88999 DISCUSSION PAPER NO. 1409 Lesotho A Safety Net to End Extreme Poverty W. James Smith, Emma Mistiaen, Melis Guven and Morabo Morojele Africa Social Safety Net and Social Protection Assessment Series June 2013 LESOTHO A Safety Net to End Extreme Poverty W. James Smith, Emma Mistiaen, Melis Guven and Morabo Morojele June 2013 Africa Social Safety Net and Social Protection Assessment Series Until recently, most countries in Africa implemented safety nets and social protection programs only on an ad hoc basis. In the wake of the global economic, food and fuel price crises starting in 2008, however, policymakers in Africa began to increasingly view safety nets as core instruments for reducing poverty, addressing inequality, and helping poor and vulnerable households to manage risk more effectively. During FY2009-2013, to support governments in their quest to understand better how to improve the efficiency and effectiveness of safety nets in their countries, the World Bank’s Africa Region undertook social safety net or social protection assessments in a number of countries in Sub-Saharan Africa. By 2014 assessments have been completed or are under preparation for over 25 countries in sub-Saharan Africa. These assessments analyze the status of social protection programs and safety nets, their strengths and weaknesses and identify areas for improvement, all with the aim of helping governments and donors to strengthen African safety net systems and social protection programs to protect and promote poor and vulnerable people. They were all carried-out with the explicit aim of informing governments’ social protection policies and programs. With the results of analytical work like these assessments and other types of support, safety nets and social protection programs are rapidly changing across Africa. For a cross-country regional review, please see "Reducing Poverty and Investing in People: The New Role of Safety Nets in Africa," which pulls together the findings and lessons learned from these assessments and other recent studies of safety net programs in Africa. Abstract This report shows that while more inclusive growth is the ultimate solution to poverty in Lesotho, the country can and should use selective social transfers to reduce poverty more rapidly among the extreme poor. But because the majority of the transfers are received by people who are not among the extreme poor there is room for increasing the efficiency and effectiveness of spending on safety nets which. These programs should be productive and concentrate on the extreme poor Basotho. In the long run, Lesotho should move towards a more consolidated safety net and strengthen existing programs, such as the Child Grants Program, that already provide some important and positive outcomes and enjoy strong popular and political support. JEL Classification: I32, I38, J32, H53 Key Words: social protection, systems, safety nets, social assistance, welfare, administration, public policy, public sector reform, developing countries. GOVERNMENT FISCAL YEAR April 1 - to – March 31 CURRENCY EQUIVALENTS Currency Unit : Maloti 1 US$ : M 7.5 WEIGHTS AND MEASURES Metric System ABBREVIATIONS AND ACRONYMS AE Adult Equivalent BoS Bureau of Statistics CC Community Council CCTs Conditional Cash Transfers CGP Child Grants Program CWMS Continuous Welfare Monitoring Survey DHS Demographic and Health Survey FAO Food and Agricultural Organization FNCO Food and Nutrition Coordination Office GDP Gross Domestic Product HIES Household Income and Expenditure Survey HH Households LVAC Lesotho Vulnerability Assessment Committee M. Maloti MAFS Ministry of Agriculture and Food Security MHSW Ministry of Health and Social Welfare MoDP Ministry of Development Planning MoET Ministry of Education and Training MoF Ministry of Finance MoFDP Ministry of Finance and Development Planning i MSD Ministry of Social Development NHDR National Human Development Report NISSA National Information System for Social Assistance NMDS National Manpower Development Secretariat NSDP National Strategic Development Plan OAP Old Age Pension OPM Oxford Policy Management OVC Orphans and Vulnerable Children PWP Public Works Program SACU Southern African Customs Union SFP School Feeding Program UNDP United Nations Development Program UNAIDS United Nations Program on HIV/AIDS UNICEF United Nations Fund for Children WFP World Food Programme ii ACKNOWLEDGEMENTS This study benefitted from the input and advice provided by many people in Lesotho, including Moepi Sematlane, Malefetsane Masasa, Rapelang Ramoea, Limakasto Chisepo, Selloane Qhobela, Malehloa Molato, Tsele Moloeli, Mkwetsa Tsitsi-Mphuthing, Rui Possolo, Mofihli Motsetsero, Thabo Sophonea,Victoria Diarra, Mohammad Farooq, Lechoko Noko, Maope, Nkhala Sefako, Mopelasa Moshokho, Thaba Tekla, Thuto Mokhehle, Victoria Arboleda, Thabo Ramochela, Brenda Yamba, Farida Noureddine, Habofanoe Matsipa, Douglas Gumbo, Napo Ntlou, Bile Khalif, Mariam Homayoun, Alka Bhatia, Mating Mahooana, Lerato Sello, Tsepang Linko, Matumelo Monoko, Russel Armstrong, Lebakiso Mathlo, and Matsepo Moletsane. The report was prepared by Emma Mistiaen and Melis Guven (Task Team Leaders), W. James Smith (Lead Consultant and principal author) and Morabo Morojele (Consultant). The team benefitted from peer reviews and advice from Margaret Grosh and Damien de Walque. Within the World Bank, valuable suggestions and advice were provided by Lynne Sherburne-Benz (Sector Manager), Anush Bezhanyan, Macmillan Anyanwu, Ashish Narayan, Christian Gonzales, Maniza Naqvi, Victoria Monchuck, Christina Panasco Santos, Louise Fox, and Frauke Jungbluth. Melanie Allwine, Shinya Takamatsu, and Nobuo Yoshida processed the 2010 household survey data used for part of the poverty analysis in Chapter 2, Elina Sheja undertook a background review of poverty in Lesotho, and Yasmeen Tabbaa provided valuable input to the report. Fiona Macintosh edited the report and Ana Lukau processed the report in Washington D.C. Finally, the team wishes to express its thanks for the unstinting support provided by the World Bank Office in Lesotho. iii Table of Contents EXECUTIVE SUMMARY ............................................................................................................. vii INTRODUCTION ......................................................................................................................... 1 CHAPTER I: COUNTRY AND FISCAL CONTEXT ........................................................................... 3 CHAPTER II: POVERTY AND SAFETY NETS ............................................................................... 14 2.1 Poverty Profile .......................................................................................................... 16 2.2 Basic Characteristics of the Poor and the Poorest ................................................... 19 2.3 The Geographic Dimensions of Poverty ................................................................... 21 2.4 Livelihoods of the Poor ............................................................................................ 23 2.5 Food Insecurity ......................................................................................................... 27 2.6 Health, Nutrition, and Education ............................................................................. 28 2.7 Some Specific Groups of the Poor............................................................................ 31 CHAPTER III: EXISTING TRANSFER PROGRAMS ....................................................................... 42 1. Overview ...................................................................................................................... 42 2. Program Descriptions .................................................................................................. 42 CHAPTER IV: INSTITUTIONAL ISSUES, PROGRAM DELIVERY, AND TARGETING...................... 92 1. Current National Social Protection Policy and Institutions ......................................... 92 2. Program Delivery and Capacity.................................................................................... 93 3. Targeting Issues and Choices ....................................................................................... 99 CHAPTER V: CONCLUSIONS AND RECOMMENDATIONS ...................................................... 106 1. The Role of Safety Nets and Transfers in National Development Strategy ............... 106 2. The Way Forward – Towards a Unified National Program. ....................................... 110 3. Short- to Medium-term Recommendations for Improving Existing Programs ......... 119 ANNEXES ............................................................................................................................... 130 Annex 1: Geographical Distribution of Poverty and Assets of the Poor ............................... 130 Annex 2: Health and Education Indicators ........................................................................... 133 Annex 3: Beneficiaries and Expenditure of Public Assistance .............................................. 135 Annex 4: Distribution of Fertilizer Subsidy ........................................................................... 137 Annex 5: School Meal Composition ...................................................................................... 138 Annex 6: Generosity of Cash Transfer Programs in Various African Countries .................... 139 MAP 33434 ............................................................................................................................ 141 iv Tables Table 1: Revenue and Expenditure, 2011/12 ........................................................................... 9 Table 2: Trends in Spending on Major Transfer Programs (M. millions) ................................ 12 Table 3: Social Safety Net Expenditures in Selected African Countries .................................. 13 Table 4: Lesotho - Estimated Poverty Lines 2010/11 ............................................................. 16 Table 5: Estimated Number of Poor People in Lesotho.......................................................... 17 Table 6: Depth of Poverty 2002/03 and 2010/11 ................................................................... 18 Table 7: Demographic Characteristics of the Poor, 2010/11 ................................................. 20 Table 8: Poverty Indicators by Geographic Zone .................................................................... 21 Table 9: Main Sources of Income by Poverty Status .............................................................. 24 Table 10: Calories from Own-Production by Wealth and Location, 2004 .............................. 28 Table 11: Nutritional Status of Children by Wealth Quintile, 2009 ........................................ 29 Table 12: Orphanhood by Wealth Quintile............................................................................. 32 Table 13: Seasonality in Rural Agriculture, Food Security, and Poverty................................. 38 Table 14: Summary of Main Existing Transfer Programs in Lesotho ...................................... 43 Table 15: Approximate Share of Spending of the SFP Going to Poor and the /Non-poor ..... 58 Table 16: Nutrition Support Program, 2012 Coverage ........................................................... 63 Table 17: Estimated Cost for Planned Beneficiaries and Cost per Meal ................................ 65 Table 18: Agricultural Fair, Subsidized, and Retail Prices for Selected Agricultural Inputs .... 76 Table 19: Impact of the AIF Program on Household Maize Production ................................. 79 Table 20: Distributional Impact of the Fertilizer Subsidy Scheme, Some Illustrative Examples ................................................................................................................................................. 81 Table 21: Coverage of Existing Transfer Programs ................................................................. 88 Table 22: Benefit Levels of Existing Programs ........................................................................ 90 Table 23: Costs and Cost-effectiveness of Current Programs ................................................ 91 Table 24: Benefits and Drawbacks of Universal and Targeted Programs............................... 99 Table 25: Some Illustrative Transfer Amounts (2011 Maloti per adult equivalent per month) ............................................................................................................................................... 112 Table 26: Possible Scale and Costs of a National Safety Net ................................................ 114 Table 27: Key Observations and Recommendations for Each Program ............................... 120 v Figures Figure 1: SACU Revenue, 2000/2001 to 2012/2013 (M. millions).......................................... 10 Figure 2: Estimated Distribution of Consumption in Lesotho, 2010/11 ................................. 18 Figure 3: Map 1-Lesotho’s Ecological Zones ........................................................................... 23 Figure 4: Composition of Farm Household Income ................................................................ 25 Figure 5: HIV Prevalence among 15-49 Year Olds by Wealth Quintile, 2009 ......................... 33 Figure 6: Distribution of Poverty by Age Group (% of Total Poor) ......................................... 34 Figure 7: Existing Programs, Estimated Coverage of the Very Poor ....................................... 89 Boxes Box 1: Some Key Productive Benefits of Safety Nets ............................................................... 2 Box 2: Main Data Sources and Caveats..................................................................................... 4 Box 3: Cost and Effectiveness of School Feeding Programs ................................................... 62 Box 4: Using Cash Transfers to Achieve Better Nutrition Outcomes in Peru ......................... 66 Box 5: Public Works to Support Rural Development - An Example from Ethiopia ................. 72 vi EXECUTIVE SUMMARY Main Messages  While more inclusive growth is the ultimate solution to poverty in Lesotho, the country can and should use selective social transfers to reduce poverty more rapidly among the extreme poor .  Lesotho is already spending a lot on transfer programs (about US$197 million per annum or 9 percent of GDP), mostly on the universal Old Age Pension, school feeding, and tertiary bursaries. While these programs yield positive outcomes and enjoy strong popular and political support, the majority of the transfers are received by people who are not among the extreme poor.  There is room for increasing the efficiency and effectiveness of spending by: (i) coordinating or consolidating programs working towards the same objective; (ii) making adjustments to the design and implementation of programs; and (iii) removing those programs that are not contributing to the overall objective of the safety net.  The safety net should concentrate on the extreme poor Basotho – those who do not have enough to feed themselves. At a rough estimate, this means about 40,000 to 80,000 households or 10 to 20 percent of the population.  The safety net should also be more promotive (to help the poor to rise out of long-term poverty by linking them with services to build their human capital and improve their nutrition) and more productive (for example, by resulting in better maintained roads or greater agricultural productivity).  In the long run, Lesotho should move towards a more consolidated safety net that might consist of a combination of: (i) regular monthly cash transfers to the truly destitute and to those in some categorical groups (such as child-headed and grandparent-headed households) and (ii) public employment opportunities and agricultural productivity-linked transfers for the able-bodied extreme poor.  A safety net along these lines would cost about M.410 million (US$55 million) per annum or about 4 percent of current public spending. However, the government can use existing spending to pay for such a safety net (for example, road maintenance expenditure could finance the employment of the poor and savings could be made on the school feeding program and other large untargeted transfers such as the general agricultural input subsidy and tertiary bursary schemes).  There are existing programs that could be used as the building blocks of a national safety net . The Child Grants Program could provide the basis for the proposed cash transfer scheme for the destitute, and the Watershed Management Program could provide a framework for a public works program that would employ the very poor on re-forestation, environmental works, and rural road maintenance.  The proposed National Information System for Social Assistance (NISSA) has the potential to increase coordination and avoid overlaps between existing programs by providing the basis for a unified registry and/or a common targeting system for a unified national program.  The safety net should be adjusted to enable the government to respond better to future crises and to be used as a counter-cyclical instrument. vii 1. The objective of this study is to help the government to decide what role safety net and transfer programs should play in the coming 5 to 10 years. It seeks to answer three questions: (i) Can increased spending on transfers accelerate poverty reduction in the medium to long term? (ii) Which groups and aspects of poverty would it make sense to target with transfers? and (iii) Which programs would have the greatest impact at an affordable cost? 2. Lesotho has experienced significant economic growth during the last two decades, but deep poverty remains. Between 2001 and 2010, Lesotho’s per capita GDP grew at an annual rate of 2.6 percent. The engines of this growth have been textile exports, mining, and public investment. However, agriculture, the sector in which most of the poor make their livelihoods, lagged behind, and a substantial minority of households, mostly in rural areas, have not benefitted from this growth. As of 2010 an estimated 57 percent of households lived below a Basic Needs poverty line equivalent to just over US$1 per day, and 34 percent lived below the food poverty line.1 3. Income distribution is extremely unequal in Lesotho, which had a Gini coefficient of 0.53 in 2010, one of the worst in the world. Inequality is so high partly because of the existence of two very different economies: one that consists of those who participate in the cash economy through wage employment, outmigration, and remittances and another that consists of those who do not. 4. Even though Lesotho has enjoyed rapid economic growth in the past decades, the overall increase in welfare has not been sufficient to significantly raise the consumption of the extreme poor.2 This is partly because the poorest have not participated in growth (due to their locations, skill levels, and lack of assets) and partly because many of them were 1 Preliminary government estimates based on the 2010/11 HIES (see Box 1 for a note on data sources). 2 The terms “extreme poor” and “extreme poverty” are used throughout to refer to those who live below the food poverty line. viii initially too far below the poverty line to escape poverty even if they did receive their proportional share of nationwide growth. In order for growth to lead to a substantial decrease in poverty in the future, the benefits will need to be more widely shared and more heavily skewed towards the poor. While the ultimate solution to poverty in Lesotho is more inclusive growth, the failure of growth to reach the poorest, combined with Lesotho’s very high inequality, suggests that the time is also right to consider new approaches, including increasing the role played by selective transfers to the poorest to raise their productivity and increase their consumption. 5. Experience in many countries in Africa and Latin America in recent years has shown that well-designed safety net programs can significantly decrease the rate of extreme poverty in ways that other public expenditures do not. They can do so because they immediately raise the consumption of the poorest while also increasing their long-term productivity by building their human capital and linking them with the productive economy. 