Report No. 87232-MR Building on crisis response to promote long-term development A review of social safety net programs in the Islamic Republic of Mauritania 2008-2013 December 2014 Social Protection and Labor Global Practice Africa Region World Bank Document GOVERNMENT FISCAL YEAR January 1–December 31 CURRENCY EQUIVALENTS (Exchange Rate Effective June 11, 2013) Currency Unit: Mauritanian Oughiya (MRO) 1 US$: 301MRO WEIGHTS AND MEASURES Metric System ABBREVIATIONS AND ACRONYMS ACF-E Action Against Hunger Spain CAC Community Nutrition Centers CCT Conditional Cash Transfers CPS Country Partnership Strategy CRENAM Moderate Malnutrition Recovery Centers CSA Food Security Commission ECHO European Community Humanitarian Aid Office EPCV National Survey of Household Living Conditions EU European Union GDP Gross Domestic Product GIZ Gesellschaft für Internationale Zusammenarbeit (German Cooperation) GoM Government of Mauritania GPS Global Positioning System HIV Human immunodeficiency virus IDA International Development Agency IMF International Monetary Fund JICA Japanese International Cooperation Agency MAED Ministry of Economic Affairs and Development MASEF Ministry of Social Action, Children and the Family MDG Millennium Development Goals MEFPNT Ministry for Employment, Vocational Training and New Technologies MICS Multiple Indicator Cluster Survey MRO Mauritanian Oughiya Moughataa Mauritanian administrative denomination (equivalent to Department) NGO Non-governmental organization NSPS National Social Protection Strategy ONS National Statistical Office ONSA National Food Security Observatory PRSP Poverty Reduction Strategy Paper RGPH General Population and Housing Census SAVS Community Cereal Banks SMART Standardized Monitoring and Assessment of Relief and Transitions UNICEF United Nations Children’s Fund US United States USAID U.S. Agency for International Development WFP World Food Program Wilaya Mauritanian administrative denomination (equivalent to Region) Table of Contents INTRODUCTION 1 A. RATIONALE AND OBJECTIVES 1 B. DEFINITION OF SOCIAL SAFETY NET PROGRAMS USED IN THIS STUDY 1 C. METHODOLOGY AND STRUCTURE OF THE REPORT 2 I. POVERTY AND VULNERABILITY 4 A. NATIONAL CONTEXT AND OUTLOOK 4 B. HUMAN DEVELOPMENT ISSUES 4 C. POVERTY AND INEQUALITY 7 II. SOCIAL SAFETY NET PROGRAMS IN MAURITANIA 11 A. FRAMEWORK AND POLICY CONTEXT 11 B. INSTITUTIONAL ARRANGEMENTS 11 C. EXISTING SOCIAL SAFETY NET PROGRAMS 13 (1) CASH AND NEAR-CASH TRANSFERS 13 (2) FOOD TRANSFERS 14 (3) NUTRITION PROGRAMS 17 (4) SCHOOL FEEDING 18 (5) PRICE SUBSIDIES 18 (6) LABOR-INTENSIVE PUBLIC WORKS 19 (7) FEE WAIVERS 19 III. DIAGNOSTIC OF SOCIAL SAFETY NET PROGRAMS 20 A. EXPENDITURES HAVE RISEN SHARPLY 20 B. THE PRIMARY FOCUS IS RESPONDING TO SHOCKS 22 C. THE MAIN CHANNEL IS FOOD AID 23 D. TARGETING OF PROGRAMS IS IMPROVING 25 E. THE ELEMENTS ARE IN PLACE FOR A TRANSFORMATIONAL CHANGE 28 IV. REFORMS TO INCREASE IMPACT ON CHRONIC POVERTY AND RESILIENCE 29 A. TARGET THE POOR AND VULNERABLE 29 B. ADD A NATIONAL CASH TRANSFER PROGRAM TO THE PANOPLY OF PROGRAMS 34 C. COMBINING THE TWO EFFORTS FOR A TRANSFORMATIVE REFORM 38 V. CONCLUSION 40 ANNEX 1: EVALUATION RESULTS OF THE PROGRAMME EMEL 42 ANNEX 2: SSN EXPENDITURE, INCLUDING EMEL LIVESTOCK COMPONENT 45 ANNEX 3: EXPENDITURE TABLE BY PROJECT, 2008–2013 (MILLIONS MRO) 46 ANNEX 4: COST AND FINANCING OF SOCIAL SAFETY NET INTERVENTIONS IN AFRICA 47 BIBLIOGRAPHY 48 Table 1: Progress on Millennium Development Goals in Mauritania ............................................. 5 Table 2: Poverty indices by socioeconomic category, 2008 ........................................................... 6 Table 3: Poverty in Mauritania, by Wilaya ..................................................................................... 7 Table 4: Poverty trends by area of residence ................................................................................... 8 Table 5: Consumption and income inequality, by Wilaya, 2008 .................................................... 9 Table 6: Cash and near-cash transfer programs............................................................................. 13 Table 7: Food transfer programs ................................................................................................... 15 Table 8: Fuel subsidies, percent of public expenditure (2008–2013)............................................ 19 Table 9: Spending on social safety net programs 2008–2013 (MRO million) .............................. 21 Table 10: Chronic poverty and crisis response programs (percent of total expenditures) ............ 23 Table 11: Food transfer programs (percent total safety net expenditure, excl. subsidies) ............ 24 Table 12: Share of fuel subsidies enjoyed by different population quintiles ................................ 25 Table 13: Targeting methods ......................................................................................................... 30 Table 14: Total spending on SSN programs, including Emel livestock support (MRO million).. 45 Figure 1: Social safety net programs within development policy ................................................... 2 Figure 2: Subsidy rates for energy products .................................................................................. 19 Figure 3: Social safety net spending by category (2008–2013) .................................................... 22 Figure 4: Poverty incidence and distribution of Boutiques Emel and SAVS ................................ 27 Figure 5: Social safety net expenditure, including Programme Emel’s livestock component ...... 45 Acknowledgments This report was prepared by Aline Coudouel (Lead Social Protection Economist, GSPDR) and Thomas Dickinson (Social Protection Specialist, GSPDR). The report was prepared under the guidance of Stefano Paternostro (Practice Manager, GSPDR). The team benefited greatly from technical support from Setareh Razmara (Lead Social Protection Specialist, GSPDR) and El Hadramy Oubeid (Public Sector Specialist, GGODR), and administrative assistance from Josiane Luchmun, Lydie Billey (Program Assistants, GSPDR) and Fatima Cherif (Program Assistant, AFMMR). The team is grateful to the Government of Mauritania for its support and cooperation in providing information and data reviewed in this report. In particular, valuable support from the Ministry of Economic Affairs and Development (MAED), the Ministry of Social Action, Children and the Family (MASEF), the Ministry of Health, the Ministry of Education and the Ministry of Employment, Vocational Training and New Technologies (MEFPNT). The team also wishes to thank the Food Security Commission (CSA) and the Budget Office of the Ministry of Finance for their collaboration. To prepare this report, the team worked closely with the Government counterparts and other development partners (UNICEF, WFP, the European Commission, and the IMF). Vice President: Makhtar Diop Country Director: Vera Songwe Senior Global Practice Director: Arup Banerji Practice Manager: Stefano Paternostro Task Team Leader: Aline Coudouel Summary Mauritania has enjoyed a decade of strong economic growth, and has an opportunity to substantially reduce poverty over the next decade. This requires policies for inclusive growth, and an effective social safety net for the poor. The 2011 drought spurred Mauritania to invest in a large-scale emergency response. Emergency programs succeeded in limiting the crises’ impacts, but have had limited impact on the country’s goal of reducing chronic poverty and building households’ resilience to shocks. Furthermore, subsidies and emergency programs have proven costly to sustain, reaching over 5 percent of GDP in 2012. To face its fiscal constraints and better reach the poorest, Mauritania needs to complement its emergency programs with interventions that are clearly targeted to the poorest and focused on building human and productive assets to increase households’ resilience. The report makes four key recommendations to reach this goal: (1) Develop long-term interventions that address chronic poverty and build resilience against shocks. (2) Broaden the range of safety net instruments, including cash transfers. (3) Replace universal programs with targeted ones for greater impact and reduced fiscal cost (4) Reinforce emergency response capacity with a strengthened early warning system and rapidly scalable programs i Executive Summary Mauritania is one of the high-growth countries in the region and, with the right policies and investments, could substantially reduce poverty over the next decade. GDP per capita is estimated at US$1,160 (2012), for a population of about 3.5 million. Mauritania has enjoyed a strong growth performance in the years following the global crisis, registering an average growth of over 6 percent annually over the last decade (2003-2013), and is expected to remain at those levels in coming years. Growth is bolstered by its increasing mineral production, which account for 80 percent of Mauritania’s exports by value, and about a third of total fiscal revenue. Despite strong growth, Mauritania is exposed to several key risks, including international price fluctuations, climate shocks, and the spillovers of an unstable regional security situation, to which the poor are particularly vulnerable. Mauritania imports most of its fuel and 70 percent of its domestic food consumption and is thus extremely exposed to international food and fuel price fluctuations. Similarly, any drop in international demand for minerals, on which the country is highly dependent, would have direct consequences on Mauritania’s fiscal balance and growth outlook. With its mostly arid, Sahelian topography, the country is also very vulnerable to climate shocks. Mauritania must also contend with the political instability in the sub-region. Poverty and repeated climate shocks have driven a rapid transition from a largely rural and nomadic society towards sedentarization and urbanization. The national poverty level stands at 42 percent (2008), but is much higher in rural areas, where 59.4 percent of the population is poor. The capital Nouakchott has seen the second-highest urban expansion in Africa, growing by 10.4 percent per year between 1950 and 2010, and now accounts for a quarter of the country’s population. The fertile zone along the Senegal River in the south is the other area of dense population, with the remainder of the country mostly arid and sparsely populated. This evolving geographic and demographic distribution calls for strong investment of the dividends of growth to promote greater inclusion and ensure sustainable long-term growth. Mauritania’s population is young, with 50 percent of the population below eighteen and 30 percent under fifteen. Beyond the strong export performance, sustainable growth will require investing in the human capital of its population. Health and education services are lacking in both quantity and quality. Repeated shocks have spurred Mauritania to invest in large-scale emergency programs. Mauritania responded to rising international fuel prices with a universal fuel subsidy in 2009. A critical drought affected the entire country in 2011, which reduced domestic food production by 75 percent. As a result, expenditure on social safety net programs has tripled in nominal terms from MRO8.3 billion in 2008 (1.0 percent of GDP or 3.9 percent of Government expenditure, including fuel subsidies) to over MRO67.9 billion in 2013 (5.4 percent of GDP or 18.1 percent of Government expenditure). Expenditures are forecast to remain high for 2013, at MRO39.2 billion (2.9 percent of GDP, or 9.7 percent of Government expenditure). Unlike many other countries of the region, Mauritania’s social safety net expenditure is mostly financed on domestic resources. Including fuel subsidies, between 2010 and 2013, Mauritania self-financed 84 percent of its safety net programs, with foreign contributions accounting for the remaining 16 percent. ii Social safety net spending by category, including Emel livestock support (2008–2013) 8.0 % GDP 7.0 6.0 Food transfers 5.0 EMEL Livestock support 4.0 Cash and near-cash transfers 3.0 Public works Fee waivers 2.0 Fuel subsidies 1.0 - 2008 2009 2010 2011 2012 2013* Source: Government and program data, World Bank staff calculations. Mauritania’s emergency response, via the Programme Emel, succeeded in averting serious crises. The core objectives of Mauritania’s emergency interventions were to mitigate the impact of crises for vulnerable populations and keep food prices low. Analysis confirms that the programs succeeded in reaching this objective. The three interventions under the Programme Emel (Arabic for hope) are food- based programs— a national network of subsidized food shops (the Boutiques Emel), restocking of cereal banks (the SAVS) and an emergency food distribution program. Programme Emel also included a large-scale livestock support program. While it was designed as a short-term response, much of the Programme Emel remains in place years after the crisis, leaving Mauritania with a mismatch between its programs, which are geared towards crisis-response, and its long-term poverty reduction objectives. Recent evaluations show that Programme Emel provides tangible benefits to Mauritania’s poor, though few of these benefits reach the poorest of the poor who lack the resources to purchase the products sold in the Boutiques Emel. While the Boutiques Emel, which account for 80 percent of Emel expenditure, were initially designed for a duration of 5 months, they remain firmly implanted across the country three years later. Costs tied to the Boutiques Emel have also increased sharply as the program has expanded, and the program’s appreciation by the population has made winding it down difficult. Mauritania needs to contain the fiscal impact of emergencies while reducing chronic poverty and building the resilience of the poorest households to future crises. For this, Mauritania needs to complement its emergency programs with national, targeted, long-term programs to support the chronically poor and help them invest in their human and productive capital to increase their income- generating capacity and build their resilience. It will also need to reduce untargeted programs or subsidies, which fail to protect the effectively reach the poor but have a significant fiscal cost. The elements are in place for transformational change in Mauritania. Between 2013 and 2014, Mauritania adopted the strategic frameworks and established the institutions that can develop and implement national social protection and poverty reduction programs. New targeted programs, such as cash transfers and food vouchers, have been developed and tested since 2011, and show considerable promise. Furthermore, the country’s favorable export and growth outlook give it the means of its ambitions. This report underscores the importance of four core actions for this transformation to take place: iii (1) Develop long-term interventions that address chronic poverty and build resilience against shocks. It is essential to complement existing emergency programs with long-term interventions that address chronic poverty and build resilience against shocks. The development of programs, benchmarks and instruments should follow the national vision laid out in the NSPS and PRSP III frameworks. In particular, an effort should be made to strengthen households’ capital – both the human capital of their members and in particular their children, and the productive assets that can improve their income-generating capacity – so as to increase their resilience and promote inter-generational social mobility. (2) Broaden the range of safety net instruments, including cash transfers. To better address chronic poverty and build resilience among the poorest, the Government should broaden the range of programs to include permanent safety net interventions, such as cash transfers. Long-term cash transfers that are well-targeted to households suffering from chronic poverty, and promote investments in the human capital of their children, can be particularly adapted to Mauritania’s situation. This would need to be complemented, more generally, with efforts to strengthen the supply of basic social services to these households. Furthermore, there are significant distortions in Mauritania’s f ood markets, which suggests large potential gains for the poorest if these can be corrected. (3) Replace universal programs with targeted ones for greater impact and reduced fiscal cost Universal subsidies tend to benefit the wealthy and have not contributed significantly to poverty reduction in Mauritania. At their current level, eliminating subsidies could save over 3 percent of GDP, far more than the cost of all existing programs targeted to the poor and vulnerable. The reallocation of a fraction of the resources currently dedicated to universal subsidies towards targeted programs would be a significant improvement in the efficiency and equity of the allocation of public resources, and lead to greater impact both in supporting domestic consumption and in reducing poverty. Improving the targeting of social programs, and improving their coordination, can build upon a national registry of poor households that adequately collects information on the poorest households and their characteristics. In the context of Mauritania, a combination of targeting methodologies can be designed to effectively identify those most in need across the territory. (4) Reinforce emergency response capacity with a strengthened early warning system and rapidly scalable programs Mauritania’s exposure to shocks places a high priority on an effective emergency response mechanism. The Government should consider improving its early-warning system and developing contingency planning and financing for social protection response to crises. This would allow for a more specific and rapid response to crises, which can increase the impact of emergency resources on human welfare and households’ resilience. The national registry of poor households can also play a critical role in the rapid implementation of programs triggered by crises, including food-based emergency programs, by pre-identifying households most likely to be affected by specific shocks. Mauritania could also consider complementing its typical emergency response with mechanisms to temporarily scale-up long-term safety net programs, so as to rapidly channel resources to crisis-stricken households. iv Introduction A. Rationale and Objectives 1. Mauritania is one of the high-growth countries in the region, and could, with the right policies and investments, substantially reduce poverty and accelerate growth over the next decade. However repeated shocks since 2008 have led the government to invest massively in large- scale, emergency programs, leading to sustainability and efficiency concerns. Furthermore, the Mauritanian Government is keen to ensure the package of interventions yield the maximum impact on the well-being and resilience of its vulnerable population. 