64631 S P D I S C U S S I O N PA P E R NO. 1117 Social Safety Nets in Fragile States: A Community-Based School Feeding Program in Togo Colin Andrews, Elena Galliano, Carolyn Turk and Giuseppe Zampaglione August 2011 Social Safety Nets in Fragile States: A Community-Based School Feeding Program in Togo Colin Andrews, Elena Galliano, Carolyn Turk, and Giuseppe Zampaglione August 2011 i Abstract This paper reviews a small community-based school feeding program launched in Togo in response to the 2007/08 food price crisis. The discussion focuses on the operational and policy lessons emerging from the program, to better understand opportunities for scale up and sustainability in the future. A focus of the discussion is how to build safety nets in fragile states and in situations where there is weak and fragmented government capacity to deliver services to disadvantaged and vulnerable communities. In this context school feeding is explored as an entry point through the use of informal mechanisms based on the commitment of communities and civil society. The analysis is premised on quantitative and qualitative analysis carried out at program sites. The discussion identifies the operational challenges and opportunities in customizing school feeding within Togo with an emphasis on targeting, cost effectiveness, procurement and institutional aspects. Evidence on the economic and social benefits of the program is also presented, focusing on dietary impacts, as well as household and local community effects. The objective of the discussion is to share lessons learned from evaluation findings so that they can be useful for implementing similar programs in the future in Togo itself or in other countries. Findings from the analysis highlight the possibilities of implementing school feeding in a low capacity setting and the scope for using the program as a springboard towards a broader and more comprehensive social safety net. JEL Classification: H53 — Social Security and Public Pensions; I38 — Government Policy; Provision and Effects of Welfare Programs. Keywords: School feeding, community driven development, informal approaches, dietary diversity. i Social Safety Nets in Fragile States: A Community-Based School Feeding Program in Togo * Human Development Network, The World Bank; Email: candrews@worldbank.org, elena.galliano@gmail.com, cturk@worldbank.org, Gzampaglione@worldbank.org Address for Correspondence: The World Bank, 1818 H Street NW, Washington DC 20433, USA ii Acknowledgements In preparing the paper we have received excellent comments and feedback from Andy Chi Tembon, John Elder and Alexandra Marc. We are especially thankful to Elena Celada for helpful support in all aspects of the project and background evaluations as well as to Herve Assah and Sylvie Nenonene for their review of this work. We are grateful to many counterparts who have provided feedback and ideas and contributed to various background materials for the paper, including Bouty Yawavi Agboka, Technical Secretariat of the Community Development Project, and Egbenovi Agbo, Yaovi Ameyapoh, Kokou Amouzouvi. This work benefited from support from the Trust Fund for Environmentally & Socially Sustainable Development (TFESSD) made available by the governments of Finland and Norway. The authors alone are responsible for the contents of this final version. Moreover, the opinions expressed here are those of the authors and do not necessarily reflect those of the World Bank, its executive directors or the countries they represent. iii Contents Abstract ............................................................................................................................................ i  Acknowledgements ........................................................................................................................ iii  Acronyms ....................................................................................................................................... vi  I.  Introduction ............................................................................................................................... 1  II.  Purpose of the Paper and Methodology .................................................................................... 3  III.  Context ...................................................................................................................................... 6  3.1  General Background .......................................................................................................... 6  3.2  Fragile State: Poverty and Exogenous Factors .................................................................. 7  3.3  Response to the Food Crisis .............................................................................................. 9  3.4  Social Protection: Linkages to the Broader Agenda ........................................................ 10  IV. Main Operational and Implementation Issues ........................................................................ 13  4.1  Basic Design Elements .................................................................................................... 13  4.2  Institutional Arrangements .............................................................................................. 14  4.3  Program Funding ............................................................................................................. 18  4.4  Monitoring ....................................................................................................................... 18  4.5  Targeting .......................................................................................................................... 19  4.6  Procurement of Food ....................................................................................................... 21  4.7  Costs ................................................................................................................................ 22  4.8  Benefits to Children ......................................................................................................... 26  4.9  Economic and Social Benefits to Households and Communities .................................... 30  V.  The Community-Based School Feeding Program as a Safety Net: A Model for Fragile, Food Insecure States ........................................................................................................................ 41  5.1  Genesis and Rationale...................................................................................................... 41  5.2  Is this Safety Net Replicable and Scalable ...................................................................... 43  VI. Conclusions and Recommendations ....................................................................................... 49  Annex 1.  Yearly Costs of School Feeding Program in TOGO (US$) ...................................... 55  Annex 2.  Model for Standardization of School Feeding Costs ................................................ 57  VII. References ............................................................................................................................. 59  iv List of boxes Box 1: General Objectives of School Feeding Programs ............................................................ 12 Box 2: Preparation and Distribution of Meals ............................................................................. 16 Box 3: Local Procurement ........................................................................................................... 22 Box 4: Focus Groups about Savings ............................................................................................ 33 List of graphs Graph 1: Changes in School Attendance (%) ................................................................................ 4  Graph 2. Cost Comparison ............................................................................................................ 41  List of tables Table 1: Change in Attendance between School Years 2008-2009 and 2009-2010.................... 12  Table 2: National Statistics on Schools and School Enrolment (2006–2007) ............................. 20  Table 3: Comparison of Prices with Other School Feeding Programs in Africa ......................... 25  Table 4: Recommended Nutritive Value of Rations for Primary Schools................................... 27  Table 5: Example of Weekly Menu in Two Regions, Savanes and Maritime ............................. 29  Table 6: Summary of Nutritional Characteristics of Four Types of Meals ................................. 29  Table 7: Injection of Funds into Communities, Year 1 (2007–2008) ........................................... 32  Table 8: Number of Femmes-Mamans in Sample Villages ......................................................... 35  Table 9: Reinvestment of Profit in Sample Regions: Maritime and Savanes .............................. 36  Table 10: Local Production and Imported Food Items ................................................................ 36  Table 11: Resources Injected in the Communities, Years 1 and 2 .............................................. 38  Table 12: Areas of Reinvestment of Income or Savings from School Meals Program ............... 40  v Acronyms AFSSA Agence française de sécurité sanitaire des aliments (French Agency for Food, Environmental and Occupational Health Safety) AGAIB Agences d’appui aux initiatives de base (Independent and Nonprofit Regional Agencies) CBSF Community-Based School Feeding CDD Community Driven Development EDIL Local Initiative School EFMFTI Education For All Fast Track Initiative EPPR Emergency Program for Poverty Reduction FPCR TF Food Price Crisis Response Trust Fund GFCRP Global Food Crisis Response Program LICUS Low Income Countries Under Stress M&E Monitoring and Evaluation MICS Multiple Indicator Country Survey PRSP Poverty Reduction Strategy Paper PTAs Parent-Teacher Associations QUIBB Questionnaire with Base Indicators SMART Standardized Monitoring and Assessment of Relief and Transitions TS Technical Secretariat (Secretariat technique) WFP World Food Program WHO World Health Organization vi I. Introduction 1. This paper reviews a small community-based school feeding program, as a sub-component of a community development project, and associates its strengths and weaknesses to the larger agenda of social safety nets and social protection in the context of a food insecure fragile country. It is premised on the quantitative and qualitative analysis carried out at program sites, the data of which indicate: (i) in the Togolese context of food insecurity and weak government, an informal approach to safety nets is providing good results in terms of targeting, nutrition to children, and benefits to households; and (ii) this type of safety net can be scaled and replicated in other fragile and food insecure areas (and states) that feature, like Togo, minimum levels of food production and cohesive communities in the wider target area complemented by informal self help and mobilization mechanisms. 2. The paper makes the case that, building safety nets on existing informal mechanisms based on the commitment of communities and civil society can be an adequate solution in fragile states and in situations where there is weak and fragmented government capacity to deliver services to disadvantaged and vulnerable communities, a limited public and formal sector presence in many of the most disadvantaged regions, and often an uncoordinated presence of donors. In particular, the Togo Community-Based School Feeding program provides tangible lessons learned in rapidly responding to on the ground needs and simultaneously shifting the actors toward an inclusive national social protection policy. 3. The paper first examines how one particular approach to helping households build or consolidate their livelihoods in a fragile state environment is working well in Togo by strengthening households’ resilience and minimizing the impact of the food crisis. The community-based school feeding program is indeed focusing on strengthening livelihoods support by helping households access necessary resources. Livelihood is defined as “the set of assets and resources that a person needs to maintain a healthy and productive life.� Livelihoods support and protection may involve not only income-generating activities but also facilitate access to education, health care, market support, and even food aid as part of 1 an integrated strategy to improve people’s basic economic security.1 The Togo school feeding program was set up to respond to the food security challenges affecting most of the rural regions, by providing one nutritious lunch each day to primary school children of selected communities while making additional resources (approximately US$8-9 monthly) available to households under stress and allowing a number of women per community selected to cook and prepare meals to make a regular profit. 4. Second, it looks at how this safety net approach has been developed building on informal existing mechanisms (community members offering a service of meal preparation and delivery to primary school children, local procurement of food, local NGOs, and parent-teacher associations monitoring school operations) in the absence of strong formal stakeholders; how it provides a solid base for local ownership and empowerment; and how, as a test ground for mainstreaming the approach, partnerships are cultivated between formal and informal stakeholders. 1 U.S. Agency for International Development. Speaker’s Corner: Protecting Livelihoods in Conflict and Fragile States. An online forum January 22-24, 2008 hosted and moderated by Dr. Laura Hammond, Washington DC: USAID. http://www.povertyfrontiers.org/ev_en.php?ID=2105_201&ID2=DO_TOPIC 2 II. Purpose of the Paper and Methodology 5. The objective of the paper is to outline operational lessons from the implementation of the Togo Community-Based School Feeding Program. Launched in 2008 as a response to the food crisis, the program was implemented in primary schools located in food-deprived rural areas, benefitting 84 schools and 16,800 children in the first year and 92 schools and 21,300 children in the second. It had tangible multiplier effects on household savings, direct cash inflows into the local economy, and reinvestment of earnings into the local economy. This section provides information and evidence to support scaling up and mainstreaming the program. 6. This paper draws most of its information and evidence from the background evaluation reports and specific findings of an evaluation mission. Both have provided important information and suggestions helping to form an understanding of benefits, advantages, limits, and challenges of the program while tracking progress of the current operation and further-addressing implementation gaps and institutional challenges. 7. To this end, the paper takes stock of all previous assessments and draws lessons and evidence in particular from three studies: (i) a quantitative evaluation through a survey of 1,050 households in 35 villages among the five regions carried out using a multi stage cluster sample regionally stratified and aimed at assessing the economic and educational benefits of the projects, as well as the institutional setup of the program; (ii) an assessment at the national level of the strategic and administrative context for social protection in Togo; and (iii) a qualitative assessment of the perceptions of program beneficiaries. The qualitative assessment was carried out mainly through focus groups with female household members, femmes-mamans, students (girls and boys), and groups of community members (men and women) but also through semi-structured interviews with small traders. The sample includes six villages in two regions (Savanes and Maritime) that were selected from the sample of villages of the quantitative evaluation. 8. Togo is in the process of setting the agenda for its social protection activities. When the school feeding program was designed, the country was just emerging from years of isolation and little donor presence. The paper brings together the main elements of the 3 background work and by analyzing the main findings and the evidence gathered, makes the case that in a country like Togo, defined by low governance levels and many years of donors absence, tapping into local resources and informal coping mechanisms by supporting pre- existing community-based mechanisms while entering in partnership with the formal government and structures of education can become an effective way to protect livelihoods (Graph 1). Graph 1: Changes in School Attendance (%) Source: Authors (2011). 9. Therefore the paper responds to the following questions: 10. Is the school feeding program successful in achieving its main objectives of alleviating hunger among primary school children in beneficiary communities while making additional resources available to households during times of food insecurity and economic stress? 11. What are the other elements of success and overall benefits of this program? 12. What are the behavioral changes in the household as a result of the school feeding programs? 13. What are the main operational challenges as perceived by the various stakeholders and implementers (e.g., targeting, procurement, monitoring of program, type of school feeding implementation scheme, etc.)? 14. What are the characteristics of the institutional and operational setup for this type of multi sector program? 4 15. Furthermore, the program is implemented through a community-based approach and meals are prepared and served by women of the village. With this scheme, benefits are intended not only for the direct beneficiaries of meals but for the households and the community as a whole. The paper is therefore an opportunity to understand the specific advantages and the range of potential benefits to this approach as opposed to other traditional, centralized school feeding programs. 