6. Lesotho faces serious fiscal challenges as a result of very high public expenditure (which had reached an unsustainable 67 percent of GDP in 2009/10) and of excessive reliability on Southern African Customs Union (SACU) receipts, which are volatile, and make budget management challenging. The government’s strategy in response has been to: (i) contain recurrent spending; (ii) reduce the burden of the wage bill over time; and (iii) increase the share of investment in the budget while at the same time improve the quality of spending on investments. In this environment, safety net transfers will have to be particularly effective to be justified. However, Lesotho currently spends approximately 16 percent of its public expenditure on various transfer programs, most of it on large-scale programs (such as agricultural subsidies, tertiary bursaries, and school feeding) that mostly do not benefit the poorest. By our estimate, less than 25 percent of the US$197 million spent annually on transfers go to the very poor. ix Poverty and Safety Nets 7. Within the 57 percent of the population who fell below the basic needs poverty line in 2010, there is a need to distinguish those people who are truly destitute and in need of regular cash transfers, those who need cyclical assistance, and those who can be raised out of poverty with other forms of access and assistance. This is not easy in rural Lesotho for which little information exists on household incomes and where there is often very little to distinguish the poorest from the poor in general. 8. There are few demographic or other characteristics that differentiate the extreme poor from other poor households. While poor households tend to be larger than non-poor ones and to contain slightly more elderly people and children, there is not much difference between the poor and extreme poor. For example, the proportion of female-headed households is no greater among the very poor than among the population as a whole. There are surprisingly few differences in the ownership of key assets between the poor and the non-poor, except for obvious high-end items such as cars and electrical goods. Groups that might be considered to be disadvantaged such as orphans and vulnerable children (OVCs), people living with HIV/AIDS (PLWH), the elderly, and the disabled are not disproportionately poor (many live in households that can support them), so special targeting methods would need to be used to identify which individuals within these groups are truly destitute. 9. Poverty in Lesotho is concentrated in rural areas, where extreme poverty rates are more than double those in urban areas.3 Rural livelihoods are a combination of low-level agriculture and remittances from family members engaged in the urban cash economy, with mountain areas being typified by livestock grazing and other areas focusing on a combination of cropping and livestock. However except for the rural-urban divide, there is no clear geographical pattern to poverty (for example, inequality and malnutrition tend to 3 Based on estimated food poverty headcount rates of 37.8 percent in rural areas and 18.3 percent in urban areas (according to preliminary government estimates based on the 2010/11 HIES). x be spread across all of the districts) so any geographical targeting of safety net programs will have to identify small local pockets of poverty. 10. According to the 2009 Labor Force Survey, only about 27 percent of the economically active population is employed in the cash economy. The poor inevitably engage in multiple activities to compensate for the limited opportunities and under- employment in farming, which is highly seasonal, and often does not require labor for parts of the year. The problem is not so much open unemployment as underemployment and the low returns to the labor of the poor. This can be addressed by introducing productive safety net programs that provide poor rural households with cash to engage in small off-farm businesses or to buy agricultural inputs or by providing them with temporary employment during the slack season through public works programs. 11. Agricultural production has been stagnant or declining over the past decade, with declining yields and little growth in livestock herds. While only a relatively small proportion of the poor are landless, the quality of that land varies widely. Also, there are often not enough able-bodied workers in poor households to use the land effectively, and the downturn in remittance income from migrants in the mining industry in South Africa has meant that many households can no longer afford to buy fertilizer and other inputs. In addition, the food security situation in Lesotho worsened due to heavy rains in 2010 followed by a drought in 2011/12 and as a result, the Prime Minister recently declared an emergency food crisis. 12. Not having enough to eat is a very good measure of the need for safety net assistance. While the numbers change from year to year, our best estimate is that about 200,000 to 400,000 people, representing 10 to 20 percent of the population, are chronically food-insecure. Almost two-thirds of food consumed in Lesotho is imported from South Africa. On the one hand, this means that supply is more reliable and grain is almost always xi available. On the other hand, however, it means that people have to have the cash to be able to buy it, and this is a particular challenge for the poor. 13. Lesotho has high levels of child malnutrition for a country at its income level, and this has a lifetime impact on the productivity and poverty status of the country’s population. Safety net interventions that address early childhood malnutrition (such as making cash transfers conditional on participation in growth monitoring, nutrition education, and therapeutic feeding) should be an important part of the national strategy. 14. Most of the extremely poor in Lesotho do not move in and out of poverty but are chronically poor because of inherent conditions such as their low productivity and the low returns to their labor in the subsistence economy or their lack of basic assets (including health, skills, and education) or the destitution of their households (the elderly, street children, the disabled, or the incapacitated who are not in families that can support them). However, poor households are also subject to a range of shocks including theft (primarily of livestock), unemployment, drought, price increases, and illness and death due to disease. They also face annual seasonal food shortages. With a short growing season and limited land and labor, the very poor only produce enough maize to last a few months after the harvest and must then buy their food from the market at a time when prices are highest. This seasonal shock happens every year and accounts for a significant component of extreme poverty in Lesotho. The national safety net system should be designed to address it. Existing Transfer Programs 15. This study examines all programs that transfer money or in-kind assistance to individual households in Lesotho. Not all of them are explicitly designed as safety nets to reduce extreme poverty (the Tertiary Bursary scheme, for example, is intended to support investment in higher education and the National Fertilizer and Input Subsidy to boost xii agricultural output), but together the 10 existing programs represent a very large share of public spending that is being transferred directly into the pockets of individuals. The study aims to discover the impact that each of them is having in terms of reducing extreme poverty to enable policymakers to know whether they are getting value for their money. 16. Child Grants Program (10,000 households costing M.16 million or US$2.2 million annually) – This is a non-conditional social cash transfer program targeted to poor households with children. At present the program covers roughly one-sixth of the country but is being expanded. The program is relatively new and is still very dependent on donor financing and technical assistance. Evaluations have shown that the program has had a positive impact and has reasonably good targeting, although only 50 percent of benefits go to the extreme poor. 17. Public Assistance (9,500 beneficiaries costing M.16 million or US$2.2 million annually) – This provides support in cash or in kind (for example, as medical exemptions and food packages) to about 9,500 destitute people annually. Coverage is limited by budget availability. Beneficiaries are identified by the Department of Social Welfare. 18. Old Age Pension (83,000 beneficiaries costing M.371 million or US$49 million annually) – This is one of the largest safety net programs in Lesotho. It is a universal, non- contributory, unconditional cash transfer program, available to everyone over the age of 70. Since only 6 percent of the poor are estimated to be older than 64, any age-based program is not going to directly reach many of the poor. 19. School Feeding Program (389,000 students costing M.236 million or US$31 million annually) – This is the transfer program with the largest number of direct beneficiaries in Lesotho. Funded by the government and the World Food Programme (WFP), it provides all children in public primary schools with free meals to improve their health and educational outcomes. Its unit costs (US$65 per student per year) are high compared with those of xiii similar programs elsewhere. The program has had a positive educational impact, but about 60 percent of the benefits go to households that are not poor. 20. OVC Bursary Program (coverage 20,000 costing M.48 million or US$6.4 m. annually) – This covers tuition and boarding fees for secondary students identified as OVCs and is financed by the government and the Global Fund. Beneficiaries are selected based more on their orphan status than on their poverty. The program makes important investments in education, but its poverty-targeting effectiveness is not known. 21. Nutrition Support Program (intended coverage 85,000 people at an estimated cost of M.47 million or US$6.3 million annually) –This program aims to reduce malnutrition among vulnerable groups in four districts where malnutrition is particularly high. It is new and is entirely dependent on WFP funding. 22. Agricultural Input Fairs (18,600 farmers at an estimated cost of M.17 million or US$2.3 million annually) – AIFs provide selected poor farmers with input vouchers worth about half the price of the seeds and fertilizer needed to cultivate a half-hectare of maize, beans, or other basic crops. The program relies on donor financing. Based on only limited data, the program appears to have had a significant positive impact on household food security and consumption. 23. Integrated Watershed Management Public Works Program (96,000 people employed annually at a cost of M.112 million or US$15 million) – The Ministry of Forestry operates this program, which employs villagers to plant trees and carry out other environmental conservation work. Although not specifically targeted to the poor, it provides the basis for a larger cash-for-work program employing the poorest. 24. National Fertilizer and Input Subsidy (Annual cost of M.44 million or US$5.9 million for an unknown number of beneficiaries) – The government provides an untargeted subsidy xiv on fertilizer and other agricultural inputs. The subsidy has varied between 30 percent and 50 percent of the retail price. The impact on output is not evident, and the subsidy does not benefit the poorest. 25. Tertiary Bursary Scheme (16,200 annual beneficiaries at an annual cost of M.575 million or US$76 million) – The largest single transfer program operating in Lesotho in budget terms, the scheme pays the fees and living expenses for students who attend universities and other tertiary education institutions. Although in theory a loan scheme, there is almost no repayment, so it is effectively a pure transfer program. It has been estimated that only 1 percent of benefits go to the extreme poor. 26. The following general observations can be made on the current collection of transfer programs in Lesotho:  Total spending on all transfers is high. Total spending on transfer programs amounts to about US$197 million annually or roughly US$104 per person. If this amount were targeted to the poorest 25 percent of the population, it could provide a transfer of over US$340 per person, which is almost exactly equal to the estimated poverty line income.4  Only a small share of the poor is being reached even though a large amount is being spent on transfers. Most of the benefits of the large-scale programs (agricultural subsidies, tertiary bursaries, and school feeding) go to the non-poor, while the programs specifically targeted to the poor are very small-scale, typically reaching only 3 to 4 percent of the extreme poor (see Table 24 and Figure 8).  The current package of programs misses some important aspects of poverty. Although the current range of safety nets is addressing some of the aspects of 4 Based very roughly on 475,000 people, a rough poverty line of US$325 (a recent estimate by the World Bank and the Bureau of Statistics of M.243 per adult equivalent per month in 2010/11 Maloti) and program operating costs of about 15 percent. xv poverty discussed in Chapter 2, it is missing others such as seasonal poverty and malnutrition.  There is no overall framework for coordinating transfers. As a result, overlaps and duplication occur, with some needy households (or areas) benefitting from multiple programs, while others that are equally in need benefit from none.  The existing range of programs cannot be easily scaled up or down in response to changing economic conditions or to shocks such as drought.  With the exception of the Child Grants Program, none of the existing programs has been rigorously evaluated either in terms of who they are reaching or of their impact on poverty or consumption. This makes it difficult for policymakers to know which programs are making effective use of resources in reducing poverty.  There is scope for increasing the promotive impact of programs. While some of the existing programs aim to raise the poor out of poverty in the longer term (for example, the OVC Bursary Scheme and Agricultural Input Fairs), most do not explicitly include any measures to permanently increase the productivity of the poor.  The lack of an overall strategy and agency means that there is no framework for political leaders or the Ministries of Finance and Development Planning to evaluate trade-offs between programs or even to get an overall view of total spending on transfers.  The weakness of data, both on household poverty and on the coverage and impact of programs, makes it difficult for policy-makers to know which programs are effective at reducing poverty, and to determine if they represent value for money. 27. The study makes a number of cross-cutting recommendations that apply to any future national safety net system in Lesotho:  Increase the predictability of transfers. Experience worldwide has shown that, to have a permanent impact on a household’s poverty level, transfers have to be xvi sustained and reliable. Households need to be certain that the money they are expecting will indeed be available every month as this will enable them to plan ahead and take more risks and make more productive investments. Some programs are successfully providing predictable transfers (such as the Old Age Pension and the Child Grants Program), while others are less successful in doing so (such as the AIF, the OVC bursaries, and the public works programs).  Improve targeting and coordination. Like most countries, Lesotho cannot afford to waste money by providing transfers to people who do not need them. One way of avoiding this is to strengthen targeting and coordination to ensure that benefits only reach those most in need. The recent efforts to target the Child Grants Program and to establish a national registry of beneficiaries (NISSA) have the potential to improve targeting and coordination.  Evaluate programs more rigorously. It is important to carry out rigorous evaluations of existing programs to enable policymakers to assess which programs represent the best value for money in reaching the poor. This report is a first attempt to evaluate the costs and effectiveness of Lesotho’s safety nets, but more rigorous examination of the poverty levels of beneficiaries and of the programs’ impact on their consumption should be a priority.  Increase the promotive aspect of programs. Wherever possible, safety net transfers should improve the conditions of the poor over the long term. For example, cash transfer programs could include conditions to encourage recipients to take actions to build their human capital and productivity, such as keeping their children in school, taking advantage of health care, or participating in nutritional interventions.  Increased productive aspect of programs. Spending on transfers can be made more productive, for example, by employing the poor to maintain roads, plant trees, or improve land management or by linking transfers to increased agricultural productivity as is done in the AIF program. xvii  Move towards a unified national program. It makes sense to move towards a unified national program that may still consist of several distinct interventions but will operate within an integrated approach to objectives and targeting.  Evaluate impact of programs and improve consumption data . Systematically and independently evaluate the impact of transfer programs, and improve the quality and frequency of household consumption data which is needed to analyze targeting efficiency. 28. Given Lesotho’s tight fiscal situation, in rationalizing its expenditures on safety net transfers, the government should seek to:  Select the most cost-effective instruments.  Do so in a way that is fiscally sustainable.  Wherever possible, use existing expenditures to minimize the need for new spending.  Narrowly target those aspects of poverty for which transfers are the most appropriate solution (this is likely to include some forms of categories related to HIV/AIDS, child-headed and grandparent-headed households, early childhood malnutrition, and seasonal food poverty, as well as geographically targeting of pockets of poverty).  Ensure that programs and investments can be scaled up and down in response to evolving fiscal conditions and changing needs, rather than becoming permanent entitlements. Institutional Issues, Program Delivery, and Targeting 29. The government is in the process of finalizing a National Social Development Policy (NSDP) that emphasizes the critical role of safety nets in combating poverty. The draft NSDP states that the goal of Government policy is to prevent and reduce poverty with a view to xviii ensuring that all of the Basotho enjoy an acceptable minimum standard of living. In order to achieve this objective, the draft NSDP emphasizes the need to develop and implement a comprehensive social protection system, including safety nets. Building on the NSDP, the government intends to develop a social protection strategy to articulate the overall objective of social protection (including social safety nets) and to guide reforms. 30. Lesotho’s transfer programs are currently implemented by several different ministries, with donors being significantly involved in the design of several programs. 5 The government recognizes the need for developing a social protection strategy to improve institutional coordination, as well as to harmonize legal frameworks and consolidate existing safety nets. The recent creation of the new Ministry of Social Development demonstrates the growing emphasis that the government is putting on social protection in Lesotho, but the effective coordination of efforts will be one of the main determinants of success. One way to improve coordination and increase the efficiency of safety net programs would be to create a central agency responsible for coordinating all safety net programs. The establishment of such an agency (and the building of its capacity) would not only improve coordination, thereby decreasing the risk of any overlap of beneficiaries between programs, but would also make it easier to monitor their impact. As such, it is expected that the newly established Ministry of Social Development will take the lead on the social protection agenda moving forward. 31. In trying to make Lesotho’s transfer programs more cost-effective, policymakers will need to find ways to tighten targeting. The challenge is to find targeting mechanisms that 5 The Department of Social Welfare in the Ministry of Health and Social Welfare had been leading the social protection agenda in the country up until the creation a separate Ministry of Social Development in June 2012. The Child Grants Program, the Nutrition Support Program, and Public Assistance are all being implemented by the new ministry. Most of the larger transfer programs are implemented by other ministries. For example, the Old Age Pension is implemented by the Pensions Unit in the Ministry of Finance and Development Planning, the School Feeding Program and the OVC Bursary Scheme are both implemented by the Ministry of Education and Training, and the National Fertilizer and Input Subsidy and the AIFs are operated by the Ministry of Agriculture and Food Security. Other safety net programs are implemented by the Ministries of Public Works and Transport, Forestry and Land Reclamation, and Local Government and Chieftainship. xix can work within the existing constraints and to do so for a reasonable cost. Targeting is particularly difficult in a country like Lesotho where: (i) a large share of the population is poor and there are few obvious differences between the conditions of the ultra-poor and the poor; (ii) there is little up-to-date information on household income; and (iii) administrative capacity at the local level is limited. One way to increase the efficiency of the safety net system is to have one common targeting mechanism (or at least a common registry) for all social transfer programs in the country. The current efforts under the Child Grants Program that combine community targeting with proxy means testing have shown some promise but need refining. Also, given the difficulty of distinguishing the able-bodied poorest from the poor in general, some form of self-targeting is likely to be required. 