2. In light of concerns on sustainability and impact, the Government of Mauritania has requested a review of the country’s social safety net programs, their cost and impact. This review is framed within the vision presented in Mauritania’s National Social Protection Strategy, of providing long-term support to the country’s chronically poor and vulnerable population, as well as the third National Poverty Reduction Strategy (PRSP III), which outlines Mauritania’s overall vision for poverty-reduction. B. Definition of Social Safety Net Programs Used in This Study 3. In this document, “social safety net” programs refer to non-contributory transfer programs targeted to the poor or vulnerable. Social safety net programs aim to increase households’ welfare and consumption of basic commodities and essential services, either directly or through substitution effects. Social safety net interventions are targeted to the poor and vulnerable, that is individuals living in poverty and unable to meet their own basic needs or those in danger of falling into poverty, either because of external shocks or socioeconomic circumstances, such as age, illness, or disability. With this definition, social safety net interventions may serve one or a combination of the following groups (Grosh et al. 2008): (1) chronic poor, defined as people who lack the assets to earn sufficient income, even in good years; (2) transient poor, defined as people who earn sufficient income in good years but fall into poverty, at least temporarily, as a result of idiosyncratic or covariant shocks ranging from illness in the household or loss of a job to drought or macroeconomic crisis; (3) vulnerable groups, commonly including—but not limited to—the disabled, the elderly, orphans, the displaced, refugees, and asylum seekers; and (4) groups who stand to be negatively affected by reforms or policies. In this report, programs are classed into five categories of interventions: (1) Cash or near-cash transfers. These can be conditional or unconditional (2) Food transfers. In-kind transfers in the form of food distribution, nutrition programs, school feeding, targeted subsidized food sales, etc. (3) Universal subsidies. On food or fuel prices. (4) Labor-intensive public works. Including cash for work programs. (5) Fee waivers. Fully subsidized basic services in health, education, etc. 4. Social safety net programs form a subset of the broader field of social protection programs and social policies. Social safety net programs include only non-contributory programs, whereas social protection includes social insurance (contributive programs and insurance) and labor market programs. Furthermore, there are other non-contributive programs that sometimes target the poor, for example programs supporting production (e.g., support for agricultural inputs) or programs for creating income (e.g., microcredit), as well as transfers targeting groups or communities (thus not targeted toward individuals or households). Ideally, social safety net programs are built to be complementary and coordinated under a single system, which itself is articulated with overall social 1 policy. A good social safety net system is more than a collection of well-designed and well- implemented programs. The social protection “systemic effect” can trigger more than the sum of the individual social programs.1 For example, social safety net programs contribute to poverty reduction, but also help vulnerable populations manage risks and promote social equity and access to essential services such as health and education (Figure 1). In recent years, the concepts of social protection and social safety net have increasingly become a central component of poverty reduction and food security strategies in developing countries. Figure 1: Social safety net programs within development policy Source: Grosh et al. (2008) C. Methodology and Structure of the Report 5. The analysis of Mauritania’s social safety net programs is based on national data, consultation of program documents, and interviews undertaken during World Bank missions in 2013. Program information is for the period 2008–2013. As a large share of overall social safety net interventions are under the form of food aid that is often tracked in terms of volume rather than value, cost estimates for food-related programs are based on international food prices of that year, and administrative costs estimated on the basis of volume. Absent robust evaluations of programs (both of processes and results), the conclusions presented here are founded on stakeholder and World Bank analysis, and comparison with international experience in social safety net program implementation. The discussion focuses on the main, Government-led interventions, particularly the Government’s flagship emergency-response program (Programme Emel), which accounts for the largest share of non- subsidy expenditure in Mauritania. International development partners working closely with the Government and their contributions to national programs are also taken into account in this analysis. NGOs, both national and international, have been included according to their involvement with these national programs. Small-scale, independent programs have not been included. 1 See Mauritania National Social Protection Strategy (2011) for a detailed discussion of the vision for a social protection system for Mauritania. 2 6. This study is structured around four chapters, focusing on overall context; a description of institutions, programs, and expenditures; a diagnostic of the existing programs; and a proposed set of options for improving the impact of safety net interventions. The study begins with an overview of context, focusing on poverty and vulnerability. Chapter two details the policy context and institutional arrangements for safety net programs in Mauritania and describes the country’s main programs. Chapter three proposes a diagnostic of the existing programs and policies, focusing on analyzing program expenditures, objectives, targeting, and overall efficiency. Chapter four explores potential reform options to build on crisis response to promote long-term development. 3 I. Poverty and Vulnerability A. National Context and Outlook 7. With an average annual growth of over 6 percent over the last decade (2003-2013), Mauritania is one of the high-growth countries in the region. GDP per capita is estimated at US$1,160 (2012), for a population of about 3.5 million. The country has enjoyed a strong growth performance in the years following the global crisis, and is expected to remain at those levels in coming years. Growth is bolstered by high prices for its increasing mineral production, which account for 80 percent of Mauritania’s exports by value, and about a third of total fiscal revenue. The country also has significant potential for further growth in mineral production, particularly iron ore, of which Mauritania is already one of Africa’s largest exporters, and gold, production of which is forecast to triple after 2017. Mauritania is also a modest oil producer, has strong potential for gas production, and abundant fish stocks. 8. Mostly a desert country, Mauritania has undergone a very rapid transition from a largely rural and nomadic society towards sedentarization and urbanization. The capital Nouakchott has seen the second-highest urban expansion in Africa, growing by 10.4 percent per year between 1950 and 2010, and now accounts for a quarter of the country’s population. The fertile zone along the Senegal River in the south is the other area of dense population, with the remainder of the country mostly arid and sparsely populated. 9. Despite its strong growth, Mauritania faces several serious challenges, and is particularly vulnerable to international food and fuel price increases, repeated climate shocks, and regional security risks. Mauritania imports 70 percent of its domestic food consumption and is thus extremely exposed to international price fluctuations. With its mostly arid, Sahelian topography, the country is particularly vulnerable to climate shocks, to which it is regularly subjected. A rising threat over recent years has been the emergence of conflicts and instability in the sub region. The national poverty level stands at 42 percent (2008), but is much higher in rural areas, where 59.4 percent of the population is poor. In light of these factors and population growth, it is unlikely that Mauritania attains its objective of reducing poverty to 25 percent by 2015. The results of the 2014 national household survey will provide an important update in this respect. B. Human Development Issues 10. Mauritania ranks 155th out of 187 countries on the 2013 human development index, but has made progress on a number of Millennium Development Goals (MDGs). Mauritania is likely to achieve objectives in basic education, fighting HIV and malaria, and in ensuring access to water. Mauritania has also made advances on gender equality, particularly by improving access to primary education and by establishing minimum quotas on women’s political participation2. Mauritania has even been ranked among the top 20 performers for absolute progress towards the MDGs (ODI, 2010), despite falling behind in key objectives such as poverty. In fact while Mauritania saw a large drop in poverty of 15 percentage points between 1990 and 2008, poverty has remained static at 42 percent since 2008 and falls far short of the target MDG objective of 28.3 percent by 2015. The 2011 Multi- 2 The implementation of this policy has met challenges and has yet to effect tangible benefits in women’s daily lives, but has resulted in an increase in the share of seats held by women in the parliament (18 percent) and municipal council (30 percent). 4 Indicator Cluster Survey (MICS) also revealed the persistence of high rates of maternal, infant, and child mortality. Maternal and child mortality in 2011 stood at 626 per 100,000 live births and 218 per 1,000 live births, respectively. Table 1: Progress on Millennium Development Goals in Mauritania Main MDGs Actual performance* Targets for 2015 1990 2010 MDGs 1. Eradicate extreme poverty and hunger Poverty incidence (%) 56.6 42 28.3 Prevalence of child malnutrition (%) 47.6 39.4 23.5 2. Provide primary education for all Gross primary enrollment rate (%) 49.0 73 100.0 Retention rate at the entrance of the 5th grade (%) 73.8 49.3 100.0 3. Promote gender equality and empower women Ratio of girls to boys in primary education 0.72 1.02 1.00 4. Reduce child mortality Child mortality (per 1,000 live births) 137.0 122 45.0 5. Improve maternal health Maternal mortality (per 100,000 live births) 930 686 232 6. Combat HIV/AIDS, malaria, and other diseases HIV/AIDS infection rate (%) 0.3 0.6 – 7. Ensure environmental sustainability Share of population with access to potable water (%) 37.0 62.0 82.0 Source: Islamic Republic of Mauritania, “Rapport sur les progrès 2010 vers l'atteinte des objectifs du millénaire pour le développement (OMD) en Mauritanie,” 2010 Health and Education 11. Mauritania’s healthcare system struggles to cope with staffing and geographic distribution of the population. Mauritania suffers from a chronic shortage of health workers (nurses, midwives, doctors, etc.) at all levels as well as the sometimes poor training of staff, and struggles with ensuring access, with 21 percent of the population living more than 5 km from an operational healthcare unit (Multiple Indicator Cluster Survey [MICS] survey, 2011); . The health system is also adversely affected by a multitude of external factors, including hygiene (safe drinking water and the removal of household waste); problems relating to sanitary habits among the population; and problems relating to food and nutritional security and natural disasters. 12. Health indicators are improving but remain precarious, with strong discrepancies along economic and geographic lines. Prenatal consultations progressed from 80.2 to 87.4 percent between 2004 and 2008. Assisted delivery also progressed slightly, rising from 56.5 to 60.2 percent over the same period, but only stands at 27.4 percent among the poorest quintile (against 92 percent for the wealthiest). The weakest areas in the country in this respect are the southern regions of Guidimagha (24.2 percent), Hodh El Charghi (28.9 percent) and Gorgol (35.8 percent). Major epidemics and diseases (AIDS, tuberculosis, and malaria) are under control, and HIV/AIDS prevalence is low (estimated at below 1 percent in 2009). Immunization, however, recorded a decline between 2004 and 2008, sinking from 79 to 68.8 percent, with even lower levels for the poor (58.4 percent), rural inhabitants (62.9 percent), and in the regions of Tagant, Gorgol, and Hodh el Gharbi (41.8, 43.4, and 43.8 percent respectively). 13. Progress has been made in education, particularly enrollment at the primary level, but this rapidly falls off from the secondary level. The gross enrollment ratio at the primary level, which was 76.6 percent in 2004, rose to 98.9 percent in 2011, close to the target of 100 percent set for 2015. 5 Despite this strong progress at the primary level, there still remain significant discrepancies between rural and urban areas, between Wilayas, and between economic quintiles. At the lower secondary level, the gross enrollment ratio improved slightly from 26.6 percent in 2006 to 29.6 percent in 2011. The differences in participation remain very high between urban (53.5 percent) and rural (12.8 percent) areas (2008 data). Overall, one child out of two does not complete the basic education cycle. Enrollment drops even further at the higher secondary level. Student populations at the upper- secondary level declining by 20.7 percent between 2006 and 2011, with a gross enrollment ratio of 13.6 percent for boys and 17 percent for girls. 14. Quality remains low, and does not appear to be on a favorable trend. Gross enrollment ratios do not provide indications on either quality or attendance, which remain highly problematic across the board. The retention rate is low, and only about 65.2 percent of pupils finish primary school. The probability of completing lower secondary school is higher for boys (42 percent) than it is for girls (34 percent). These results stem from several factors, including the reform of the sector, which lengthened the lower-secondary-school level, the low quality of the educational services provided as well as frequent changes between French-based and Arab-based curriculum and educational policies. As a result, educational attainment has been suffering a long-term decline. 15. Despite success in reversing the enrollment gap at primary level between girls and boys, completion rates and overall literacy remain lower for girls. Girls’ participation in primary education increased from 79 percent in 2004 to 93.5 percent in 2008. At the secondary level, however, participation is higher for boys (32.7 percent) than for girls (28.4 percent). Overall however, literacy remains lower for women over 15 (54.4 percent) than for men (70.3 percent). Literacy also varies according to rural (50.3 percent) or urban (73.3 percent) residency, as well as by economic level (66.6 percent among the non-poor against 46 percent among the poor). 16. Education exerts a strong influence on employment and poverty outcomes. Poverty generally decreases significantly when the education of the household head improves. More educated socioeconomic categories, such as public sector employees, therefore show the lowest incidence of poverty (21.8 percent), while the less-educated categories such as independent subsistence farmers show far higher poverty rates (69.7 percent). Other socioeconomic groups particularly vulnerable to poverty are the unemployed, apprentices, and family aides (44.4 percent each). However, in absolute numbers, the largest group in the poor is made up of the “inactive” (32 percent of the poor), followed by independent, non-agricultural workers (21 percent of the poor). Table 2: Poverty indices by socioeconomic category, 2008 Contribution to Poverty rate (%) total poverty (%) Public sector employees 21.8 6.2 Private sector employees 39.3 13.2 Independent agricultural workers 69.7 11.1 Independent non-agricultural workers 38.1 21.2 Apprentices/family aides and other workers 44.4 12.4 Unemployed 44.4 3.6 Inactive 47.3 32.2 Total 42 100 Source: EPCV, 2008 6 17. Unemployment affects an estimated 10 percent of the population and almost 17 percent in urban areas.3 However, there is also significant under-employment, as well as many who are no longer looking for work. One-third of youth (15–34 years old) are inactive, neither working nor in school. Capital-intensive mining projects are unable to absorb this rapidly growing and low-skilled population. The informal economy and low value-added activities have been the main drivers of job creation in recent years, and account for 85 percent of the total labor force. 18. Despite progress, nearly half of the population of Mauritania does not have access to safe drinking water. The safe drinking water supply rate stood at 52 percent nationally in 2011. In urban areas (towns with more than 5,000 inhabitants), the rate of access to private water main connections was 35 percent although it varied significantly from town to town. In rural areas, 60 percent of households had access to safe drinking water. Access to water also varies strongly according to household wealth, with 87 percent of the wealthiest households having access to safe drinking water, against only 30 percent of the poorest. In 2008, 45.7 percent of households did not have toilets in their house, down from 48 percent in 2004. C. Poverty and Inequality 19. A large share of Mauritania’s population remains poor. According to the most recent poverty assessment, the Enquête Permanente sur les Conditions de Vie des ménages (EPCV 2008), poverty (defined at MRO129,000 per household per year) remained at about 46.5 percent between 2000 and 2004 before declining to 42 percent in 2008. Poverty incidence is unlikely to have improved since 2008, given the slow rate of GDP growth resulting from the 2009 political shock and the 2011 climatic shock. Using the growth elasticity of poverty for the 2004–08 period, World Bank calculations indicate that poverty incidence would be 41 percent in 2012, leaving the 2015 target of 25 percent likely beyond reach4. Table 3: Poverty in Mauritania, by Wilaya Wilaya Poverty rate (%) Ext. poverty (%) 2004 2008 2004 2008 URBAN 28.9 20.8 13.8 7.7 RURAL 59 59.4 39.1 40.8 Hodh El Chargui 50.4 58.4 26.8 40.2 Hodh El Gharbi 47.6 45.2 30.5 26.5 Assaba 44.1 56 27.5 37.5 Gorgol 68.5 66.5 47.3 43.4 Brakna 65.2 63.9 43.1 45.3 Trarza 51.9 37.1 33.2 22.3 Adrar 39.6 57.1 22.4 35.5 Dakhlet-Nouadhibou 20.7 18.6 10.8 3.8 Tagant 70.3 67.8 53.9 50.2 Guidimakha 62.7 56.7 44.5 37.5 Tiris-Zemmour 30.4 17.8 12 10.7 Inchiri 50.9 31.7 22 22.8 Nouakchott 25.9 15.6 10.9 4.8 TOTAL 46.7 42 29 26 Source: EPCV, 2008 3National Statistics Office, Enquête sur l'Emploi en Mauritanie, 2013. 4 World Bank Mauritania Country Partnership Strategy (2013). 7 20. There is a diverging trend between urban areas, where poverty has fallen significantly, and rural areas, where poverty has worsened. Overall, poverty is increasingly a rural phenomenon in Mauritania: 59.4 percent of rural inhabitants live below the poverty line, against 20.8 percent in urban areas. Nouakchott itself, the main urban area in the country, accounts only for 7.9 percent of poor households. As presented in table 4, the poverty gap is greater in rural areas (22.3 percent) than in urban ones (4.9 percent). In turn, the severity of poverty is greater for rural inhabitants (11.1 percent) than urban dwellers (1.7 percent). Table 4: Poverty trends by area of residence Poverty indices (%) Contribution to total poverty (%) incidence gap severity incidence gap severity URBAN 20.8 4.9 1.7 22.3 15.2 11.3 RURAL 59.4 22.3 11.1 77.7 84.8 88.7 TOTAL 42.0 14.5 6.9 100.0 100.0 100.0 Source: EPCV (2008) 21. Poverty is concentrated in the Wilayas in the Southern part of the country. Poverty remains very high in certain Wilayas, and has even increased in three between 2004 and 2008 (Hodh El Chargui, Assaba, and Adrar). In its 2008 analysis of poverty rates, ONS classified regions in four main groups: (i) Wilayas with poverty rates of 60 percent or above (Tagant, Gorgol, and Brakna), (ii) Wilayas with poverty rates between 55 and 60 percent (Hodh El Charghi, Adrar, Guidimagha, and Assaba), (iii) Wilayas with poverty rates between 30 and 50 percent (Hodh El Gharbi, Trarza, and Inchiri) and Wilayas with poverty rates below 20 percent (Nouadhibou, Tiris-Zemmour, and Nouakchott). In 2008 there were 15 Moughataa where poverty rates were above 65 percent, and these areas were home to nearly a third of the country’s population. 22. A quarter of the population lives in extreme poverty. A little over a quarter (25.9 percent) of Mauritania’s population is estimated to live at or below the level of extreme poverty (set at an annual MRO96,400 for 2008). Between 2004 and 2008, extreme poverty fell by 3 percentage points, a smaller reduction than the overall fall in poverty registered over the same period. Rural areas again have a higher level of extreme poverty than urban areas (40.8 percent in rural areas versus 7.7 percent in urban) and account for the overwhelming majority of extreme poor: 86 percent of the extreme poor are rural inhabitants. Extreme poverty is also concentrated, with the most affected Moughataa being Monguel (71.2 percent, Gorgol), Moudjeria (65.3 percent, Tagant), M’bagne (61 percent, Brakna), Barkéwol (60.4 percent, Assaba) and M’bout (59.9 percent, Gorgol)5. 23. The Government has set a target of 3 percent extreme poverty by 2030, which would require an average annual growth in per capita income on the order of 10 percent6. If Mauritania can maintain currently projected growth rates, it could, conceivably, halve extreme poverty by 2030. However, this would also require better redistribution of wealth and a more inclusive growth model as well as a strong social safety net designed to support the poorest. 24. Mauritania’s vulnerability to environmental shocks and high levels of food insecurity leaves many non-poor households at risk of falling into transient poverty. In December 2011, an estimated one quarter of rural households were food insecure, and half of these, severely so. As we have seen, the 2011 drought led domestic agricultural production to fall by 75 percent, which severely 5 For a detailed and in-depth discussion of poverty in Mauritania, see the National Social Protection Strategy (2012) 6 World Bank, Mauritania Country Partnership Strategy (2013) 8 impacted agricultural communities, and caused losses of cattle, which increased vulnerability among pastoral communities. For even the non-poor therefore, environmental or economic shocks risk bearing down with an immediate and direct impact and tipping them over the threshold into poverty. 