16. Finally, the paper and its conclusions illustrate (i) how in the context of fragility, safety net mechanisms can be successfully implemented tapping into local resources and leveraging informal coping mechanisms; (ii) the appropriateness of school feeding as a safety net mechanism in the context of Togo’s food insecure areas and the main advantages of the community-based approach; and (iii) the issues relevant to scaling up the program. 5 III. Context 17. This section situates Togo in a geographic, socio-economic, fiscal, and political context; provides a synopsis of the role poverty and exogenous factors on this fragile state; summarizes the state’s response to the food crisis; and links the response to the broader social protection agenda. 3.1 General Background 18. Togo is a small country located in West Africa. The length of the country spans six distinct geographic regions and five administrative regions — Savanes, Kara, Centrale, Plateaux, and Maritime, north to south respectively. Agriculture represents nearly 60 percent of all livelihoods, and widespread slash-and-burn agricultural practices have been attributed to deforestation. Unlike many other countries in West Africa, Togo is largely self sufficient in terms of food.2 19. The country is composed of 37 tribes, of which the Ewe, Mina, and Kabre are the most important, and five main languages, French (official, the language of commerce), Ewe and Mina (the two major African languages in the south), and Kabye/Kabiye and Dagomba (the two major African languages in the north). Three religions are practiced, Muslim (20 percent), Christian (29 percent), and indigenous beliefs (51 percent). The latter affects the education rate of girls.3 20. The median age of the population is 19.3 years old, with 39.9 percent between the ages of 0 and 14 years old. Gross National Income per capita (in current US$) using the Atlas method is US$440. Malnourishment touches approximately 30 percent of the 2 International Food Policy Institute. 2004. Agricultural Science and Technology Indicators. ASTI Country Brief No. 16. Washington DC: IFPRI. http://www.ifpri.org/sites/default/files/publications/Togo_CB16.pdf 3 Girls are known to enter Vodum Convents at the age of four and complete scant education. “UNICEF Executive Director’s Togo Visit Focuses on Education.� Press Release of September 2008, United Nations, New York. 6 population while the risk factor of infectious disease is high.4 Between 1980 and 2010 Togo's HDI rose by 0.7% annually from 0.347 to 0.428 today, which gives the country a rank of 139 out of 169 countries with comparable data. The HDI of Sub-Saharan Africa as a region increased from 0.293 in 1980 to 0.389 today, placing Togo above the regional average.5 21. The total net primary school enrolment rate was 93.6 percent in 2009. Among all grades of primary education in 2009, the total number of repeaters was 22.9 percent and the total primary completion rate was 61.4 percent. The most recent figures for the total primary education dropout was 55.5 percent in 2006.6 The number of primary age school children out of school in 2009 was 65,410 children, of which 9,808 males and 55,602 females. 22. Real GDP growth is forecast to accelerate to 3.6 percent in 2011 and 3.9 percent in 2012, backed by a rising level of foreign investment and assistance, investment in infrastructure, and the government's program of structural and financial reform.7 23. After decades of single-party rule, internal civil strife, and a deeply impoverished society, elections were held in 2005 and the president, Faure Gnassingbé, was re-elected for a second five-year term in March 2010. Local elections will be held in 2011 and a legislative elections in 2012. 3.2 Fragile State: Poverty and Exogenous Factors 24. A fragile state is a low income country characterized by weak state capacity and/or weak state legitimacy, leaving citizens vulnerable to a whole range of shocks.8 For years the country lacked the capacity to devise the policies and build the institutions that sustain a comprehensive and inclusive regulatory framework. Togo is included among the list of fragile states in Africa. 4 FAO http://www.fao.org/economic/ess/food-security-statistics/en/ 5 http://hdrstats.undp.org/en/countries/profiles/TGO.html. 6 http://web.worldbank.org/WBSITE/EXTERNAL/TOPICS/EXTEDUCATION/EXTDATASTATISTICS/EXT EDSTATS/0,,contentMDK:22614780~menuPK:7196605~pagePK:64168445~piPK:64168309~theSitePK:3232 764,00.html?TGO,206 7 Economist Intelligence Unit. 2011. http://country.eiu.com/Togo. 8 Wikipedia 7 25. The main factors that led to this deterioration of trust between the citizens of Togo and their leaders, and a drastic reduction in the number of donors, were an authoritarian and undemocratic regime, political instability, poor governance, weak institutional capacity, and lack of transparency in public resource investments. This protracted socio-political impasse severely affected the economy of Togo and hampered the capacity of the government to deliver basic social services to the majority of its population. 26. In fact, over 60 percent of Togo’s 6 million citizens now live below the poverty line, and poverty is strongly correlated with under nutrition to the extent that 64.2 percent of the poor community is undernourished. Rural areas have a much high incidence of poverty than do urban, with three households out of four being poor as opposed to two out of five, respectively. Almost half of the rural population remains food insecure, while 37 percent is at risk.9 27. Recent exogenous factors have only deepened Togo’s impoverishment. Rains and floods devastated Togo in 2007, and in 2008 the food and fuel price crisis affected Togo more than most countries of Sub-Saharan Africa. The combined effect of higher food prices and loss of agricultural production due to the floods left 13 percent of households in the Savanes (2008) severely food insecure and threatened the livelihoods of more than a third of the households in the Savanes, Kara, and Plateaux (300,000 people). Informal coping strategies were a last resort; adults reduced their daily food intake in favor of children, the nutritional quality of the main meal was downgraded, seeds stocked for the next season were eaten, and livestock sold.10 28. The recently released World Bank food price index is 36 percent above its level of 2010 and close to its peak in 2008. Severe weather events, export restrictions, the increasing use for biofuel production, low global stocks, and surging fuel prices (crude oil increased 21 percent in the first quarter of 2011as a result of unrest in the Middle East and North Africa) are attributed to the spikes. For example, maize makes up 65 percent of total 9 World Bank Report No: 45237-TG. Togo. Food price crisis response trust fund Project Paper on a proposed additional financing grant under the global food crisis response program, Sept. 30, 2008. 10 World Food Program. Togo - Enquête rapide sur la sécurité alimentaire des ménages, April 2010, Rome. 8 staple production in Togo. In 2007, it alone rose by 42 percent. In 2010, it marked one of the key increases in the food price index at 74 percent over the prior year.11 The near-future portents for Togo are no less daunting. 29. Children are among the major net losers of this scenario. According to the most recent Multi Indicator Cluster Survey (2006), one out of four children in Togo is undernourished. Approximately 50,000 children under five suffer from malnutrition, with the respective rate of severe malnutrition exceeding the critical level set by the World Health Organization.12 Similar percentages are found among children attending the first two years of primary school. 30. The two northern-most regions, Savanes and Kara, were reported to have acute malnutrition rates of 32 and 24 percent respectively,13 and the Government of Togo endorsed these findings. Inputting features were: failure to feed children appropriate foods in quantity and quality; lack of access to essential health services, water, and sanitation; and increasing levels of poverty and vulnerability. 3.3 Response to the Food Crisis 31. In December 2010, the Government of Togo adopted a fuel pricing mechanism to relieve pressure on the budget by linking domestic prices to world prices, albeit with a built-in stabilizer to smooth fluctuations and cushion the effect of price rises on consumers.14 32. In April 2010, Togo was one of eight countries to be selected to participate in the first round of the multi donor Global Agriculture and Food Security Program (GAFSP). 11 World Bank. 2010. High and Volatile Food Prices Continue to Threaten the World’s Poor Press Release of April 14, 2011,Washington DC. 12 http://www.unicef.org/wcaro/WCARO Togo Factsht Malnutrition.pdf 13 The Government of Togo endorsed the study. UNICEF 2007. Revised Country Program Document for Togo (2008-2012). New York: United Nations. http://www.unicef.org/about/execboard/files/07-PL43-Togo(1).pdf 14 Economist Intelligence Unit. 2011. http://country.eiu.com/Togo. 9 The GAFSP was set up by the World Bank at the request of the G20 to support country-led agriculture and food security plans and to help promote investments in smallholder farmers.15 33. In 2009, the government adopted a full Poverty Reduction Strategy Paper covering the period 2009–2011. Its main pillars are: (i) improving economic governance and transparency; (ii) promoting economic recovery and sustainable development; and (iii) addressing urgent social needs. Social protection is one of the priority areas of intervention for the achievement of human capital under Pillar 316 and social safety nets one of the privileged channels of resources both in terms of risk management as well as mechanisms of redistribution of resources among the most vulnerable. 34. The government has also been preparing a full strategy for the agricultural sector, centered on the need to produce sufficient food for domestic consumption, and looking at the entire production, transformation, and marketing cycle so as to respond to the food price crisis in a systemic way. 35. More specifically on the education side, a Country Status Report on Education was prepared by the government and has received support for the Education For All Fast Track Initiative (EFMFTI).17 It is considered the main vehicle for international support to primary education in Togo. 3.4 Social Protection: Linkages to the Broader Agenda 36. In Togo, the agenda of social protection — aside from the most traditional areas pertaining to labor, pensions, and health insurance — is new. Social safety nets are covered almost entirely by external resources; they are channeled to rehabilitate vulnerable 15 World Bank. “High and Volatile Food Prices Continue to Threaten the World’s Poor.� Press Release of April 14, 2011,Washington DC. 16 The “Development of Human Capital� aims to: (i) promote the education and training system; (ii) develop the health system and health services; (iii) improve the nutritional level of the general public; (iv) improve access to safe water and to sanitation infrastructure; and (v) promote gender equity and social protection. The intent being to provide an effective response to chronic poverty and to facilitate access to basic social needs for the chronically poor. 17 In 2009, the Education for All-Fast Track Initiative Secretariat and Partnership (UNESCO, UNICEF, WFP, WHO, and the World Bank) are working with all these partners (FRESH, Deworm the World) to respond to country demand for quality, school-based deworming programs led by the education sector, http://www.educationfast-track.org. (Bundy et al. 2009) 10 communities and to start a recovery process, one where pro-poor development strategies act in complement. 37. The immediacy of human and social development needs has brought multi and bilateral donors to rely on international actors, local NGOs, or other informal community-based mechanisms for their capacity to deliver promptly. In the case of the school feeding programs, the main channels of services delivery are built on local community mechanisms and program implementation is the responsibility of local NGOs. 38. The Togo approach leverages the strong formal experience of Parent-Teachers Associations (PTAs) and regional NGOs in managing and supporting rural schools. It also builds on the existing informal system of the femmes-mamans (women from the community who prepare/sell the snacks and meals). The training provided to ensure compliance with standards in terms of quantities, nutritional contents, and safety has benefits that go well beyond the school walls. 39. “Safety nets allow households to take up investment opportunities that they would otherwise miss — both with regard to the human capital of their children and the livelihoods of household earners. Specifically, safety net programs can contribute to capital accumulation among the poor by preventing the negative outcomes of malnutrition and underinvestment in education, and by enabling investment in productive assets.� (Grosh et al. 2008). 40. School feeding programs as one type of in-kind food transfers provide additional resources to households by making food available when needed. The most likely beneficiaries include poor children and their families that do not have sufficient income to purchase enough of the right foods and are more likely to achieve a better diet if they can receive specific foods or purchase them at a subsidized cost (Box 1). 11 Box 1: General Objectives of School Feeding Programs  Meet the immediate food needs of children.  Alleviate short-term hunger and improve learning capacity.  Increase access to education (i.e., enrolment, attendance, retention, and completion).  Reduce gender and social inequalities. 41. The Government of Togo believes school feeding programs to be, in the short term, important safety nets mechanisms to help reduce vulnerability especially in households further impoverished by the food price crisis and flooding (Table 1). In the long run, school feeding programs can constitute a key element of the country’s social protection agenda and policy support to universal education. In fact, among the long-term measures of the EFMFTI is the establishment of sustainable school feeding programs. Table 1: Change in Attendance between School Years 2008-2009 and 2009-2010 Regions Boys Girls Total Maritime Beneficiary village 2.3% 11.9% 6.7% Non beneficiary village 2.4% 3.6% 3.0% Plateaux Beneficiary village 13.4% 26.6% 19.7% Non beneficiary village -13.5% -2.8% -15,8% Central Beneficiary village -14.6% -15.8% -15.2% Non beneficiary village -22.7% -20.7% -21.8% Kara Beneficiary village 10.9% 15.1% 12.6% Non beneficiary village 4.7% 10.6% 7.5% Savanes Beneficiary village 23.6% 34.2% 27.9% Non beneficiary village 37.7% 21.9% 30.3% Total Beneficiary village 6.8% 12.6% 9.4% Non beneficiary village -9.1% -5.5% -7.5% Source: Agbo E., 2010b in Evaluation de l’opération de fourniture de répas aux écoliers des zones vulnerables du Togo, Analyse qualitative. 12 IV. Main Operational and Implementation Issues 42. This section of the paper describes the operational characteristics and the implementation arrangements of the program and analyzes the main findings and conclusions of the program field evaluation, specifically in the areas of targeting, sources of food, the nutritional, educational, and social benefits to children, and the socio-economic benefits to households and other community members. 4.1 Basic Design Elements 43. The Togo Community-Based School Feeding program financed one diversified meal daily for primary school children across the five regions of the country over a two-year period. The primary motivation for this intervention is to improve the nutritional well-being of children during times of food insecurity; improve school enrolment, attendance, and performance; bring about behavioral change with regard to hand-washing and hygiene; and possibly improve community cohesion and capacity to organize. An additional benefit includes a transfer of resources to households under stress. 44. A US$2 million fund supported the preparation and distribution of daily lunches to 16,800 primary school students in 84 schools in the first year and 21,300 students in 92 schools in the second year in rural and poorest areas. The program began delivering meals in the 2008–2009 school year. 45. The operational and institutional setup made available under the Emergency Program for Poverty Reduction (the first concrete World Bank response to the crisis) was absorbed. The Technical Secretariat, established to support the implementation of the Bank- financed community development projects, was equipped to manage additional funds and activities. 46. The program is a component of the Community-Driven Development Project initiated in Togo to provide poor and vulnerable communities with improved basic socio- economic infrastructure and income-generating activities by financing at least 350 sub- projects identified and implemented directly by communities. The school feeding component 13 was financed by the Food Price Crisis Response Trust Fund, which was endorsed by the Board in May 2008 as part of the Global Food Crisis Response Program.18 47. Prior to this Bank-funded school feeding program, no formal school meal program or canteen existed in Togo. Children were usually given a small allowance by their parents (about 25–50 CFA) to buy snacks and hot meals prepared the village women (femmes-mamans) authorized by the school to sell food.19 This system is relatively efficient but became increasingly expensive because of the food price crisis. By the end of the 2007– 2008 school year, the cost of a basic meal (i.e., 120 grams of rice with fish sauce) had increased by almost 50 percent.20 4.2 Institutional Arrangements 48. The program operates through three main groups of stakeholders: the national government, non governmental organizations, and local communities. The communities are active in providing food-related services (in terms of the supply, purchase, and preparation of the food as well as serving the meals) but also in responding to the program by sending children to school, learning new skills, and making decisions at the village level. Regional and local NGOs, including the PTAs, are responsible for program implementation and daily on-site monitoring (Box 2). The national government, via the establishment of a Steering Committee, is providing overall guidance to the program. 