32. Lesotho is in the process of establishing a National Information System for Social Assistance (NISSA), that could go a long way towards both improving targeting and reducing the overlap of beneficiaries between programs. NISSA is currently being used to target the beneficiaries of the Child Grants Program and has registered about 30,000 households, but the eventual aim is to transform it into a national registry for all of Lesotho’s social assistance programs. At a minimum, it would be a unified database that policymakers could use to determine who is benefitting from which program and to coordinate coverage. 33. Therefore, the following actions are recommended: - Establish a national safety net strategy for the next 5 to 10 years that spells out the broad poverty reduction objectives, desired target groups and coverage, and program choices. The forthcoming formulation of a national social protection strategy represents a unique opportunity to do this. - Eliminate duplication of coverage (for example between the Old Age Pension, the Child Grant Program, and Public Assistance) by creating an integrated beneficiary registry. xx - Establish a central body with policy oversight and expenditure planning authority over all transfer programs. - Continue the gradual expansion of the NISSA once a clear vision for the system has been developed. Careful consideration has been given to how the expansion should take place. The system should also be continuously monitored and evaluated to determine if it is in fact an effective way to improve coordination and targeting. Conclusions and Recommendations 34. The report makes two sets of recommendations. The first set covers options for implementing a broad national safety net system over the coming decade that is affordable and that has the potential to significantly reduce extreme poverty. The second set provides specific measures in the shorter term to strengthen existing programs and put Lesotho on the path towards creating a unified national system. What Role Can Safety Net Transfers Play in Lesotho’s National Development Strategy? 35. The solution to rural poverty in the long run will almost certainly be the continued migration of many rural residents to urban areas combined with increased agricultural productivity and greater connections to markets for those who continue in rural areas. However, in the short to medium term, there are legitimate arguments to be made for providing targeted transfers to those most in need. 36. First, the high level of inequality provides a compelling rational for greater use of safety net transfers: because arithmetically, taking a relatively small amount from the top end of the economy can fund a large proportional increase in the consumption of the bottom end; and because such high and persistent inequality is the clearest sign that other approaches tried to date have not reached the poorest. Second, shortages of cash income in rural areas appear to be constraining smallholder agricultural production and are limiting xxi effective demand that hinders the development of markets. Third, there are a large number of people in some specific categories who are incapable of supplying their own minimum levels of consumption (for example, child-headed and grandparent-headed households). Finally, safety net transfers represent an injection of funds straight into the lowest level of the economy, where they have large multiplier effects, and have a more immediate impact on poverty status than almost any other expenditure the Government can make. 37. In this report we argue that the overarching objective of a national safety net system should be to protect the consumption of the poorest Basotho at the food poverty line . This provides a clear, unambiguous focus, and limits the potential scale of transfers. Depending on the estimates one believes, this could vary between 10 and 20 percent of the population. It is further recommended that the national program be linked to addressing some basic national priorities: addressing early childhood malnutrition and the inter-generational poverty effects of HIV/AIDS are two potential ones. This will both give a clear focus to the program, and increase social and political support. 38. The question that follows is what type of safety net to provide and how to make it as cost-effective as possible. The analysis in this report points to the need to provide a combination of categorical support to some very specific groups of the ultra-poor in the poorest areas of the country and transitional productive support to those who are either being left behind by growth or who live in low-productivity areas. 39. Based on the analysis in this study, we suggest that a national package of interventions might consist of:  A single cash grants program targeted mostly to categorical groups of the extreme poor such as child-headed households, grandparent-headed households with no working aged adults, and the infirm and disabled who are not living with relatives. xxii This program could be linked with interventions to reduce chronic child malnutrition in beneficiary households. The existing Child Grants Program would provide a basis on which to build this proposed cash transfer scheme for the destitute.  Some form of self-targeted public works program for the able-bodied poor. Lesotho is unusual in not having any large-scale cash-for-work or food-for-work program. Existing expenditures on public works and maintenance could be used to employ the rural poor during the agricultural slack season, and the existing Watershed Management Program could provide a framework for a public works program employing the very poor on re-forestation, environmental works, and rural road maintenance.  A more focused and effective version of the Agricultural Input Fairs, though a decision on this should depend on a more rigorous assessment of the costs and benefits of the program and the extent to which it has been successful in increasing the consumption of the poor. 40. There needs to be a national debate on the coverage of these interventions. Based on the analysis in this report, we would recommend that the target group should be those with insufficient food consumption – perhaps 200,000 to 400,000 people, representing the poorest 10 to 20 percent of the population (or 40,000 to 80,000 households). The cash grants program would be targeted to households with no able-bodied members, and the public employment scheme would cover those with able-bodied members. 41. The amounts required to finance such a program would be affordable in the context of Lesotho’s current fiscal position. Depending on the coverage of the program, the costs might amount to between M.200 million and M.615 million a year (US$27 to $81 million). A medium variant, as can be seen in Table 26, would cost about M.411 million (US$ 55 million) annually, representing about 4 percent of total public expenditure or 2 percent of GDP. To put this in perspective, total spending on transfers in Lesotho in 2010/11 amounted xxiii to about M.1,500 million (US$197 million) annually, or roughly 9 percent of GDP. The government could use some existing spending to pay for a national pro-poor transfer program (for example road maintenance expenditure could finance the employment of the poor while savings could be made on the School Feeding Program and other large untargeted transfers such as the National Fertilizer and Input Subsidy and the Tertiary Bursary Schemes). 42. In the current fiscal environment it does not appear either sensible or feasible to advocate for a large-scale expansion of cash or in-kind transfers. However, there is scope for making more use of narrowly targeted transfers to reduce extreme poverty, for redirecting expenditures, and for improving the performance of existing programs. Any efforts to expand existing universal transfer programs (for example, by lowering the pension age or enhancing the school meals package) should be resisted. Also, over time policymakers should consider ways to improve the targeting of the School Feeding Program. What Measures Can be Taken in the Short to Medium Term to Strengthen Existing Programs? 43. Finally, in this report, we make various program-specific recommendations both to increase efficiency and to put move towards the more unified program that we proposed above. These are summarized in the table below. xxiv Key Observations and Recommendations for Each Program Program Key Observations Recommendations (Short Term= 1-2 years and Medium Term= 3-5 years) Safety Net Programs Child Grants The CGP has well-defined operational features as well as the Short Term: (a) Rigorously assess the effectiveness of current targeting Program necessary operational systems in place to implement the and the impact of the program; (b) hold policy-level debate on program. It is also one of the few programs that make a intended national scale and willingness to finance long term; and (c) systematic effort to identify the most vulnerable. The consider alternative criteria and methodologies for targeting, based on effectiveness of its targeting has not yet been evaluated evidence from evaluation. rigorously, and it is heavily dependent on external support. Medium Term: (a) Expand to a general program to reach the most destitute and (b) consider introducing co-responsibilities for the receipt of the cash transfers to encourage human capital development; particularly related to malnutrition. Public Assistance Coverage is too small to affect poverty nationally, and Short Term: (a) Consider integrating PA with an expanded nationwide payments are too sporadic to affect poverty at the cash grants scheme based on the CGP model (above). household level. Old Age Pension The OAP benefits many deserving poor people and there is Short Term: (a) Do not lower pension age at this stage; (b) exclude some evidence that the consumption and education of people who receive pensions from the private sector that are higher recipients has increased as a result of the pension. than the OAP; and (c) continue to explore ways to make payments However, the program also provides benefits to many non- electronically. poor elderly. Medium Term: Analyze the sustainability of the program in the long run. School Feeding While successful in reaching all students nationwide, the Short Term: (a) Continue to rationalize the program into a unified Program SFP is an expensive way of making transfers, and many of home-grown feeding program but make every effort to reduce costs – the benefits go to households that do not need them. possibly by aiming to provide a more modest meal; (b) increase operational efficiency; and (c) strengthen implementation capacity Medium Term: Eliminate free meals in better-off schools. Nutrition Support This program could be used to link the safety net with Short Term: (a) Continue implementation of the program but with Program malnutrition. Given its early stage of implementation, it is close monitoring and evaluation and assess its costs and effectiveness. too early to assess its effectiveness or impact. Medium Term: (a) Consider introducing an expanded nutrition program; (b) review experience with any other previous programs and successful international experience; and (c) determine what scope there is for expanding it as part of national safety net in addition to the unified cash transfer and public works programs. xxv OVC Bursary The program is an important effort to help deserving Short Term: (a) Ensure predictability of support; (b) consider Scheme children to go to secondary school, but does not seem to be alternative targeting methods to improve targeting of worse-off particularly pro-poor and the amount provided seems students; and (c) develop clear targeting and operational guidelines. inadequate and unreliable. Medium Term: (a) Consider giving additional support to students from worse-off families and (b) explore better links with other programs or potentially merging targeting with a national cash transfer scheme. Agricultural Input The program has the benefit that it targets smaller and Short Term: (a) Rigorously evaluate the program’s impact on Fairs more vulnerable farmers, but there is limited evidence on consumption before deciding whether to continue it. who it is reaching or on the impact it is having on the Medium Term: (a) Subject to the evaluation either make the program consumption of beneficiaries. Its funding is unpredictable part of the overall safety net or discontinue it when donor funding because of its donor dependence but provides a possible ceases and (b) ensure that any expanded program is more integrated basis for expansion into a wider-coverage program of into the routine operations and financing of the government. agricultural input support for the poorest among rural farmers. Other Transfer Programs Integrated The program shows some promise as a works-based Short Term: (a) Expand spending on low-level road maintenance; (b) Watershed transfer scheme that also improves the environment. It maximize labor-intensity & employment of local women and men; (c) Management Public could employ more people if expanded beyond critically evaluate experience with the forestry program and assess Works Program reforestation to other land management and infrastructure scope for expanding works-based transfers; and (d) carefully pilot and maintenance works and if it were more explicitly targeted monitor the WFP cash-for-work/food-for-work program to the poor through a combination of a low wage rate and Medium Term: Expand labor-intensive works-based transfers based on geographical targeting. Continuing annual employment for experience and evidence from ongoing and previous programs and of the same poor villagers is needed rather than one-off public employment for the very poor in other countries. projects. National Fertilizer The program is untargeted, there is no evidence as to its Short Term: (a) Consider more intense targeting in poorer areas; (b) and Input Subsidy output impact, and it seems unlikely that it reaches the reduce the scope of the National Fertilizer and Input Subsidy and retain poor. only for poor small-scale farmers; and (c) consider ways to prevent recipients from re-selling the inputs at the commercial price. Medium Term: Consider eliminating the program entirely but strengthen the targeted subsidy under the Agricultural Input Fairs instead. Tertiary Bursary The program as it is currently structured suffers from a Short Term: Consider how to restructure the program over time. Scheme number of fundamental flaws. It is costly, it benefits mostly Options include better targeting (possibly some form of means testing) households who do not need a public transfer, and it needs and shifting to a more commercially oriented program that is a more to be more financially sustainable. genuine loan scheme. Medium Term: Restructure the program. xxvi INTRODUCTION 1. The main objective of this study is to help the Government of Lesotho and its development partners to decide what role safety net and transfer programs should play in the coming 5 to 10 years. The following are the questions it seeks to answer: (i) could increase spending on transfers be used to accelerate poverty reduction in the medium to long term? (ii) Which groups and aspects of poverty would it makes sense to target with transfers? And (iii) which programs would have the greatest impact for the most affordable cost? 2. Safety net transfer programs are generally understood to include any programs that transfer resources – either in cash or in-kind – to individuals or households with the express intent of increasing their consumption. Examples include social cash transfer programs (such as child support payments and old age pensions), public works employment schemes for the poor, school feeding programs, and educational scholarships for poor students. This study examined such programs in Lesotho – both ones that currently exist, and the potential scope for using instruments that might not currently exist. It also widened its scope to examine all programs that transfer public resources directly to households – such as agricultural subsidies and university bursaries. It analyzed them through the lens of their impact on the poor because it is important to evaluate public spending and trade-offs in the context of where public funds are currently being spent and what distributional impact they are having. The assumption underlying this analysis throughout the report is that the over- riding objective of both the government and its development partners is to reduce the high levels of poverty and inequality that persist in Lesotho. 1 Box 1: Some Key Productive Benefits of Safety Nets Safety net programs, if properly designed, can not only protect and increase the immediate consumption of beneficiaries but also have longer-term positive effects such as breaking the intergenerational transmission of poverty and permanently raising incomes.  Safety net transfers have an immediate impact on inequality and extreme poverty . Simply by providing transfers to the poor, safety nets have an immediate impact on extreme poverty and reduce inequality by directly redistributing resources to the worse off.  Safety net transfers can enable households to make investments in their future that they would otherwise not be able to make. Safety nets help families to increase their human capital (via improved nutrition and greater use of education and health services) and enhance their livelihoods (by enabling them to invest in productive assets).  Safety nets can help the poor to manage risk by making it possible for households to avoid falling permanently into destitution (for example, as a result of selling their assets, such as land or livestock, during prolonged periods of droughts or floods). It also gives them the flexibility to take on higher-risk, higher-return activities, which in the context of Lesotho might mean greater use of fertilizer and other purchased agricultural inputs or diversifying cropping.  Safety net transfers can help the poor to increase their income over the long term as the cash that they receive from transfer programs enables them to build up small amounts of capital and diversify their activities.  Safety net transfers inject funds directly into the lowest level of the economy where they increase demand and promote growth opportunities at the village level.  Public works temporary employment schemes not only provide the able-bodied poor with transfers but also create assets that will foster longer-term growth, for example, by building roads that increase access to markets and enable market integration or through soil and water management, reforestation, and irrigation works. 3. Chapter 1 explores the country context and fiscal space for public expenditure on social safety nets. Chapter 2 examines the dynamics of poverty in Lesotho and the characteristics of the major vulnerable groups and discusses which of these groups it would make most sense for the social safety net to target. Chapter 3 analyzes Lesotho’s existing transfer programs. Chapter 4 reviews institutional and capacity issues. Chapter 5presents options for a long-term package of social safety nets and associated costs, makes some recommendations for strengthening existing programs, and suggests which type of programs it might be cost-effective to expand. 2 CHAPTER I: COUNTRY AND FISCAL CONTEXT 4. The defining characteristics of Lesotho are that it is surrounded by South Africa, it is a heavily bifurcated economy (split between those who are engaged in wage employment and those who remain in subsistence agriculture with few off-farm opportunities), and many households rely heavily on both internal and external remittances. Lesotho has a population of approximately 1.9 million and a per capita income of about $1,000.6 More than two-thirds of the country consists of mountainous highlands, and less than 10 percent of the land is suitable for cultivation. The most abundant natural resources are water, diamonds, and hydro-potential. In the mountains in particular, wealth and agricultural activity revolve primarily around livestock ownership. 