25. Overall, almost three-quarters of the population (73 percent) considered that their economic situation had deteriorated in 2008, against 44 percent in 2004. The determinants of poverty were perceived as mostly lack of money to buy food (59.7 percent of households) and unemployment (51.3 percent) according to the 2008 data. Urban respondents invoked lack of means to buy food and unemployment as the main drivers of poverty, while rural respondents invoked absence of land or livestock and the inability to respond to shocks. Among the poor themselves, the main drivers of poverty are reported to be lack of money for food (60.2 percent), lack of livestock (57 percent) and unemployment (56.6 percent). Among non-poor households, the main reported difficulties pertain to housing (27.6 percent) and challenges in accumulating savings (12.9 percent). Income Disparities and Household Expenditures 26. Overall inequality in Mauritania remains moderate compared to many lower/mid- income countries, but may well rise if the dividends of current economic growth remain concentrated among a few. The Gini coefficient of poverty was estimated at 0.38 in 2008, similar to 2004 (0.39). Between 2000 and 2004, the bottom 20 percent enjoyed significant increases in consumption—higher than increases for the non-poor—suggesting an inclusive pattern of growth. However, this pattern was reversed in the 2004–08 period, as growth was driven more by extractive industries, which did not contribute to large-scale employment creation or broad-based growth. The poorest quintile accounts for only 6.3 percent of national household expenditure, seven times less than the richest quintile (44.2 percent). By decile, Mauritania’s wealthiest 10 percent account for 28.1 percent of national expenditure, 11 times more than the poorest decile (2.5 percent). Furthermore, the poorest are growing poorer, with expenditure for the bottom decile falling from 2.7 to 2.5 percent of national household expenditure between 2004 and 2008. Looking forward, it will be essential for Mauritania to seriously invest in better basic services or redistributive policies for its population to mitigate the likely widening inequality gap brought on by the current boom in extractive industries. Table 5: Consumption and income inequality, by Wilaya, 2008 Share of total consumption by quintile Gini coefficient 1st 5th 2004 2008 URBAN 7.4 40.8 RURAL 7.5 40.5 Hodh El Chargui 6.5 42.6 0.31 0.36 Hodh El Gharbi 8.2 39.5 0.38 0.31 Assaba 7.4 42.3 0.41 0.35 Gorgol 8.3 38.1 0.34 0.30 Brakna 7.4 41.5 0.31 0.34 Trarza 7.5 39.0 0.36 0.31 Adrar 7.7 41.3 0.38 0.33 Dakhlet-Nouadhibou 6.8 40.3 0.39 0.34 Tagant 7.9 40.3 0.33 0.32 Guidimakha 7.5 40.7 0.32 0.33 Tiris-Zemmour 8.5 38.0 0.27 0.27 Inchiri 8.5 37.7 0.24 0.30 Nouakchott 7.9 39.6 0.40 0.31 TOTAL 6 44 0.39 0.38 Source: EPCV 2004, 2008 9 27. Because Mauritania imports 70 percent of its food, high international food prices have been a heavy burden for poor households. Food expenditures showed a strong rise from 51.7 percent to 57.8 percent of average household expenditure between 2004 and 2008, resulting from spiking international food prices at the time. As Mauritania imports up to 70 percent of its total food needs, the increase in household food expenditures can be largely attributed to the sharp increases in food products at the time of the national household survey. High food prices are a particularly heavy burden for poor rural households, who spend on average 66.5 percent of their income on food (against 52.4 percent for poor urban households). Extreme poor households suffer most of all, as food represents up to 69.8 percent of household expenditure. Nevertheless, even the wealthiest households still dedicated nearly half (49.7 percent) of their total household income to food products in 2008. 28. The higher cost of food may have affected households’ expenditures in other areas, notably education and health. Education and health expenditures have fallen from 1.5 and 6.1 percent of total household expenditures in 2004, to 1 and 3.8 percent in 2008. However, the national EPCV report cautions that this sharp fall may also be in part attributable to the season in which data were collected, reportedly one of low health expenditure. Food Insecurity and Malnutrition 29. According to latest available data (2011), 26 percent of the population suffers from food insecurity. Again, this situation is related to the overall level of poverty; lack of investment in the social and productive sectors, in particular agriculture; and the harsh climate. Food insecurity was further aggravated by the rise in the price of food staples and by the 2011 drought. Anecdotal evidence also points towards an increase in food insecurity in urban contexts. 30. Chronic malnutrition affects 40 percent of Mauritania’s children. Chronic malnutrition remained stable between 2004 (40.3 percent) and 2008 (40.l percent). Acute malnutrition however increased from 12.2 to 15.6 percent over the same period. Once again, the poor, particularly the rural poor, suffer from higher levels of malnutrition and stunting. Recent evidence suggests that levels of malnutrition have remained high: the latest vulnerability survey by the Food Security Observatory shows that 21.1 percent of households have an irregular and unbalanced diet; i.e., more than 428,000 people in rural areas and 25,000 in peri-urban areas. Moreover, recent screenings by the Ministry of Health show a global acute malnutrition rate of 10.9 percent and a severe acute malnutrition rate of 1 percent among children between 6 and 59 months old. The prevalence of global acute malnutrition in the Wilayas of Guidimakha, Assaba, Gorgol, and Brakna is higher than the emergency threshold of 15 percent. Hodh El Chargui shows a global acute malnutrition rate close to the emergency threshold (14.8 percent). 10 II. Social Safety Net Programs in Mauritania A. Framework and Policy Context 31. Social protection has been part of Mauritania’s poverty reduction strategy since 2006. Mauritania’s first Poverty Reduction Strategy (Cadre Stratégique de Lutte contre la Pauvreté – CSLP), covering the period 2001–2004, included a first discussion of social protection, which was fully integrated by the 2006–2010 CSLP II. The current CSLP III (2011–2015) defines social safety net programs as central to poverty reduction and puts forward a set of measures promoting economic inclusion and productivity among the poor. CSLP III also calls for the harmonization of the institutional and regulatory framework for social protection. 32. Mauritania adopted a National Social Protection Strategy (NSPS) in 2013. The NSPS was developed under the joint leadership of the MAED and the MASEF, with technical support from UNICEF. The NSPS offers a long-term vision for the country and a roadmap for creating a coherent and integrated national social protection system. The NSPS has five main pillars: (i) food security and nutrition, (ii) access to health and education, (iii) social security and employment, (iv) improving the living environment, and (v) social assistance and support to vulnerable populations. 33. The country has strategic frameworks for health and education, which the NSPS supports, especially by promoting access. The National Health Policy (Politique Nationale de Santé – PNS 2006–2015) frames health sector objectives, particularly in terms of the availability of services, of fairness and access for vulnerable groups. These directions are developed in the National Health Development Plan (Plan National de Développement Sanitaire – PNDS 2012–2020). National education policy is described in the National Development Program for Education (Programme National de Développement du Secteur Educatif – PNDSE), a 10-year plan covering all sector elements: childhood, primary education and literacy, basic and secondary education, technical and vocational training, and higher education. 34. Mauritania also has a strategic framework for fighting against malnutrition. The National Development Council for Nutrition (Conseil National de Développement pour la Nutrition – CNDN), created in 2010, has developed a Cross-Sectoral Nutrition Action Plan (Plan d’Action Intersectoriel de Nutrition – PAIN) to reach the MDG, targeting a 50 percent reduction in the proportion of people suffering from hunger. At the decentralized level, the CNDN is supported by Regional Development Commissions for Nutrition. B. Institutional Arrangements 35. The Government has set up an institutional arrangement for implementing the National Social Protection Strategy. The MAED and the MASEF hold joint responsibility for implementing the NSPS7. The institutional arrangement chosen to supervise implementation of the Strategy includes: (i) A Steering Committee, (ii) A Technical Committee, and (iii) A Social Protection Unit (Cellule de Protection Sociale – CPS). The Steering Committee manages and supervises all activities tied to the NSPS. The Steering Committee is chaired alternately, every six months, by the MAED and the MASEF. The Technical Committee is a technical body whose mission is to support and advise the 7 Joint order Number 94 MAED/MASEF (January 8, 2014) 11 Steering Committee. The Social Protection Unit, attached to the MAED, instigates, coordinates, and manages the NSPS’s daily activities. 36. The adoption of the NSPS and the definition of its institutional anchoring provide a solid foundation for implementation. In the past, responsibility for the social protection portfolio had passed several times from one ministry to another, especially during times of Government reorganization. Social protection was first housed under the Ministry of Health and Social Affairs, then the Commissariat à la Sécurité Alimentaire et à la Protection Sociale (now the CSA), before being returned to the Ministry of Social Affairs, Children, and the Family (MASEF). The NSPS was developed under the aegis of the MAED, which was responsible for the preparation of the NSPS and is now tasked with coordinating its implementation in collaboration with the MASEF. 37. Several ministries currently implement programs that fall under the crosscutting theme of social protection. The MASEF (programs for the handicapped, beggars, and orphans), the Ministry of Health (nutrition programs), the Ministry of Education (school feeding), and the Ministère de l’Emploi (public works) all implement programs are targeted to the poor and vulnerable populations that fall under the five pillars of the NSPS. 38. Mauritania also has two national agencies dedicated to implementation of programs that include social protection. The Commissariat à la Sécurité Alimentaire (CSA) has the overall responsibility for emergency humanitarian interventions, particularly in cases of drought and food insecurity, It is responsible for managing, supervising, and coordinating emergency interventions nationwide and also manages national emergency food stocks. Thus, the CSA has a key operational role in the Government’s National Emergency Programme Emel (detailed below). The CSA is also responsible for the monitoring of food security and nutrition (via the Observatoire National de la Sécurité Alimentaire, ONSA) and has recently set up a community nutrition unit specifically tasked with addressing chronic malnutrition (the CAC and CRENAM programs, described below). The second institution is the Agence Nationale Tadamoun de Lutte contre les Séquelles de l’Esclavage, l’Insertion et la Lutte contre la Pauvreté, also known as the Agence Nationale Tadamoun (Solidarity in Arabic). Under the direct supervision of the Presidency, the Tadamoun Agency was created in 20138 but inherited the staff and programs of the pre-existing Agence Nationale d’Appui et d’Insertion des Réfugiés (ANAIR). Its objectives are to: (i) eradicate the after-effects of slavery, (ii) integrate repatriated refugees, and (iii) implement programs targeting the eradication of extreme poverty. 39. Technical and financial partners also support several social protection programs in Mauritania. The World Food Program (WFP) works closely with the CSA via its focus on emergencies and food security, and it has been involved in implementation of the Programme Emel. The WFP also contributes to the nutrition and school feeding programs led by the Ministry of Health and the Ministry of Education. UNICEF has supported the creation of the NSPS and collaborates closely with the Ministry of Health, especially on maternal and child health and nutrition programs, as well as on a CCT targeted to mothers and young children. Between 2011 and 2013, the European Union financed a series of pilot cash transfer projects, which were implemented by national and international NGOs. 8 Decree Number 048/2013 12 C. Existing Social Safety Net Programs 40. This section presents the main social safety net programs according to five categories of intervention: (1) cash or near-cash transfers, which can be conditional or unconditional, (2) food transfers, including in-kind transfers in the form of food distribution, nutrition programs, school feeding, subsidized food sales, etc., (3) universal subsidies on food or fuel prices, (4) labor-intensive public works (cash for work programs), and (5) fee waivers for basic services in health, education, etc. (1) Cash and Near-Cash Transfers 41. Several pilot cash transfer programs were implemented between 2011 and 2013, financed by European Union grant. These programs were the first significant experience of cash transfers in Mauritania. The pilot programs were implemented by the WFP/CSA, and a number of international and Mauritanian NGOs (including the French Red Cross, ACF-Spain, World Vision and GRET). The pilots were designed to have a strong focus on knowledge-creation, and each followed an independent methodology and focused on using cash transfers as an incentive and catalyst for results in food security, nutrition, health and livelihoods. The WFP/CSA program was the largest, shared between Nouakchott and several rural areas, while the other pilots, carried out by NGOs, were smaller and more localized. Following this, the WFP and the CSA also implemented a transfer program based on food vouchers (near-cash transfers). The Ministry of Health, in collaboration with UNICEF, is also preparing a cash transfer program focusing on child and maternal health, due to start implementation in 2014. Table 6: Cash and near-cash transfer programs EU Transfer Program Food Voucher Cash Transfer WFP/CSA Pilot Cash Programs Program Program Cash Program NGOs Implementation WFP/CSA 1. Red Cross, CSA (WFP) Ministry of Agency (+ technical 2. ACF Health support) 3. World Vision (UNICEF) 4. GRET Target groups Extreme poor 1 & 2. Extreme poor Extreme poor households households, in food households Mothers with insecurity, 3. Pregnant young children women and/or children (0-24 months) 0–2 years old Geographical areas Nouakchott 1. Gorgol Nouakchott Assaba, Brakna, 2. Guidimakha Gorgol, Guidimagha 3. Brakna Indicative number 2011: 30,000 (12,400 in 1. 900 2011: 3,800 of beneficiaries Nouakchott, 18,000 in 2. 400 2013: 12,000 Gorgol) 3. 450 Years of activity 2011–present 2011–2013 2011, 2013 Indicative annual 2011: MRO150 million 2013: MRO540 spending (MRO 2012: MRO2.6 billion million (transfer million) 2013: MRO2.1 billion value) Funding EU, USAID, UK EU WFP Benefits MRO15,000/month for Between MRO39,000 - MRO15,000 face 3 months per year 70,000 (over 3 value (real transfers) MRO23-30,000) Source: Government and program data 13 42. The largest cash transfer program to date was implemented by WFP/CSA. The intervention targeted extreme-poor households (falling under the national threshold of US$1/person/day), who received a transfer of MRO15,000 (equivalent to US$50) per month for three months during the pre-harvest gap season. This amount was set based on the cost of a one-month food basket for a household and represented approximately half a household’s average food expenditure. Beneficiaries were targeted following a combination of geographic and community targeting, subsequently verified by survey and proxy-means testing (PMT). The WFP/CSA intervention was carried out in both urban and rural areas, making transfers to 12,400 extreme-poor households in Nouakchott and 15,000 households in four vulnerable Wilayas (Assaba, Brakna, Gorgol, and Guidimagha). Despite technical difficulties, the program succeeded in transferring MRO1 billion (US$3.3 million equivalent) to poor and vulnerable households. Difficulties in setting up a functioning payment system in the first year caused the program to scale down benefits (from six payments to three). Over time however, the program has succeeded in developing key targeting, beneficiary registration, and payment instruments. 43. Interventions by NGOs followed various methodologies and concentrated on creating knowledge. These were small-scale interventions (between 400–900 beneficiary households) and were carried out in very isolated areas of the country. The objectives focused mainly on food insecurity and nutrition of the poorest households (and, for the GRET program, on pregnant women and mothers of young children). Livelihoods activities were also tested in complement to the transfer (World Vision). Transfers were concentrated in the six months around the pre-harvest gap and were complemented with transfers in kind and information sessions. 44. A food voucher pilot program (near-cash transfers) was tested as an alternative approach to cash. A first pilot food voucher program was implemented in Nouakchott from 2011, providing two months’ support to 3,800 beneficiaries in vulnerable neighborhoods. A second, larger- scale food voucher pilot was scheduled for 2013, targeted to the 12,000 households identified as extreme-poor through the targeting mechanism developed for the cash transfer program. The vouchers were to have a face value of MRO15,000 and be redeemable in the Boutiques Emel, thus reaching a market value of MRO23,000–30,000 (counting the supplementary 30–50 percent subsidy provided in the Boutiques). (2) Food Transfers 45. Food transfers are currently, by far, the main safety net in Mauritania. They include a variety of programs—free distribution, sale of subsidized food items targeting the poor, nutrition programs, school feeding programs, and social programs whose support includes food packages. As part of its drought-response, Mauritania also implemented a large-scale program in support of livestock structured around the distribution of feed - a consequence of the importance of pastoralism to Mauritania’s rural economy, as well as the large number of Mauritanians dependent on this mode of production for their livelihood. Table 7 below summarizes the main aspects of the eight principal food transfer programs. The Emel livestock component is described in annex 2. 14 Table 7: Food transfer programs EMEL Strategy Secours aux Sinistrés Boutiques Stocks Alimentaires Emergency food (Emergency Relief Emel Villageois de Sécurité distribution program Program) (SAVS) Implementing CSA (rural areas) CSA (WFP) CSA (WFP) CSA agency Ministry of Commerce (technical support) (urban areas) (WFP) Food insecure poor Food insecure rural Rural extreme poor Disaster-affected Target groups communities communities Geographical areas Nationwide All rural regions Nationwide rural Nationwide Objective: 40 percent 2010: 15,561 Number of of the national 2011: 7,592 Beneficiaries population (2012) 2012: 585,000 Years of activity 2011–present 2003–present 2008–2012 2010–present 2010 900 million Annual budget 2011 8,000 million 140 million 700 million 232 million (MRO 2012 22,000 million 1,580 million 5,000 million million) 14,400 million n/a Discontinued 2013 GoM, WFP GoM, WFP, WB GoM, WFP, SNIM Funding source Foundation, Japan Daily ration of 4 food 15,000 tons 4-17,000 tons of Up to 6 months of Benefits/allocation items at 50-70 percent distributed/year cereal/year food + tent of market prices National nutrition School Feeding Centres de Centres program* Récupération d'Alimentation Nutritionnelle Communautaire (CRENAM) (CAC) Implementing Ministry of Health Ministry of Education CSA (WFP) CSA (WFP) agency (UNICEF) (WFP) (technical support) Vulnerable young School children in Chronic malnourished Moderately Target groups mothers and children vulnerable and food young mothers and malnourished insecure areas children children 9 food insecure National, 637 centers Tagant, Hodh El Geographical areas Wilayas Chargui, Assaba Number of 150,000 school 2010: 31,882 beneficiaries children/year Years of activity 2010–present 2011–present 2010 38.6 1,162 830 - Annual budget 2011 1,026* 1,162 377 410 (MRO 2012 1,243 1,220 480 606 million) 1,544 1,220 480 447 2013 EU, USAID, Japan, WFP, EU, IDA GoM, WFP, Saudi Italian Cooperation Funding source ECHO Arabia Two meals a day 15 days of dry rations Wet rations twice a during the school year to pregnant / lactating day for 2x3 months Benefits/allocation women and children to children aged 6- aged 6-59 months 59 months * Budget data from UNICEF (no figures were obtained from the Ministry of Health). Total program budget is therefore likely higher. Source: Government and program data. 15 The Programme Emel 46. The Boutiques Emel are a network of about 1,200 shops selling basic food items (wheat, rice, oil, sugar, pasta) at subsidized prices. The Boutiques Emel, initially designed to operate for 5 months in response to the drought of 2011, met with very strong demand, and continue to operate in 2014. The Boutiques Emel grew rapidly with a budget that went from MRO8 billion in 2011 to MRO22 billion in 2012 and forecast at MRO14 billion in 2013 (US$26.6 million, US$73.3 million, and US$46.6 million respectively). The number of Boutiques has grown from 700 shops (278 in Nouakchott and 322 in the rest of the country) in 2011 to about 1,200 (300 in Nouakchott, 900 in the rest of the country) in 2013. The program has three main objectives: distribute subsidized food to the poor, generate employment, and hold down food prices. The program sells a daily ration of four basic food products (wheat/pasta, locally produced rice, sugar, and cooking oil) at subsidized prices equivalent to 50–70 percent of market prices. Hence, this program is a combination of subsidy and food transfer. The program also had an additional objective of creating a national market for domestic rice production, which had hitherto not found favor with Mauritanian consumers. 