49. Of these three groups comes six main actors: 50. Ministry of Grassroots Development is responsible for the overall program. At the central level, a Steering Committee chaired by the Ministry of Primary Education oversees program implementation and provides overall strategic guidance. The program is then coordinated by the Technical Secretariat, which. 18 The objectives of the GFCRP are to: (i) reduce the negative impact of high and volatile food prices on the lives of the poor in a timely way; (ii) support governments in the design of sustainable policies that mitigate the adverse impacts of high and more volatile food prices on poverty while minimizing the creation of long-term market distortions; and (iii) support broad-based growth in productivity and market participation in agriculture to ensure an adequate and sustainable food supply response. 19 The femmes-mamans are required to have an updated cahier de santé, which is a menu of vaccines and exams that is administered by regional Departments of Health. 20 World Bank. 2010b. Togo Community-Driven Development Project, Paper No. 54513-TG. Washington, DC: World Bank. 14 51. Technical Secretariat (TS) liaises with the Steering Committee, the Ministry of Grassroots Development and sectoral ministries (Health and Education) to coordinate the program. A working relationship has been built with the Ministry of Health to run deworming campaigns as well as other activities in relation to nutrition and health prevention and protection. The TS retains overall implementation responsibility of all sub-components: overall management and program funding; and, selecting, recruiting, and transferring funds to regional non governmental organizations. 52. Regional non governmental organizations (NGOs) organize and implement school feeding within their geographic purview. Their responsibilities include: (i) signing contracts with and transferring operating funds to the local Parent–Teacher Associations, and ensuring on-site implementation proceeds as designed; (ii) verifying the number of students daily registered in school; (iii) monitoring the quality and quantity of meals; (iv) reporting on financial issues to the TS; (v) providing information to teachers and parents about the program as well as parallel health programs like deworming; and (vi) ensuring compliance with food handling norms. 53. Parent–Teacher Associations (PTAs), because of the protracted institutional and governance crisis, played a major role mostly on a voluntary basis in managing rural schools by organizing, financing, and staffing them. Strong and well established, PTAs are networked across regions and at the national level. For this school feeding program, they make sure that the program at the school level is adequately and transparently managed and implemented. Within the program, PTAs are responsible for daily implementation in each school and for monitoring meal quality and quantity by: (i) making sure that quality of meals as well as all distribution procedures are carried out according to standards, and (ii) providing weekly payments to the femmes-mamans. 54. Femmes-mamans directly receive the cost of an individual meal (US$0.31 per day) to pay for food purchases, transportation (from the market to the school), and their labor, with daily income estimated at approximately US$1-3 per day. In general, there are five to eight femmes-mamans per school, and each cooks for approximately 35–40 children. The femmes-mamans may decide to work individually or in teams, depending on the total number of students and the agreements within community members, and they are selected 15 among those who have a regular and updated health record released by the regional health department. Their health status is checked every six months, especially for transmissible and infectious diseases. Their standings in the community, earning money consistently throughout the school year, have not gone unnoticed. 55. School administrative bodies are very much involved in daily program procedures and are responsible for counting the number of children who are present daily in school and informing the femmes-mamans of the number. Box 2: Preparation and Distribution of Meals Meals are prepared by selected village women or femmes-mamans. A committee of femmes- mamans, school authorities, PTA, regional NGO, and local representatives of the Ministry of Education meets every two weeks to determine the meal plan. The factors taken into consideration are: (i) local availability of food products, (ii) overall caloric content; (iii) local tastes, and (iv) need for a diversified and balanced diet. However, once the menu is determined, all schools in a given region must adhere to it. The femmes-mamans receive training on compliance with basic food safety and hygiene norms as well as on basic accounting and business management. In providing these services, the program vis-à-vis the NGOs serves to deepen human capital and strengthen communities. 56. The program builds on the operational and institutional experience of the Community-Driven Development program and the Emergency Program for Poverty Reduction. For the latter, the new government instituted Boards composed of civil society, non profit organizations, and government representatives to oversee newly created independent, non profit, regional agencies (AGAIBs). This construct helped to build consensus across sectors at the national policy level that school meal programs are (at least in the long run) a measure to assist vulnerable households in food insecure areas, and lent to the school feeding program structure by rapidly delivering services with a minimum of new skills and capacity transfers. 57. To this end, the program has leveraged the thorough experience of Parent– Teacher Associations (PTAs) in managing and supporting rural schools and transformed the 16 extant network of femmes-mamans into the main service providers; program delivery occurred without major glitches. This represents an important transfer of capacity resulting in empowerment of local structures and community members. 58. Implementation at the local level solidified working relationships between PTAs, primary school administrators, community members, and grassroots organizations. In return, these actors benefited from continuous capacity building and institutional strengthening in such areas as efficient delivery, good hygiene, monitoring, accounting, and participatory techniques. In brief, more cohesive communities and the outlines of a social safety net. 59. Not without shortcomings, the program has yet to develop sufficiently adequate mechanisms of coordination and partnerships with the local governments, relevant sectoral ministries, and donors. In part, due to prolonged period of weak government, low public capacity, and scarce donor presence and in part, for lack of supporting nascent government actors and educating on the benefits of social safety net structures. 60. A prime example of a governmental cross-support would be working with Ministry of Water and Mining and furthering work with the Ministry of Health. For, the majority of primary schools lack appropriate water infrastructures, compromising basic hygienic conditions and the participation of girls. The Ministry of Water and Mining has indicated a willingness — though not acted — to build basic water infrastructures in communities where the school feeding programs exist but has not. 21 And, at present, the deworming campaigns run by the program in collaboration with the Ministry of Health are neither systematic nor built into the primary school calendar. 61. The program envisages reinforcement of the central Technical Secretariat via the recruitment of a project coordinator in charge of, among other things, assuring appropriate coordination with complementary interventions in the expansion phase. Given 21 An additional grant under the pilot crisis response window for the Community-Driven Development Project, per World Bank Paper No. 54513-TG (2010). Also, under the Interim Strategy Note (2008), the World Bank supported two infrastructure projects. The second was scheduled for US$10 million but was granted US$26.2 million. One of its three pillars, even though in an urban area, is the provision of clean drinking water. Indicative of the government’s commitment, work has begun. 17 the centrality of Technical Secretariat’s role to the school feeding program, this should better support collaborations. 62. The outputs and outcomes of the program’s institutional arrangements would seamlessly support scale-up. Stronger coordination and partnerships with local government services, donor organizations, and UN agencies, however, would need to be factored into design. 4.3 Program Funding 63. The Technical Secretariat (TS) is responsible for overall funding of the program. It distributes funds to the six international and national non governmental organizations (NGOs) that support the project at the regional level. The transfers occur quarterly and cover meals for three months. 64. These NGOs work with the Parent–Teacher Associations (PTAs) and the broader community to select a group of women from the village to prepare the school meals, usually on a rotating basis. The funds are transferred to the PTA on a monthly basis, and the PTA pays the femmes-mamans each Friday. 65. Funds are replenished on the basis of implementation reports prepared by the regional NGOs and countersigned by the PTAs and school representatives. 4.4 Monitoring 66. Given the decentralized nature of the program and the simplicity of its schematics, oversight of progress and performance is progressively pushed downstream. In effect, this serves to minimize gaps in the transparency of operations and glitches in the quality of performance. 67. In each community, the Parent–Teacher Association (PTA), school management, and teachers supervise the delivery of meals to students. This entails oversight of the femmes-mamans and review for dietary diversity, triggering upfront training of all participants and formation of long-term human capital. In addition, the benefits brought to the community at-large, inflows of capital to small-scale farmers, traders, and suppliers, create de facto monitoring. 18 68. These formal and informal monitoring mechanisms have been tested on the ground for two years and produced invaluable lessons learned. For example, the system requires tightening to ensure that meals are provided every day with the agreed ingredients and quantities to inhibit profiteering at the expense of quality. To this end, the capacities and monitoring functions of the PTAs should be reinforced and a clear profit ceiling should be set. Women should be encouraged to keep a simplified account sheet to track their weekly expenses and a sample sheet should be added to the existing operations manual of the Emergency School Feeding sub-component. 4.5 Targeting 69. Geographic targeting was deemed best suited for the purposes of the Community-Based School Feeding program. This system is used mainly where programs are relatively small and when the probability of homogeneity for poverty incidence within certain areas is high.22 It also can result in most of the benefits going to the poor with few errors of inclusion. 70. The objective of this program related to the alleviation of household food insecurity in times of economic crisis and external shocks (flooding). Targeting therefore was based on four clear criteria: vulnerability to floods, exposure to the food price crisis, targeted by LICUS operations,23 and selected by the Community-Driven Development project, which made use of poverty criteria derived from the QUIBB and used in the Poverty Reduction Strategy. Within districts, schools also were ranked according to the incidence of poverty.24 71. In the first and second year of operation, 84 (of which at least 39 in the two flood-affected regions) and 92 primary schools were selected, respectively. The resulting figures for direct beneficiaries were based on the totality of children attending the targeted school: the first and second year covered, respectively, 16,800 and 21,300 children (Table 2). 22 In Togo, three quarters of the rural population is poor with headcounts of: Savanes 92.4 percent, Centrale 84.0 percent, Kara 80.0 percent, Plateaux 60.2 percent, and Maritime 71.1 percent (IMF 2010). 23 Low Income Countries Under Stress (LICUS) is a World Bank Trust Fund aimed at assisting the most marginalized and fragile states, those experiencing the most severe conflict and institutional problems. 24 The incidence of poverty: Savanes 90.5 percent, Centrale 77.7 percent, Kara 75.0 percent, Maritime 69.4 percent and Plateaux 56.2 percent (IMF 2010). 19 Table 2: National Statistics on Schools and School Enrolment (2006–2007) Regions  Number of Schools  Number of Students Average Number of Enrolled  Students per School  Lomé Golfe  996  207,721  209  Maritime   1,099  210,894  192  Plateaux  1,514  241,670  160  Centrale  660  120,291  182  Kara  745  126,942  170  Savanes  572  114,099  199  Total National   5,586  1,021,617  183  Source: Operational manual (April 2008) - Ministère de la Coopération, du Développement et de l’Aménagement du Territoire, Direction Générale, Secrétariat Téchnique 72. Across the 20,000 children averaged each year, the characteristics of poverty were similar. The median direct beneficiary lives in a poor household where the parents can barely afford transportation costs, pocket money for meals, or education costs. In addition, they often work on the farm or in the household; some were initially enrolled in school but then dropped out. 73. Quite apart, girls’ attendance increased even though no specific effort was made during project design to ensure a higher level of female participation. Interviews and discussions with parents have suggested that the incentive of school meals and the fact that children are protected within the school are factors for higher female participation. 74. The net transfer of the program, through meal distribution to children, can be estimated at about 10 percent of total household annual consumption for an average household with two children at school or about 20 percent of its annual food consumption, based on average rural household consumption derived from the QUIBB (2006). 75. In fact, 70 percent of households had at least one child enrolled in a primary school that benefited from the program, with an average of two children per beneficiary household. The number of beneficiary children increased proportionally with household size, a strong indicator of the poverty status of the household. The program therefore is effective in distributing resources in alignment with the poverty levels of rural villages. 20 76. Tangible and intangible effects have accompanied the provision of meals to primary school children. The access to education services has been improved and in so doing, the opportunities for income-generating and productive activities among household members were augmented. The predictable demand for food has increased local market access, and communities have become increasingly empowered. In some instances social cohesion among household and community members was solidified and in others, strife-riven. Although targeting school children as primary beneficiaries, the program was set up to and did indeed reduce vulnerability of households in food insecure areas. 4.6 Procurement of Food 77. The program relies on a decentralized food procurement approach, where the village women in charge of preparing meals (femmes-mamans) are also responsible for all food purchases. The approach is informal with no coordination with national or local government institutions, regional or local NGOs, or catering services. The school management and the Parent–Teacher Associations (PTAs) have no jurisdiction over the food purchasing process. 78. This simple procurement approach, based on individual purchasing of a limited amount of food, allows for greater ownership and accountability at the grassroots level. This plays an important role in deterring obstacles or mistakes that could jeopardize feeding school children. At the same time, it lends to difficulties in predicting food quantities, controlling food quality, and determining the provenance of food supply. In fact, the program does not have complete information on the sources of food. In other words, that the program is building community cohesion by capturing the largest swath of the local supply chain, from producers to manufacturers, is not measured (Box 3). 21 Box 3: Local Procurement An assessment addressing the purchasing process that supplies food to the school feeding program would allow understanding the program source of food (how local is local supply), how the supply chain is organized and who benefits from it, the conditions for small farmers to access the market generated by the school feeding program, and the impact on local agricultural production and local development. Following are some of the main areas of investigation and questions. Food Quantities Food Source  How much food is purchased  Where does the food purchased for the school individually by each village woman feeding program come from? (broken down by ingredients/product and  Do food insecure areas have problems in quantity/weight)? supplying the school feeding program needs?  