5. Lesotho has experienced significant economic growth during the last two decades, but deep poverty remains. The country’s real per capita GDP growth between 1981 and 2008 (averaging 3.4 percent) was above the average for Sub-Saharan Africa. Between 2001 and 2010, per capita GDP grew at an annual rate of 2.6 percent. The engine of economic growth between 2001 and 2007 was manufacturing exports, and between 2008 and 2010 it was public investment. Agriculture has lagged behind, and a substantial minority of households, mostly in rural areas, have not benefitted from this growth. In 2010 an estimated 57 percent of the population lived below the national poverty line. 7 6 World Bank (2010b), p. 1. 7 Preliminary government estimates based on the 2010/11 HIES. 3 Box 2: Main Data Sources and Caveats Numerous data sources have been used to examine the aspects of poverty most relevant to designing a safety net in Lesotho. These include: the annual Lesotho Vulnerability Assessments; the 2009 Demographic and Health Survey (DHS); the quarterly Continuous Monitoring Surveys conducted since 2009; the Lesotho Integrated Labor Force Survey of 2008; the Agricultural Census of 2009/10; and the World Bank’s own Poverty Assessment of 2010. The main sources of household consumption data used to calculate poverty indicators are the consecutive Household Income and Expenditure Surveys (HIES) conducted in 1994/95, 2002/03, and 2010/11. In 2010/11 the update of the HIES was done by adding a consumption module to the CMS (hereafter referred to as the 2010/11 HIES). While a formal poverty assessment update has not yet been issued on the basis of this data, preliminary analysis has been undertaken by the Bureau of Statistics, and these preliminary poverty estimates are used in this report to provide the most recent available picture of poverty. 6. The economy has undergone a significant structural shift away from agriculture and livestock since the early 1990s. The industrial sector grew 40 percent faster than the economy as a whole because of the development of the textile and mining sectors. The services sector has also been flourishing thanks to transport, communication, and financial services. As a consequence the share of the agricultural sector in GDP has fallen to 8 percent while industry has risen to a share of 34 percent and services to a share of 58 percent.8 This structural transformation has created new employment opportunities for many, but as the transition has only been partial and disproportional to the structure of the labor force, a large share of the working-age population is still stuck in the agricultural sector with limited growth prospects and few income-earning alternatives. 7. The recent rapid growth of the economy has not been pro-poor. The economic growth of the past 20 years was not accompanied by high population growth, which should have given Lesotho an opportunity to reap the dividend from its demographic transition.9Nevertheless, even though Lesotho has enjoyed rapid levels of economic growth in the past decades, the overall increase in welfare has not been sufficient to significantly lower the level of poverty. 8 World Bank (2011d). 9 World Bank (2010b). 4 8. The poverty elasticity of growth has been estimated at -0.579, implying that a 1 percent increase in GDP resulted in a 0.6 percent decline in the incidence of poverty. 10 In addition, many of the poor are living too far below the poverty line to escape poverty even if they did get their proportional share of growth.11 Thus, in order for growth to lead to a substantial decrease in poverty in the future, the benefits will need to be more widely shared and more heavily skewed towards the poor. 9. Inequality is high and there is a large urban-rural gap. The Gini coefficient for Lesotho was 0.53 in 2010, which compares with levels of 0.54 in highly unequal economies such as Brazil and of 0.38 in fairly egalitarian societies such as Vietnam and Tanzania. There is also a large urban-rural gap. As of 2006, 76 percent of the population of Lesotho lived in rural areas,12 and extreme poverty was more than twice as high in rural areas than in urban areas.13 For some households, the loss of income from remittances combined with decreasing agricultural productivity has further increased their poverty.14 The economy is vulnerable to both natural shocks (in agriculture in 2001, 2005, 2010, and 2011) and external economic shocks, which resulted in the contraction of wage employment and remittance earnings in both 2005 and 2009. 10. Inequality in Lesotho is only partly an urban-rural phenomenon. Decomposing the Thiel Index, which also expresses inequality, shows that only 4 percent of overall inequality can be explained by the urban-rural divide. The remaining 96 percent was explained by intra-urban and intra-rural inequality.15 10 Using the data on GDP per capita and the incidence of poverty in the 1994/95 and 2002/03 household budget surveys in, a partial poverty elasticity of -0.579 (UNDP, 2007, p. 23). 11 World Bank (2010 b) p. 27. 12 BoS (2009a). These are 2006 Census results, while government preliminary analysis of the 2010/11 HIES classifies about 19 percent of the population as living in urban areas, and 81 percent rural, although definitions of what constitutes an “urban” area is likely to differ from the Census. 13 Based on estimated headcount poverty rates of 37.8 percent in rural areas and 18.3 percent in urban (food poverty line, using government preliminary estimates of the 2010/11 HIES data). 14 World Bank (2010b). 15 World Bank (2010b). 5 11. Why is inequality so high? The answer is partly because of the existence of two very different economies within Lesotho, one consisting of those who have participated in the cash economy through government employment, outmigration, or employment in services and industry and the other consisting of those who have not. 12. The persistence of high inequality is an argument in favor of pursuing a more inclusive growth path and adopting policies that will overcome inequality in time (emphasizing education and the integration of markets and remote areas), but also for the creative use of safety net transfers to reach those left far behind. The challenge is how to provide such safety net support in ways that increase productivity and avoid giving people incentives to remain in low-productivity areas or activities. 13. The HIV/AIDS epidemic has resulted in a large number of orphans and a fall in the country’s already low life expectancy. Lesotho has the third highest HIV adult prevalence rate in the world at 23 percent (after Swaziland with 25.9 percent and Botswana with 24.8 percent).16 At the end of 2008, an estimated 280,000 people were living with HIV, of whom 21,000 were children. As a result, Lesotho is one of nine countries where life expectancy at birth has fallen below the level in 1970.17 AIDS has resulted in a loss of able-bodied labor, has undermined economic productivity, and has created large numbers of orphans. According to the Bureau of Statistics, in 2011 there were 238,000 orphans and other vulnerable children in the country.18 14. Lesotho is facing serious fiscal challenges, and increasing the efficiency of public spending has become imperative. Lesotho’s economy is heavily dependent on the revenues from the Southern African Customs Union (SACU), which have declined sharply as a result of the global economic downturn. Since Lesotho spends about 17 percent of its revenue on 16 MHSW (2010) and UNAIDS (2010), p. 181. 17 UNDP (2010). 18 BoS (2011b). 6 various transfer programs19 and given that the programs are almost entirely funded by general tax revenues, it has become increasingly important to examine the efficiency of this spending and to ask what role spending on different types of safety net programs should play. 15. The Government of Lesotho recognizes the need to support the poor and vulnerable through social protection interventions, but it is concerned about creating new entitlements in the current fiscal circumstances and wants to ensure that these programs are operationally efficient. The government has adopted a medium-term program designed to restore macroeconomic sustainability and achieve sustained growth that will reduce poverty, and one of the pillars of this reform program is containing public expenditure while protecting poor and vulnerable groups.20 16. Exploring options for putting together a rational package of social safety net programs is particularly relevant in the current context of high inequality, persistent poverty, and HIV/AIDS prevalence. Well-designed safety nets can have an immediate impact in terms of reducing inequality and extreme poverty. They can also enable households to make better investments in their future, including in the human capital development of their children and livelihoods, and help them to manage risk.21 Because of the urgency of the need to ensure efficient spending and operational effectiveness, now is the right time to review Lesotho’s existing social safety net programs to assess whether they are succeeding in helping the most destitute. This report reviews the main safety net programs in Lesotho and discusses a range of options for putting together a package of social safety net interventions to accelerate the pace of poverty reduction. 19 Based on estimate of approximately US$197 million equivalent on transfers (Chapter 3) and revenue of $1,250 million in 2010/11 (Chapter 4). 20 As is clear in the Interim National Development Framework, the government is committed to social protection for poor and vulnerable populations. 21 Grosh et al (2008) 7 Fiscal Space 17. Facing limited opportunities for agriculture and domestic economic production, the Government of Lesotho has over the last 30 years adopted what is essentially a redistributive public policy stance. It has used revenues from water sales to South Africa and SACU receipts to fund: (i) employment in the public service; (ii) widespread free public education and health services; and (iii) selected transfers (including the Old Age Pension and the School Feeding Program). 18. A perfect storm of conditions has undermined this strategy in the recent past:  The collapse of employment in the mines of South Africa (from about 100,000 Basotho in the 1990s to about 41,000 in 2011).22  A long-term decline in agriculture (with Lesotho going from maize self-sufficiency in the 1970s to producing only about 30 to 50 percent of its requirements today).23  The impact of HIV/AIDS (an incidence of 23 percent, the third highest in the world), which has decimated the labor force and created an estimated 200,000 orphans.  A fiscal crisis threatened the country’s ability to finance its public employment and programs. 19. Lesotho has historically, and recently, had very high levels of public spending . On the back of windfall SACU revenues, government spending had risen to an unsustainable 67 percent of GDP by 2009/10. These spending levels financed the growth of a large civil service, with the wage bill rising to a fixed commitment of 21 percent of GDP in 2009/10.24 Table 1 gives a broad overview of Lesotho’s fiscal situation in 2011/12. 22 2011 estimate from World Bank (2011d) 23 Varies depending on year. Estimates from 2010 and 2011 Food Security and Vulnerability Monitoring Reports (LVAC/DMA, 2010 and 2011). Also, Agricultural Situation Reports (MoAFS, 2010 and 2011) show long- term secular declines in maize output and yields from 2000/01 to 2008/09. 24 The wage bill has since declined somewhat, to about 19% of GDP in 2012/13 IMF (2013). 8 Table 1: Revenue and Expenditure, 2011/12 a/ Maloti millions US$ million Total Revenue 9,616 $1,282 millon Of which, SACU 2,752 $366 million Total Expenditure 11,497 $1,532 million Selected Sector Spending % of Total Spending b/ (2010/11) Education 1,965 $262 million 20.9% Health& Social Welfare 1,654 $220 million 17.6% PW and Transport 688 $92 million 7.4% Approx.Transfers & Safety 1,470 $197 million 16% c/ Nets Spending Sources: IMF data 2011/12 (2013), Budget Document for sector spending (MoFDP {(2011a}). Notes: a/ Based on exchange rate of 7.5 for 2011/12. b/ Recurrent plus capital; note sector spending composition is reported for 2010/11, to maintain comparability with the spending on transfer programs tabulated in this report, for which 2010/11 budget year represented the most recent available data on most programs. c/ As estimated in this report; includes some off-budget aid financed programs and is thus not directly comparable to other sectors. 20. These very high spending levels combined with the global economic contraction, and local and regional pressures have resulted in significant fiscal pressure. SACU revenues, which accounted for over half of Lesotho’s total revenue up until 2009/10 (33 percent of GDP), declined by about 18 percentage points of GDP in 2010/11 because of the global slowdown and the fall in South African imports. In total SACU revenue fell by US$322 million25 (see Figure 1), and the government moved from a position of surplus in 2008/09 to a deficit of 10.3 percent of GDP in 2010/11.26 The situation has since improved, and with an anticipated doubling of SACU revenues in 2012/13 (due in part compensation for underpayments in previous years); nonetheless the volatility of SACU transfers has made budget management challenging.27 25 Based on decline from M.4,918 million in 2009/10 to 2,627 million in 2010/11 (MoFDP, 2011b). 26 MOFDP (2011b). 27 See for example World Bank (2013b). 9 21. With heavy fixed expenditure commitments, when SACU revenues collapsed the government found itself in an extremely tight fiscal situation and was faced with the need to undertake a significant rationalization of expenditures. This is thus generally not a good time to be considering the expansion of programs, including safety net transfers, and any new initiatives will have to be financed by rationalizing existing spending. The government’s overall strategy in response to the crisis has been to: (i) try to contain recurrent spending; (ii) reduce the burden of the wage bill over time; and (iii) increase the share of investment in the budget, while at the same time improving the quality of spending on investments. Figure 1: SACU Revenue, 2000/2001 to 2012/2013 (M. millions) 7000 6000 5000 4000 3000 2000 1000 0 Source: MoFDP (2010) Note: Data for 2010/11-2012/13 are estimates. 22. In response to the fiscal pressures and volatility of SACU transfers, the Government has embarked on an ambitious fiscal consolidation program. While this has had success in closing the fiscal gap, there seems little prospect for a return to the very expansionary spending of the past. Under these circumstances, the government needs to be cautious about adopting any new transfer programs that will involve permanent recurrent spending commitments, and the case is strong for rationalizing existing spending on transfers. At the same time, there are strong arguments for providing targeted transfers to those most in need. The unequal distribution of income and the nature of poverty continue to argue in favor of a broadly redistributive strategy, which would support the consumption of the very poorest at an acceptable level (see further discussion in Chapter 5). 10 23. Recognizing this, the government has ring-fenced spending on social services (including social protection) in the adjustment process. In addition, the recovery program that the government has agreed with the IMF includes a floor for social spending. However, this does not obviate the need to strictly discipline existing social spending. 24. The implications for transfers and safety nets strategy moving forward should be to:  Select the most cost-effective instruments.  Do so in a way that is fiscally sustainable.  Wherever possible, use existing expenditures to minimize the need for new spending.  Narrowly target those aspects of poverty for which transfers are the most appropriate solution (this is likely to include some forms of categories related to HIV/AIDS, child-headed and grandparent-headed households, early childhood malnutrition, and seasonal food poverty as well as geographically targeting of pockets of poverty).  Ensure that to the extent any possible programs and investments that can be scaled up and down in response to evolving fiscal conditions and changing needs are indeed expanded and contracted accordingly rather than becoming permanent entitlements. The possible shape of such a program is discussed in Chapter 5. 25. Expenditure on transfers is dominated by three large programs and has been growing in recent years. Total expenditure rose in 2009/10 because of increases in expenditure on the Tertiary Bursary Scheme and the Old Age Pension due to increased benefit levels. Spending on the minor programs that are targeted to the poor has also been rising – primarily as a result of the introduction and scaling up of the Child Grants Program – but these programs do not account for a significant amount of public spending compared 11 with the larger schemes. As indicated in Table 2, the larger programs represent over 90 percent of overall spending on transfer programs while the smaller, poverty-targeted ones account for only 7.4 percent. Table 2: Trends in Spending on Major Transfer Programs (M. millions) 2007/08 2008/09 2009/10 2010/11 % of spending on transfers 2010/11 Old Age Pension 205 208 320 371 29% School Feeding Program 153 174 202 236 18.5% Tertiary Bursary Scheme 372 390 625 575 45% Sub-total 730 772 1147 1182 92.6% Smaller Programs Agricultural Input Fairs n.a. n.a. 59 17 1.3% OVC Bursary Scheme 44 33 22 48 3.8% Child Grants Program n.a. 16.5 16.5 16.5 1.3% Public Assistance n.a. n.a. 13 13 1% Sub-total 44 49.5 110.5 94.5 7.4% Total 774 821.5 1257.5 1276.5 26. There is scope for re-orienting expenditures on transfers away from those that benefit the well-off towards those that would more directly benefit the poor . Obvious examples of the former include the Tertiary Bursary Scheme, (M.575 million or US$76.7 million annually) and the National Fertilizer and Input Subsidy (M.44 million or US$5.9 million annually). 27. As will be discussed in Chapter 3, a lot is already being spent on transfers of one kind or another (approximately US$197 million per year). This represents about 16 percent of public spending or 9 percent of GDP.28 However, the total amount spent on safety net programs is lower, representing about 8 percent of public expenditure and about 4.6 percent of GDP, which is still high compared to many other countries. Table 3 presents a list of estimated expenditure as a percentage of GDP in a number of other African countries. Although there is no one ideal level of expenditure for safety nets (since it depends on 28 Based on 2010/11 figures of M.9,381 million total expenditure (US$1,251 million) and a GDP of M.16,310 million (US$2,175 million). 12 variables such as the size of the population in need, the efficiency of the programs, and the government’s ability to finance them), this table provides an indication of how much is being spent in other African countries and demonstrates that Lesotho is among the countries with the highest expenditures.29 Table 3: Social Safety Net Expenditures in Selected African Countries Country/region Year (% of GDP) Source Malawi 2003-2006 6.5 World Bank (2007) Lesotho 2010/11 4.6-9 This report Botswana 2010 2.2 World Bank (ongoing) Government of Swaziland and World Bank Swaziland 2010/11 2.2 (2012a) Mozambique 2010 1.7 World Bank (2011a) Kenya 2010 0.8 World Bank (2012b) Tanzania 2010 0.3 World Bank (2011d) Zambia 2010 0.23 World Bank (2013a) Source: Authors’ compilation. 29 As indicated in Grosh et al (2008) “Because indicating precisely what countries should spend on safety nets based on theory is so difficult, analysts often seek guidance based on benchmarks. Benchmarking is, of course, imperfect, as there is nothing to indicate whether the countries used in the benchmarks are spending the ‘right’ amount.” 13 CHAPTER II: POVERTY AND SAFETY NETS 28. Lesotho has historically been characterized by intense rural poverty due to low productivity, harsh terrain, poor soils, minimal integration of these areas with markets, and population pressures. In the past, the solution to this poverty has consisted of a combination of: (i) urbanization; (ii) out-migration from Lesotho; (iii) transformation from agriculture to the modern sector by virtue of Lesotho’s proximity to the South African economy and, more recently, the influx of textile producers;30 and (iv) employment in the mines of South Africa. All of these provided a release valve from the pressures on a weak agrarian base. Nonetheless, significant deep rural poverty remains. Also, in recent years, these escape valves have functioned less effectively, with declining employment in the mines of South Africa, slowing modern sector growth due to the global economic downturn, and weak performance in agriculture. 