47. The free emergency food distribution program was set up as a first line of response for the most vulnerable households. The program was designed for the poorest rural populations, especially small farmers, and 117,043 households benefited from free food distribution in 2012. Overall, about 585,000 people, without income and severely affected by the drought, received a distribution designed to cover a household’s food consumption for a period of three months. Several donors supported this intervention, the most important being the SNIM Foundation,9 the Government of Japan, and the WFP. The program was discontinued in 2013 due to favorable rains and harvests. 48. The cereal banks (Stock d’Alimentation Villageois de Securité, SAVS) program is a network of village cereal banks whose objective is to increase food security and halt speculation. Created after the food crisis of 2003, the SAVS program was designed as a long-term response to food insecurity in rural areas. Cereal banks are self-managed by village management committees, and use proceeds from the sale of grain to restock SAVS during the following harvest (when prices are low). In practice, village communities have often distributed the grains provided instead of selling them, leaving many SAVS rapidly depleted of cereal and capital. The number of functioning SAVS is estimated at 1,500–2,000 (out of 3,000 created). The SAVS were initially created with WFP support and have recently received support from other development partners, including the World Bank (US$5 million in 2012). The overall budget for SAVS was MRO140 million in 2011 (US$466,000) and MRO1,580 million in 2012 (US$5.3 million). 49. The second main component of the Programme Emel focuses on supporting livestock, an important factor in a country whose rural economy and society are highly dependent of pastoralism. The livestock component of the Programme Emel was extremely large-scale, budgeted at nearly MRO22 billion, or US$73 million equivalent, in the emergency plan for 2012. Such livestock support programs do not normally fall within the classification for safety net interventions. However given the social and economic reality of Mauritania, half the population of which is estimated to be considered pastoralist, the logic of livestock support as a safety net becomes evident, even though its efficiency remains open to discussion. The livestock support was designed to protect the country’s livestock herds from depletion, and the rural economy, greatly dependent on pastoral activities. Furthermore, in the many areas of the country where banks are scarce, livestock represent the main 9 Charitable arm of Mauritania’s national mining company, the Société Nationale de l’Industrie des Mines (Mauritania). 16 reserves of capital for the rural economy. Loss of livestock would therefore be a severe shock to local employment and rural financial systems in Mauritania. 50. However, the majority of poor themselves do not have their own herds, and the Emel livestock support component likely benefitted wealthier individuals in the pastoralist communities, those with extensive herds tended by the poor. In the absence of evaluation it is difficult to emit any conclusions on the value and impact of this program, yet while it is plausible that the support to livestock may have succeeded in preserving the stability of the rural financial system, the value of such an intervention as a social safety net may have been limited. 51. Evaluations of the Programme Emel10 suggest that it managed to provide essential support to destitute communities and helped avoid a potential humanitarian crisis. However, the Government’s evaluation of the Programme Emel also identified some limitations, notably in the methodology used to select beneficiary communities. A recent World Bank’s evaluation also noted some weaknesses in the distribution of the Boutiques over the territory, but found that the overall targeting of the program was rather positive (Section III, C has a more detailed analysis of the Boutiques Emel’s targeting). Furthermore, the evaluation also noted that the Government’s large-scale distribution operation through the Boutiques Emel appeared more cost-effective than distribution by the private sector of the same food items both outside of and in Nouakchott, indicating inefficiencies and distortions in local food markets. Further investigation would be required to clarify this potentially important issue. (3) Nutrition programs 52. Nutrition programs in Mauritania are growing, from a total budget of MRO869 million (US$2.9 million) in 2010 to an estimated MRO2.4 billion (US$8 million) in 2013. The country has four main nutrition programs, led by the Ministry of Health, the CSA and the MASEF. 53. The national nutrition program (Programme National de Nutrition), run by the Ministry of Health, targets malnutrition and micronutrient deficiencies. The Ministry of Health also leads the Conseil National de la Nutrition. Interventions focus on infant and young child feeding practices, good nutrition, micronutrients, management of severe acute malnutrition, and capacity building (training, coordination, and emergency response). The program covers the most vulnerable Wilayas in the country, and resources are allocated according to the annual food insecurity and vulnerability survey. The program is supported by UNICEF, the EU, USAID, JICA, and ECHO. The program saw an important scale-up in 2011, reaching MRO1 billion (US$3.3 million), and has continued to grow in 2012 (MRO1.2 billion, or US$4.1 million) and 2013 (MRO1.5 billion, or US$5.1 million). 54. The CSA manages two programs, the CACs and the CRENAM. The Centres d’Alimentation Communautaire (CAC) is structured around treatment centers for malnutrition among children and pregnant or lactating women. In 2010, the CACs formed a network of 65 nutrition centers in four Wilayas of the North (Adrar, Zouerate, Inchiri, and Nouadhibou). The program supported 1,884 children with supplementary feeding for a nine-month period (2010). Twelve CACs were also opened in Nouakchott to distribute additional food to vulnerable children aged 6–59 months for a period of three months. The CSA manages the centers with technical and financial support from the Italian Cooperation and the WFP. The program’s annual budget was an average MRO487 million (US$1.6 million) for the 2010–2013 period. The CRENAM are treatment centers for moderate 10 A 2012 Government evaluation and a 2013 World Bank evaluation of the boutiques. 17 malnutrition among children and pregnant or lactating women. The CRENAM (Centres de Récupération Nutritionnelle en Ambulatoire pour la malnutrition Modérée) are mobile and temporary centers, located in a house or location made available by the village and providing the equivalent of 15 days of dry rations to pregnant/lactating women and to children aged 6–59 months. The program is supervised by the CSA with WFP technical and financial support and implemented through a network of national and international NGOs. It is implemented in eight vulnerable regions with special focus on agro-pastoral areas, and Nouakchott, with a total of 637 centers. The centers’ distribution responds to annual vulnerability studies. The program’s total budget was an average MRO530 million (US$1.8 million) over the 2010–2013 period. 55. The Programme d’Appui et de Suivi de la Nutrition (PASN) is a community-based outreach program managed by the MASEF. The program is implemented in eight Wilayas and is structured around 233 community nutrition centers (CNC) in which a network of community nutrition agents and volunteers provide regular training sessions on good practices in mother and child nutrition, cooking lessons and provide regular distribution of food complements. (4) School feeding 56. The national school-feeding program targets nine rural Wilayas marked by low attendance, high food insecurity, malnutrition, and poverty. Nominally, it covers 50 percent of the schools in the country and focuses on primary schools with 50 to 300 pupils. The program is managed by the Ministry of Education through its Direction Nationale de la Nutrition et de l’Éducation Sanitaire. The WFP provides food inputs and logistics. School management committees are responsible for program supervision and food stocks. 57. Once a major element of Mauritania’s social safety net system, the national school feeding program is being superseded by other interventions. With a flat budget since 2009 (approximately an annual MRO1.2 billion, or US$4 million), the program has continued to support the same number of schools, but has had to reduce the number of meals served (and the number of days) over the school year to deal with price increases. In years of particularly high food prices (2008 and 2012, for example), meals were only provided for a period of 80–100 days in a school year of 180 days. The program is therefore not capable of achieving its objective of providing two meals a day to 150,000 pupils. The WFP is scheduled to withdraw from the program in 2016 and the Government is developing a national strategy for school feeding strategy that should hopefully help address the looming financing gap. It is useful to underline that school feeding is a core instrument in favor of educational achievement, with proven effectiveness at keeping children in school. (5) Price Subsidies 58. Energy prices are subsidized in Mauritania, costing up to 8.2 percent of national public expenditure or 2.5 percent of GDP at their peak in 2012. Before reform of the gasoline subsidy, the cost of subsidies was shared equally between oil products, natural gas, and electricity. In 2012, the cost to the Mauritanian Government was about MRO9 billion for oil products (gasoline and diesel), MRO11.7 billion for natural gas and MRO10 million for electricity. Elimination of subsidies on gasoline have brought the overall subsidy bill down, but costs remain high, with a forecast of 3.7 percent of public expenditure for 2013. Subsidies are still applied to petroleum products (diesel, kerosene), natural gas, and electricity. However, a reduction of about 30 percent is planned for the subsidy bill for natural gas and electricity between 2012 and 2013 (provisional figures). 18 Table 8: Fuel subsidies, percent of public expenditure (2008–2013) 2008 2009 2010 2011 2012 2013* Petroleum products - - - 3,500 9,000 1,000 8,000 Natural Gas 611 3,327 4,354 9,070 11,700 7,060 Electricity - 8,900 8,690 10,000 6,000 TOTAL (percent nat. expenditure) 0.3% 5.9% 4.3% 6.4% 8.2% 3.7% TOTAL (percent GDP) 0.1% 1.5% 1.1% 1.8% 2.5% 1.1% *Provisional figures. Source: Bank staff calculations based on Government data 59. Subsidy levels vary by product. Figure 2 shows an estimate of the level of price subsidies as of July 2011. Gasoline subsidies (which amounted to about 5 percent in 2011, prior to their elimination) were far below subsidies for electricity and natural gas (28 and 59 percent respectively). Diesel fuel is subsidized at 12 percent. Strangely, the price for kerosene is 11 percent below cost. Figure 2: Subsidy rates for energy products Sources: Government data, IMF calculations (2011) (6) Labor-Intensive Public Works 60. There are several small cash-for-work or labor-intensive public work programs in Mauritania. The main programs are: (i) the Programme national d’insertion et d’appui a la micro- entreprise (PNIM, under the Ministry of Employment) which implements small-scale projects; (ii) the micro-projects and cash-for-work program which implements small community development projects in rural areas (under the CSA); and (iii) an income generating activity program (under the MASEF). CSA also implements a food-for-work program, which provides work for 25,000–50,000 beneficiaries a year in the nine most vulnerable Wilayas of the country, with particular focus on agro-pastoral areas. (7) Fee Waivers 61. There are a number of small existing fee waiver programs in Mauritania under the purview of MASEF. Several programs are focused on improving health access to the poor (Programme d’appui national aux hopitaux pour indigents, programme pour disponibiliser l’hémodialyse, and Programme évacuation des indigents à l’etranger). The hospital access program is a pilot program for access to free basic services that is being tested in two Wilayas and is based on a certification of poverty. A flat-rate obstetrical program is also under development with the Ministry of Health, with the support of the Agence française de Développement (AfD). While this program does not provide free services, it supports staggered payments in anticipation of obstetrical costs and includes an element of price subsidy. 19 III. Diagnostic of Social Safety Net Programs 62. The following section provides an overall analysis of the safety net programs in place in Mauritania. Five main observations are drawn. Firstly, Mauritania has seen overall expenditures on safety net rise sharply since 2008. Secondly, these increases have mostly been in response to shocks. Thirdly, the vehicle for support remains fundamentally focused around the distribution and subsidy of food. Fourthly, most programs (except universal subsidies) follow some form of targeting of resources, some of which have performed quite well. Finally, with a significant growth potential and a new strategic orientation, Mauritania today can effect a transformational change in its approach to social protection, which could have significant impact on resilience and chronic poverty. A. Expenditures have risen sharply 63. Mauritania has greatly increased its expenditure on social safety net programs since 2008. This rise results from the Government’s large-scale subsidy program and emergency response implemented from 2008. Expenditure has risen sharply from MRO8.3 billion in 2008 to an estimated MRO39.2 billion for 2013 (including fuel subsidies). Over this period, spending on social safety nets rose from 1.0 percent to 2.9 percent of GDP, and from 3.9 percent to an estimated 9.7 percent of Government expenditure. Excluding the costs of universal food and energy subsidies, expenditure increased from 0.9 percent of GDP in 2008 to an estimated 1.8 percent in 2013. Table 9 below sets out the structure of spending between 2008 and 2013. 64. The peak in spending was reached in 2012, reflecting programs designed in response to the 2011 drought. In 2012, spending on social safety nets climbed to MRO67.9 billion, or 5.5 percent of GDP and 18.9 percent of Government spending (including fuel subsidies). Table 9 shows that food transfer programs, mostly linked to Programme Emel, accounted for almost half of expenditure that year (MRO32 billion). Of this amount, two-thirds are attributable to the Boutiques Emel (MRO23 billion). The second major area of expenditure this year was on universal fuel subsidies (MRO30 billion). 65. The large drop in spending in 2013 is linked to the winding down of emergency food distributions tied to the Programme Emel and the elimination of gasoline subsidies. The elimination of the gasoline subsidy brought the cost of oil product subsidies (gasoline, diesel fuel and kerosene) to drop from MRO9 billion in 2012 to MRO1 billion in 2013 (see table 8 above). The good rainfall of 2013 also enabled the Government to reduce spending on emergency programs, eliminating emergency food distributions and reducing budgets for the Boutiques Emel, which saw their budget drop from MRO23 billion in 2012 to MRO14 billion in 2013. 66. Most of the funding for social safety nets comes from Mauritania’s government resources. The country relies to a large extent on national resources to finance its programs, covering 84 percent of the total cost of social safety nets (including fuel subsidies). If fuel subsidies are excluded from the calculation, Mauritania self-finances 64 percent of its social programs. This figure is notably higher than other countries in the region, which only cover on average 30 percent of the cost of their social safety net expenditures, the remainder being covered by donor support. This reflects the deliberate decision by the Government to take charge of the response to the 2011–2012 drought. 20 67. The elimination of gasoline subsidies has enabled Mauritania to bring its expenditures on subsidies under control. The total bill for fuel subsidies fell from 2.5 percent to 1.1 percent of GDP between 2012 and 2013. In light of the often complicated and politically delicate nature of such energy reforms, this is a notable success11. Indeed, it is estimated that half of African countries maintain fuel subsidy programs, which cost an average of 1.4 percent of GDP. In particular, energy exporters (such as Mauritania) assign on average at least 2.6 percent of GDP to these programs.12 This reduction in subsidies potentially frees up resources for more effective, targeted programs. Table 9: Spending on social safety net programs 2008–2013 (MRO million) 2008 2009 2010 2011 2012 2013* Type of program 1. Targeted cash transfers (1) - - 101 266 2,250 3,900 2. Targeted food transfers 7,388 3,235 3,725 12,614 33,226 19,075 2.1 Boutiques Emel - - - 8,000 22,082 14,384 2.2 SAVS 4,702 304 112 141 1,583 - 2.3 Emergency distributions 1,443 1,688 916 707 5,012 - 2.4 Support for disaster victims - - 214 232 - - 2.5 Support for vulnerable groups (2) - - 452 558 1,000 1,000 2.6 Nutrition (3) 43 43 869 1,814 2,329 2,471 2.7 School feeding 1,200 1,200 1,162 1,162 1,220 1,220 3. Universal fuel subsidies 611 12,227 11,414 21,260 30,700 15,000 4. Public works 361 351 973 1,096 1,600 1,250 5. Fee waivers (health) (4) 408 931 740 923 789 Total (including universal subsidies) Million MRO 8,360 16,221 17,144 35,976 68,699 40,014 Percent of GDP 1.0 2.0 1.7 3.0 5.5 2.9 Percent of Government expenditure 3.9 7.9 6.5 10.9 18.4 9.9 Total (excluding universal subsidies) Million MRO 7,749 3,994 5,730 14,716 37,999 25,014 Percent of GDP 0.9 0.5 0.6 1.2 3.1 1.8 Percent of Government expenditure 3.6 1.9 2.2 4.4 10.2 6.2 (1) including voucher programs for 2013 (near-cash transfers) (2) Programs in support of disabled people, early childhood development center for the protection and integration of children in difficult situations (MASEF) (3) National nutrition program (Ministry of Health), CAC and CRENAM Programs (CSA/PAM) (4) Programs for the evacuation of indigents to other countries, for provision of hemodialysis, for support of national hospitals for indigents, and social assistance program (MASEF). Source: Government and program data, World Bank staff calculations *: Budget forecasts. SAVS expenditure was not included in the 2013 data. 68. Mauritania’s social safety net expenditure is slightly above the African average. With 6 percent of national expenditure and 1.8 percent of GDP spent on social safety nets, Mauritania’s expenditure (excluding fuel subsidies) is slightly above the African average of 1.6 percent of GDP (and 4.4 percent of Government expenditure). The peak of 3 percent of GDP (and 9.9 percent of Government expenditure) in 2012, while tied to exceptional circumstances, is even above the average observed in middle-income countries in Africa, where average spending stands at around 2.3 percent of GDP and 7.0 percent of total Government spending. 11 Brazil coordinated the elimination of its national gas subsidy with an unconditional cash transfer for poor families. The transfer was relatively low, equivalent what the poorest quintile spends on gas, which limited distortions. 12 World Bank (2013) “Africa’s Pulse,” vol. 5. 