Is the amount of food required for the  When and what are the items in the food basket school feeding program always available not adequately accessible and available in food locally insecure areas?  How far does each woman travel in order  What share of the school feeding program supply to purchase the entire ingredients of the is bought from small-scale farmers? food basket for meal preparation?  How far is the nearest market from the school?  Prices and market trade modalities?  How are small-scale farmers organized?  Is the demand for food by the school  How could the program create an enabling meal program causing an increase in environment for small-scale farmers to access the food prices? school feeding market?  Are farmers directly selling on the  What are the conditions for small-scale farmers market? How are they organized? Are to access this market and how can this be there intermediaries? Who are the supported by the government, donors, or the traders? private sector?  Do traders apply an interest rate to the  What would be the most favorable procurement price of food when women buy credit? approach?  What is the cost of direct procurement from individual small-scale farmers or farmer groups? Source: Authors (2011). 4.7 Costs 79. The full cost of the on-site meal program in Togo includes the administrative costs for running the program, commodity costs, labor, and other costs. Comparison to a literature survey and four other Sub-Saharan Africa programs shows high Kcals to be the main driver of cost in Togo’s program; when meal caloric value is standardized, Togo falls in the low end of the range of school feeding prices. 80. Administrative Costs. The program runs with 15 percent administrative costs to cover administrative–monitoring (10 percent) and transportation (5 percent) expenses, 22 resulting in a very cost-effective operation. This is due mainly to the active involvement of the community in key monitoring and service delivery functions. 81. Meal costs. Meal costs account for 85 percent of the total on-site program cost, and have no set up costs. Commodities — the costs to purchase, transport, and store the food — were calculated at 60 percent while the labor to cook (inclusive of expenses for water and wood) and provide handling services represented 25 percent. Utensils and pots were mainly brought to school by the femmes-mamans. The meal cost per child (without administrative costs) in Togo in the first year was approximately US$44 (based on 16,800 children and 140 school days) and US$56 in the second year (based on 21,000 children and 180 school days) whereas the overall full cost (administrative plus meal costs) was US$51 and US$64, respectively.25 When controlled for 700 Kcals and using Year 2 data, the figure drops to US$28 (Table 3). 82. Meal unit price. One meal costs US$0.31 (US$0.36 with administrative costs), and each femme-maman is responsible for the preparation of approximately 35–40 meals. Earnings very much depend therefore on the availability and prices of products at the nearby markets and on observance of the agreed quantities for each meal portion. Currently, the diversity of the menu is being reviewed for additional sources of vitamins and means to fortify meals. A revision along these lines could make the program more costly and raise the meal unit price. Incorporating one fruit per day while reducing the quantities of the remaining commodities could serve as an offset. 83. Income level of femmes-mamans. These earnings are in line or slightly higher than the minimum wage levels in the informal and agricultural sector.26 Factors that cause the women’s earnings to fluctuate are food pricing and availability as well as transport and distance to market. With the exception of those earning up to 30 percent of the meal costs (in some cases up to US$4/2,000 CFA per day), the majority of surveyed femmes-mamans and 25 In the findings section, the costing issue will be elaborated with a comprehensive table and discussion on overall program costs. 26 The minimum wage is fixed at US$28.70 (13,757 CFA) per month in the formal sector. In the informal sector, wages are lower than the minimum wage level varying from US$12.30 to US$20.60 (6,000 to 10,000 CFA) per month and US$0.55 to US$1 (270 to 455 CFA) per day. In the agricultural sector, the minimum wage is a little more than US$1 (500 CFA) per day with large regional disparities. 23 the project coordinator agree that an increase in the unit price would guarantee the maintenance of quality meals and a better appreciation of the work. The general opinion of the program management is that about a 10 percent profit should be the maximum allowed by the project in order to maintain the quality and quantity of food served. 84. Comparison programs. Grosh et al. (2008) found the average administrative cost for on-site school meal programs to account for about 30 percent (ranging from 10 to 55 percent) of total program costs while Galloway et al. (2009) placed costs in four World Food Program (WFP) school feeding programs in Sub-Saharan Africa — The Gambia, Kenya, Lesotho, Malawi — at 40 percent. 85. The estimated full costs of on-site meal programs in The Gambia, Kenya, Lesotho, and Malawi ranged from US$28 to US$63 per child per year (weighted average US$40 per child per year).27 On average, commodity costs accounted for 59 percent of the total expenditure. The contribution from local communities averaged 5 percent of the total cost (varying from 0 percent in Lesotho to 15 percent in Kenya), or about US$2 per child per year on average (Table 3). 27 Program costs were standardized using a typical 200 feeding-day school year and a daily ration of 700 Kcal, and adjusted for breaks in the food delivery pipeline. 24 Table 3: Comparison of Prices with Other School Feeding Programs in Africa28 Variables Kenya Lesotho Malawi Gambia Togo Number of children 1,156,000 390,000 214,000 113,000 21,000 Number of days per year 195 180 183 146 180 Kcal per day on average 703 n.a. 376 551 1,150 Donor costs (%) 71 13 92 82 100 Government costs (%) 14 87 3 8 0 Community costs (%) 15 6 5 10 0 Total cost per year 21,935,000 21,556,000 4,962,000 2,787,000 854,878 Cost per child per year 19 55 23 24 64 Standardized cost 28 62 49 43 28 per child per year* Note: *Adjusted to a 700 Kcal per day and a 180-day school year. See Annex 1 for the model used to calculate school feeding prices. Source: Galloway et al. (2009) and Author‘s calculations (2011). 86. The breakdown of costs in the two sets of programs, Togo and the reviewed group, is based on different assumptions in terms of the number of school days, quantities of food, and caloric intake of meals. The Togo program offers the largest food basket with an average daily caloric intake of about 1,200 Kcal. In the reviewed school feeding programs, costs were standardized using a 700 Kcal daily ration and 180 school days. By comparision the Togo school feeding programme remains a relatively more expensive intervention. Costs are largely influenced by the programs decentralized community development mechanism which means that purchases are done for individual schools on a weekly or daily basis and commodities are purchased on the nearest markets. There are clearly more cost-effective ways to feed children in schools for example simply providing them with fortified snacks as opposed to a more traditional school meal (school lunch). And this would significantly cut the preparation time and overall costs. However, the wider aspirations of the program must also be taken into account, including efforts to involve the community and general income for women through involvement in various parts of program implementation. 28 School feeding: Outcomes and costs, by R. Galloway, E. Kristjansson, A. Gelli, U. Meir, F. Espejo, and D. Bundy, in Food and Nutrition Bulletin, vol. 30, no. 2 - 2009, The United Nations University 25 87. In a largely fragile country, the program has consistently delivered meals and provided real sources of income. Its ability to set and uphold the meal unit price speaks to its skeletal structures. Nevertheless, a worst-case scenario for recurrent food price spikes should be factored into the design, in order to protect the newly perceptible social protection structure. 4.8 Benefits to Children 88. School enrolment and attendance. The program has made an important contribution in attracting and retaining beneficiary children in school, in providing access to school to children who are older and have not yet enroled, and in increasingly attracting girls. Results on enrolment, dropout, and absenteeism rates and a reduction of the age at entry in primary school 29 in all regions are particularly positive among girls. An increase in new enrolments in beneficiary schools in 2009–2010 was 16 points higher than in the group of control schools (World Bank 2010). The dropout rate was 0.9 percent in beneficiary schools compared to 1.4 percent in control schools. Absenteeism is lower among children enroled in beneficiary schools, and the retention rate expressed in terms of percentage of children attending school every day is 2 percent higher than in control schools. 89. Despite the fact that the 2008–2009 school year had a general increase in enrolment and attendance due to the government initiative of abolishing school fees, school enrolment and attendance data in the evaluation of sample communities increased more significantly in schools served by the program than those without any program, particularly among girls. Village teaching staff, it turns out, still are largely funded by the parents. In the surveyed areas, two to three teachers work in parallel with the state-funded teachers. In the rural village schools, teachers funded by the communities generally outnumber state-funded teachers. In this sense, wherever positive changes in enrolment and attendance have been reported in the beneficiary schools, these result undeniably from the incentive approach of the school feeding program. 90. Although changes in the education performance were not examined by the field evaluation, children feel that their presence in school has changed, that they are more 29 The average age of girls in first grade was 6.9 in beneficiary schools and 7.6 in control schools. 26 concentrated and less tired, that they have more time to do their homework, and that they are able to spend more time within the school perimeter thus allowing them to make friends. Not only is the school lunch addressing short-term hunger, it also is providing income replacement for those children who worked during the lunch hour or who walked considerable distances home for the meal. 91. Dietary Intake. This program is active in areas where securing food is a real problem. More than 70 percent of children in rural Togo do not satisfy their daily caloric intake. The program provides students with one large meal per day and approximately the right amount of daily calories in a full-day school program. A nutritional assessment of food served at beneficiary schools shows that the average caloric intake per meal is approximately 1,200 Kcal per day and that overall meals are providing between 60 and 70 percent of daily caloric intake needed for primary school age children. Table 4 provides a breakdown of the general energy content of in-school meals. Table 4: Recommended Nutritive Value of Rations for Primary Schools School Type Acceptable Range of Nutritive Value per Ration for Children 6-12 Years Old Energy (Kcal) Protein (gram) Fat (gram) Half-Day 600–900 (30–45%) 16–24 (40–60%) 7–11 Full-Day 1,200–1,500 (60–75%) 28–36 (70–90%) 14–17 Boarding Up to 2,000 (Up to 100%) At least 40 (Up to 100%) >23 Source: Bundy et al. (2009). 92. The most cost-effective nutrition interventions are those that target the first 24 months of life. From this perspective, providing food to school-age children cannot reverse the damage of early nutritional deficits. School feeding can improve school participation; alleviate short-term hunger; and increase children’s ability to concentrate, learn, and perform specific tasks. These effects are not limited to but are greater among children who are also chronically undernourished. If the food is fortified and combined with deworming, there may be additional benefits for children’s cognitive abilities and educational achievement.30 30 Bundy, D., C. Burbano, M. Grosh, A. Gelli, M. Jukes, and L. Drake. 2009. Rethinking School Feeding: Social Safety Nets, Child Development, and the Education Sector. Washington, DC: World Bank. 27 93. According to the recent assessments, the majority of children are satisfied with the quality and quantity of meals. Parents have pointed out that meat or fish is not provided on a daily basis, raising concerns as to adequate protein intake. In fact, the femmes- mamans have reported that the fixed rate of US$0.31 does not afford adequate amounts of meat per child (Table 5). 94. Meals consumed in schools are also more diversified than those served at home, as they include animal proteins usually absent from the children’s diet (Table 6). For the moment, only anecdotal information from teachers and parents suggests that children have been benefiting from an improved and diversified diet. In particular, the number of children has not only doubled in the visited schools, but children tend to be ill less. Since the beginning of the program, little or no absences have been reported. 95. However, adjustment to the diet needs to be considered as meals seem not to be nutritionally well balanced: for example, there is a predominance of staple foods and a scarcity of vitamins, calcium, and protein (fruit is not yet provided although recommended in the project document). The program is also considering options for micronutrient fortification and complements, including locally-processed food fortifiers or locally-grown food fortifiers.31 Moreover, the program concentrates feeding into one large meal; for the purpose of learning, it is not the best choice. Dividing feedings into two smaller meals, breakfast (e.g., providing a caloric snack) and lunch, would better facilitate learning. Changes in the schedule and type of meals offered would clearly entail relevant operational consequences in terms of costing, availability of adequate ingredients for breakfast and lunch, and a more complicated time and workload distribution among the femmes-mamans. 31 Moringa is a plant currently being promoted as a means to combat poverty and malnutrition. It grows quickly in many types of environments, and much of the plant is edible. The leaves contain all essential amino acids and are rich in protein, vitamins A, B, and C, and minerals. Togo grows this plant therefore a commission at the University of Lome is studying its qualities, the quantities needed to fortify food in relation to individual ingredients, modalities, and costs. 28 Table 5: Example of Weekly Menu in Two Regions, Savanes and Maritime Monday Tuesday Wednesday Thursday Friday Témanoukopé (Gboto District in Maritime Region) Corn noodles + Beans + cassava "Ayimolou" Noodles + Bread Rice + sauce vegetable sauce + oil or sauce + (Rice + beans) + + fish (with fish) with fish/meat fish fish Shalom (Dapaong Town in Savanes Region) Rice + tomato Noodles + gari + Beans + noodles Rice noodles + "Ayimolou" (rice sauce (with meat fish + fish vegetable or + beans) + grilled or fish) peanut sauce fish (with meat) Source Agbo E., 2010b in Evaluation de l’opération de fourniture de répas aux écoliers des zones vulnerables du Togo, Analyse quantitative. Table 6: Summary of Nutritional Characteristics of Four Types of Meals Type of Meal Calories (Kcal) Proteins Calcium Iron (g) (mg) (mg) Meal based on rice 1,951.40 80.72 284.74 23.48 Meal based on corn noodle, millet, and 1,295.07 48.14 841.65 21.83 Meal based on rice and niebe’ (black 1,011.50 63.61 202.31 15.49 ) Meal based on simple noodles 810.52 41.57 55.61 5.73 (macaroni) Criteria OMS and *AFSSA 2007 1,457.80–2,325 17.53–34.89 700–1,200 7–10 (children 6–21 years) per daily criteria Criteria for lunch meals: 60% of daily 874.80–1,395 10.50–20.90 420–720 4.2–6.0 criteria Notes: *Agence française de sécurité sanitaire des aliments (French Agency for Food, Environmental and Occupational Health Safety). Source: Agbo E., 2010b in Evaluation de l’opération de fourniture de répas aux écoliers des zones vulnerables du Togo, Analyse quantitative. 96. In order to better monitor the benefit of nutrition in the school feeding program, anthropometric measures (weight/height) on primary school children have been collected in a sample of beneficiary and non-beneficiary schools and compared to a regional index of malnutrition. A second round of measurements will be collected at the end of the program. 97. Finally, qualitative interviews of the parents of beneficiary children indicate that parents tend to buy and consume the same amount of staple food (corn or rice), implying 29 that food is better distributed within household members allowing members of the family especially younger children to take advantage of this relative increase in food availability within the household. 98. Substitution effect. The evaluation has not looked into the possible negative consequences of substitution effect. Meaning, a child served with a meal at school forgoes an additional meal at home because parents simply provide this child with less food at home, thus reducing the associated benefit of the meal program. Some of the interviewed parents of beneficiary children claimed that the school meal is by far the largest—and in many cases the only—meal of the whole day, and that the child is not hungry or demanding food in the evening since portions are generous in school. 