29. The analysis in this report is based on a wide range of sources that, taken together, paint a comprehensive picture of the nature and type of deprivations faced by the poorest in Lesotho (Box 2). According to preliminary analysis of the 2010/11 HIES, about 57 percent of the population fell below the basic needs poverty line, and were classified as poor. Yet, for policymakers, noting that more than half the population is poor is not very helpful. For the purposes of defining a safety net strategy, it is necessary to distinguish those people who are destitute and in need of regular cash transfers, those who need cyclical assistance, and those who can be helped to rise out of poverty with other forms of access and assistance. 30. While the most recent data (from the 2010/11 HIES) suggest that the overall poverty rate has barely changed since 2002/03, there are several underlying factors suggesting that 30 In May 2000, the U.S. Congress approved legislation known as the African Growth and Opportunity Act or AGOA. The purpose of this legislation was to assist the economies of Sub-Saharan Africa and to improve economic relations between the United States and the region. 14 conditions may have deteriorated for some of the poorest Basotho in the short term. These include: (i) an apparent collapse in agricultural production (for example, average maize yields have decreased by almost 50 percent from the levels of the early 2000s, while the size of cattle herds has stagnated or declined substantially since 2002/03; (ii) little growth in the labor-intensive sectors; and (iii) inflation of about 50 percent in food prices (between 2007 and 2011), which appears unlikely to have been matched by increases in households’ purchasing power, especially among the poor and those only marginally attached to the cash economy. At the same time, there has been a continuing shift from subsistence farming to wage employment in urban areas, which should help to reduce poverty. Therefore, a stagnant picture of poverty is emerging from preliminary analysis of the 2010/11 data, with an accompanying sustained high inequality that implies greater relative impoverishment of those at the poorest end of the income distribution. 31. The high levels of poverty in Lesotho do not seem consistent with the relatively high average national income (for example, in Ethiopia, a far poorer country, 39 percent of the population is estimated to be “poor”31). However there is high inequality in Lesotho and many Basotho are undeniably living in conditions of extreme deprivation. The balance of this chapter is concerned with exploring in more depth whether there is some way of identifying these distinct groups of ultra-poor, how many of these people there are in various groups that could benefit from safety net programs, and whether there are particular aspects of poverty that it might make sense to tackle using transfer programs that promote productive activities. 31 Based on Ethiopia’s 2004/05 Household Income and Consumption Expenditure Survey HICE and an average for 1994 to 2008 as reported in the MDGs. This came at a time when per capita GDP in Ethiopia was only US$189, about one-sixth of Lesotho’s. Part of the explanation of course is that poverty is a relative concept and varies with the accepted consumption norms of a society, and the cost structure of an economy, which determines how much is needed to purchase the basic basket of necessities. Yet another part of the explanation has to do with the highly unequal distribution of income in Lesotho (discussed in section 2.2), as a consequence of which a significant share of the population is living at extremely low income levels despite the relatively high average national income. 15 2.1 Poverty Profile Income Poverty: What Do We Mean by Poor? 32. To arrive at a meaningful estimate of how large a safety net transfer needs to be to be effective, it must be evaluated relative to the poverty line income and to the current (pre-transfer) incomes of the poor. The poverty lines used in the 2010/11 HIES are presented in Table 1. The basic needs poverty line – the amount required to satisfy food requirements and consume a basket of essential non-food items – was equivalent to M.243 per adult per month. Those with consumption below this level are classified as poor. The food poverty line is the minimum amount required to satisfy the daily energy needs of an individual, so households living below this level are considered to be ultra-poor. Table 4: Lesotho - Estimated Poverty Lines 2010/11 (Maloti per adult equivalent per month) National Food Poverty National Basic Needs Line Poverty Line Estimated 2010 137 243 US$/day Equivalent $0.61 $1.08 Sources: Preliminary government estimates based on the 2010/11 HIES. Notes: The poverty lines used for preliminary analysis were expressed in 2002/03 prices to maintain comparability with 2002/03 analysis. In 2002/03 prices they were M.84.4 (Food) and M.149.4 (Basic Needs) per adult equivalent per month. These were updated to 2010 prices for this table in consultation with the Bureau of Statistics using CPI changes since 2002/03. Note consumption is defined throughout in adult per capita equivalents, which affects the levels of support that might be needed under safety net programs. Programs need only provide an amount that covers the adult equivalent size of a household, not its total number of members. For adult equivalence scale, see annexes 1.5 and 1.6. The average household consists of about four adult equivalents. How Many Poor People Are There? 33. Table 5 shows the poverty levels and number of poor people living below each of these different poverty lines as of 2010/11. 16 Table 5: Estimated Number of Poor People in Lesotho National Food National Basic Needs Poverty Line Poverty Line Poverty Headcount 34.1% 57.1% Estimated number of poor (2010/11) 648,000 1,083,000 Sources: Preliminary government estimates based on the 2010/11 HIES, and author’s calculations based on population of 1.9 million. 34. If scarce resources are to be used to finance transfer to the poor, it is important that they be targeted to the most destitute. Therefore, it is worth examining in more detail how truly poor the poor are in Lesotho. Figure 1 shows the distribution of consumption and clearly illustrates that the incomes of the poorest 20 percent of the population are substantially below those of the next poorest 20 percent. This suggests that there is a group of ultra-poor whom it is worth considering as the target group for some form of social assistance. Unfortunately, however, the data are not nuanced enough (nor up-to-date enough) to say with certainty whether the group in dire need represent, say, 10 percent of the population or – as the preliminary processing of the HIES 2010/11 suggests – almost a third of the population.32 To make this distinction, it is necessary to turn to other evidence. 35. There are only limited data on food poverty and extreme poverty in Lesotho. A 2007 nutrition survey estimated about 7 percent of the population had inadequate food consumption.33 The Lesotho Vulnerability Assessment Committee makes an annual estimate of the number of people in need of humanitarian assistance.34 This number has averaged about 650,000 over the past six years, but it varies a lot from year to year because it measures the number affected by natural disasters and temporary shocks (such as drought) as well as those in chronic food insecurity. An informal estimate by the World Food 32 Unfortunately the 2002/03 HIES was not analyzed by decile but only by quintile, thus depriving of us an additional degree of precision in examining the differences among the very poor. 33 FNCO (2007). 34 The Lesotho Food Security and Vulnerability Report is generated annually by the Lesotho Vulnerability Assessment Committee (LVAC) encompassing the Disaster Management Authority, the Bureau of Statistics, and the World Food Programme. 17 Programme (WFP) puts the number of people who are habitually food-poor at around 200,000, which would imply about 10 to 15 percent of the population. Figure 2: Estimated Distribution of Consumption in Lesotho, 2010/11 Average household consumption per adult equivalent 2010/11 Maloti per 600 400 200 month 0 Poorest 2 3 4 Richest Consumption Quintile - from Poorest to Richest Source: Preliminary government estimates based on the 2010/11 HIES. How Poor Are the Poorest? 36. Table 6 shows estimates from 2002/03 and 2010/11 of the depth of poverty in Lesotho. The depth of poverty (or the poverty gap) measures how far, on average, the incomes of the poor fall below the poverty line.35 Table 6 tells us that on average the incomes of the poorest were about 15 percent below the food poverty line in 2010/11, and this gives us an aggregate measure of how much would be required to bring them up to a basic minimum level of food consumption. Table 6: Depth of Poverty 2002/03 and 2010/11 Depth of Basic Needs Poverty Depth of Food Poverty 2002/02 29.0 11.2 2010/11 28.7 14.9 Source: Preliminary government estimates based on the 2010/11 HIES; for 2002/03: National Human Development Report (UNDP (2007) Table 3.1.1 p.23; based on HIES. 35 The poverty gap (depth) measures the mean aggregate income or consumption shortfall relative to the poverty line across the whole population. It is obtained by adding up all the shortfalls of the poor (considering the non-poor having a shortfall of zero) and dividing the total by the population. 18 37. According to these preliminary estimates the depth of poverty has been stagnant over time, suggesting that the poor on average are no better off in relative terms than they were in 2002/03. However the depth of extreme poverty has increased, suggesting that the very poorest may in fact be even worse off. What Do We Know About Current Incomes? 38. According to preliminary results from the 2010/11 data, the average household consumption in rural areas was M.857 per month,36 which, based on an adult equivalent household size of about four, would imply an average per capita income of about M.210 per month (93 cents per day), or roughly equivalent to the calculated poverty line income. 39. Data on casual wage rates collected recently as part of Lesotho’s annual Vulnerability Assessment process revealed that the daily rate for unskilled agricultural labor is about M.20 for a five-hour workday (about US$2.50).37 Someone working for 20 days a month at this level would earn about M.400 per month. Based on this recent evidence, the average income of a typical rural Basotho seems to be around the basic needs poverty line. 2.2 Basic Characteristics of the Poor and the Poorest 40. The characteristics of poor households such as their size and composition and the assets that they own can be useful in targeting the ultra-poor and can reveal whether they have any particular attributes that it might be cost-effective to tackle with safety net interventions. 41. In terms of demographic characteristics, very poor households tend to be larger than non-poor ones and – as might be expected – to have higher dependency ratios as they 36 Preliminary estimates by the Bureau of Statistics from the 2009 CMS as reported in Kardan et al (2011). 37 LVAC/DMA (2011), p.13. 19 contain larger than average numbers of elderly and children. The proportion of female- headed households is no greater among the very poor than among the population as a whole. The NISSA registration system found about 8 percent of the households in the areas surveyed consisted only of elderly people and/or of elderly people and children (such as grandparent-headed households). The system does not contain any data on the poverty status of these households, but given that most were not selected for the Child Grant Program by a proxy means test or community targeting, this implies that they were not found to be poorer than the average.38 Table 7: Demographic Characteristics of the Poor, 2010/11 Non-poor Poor Very Poor Household Size 4.1 5.5 5.7 Dependency Ratio 0.78 1.07 1.10 Average Number of Children < 16 1.38 2.05 2.17 Average Number of Elderly >59 0.36 0.49 0.51 Average Age of HH Head 50 54 55 % of Female-headed Households 37.6% 36.6% 37.3% Source: Preliminary government estimates based on the 2010/11 HIES. 42. Differences in asset ownership between the very poor and the rest of the population can be used to help to identify those households that are in need of transfers, especially in situations where there is no information on household income. Unfortunately in Lesotho there are surprisingly few differences between the poor and the non-poor in the ownership of key assets, except for obvious high-end items such as cars and electrical goods (see more in Annex 1). The value of the household survey data is limited since it reports the proportion of households owning each asset and not the actual quantity of the asset that they own and, in many cases, the assets it measured were those that only tend to be associated with urban households or those receiving significant remittances. 38 Ayala Consulting (2012) page 8. However, this system has so far only been implemented in selected poor areas. 20 2.3 The Geographic Dimensions of Poverty 43. Understanding the geographical distribution of poverty helps policymakers to determine where to target safety net interventions and to assess the appropriateness of the current geographic spread of programs. Poverty in Lesotho is concentrated in rural areas (where 88 percent of the extreme poor live), and its character varies depending on which of the four distinct geographic areas of the country people live in. 44. Livelihoods in all of the four rural zones are a combination of low-level agriculture and remittances from family members engaged in the urban cash economy, with a focus on livestock grazing in the mountains and a combination of cropping and livestock elsewhere. The conventional wisdom is that extreme poverty is concentrated in the mountain regions, although the poverty data do not show an overwhelming concentration in any one zone. Inequality, as estimated in 2002/03, was almost identical across all of the zones.39 Table 8: Poverty Indicators by Geographic Zone Geographic Zone Headcount Estimated Estimated “Very Mean Annual Poverty Number of Poor” 2009 HH a/ c/ (2010/11) Extreme Poor Population Expenditure b/ (2011 LVAC) Urban 0.18 76,300 n.a. 20,131 Rural Lowland 0.38 278,300 122,414 13,964 Rural Foothills 0.42 105,500 46.011 7,210 Rural Mountain 0.35 147,500 36,365 11,167 Senqu River Valley 0.32 30,100 25,467 10,815 Notes: a/ Author’s calculations based on rough approximation of current population of 1.9 million, and preliminary estimates of the 2010/11 HIES food poverty headcount ratios and population distribution as given in World Bank 2010b, Table A2.5 p.206. b/ LVAC/DMA (2011), p.7-8. c/ Data from 2009 CWMS in BoS (2010b), Table p.5.4, p.34. d/ 0 .17 in Maseru; 0.25 in other urban areas. 39 With a gini coefficient of about 0.52 in all areas (both urban and rural). Based on 2002/03 HIES data in Table 3.8.2 of the National Human Development Report (UNDP, 2007). 21 45. The tables in Annex 1 show the breakdown of poverty, malnutrition, and food security indicators by Lesotho’s nine districts, which have an average population of about 200,000 each. No clear pattern emerges – in part because district boundaries are administrative constructs and do not necessarily correspond to distinct poverty or economic zones. Districts that have high food poverty are not necessarily the same as those with the highest poverty levels or the worst malnutrition rates, and some areas with relatively low poverty rates still have large numbers of very poor people because they are more heavily populated. The implication of this is that district-level targeting alone will not be sufficient, so that any geographical targeting will have to be at a more micro level. Using the 2010/11 HIES and data collected under NISSA to undertake a more nuanced assessment of local level poverty rates should be a priority. 22 Figure 3: Map 1-Lesotho’s Ecological Zones 2.4 Livelihoods of the Poor 46. Lesotho is essentially two countries – at least in a poverty-economic sense – one that consists of those who participate in the cash economy through wage employment, outmigration, and remittances and another that consists of those who rely on a low- productivity, low-return agricultural activities. 47. The poorest in Lesotho still tend to rely more on agriculture than the non-poor, but, unlike in most of the rest of sub-Saharan Africa, they are not overwhelmingly subsistence farmers. According to the 2009 Labor Force Survey, about 27 percent of the economically active populations are employed in the cash economy, about 14 percent are subsistence farmers, 11 percent are casual workers, and 12 percent are either self- 23 employed or work without pay within their family. The remaining 29 percent are unemployed. Many of the unemployed are young people who are transitionally unemployed between leaving school and starting a family, while others are chronically unemployed, having left subsistence agriculture but not having been absorbed into wage employment. While data from 2002/03 HIES showed that 23 percent of the labor force defined themselves as unemployed, a World Bank study from 2010 argued that only 5 percent actually were unemployed in a more traditional sense.40 Culturally the meaning of unemployment in Lesotho often refers to the lack of formal employment, and thus many farmers and self-employed are likely to define themselves as unemployed while in fact they are working in the informal sector. 48. The poor inevitably engage in multiple activities to compensate for the limited opportunities and under-employment in farming, which is highly seasonal and often does not require labor for much of the year.41 According to Lesotho’s annual Food Security and Vulnerability Assessments, rural households classified as “poor” or “very poor” rely on casual labor for up to 40 percent of their incomes, and the daily wage rates for such work have remained stagnant for the past several years. Table 9: Main Sources of Income by Poverty Status Main Source of Income Very Poor Poor Non-poor Farming 39.3% 29.3% 17.8% Public Sector 5.4 % 9.4% 17.6% Private Sector 19.7% 25.3% 31.5 % Remittances 11.2 % 12.0 % 10.0 % Non-farm HH Business 7.4 % 9.3 % 9.8 % Pensions 2.0 % 2.2 % 2.5 % Other 15.0 % 12.6 % 10.8 % Sources: 2006 National Human Development Report UNDP (2007) p.34. 40 World Bank (2010b), p.54. 41 For example, only 10 percent of women and 34 percent of men engaged primarily in agriculture report that they have year-round employment (the rest being either seasonal or occasional), whereas 73 percent of those involved in non-agricultural work were employed all year round (2009 DHS, in MHSW, 2010, Table 3.7.1 p.40). 24 49. Agriculture in Lesotho generally results in low productivity because of the poor soil, unpredictable rainfall, and difficult topographical conditions. Overall agricultural production has been stagnant or declining over the past decade,42 with declining yields and little growth in livestock herds. There has been some increase in commercial agriculture, but this has had little impact on the poor as it is not particularly labor-intensive and tends to take place in higher-productivity areas where the poor are not concentrated. Figure 4: Composition of Farm Household Income 6.3% 19.6% Subsistence Farming Cash Crop Livestock 5.0% 57.8% Mine Remittance Wage/Salary 4.0% 1.8% Pension Fund Source: Author’s calculations from 2009/10 Agricultural Census (BoS (2012b), Statistical Table 10 50. While relatively few of the poorest are completely landless, their land is often of low quality and/or they lack the labor to cultivate it or the cash income to purchase fertilizer and other inputs.43 In the mountains, wealth and agricultural activity revolve primarily around livestock ownership. However, herd sizes tend to be small (for example two-thirds of farming families that own cattle own fewer than five cattle).44 While many of the poor in mountain regions own livestock, the ownership of livestock does not appear to protect 42 BoS (2010c) p.19-23 and MAFS (2010 and 2011) 43 See discussion in Annex 1. 44 MAFS (2010), Table 29. 25 them against consumption poverty. This is in part because livestock are regarded as a critical store of wealth and are only very infrequently sold or slaughtered.45 Thus, the poor may have relatively large livestock holdings but still have extremely low consumption levels. In addition, very high levels of theft lower the returns to owning livestock, and livestock management is very labor-intensive. Some of the poorest people in Lesotho are the many young men engaged as cowherds. Non-agricultural Employment and Incomes 51. In general the very poor tend not to participate in wage employment in the modern sector. However, they do have substantial links with the sector, mainly in the form of remittances from family members who migrate for employment (see discussion below) and, more recently, through the employment of poor women in textile factories. In 2002/03, 19.4 percent of rural households in the poorest quintile reported having a textile worker in their household.46 52. There has been a continuing shift in Lesotho over the past two decades from a subsistence economy dependent on remittances to a middle-income exporter of natural resources and manufactured goods.47 This has been accompanied by a sustained migration from rural to urban areas, which means that working for pay has become increasingly common, especially among younger Basotho. This has almost certainly led to a reduction in extreme poverty for some households, although recent developments may have contributed to deterioration in the past few years. 