21 Figure 3: Social safety net spending by category (2008–2013) A. Safety Net Spending 8.0 % GDP 7.0 6.0 Food transfers 5.0 Cash and near-cash transfers 4.0 Public works 3.0 Fee waivers 2.0 Fuel subsidies 1.0 - 2008 2009 2010 2011 2012 2013* B. Safety Net Spending including Emel livestock support 8.0 % GDP 7.0 6.0 Food transfers 5.0 EMEL Livestock support 4.0 Cash and near-cash transfers 3.0 Public works Fee waivers 2.0 Fuel subsidies 1.0 - 2008 2009 2010 2011 2012 2013* Source: Government and program data, World Bank staff calculations. B. The primary focus is responding to shocks 69. The rise in expenditure on social safety nets is primarily linked to crisis response. Mauritania’s largest programs were implemented in response to external shocks: the fuel subsidies program was launched in response to the crisis in world energy prices that began in 2008, and the Emel program in response to the drought that hit the country from 2010 –2011. The large increase in spending on social safety net of recent years has thus been mostly focused on responding to shocks: 90 percent of social safety net spending in 2013 was connected to crisis response. In 2009, when the fuel subsidies program was created, expenditure on crisis response as a share of total spending on social safety nets climbed from 19 percent to 86 percent. In 2012, when the Emel program was at its peak, emergency response accounted for 91 percent of total expenditures (see table 10). The remainder, between 8 – 25 percent, accounted for the sum of safety net programs focusing on chronic poverty. 22 Table 10: Chronic poverty and crisis response programs (percent of total expenditures) 2008 2009 2010 2011 2012 2013* Chronic poverty programs 81 14 25 14 9 15 Crisis response programs 19 86 75 86 91 85 *estimates Crisis response programs: EMEL (emergency food distribution, Boutiques Emel, SAVS 2012), fuel subsidies, disaster relief program (CSA). Permanent programs: nutrition and school meals programs, CSA programs (food-for-work, micro-projects), MASEF social programs, HIMO program. Source: Government and program data, World Bank staff calculations. 70. Mauritania’s spending on social safety nets, despite its significant growth, has not resulted in the creation of programs focused on long-term, chronic issues. Apart from crisis response programs, Mauritania has made no significant new investments in social safety net programs, despite the vision and roadmap provided by the National Social Protection Strategy and the PRSP III. Ongoing interventions, such as school feeding or nutrition programs, have received limited increase in spending over this period. The recent donor-supported cash transfer programs have also mostly been designed with cyclical shocks in mind. 71. Universal, untargeted, subsidy programs in response to price shocks are costly and bring little benefit for the poor. Recourse to subsidies can be an attractive option in times of crisis because it offers a solution that is both fast and relatively simple to implement, and quite effective in the short term. However, these subsidy programs, which were originally designed as temporary interventions, have expanded to take up a significant portion of national expenditure, despite the gasoline subsidy reform of 2012. Furthermore, except when targeted, the benefits of subsidy programs tend to be overwhelmingly captured by the non-poor. 72. Crisis response programs do not focus on reducing the long-term poverty or improving the resilience of vulnerable households. A crisis response program is suitable over the short term in specific conditions (drought, flood, a surge in fuel/food prices). However, Mauritania’s current focus on subsidized food distribution has extended well beyond short-term crisis response. International evidence repeatedly shows that emergency programs have limited effect on resilience or poverty levels among the population. In fact, long-term food subsidy and/or distribution programs may even increase vulnerability by reducing the incentive for the private sector to maintain functioning food markets. 73. It is worth noting, however, that the Boutiques Emel also has the objective of stabilizing food prices, which goes beyond immediate crisis response. The Boutiques Emel were designed to ensure access to affordable food for the country’s poor, but also stabi lize prices at a time of high price volatility. The Boutiques Emel program also supported domestic rice production by distributing nationally produced rice as a subsidized good. C. The main channel is food aid 74. Most social safety nets interventions in Mauritania are structured around food aid. Setting aside the universal subsidies for fuel and pilot cash transfer schemes, the principal interventions in social safety nets are all structured around the distribution of food. Even the remaining programs such as the small labor-intensive public works program and the MASEF social assistance programs focused on vulnerable groups rely on food aid (food-for-work, food benefits in social assistance programs). It is also worth noting that the large-scale cash transfer program managed by the WFP and the CSA was partly implemented in the form of food vouchers (to be redeemed in Emel boutiques). 23 75. Between 2008 and 2013, excluding fuel subsidies, food transfers represented, on average, 82 percent of total spending on social safety nets (see table 11 below). In 2012, at the peak of the program, Emel accounted for 75 percent of total expenditure on social safety nets in Mauritania. Of this, the Boutiques Emel have become by far the most important channel for food-based support in the country, accounting for over half of the total spending on social safety nets. The SAVS programs and emergency food distribution have seen their budgets fluctuate according to circumstances and the availability of funds from their financial partners. It is also worth noting that the budgets for nutrition and school meal programs have remained fairly constant, though their share of total spending has fallen as this has increased over the period. Table 11: Food transfer programs (percent total safety net expenditure, excl. subsidies) 2008 2009 2010 2011 2012 2013 Boutiques Emel - - - 54 58 58 SAVS 61 8 2 1 4 - Emergency food distributions 19 42 16 5 13 - Support for disaster victims - - 4 2 - - Support for vulnerable groups - - 8 4 3 4 Nutrition programs 1 1 15 12 6 10 School meals programs 15 30 20 8 3 5 Targeted food transfers (% total) 95 81 65 86 87 76 Source: Government and program data, World Bank staff calculations. 76. However, large food subsidy and distribution through the Programme Emel is unsustainable over the long term. Despite the good results obtained by the Boutiques Emel, the stakeholders involved recognize that the costs linked to setting up and maintaining the stores are not sustainable. The Boutiques Emel program must not only cover the cost of food subsidies but also all the outgoings connected to the supply and maintenance of a national store network, together with their employees. To these may be added the fees charged by intermediaries (the Fédération du Commerce in the provinces, Sonimex in the capital). For example, for the emergency food distribution programme, the operating costs reported by the CSA amount to 40 percent of the total value of the program (similarly, a recent evaluation of the World Food Program in Burkina Faso concluded that food vouchers are 40 percent cheaper than distributing food). The goods distributed can also produce distortions in local markets. Overall, evidence shows food-based programs to be, comparatively to cash or voucher programs, more costly in logistical terms (transport, warehousing, and distribution costs), and with high risks of loss (spoiled food, losses or theft at different points along the supply chain). Even if these costs may be legitimate in isolated areas or periods of food availability crisis, they are harder to justify outside of emergency contexts, and when local markets are functioning. 77. While food-based programs have an important role to play during crises, they cannot be the only basis for an effective national safety net system. Food aid is important as a rapid response to certain acute types of crisis, and can be justified in contexts when access to basic goods is disrupted, even for non-poor households, for example in isolated areas during a drought or in a natural disaster situation. Food-based programs such as school feeding and nutrition programs also play important roles in supporting the poor and vulnerable. However, in terms of chronic poverty reduction and building resilience among the population, long-term support programs (provision of access to health and education services) or cash transfers are more effective in transferring resources to households. 78. In this respect, Mauritania’s situation bears some parallels to that of Ethiopia, which in the 1980s had to cope with recurring droughts and repeated humanitarian intervention. However, despite the food distribution programs and emergency interventions, households gradually depleted their assets and fell into ever-deeper poverty with each passing year. This situation drove 24 Ethiopia to change its mode of intervention. It did so by adopting a system focused on addressing the root causes of household food insecurity with long-term programs, particularly with predictable transfers made through several programs (rural productivity and soil improvement, food-for-work, food distribution, cash transfers, etc.), as a complement to its emergency responses. D. Targeting of programs is improving 79. Social safety net programs in Mauritania are all (setting aside universal subsidies), in some form, targeted at vulnerable populations. However, these programs suffer from deficiencies in the implementation of the targeting, and thus are weakened in their effectiveness and impact. 80. Mauritania’s least effective intervention is the national fuel price subsidy program, which is not targeted. Fuel subsidies are universally recognized as an ineffective method of addressing the needs of vulnerable populations. While they are officially implemented to protect the level of consumption of entire population, in practice such programs strongly favor industry, the wealthy, and the urban middle classes, and international experience clearly demonstrates that it is the rich and not the poor who take the lion’s share of such subsidies. Moreover, as subsidy programs become established they tend to raise demand and become more costly, crowding out investments that could provide greater benefit to the poor. A caveat should be noted with regards to gas subsidies, which may contribute usefully to desirable social objectives, such as drawing the population away from unsustainable energy sources such as charcoal. Such subsidies may be refined to be better targeted to the poor, for example through the subsidy of small gas canisters (‘bonbonnes’) but not large-scale consumers. 81. IMF analysis (2011) shows that almost all the subsidy programs in Mauritania benefited the most affluent sector of the population. According to the study, the most affluent quintile of the population received 40 percent of the total value of subsidies, while the poorest quintile only received 5 percent. Mauritania’s subsidy program is thus highly regressive, since it clearly benefits the rich more than the poor it is intended to support. Table 12: Share of fuel subsidies enjoyed by different population quintiles Quintiles Total 1 2 3 4 5 Diesel fuel 8.0 12.9 17.7 23.5 37.9 100 Direct effect 1.1 2.4 6.4 13.3 76.8 100 Indirect effect 9.6 15.3 20.3 25.8 29.1 100 Butane gas 2.3 9.0 17.8 28.8 42.1 100 Electricity 4.9 10.0 17.1 26.1 41.9 100 Direct effect 1.9 7.7 12.7 25.0 52.8 100 Indirect effect 5.8 10.8 18.3 26.3 38.8 100 Total fuel 5.4 10.6 17.3 25.6 41.1 100 Source: IMF (2011) “Reform of Fuel and Food Subsidies in Mauritania” 82. The various interventions of the Programme Emel have a broader set of social and political objectives, beyond support to those affected by crises. Of note, for example, is the objective of stabilizing food prices on domestic markets, as well as the creation of a distribution network and market-creation for domestic rice production. As such it is not appropriate to judge the program purely on its efficiency to support the poorest. 25 83. Furthermore, each intervention is targeted to a different segment of the population. The emergency free food distribution is targeted at the very poorest households, the Boutiques Emel are targeted at poor households, and the SAVS are targeted at poor agricultural producer households. The livestock assistance program does not specifically target the poor (and mainly benefits those with larger herds) but serves to protect accumulated capital and economic activity in rural areas and the traditional livestock-based financial system. 84. Overall, the Programme Emel has demonstrated few inclusion errors in areas where it has been implemented. The evaluation carried out at the end of 2013 noted that the program’s targeting of the poorest segments of the population has been relatively successful. The allocation of resources (restocking of SAVS, setup and stocking of the Boutiques Emel, and free food distribution) is carried out on the basis of the zones identified as being most at risk. For SAVS and free food distribution, the allocation of resources at the household level is carried out on a community basis. The Boutiques Emel also possess additional targeting mechanisms, including imposition of a private cost in terms of waiting times, daily rationing, and the quality of products sold (which are generally of lower quality). In rural communities, a practice of redistribution in favor of the most vulnerable households was also observed. 85. Nevertheless, Programme Emel presents a number of issues tied to exclusion. The World Bank assessment (summarized in Appendix 1) notes that the poorest population segment lacking cash resources to cover their food needs has only been able to fully benefit from the free distribution part of the program, which was limited, and did not have the resources to purchase food supplies, even at the subsidized prices of the Boutiques. Also, a significant urban bias was observed in the design and execution of the program, both in terms of location and supply of the stores. Thus, the program disproportionately benefits Nouakchott, which is home to one quarter of the Boutiques Emel in the country but only 8 percent of the poor (see figure 4). At the other extreme, Brakna is home to 16.1 percent of the country's poor, but only 8.9 percent of Boutiques. Even in rural areas, the Boutiques are located in the main towns, which can be far away from the most impoverished families. This is less a failure of targeting by the intervention than a fundamental limitation of the model: it is not possible, logistically or financially, for a network of subsidized public boutiques to reach the level of penetration necessary to attend to the most isolated and vulnerable households in the country. 86. Overall, the targeting of Programme Emel was a success in the short term, but has important limitations. The targeting of Programme Emel worked as a specific response to the emergency and to the crises. The targeting power of Programme Emel lies in its relative simplicity (geographic distribution of resources according to food insecurity indicators, allocation of resources by communities themselves and limitations placed on access to the Boutiques Emel to discourage their use by the well-to-do). However, the targeting of Emel interventions passes over too many poor households to be considered a success as a targeting mechanism that satisfies the objectives of social protection at a national level, over and above specific crisis response purposes. 87. National school meals and nutrition programs are relatively well targeted. These types of programs tend to rely on a combination of geographical and categorical criteria for targeting beneficiaries. Priority zones are identified through food insecurity and vulnerability surveys, and resources are allocated on this basis towards young mothers and children (nutrition programs) or school-age children (school meals). 26 Figure 4: Poverty incidence and distribution of Boutiques Emel and SAVS Boutiques Emel: situation in 2013; SAVS: cumulated tonnages distributed in 2012 and 2013. Sources: ONS 2009, Profil de la pauvreté 2008 – CSA 88. In practice, however, these programs also entail significant risks of exclusion. School meals or nutrition-type programs also hold significant risks of exclusion, arising from two principal factors. First of all, the under-financing of these programs sets limits on their scope. The number of school meals or food rations for malnourished children remains well below requirements (constraints on the supply of services), even more so due to combination of flat budgets and the increases in international food prices that have gradually reduced the volumes of food distributed through these programs (see previous section). The poorest households tend to have the least access to health and education services (due to isolation, distance or cost considerations). The school feeding program targets schools of average size (50 to 300 pupils), which may favor children in urban areas or larger villages, and penalizes smaller schools, which are often found in the most remote—and the poorest— areas. Finally, without actively identifying the poorest on an individual or household basis, it is inevitable that many will fail to be included in social programs. 89. The targeting methodologies implemented by recent cash transfer and food voucher programs are an important step forward. The methodologies developed by the recently implemented pilot cash transfer programs reveal a shift in perspective towards treating the household as the focus of targeting. The creation of registers and databases has been an important step towards rigorous identification of poor and vulnerable households, and has already enabled programs to coordinate transfers with additional health, nutrition or livelihood services. Identification and registration instruments also allowed programs to keep track of beneficiaries through time. For example, the food voucher program carried out in Nouakchott in 2012 was based on the lists of beneficiaries identified by the 2011 cash transfer program. Moreover, these mechanisms have managed to establish transparent selection procedures anchored in objective data and criteria. 27 E. The elements are in place for a transformational change 90. The elements are in place in Mauritania for transformational change. Between 2013 and 2014, Mauritania has adopted the strategic frameworks and established the institutions that could serve to implement national safety net and poverty reduction programs. With strong growth forecasts and mining revenues providing continued windfall gains and fiscal space, Mauritania is in a unique position to make real progress in the fight against poverty. 91. The national Emel program has shown the country’s capacity to rapidly implement large-scale programs. Through its programme Emel, Mauritania has proven both its capacity to mobilize significant resources to provide effective crisis response and its capacity to coordinate large- scale, national programs. There are many lessons to be drawn from the successes and challenges faced by the Emel program – for example on the allocation of resources following objective data and criteria, avoiding the pitfalls of political interference and ensuring careful monitoring and oversight – which can guide the development of national, long-term, targeted safety net programs. The recent implementation of cash transfer programs has also created a valuable experience base for large-scale programs. 28 IV. Reforms to Increase Impact on Chronic Poverty and Resilience 92. In light of the analysis highlighted above, and the objectives outlined in the Government’s strategic documents, it may be useful for the Government to consider pursuing two paths: setting up a targeting mechanism and creating a cash-transfer program on a national scale. The following discussion lays out some elements to consider for each of these.13 A. Target the Poor and Vulnerable Why Target? 93. Focusing resources on the poor or vulnerable increases the services that can be provided with a fixed budget or reduce the cost necessary to achieve a given impact. The gain attributable to targeting is significant. For example, if all the benefits of a transfer program were paid to the poorest quintile rather than spread uniformly across the entire population, the budgetary savings or the difference in terms of impact on poverty for a given budget would be five times greater. In practice, however, the gain is never quite so high, because targeting is not an exact science and has its own costs, including administrative expenses, social costs, the transaction costs incurred by the beneficiaries of the program, the cost of unintended side effects, and the political costs that may erode support for the program. The extent of targeting errors and costs generally varies by context and by the targeting methods used. These aspects must therefore also be evaluated carefully when developing a program or policy. 