4.9 Economic and Social Benefits to Households and Communities 99. Overall the program has encouraged positive household changes in: (i) the behavior of children; (ii) the alleviation of daily household expenses for food, accounting for up to US$8–10 per month, as at least one or more meals per day are subsidized by the program and income for selected women hired by the program; (iii) increased awareness of hygienic practices at home as drivers of good health; (iv) an improved capacity for planning household expenses and management of resources; (v) an achieved sense of pride for sending children to school while saving money at the same time; and, (vi) a larger commitment to and interest in school matters from both parents. 100. The program is perceived as positively bringing both social and economic benefits to households. In fact, perceived benefits go well beyond the immediate substitution effect of providing meals to children at subsidized prices. Poor households save on the daily allowance given to the child to buy a meal and have the opportunity to use such savings for productive purposes or, more likely, to improve overall household nutrition and health levels. Enrolment and attendance at beneficiary primary schools have improved, especially among girls, particularly in those areas where such rates are low, partly because of the high poverty levels and high food vulnerability and partly because parents take children out of schools for additional help in the fields. The elimination of school fees, instituted by the government in 2008, is not yet sufficient to make sure that children are actually attending school. Households are therefore in charge of supporting the education of their children by paying 30 partly for the village teachers. Often payment of teachers becomes a reason to pull children out of school in periods of economic difficulty. To this end, the school meal program becomes an important incentive for parents to enroll their children in school and not to pull them out during the school year as a coping mechanism when financial difficulties occur. 101. Economic benefits, although limited, are tangible and take the form of savings and income. Household savings have been estimated at an average of US$8–10 per month. The femmes-mamans make a daily income of US$1–3 and small traders have in some cases doubled their supply of food on the market. The program represents a predictable and ongoing demand for services and food. The demand is defined by the number of school children to whom lunches are provided daily and by the menu and food basket. 102. Overall, the local economies reflect the chain of actions generated by the preparation of meals, positively affecting a number of local stakeholders. The program injected about US$1,400 per month in each rural community in the first year and US$1,600 per month in the second year. This represents a considerable amount of money for small villages where the volume of local trade transactions is very limited. 103. Even though it is difficult to quantify the program’s impact on local trade and its multiplier effects on the local economy, results from focus groups with local traders and households suggest that a local economic dynamic has indeed been set in motion: Trade transactions increased, informal saving mechanisms were created, and new income- generating activities were set up. In addition, the project created a stable income-generating activity for about 600 femmes-mamans. After prolonged social stagnation, these women are the first embodiment of long-term human capital, learning to organize and cater large quantities of food, cook using healthy and hygienic practices, and budget for the meals. Their earnings, in addition, often are invested in social assets for their children (school furniture, health services) or income-generating activities. 104. The income-generating benefits accrue mainly to suppliers of cooking services and to suppliers of food and other complementary services (femmes-mamans, small traders, small-scale farmers, and transportation). However, when looking at the beneficiary communities as receptors of school feeding benefits, there is growing evidence that these 31 programs can help create a stable demand for food at the local level, which, in turn, has important multiplier effects on the local economy and the local community (Table 7). 105. The Togo school feeding program the yet unexplored potential for the generation of local development outcomes. The working hypothesis is that when meals are supplied locally and food is locally produced by small-scale farmers, and purchased from the community or the nearby markets, then the school feeding program represents a stable and predictable demand for food that is locally satisfied. In this perspective, small-scale farmers and their families could be participating in the access to wider market opportunities generated by the school feeding program. Interviews with small traders (which are often the wives or relatives of small farmers) have indicated that since the program, they are bringing double their normal supply of goods to market. In fact, each week food is purchased on the market to provide daily meals to an average of more than 100,000 children in more than 90 communities. To date, there is no quantified evidence of the increased opportunities for small-scale farmers. The program evaluation should prioritize the gathering of data to determine the standing of small-scale farmers. However, there is no doubt that each day the school feeding program represents a predictable, stable (in terms of quantities), and year-long demand for food. Table 7: Injection of Funds into Communities, Year 1 (2007–2008) Number of Cost of Meals per Year Cost of Meals per Month Number of Regions Meals per Schools CFA* US$** CFA US$ Year Maritime 20 560,000 84,000,000 179,158.05 13,200,000 28,153.4 Plateaux 19 532,000 79,800,000 170,212.57 12,540,000 26,747.6 Centrale 13 364,000 54,600,000 116,437.24 8,580,000 18,297.28 Kara 15 420,000 63,000,000 134,350.66 9,900,000 21,107.5 Savanes 17 476,000 71,400,000 152,268.04 11,220,000 23,927.8 Total 84 2,352,000 352,800,000 752,426.56 55,440,000 118,233.5 Notes: * The yearly cost of meals includes all costs associated with meal preparation from commodities to labor, food, transportation, etc. This is based on 84 schools, a daily cost of US$0.31 per meal, a year of 140 days and 22 school days in each month. ** US$ figures may change depending on currency exchange rate. Author’s calculations based on project costs included in the Operational Manual (2011) 32 106. Perceived economic benefits. According to the assessments, the program has provided economic benefits to the main program stakeholders while launching a dynamic of possible local economic development. 107. Parents. The parents of beneficiary children admit that not having to provide for one or more (average household has two children in school) meals per day alleviates daily household expenses. It is likely that the program affords them the opportunity to make food more available in the household and to buy food for the remaining household members (Box 4). Box 4: Focus Groups about Savings In the Maritime Region (Temanou Kope), a father explained that on school days he would spend 300 CFA (US$0.60) on food per day while on the weekends he spends almost twice as much. Another respondent declared saving about 10 “bols� of corn per month, equivalent to more than US$10 (5,000 CFA) per month. In Klologo (Maritime), another respondent remarked that before he would spend more than US$16 per month and would consume up to 40 bols of food and that with the project he spends approximately US$10 per month and 20–25 bols per months. In Kpembona (Savanes), a mother explained that before she would use one bag of corn (equivalent of 100 kg) for 12 household members and would give approximately US$2 per week (1,000 CFA) to her child to buy lunch in school. Source: Agbo E. (2010b). 108. Saving correspond to approximately US$8–10 per month and almost half of the daily food meal portion. Savings are used to: (i) better distribute food within the household, allowing for younger children to benefit from a proper amount of food; (ii) enhance productivity like buying fertilizers, investing in peanut production, and buying breeding animals; (iii) in some cases, respond to the increasing needs of the school meals program (developing vegetable gardens, selling wood and water); and (iv) support education and health expenses as well as help relatives in need. The net transfer of the program through meals can be estimated, based on the 2006 QUIBB survey, at about 20 percent of the yearly food consumption of households. 33 109. Femmes-mamans. The program has hired approximately 600 femmes-mamans in two years. The women are now earning a daily income of US$1–3, which amounts to a monthly income of up to over US$60. Each school requires the employment of approximately 5–10 femmes-mamans. Depending on the community, a single group works the entire school year or multiple groups rotate throughout the school year. The women receive a fixed fee of US$0.31 for each child; with this money, they must purchase the food and prepare/distribute the meals. Preparation includes: purchase of staple food, meat, fish, and vegetables on the market, transportation, storage when needed, purchase of wood and water in the community, cooking of meals. On average, one woman cooks for approximately 35–40 children and works approximately 5 hours in the morning–early afternoon. This represents a job with a reliable income and given the scarcity of employment opportunities, complaints and conflicts have arisen. 110. Income variation depends on a number of factors, including: (i) the number of children they are preparing meals for. Most femmes-mamans prepare meals for approximately 35-40 children but work in teams of differing sizes (Table 8). In the same communities, teams operate on a rotation basis; (ii) the number of femmes-mamans sharing the work load for the preparation of daily meals (femmes-mamans are preferably organized in teams); (iii) the availability on the nearest market of food items; (iv) the price variations according to seasonal availability; and (v) regional differences and fluctuations in prices. Finally, some differences have been reported on the composition of the meals and use of ingredients. Basically, there have been reported cases of femmes-mamans who, in order to make a larger profit, chose to reduce the amount per serving of the most expensive ingredients (often meat and fish) while serving larger portions of staple food. In this case, the caloric amount remained somehow similar but the level of protein was reduced. The program should strengthen its monitoring tasks, reinforce the role and capacity of PTAs in controlling the quantity and quality of meals, and check for those cases where a higher profit has been made at the expenses of the quality. Women are also encouraged to keep a simplified account sheet (a sample sheet should be available in the Operational Manual) tracking their weekly food expenses. 34 111. In addition to monetary gains, the program offers the femmes-mamans with the opportunity to learn a profession where they gain the culinary skills and learn to budget for large groups of people. Table 8: Number of Femmes-Mamans in Sample Villages Region/Villages Number of Femmes-Mamans Status Maritime Klologo 3 Cooked for the school prior to program Kovie 6 Cooked for the school prior to program Temanou kope 12 Trained by a regional NGO Savanes Dapaong 15 (rotate) Cooked for the school prior to program Kpembona 30 (rotate) Tierou 12 (rotate) Trained by a regional NGO Source: Agbo E. (2010b), Evaluation de l’opération de fourniture de répas aux écoliers des zones vulnerables du Togo, Analyse qualitative. 112. Femmes-mamans are using their income to respond to social needs (health and education) and to re-invest in agricultural activities, especially purchasing fertilizers or paying in-farm labor costs in the fields. In fact, their daily routine leaves little time to work in the fields, so in many cases the women support their husbands with expenses needed to increase agricultural productivity. Finally, some women have used their earnings for income- generating activities like baking scones and brewing beer, buying and breeding animals, setting up vegetable gardens, or purchasing and storing such agricultural products as corn, beans, and millet (Table 9). 35 Table 9: Reinvestment of Profit in Sample Regions: Maritime and Savanes Frequency of Similar Responses Activity Maritime Savanes Kovie Temanu Dapaong Kpembona Purchase clothing and medicines, and 2/6 5/7 8/13 24/26 payment of school items and fees* Agriculture and product** 2/6 2/7 3/13 20/26 Household utensils*** 1/6 2/7 18/26 Contract for apprenticeship 1/6 2/13 Tontine 22/26 Note: *Childcare includes, in order of importance: health care, contribution to village teacher, secondary school and college fees, school equipment, clothing and special event dresses, food, driving lessons for older children; **Sharecropping, peanut production, purchasing fertilizers, breeding animals; production of local beer and bread; ***Dishes, pots, oil, soap, and salt. Source: Focus group discussions (Agbo E., 2010b in Evaluation de l’opération de fourniture de répas aux écoliers des zones vulnerables du Togo, Analyse qualitative). 113. Small traders. In general, the food is purchased from markets in the main district towns unless items become available in the small market of the beneficiary community or in nearby villages. Small traders supplying commodities for the school feeding program can count on the demand for food generated by the program. Each school requires meals for approximately 200 children every day of the school year, thus this demand for food, although small in scale, is predictable. Purchases are done weekly for imperishables while fresh ingredients are bought daily. Femmes-mamans do not have regular suppliers but pick and choose the best deals on the market. This allows sellers to take advantage of the accrued purchases on the markets. Interviews with small trader indicate that the demand for certain items has doubled since the program began and that they are bringing double their normal supply of goods to market (Table 10).  Table 10: Local Production and Imported Food Items Ingredients Purchased on Local Markets Imported Ingredients Corn, maize, millet, beans, fresh vegetables Rice, oil, salt, broth cubes, wheat pasta, and (mainly okra, peppers, and tomatoes), cassava, tomato paste paprika, fish, and meat Source: Interview with the Director of Secrétariat Téchnique (2011). 36 114. Small-scale farmers. The program has the opportunity to play an important role in allowing small-scale farmers to access markets with the purchases generated by the school feeding program. In fact, each week products are sold on the markets of the nearby school communities and these are used to provide daily meals to more than 21,300 children in 92 communities. 115. To date, there is no quantified evidence of the increased opportunities for small farmers, as well as no exact information on the sources of procured food. Anecdotal information suggests that staple products as well as beans and condiments for the preparation of school meals come mainly from local small-scale farmers and that these are usually the relatives (often the husbands) of the traders at the market. Moreover, weekly supply is done individually by the femmes-mamans, who buy the quantity of products needed for their share of meals from different sellers according to the best price. Although there is no doubt that the school feeding program represents a predictable, stable (in terms of quantities), and year-long demand for food, this is still an area where investigation is required in order to understand the potential earnings for small-scale farmers, tracing the provenance of local products to determine where are the producers located, how much of this local production is local and adjacent to the program areas (e.g., meals are also composed of a number of imported items). 116. Finally, it would be important to analyze if the program could be successfully linked to other programs designed to increase agricultural production or activities that promote the creation of an enabling environment for small-scale farmers. This would be particularly important for those schools that are located in remote areas not connected by roads and thus not accessible by car or motorcycle. In this case, access to markets becomes very time-consuming and expensive, and diminishes the income made by the femmes- mamans because of the additional time demands and taxi fees for longer distances. In these areas (i.e., Plateaux features 16 schools in remote areas), the program could look into options to become partly self-sustained by finding ways to support local agricultural production. 