45 See, for e.g., “The ‘Bovine Mystique Phenomenon” on p.35 in UNDP (2007) 46 World Bank (2010b), Table 3.17, p.80 47 These exports consist primarily of textiles, diamonds, and water resources. 26 2.5 Food Insecurity 53. The extent to which people do not have enough to eat is a very good measure of the need for safety net transfers. If the food poverty line is an accurate measure of the consumption of a minimum basket of calories,48 then in 2010/11 just over one-fifth of the population did not have enough to eat. Estimates based on the LVAC system vary widely from year to year (Figure 3 and Annex 1) depending primarily on the weather. In 2010/11 there was a significant decline in grain production due to flooding, and a poor harvest was experienced again in 2011/12 because of late and insufficient rains, resulting in a short-term crisis in food availability affecting an estimated 514,000 people or 27 percent of the population. As a result, the Prime Minister recently declared an emergency food crisis in Lesotho and appealed for support from development partners. 54. The LVAC estimates however are not based on actual measurements of household consumption. The 2007 nutrition survey estimated that about 7 percent of the population, or 123,000 people, had “poor” adequacy of food consumption, while about one-third had “borderline” food inadequacy.49 The best guess of people working on food issues in Lesotho is that about 200,000 people (or about 10 to 15 percent of the population) – mostly in the mountains – are routinely chronically short of food.50 The figure of 200,000 is also the lowest level of food insecurity measured by the annual Vulnerability Assessments, suggesting that it represents a minimum floor of the number of food-insecure. 55. Unlike in most of southern and eastern Africa, food availability and hunger in Lesotho do not depend primarily on the level of domestic grain production in a given year . Almost two-thirds of the food consumed in Lesotho is imported from South Africa. On the one hand, this means that supply is more reliable and grain is almost always available. On 48 In 2002/03 the caloric definition of the food poverty was 2700 kcal per adult (p.95, BoS {2006}.) 49 FNCO (2007). 50 Personal communication to the authors from WFP staff. 27 the other hand, however, it means that people need to have enough cash to buy it, and this is a particular challenge for the poor. As Table 10 shows, the poor tend to produce less of their own food than do the non-poor and thus to rely more on purchases from the market. The implication of this is that they are more susceptible to price shocks and that agricultural interventions (such as input programs) will be less relevant for the poorest than in other parts of southern and eastern Africa; while cash transfers, and/or measures that stabilize consumer food prices may have more impact on the poor (although given that prices are determined by the South African market, there is relatively little scope for price-related interventions). Table 10: Calories from Own-Production by Wealth and Location, 2004 Poor Middle Well-Off All Lesotho 21% 40% 49% Foothills (Most among the poor) 26% 36% 49% Mountains (Least) 14% 42% 55% Source: World Bank (2010 b) Table 4.3 p.103). 2.6 Health, Nutrition, and Education 56. Overall indicators for public health have been declining in Lesotho due in part to the HIV epidemic, which has dramatically reduced life expectancy levels. The high prevalence of chronic poverty and food insecurity is influencing and being aggravated by the pervasive and growing HIV epidemic in the country (see the relevant sub-section below). Life expectancy at birth has declined drastically in Lesotho from 60 years in 1991 to 48 in 2011 (see Annex 2). Under-five mortality rate (per 1,000 live births) was estimated to be as high as 86 deaths per 1,000 live births in 2011 (almost the same as in 1992).51 57. While costs represent a barrier to using health services by the poor, the number one reason for people not seeking treatment (in all income groups) is not the ability to 51 World Bank Data Bank (http://data.worldbank.org/country/lesotho). 28 pay, but the expected unavailability of services and drugs at health facilities . This became clear from the most recent DHS in 2009. The poor also suffer from the fact that they tend to live in more remote areas where there are fewer services and service providers. 58. Lesotho suffers from high levels of child malnutrition for a country at its income level. This is partly a function of the unequal distribution of income. As can be seen in Table 11, malnutrition declines as wealth increases but it is nevertheless a problem for households at all income levels. This suggests that it is related not only to income and food consumption but also to dietary composition, feeding practices, and poor retention of nutrients due to parasites and disease. Although there is some regional variation in malnutrition, it is not dramatically higher in one area than the others and thus does not provide a pointer to the sort of severe geographical food poverty that would help policymakers to target safety nets.52 Table 11: Nutritional Status of Children by Wealth Quintile, 2009 Height-for-age Weight-for-height Weight-for-age Severe Moderate Severe Moderate Severe Moderate stunting stunting wasting wasting undernutrition undernutrition Wealth quintile st 1 18.9 45.6 2.3 5.6 4.4 17.8 nd 2 17.4 46.9 1.4 3.0 2.3 13.5 rd 3 14.5 40.3 0.9 3.1 1.3 11.9 th 4 11.7 28.7 0.7 3.1 1.7 11.5 th 5 8.5 28.3 2.0 3.8 1.4 9.2 All Lesotho 14.8 39.2 1.4 3.8 2.3 13.2 Source: Data from 2009 DHS in MHSW (2010) Table 11.1, p. 133. 59. Most damage from malnutrition occurs in the first year or two of life but has lifetime effects on productivity and well-being. Evidence worldwide shows that improved nutrition in childhood increases an individual’s years of schooling and school outcomes and can raise his or her long-term productivity and lifetime earnings. Addressing early childhood malnutrition therefore has long-term effects because it can break the poverty-malnutrition 52 Data from 2009 DHS (MHSW, 2010). See also Annex 1 tables on malnutrition by district. 29 cycle. Cash transfer programs can be used to encourage behavior that will lead to decreased malnutrition rates.53 Education 60. Lesotho has a fairly well-educated population although challenges remain in terms of achieving universal primary education and increasing access to the secondary level. Primary participation rates have increased significantly since 2000 when the government introduced its free primary education policy, but net enrollment rates were still only 87 percent for boys and 94 percent for girls in the poorest quintile in 2009 compared with 98 percent for boys and 97 percent for girls in the richest quintile (see Annex 2). Repetition and dropout rates are particularly worrisome for boys, and this takes its greatest toll on the poor. For instance, boys in the lowest quintile had a dropout rate of 5.1 in 2004 compared to 0.8 for the richest quintile (see Annex 2). Poor boys in rural regions have particularly low attendance rates, which are partially explained by the “herd boy” syndrome in which young boys are pulled out of school to watch their family’s cattle, especially in the mountainous regions. 61. Although secondary enrollment rates have increased from 21 percent in 2001 to almost 30 percent in 2010, secondary education is still effectively a privilege enjoyed only by the better-off. Only 7 percent of boys and 15 percent of girls from the poorest quintile attended secondary school, while 60 and 62 percent respectively from the wealthiest quintile did so.54 Such high discrepancies in attendance rates reinforce current income inequalities and make it even more difficult to achieve more equitable growth in the future. 53 Some cash transfer programs have successfully linked the receipt of benefits to nutrition interventions, including growth monitoring, regular health center visits by pregnant women and children under the age of 5, the provision of micronutrients, and nutrition education. An example where a country has attempted to use a cash transfer program to achieve better nutrition outcomes is the Juntos program in Peru and another example is the Nutrition Enhancement Program in Senegal (see further discussion under the malnutrition program in chapter 3). 54 MHSWS (2010). 30 A significant reason for poor attendance is the fees charged for secondary education, which was identified as the main obstacle by over 60 percent of males not attending secondary school. However the lack of education services in remote areas where the poor tend to live combined with lower parental education and literacy levels in those areas makes it more difficult for the poor to access the limited number of secondary school places available. Finally, the inequalities are aggravated at the tertiary education level where only 1 percent of students come from the poorest quintile. Despite this, the government continues to heavily subsidize tertiary education with bursaries and loans that are generally not repaid. As only the wealthy households make it through the secondary school level, such disproportionate spending on tertiary education subsidies is highly regressive. 55 2.7 Some Specific Groups of the Poor Orphans and Vulnerable Children 62. Lesotho has suffered a crisis of orphanhood, largely as a result of the HIV/AIDS epidemic. It is estimated that some 28 percent of children under the age of 18 have lost one or both parents, and about 8 percent (or some 32,000 children) have lost both parents.56 However, not all orphans are poor. While some are living in conditions of extreme destitution, others have been absorbed into well-functioning households that can support them. 63. The available data show that there is little discernible relationship between orphanhood and poverty (Table 12). This is a pattern that is common globally, and in these circumstances, targeting benefits on the basis of being an orphan alone is not a rational strategy. The challenge is to distinguish very poor orphans from orphans in general, and 55 See discussion of tertiary bursaries in Chapter 3. 56 Based on a total of 339,000 children according to the Bureau of Statistics’ OVC survey. Note that estimates of the number of orphans vary widely, from a low of about 138,000 (Bureau of Statistics OVC Survey; (BoS (2011b) to a high of 365,000 (DSW, 2011). 31 doing this requires local knowledge (suggesting some form of community targeting) and/or the use of other categorical attributes that may indicate the orphans’ degree of vulnerability. For example, some 3.4 percent of children are estimated to live in households headed by an isolated elderly caregiver, and about 0.3 percent are estimated to live in child- headed households.57 The recent Situation Analysis by UNICEF and MHSW (MHSW (2011) also found no significant differences in asset ownership between households with orphans and those without, confirming that orphanhood alone is not a good basis for targeting the poor.58 Table 12: Orphanhood by Wealth Quintile Wealth Quintile % of Children Who Have Lost One or Both Parents Poorest 27.9 % nd 2 27.8 % rd 3 30.6 % th 4 26.8 % Wealthiest 24.8 % Source: Data from 2009 DHS in MHSW (2010) Table 2.3 p.12. 64. Because there are many non-poor orphans and many other children who are not orphans but are nonetheless living in extreme poverty and vulnerability, Lesotho has adopted the concept of “orphans and vulnerable children” (OVCs). Vulnerable children are defined as being an orphan (having lost one or both parents), or being HIV positive, or living without adequate adult support.59 However, this categorization still includes many children who do not necessarily come from poor households. 57 DSW (2011) 58 DSW (2011) 59 Note this also includes children living in households that have a chronically ill parent, have experienced a recent death from chronic illness, or are grandparent-headed (DSW, 2011). 32 People Living with HIV/AIDS (PLWH) 65. The HIV prevalence rate was 23 percent in 2009,60 the third highest in the world. By this estimate, some 145,000 people are living with HIV/AIDS in Lesotho.61 While HIV/AIDS is a national disaster and has dramatically reduced life expectancy and national productivity, households affected by HIV/AIDS do not appear to be poorer on average than those who are not affected (Figure 5). Figure 5: HIV Prevalence among 15-49 Year Olds by Wealth Quintile, 2009 35 30 25 20 Women 15 Men Total 10 5 0 Poorest 2nd 3rd 4th Wealthiest Source: Data from 2009 DHS in MHSW (2010) Table 13.4 p.205. 66. It is difficult to assess the relationship between HIV/AIDS and consumption poverty. Clearly the loss of household labor (for example in agriculture) or of a breadwinner can impoverish a family. At the same time, many PLWH come from families who can support them or are receiving ARTs and continue to live productive lives. The situation in Lesotho is similar to that in many other countries where AIDS is often a disease of the non- poor (because the poorest tend to live in villages and to be less mobile and less socially interactive than the urban and middle-class populations). This does not mean that there are 60 Data from 2009 DHS in MHSW (2010) Table 13.4. 61 Based on 23 percent prevalence and the fact that 32.9 percent of the population are aged 19 to 49 in a total population of 1.9 million (data from 2009 DHS in MHSW, 2010, Table 2.1). 33 not people suffering from HIV/AIDS who are in desperate situations and need safety net support. The issue from a targeting point of view is that they need to be identified on a case-by-case basis and that providing transfers to all people who are HIV-positive is not a useful poverty reduction measure. The Elderly 67. The proportion of elderly is higher in Lesotho than in other Sub-Saharan African countries, reflecting the outmigration of young people and beginnings of a demographic transition towards an older society. In Lesotho, 8.3 percent of the population (about 165,000 people) are over the age of 65, and 4.1 percent, or some 82,000 people, are over the age of 70. 68. Since consumption is measured at the household level, it is difficult to say whether the elderly are on average poorer than the population as a whole. In many countries, the better-off tend to live longer than the poor, which would imply that a smaller portion of the elderly are from poor households, but this may not be true in Lesotho due to the impact of the AIDS epidemic. Figure 6: Distribution of Poverty by Age Group (% of Total Poor) 6% 40% 54% Children aged 0-15 Adults aged 16-64 Elderly aged 64 and above Source: World Bank (2010b), Figure 1.1, p. 3. 34 69. Poor households tend to have more elderly members in them than average, so targeting transfers to the elderly will tend to reach slightly more poor households than would random targeting. As we have seen in Table 7, the average age of the household head is slightly higher in poor families than in non-poor families (55 compared to 50), and the estimated food poverty rate in households containing people over age 59 (39.3 percent) is somewhat higher than in the population as a whole (34.2 percent).62 At the same time, the elderly make up a fairly small proportion of the poor (Figure 6), so targeting them alone would miss most of the poor. For example, if people over 70 make up 4 percent of the population, then the Old Age Pension can indirectly benefit at most only about a quarter of the population.63 70. Many of the elderly live in non-poor households where they are adequately catered for by relatives, while at the same time others live on their own in desperate conditions. The challenge for policymakers is to identify those who are poor and vulnerable. As with PLWH and orphans, this probably can only be done using local knowledge through some form of community targeting. 64 The Disabled 71. In 2006 it was estimated that about 69,000 people in Lesotho were disabled, representing about 3 percent of the population.65 Of these, most either suffered from mobility problems or were partially or fully blind.66 There is no information on the poverty status of the disabled. As with the other categorical groups examined here, some of the 62 Preliminary government estimates based on the 2010/11 HIES. 63 Based very roughly on a maximum household size of six for households with elderly members. 64 The discussion here relates only to those elderly who are extremely poor, not supported by households, AND are not yet old enough to receive the Old Age Pension. 65 This section draws primarily on the survey of disability carried out as part of the 2006 Census and reported on in Chapter 6 of BoS (2009a). 66 Note that about 18,000 or about a quarter of the disabled were above the age of 65, and thus are also included in the category of “the elderly.” 35 disabled are able to function well economically or live in households that can adequately support them, while others are alone and destitute. There are some proxy indicators that shed light on the status of the disabled. For example, the Census results show that about half of all disabled adults are married. Their education levels tend to be somewhat lower than those of the population as a whole67 and, not surprisingly, they are less economically active than average. 2.8 Shocks and Movements In and Out of Poverty 72. The poor, and in fact all Basotho, are exposed to multiple shocks . Understanding how much and which elements of extreme poverty are due to shocks is crucial for determining the extent to which a safety net strategy should be attempting to address short-term falls in consumption as opposed to chronic insufficiency of incomes. 73. The main shocks, as ranked by the poor in village interviews, include theft (primarily of livestock), unemployment, drought, price increases, and illness and death due to disease.68 In Lesotho’s rainfed agrarian economy, the Basotho face regular shocks due to weather. For example, floods in 2010 resulted in estimated declines in crop output of 25 to 80 percent,69 fuelling further consumption drops and food price increases. This was exacerbated by further poor rains and harvest in the 2011/12 season, resulting in a short- term food production crisis. The number of people in need of seasonal humanitarian assistance as a result of either drought or flood is estimated to have fluctuated between 200,000 and 500,000 over the last few years.70 Further agriculture-related shocks include the theft of livestock, which was cited in village interviews as the single greatest cause of 67 BoS (2009a) Table 6.4. 68 Findings of these interviews, conducted as part of the PRSP process, are reported in the World Bank’s 2010 Poverty Assessment (World Bank {2010b}). 69 Relative to the previous year. The declines vary from district to district (World Bank, 2011d). 70 LVAC/DMA (2010 and 2011). 36 impoverishment,71 and the inability to finance agricultural inputs partly because of the steep price increases for fertilizer in recent years and partly because of the decline of cash remittances from miners, which used to finance agricultural activities. In addition, households face idiosyncratic shocks, mostly related to disease and the impact of HIV/AIDs, which deprive families of a principal breadwinner or reduce the labor available for cultivation or other income-generating tasks. Widowhood is a particular source of impoverishment, as widows often lose control of their productive assets (such as land, ploughs, tools, and housing) to relatives of their husbands.72 74. Unfortunately there is no analytical work on the incidence of these shocks or – more importantly – on the quantitative impact that they have on the consumption of the poor. Thus there is no firm evidence for policymakers to use to decide how to direct safety net interventions. The only feasible conclusions that can be made are that most of the extremely poor in Lesotho are poor because of inherent conditions of low productivity and low returns to their labor in a subsistence economy or because they lack basic assets (including health, skills and education) or are living in household conditions of destitution (the elderly, street children, disabled, or incapacitated who are not in families that can support them), rather than as a result of specific shocks. The only shocks that do have a significant effect on welfare are: (i) the temporary effects of price shocks and drought; (ii) exogenous shocks that affect employment and earning opportunities; and (iii) annual seasonal food shortages (discussed below).The current (2012) food crisis illustrates the need for flexible safety nets that can be expanded in response to need. 71 Leboela and Turner (2003) as cited in World Bank (2010b) p.41. 