94. International experience shows that targeted programs can succeed in focusing their resources on the poor without presenting unacceptable exclusion errors and costs. It is difficult to take into account judgments about social and political costs, partly because the measurements used vary greatly, and partly because the debates on the question are often polemical. The widespread and increasing interest of policymakers in targeting, however, suggests that these costs are not crippling. Core questions 95. To use targeting effectively, a consensus must be found on four questions. These questions, summarized below, must take into account complex political and ideological considerations specific to each country context. (1) Whom to target? Distinguishing the poor from the non-poor is not easy. Despite the fact that there is a consensus in Mauritania on how to statistically monitor poverty using the poverty line, in practice many programs apply their own standards and criteria. Different notions often coexist—poor, indigent, vulnerable, at risk, or excluded. And different actors can be interested in differing aspects of poverty—food insecurity and or malnutrition, monetary poverty, belonging to a disenfranchised community, physical handicap, etc. The issue of preference for the “deserving” poor over the “lazy undeserving” is also politically sensitive. Programs that specifically target the poor may sometimes perceived as being at the expense of the middle class, or as necessary investments to promote inclusion and preserve social stability. The 13This section draws upon the analysis submitted to the Moroccan authorities by the World Bank in its report "Ciblage et Protection Sociale au Maroc: Note d’orientation stratégique" (Strategy Note on Targeting and Social Protection) published in 2002. We thank the project managers, Bénédicte de la Brière and Nadine Poupart, for permission to use it. 29 reluctance to abandon the middle class has led to the idea of “inverse targeting,” which means excluding the richest (top 20 percent) from the benefits of universal programs while keeping the programs available for everyone else. (2) What to target? Which programs should be targeted programs provide depends on the programs in place and their objectives. In Mauritania, considering that most of the social programs reviewed (transfers, food support, SAVS, etc.) above are already targeted in some form or another, the question is more on how better to target existing and new programs. (3) How to target? This question depends on program characteristics and objectives, and hinges on technical implementation and good governance. The challenge lies not in being able to distinguish between the poor and non-poor, but in establishing a mechanism for registering potential beneficiaries and the process for accepting and rejecting cases, the capacity to collect information, and mechanisms for handling complaints and appeals, and inter-institutional coordination. The possibility of decentralizing certain targeting activities raises questions about the techniques used (computerization) and the advantages and disadvantages of physical proximity to beneficiaries including the risk of being hijacked by local interests (registration, determination of eligibility, registry management). (4) Who does the targeting? This question is as context-dependent as “Whom to target?” In Mauritania, the Social Protection Steering Committee is currently tasked with supervising the development of an appropriate institutional structure to bear responsibility for targeting in the country. Targeting Methods: Overview of International Practices14 96. Several approaches can be used to direct public spending and services to specific populations. Some require a prior assessment of the applicant’s eligibility (individual or household). Others restrict eligibility to certain categories of people, for example the residents of a given geographical region or people of a particular age. Still others are designed to discourage individuals who do not need the program’s benefits, without, however, explicitly prohibiting them from participating (Table 13). Table 13: Targeting methods Method Description Only people below a specified threshold of resources are eligible. Resources Declaration of resources, may be verified by independent sources (e.g., tax office, social security verified or not office), or by documents submitted by the people themselves (e.g., pay slips), or may not be verified at all (people taken at their word). Individual or A score is calculated for each candidate based on a small number of Assessment of resources household characteristics that are easily observable and difficult to manipulate, to by multidimensional targeting based which a weight is assigned (ideally obtained from factor or regression eligibility tests (“proxy on indicators analysis of household data). Eligibility is determined by comparing the score means tests”) against a predetermined cutoff. A community leader or group of community members whose principal Community targeting functions in the community are not related to the transfer program identifies who in the community will receive benefits. Categorical Geographic targeting Eligibility is determined by place of residence. 14 See Grosh et al., 2008. 30 targeting Eligibility is determined by age, gender, or some other demographic Demographic targeting characteristic (widowhood, etc.). Specific vulnerable Selection by category, such as people with disabilities, orphans, the groups unemployed, etc. The use of low pay or limited working hours (on labor-intensive public Limited pay or working works schemes, for example) to attract only individuals with a low hours opportunity cost. Self-selection Low-quality The free or subsidized goods or services supplied are “inferior” (for goods/services example, poor-quality wheat). Source: Adapted from the review by David Coady, Margaret Grosh, and John Hoddinott, 2004 Advantages and Disadvantages of the Various Targeting Methods 97. Most of the above targeting methods may be applied to most programs, but some go hand-in-hand with certain types of intervention. For example, the low-pay-based self-selection method applies only to labor-intensive public works schemes. In other cases, however, an appropriate method or combination of methods has to be chosen. Often, a program will use multiple methods, which generally leads to better targeting than exclusively using a single method. 98. The selection of a household targeting method depends on several factors. Specifically, the main factors are: (i) cost and administrative capacity, (ii) technical feasibility, given the extent of the informal sector of the economy, and (iii) political acceptability. The targeting method chosen must maximize accuracy at an acceptable cost and ensure transparency. Endorsement and ownership by the population is also an important aspect to consider. Individual or Household Targeting 99. Verified declaration of resources. This is the gold standard in terms of accuracy. Thoroughly verified information also contributes to credibility and transparency (provided the checks are standardized, and all potential beneficiaries receive the same treatment). However, it can be expensive and administratively complex to verify resources, and because this method depends on revenue and tax declarations, it is often impossible in countries where the employment market is largely informal, such as in Mauritania. Verified declaration is the most appropriate method when the declared income is checkable, when a form of self-selection filters out non-target groups, when there is substantial administrative capacity and/or when the benefits are sufficiently high to justify the administrative cost of checking. 100. Qualitatively verified or non-verified declarations of resources. Sometimes no checks are carried out at all: the program agent simply records what the candidate says. However, a social worker may visit the individual or family at home to verify that their observable standard of living (which reflects income/well-being) is consistent with the numbers declared. The factors taken into account will more or less exhaustively reflect the household's needs and means, without quantifying them. 101. Assessing resources via a proxy means test (PMT). This is often considered a worthwhile alternative in developing countries with a largely informal employment market. This targeting approach was pioneered in Latin America and forms the basis of the large-scale transfer programs in those countries. Using PMT, evaluations find that between 80 percent and 90 percent of the benefits of the Chilean and Mexican programs (Puente, Chile Solidario, and Opportunidades) successfully reach the poorest two quintiles of the population. Moreover, the cost per interview varies between US$2.3 31 and US$8.4 (between 9 percent to 34 percent of the cost of checking resources in the United States 15). The administrative demands of PMT, while still complex, are more within reach of developing countries than with detailed verifications of declarations. 102. Community targeting. This passes the responsibility of targeting to the communities themselves. Community targeting gives voice to the poor, reflects the local priorities and knowledge of household conditions, and can, under the right conditions, achieve good results. Community targeting may also reinforce social problems such as major divisions (along religious, ethnic, or caste lines), and be seized on by local elites for discriminatory purposes16. The conclusions of a comparative study of community targeting and proxy means testing are also mixed: community targeting can be less efficient than PMT, particularly for the population strata earning close to the poverty line, but provides greater satisfaction to beneficiary communities17. Categorical Targeting 103. Demographic targeting. Demographic targeting is usually age-based. Family allowances and pensions generally use this method. The administrative burden of demographic targeting is lower, and is politically popular as it is associated with the principle of universality, and does not stigmatize the beneficiaries of such programs. The potentially low correlation of demographic characteristics (age, for example) with poverty is, however, a limiting factor. Demographic targeting is thus a least costly method, suitable for situations where a particular demographic characteristic is closely correlated to income or when it is combined with self-selection. For example, food supplements can be distributed to children who use public health services when the wealthier categories use private health facilities. 104. Geographic targeting. This is a method whereby eligibility is determined by locality: people living in an identified region are eligible, whereas those living outside it are not. The advantage of geographic targeting is its administrative simplicity. There is also no risk of creating stigmatization, as both the poor and non-poor living in a given region benefit from the program. However, the effectiveness of geographical targeting is low when poverty is not concentrated, and is dependent on the accuracy of the poverty maps. Effectiveness also increases as estimating techniques improve and are applied to smaller areas. Geographic targeting is pertinent when living standards vary significantly from one region to another, when administrative capacity is limited, when the value of the benefits is low, and/or it can be supplemented by self-selection. For example, in 2004 the World Bank carried out an assessment of targeting in Morocco which concluded that, in rural areas, the use of geographic targeting at community level or infra-community level reached the poor while minimizing spillover onto the non-poor, particularly if this method was coupled with other targeting methods. In contrast, in urban areas, geographic targeting at the community level was not effective and required other methods. 105. Geographic targeting is often combined with other forms of targeting. A review of 31 conditional cash transfer programs and found that 27 of them used geographic targeting, which, in 22 of those cases, was combined with other methods, generally proxy means tests or resource declaration, and in 7 cases combined with community targeting. 15 Castañeda and Lindert 2005 16 Conning and Kevane 2000 17 Alatas, Banerjee et al. 2012 32 106. Targeting to specific, traditionally vulnerable groups. Certain groups are particularly likely to have low levels of education, be socioeconomically marginalized, and have limited means. They can also be victims of discrimination, which hampers their capacity to generate independent income. Other groups face specific problems: the elderly can see their health decline, people with disabilities are hampered by physical and social barriers, and immigrants may not have access to the full range of services, or if they are clandestine or have no papers, they may hesitate to use the services available to them. Households that include individuals who fall into specific categories tend to be poorer than households that do not include such individuals, however this correlation is not universal, so care is to be taken when considering this approach. Self-Targeting 107. Self-selection or self-targeting. Self-selection by offering low pay or low-quality work coupled with geographic targeting is often used for certain types of program, particularly labor- intensive public works. Self-selection, however, is not always an ideal solution as it can have social costs: stigmatization, awkward and/or humiliating procedures, notably the distribution of goods and services in out-of-the-way centers and/or after a long time waiting in line, or the poor quality of the subsidized products. Although administrative costs are lower, it has been shown that these methods involve stigmatization and a deprivation of important benefits18. Way forward 108. Beyond the choice of targeting method, it is the quality of the concept and the quality of implementation that determine results. Awareness campaigns can help to keep exclusion errors to a minimum. To limit errors of inclusion, eligibility criteria must be carefully defined to distinguish the poor from the non-poor, and to do so at an acceptable cost. Targeting systems should be dynamic and flexible, and be capable of including newly-poor households and excluding those who are no longer eligible. To obtain good targeting results also requires trained staff, well-defined rules and procedures, clearly assigned and coherent institutional roles, adequate information systems, and a monitoring and evaluation system. Developing such systems is costly, complex and time-consuming. 109. The targeting method must then be evaluated to see if benefits are actually reaching the intended target. Key questions to be answered include: What proportion of the beneficiaries of a program are actually poor? What proportion of the poor are covered by the program? Do changes in eligibility rules actually reduce the proportion of non-poor beneficiaries? To what extent does expanding the program also expand coverage of the poor? 110. Targeting represents a significant change of mindset, and can be an important progress in the social contract between the state and its people. Such a change is important, and must be based on a national consensus and an official definition of household wellbeing/welfare. Once a national registry of targeted households is in place then it is important for social programs to align themselves to and apply that single concept of welfare. Each program can of course, on the basis of the registry information, add filters and/or change household eligibility thresholds to reflect its priorities and objectives. 18 Alderman and Lindert 1998 33 B. Add a National Cash Transfer Program to the Panoply of Programs 111. In recent years, many countries have adopted long-term cash transfer programs, in Africa and around the world, as an effective means of fighting extreme poverty. A review of cash transfer programs in Africa19 identified over 120 programs implemented across 39 African countries between 2000 and 2009, with many more currently underway. Similarly, practically all the countries in Latin America have set up programs of this type. There are also large-scale programs in South and East Asia. The interest focused on programs that use cash transfers to encourage parents to invest in their children’s education over the long term, has even spread from developing countries to developed ones, as demonstrated by the recent experiences in New York and Washington and the pilot program in France Impact on poverty 112. Cash transfer programs are effective instruments in poverty reduction. In Africa, evaluations show Zambia’s CGP program has led to a 5 percent reduction in poverty, 11 percentage point reduction in the poverty gap and 11 percentage point reduction in the severity of the poverty gap (using severe poverty line). Kenya’s cash transfer program for orphans and vulnerable population led to a 13 percentage point reduction in poverty. South Africa’s transfer programs together reduced the poverty gap by 45 percent. Impact on human development 113. In the last decade, cash transfer programs have become one of the most widely used tools of social protection in developing countries to support families living in poverty. Cash transfers are increasingly being used for income support and human capital promotion among the poor. Programs provide money transfers to poor households over a number of years, conditional on compliance with a number of health, education, and nutrition conditionalities. Latin America is the leading region in the world in the use of this type of program, with currently over 90 million people benefitting from regular transfers on the continent. Countries such as Brazil, Mexico, and Colombia were pioneers in the implementation of long-term cash transfer programs and in the use of more precise targeting and monitoring systems. Following on this, many African countries today are preparing or implementing programs inspired by the models of Latin America and adapted to local realities. 114. International experience has demonstrated that cash transfers can have significant impacts on human capital development over the medium and long term. Cash transfer programs can play an important role in improving the distribution of growth outcomes across the population, and improving social indicators. Studies show that, over time, such social safety nets increase the quantity and quality of households’ consumption; improve children’s nutritional intake, education and health outcomes; and increase expenditure towards productive activities necessary to strengthen the resilience of the poorest.20 115. Recent experience in neighboring Senegal, showed the positive impact of cash transfers on the poverty, health, and education of children. The impact evaluation of the cash transfers provided through the NETS child nutrition project (Projet Nutrition ciblée sur l’Enfant et Transferts 19Garcia & Moore, “The cash dividend: The Rise of Cash Transfer Programs in Sub-Saharan Africa” (2012) 20See Independent Evaluation Group (2011) and “State of Social Safety Nets, 2014” World Bank (2014). It is important to note that in no evaluated country did cash transfers ever lead to increased expenditures on alcohol or tobacco. 34 Sociaux), found that, at the household level, benefits were predominantly dedicated to food purchases, leading to increases in the number of meals reported and reductions in negative coping strategies. 21 Mothers showed increased awareness of reproductive health care, through trainings provided as part of the program. The intervention led to: (i) more diverse diets for children 0 to 23 months old, (ii) the adoption of better eating habits for children under the age of 5, (iii) the recommended number of meals for children 6 – 23 months old, (iv) a reduction in infant morbidity, particularly diarrheal disease, and (v) improved vaccination coverage. The evaluation of the program of education subsidies for vulnerable HIV/AIDS-affected children shows positive impacts on school enrolment and on civil registration. 116. Such results are confirmed internationally, with the Bolsa Familia in Brazil resulting in reductions in food insecurity and increases in food consumption among participating poor families. Consumption increased as a share of total spending among beneficiary families, especially in expenditure on food, education, and children clothing. For each real received as a transfer, 62 cents were spent on food, as opposed to 24 for non-beneficiaries. For 74 to 85 percent of families, in addition to food consumption, the quality and variety of food consumed, the number of daily meals, and the quantity of children’s food increased. The beneficiary families’ diets were 5.1 percent more diverse than that of non-beneficiaries22. 