117. Generation of economic dynamics. In the first year of operations (2008–2009) when 84 beneficiary communities were covered, the program injected cash for approximately US$1,400 (the equivalent of an average 738,672 CFA). In the second year, 92 schools were covered and the amount increased to US$1,600. Daily meals have a fixed unit cost across all 37 regions of US$0.31 (150 CFA). Purchases for the school meal program can generate a snowball effect as additional income is provided to the small traders, small-scale farmers, vegetable producers, service providers such as taxi drivers, and cooking utensils suppliers (Table 11). Table 11: Resources Injected in the Communities, Years 1 and 232 Number Year Number Overall Funding Funding per of of Community Students Schools per Month (US$) Daily Daily Weekly Weekly Monthly (CFA) * (US$) ** (CFA) (US$) (US$) 16,800 2008- 84 2,520,000 5,371.05 12,600,000 26,855.20 118,163.10 1,406 2009 21,000 2009- 92 3,150,000 6,708.74 15,750,000 33,539.70 147,592.20 1,604 2010 Totals 265,755.3 Additional Additional Additional Resources in Year 2 in the Total Number of Communities**** Students in Schools in Daily Daily Weekly Weekly Monthly Monthly Year 2 Year 2 (CFA) (US$) (CFA) (US$) (CFA) (US$) 4,300 8 630,000 1,337.69 3,150,000 6,684.50 13,860,000 29,429.10 Notes: *CFA/US$ that the program allocates daily for the preparation of school meals in the totality of beneficiary school communities. This amount will then in turn be used to purchase food in the market, pay for transportation, etc. **US$ values can vary according to the currency exchange rate. ***US$ that the program allocates each month on average (counting equal numbers of students per school) in each community. ****The additional amount of funds injected in Year 2 daily, weekly, and monthly as a result of the extension in coverage of eight additional schools. Source: Author’s calculations based on operational manual 118. Empowerment for children, women, and households. Children enjoy their lessons, learning, and the time spent in school with their peers. They become agents of change in their homes and with their peers as they gain knowledge of good hygiene, nutrition, and the importance of being able to concentrate in class. A large groups of community members is involved in program implementation at all levels, acquiring new skills, participating in important decision-making about their children’s education and community welfare. Moreover, benefits of the program have been spread throughout the 32 Based on data from the assessment: Evaluation de l’opération de fourniture de répas aux écoliers des zones vulnerables du Togo (analyse quantitative). 38 community, helping to relieve households of financial and food concerns and become better able to provide for their children’s education while being able to better plan resources in order to respond to economic and social needs. Communities have also learned new skills, becoming aware of the importance of hygiene and becoming active counterparts of the local political and administrative institutions (Table 12). 119. Institutional strengthening. This safety net program is largely based on the role of informal institutions, regional NGOs, grassroots organizations, and pre-existing community mechanisms (i.e., the Parent–Teachers Associations (PTAs) are very much involved in the overall implementation and supervision of the in-school meal supply program). Moreover, the community-based approach has promoted wide community empowerment and ownership, increasing parents’ participation in collective matters related to school, health, and community well being. In order to improve performance, the program requires the participation and active involvement of some local institutional stakeholders such as the health and sanitation department, the water district and community committees, the primary school inspection authorities, all of which are starting to have clear mandates and to put capacities in place. The program with time is becoming an opportunity for the creation of new institutional dynamics, promoting new roles and responsibilities in the local administration. 120. Within the formal institutional setting and central authorities, the program management unit (Technical Secretariat), works efficiently but needs strengthening in terms of its capacity to coordinate with other sectors and other donor programs. Finally, the fact of piloting a program that centers program implementation and service delivery on informal structures while bridging to the formal structures of social protection and other sectors, provides important learning in the areas of innovative tools for social protection in vulnerable food insecure areas. The merit of the school feeding program is therefore not only to have responded promptly to a national crisis but also to build capacity both at the national and decentralized levels, while testing implementation mechanisms to ensure best and quick results. 39 Table 12: Areas of Reinvestment of Income or Savings from School Meals Program Parents Femmes-Mamans Traders Village school fees Village school fees Household needs Medicines, vaccines, other health- Medicines, vaccines, other health- Funeral and other related needs related needs special events Clothes, household utensils, motto- Support to agriculture by payment of Tontine to mobilize taxi children to school labor, fertilizers, share crops funds Tontine to mobilize funds Production of local beer, bread, etc. Support to relatives in need Clothes and household utensils Fertilizers Contract for apprenticeship Peanuts, local beer, and other Tontine to mobilize funds forms of local production Buying breeding animals Vegetable gardens (not in Savanes) Wood, water, and chickens to sell to the program Source: Agbo E., 2010b in Evaluation de l’opération de fourniture de répas aux écoliers des zones vulnerables du Togo, Analyse qualitative. 40 V. The Community-Based School Feeding Program as a Safety Net: A Model for Fragile, Food Insecure States 121. This section assesses whether the Community-Based School Feeding program is an adequate safety net for fragile food insecure states. First, it summarizes the genesis and rationale of this particular safety net and summarizes its key features. Second, it looks at the advantages of the community-based approach versus traditional school feeding. Third, it assesses under which conditions this safety net program is replicable and scalable. 5.1 Genesis and Rationale 122. The school feeding program was designed as a complementary intervention to the community development project in areas with very high food security problems. It was conceived as a response to the food security crisis, within the context of weak institutional environment and poor donor presence. Implementation was decentralized and involved multiple local stakeholders. The rationale for using an informal approach through community-based mechanisms derives from the need for immediate results and ready-to- deliver solutions in a situation of absence or weakness of formal safety net systems. These dictated the requirement for implementation arrangements that could build on what existed on the ground so that the program could start disbursing and delivering in a minimum amount of time and with a certain degree of success. This was particularly important in a situation where the institutional base and capacities were very low. Graph 2. Cost Comparison 41 Source: Elena Galliano (2011). 123. Following are the main characteristics of the community-based approach to school feeding: 124. The program is built on existing community interventions (CDD and EPPR). This allows for synergies and for using as much as possible the organization and implementing structures already existing on the ground (the PTAs and communities are well prepared to consult on program strategic decisions as well supervising operations, since the CDD approach is based on a demand-driven approach); 125. Implementation is based on informal existing community practices. The program has basically formalized the existing system of femmes-mamans, who already prior to the program were preparing and selling food/meals in the schools by: subsidizing their work, providing them with a structure in terms of quality and quantity standards to meet (menu are discussed every two weeks and have to have a given amount of caloric intake and weight), and rules about working conditions (hygiene) as well as training on simple accounting procedures, etc.; 126. Implementation mechanisms are linked to existing community-based organizations and institutions. Program implementation is mainly done within the school; therefore management and coordination on the ground are both carried out by the school management and the Parent-Teacher Associations. These are strong grassroots organizations, especially important and vocal in the village; 127. Decision-making belongs to the community. Decisions on how to improve program implementation are made within the school community (PTAs, parents, teachers, involved local NGOs). For example, the majority of the surveyed schools decided after some time that it was preferable to set up a proper and cleaned up premise within the school perimeter, instead of relying solely on home cooking for the entire school. This has helped the femmes-mamans to work more efficiently. In fact, they save time by remaining within the school after they have put together ingredients and fetched water and wood, and by not having to walk home to cook and come back to school to deliver meals. In addition, having 42 organized teams of cooks has facilitated their work as they can share utensils, benefit from some economy of scale (dressing, sauces, etc.), help each other when needed, etc. 128. Program supervision and monitoring is done with a bottom-up approach. The program is monitored using a network of informal and formal local (regional and community) institutions to ensure good coordination and monitoring of activities and quality of the program. At the community level, the PTAs, together with the overall supervision of the school management and the village committees, are providing day-to-day supervision of the program. This requires and ensures a high level of transparency and accountability. At the regional level, NGOs are selected to monitor the program, channel funds to the PTAs based on monitoring reports, and liaise with the national program coordinator (TS). 5.2 Is this Safety Net Replicable and Scalable 129. Based on the evidence collected and the experience in Togo, the school meal program is likely to be replicable in other food emergency contexts provided that there is an enabling environment and that at least three criteria are met. 130. Availability of community social capital. In particular, what has worked well in Togo is the fact that a safety net was built up based on pre-existing informal mechanisms of self-help and mobilization with the women cooking for the entire school and other forms of partnership between school and communities. Essential capacity to run the program was in fact already there. Thus replicability of this model hinges on the capacity of target communities to mobilize themselves, to use skills and established relationships at the community level as well as on all other forms of internal organization and coping mechanisms that have been developed as alternative to an absent or weak government. 131. Community cohesion and organization. This approach seems to work well in areas where communities enjoy a relatively strong degree of internal organization and cohesiveness to allow for fast start-up of the operations, good team work, and openness to the internal supervision and monitoring mechanisms that finally allow for greater accountability. Fragmented communities could become bottlenecks to efficient community development programs as they would probably take additional time to find positive links and reasons for forming partnerships and functioning jointly in a transparent fashion. 43 132. Some local food production. A minimum of local agricultural production must be available in the communities or immediate adjacent target areas. If local production were scarce or nearly inexistent, the program would have hard time to guarantee efficient and consistent supply of food throughout the year, to respond to the needs of the local schools. Procuring food from markets other than the local ones would require a higher level of program organization, with specific procurement rules, incentives, and a larger involvement of government or private sector stakeholders. 133. Scalability. Scalability of the program is a complex issue that still leaves many open questions. The program is still relatively small in scale but has started to expand its coverage with the second phase. The pilot phase has generated good results, had no major implementation problems, and has achieved a good satisfaction level from all beneficiaries. Where a pilot has already succeeded, scaling up seems to be the next logical step. In fact, the need to assess the scalability of this operation derives from the end of the emergency mode, the recognition of the success of the small-scale pilot intervention, and the willingness to extend the benefits to many more children. Thus scalability happens when program sponsors are able and willing to move the program from being “islands of excellence that serve and empower a handful of communities to operating at a larger regional scale, where many more communities can benefit from the program’s approach� (Binswanger and Aiyar 2003). 134. Scaling up of the Togo program means primarily a geographic expansion (more communities and more schools in all regions). With this lens, the main consideration when looking at scalability is about targeting. Extending geographic coverage to more schools opens up the inevitable question about who is the program really meant for, who is it supposed to benefit, does it benefit people who do not really need subsidized food, and would this go at the expenses of the poorest? As the program increases and approaches universal coverage, it may include higher proportions of non-poor children (Bundy et al. 2009). The issue in Togo is therefore looking at the inevitable inclusion errors that occur when applying geographic targeting instead of other more selective forms of targeting (which however entail other issues such as the potential social costs from stigmatization) and to look at how big that margin of error is. Further thinking is required on the type of approach to program implementation and specifically the likeability for this program to maintain the 44 community development and informal approach. In fact, this approach may not be the best all over the country. Some areas may not be capable of replicating and offering adequate informal community mechanisms to leverage for program implementation. For example, in urban areas this approach may not be applicable. 135. Another way of scaling up is for this program to integrate with other complementary programs. This implies new partnerships with other donors (UNICEF, WFP, UNDP, international NGOs, etc.), with local social services (in the case of Togo these are not yet part of local administrations but are mainly the district sectoral offices of the main ministries) and the private sector. The program has a great potential for attracting partnerships. In fact, linkages with other programs are important when they can cumulate benefits on one another and over different parts of the life cycle, so that benefits for one age group derive from interventions in an earlier age group. In Togo, the program could coordinate with the UNICEF Early Childhood Development program that tackles malnutrition and other issues in preschool children,33 or with WFP on issues related to food fortification. In addition, school feeding programs can be part of a comprehensive package of interventions so that by addressing the issue of access to education, they can take advantage of the synergistic effect that nutrition, health, and education interventions have on the overall well being of children and households. In Togo, conditions and characteristics of partnerships still have to be finalized. 136. A third way of looking at scalability is to consider an increase or improvement in the degree of integration of the program with government structures. Especially in programs aiming towards extended regional or even national coverage, the level of ties with the formal central and/or local structures may bring a greater legitimacy and recognition to the program, especially when this means that a government becomes more responsive to the demands and needs of communities. This happens when policy and programmatic goals are defined, together with the corresponding responsibilities and accountability criteria at both ends and allocated funds. In this case, the lens of scalability opens up on the issue of legitimizing programs into policy frameworks and this would finally lead to the core issue of 33 UNICEF is providing emergency nutritional assistance to malnourished children under five in the Kara, Maritime, and Savanes Regions. 45 program funding and sustainability. The Government of Togo has yet to contribute to the program. Yet, in the event that the program becomes more formally endorsed and owned, there may be risks leading to the possibility that (i) a community-driven development program based on informal implementation mechanisms may evolve into one with a more structured implementation where the original community and informal mechanisms are less relevant and central to the program functioning, and (ii) the program may risk becoming vulnerable to political choices and the programmatic platform of constituencies. 137. Moreover, the scalability of programs may entail a transformation in the original design as needs may change and stakeholders do too. For example, the community school feeding program in Togo was conceived as a safety net during an emergency and food crisis and according to collected evidence, has fulfilled its role well. Now that the emergency is over, that the government is defining its political agenda for development and future commitments, and that the school feeding program aims to extend its coverage, a few important questions remain to be discussed, such as: 138. Will the Togo school feeding program maintain its characteristics as a safety net, and who will the program benefit in case Togo graduates from the state of food insecurity and political fragility? 