72 Based on communications from NGOs in Lesotho during consultations for this study, and World Bank (2010b). 37 Seasonality in Poverty 75. There is a distinct seasonality to the cycle of rural poverty in Lesotho . With a short growing season and limited land and labor, the very poor only produce enough maize (the staple food that constitutes much of their consumption) to last a few months after the harvest and then must buy food from the market. Table 13 summarizes the seasonality of food, labor, and poverty in rural Lesotho.73 76. Because stocks of food run out in about December each year, local prices also tend to increase at the same time as the poor need to purchase maize meal. This phenomenon is not as extreme as it is in other countries of the region because imports from South Africa tend to smooth these seasonal price variations. Nevertheless, the poor suffer doubly from this combination of need and prices, especially in the mountains where transport and distribution networks are weak. Table 13: Seasonality in Rural Agriculture, Food Security, and Poverty Agricultural Calendar Activity Labor and Weather Food Situation of the Poor September-December Land preparation Sporadically labor-intensive Good to moderate, Spring; dry season depending on landholding and harvest. December Planting Labor-intensive Beginning of lean season Early rains Poor food availability January-March Weeding Rains; Sporadically labor- Lean Season - Almost no intensive food available May-July Harvest Sporadically labor- No food available until intensive. harvest; then abundant July-September Winter Much of highlands under Generally good food snow, limited work availability 77. This seasonal shock is a predictable one – it happens every year, and is a significant component of extreme poverty. This suggests that any safety net needs to be designed to address this annual seasonal shortfall. It also suggests a need for public works schemes that 73 More detailed calendars are presented for each of the different ecological zones in Annex 1. 38 could employ the surplus labor during the slack season in building or repairing community infrastructure.74 Price Increases and Shocks 78. While Lesotho is somewhat less susceptible to wild fluctuations in domestic food prices than other countries due to its integration with South Africa, it has nonetheless been hit by the same large increases in global prices that have affected the poor elsewhere. As can be seen in Annex 2, there was substantial inflation in 2007 to 2009, during which time the prices of basic foodstuffs increased by an average of 37 percent.75 There is no hard evidence on the impact of this on the poor, but the available data on casual wage rates and incomes (from the LVAC and the CWMS) suggest that it is unlikely that the incomes of the poor increased by the same amount over this period, so it is likely that their purchasing power was significantly eroded by this inflation. The Openness of Lesotho’s Economy: Exposure to Positive and Negative Shocks 79. The openness of the economy and its position vis-a-vis South Africa leave Lesotho prone to exogenous shocks that can quickly undermine (or increase) the consumption of the poor. The recent growth of and then the loss of employment opportunities in garment manufacturing is one example, and the expansion and decline of employment in South Africa’s mines is another. 80. Employment in the mines of South Africa was the main source of employment and income for Basotho families for many years and raised many of them out of poverty . At the peak of this employment, each miner from Lesotho supported between 7 and 9 people 74 Public works would generally be a permanent feature of a safety net system, but be expanded in coverage during the lean season – provided that the seasonality of weather and labour demands on the poor permits. 75 Change in the “Food and Non-alcoholic Beverages” component of the CPI between January 2007 and July, 2009. 39 back home. While remittances from men going to work in the mines of South Africa have declined (Box 2 and Annex 1), in 2010 the number of migrants still amounted to 428,000, or about one-fifth of the population. At the same time there has been a major expansion of employment opportunities for women in textile production for export.76 These textile factories have created some 45,000 jobs, and although they pay relatively modest wages (about M.800 per month or US$100), the amounts involved are a significant source of cash income for poor families. The opportunities created by such economic openness can create both precipitous rises and declines in the fortunes of the poor. For example, with the contraction of global demand, employment in textiles dropped by about 5,000 in 2009. This openness to the ebb and flow of external developments is always going to be an important part of Lesotho’s economic and poverty landscape. Its implications for public policy are that transfer or safety net programs should be in place that the government can expand or contract in response to these developments and that careful attention needs to be paid to the role played by remittances in assessing the incomes of the poor for targeting transfers. Movements in and Out of Poverty 81. The extent to which deep household poverty is transient or permanent is an important factor for determining the objectives of a national safety net strategy. For example, if people fall into and rise out of poverty, then programs need to be designed to provide temporary support to a shifting population of beneficiaries, whereas if they are permanently mired in deep poverty, then policymakers need to choose programs that will support the same families over the long run. 82. Very limited panel data exist on households’ movements in and out of poverty in Lesotho. One of the few attempts to estimate these inter-temporal movements is presented in Watson and Hall (2004), who used a small, non-random rural panel to 76 Garments are produced primarily by producers who have located in Lesotho to take advantage of the AGOA trade preferences. 40 conclude that 58 percent of the households stayed at the same level between 1993 and 2002 with 26 percent of those being chronically poor while 32 percent had never been poor. Meanwhile, 42 percent of the households in the panel changed poverty status, 14 percent changing from poor to non-poor and 28 percent changing from non-poor to poor. 83. Based on the limited available evidence, it seems reasonable to conclude that there is relatively little movement in and out of poverty and that most of the extreme poor are mired in long-term poverty, either because they have insufficient assets to participate in the productive economy or are destitute as a result of lifecycle poverty. The latter group probably consists of the elderly disabled living alone, grandparent-headed households, and double orphans, all of whom will be in poverty for many years. On top of these are the households that fall into poverty every year in the lean agricultural season and then rise out of it again when the season is over. 41 CHAPTER III: EXISTING TRANSFER PROGRAMS 1. Overview 84. Lesotho is implementing a number of transfer and safety net programs . Some are funded purely by the government, and some are supported by development partners. Several programs are targeted to specific categorical groups, others include mean-tested targeting, while yet others are universal. This chapter examines all of the main transfer programs that provide cash or in-kind benefits directly to households in Lesotho (independent of whether or not they are intended to target the poor) and assesses them through the lens of their impact on extreme poverty as a way of informing the debate about future policy choices. Table 14 summarizes the main programs and their respective spending. 2. Program Descriptions 85. This section describes the coverage, benefit levels, implementation arrangements, and expenditures for each of the existing programs. To the extent that data are available, it assesses their impact and effectiveness. However, it should be noted that there is no systematic way of discovering the actual poverty status of program beneficiaries in Lesotho, so this analysis is based on evaluations of individual programs – which are of varying quality – and rough simulations of incidence by the study team in cases where the data allow this. Recommendations for short- and medium-term measures to strengthen these programs are made in Chapter 5. 42 Table 14: Summary of Main Existing Transfer Programs in Lesotho Program Direct Beneficiaries (annual) Expenditure (annual) Average Transfer per Target Group Participant Universal Old Age Pension 83,000 M.371 m. ($49.5 m.) M.350/month Universal to all over the age of 70 Child Grants Program a/ 10,000 households (30,000 M.16.5 m. ($2.2 m.) M.120/month/HH Poor households with OVC children) Public Assistance 9,600 M.13 m. ($1.7 m.) M.100/month +/- Destitute people School Feeding Program b/ 389,000+/- M.236 m. ($ 31.5 m.) One meal per day 180 Primary school students days/yr. OVC Bursary Program c/ 20,000 M.48 m ($6.4 m.) M.2,381 p.a./beneficiary OVC secondary students Nutrition Support 85,000 M.47 m. ($6.3 million) n.a. Children at Early Childhood Program(WFP) d/ Development centers malnourished infants, and pregnant and lactating women Agricultural Input Fairs e/ 18,600 households Est. M.17 m. ($2.3 m.) M.650 each Vulnerable but viable famers National Fertilizer and Input Unknown M.44 m. ($5.9 m.) M.140/50 kg. bag Untargeted Subsidy f/ Tertiary Bursary Scheme g/ 16,200 M.575 m. ($76.7 m.) M.30,560 (see footnote) Tertiary students Integrated Watershed 58,000-115,000 M.112 million ($ 14.9 m.) M.960 Able-bodied rural people Management Public Works Program h/ 43 Notes: a/ Based on numbers from the DSW and UNICEF in June 2012. This cost excludes the investment costs to set up the program (approximately US$5 million over three years). b/ Total of government-funded beneficiaries (309,000) and WFP-funded (80,000); expenditure based on government budget of M.218 million (2010/11) and WFP budget of US$ 2.4 million (M.18 m.). c/ Based on average number of beneficiaries per year past five years (as per ministry data) – since total varies from year to year; total cost based on (i) for government average of past five years (M.175 m./5=M.35 million p.a)) as reported in template – since varies a lot from year to year because of carry-over; PLUS (ii) Global Fund spending of M.13 million past year; as reported by MoE. Note 20,000 beneficiaries include 5,200 supported by Global Fund in 2010/11, but budget reported by MoE in template does not include Global Fund spending. d/ Nutrition Support: Based on numbers from WFP, June 2012. e/ Total cost estimated based on data from FAO Lesotho. Beneficiaries based on average of 22,550 in 2009/10 and 13,185 in 2010/11 due to large annual variance. f/ Expenditure based on government estimate net of repayments; number of beneficiaries of 16,200 based on World Bank analysis of program. g/ Based on budget of M.112 m. 2010/11, estimate of 9,600 per month employed; annual total varies depending on whether work 1 or 2 months; M 960 per beneficiary based on one month employment limit; 20 days est. and M.48/day. h/ Costs based on average of 2009/10 M.59 million and 2010/11 (M18 m.+ supplemental of M.9.5 m = M27.5) – average M.43.7 million -- because varies from year to year due to carry-over of stocks. Average subsidy based on reported prices of M.110 subsidized and M.250 commercial retail per 50 kg, bag of fertilizer (Ministry of Agriculture and Food Security). 44 Cash Transfer Programs Child Grants Program 86. The Lesotho Child Grants Program (CGP) is a non-conditional social cash transfer program, providing support to very poor and vulnerable households that contain children. 77 The program was launched on a pilot basis in 2009 and is in the process of being scaled up. It is being implemented by the Ministry of Social Development with financial assistance from the European Union and technical support from UNICEF. 87. Coverage and Benefits. At present the program operates in 5 of the 10 districts (21 out of 128 community councils using the old demarcations) and reaches almost 10,000 households. The program will be expanded to cover all of the 10 districts (in 22 of the 65 community councils, using the new demarcations)78 and is expected to reach around 25,000 households by 2014 (roughly 60,000 children). Currently the program provides M.120 per month (about US$16) per household regardless of the number of children in each household. Starting in October 2012, the amount will be adjusted according to the number of children per household. Households with one or two children will receive M100 per month (M300 per quarter), those with three or four children will receive M.600 per quarter, and households with five or more children will receive M.750 per quarter. 88. Implementation and Targeting. The Ministry of Social Development (MSD) is responsible for the overall implementation of the program, and UNICEF is providing continued technical assistance to build the ministry’s capacity to independently run the CGP operations. A civil society organization was engaged to facilitate the community 77 Although originally intended primarily to support orphans and vulnerable children the program has now expanded to cover any extremely poor households with children under 18. 78 It should be noted that the Government of Lesotho has re-demarcated community councils (CCs), which has resulted in 65 CCs, 1 municipal area, and 11 urban councils. As a result the CGP is presently operational in 17 CCs and will be expanded to cover 22 of the 65 CCs in all of the 10 districts. 45 development component under the program. A substantial ongoing program of technical assistance aims to build the capacity of the MSD to develop and test the tools and systems necessary for successfully running the CGP and other social protection interventions. 89. The CGP targets poor households that are caring for OVC.79 Targeting is a fairly sophisticated process, including a census-style interview to collect data from all households within a given community, feeding into a National Information System for Social Assistance (NISSA) database, and thereafter categorizing households using a proxy means test (PMT), which attempts to estimate the poverty status of each household using a variety of variables. The PMT formula is used to categorize households into the following five categories (Maloti figures refer to estimated per capita monthly income): - Ultra Poor (below 15th percentile / M.47.11). - Very Poor (below 30th percentile / M.81.6). - Poor (below 50th percentile / M.139.75). - Less Poor (below 70th percentile/ M.235. 20). - Better-off (not in any of the above categories). 90. Households in the poorest two categories are deemed to be eligible for the program. Their selection is further validated by community-level Village Assistance Committees (VAC), and, only after the PMT and the VAC have both verified a household as being eligible is it included in the program. All of the information collected during the targeting process is entered into the management information system (MIS) of the program and there is a medium-term plan to expand this system into a National Information System for Social Assistance (NISSA) that can be used to identify beneficiaries for any safety net program or at least to act as a common single registry for social assistance programs (see further 79 According to the Operations Manual, the CGP is targeted to poor households with children. The target group includes poor households caring for orphans, including single orphans, double orphans, and child-headed households. The target group also includes poor households caring for children who are not orphans but who are vulnerable by virtue of their poverty status. 46 discussion in Box 8 in Chapter 4). The MIS currently captures information on around 30,000 households or about 8 percent of the population. 91. Expenditure, Financing, and Cost-effectiveness. The annual direct cost of the program is around M.16.5 million (US$2.2 million) which is currently fully funded by the EU.80 This includes an estimated M.2.25 million (US$0.3 million) for operating costs but does not include the initial investment costs to set up the program (approximately US$5 million). The current plan is for the government to pay for the entire benefit amount from the government budget by April 2013 and to completely take over the entire program by April 2014. By then the EU would only provide technical assistance and a partial contribution to the operational costs (50 percent in 2013 and 25 percent in 2014). 92. The 13.7 percent direct operating cost is relatively low, particularly given the small size of the program and the fact that the operational costs are expected to decrease with time (which is normally the case for new programs as they tend to be more expensive in the initial phase). However, it should be noted again that these costs do not include very substantial expenditures on technical assistance, which also supports implementation. 93. Impact and Commentary. Although there is no rigorous evidence on the program’s impact as yet, an initial evaluation81 found that the program resulted in recipient households being better able to meet their food and non-food needs. Over 87 percent of households reported spending the cash transfer on basic food items and other basic needs. A more recent targeting evaluation report82 found the targeting process did select households that were poorer than average in the areas covered (with monthly household consumption estimated at about 23 percent below the average83) but not necessarily the 80 This amount does not include the investment costs involved in setting up the program. 81 Kedra et al (2011). 82 Pellerano et al (2012). 83 M.700 per month for eligible households, compared to M.915 per month for the overall sample (Pellerano et al 2012), Table 3.44 p.71). 47 poorest, with 51 percent of eligible households not being from among the poorest 29 percent of the population with children.84 94. The PMT is a potentially useful targeting tool, but there is a need for a rigorous examination of its effectiveness in the CGP before expanding it or replicating it in other programs. The initial evaluation of the PMT85 highlighted a number of challenges with the existing PMT formula, including the weak and outdated data used to estimate the model and the fact that the underlying data did not contain information on key assets and income flows (such as pensions). The current process of data collection and the community verification of the beneficiaries selected by the PMT seem complex and labor-intensive, and moving forward it will be important to review these processes and to examine whether the government can continue to implement them in the future. A PMT presents challenges in a context like Lesotho of uniform poverty (the divergence between the potential beneficiaries selected by the PMT and by the VACs seems to confirm this). The government intended to update the PMT using data from the 2010/11 HIES (with technical assistance from the World Bank), but because of technical problems with the data, this was not possible.86 However, it is recommended that the current PMT be revised once new data become available and that the PMT be rigorously evaluated over the next two to three years to assess if the households that it selects are really much poorer than others. Any decision regarding whether or not to continue with the PMT should be based on evidence from such an evaluation. 84 Pellerano et al (2012),Table 4.4, p.98, OPM 2012. This initial review reports that 74 percent of the households are among the poor. 85 Pellerano et al (2012). 86 Due to technical problems with the 2010/11 HIES consumption data, it could not support our estimation of PMT regression coefficients. Revision of the PMT formula using the 2002/03 HIES data was discussed by the Bank with the government, but the costs of revision were deemed likely to have outweighed the benefits. Since such an update would probably have yielded marginal improvements in precision and implementing these changes would have incurred considerable costs (for example, in changing to the MIS, changing the data collection forms, and retraining data collectors). 