117. Cash transfers also lead to increased school enrolment and attendance, though impacts tend to vary by program. In Africa, primary and secondary school enrollment increased in Kenya, Zambia, Malawi, Ghana and South Africa, and in some cases led to increases in the highest grade completed (Kenya), reductions in grade repetition (Ghana) and absences (South Africa). In Brazil, evaluations show that Bolsa Familia increased enrollment rates by 5-7 percentage points and had small effects on dropout rates and grade promotion in the short term, and that enrollment rates increased by 13 percent in the long run.23 The impact on education outcomes has been strongest for the most vulnerable or traditionally neglected children, but impacts on actual learning outcomes have been lower than initially expected. Making Cash Transfers Conditional 118. The addition of conditionalities to a cash transfer program can help attain a series of socially desirable outcomes, most often focusing on the health and education of young children in beneficiary households. These efforts often have to go hand-in-hand with strengthening the educational services offered, training, and the quality of teaching. In the health and nutrition fields, although the supply issue (in terms of coverage) has not yet been settled, there are barriers and unsatisfied latent demand that could be stimulated by Conditional Cash Transfers (CCTs). CCTs can help change certain behaviors that are tied to high levels of child malnutrition, especially when combined with courses in infant feeding and appropriate care practices. They could also contribute to changing certain social norms regarding the mobility and independence of women. In the field of health and nutrition, the required conditions generally include a set of preventive medical visits to monitor the growth and vaccination of children under age 5, prenatal monitoring of mothers and sometimes their regular attendance at information sessions on health, nutrition, hygiene, and family development. In the field of education, conditions tend to be structured around registering children in 21 Institut Fondamental d’Afrique Noire (IFAN). “Impact Evaluation of Cash Transfers in the NETS Program of Social Transfers Targeting Children” University of Dakar (2012). 22 Ministério de Saúde. 2007. Avaliação do componente de saúde do Programa Bolsa Família (2008). Repercussões do Programa Bolsa Família na Segurança Alimentar e Nutricional. Rio de Janeiro. 23 Glewwe and Kassouf (2008, 2010) 35 school, consistent attendance, and the achievement of academic results (moving on to the next level, passing certain exams, etc.). Conditions may also include requiring participation in community work, adult literacy campaigns, or vocational development campaigns. Most programs also rely on a community support component, either by social workers or by community focal points. Most programs pay the funds to the mother or father responsible for the children, or even to the child in the case of adolescents or orphans. The relative value of transfers varies from 20 percent of average targeted household consumption (Mexico) to less than 4 percent for programs in Bangladesh, Pakistan, and Cambodia. 119. Including a health/nutrition component in a CCT program can stimulate a virtuous circle of learning. The brain of a malnourished child has less chance of developing normally, especially if the malnutrition is associated with harmful care practices. Such truncated development has long-term consequences on intelligence, and predisposes the child to chronic diseases (diabetes, obesity, high blood pressure) in the adult years. Furthermore, children who go to school hungry or with inadequate diets, or who are affected by infectious diseases, have limited learning capacity, which all have a cumulative effect during their school years. Nutrition and education go hand-in-hand and are mutually supportive. Many CCTs are supplemented by community programs to stimulate and develop early learning, such as the Atención integral a la Niñez Comunitaria strategy in Central America aimed at involving the entire community in early learning development and changes in the approach to childhood awareness. Catalytic effect on social services 120. In many countries, cash transfer programs have contributed to accelerating the modernization of social services, particularly through the creation of national social registries and the development of unified identification methods and payment mechanisms. The creation of national registries and databases of poor and vulnerable households served to support the targeting and management of beneficiaries for cash transfer programs. However, setting up these dynamic databases of poor and vulnerable households also often served as a first step in harmonizing and coordinating interventions aimed at poor households. Lastly, cash transfers require a higher degree of coordination between beneficiary databases, payment systems, and eligibility tracking. This stimulates an improvement in vertical integration—between the various administrative echelons (central, regional, and local)—and horizontal integration—between sectors (typically between the departments for social assistance and development and the departments for health, education, and sometimes labor and jobs, agriculture, and rural development). In many countries, cash transfers are at the cutting edge of administrative reform of social assistance. Cash transfers have also stimulated innovations in systems for direct payment to beneficiaries (the use of electronic cards, SMS transfers) with positive consequences for access to financial services. Impact on local economies 121. Evidence is growing about the effects of cash transfers on local economic activity.24 Cash transfers provide an immediate impact by raising the purchasing power of beneficiary households. As beneficiary households spend cash, impacts immediately spread outside beneficiary households to others inside and outside their communities. In Ghana for example beneficiaries of the LEAP program spend about 80 percent of the transfers inside the local economy, which in turn launches a new round of income increases in the community. In turn, periodic markets and purchases outside village will shift income effects to other communities. In the longer run, as programs are 24 From Protection to Production Project (FAO, 2013) and the transfer project (FAO, 2013) 36 scaled up, transfers will have direct and indirect (or general equilibrium) effects throughout the region of implementation. Cash transfers can therefore have large income multiplier effects, provided markets are functioning fluidly. If, on the other hand, production constraints limit supply response, it is conceivable that transfers may lead to higher prices and a lower multiplier effect. 122. Cash transfers can also lead to significant increases in productive investment. Cash transfer programs in Zambia led to increased investments in agricultural inputs, particularly for smaller households, as well as a shift to higher-value agricultural production (from cassava to maize or rice) and more selling at market (from 23 to 35 percent) and ownership of non-agricultural enterprise (from 22 to 39 percent). Cash transfers also led to increased livestock ownership in the Zambia, Kenya and Malawi programs. 123. Cash transfers also promote more productive allocations of labor within households. Cash transfers have been shown to favor reductions in adult agricultural wage labor and increases in time spent on household production, such as family farms (in Zambia, Malawi, Kenya, and Ghana). Evaluations also show an increase in non-agricultural labor market participation (Zambia and South Africa). Cash transfers also allowed elderly and physically disabled to “rest” (Zimbabwe). However, results remain mixed on child labor, with programs leading to reductions of child labor in Kenya and South Africa, reductions in child wage labor but increases in on-farm activity in Malawi, and no impact in Ghana or Zambia. Using cash transfers as an additional crisis-response mechanism 124. A national cash transfer program can be an important risk management instrument. Databases of beneficiaries and payment systems can be used to: (1) make temporary payments to offset price rises, such as in Mexico in 2008 (a single payment was added to the transfer, which was subject to health-related conditions for all the members of a household) or (2) raise the eligibility threshold to include households who are not necessarily in extreme poverty but are at risk of falling into it. This is the strategy that Brazil adopted during the 2008/2009 crisis. 125. In fact, by providing a regular, minimum, income in times of crisis, cash transfer programs can protect the income generation strategies of households and help them avoid resorting to harmful strategies (removing children from school, distress sales of livestock when prices are low, begging, or theft). This effect was demonstrated during the coffee crisis in Nicaragua in 2000 (Maluccio 2005); the beneficiaries of the Red de Protección Social program saw their incomes fall less than non-beneficiaries and faced less risk of having to pull their children out of school or stop going to health centers when sick. On the other hand, the “guarantee” kicks in if triggered by price inflation of life essentials that rapidly erodes the value of the transfer, or if the crisis lasts (Ethiopia experienced this phenomenon when food prices soared in 2008, combined with the drought in the Tigray region). Nicaragua also used cash transfers experimentally as a second line of response, after the humanitarian efforts, when it was hit by landslides, hurricanes, and drought. Cash transfers kept households on an even keel. Coupled with a jobs program to rebuild damaged community assets, cash transfers also kept basic services running. Catalytic effect of a cash transfer beneficiary database 126. A database created for a national transfer program can also serve as a database of potential beneficiaries for other social assistance programs. A well-constructed database of eligible households for a national cash transfer program can also provide useful information on potential beneficiaries for other social assistance programs and create a link between multiple social safety net programs with different needs and objectives. Such beneficiary databases can be a particularly useful 37 resource in case of climate shock of humanitarian crisis, when it can serve to identify beneficiaries for emergency interventions. 127. Coordinating and/or pooling beneficiary databases can bring several benefits. These benefits including (a) preventing or reducing undesirable duplication of benefits (e.g., people receiving benefits from more than one donor or program when people should not); (b) reducing duplication of administrative costs across programs by relying on a single database and (c) monitoring the evolution of living conditions of potential beneficiaries over time. The use of Cash Transfers as a compensation mechanism for subsidy reform 128. The reform of gas price subsidies in countries such as Brazil or the Dominican Republic can be pertinent examples for Mauritania. In Brazil, it consisted of replacing the subsidy with an unconditional bimonthly cash transfer (by Auxilio Gás) to offset higher butane prices, for poor families and then integrating the beneficiaries into the transfer programs. The cash transfer was relatively low, equivalent to the amount spent by the poorest quintile to buy gas. The advantage of this method was in not distorting the market and in building up a large base of beneficiaries to broaden support for the reform. After two years, the beneficiaries in urban areas were integrated into the program after verifying their socioeconomic status. Another possibility is to distribute food stamps or prepaid cards to encourage the use of public transport, as transportation is often a major expense item for the urban poor and a barrier to getting a job. The experience of the Dominican Republic is also pertinent. C. Combining the two efforts for a transformative reform 129. Setting up a Conditional Cash Transfer (CCT) program is an ideal opportunity to push reforms, in particular the adoption of transparent and effective targeting methods, the construction of a national registry, and the coordination of the institutions involved. Indeed, CCT programs have often been the basis for innovations in social program management, which were are then extended to the social system. For example, in most countries, the databases created as part of the national cash transfer program have formed the basis of a single national registry, which is then used by other programs for their own targeting purposes. Similarly, the rolling out of CCT programs has often been the motivating for the definition and adoption of targeting criteria and methodologies, which are then adopted (and adapted) by other programs. Lastly, when rolling out CCT programs, it takes considerable coordination effort to ensure that the services offered are actually available, and to ensure that the eligibility conditions can be verified. And this collaboration can be a springboard to more comprehensive coordination among the actors. 130. It is essential to build a strong institutional mechanism when setting up a CCT program, a targeting methodology, and a single registry of poor households. Given the coordination challenges, some countries have opted to set up a separate agency/structure for their CCT program. In such cases it usually reports directly to the prime minister or head of state (initially Colombia and Brazil). This has the advantage of giving the new program immediate visibility, but the new structure can have problems building its legitimacy vis-à-vis sectoral Government departments, such as education and health. Other countries have preferred an existing Government agency to implement the CCT program. The benefit of that is the ability to take advantage of the institution’s existing resources, although this option can restrict the ability to roll out the program in other sectoral departments. If the Government department supervising it is weak or has a vast mandate, it can exceed its capacity or the CCT program can be drowned. Other countries have split registration/targeting functions from the actual implementation of the CCT program, sometimes from the outset, sometimes after a while 38 (Dominican Republic, Colombia, Brazil), so that the single register is not associated with a single program (thus less accepted by other institutions) but as a national tool. 131. In all cases, strong political support at the highest levels is one of the keys to the success of registry systems, targeting methods, and CCT programs. As mentioned above, setting up these tools and programs takes massive coordination between the various social sector departments and central and external services. These efforts must also be accompanied by a paradigm shift: from universalism to targeting, from aid-in-kind to direct financial aid, from input management to results- based management. For these changes to happen, strong coordination mechanisms must be set up, supported by the necessary legal framework, and reform must be demonstrably supported at the highest political levels. 39 V. Conclusion 132. Mauritania’s emergency response, via the national Programme Emel, succeeded in averting serious crises, but does not match the stated national objective of reducing chronic poverty and reducing vulnerability to shocks. The core objectives of Mauritania’s emergency safety net interventions were to mitigate the impact of crises for vulnerable populations and keep food prices low, and analysis confirms that they largely succeeded in reaching these objectives. However, despite being designed as a short-term response, the Emel program remains in place years after the crisis. The Boutiques Emel were initially designed to for a duration of 5 months, but are, three years later, firmly implanted across the country and in the habits of the population. Costs tied to the Boutiques Emel program have increased sharply as the program has expanded, and the program’s adopti on by the population has made winding down the Boutiques difficult. This leaves Mauritania with mismatch between its programs, which are geared towards crisis-response, and its chronic poverty reduction objectives. 133. To contain fiscal costs while achieving its long-term goal of reducing extreme poverty to 3 percent by 2030, the Government will need to complement its emergency food-based programs by investing in comprehensive, long-term programs focused on responding to structural causes of poverty. Recent evaluations show that the current Programme Emel is popular and has effectively provided tangible benefits to many Mauritania’s poor. But few of these benefits have reached the poorest of the poor, who lack the resources to pay for the products sold, and the fiscal cost of the program is not sustainable in the long run. An effective poverty-reduction policy will require a transition away from the current focus solely on emergency programs, with the introduction of targeted long-term programs to support the chronically poor and help them invest in their human and productive capital to increase their income-generating capacity and build their resilience. It will also need to reduce untargeted programs or subsidies, which fail to protect the effectively reach the poor but have a significant fiscal cost. 134. Mauritania has adopted several strategic frameworks that provide a comprehensive vision for social protection and poverty reduction, and has some experience to build upon. Mauritania has recently adopted its National Social Protection Strategy (NSPS), and includes social safety net programs in the third Poverty Reduction Strategy Paper (PRSP III). Both these frameworks hinge on increased coordination and on long-term approaches to poverty reduction, implemented through targeted programs. Also, new targeted small-scale programs, such as cash transfers and food vouchers, have been developed and tested since 2011, and show considerable promise. For example, they have led to the development of targeting methodologies and beneficiary registries, as well as electronic payment systems for transfers. These experiences are valuable bases on which to develop effective national social safety net programs in Mauritania. 135. Overall, the elements are in place for transformational change. Bearing in mind the important efforts and results achieved, this report underscores the importance of four core actions. Implementing these recommendations would require a series of additional analytical work and operational testing, in particular to design an effective targeting methodology (selecting the combination of targeting mechanisms, estimating poverty levels in all localities, etc.), to better understand how food markets can be made more efficient and with fewer distortions, to design early warning systems that can be used to trigger the range of social protection interventions, to identify the reforms of universal programs – in particular subsidies – that can be programmed to reduce fiscal costs and finance targeted programs, and designing programs which can effectively increase resilience among the poorest and most vulnerable households. The four proposed recommendations are: 40 (1) Develop long-term interventions that address chronic poverty and build resilience against shocks. It is essential to complement existing emergency programs with long-term interventions that address chronic poverty and build resilience against shocks. The development of programs, benchmarks and instruments should follow the national vision laid out in the NSPS and PRSP III frameworks. In particular, an effort should be made to strengthen households’ capital – both the human capital of their members and in particular their children, and the productive assets that can improve their income-generating capacity – so as to increase their resilience and promote inter-generational social mobility. (2) Broaden the range of safety net instruments, including cash transfers. To better address chronic poverty and build resilience among the poorest, the Government should broaden the range of programs to include permanent safety net interventions, such as cash transfers. Long-term cash transfers that are well-targeted to households suffering from chronic poverty, and promote investments in the human capital of their children, can be particularly adapted to Mauritania’s situation. This would need to be complemented, more generally, with efforts to strengthen the supply of basic social services to these households. Furthermore, there are significant distortions in Mauritania’s food markets , which suggests large potential gains for the poorest if these can be corrected. (3) Replace universal programs with targeted ones for greater impact and reduced fiscal cost Universal subsidies tend to benefit the wealthy and have not contributed significantly to poverty reduction in Mauritania. At their current level, eliminating subsidies could save over 3 percent of GDP, far more than the cost of all existing programs targeted to the poor and vulnerable. The reallocation of a fraction of the resources currently dedicated to universal subsidies towards targeted programs would be a significant improvement in the efficiency and equity of the allocation of public resources, and lead to greater impact both in supporting domestic consumption and in reducing poverty. Improving the targeting of social programs, and improving their coordination, can build upon a national registry of poor households that adequately collects information on the poorest households and their characteristics. In the context of Mauritania, a combination of targeting methodologies can be designed to effectively identify those most in need across the territory. (4) Reinforce emergency response capacity with a strengthened early warning system and rapidly scalable programs Mauritania’s exposure to shocks places a high priority on an effective emergency response mechanism. The Government should consider improving its early-warning system to monitor a broader range of risks and risk factors, and developing contingency planning and financing for social protection response to crises. This would allow for a more specific and rapid response to crises, which can increase the impact of emergency resources on human welfare and households’ resilience. The national registry of poor households can also play a critical role in the rapid implementation of programs triggered by crises, including food-based emergency programs, by pre-identifying households most likely to be affected by specific shocks. Mauritania could also consider complementing its typical emergency response with mechanisms to temporarily scale- up long-term safety net programs, so as to rapidly channel resources to crisis-stricken households. 