139. Will the program shift its focus, becoming primarily an education tool important in supporting the education agenda? 140. Will the school feeding program have to redefine its objectives solely according to educational outcomes of children? 141. Will the program continue to rely on community-based informal mechanisms? 142. Can it be scalable in urban areas and would such a program make sense in non-rural areas? 143. But above all, is the question of who would support the scaling up of this program, and how expensive would this be according to different scenarios (regional, national, etc.). Paramount to this discussion is the definition of the role of safety nets in the social protection agenda of Togo, the role of informal safety nets and their position vis-à-vis 46 government relationships and partnerships, and the role of school feeding in the sectoral agenda of education, health, and social protection. 144. Collected evidence demonstrates how this program has attracted additional students to school and how families are at the same time economically relieved. Basic education is known to be one of the pillars of human development and a fundamental right. On this, a large body of evidence exists on the benefits of education and on the positive linkages between education and poverty reduction. If the program maintains its capacity to attract and retain children in school, then this is to be considered an important investment in human capital. The government and development partners are indeed interested in maintaining and enlarging the coverage of this program. In fact, the government has manifested a clear political endorsement of school meal programs by stating (PRSP) that in the long run these can constitute an important element of the country social protection agenda and policy support to universal education. Furthermore, also among the long-term measures of the EFMFTI is the establishment of sustainable school feeding programs. 145. At the same time, the school feeding program is still very much of a safety net program as the opportunities for education are linked to the capacity of household members to actually send their children to school. In fact, the extent to which children can benefit from education depends to a great degree on whether the child is actually enrolled and participating in school (do children have the opportunity to learn? do they have access to education?) and whether the child is able to learn what he or she is being taught in school, which means for children to be fed and rested and for their families to be economically relieved so that they are actually able to send their children to school. This program is therefore an intervention that has both the features of safety nets and of a program able to foster educational improvements. As such, the educational and socio-economic opportunities of this program call for a need to extend coverage to many more vulnerable areas and schools. 146. However done, scalability is linked closely to cost and budgetary constraints (specific costing considerations are not yet available). For the time being, the government does not have the financial means to support this line of activities. Therefore, scalability may rely on development partners such as the World Bank, other UN agencies or bilateral donors, 47 and if possible an initial contribution from the government. As far as the community-based approach, where implementation and monitoring responsibilities remain anchored in grassroots and informal community-based mechanisms, this would most likely be maintained as much as possible because it guarantees a fast start-up of activities without initial investment costs and because it sets in motion important socio-economic dynamics in the vulnerable target areas. Finally, if the community development approach is maintained and strengthened, this would imply that areas of intervention in the social protection agenda as well as in some aspects of the education agenda are still led through informal safety nets by civil society and communities. The Government of Togo would possibly agree to this for at least two good reasons. First, vision. School feeding is an investment in human capital; and the government may decide to support the success of the school feeding program and to shift responsibilities to communities and grassroots organizations with the right amount of incentives, until the government is fully capable of other expenses. Second, lack of adequate capacity. The program is successfully led at the local level with guidance from the Steering Committee but it is �outsourced� to grassroots groups and communities in terms of allowing the commitment, knowledge, and experience on the ground of the PTAs, local NGOS, and femmes-mamans to successfully run the program. Training national bodies and assigning responsibilities for implementation to central and local authorities or other formal channels requires a large amount of time and resources for capacity transfer, as well as some degree of political hassle. 48 VI. Conclusions and Recommendations First, in the context of fragility there is scope to build safety nets on existing mechanisms that are typically based on the commitment of communities as well as civil society. 147. After two years of program implementation carried out through informal community-based mechanisms, the Togo Community School Feeding program has proven to have a number of advantages over more traditional and formally centralized approaches. Given weaknesses in government capacity, tapping into local resources enabled an immediate response with minimal organizational delays, while also providing the elements to start supporting the local development processes. 148. The institutional approach underpinning the program promoted adaptability and ownership. As a locally run program, community adaptation was afforded. The community-based approach promotes program ownership and sets the basis for sustainability by involving communities at different levels of program implementation (the process of preparing menu plans; cooking and purchases; monitoring; etc). Some of the qualitative evidence pointed to the promotion of social cohesion, evidenced through the increased social interaction of women cooking, community members involved in programs or schools committees. Slowly the process has helped to establish positive linkages with the local decentralized structures that are keen in providing their support to the operation. 149. While the experiences of the school feeding program underlines the importance of informal mechanisms, the robustness of these structures as the program consolidates and scales up may be a challenge. Already in the first two years of implementation the program has focused the attention of stakeholders toward thinking through an inclusive national social protection policy. This may present challenges as policy dialogue goes forward in terms of thinking through the overall appropriateness of school feeding vis-à-vis other instruments, and the rules and responsibilities of different actors should the program move from an emergency response to a more permanent intervention. 49 Second, the appropriateness of school feeding as a safety net intervention in Togo is widely supported, given the prevalence of food insecurity and poor access to basic services. 150. The program proved quick to scale up, with a number of simple as well as innovative implementation features that could be executed with little delay (targeting, transfer delivery, procurement). First, the program was easily targeted on a geographic basis and given the high poverty context, there seemed to be few concerns of exclusion error. Delays in transfer delivery were minimal given the emphasis on local service supplies and the role of local communities both in food preparation and production. One of the innovative features of this program is the fact that meals are prepared using, as much as possible, locally-produced food. The menu is based on local tastes and seasonal availability, and it is for a large part supplied with local production (small-scale farmers living in the community or nearby villages). Local production seems sufficient to cover a large part of the food basket ingredients, for example: corn, beans, vegetables (although very limited as the food basket is not including enough fresh vegetables), meat, fish, and spices. 151. Both quantitative and qualitative evidence suggest a number of important benefits of the school feeding response, some expected and others more unintended. The most direct effect of the program has been on the enrollment and attendance of children coming to school. The program has reported good results in terms of attracting children to school and ensuring that they attend class for the full day while feeding them with a nutritious meal that they would not have otherwise at home. Data presented also indicates the unexpected and accelerated attendance by girls in rural communities. In this context, the intervention is providing an entry point to improve education access and dietary intake. As a response mechanism, it offers the possibility for future adaptation typically in the areas of access to complimentary health services (de-worming). Evidence presented also indicated the direct repercussions at the household level, including US$8-9 annual savings on food, the transfer of good practices in hygiene, and increased participation of parents in the school/community. 152. The spillover or multiplier effects in the local economy were a point of emphasis underscored in the qualitative analysis. The program implementation arrangements 50 fostered household savings, direct cash inflows into the local economy, and reinvestment of earnings into the local economy. Indirectly the program has provided a means of dependable income for nine months of the year to the parent–teacher association and femmes-mamans, as well as revenue for local traders, small-scale farmers, taxi drivers, and merchants. Both household savings and profits from the preparation of meals by the femmes-maman are being mainly re-invested in productive activities, or used to finance education and health services or to assist relatives in need. In the context of crisis-affected areas, such multiplier effects seem especially significant. The incidence of economic benefits to small-scale farmers and local traders within the program catchment areas is higher than in any centralized approach, where food may be totally imported or come from large traders. Furthermore, the school feeding program runs for a fixed number of days a year and has a predetermined food basket of locally-produced items that are purchased once a week from the nearest markets; thus it generates a stable and predictable demand of food. Based on the anecdotal information collected during the field evaluation, small traders are usually wives or relatives of farmers, and they have learned to adapt their supply for the needs of the program based on the increased demand of food. Transportation is sometimes needed to bring all food items from the market to the houses or schools. Local taxi drivers are available and earning additional funds. 153. An unresolved challenge remains the dietary composition of school meals. Meals are abundant and nutritious and children appreciate them; yet, the option of introducing a more balanced diet by introducing daily fresh fruit and vegetables and even fortifying meals has been raised. Fortification specifically may become important in view of the lack of regular and affordable presence on the local markets of fresh fruits and vegetables. Some options are being considered and they include: (i) the utilization of a local herb (fresh or dried) called “moringa,� known worldwide for its nutritional contents; it is currently being studied by various research institutions in Lomé but requires further inquiry; and (ii) pursuing possible partnerships with the World Food Program or other international or local NGOs to seek possible complementarities to the program on the activities of food fortification supply. 51 Third, the analysis identified a number of issues relevant to the future scale up and replication of the school feeding program. 154. The scalability of the program is a complex issue that leaves many open questions. The program is still relatively small in scale but has started to expand its coverage with the second phase. The pilot phase has generated good results, had no major implementation problems, and has achieved a good satisfaction level from all beneficiaries. Where a pilot has succeeded, scaling up seems to be the next logical step: In this program it will most likely be approached through geographic expansion (more communities and more schools in all regions). With this lens, the main consideration when looking at scalability is targeting. Extending geographic coverage to more schools opens up the inevitable question about who is the program really meant for, who is it supposed to benefit, does it benefit people who do not really need subsidized food, and would this go at the expenses of the poorest? 155. Scalability is also linked closely to cost and budgetary constraints. For the time being, the government does not have the financial means to support this line of activities. Therefore, scalability may rely on development partners such as the World Bank, other UN agencies or bilateral donors, and if possible an initial contribution from the government. Moreover, the scalability of the program may entail a transformation in the original design as needs and stakeholders may change. For example, the community school feeding program in Togo was conceived as a safety net during an emergency and food crisis and according to collected evidence, has fulfilled its role well. Now that the emergency is over, that the government is defining its political agenda for development and future commitments, and that the school feeding program aims to extend its coverage, the discussion necessarily shifts to the appropriateness of school feeding vis-à-vis other interventions, especially in non-rural areas. 156. The program represents a relevant reliable and predictable safety net. This is very important in a country defined as fragile because of its institutional weaknesses and its crisis and post-crisis situation. The program runs every day of the school year for all students of the most disadvantaged schools in the crisis affected areas. That is, households can count on the savings from meal preparation for an entire school year, allowing them to better plan 52 resources within their own livelihood strategies and coping mechanisms. In addition, for some members of the communities (femmes-mamans, local market traders, and small farmers supplying food for the school feeding purpose), the program represents a reliable source of income. 157. Based on the evidence collected and the experience in Togo, the school meal program is likely to be replicable in other food emergency contexts provided that there is an enabling environment and that at least three criteria are met: availability of community social capital, community cohesion and organization, some local food production. Operational Recommendations Going Forward 158. There is a need to strengthen basic monitoring and evaluation mechanisms including the monitoring of meal quality and of the consistency in meals composition as according to guidelines. 159. The operational guidelines could introduce mechanisms to support improved accounting and payment procedures with the routine use of budget sheets for femmes- mamans. This may require some initial training. 160. The program could benefit from a review of costs and comparisons with other similar programs in the regions, specifically those in other fragile contexts where operating considerations will be similar. 161. Partnership opportunities could be explored to create additional synergies and maximize results in the areas of education, health, agricultural development, and governance. 162. The program could mobilize greater expertise from nutrition specialists to consider the dietary and nutritional aspects. Of specific interest is food fortification. 163. From an evaluation purpose, several areas remain to be further investigated, including the balance between direct and indirect interventions, the effect of school feeding on educational outcomes, the impacts of the innovation designs introduced in the program like local food procurement. 53 164. In the context of any future scale up, the program should consider different modalities and scenarios in program development. This would address adaptations in implementation arrangements; the degree to which local implementation mechanisms would be featured in future roll out; and, critically, different financing modalities that would most likely include external financing. 54 Annex 1. Yearly Costs of School Feeding Program in TOGO (US$)34 YEAR 1 (US$) Number Total Meal Remuneration Food NGO Costs x Program Total of Costs (I) of Femmes- Costs Service Coordination Program Schools Mamans (II) (I-II) Costs (III) (IV) Cost (I-IV) Maritime 20 173,600 41,667 131,933 17,360 8,680 199,640 Plateaux 19 164,920 39,583 125,337 16,492 8,246 189,658 Centrale 13 112,840 27,083 85,757 11,284 5,642 129,766 Kara 15 130,200 31,250 98,950 13,020 6,510 149,730 Savanes 17 147,560 35,417 112,143 14,756 7,378 169,694 Total 84 729,120 175,000 554,120 72,912 36,456 838,488 Percent of total 87% 21% 66% 9% 4% 100% Per day/child 0.31 0.07 0.24 0.03 0.02 0.36 Per year/child 43 10 33 4 2 50 Assumptions Number of women 500 Number of children 16,800 Number of school days 140 Women profit/day US$1-4 (max-min) Women’s average US$2.5 profit/day Schools 84 Cost/child/meal 0.31 NGO services 10% Program coordination 5% Average number children/school 200 34 Author’s calculations based on data from Agbo E., 2010b in Evaluation de l’opération de fourniture de répas aux écoliers des zones vulnerables du Togo, Analyse quantitative. 