41 Annex 1: Evaluation Results of the Programme Emel In 2011, Mauritania was struck by a severe, nationwide drought. The Mauritanian Government responded by launching a nationwide program called Emel (Arabic for hope) to support the population through the drought. The Programme Emel has two components: (1) a ‘support to the population’ component, including emergency food distribution, the Boutiques Emel, and the restocking of the SAVS (cereal banks), and (2) a livestock support component, including subsidized distribution of livestock feed, animal health activities, and investment in pastoral water projects and credit for livestock farming. At the end of 2013, the World Bank undertook an evaluation of the first component (see World Bank, Evaluation des Filets de Securité Alimentaires du Programme EMEL, 2014). The results are summarized here. Activities for this first component represented 3.6 percent of the national budget and 1.1 percent of GDP in 2012. The population support component was designed to cover 92 percent of the cereal needs of poor and extreme-poor populations in 2012 and 63 percent in 2013 (or 41 percent and 28 percent respectively of national requirements). Therefore the cereals provided by the program were intended to cover in excess of the needs of extreme-poor populations in 2012 and approximately match them in 2013. In practice, the three activities of the population support component managed to cover an estimated 53 percent of the needs of the poor (85 percent of the needs of the extreme-poor population only) and 24 percent of national consumption in 2012. These figures are estimated to be similar for 2013—49 percent, 79 percent, and 22 percent respectively. The shops (or Boutiques) were the core mechanism of the Programme Emel’s food distribution component, with 83 percent of distributed cereals in 2012 and 92 percent in 2013. Free emergency distributions and the SAVS were a small part of the program (11 percent and 7 percent respectively of cereals actually distributed in 2012 and 2 percent and 6 percent respectively in 2013). This aspect of the Programme Emel, essentially through the Boutiques, had a significant, positive impact on food security of the population. The Boutiques enabled the population to achieve substantial savings on the purchase of basic food items and contributed to the stabilization of prices initiated in 2011 with the National Solidarity Program (Programme de Solidarité Nationale).25 In general, it is estimated that, where implemented, the Programme Emel’s food distribution component has had few inclusion errors. With the exception of a small number of inclusion errors (e.g., a 2012 wholesale operation in Nouakchott), it appears that, in general, where implemented, the program’s targeting of the poorest sections of the population may be considered relatively successful. This is due to a number of implementation choices as well as sociological factors: (a) setting up the shops mainly in poor areas; (b) rationing and quality26 of products; (c) personal costs (waiting lines and visible expression of social status tied to patronage of the Boutiques); (d) social pressure 25 The World Bank did not evaluate the livestock-support component, but a Government evaluation underscored that “the availability of livestock feed at affordable prices limited animal mortality and maintained female fertility. Growth estimates for 2012 are based on a livestock growth of -2.6 percent.” (Islamic Republic of Mauritania, April 2013). 26 For rice in particular, it was decided to use the shops for distribution to promote national production. Imported rice is said to be of higher quality and observation at shops visited did in fact reveal that the quality of local rice is still extremely uneven (there were significant impurities in some batches). 42 preventing the less needy from participating; (e) targeting by the communities themselves for the SAVS and free-food distribution. However, those most in need, the extreme poor, did not have the means to buy subsidized products at the Boutiques Emel, and were thus excluded from the largest component of the government emergency response. The extreme poor were only able to benefit fully from Emel’s free emergency distribution activity, which was small scale and incompletely implemented. Thus, households in extreme poverty were not able to benefit fully from the Boutiques program, which, in its current form, does not allow for systematic covering of food needs of the most vulnerable. On the other hand, the shops, and to a lesser extent the SAVS, allowed the vulnerable population (those not completely without resources) to cover their food needs better, directly via the increased purchasing power of their cash resources and indirectly via the strengthening of existing solidarity and redistribution mechanisms within communities. There is a noticeable urban bias in the conception and execution of the program, both in terms of the location and endowment of the Boutiques. Total population per shop is slightly lower in rural areas, but because there is a greater incidence of poverty in rural areas (59.4 percent compared to 20.8 percent in urban areas—15.6 percent in Nouakchott), the poor population per shop is respectively 2.5 and 4 times higher in rural areas than in urban areas and Nouakchott. Far from correcting this imbalance, the increased number of shops and the changes in distributed rations in 2013 further favored urban areas, this time outside of Nouakchott. The SAVS and free food distributions, implemented mostly in rural areas, accounted for only a small proportion of the total volume mobilized by the program and did not therefore correct this strong bias in favor of urban areas. The Boutiques have allowed households to achieve significant savings. The savings achieved by beneficiaries in comparison to market prices are estimated at about MRO21 billion (equivalent to US$70 million in September 2013) for each of the two years of implementation. Under the ideal hypothesis that only the poor populations benefited from the program, this figure corresponds to an average savings, compared to market prices, of approximately MRO5,000 per month per poor household outside of Nouakchott, both in 2012 and 2013. For poor households in Nouakchott, the average savings were MRO15,000 per month in 2012 and MRO12,000 per month in 2013. It appears that the program also contributed to the price stabilization that began in 2011 with the Programme de Solidarité Nationale (PSN). Study of the market price lists for wheat, sugar, and oil shows that prices for these items were relatively stabilized by 2011, after rising significantly in 2010, and that this continued after the Programme Emel was implemented. On the other hand, the SAVS did not generally play their price-stabilizing role for the most vulnerable because beneficiaries often misunderstood how they were to be used. In the majority of cases, the SAVS were seen as emergency interventions (and thus distributed among the community) and not as a potential tool to improve resilience among the most vulnerable within the community. Paradoxically, the Government’s large-scale food distribution operation via the Boutiques Emel appears to have been more cost-efficient than distribution by the private sector of the same food items both outside of and in Nouakchott. In other words, it would have cost the Government more to achieve the same result using the country’s private distribution network (for example with a system of vouchers usable in private shops) than through setting up a network of 1,200 shops itself. This result confirms the Government’s chosen strategy in the current market and implies that implementation of the shops by the CSA and the SONIMEX was relatively effective. Conversely however, this finding clearly indicates that local markets are not efficient (inefficient import and distribution of basic food 43 items and/or high margins by the private sector resulting in high and uncompetitive prices for households). In light of this finding, it would be desirable to make a more thorough study to examine this important question, since resolving some of the imperfections of the market in basic food items could have a very strong impact on poor households and the overall rural economy. Addressing market imperfections (by increasing competition, developing access infrastructure, etc.) could be an important mechanism in improving the population’s food security, as well as having far greater multiplier effects for the national economy than Government-subsidized distribution of food. It would also be desirable for this question to be clarified before considering a move away from the Boutiques Emel to other systems with potential to strengthen the food security of the most vulnerable by making greater use of the private sector (e.g., cash transfers or distribution of vouchers for use in private stores). Otherwise, part of the subsidies originally aimed at the most vulnerable in these new systems stand to be absorbed through the high margins practiced by private distribution networks. The lessons that can be learned from the two years of implementation indicate six courses of action. The following recommendations would help increase and sustain food security gains while moving toward systems that reduce private costs for beneficiaries (waiting lines, rationing, limited choice, etc.) and market inefficiencies, and support the shift from an emergency rationale toward instruments promoting increased resilience: (1) After establishing a national targeting system, develop a mechanism for cash transfers and/or vouchers targeting the poorest people so that they may benefit more from the existing system, from which they have so far been partially excluded due to a lack of cash resources; (2) Rebalance the Emel program toward rural areas by strengthening the shops in those areas and running a more ambitious SAVS operation for long-term impact (see point 5); (3) As the cash transfer and/or voucher system develop and the market for basic food items becomes more efficient, withdraw the Government from distribution (gradual closure of the Boutiques Emel, starting in the areas where the cash transfer system is established and the private market is deemed to be working well); (4) Conduct an evaluation study on the effectiveness of the market in basic food items and, if need be, promote better functioning of the market (increased competition, administrative and tax simplification, rural access, etc.); (5) Create a long-term strategy for the SAVS, which appear promising but are still largely underused, to increase resilience among the poorest, while strengthening cohesion among rural communities and relieving existing solidarity mechanisms; (6) Better define responsibilities in terms of food security, since these appear fragmented at the moment, and provide the responsible body for managing national strategy with greater capacity in monitoring and evaluation (M&E) so that, e.g., potential imbalances, like those observed in this study between rural areas and urban areas, are corrected instead of compounded as in 2013. 44 Annex 2: SSN Expenditure, Including Emel Livestock Component Figure 5: Social safety net expenditure, including Programme Emel’s livestock component 90 UM billion 80 70 Food transfers 60 EMEL Livestock support 50 Cash and near-cash transfers 40 Public works 30 Fee waivers 20 Fuel subsidies 10 - 2008 2009 2010 2011 2012 2013* * 2013: estimates Table 14: Total spending on SSN programs, including Emel livestock support (MRO million) 2008 2009 2010 2011 2012 2013 Type of program 1. Cash and near-cash transfers - - 101 266 2,250 3,900 2. Food transfers 7,416 3,330 3,411 19,182 51,255 37,159 2.1 Targeted subsidized food sales 4,702 304 112 8,141 23,665 14,384 2.2 Targeted food distributions 1,471 1,783 1,268 1,076 5,181 214 2.3 Nutrition 43 43 869 1,814 2,329 2,471 2.4 School feeding 1,200 1,200 1,162 1,162 1,220 1,220 2.5 EMEL livestock support - - 6,990 18,860 18,870 3. Universal fuel subsidies 611 12,227 11,414 21,260 30,700 15,000 4. Public works 361 351 973 1,096 1,600 1,250 5. Fee waivers - 408 931 740 923 789 Total social safety net programs Million MRO 8,388 16,316 16,829 42,544 86,728 58,098 Percent of GDP 1.0 2.1 1.7 3.5 7.0 4.3 Percent of Government expenditure 3.9 7.9 6.4 12.9 23.2 14.4 Total social safety net programs, excluding fuel subsidies Million MRO 8,388 16,316 16,829 42,544 86,728 58,098 Percent of GDP 1.0 2.1 1.7 3.5 7.0 4.3 Percent of Government expenditure 3.9 7.9 6.4 12.9 23.2 14.4 45 2010 2011 2012 (est.) 2013 Institution/Programme 2008 2009 Fin. Nat. Fin. Ext. Fin. Nat. Fin. Ext. Fin. Nat. Fin. Ext. (est.) TOTAL TOTAL TOTAL Exéc. Exéc. Exéc. Exéc. Prog Exéc. Prog Commissariat à la Sécurité Alimentaire 6,145 1,992 762 2,131 2,893 8,205 2,678 10,883 23,094 10,269 33,363 20,211 Solidarité 2011 & EMEL 2012 SAVS 4,702 304 112 112 141 141 1,583 1,583 Distribution Gratuite de vivres 1,443 1,688 448 468 916 707 707 5,012 5,012 Boutiques Solidarité 8,000 8,000 18,082 4,000 22,082 14,384 Programme nutrition Distribution rations alimentaires / CRENAM 830 830 377 377 480 480 480 Centre alimentation communautaire / CAC 410 410 606 606 447 Autres programmes Vivre contre travail 100 352 452 160 398 558 1,000 1,000 1,000 Micro-projets 368 368 308 308 350 350 Secours aux sinistrés 214 214 45 187 232 Cash Transfer + Food Vouchers 150 150 2,250 2,250 3,900 Ministère de la Santé 43 43 139 139 156 987 43 1,200 1,544 Transferts Monetaires 101 101 116 116 Programme nutrition 43 43 39 39 40 987 1,027 43 1,200 1,243 1,544 Ministère de l'éducation 1,200 1,200 595 567 1,162 595 567 1,162 20 1,200 1,220 1,220 46 Cantines scolaires 1,200 1,200 595 567 1,162 595 567 1,162 20 1,200 1,220 1,220 MASEF 309 794 1,182 1,182 1,007 1,007 200 1,041 1,241 1,153 Evacuer les indigents à l'étranger 408 291 291 374 374 354 354 500 Disponibiliser l'Hémodialyse 641 641 367 367 369 369 89 Programme appui aux Handicapés 24 24 32 32 40 40 70 Centre de Promotion de la Petite Enfance 18 95 80 80 90 90 99 99 109 Intégration sociale enfants en difficultés (CPISE) 10 34 34 15 15 30 30 35 Appuyer les AGR 281 291 113 113 130 130 150 150 150 Appui hopitaux pour indigents 150 150 150 Assistance sociale - 50 50 50 Min. Emploi & form. prof. 80 60 40 40 100 - 100 100 100 100 Programme HIMO 80 60 40 40 100 - 100 100 100 100 Annex 3: Expenditure Table by Project, 2008–2013 (millions MRO) Subventions Universelles 11,440 12,227 11,414 11,414 28,250 28,250 49,560 49,560 33,870 Hydrocarbures 3,500 3,500 9,000 9,000 1,000 Gaz butane 611 3,327 4,354 4,354 9,070 9,070 11,700 11,700 8,000 Electricité 8,900 7,060 7,060 8,690 8,690 10,000 10,000 6,000 (aliments) (SAVS, food distrib, boutiques) 14,516 - - 14,990 14,990 32,717 32,717 27,500 Food subsidies excl. SAVS, distrib & boutiques 10,829 - - - 6,990 6,990 18,860 18,860 18,870 TOTAL 19,217 16,316 14,132 2,698 16,829 38,312 4,232 42,544 23,437 50,622 12,670 86,728 58,098 Annex 4: Cost and Financing of Social Safety Net Interventions in Africa Spending on social safety net programs, excluding universal subsidies Spending Total universal spending Percent of Percent subsidies Years (percent Percent total financed by (percent of Country of GDP) of GDP Government development GDP) spending partners Benin 0.3 1.1 65 0.5 0.9 Ave. 2005-10 Botswana 1.5a 9.5 0 0 1.5a Ave 05/06-11/12 Burkina Faso 0.6 <1.0 80 0.7 1.3 Ave. 2005-09 Cameroon 0.2 1.5 77 1.4 1.6 Ave. 2008-10 Ethiopia 1.2b . 100 0 1.2b 2009 Kenya 0.8 1 71 0 0.8 2010 Lesotho 4.6 8 0 4.6 2010/11 Liberia 1.5 4.4 94 0 1.5 Ave. 2008-11 Madagascar 1.1 5 . 0 1.1 2010 Mali 0.5 . 60 0.1 0.5 Ave. 2006-09 Mauritania 1.3 4.6 38 1.35 2.9 Ave. 2008-13 Mauritius 4.4 9 . 0.8 5.2 2008/09 Mozambique 1.7 . 62 1.4 3.1 2010 Niger . 1.0-5.0 67 . . Ave. 2001-06 Rwanda 1.1 . . 0 1.1 2010/11 Sierra Leone 3.5 13.1 85 2.1 5.6 2011 South Africa 3.5 . . . . 2010 Swaziland 2.1 . . 0 2.1 2010/11 Tanzania 0.3 1 . 0 0.3 2011 Togo 0.5 1.8 75 0.8 1.3 Ave. 2008-10 Zambia 0.2 . 75 1.9 2.1 2010/11 Regional averages Average Africa 1.6 4.4 70 0.5 2.1 . Ave. LICs* 1.1 3.6 75.9 0.5 1.6 . Ave. MICs** 2.4 7 50.7 0.7 2.9 . Other regions Europe & Central Asia 1.8c - 1.8c Latest ‘08-‘10 Latin America and the Caribbean 1.1d - 1.1d 2010 Middle-East & North Africa 0.7 5.7 6.4e Latest *LICS: Benin, Burkina Faso, Ethiopia, Kenya, Liberia, Madagascar, Mali, Mozambique, Niger, Rwanda, Sierra Leone, Tanzania, Togo ** MICS: Botswana, Cameroon, Lesotho, Mauritania, Mauritius, South Africa, Swaziland, Zambia, Sources: (Monchuk, 2013) 47 Bibliography Awokuse, Titus O. (2006) “Assessing the Impact of Food Aid on Recipient Countries: A Survey.” Working paper 06-11 (FAO - ESA). Barrett, C. (2006) “Food Aid’s Intended and Unintended Consequences” Cornell University. Christaensen, L et al. (2013) « The Speed of Urbanization and Economic Development: A Comparison of Industrial Europe and Contemporary Africa » (World Bank, George Washington University). Ébauche. CSA (2011) « Programme Emel (document de programme et de budget) » CSA (2011) « Rapport d’evaluation du programme national de solidarité » CSA (2012) « Note sur le Programme Emel » Dayton-Johnson, J. and Hoddinott, J. (2004). “Examining the incentive effects of food aid on household behavior in rural Ethiopia.” World Bank Working Paper. Garcia & Moore (2012) “The Cash Dividend: The Rise of Cash Transfer Programs in Sub-Saharan Africa” (World Bank) IMF (2011) “Aide-mémoire sur la réforme des subventions aux produits énergétiques et alimentaires.” IMF (2011) Poverty Reduction Strategy Paper (Country Report No. 11/252) IMF (2013) Mission Statement of Mission to Mauritania (IMF website) IMF (2012) “4 consultations for Mauritania” (Country Report No. 12/246) Levinsohn, J., McMillan, M., (2005) “Does food aid Harm the poor? Household evidence from Ethiopia.” NBER Working Paper. MAED (2006) “Profil de la pauvreté en Mauritanie 2004” MAED (2009) “Profil de la pauvreté en Mauritanie 2008” Ministère des Affaires Economiques et du Développement (2012) “Stratégie Nationale de Protection Sociale” Monchuk, V. (2013) “Review of Safety Net Programs in Africa” (World Bank, Mimeo) OCHA (2013) “Estimation de la population par Moughataa, Mauritanie” ODI (2010) “MDG Progress Report.” République Islamique de Mauritanie (2010) “Rapport sur les progrès 2010 vers l’atteinte des objectifs du millénaire pour le développement (OMD) en Mauritanie.” République Islamique de Mauritanie (2013) “Rapport d’Evaluation du Programme Emel 2012.” UNDP (2013) “Human Development Report 2013” UNICEF (2011) “Annual Report on Mauritania 2010” 48 United Nations (2012) “Report on the MGOs 2012” Watson, C. (2009) “Étude de faisabilité pour un projet pilote filet de sécurité/ transfert social en Mauritanie” (Save the Children) Watson, C. and Fah (2011) “Étude sur la Protection Sociale en Mauritanie.” UNICEF. WFP (2011) “Étude de diagnostic des SAVS.” WFP (2011) “Mauritania Country Program 2012–2016 » WFP (2012) “Emergency Operations Program (EMOP) in Mauritania 2012” WFP (2012) “Mauritania Standard Project Report 2012” World Bank (2012) “Ciblage et protection sociale: note d’orientation stratégique pour le Royaume du Maroc.” World Bank (2012) “Outline of Political and Social Analysis of Mauritania for Assessing Fragility and Conflict Risk” (SDV contribution to Mauritania 2013–2015 Preparation of the Country Partnership Strategy (CPS). World Bank (2013) “Country Strategy Partnership for the Islamic Republic of Mauritania” (draft) World Bank (2013) “Note d’évaluation des filets de sécurité alimentaire du Programme Emel” World Bank (2014) “State of Social Safety Nets, 2014” 49