55 YEAR 2 (US$) Number Total Meal Remuneration Food NGO Costs x Program Total of Preparation of Femmes- Costs Service Coordination Program Schools Costs (I) Mamans (II) (I-II) Costs (III) (IV) Cost (I-IV) Maritime 22 280,213 45,833 234,380 28,021 14,011 322,245 Plateaux 21 267,476 43,750 223,726 26,748 13,374 307,598 Centrale 14 178,317 29,167 149,151 17,832 8,916 205,065 Kara 16 203,791 33,333 170,458 20,379 10,190 234,360 Savanes 19 242,002 39,583 202,419 24,200 12,100 278,303 Total 92 1,171,800 270,000 901,800 117,180 58,590 1,347,570 Percent of total 140% 32% 108% 14% 7% 161% Per day/child 0.31 0.07 0.24 0.03 0.02 0.36 Per year/child 56 13 43 6 3 80 Assumptions Number of women 600 Number of children 21,000 Number of school days 180 Women profit/day US$1-4 (max-min) Women’s average US$2.5 profit/day Schools 92 Cost/child/meal 0.31 NGO services 10% Program coordination 5% Average number children/school 228 Source: Author (2011). 56 Annex 2. Model for Standardization of School Feeding Costs 165. In 2009, the World Food Program (WFP) estimated the costs of WFP and government-run school feeding programs in Africa and compared these costs using standardized measures. The goal was to estimate all the costs of school feeding programs in Africa, including costs to the WFP, governments, and communities. Kenya, Malawi, Lesotho, and The Gambia were selected for this analysis because their meals were similarly composed of cooked meals. All costs were provided by the World Food Program and by stakeholders in the Ministries of Education and in the communities. Lesotho has two school feeding programs: One is located in the mountainous areas and run by the WFP; the other is located in the low lands and run by the government. The cost estimates are the country-wide average costs of these two programs. The Togo example has been added in this paper by the author using the same parameters for calculations, and using the existing costing information. 166. In terms of the caloric content of meals, many WFP and government school feeding programs provide roughly one-third of the energy requirements for school-age children, as evidenced by the programs in Kenya and Lesotho where meals range between 500 and 700Kcal per day. The Togo program offers meals that account for an average of 1,200 Kcal. 167. To compare across the countries, costs were standardized with a model counting days fed and energy provided by using a 200-day school year and a 700-Kcal per day ration. In order to accurately calculate costs, the actual number of days of school feeding was provided. This number may be different from the planned number of school feeding days due to pipeline breaks, which occur when food is not delivered because of problems with supply or climate. Costs were calculated using following equation: Cs = Cpr x [200/Dsf] x [700/Kcal] 57 Where: Cs = Cost per beneficiary Cpr = Actual cost per beneficiary using total project expenditure Dsf = Number of days fed Kcal = Planned ration in kilocalories. 58 VII. References Agbo, E., A. Yaovi, and A. Kokoul. 2010a. Evaluation de l’opération de fourniture de répas aux écoliers des zones vulnérables du Togo (Analyse quantitative). Mimeo. Agbo,E. 2010b. Evaluation de l’opération de fourniture de répas aux écoliers des zones vulnérables du Togo (Analyse qualitative). Mimeo. Binswanger, H. and S. Aiyar. 2003. Scaling Up Community-Driven Development. Theoretical Underpinnings and Program Design Implications. World Bank Policy Research Working Paper No. 3039. Washington DC: World Bank. Brinkerhoff, D. 2007. Dilemmas and Directions, Capacity.org, Issue 32, 2007 “Fragile States.� Bundy, D., C. Burbano, M. Grosh, A. Gelli, M. Jukes, and L. Drake. 2009. Rethinking School Feeding: Social Safety Nets, Child Development, and the Education Sector. Washington, DC: World Bank. Economist Intelligence Unit. 2011. http://country.eiu.com/Togo. Espejo, F., C. Burbano, and E. Galliano. 2009. Home Grown School Feeding, A Framework to Link School Feeding and Agricultural Production. New York: World Food Program. Galloway, R., E. Kristjansson, A. Gelli, U. Meir, F. Espejo, and D. Bundy. “School Feeding: Outcomes and Costs.� Food and Nutrition Bulletin 30(2):2009. Grosh, M., C. del Ninno, E. Tesliuc, and A. Ouerghi. 2008. For Protection and Promotion, The Design and Implementation of Effective Safety Nets. Washington, DC: World Bank. International Food Policy Institute. 2004. Agricultural Science and Technology Indicators. ASTI Country Brief No. 16 . Washington DC: IFPRI. http://www.ifpri.org/sites/default/files/publications/Togo_CB16.pdf International Monetary Fund. 2010. Togo: Poverty Reduction Strategy Paper (2009-11). IMF Country Report No. 10/33. Washington DC: IMF. Revenga, A. Rising Food Prices: Policy Options and World Bank Response. Note prepared for the Development Committee Meeting, April 2008, World Bank, Washington, DC. Slate, R. and A. McCord. 2009. Social Protection, Rural Development and Food Security: Issues Paper on the Role of Social Protection in Rural Development. London: Overseas Development Institute. United Nations Children’s Fund. 2009. Analyse de la situation de l’enfant au Togo. New York: UN. ———. 2008. UNICEF Executive Director’s Togo Visit Focuses on Education. Press Release of September 2008. New York: UN. 59 ———. 2007. Revised Country Program Document for Togo (2008-2012). New York: United Nations. http://www.unicef.org/about/execboard/files/07-PL43-Togo(1).pdf ———. Multiple Indicator Country Survey (MICS). New York: UN. www.unicef.org/wcaro/WCARO_Togo_Factsheet_General.pdf U.S. Agency for International Development. Speaker’s Corner: Protecting Livelihoods in Conflict and Fragile States. An online forum January 22-24, 2008 hosted and moderated by Dr. Laura Hammond, Washington DC. http://www.povertyfrontiers.org/ev_en.php?ID=2105_201&ID2=DO_TOPIC World Bank. 2011. Togo Poverty Reduction Strategy Paper 2009-2011. Washington, DC: World Bank. ———. 2011. EdStats 2010. http://web.worldbank.org/WBSITE/EXTERNAL/TOPICS/EXTEDUCATION/EXTDATAS TATISTICS/EXTEDSTATS/0,,contentMDK:22614780~menuPK:7196605~pagePK:64168445~piPK :64168309~theSitePK:3232764,00.html?TGO,206 ———. 2010a. Analyse du context stratégique de mise en oeuvre des activités de protection sociale/filets sociaux de sécurité au Togo, draft June 2010. ———. 2010b. Togo Community-Driven Development Project, Paper No. 54513-TG. Washington, DC: World Bank. ———. 2008. Togo Community-Driven Development Project, Paper No. 45237-TG. Washington, DC: World Bank. ———. 2010. World Food Program. Togo - Enquête rapide sur la sécurité alimentaire des ménages, April 2010, Rome 60 Social Protection Discussion Paper Series Titles No. Title 1117 Social Safety Nets in Fragile States: A Community-Based School Feeding Program in Togo, August 2011 (also available in French) 1116 Strengthening Governance of Social Safety Nets in East Asia by Sara Giannozzi and Asmeen Khan, August 2011 (online only) 1115 International Portability of Health-Cost Coverage: Concepts and Experience by Martin Werding and Stuart McLennan, July 2011 (online only) 1114 Liberia’s Cash For Work Temporary Employment Project: Responding to Crisis in Low Income, Fragile Countries by Colin Andrews, Prospère Backiny-Yetna, Emily Garin, Emily Weedon, Quentin Wodon and Giuseppe Zampaglione, July 2011 1113 Employability and Productivity among Older Workers: A Policy Framework and Evidence from Latin America by Edmundo Murrugarra, July 2011 (online only) 1112 Cash Transfers, Children and the Crisis: Protecting Current and Future Investments by Ariel Fiszbein, Dena Ringold, Santhosh Srinivasan, June 2011 (online only) 1111 Severance Pay Programs around the World: History, Rationale, Status, and Reforms by Robert Holzmann, Yann Pouget, Milan Vodopivec and Michael Weber, May 2011 (online only) 1110 Portability of Pension, Health, and other Social Benefits: Facts, Concepts, Issues by Robert Holzmann and Johannes Koettl, May 2011 (online only) 1109 Disability and Poverty in Developing Countries: A Snapshot from the World Health Survey by Sophie Mitra, Aleksandra Posarac and Brandon Vick, April 2011 1108 Advancing Adult Learning in Eastern Europe and Central Asia by Christian Bodewig and Sarojini Hirshleifer, April 2011 (online only) 1107 Results Readiness in Social Protection & Labor Operations by Laura Rawlings, Maddalena Honorati, Gloria Rubio and Julie Van Domelen, February 2011 1106 Results Readiness in Social Protection & Labor Operations: Technical Guidance Notes for Social Service Delivery Projects by Julie Van Domelen, February 2011 1105 Results Readiness in Social Protection & Labor Operations: Technical Guidance Notes for Social Safety Nets Task Teams by Gloria Rubio, February 2011 1104 Results Readiness in Social Protection & Labor Operations: Technical Guidance Notes for Social Funds Task Teams by Julie Van Domelen, February 2011 1103 Results Readiness in Social Protection & Labor Operations: Technical Guidance Notes for Labor Markets Task Teams by Maddalena Honorati, February 2011 1102 Natural Disasters: What is the Role for Social Safety Nets? by Larissa Pelham, Edward Clay and Tim Braunholz, February 2011 1101 North-South Knowledge Sharing on Incentive-based Conditional Cash Transfer Programs by Lawrence Aber and Laura B. Rawlings, January 2011 1008 Social Policy, Perceptions and the Press: An Analysis of the Media’s Treatment of Conditional Cash Transfers in Brazil by Kathy Lindert and Vanina Vincensini, December 2010 (online only) 1007 Bringing Financial Literacy and Education to Low and Middle Income Countries: The Need to Review, Adjust, and Extend Current Wisdom by Robert Holzmann, July 2010 (online only) 1006 Key Characteristics of Employment Regulation in the Middle East and North Africa by Diego F. Angel-Urdinola and Arvo Kuddo with support from Kimie Tanabe and May Wazzan, July 2010 (online only) 1005 Non-Public Provision of Active Labor Market Programs in Arab-Mediterranean Countries: An Inventory of Youth Programs by Diego F. Angel-Urdinola, Amina Semlali and Stefanie Brodmann, July 2010 (online only) 1004 The Investment in Job Training: Why Are SMEs Lagging So Much Behind? by Rita K. Almeida and Reyes Aterido, May 2010 (online only) 1003 Disability and International Cooperation and Development: A Review of Policies and Practices by Janet Lord, Aleksandra Posarac, Marco Nicoli, Karen Peffley, Charlotte McClain-Nhlapo and Mary Keogh, May 2010 1002 Toolkit on Tackling Error, Fraud and Corruption in Social Protection Programs by Christian van Stolk and Emil D. Tesliuc, March 2010 (online only) 1001 Labor Market Policy Research for Developing Countries: Recent Examples from the Literature - What do We Know and What should We Know? by Maria Laura Sanchez Puerta, January 2010 (online only) 0931 The Korean Case Study: Past Experience and New Trends in Training Policies by Young-Sun Ra and Kyung Woo Shim, December 2009 (online only) 0930 Migration Pressures and Immigration Policies: New Evidence on the Selection of Migrants by Johanna Avato, December 2009 (online only) 0929 Ex-Ante Methods to Assess the Impact of Social Insurance Policies on Labor Supply with an Application to Brazil by David A. Robalino, Eduardo Zylberstajn, Helio Zylberstajn and Luis Eduardo Afonso, December 2009 (online only) 0928 Rethinking Survivor Benefits by Estelle James, December 2009 (online only) 0927 How Much Do Latin American Pension Programs Promise to Pay Back? by Alvaro Forteza and Guzmán Ourens, December 2009 (online only) 0926 Work Histories and Pension Entitlements in Argentina, Chile and Uruguay by Alvaro Forteza, Ignacio Apella, Eduardo Fajnzylber, Carlos Grushka, Ianina Rossi and Graciela Sanroman, December 2009 (online only) 0925 Indexing Pensions by John Piggott and Renuka Sane, December 2009 (online only) 0924 Towards Comprehensive Training by Jean Fares and Olga Susana Puerto, November 2009 0923 Pre-Employment Skills Development Strategies in the OECD by Yoo Jeung Joy Nam, November 2009 0922 A Review of National Training Funds by Richard Johanson, November 2009 0921 Pre-Employment Vocational Education and Training in Korea by ChangKyun Chae and Jaeho Chung, November 2009 0920 Labor Laws in Eastern European and Central Asian Countries: Minimum Norms and Practices by Arvo Kuddo, November 2009 (online only) 0919 Openness and Technological Innovation in East Asia: Have They Increased the Demand for Skills? by Rita K. Almeida, October 2009 (online only) 0918 Employment Services and Active Labor Market Programs in Eastern European and Central Asian Countries by Arvo Kuddo, October 2009 (online only) 0917 Productivity Increases in SMEs: With Special Emphasis on In-Service Training of Workers in Korea by Kye Woo Lee, October 2009 (online only) 0916 Firing Cost and Firm Size: A Study of Sri Lanka's Severance Pay System by Babatunde Abidoye, Peter F. Orazem and Milan Vodopivec, September 2009 (online only) 0915 Personal Opinions about the Social Security System and Informal Employment: Evidence from Bulgaria by Valeria Perotti and Maria Laura Sánchez Puerta, September 2009 0914 Building a Targeting System for Bangladesh based on Proxy Means Testing by Iffath A. Sharif, August 2009 (online only) 0913 Savings for Unemployment in Good or Bad Times: Options for Developing Countries by David Robalino, Milan Vodopivec and András Bodor, August 2009 (online only) 0912 Social Protection for Migrants from the Pacific Islands in Australia and New Zealand by Geoff Woolford, May 2009 (online only) 0911 Human Trafficking, Modern Day Slavery, and Economic Exploitation by Johannes Koettl, May 2009 0910 Unemployment Insurance Savings Accounts in Latin America: Overview and Assessment by Ana M. Ferrer and W. Craig Riddell, June 2009 (online only) 0909 Definitions, Good Practices, and Global Estimates on the Status of Social Protection for International Migrants by Johanna Avato, Johannes Koettl, and Rachel Sabates-Wheeler, May 2009 (online only) 0908 Regional Overview of Social Protection for Non-Citizens in the Southern African Development Community (SADC) by Marius Olivier, May 2009 (online only) 0907 Introducing Unemployment Insurance to Developing Countries by Milan Vodopivec, May 2009 (online only) 0906 Social Protection for Refugees and Asylum Seekers in the Southern Africa Development Community (SADC) by Mpho Makhema, April 2009 (online only) 0905 How to Make Public Works Work: A Review of the Experiences by Carlo del Ninno, Kalanidhi Subbarao and Annamaria Milazzo, May 2009 (online only) 0904 Slavery and Human Trafficking: International Law and the Role of the World Bank by María Fernanda Perez Solla, April 2009 (online only) 0903 Pension Systems for the Informal Sector in Asia edited by Landis MacKellar, March 2009 (online only) 0902 Structural Educational Reform: Evidence from a Teacher’s Displacement Program in Armenia by Arvo Kuddo, January 2009 (online only) 0901 Non-performance of the Severance Pay Program in Slovenia by Milan Vodopivec, Lilijana Madzar, Primož Dolenc, January 2009 (online only) To view Social Protection Discussion papers published prior to 2009, please visit www.worldbank.org/sp. Abstract This paper reviews a small community-based school feeding program launched in Togo in response to the 2007/08 food price crisis. The discussion focuses on the operational and policy lessons emerging from the program, to better understand opportunities for scale up and sustainability in the future. A focus of the discussion is how to build safety nets in fragile states and in situations where there is weak and fragmented government capacity to deliver services to disadvantaged and vulnerable communities. In this context school feeding is explored as an entry point through the use of informal mechanisms based on the commitment of communities and civil society. The analysis is premised on quantitative and qualitative analysis carried out at program sites. The discussion identifies the operational challenges and opportunities in customizing school feeding within Togo with an emphasis on targeting, cost effectiveness, procurement and institutional aspects. Evidence on the economic and social benefits of the program is also presented, focusing on dietary impacts, as well as household and local community effects. The objective of the discussion is to share lessons learned from evaluation findings so that they can be useful for implementing similar programs in the future in Togo itself or in other countries. Findings from the analysis highlight the possibilities of implementing school feeding in a low capacity setting and the scope for using the program as a springboard towards a broader and more comprehensive social safety net. HUMAN DEVELOPMENT NETWORK About this series... Social Protection Discussion Papers are published to communicate the results of The World Bank’s work to the development community with the least possible delay. The typescript manuscript of this paper therefore has not been prepared in accordance with the procedures appropriate to formally edited texts. The findings, interpretations, and conclusions expressed herein are those of the author(s), and do not necessarily reflect the views of the International Bank for Reconstruction and Development / The World Bank and its affiliated organizations, or those of the Executive Directors of The World Bank or the governments they represent. The World Bank does not guarantee the accuracy of the data included in this work. For more information, please contact the Social Protection Advisory Service, The World Bank, 1818 H Street, N.W., MSN G7-703, Washington, D.C. 20433 USA. Telephone: (202) 458-5267, Fax: (202) 614-0471, E-mail: socialprotection@worldbank.org or visit us online at www.worldbank.org/sp.