48 95. Beneficiaries seem to spend a large share of the cash transfer on school and hygiene-related expenses, and a majority of households have indicated a preference for receiving cash rather than food. Sixty percent of the households reported having spent the cash transfer on school uniforms, shoes, and supplies. Over 45 percent of the households reported spending part of the first three cash transfers on hygiene and personal care products.87 However, it is hard to discern any significant impact on any single dimension of well-being. It can further be noted that a majority of households report that they would prefer to receive cash rather than food. 96. The cash transfer amount for the CGP represents less than the national food poverty line, and there are concerns about the program’s sustainability because of its heavy dependence on external support. The monthly cash transfer of M.120 per household is significantly less than the national food poverty line per adult equivalent of M.137. Given that the average household size is around four to five people, this results in a relatively low transfer value.88 The program is currently heavily dependent on donor support, but the government is supposed to take over the program, which raises concerns about its sustainability, particularly in a time of fiscal constraints.89 97. Cash directly transferred to targeted beneficiaries have proven to be an efficient way of supporting poor and vulnerable worldwide, but there is a need to think carefully about which groups to target in the context of Lesotho. Evidence from many cash transfer programs has shown that they can significantly increase household consumption as well as contributing to human capital development. However, with widespread uniform poverty in 87 However, according to UNICEF, the trends of expenditure are changing and families are prioritizing their children’s needs including nutrition and clothing (after receiving the initial foo d parcel provided by the program, a majority of recipient families made their first expenditure on school-related items). 88 About 22 percent of the food poverty line (based on M.120 per household per month, and 4 adult- equivalents per household, implies about /M.137 per adult equivalent per month, or about 22 percent of the food poverty line). 89 As noted in the section on coverage and benefits, the value of the transfer may however be revised in about October 2012. 49 Lesotho, particularly in rural areas, it is not possible to support all poor people, and it is difficult to target based solely on poverty indicators (because of the country’s uniform poverty and the difficulty involved capturing remittances). Therefore, the government needs to identify which groups should be prioritized for support from safety net interventions in the context of Lesotho. 98. The CGP could be an appropriate basis on which to build a general cash transfer program for all destitute groups (such as the elderly poor living alone and the disabled). It is also worth mentioning that orphanhood alone does not seem to be a good indicator of poverty, but if it were combined with a PMT and community targeting, it could be a good basis for targeting. The approach taken in the CGP will need to be fully evaluated to help policymakers to understand how well this approach works. Categorical targeting only could otherwise be used, but would then likely include more people who are less destitute. 99. The proposed general cash transfer program could include a number of conditions to promote human capital development. The actual conditions to be used would need to be careful considered in the specific context of Lesotho. Conditions related to health, nutrition and education are commonly used worldwide, but what makes sense depends a lot on the specific country context (including its capacity to implement and what services are available on the supply side). In Lesotho a recent pilot initiative has been looking into the potential for using cash transfers to reduce the rate of sexually transmitted infections.90 Also, the government is at the initial stages of piloting a conditional cash transfer program as part of the CGP.91 Cash transfer programs can also be used to encourage behavior that will lead to decreased malnutrition rates (see further discussion under the nutrition program below) or increased child vaccinations and school enrollment and attendance rates (see Box 9). 90 Tanzania has recently implemented a similar experiment designed to expand conditional transfers to encourage the prevention of STIs. By targeting adults and by making quarterly cash payments conditional on negative tests for curable STIs, the program did not pay participants to do something but paid them not to do something, which was engage in unsafe sex. 91 With support from the EU and UNICEF. 50 Public Assistance 100. The Public Assistance program (PA) operated by the MSD provides support in cash or in kind to extremely destitute people. 101. Coverage and Benefits. Coverage is limited by budget availability, and the program currently only reaches about 9,500 people annually across the country. The PA provides cash grants and various types of in-kind support to beneficiaries, including medical exemptions and food packages. The nature and size of the benefit vary depending on the need. The PA is largely discretionary and benefits both destitute households and individuals. Increasingly the program is tending to provide cash payments of M.100 (US$14) per month. The program also provides some beneficiaries with additional limited resources as ad hoc income and livelihood support interventions. 102. Targeting and Implementation. Beneficiaries need to register for the PA at the district office of the Department of Social Welfare. The targeting assessment is done on a case-by-case basis at the discretion of officials in the district offices. Beneficiaries may also be identified by chiefs, community councilors, or the outreach officials of various departments of the government. However, the following criteria are used to target potential beneficiaries: (i) households without a regular income; (ii) households with monthly incomes of less than M.150 (US$21) per month; and (iii) households with livestock and/or fields or other assets that cannot generate an income of M.150 per month. Roughly half of the beneficiaries are children who qualify as OVC, 47.7 percent are classified as destitute adults, and 2.3 percent are disabled (see Annex 3). 103. Expenditure, Financing, and Cost-effectiveness. The PA is fully funded by the government, and the total budget in 2010 was M.13 million (about US$1.86 million). An estimated M.1.7 million of those funds were allocated to the operating costs of the program. With the recent shift from in-kind to cash transfers, the largest share of the funds 51 is now allocated to the cash paid to beneficiaries (96 percent) and only 4 percent is allocated to in-kind support (see Annex 3). 104. With a total annual cost of around M.13 million (US$1.86 million) and 9,584 beneficiaries, the average annual cost per beneficiary is M.1,356, resulting in a monthly average cost of M.113 or a rough estimated 13 percent to deliver the M.100 per month. 105. Impact and Commentary. There has been no review or assessment of the PA. Although the recent overall review of safety net programs92 includes a brief summary of the PA, it does not include any analysis on the program’s impact or effectiveness. Given the absence of any review of the program, it is difficult to determine to the extent to which the program reaches the most destitute. There also seems to be a lack of clarity on the program’s operational guidelines, including its eligibility criteria. Other challenges with the program include the subjectivity, inconsistency, and lack of transparency in selecting beneficiaries. Old Age Pension 106. The largest safety net transfer in Lesotho is the Old Age Pension program (OAP). This is a non-contributory, unconditional cash transfer paid to all Basotho over the age of 70 who are not getting any other pension from the government. 107. Coverage and Benefits. The number of beneficiaries has grown from about 65,000 when the pension was introduced in 2004 to around 84,000 in 2011 (about 4.4 percent of the population). Monthly benefits were initially M.150 (US$20) and have been raised significantly over the years and by 2012 had reached M.350 (US$47) per pensioner per month. 92 Ayala Consulting Co (2011b). 52 108. Implementation and Targeting. The program is administered by the Pensions Division of the Ministry of Finance. The pensions are paid in cash on a monthly basis through 296 pay points (including 17 pay points that are serviced by helicopter due to poor accessibility by road). The government is exploring the possibility of adopting electronic payments and of using biometrics to help to identify beneficiaries. 109. The OAP is categorically targeted to all citizens over the age of 70 except for ex-civil servants who are receiving a government pension of a higher value. As a result, no retiree is allowed to draw two pensions from the state. However, this restriction does not apply to retirees who are receiving pensions from the private sector.93 110. Expenditure, Financing, and Cost-effectiveness. The current cost of the OAP is about M.371 million (US$49.5 million) annually (about 3.3 percent of total public expenditure), including delivery costs of about M.14 million per year (US$1.9 million). Delivery costs thus represent about 3.8 percent of the total program cost, which is low in comparison to many other cash transfer programs in Lesotho and elsewhere. 111. Impact and Commentary. While there has been no systematic assessment of the impact of the OAP on consumption and poverty, two reviews have identified a number of positive poverty-related effects. The benefits are shared within households, and there is some evidence94 that consumption and educational attainment increased as a result of the pension. The same evaluation reported that the proportion of beneficiaries reporting that they never or rarely had enough food to satisfy their hunger fell from 80 percent to 40 percent after receiving the pension. Another assessment95 noted increases in self-esteem 93 A small number of beneficiaries of the African Pioneer Corps (APC) pension scheme, which is targeted to servicemen who participated in the First and Second World Wars (or their widows), are also entitled to a pension. The APC provides a pension of M.200 per month. According to a representative of the Ministry of Finance and Development Planning, approximately 600 ex-volunteers and 3,000 widows are currently receiving the pension. 94 Croome et al (2007). 95 Ayala Consulting (2011b). 53 among the elderly and indicated that a large proportion of the pension (60 percent) is being spent on food. This assessment estimated that around 20 percent is spent on dependent orphan children. 112. The benefit level of M.350 per month is about two-and-a-half times the estimated food poverty line and is 44 percent higher than the basic needs poverty line.96 The pension thus provides a fairly generous income for the single elderly (and more so if both individuals who constitute an elderly couple are receiving it). However, assuming that many of the elderly poor live in households of about four adult equivalents (about the national average), the pension would represent about half of the food poverty line income for a family. 113. The Old Age Pension is clearly an integral and important part of the social safety net in Lesotho that protects a critical group of vulnerable Basotho – the elderly – from extreme poverty. This is particularly important at the moment, as a large number of grandparent- headed households are supporting orphaned children. Having said that, the OAP only directly reaches about 4.4 percent of the population and indirectly reaches perhaps 17 percent.97 Also, as discussed in Chapter 2, only an estimated 6 percent of the poor in Lesotho are over the age of 64. 114. The poverty rate among the elderly in 2010 was estimated to be the same as among the population as a whole, implying that almost two-thirds of the OAP payments go to non- poor households. This is not to say that a universal pension is not an extremely useful social intervention, but, as a means of reducing extreme poverty in the country, the M.371 million spent annually on the OAP has only a limited impact and that, in evaluating it as part of a unified program to address extreme poverty, policymakers need to recognize that most of the transfers are going to the non-poor. Given that only 6 percent of the poor are 96 Note that these are rough estimates and the actual current poverty lines may be different. This will need to be confirmed when the Bureau of Statistics releases new estimates. 97 Based very roughly on an estimated average household size of four among those containing pensioners. However, the actual will likely be lower, as some elderly couples may receive two pensions in one household. 54 estimated to be older than 64,98 any program targeted according to old age is not going to cover many of the poor. 115. While the scale of the program’s coverage ‒ and its fiscal cost ‒ has been limited in the recent past by the significant fall in life expectancy in Lesotho, over time the pool of beneficiaries will inevitably grow as the population ages. The increasing number of beneficiaries combined with the increasing value of the pension payment has resulted in a program that now consumes about 3.3 percent of total current expenditure in 2010/11 as opposed to 2.7 percent during the first year of its operation in 2005/06. In its current form in which it serves 4 percent of the population, the pension appears to be affordable. However, there are concerns related to the long-term sustainability of the program that merit continued attention. In particular, the authorities should resist any policy changes that would significantly increase the costs of the pension in the current tight fiscal environment – for example, by reducing the age of eligibility or further increasing benefit levels.99 116. Finally, given Lesotho’s fiscal challenges, it would make sense to maximize the efficiency of the OAP. For example, given that the program currently excludes people who receive a government pension of a higher value, it would make sense to also exclude people who receive pensions from the private sector that are higher than the OAP. This would avoid any elderly person receiving two pensions, which could be the case with the current eligibility criteria. Similarly, with the advent of a unified beneficiary registry system (NISSA – discussed below), it may also make sense to coordinate benefits to ensure that, except in conditions of extreme deprivation, the same households do not receive the support under the Child Grants Program if they are already receiving the (much more generous) Old Age Pension. 98 World Bank (2010b) p.3. 99 There has been some mention of lowering the eligibility age to 65. About 1.8 percent of the population is between 65 and 69 years old (35,000 people), so including them would increase costs by about M.154 million annually (based on a benefit of M.350 per month and 5 percent overheads). 55 117. In addition, there may be operational areas where efficiency gains can be made, such as exploring the use of electronic ways to deliver the cash to beneficiaries. Furthermore, since the OAP is the largest safety net program in the country, it would make sense to undertake: (i) a proper impact evaluation to be able to adjust other safety net programs to take account of the household impact of the pension and (ii) a sustainability analysis of the cost of the program in the long and medium term to inform policymakers as they consider making any further changes in its configuration. Food and Feeding Programs School Feeding Program 118. The transfer program with the largest number of direct beneficiaries is the School Feeding Program (SFP), which is implemented nationwide by the Ministry of Education and Training with some support from the WPF. The objectives of the program are to provide all children in public primary schools with supplementary nutrition to improve health and educational outcomes. 119. Coverage and Benefits. The program provides school meals to all 389,000 students in primary schools across the country. Under the government-funded part of the program, children receive a substantial lunch (including papa, bread, vegetables, beans, eggs, and milk) for 180 days per year. In addition, the WFP provides support to about 80,000 students in 429 schools in the most remote areas of Lesotho in the form of dry rations of food to be cooked on the school premises (see detailed composition of the meals provided in Annex 5). The WFP provides these children with a mid-morning snack of maize meal and with a lunch of maize meal, pulses, and vegetable oil100 for a total of 180 days per year. 100 These calculations are based on half-day school attendance, as per the majority of the school schedule. 56 120. Implementation and Targeting. Although the SFP is implemented by the Ministry of Education and Training, the Food Management Unit (FMU) under the Office of the Prime Minister is responsible for the transportation, storage, and delivery of food to schools. In the schools supported by the government, caterers are selected locally and are required to purchase and prepare the food according to menus established by the Ministry of Health. Caterers prepare food in schools, but some schools do not have the required facilities and the food is therefore prepared in the home of the caterer. In the case of the WFP-funded part of the program, the WFP distributes food to schools, which is then prepared by locally selected and contracted cooks (paid for by the government, which also covers the costs of cooking fuel, salt, and sugar).101 121. Expenditure, Financing, and Cost-effectiveness. The government covers the main part of the cost of the program, with the WFP providing support for the delivery of food in the most remote areas. More than a quarter of the primary education budgetary allocation goes to the SPF, which had a budget allocation of M.218 million (US$30 million) in 2010/11.102 122. The WFP is supporting 80,000 students under the School Feeding Program up until December 2012. Thereafter it is expected that the government will take over the funding of the program in full. Currently the WFP’s part of the SFP program has an estimated annual budget of US$1.7 million.103 Private caterers are paid M.3.50 (US$0.50) per pupil per day to purchase, prepare, and serve food.104 According to a recent review of the SPF,105 the 101 The WFP also provides a monthly take-home ration targeted to orphans and herd boys, which is conditional on school attendance. Beneficiaries of the take-home ration must be OVCs and/or live in food-insecure households and must be under 15 years of age. There are no standard targeting criteria for this component and the NGOs that are implementing the program decide on suitable beneficiary selection rules. 102 WFP (undated), “Lesotho Development Project Number 200199 – School Meals Program” , p. 4 103 The total allocation for the WPF program is US$4.8 million for a period of two years, but this includes feeding 30,000 pre-school children. The annual cost of the SFP can therefore be calculated based on an assumption that 73 percent of the cost would go to the SFP and the rest to the pre-school children. 104 The caterers feed a maximum of 150 children per contract for a period of 12 months (to allow for other unemployed to get a chance to participate). 57 average cost per child per day under the government-run part of the program is M.3.54 compared to M.2.75 for the WPF-run part of the program. Given that the children are fed for 180 days per year, this implies a yearly cost per child of M.637 (US$85) for the government part of the program compared to M.495 (US$66) under the WFP-funded part of the program. Given that the rough estimated value of the meal to the household is M.203.4 per child per year (US$24.15),106 this would seem like an inefficient way of delivering transfers to beneficiaries (as opposed to, for example, providing the cash directly). 123. Impact and Commentary. There have been several reviews and assessments of the School Feeding Program but no comprehensive impact evaluation. It is therefore difficult to determine the exact impact of the program, although it is possible to determine some general trends in its impact and highlight some of the positive results. Furthermore, since the program is not targeted, a priori it can be expected that a large proportion of the benefits go to the non-poor. Table 15 indicates that only 43 percent go to the two poorest quintiles and that 37 percent go to the two richest quintiles. Table 15: Approximate Share of Spending of the SFP Going to Poor and the /Non-poor % of Primary School Students Approximate Value of School Feeding Coming from Each Consumption Expenditure Going to Group Decile (M. million annually) Poorest 22% 51.9 ($6.9 million) nd 2 21% 49.6 ($6.3 million) rd 3 20% 47.2 ($6.3 million) th 4 19% 44.8 ($6.0 million) Richest 18% 42.5 ($5.7 million) Source: Author’s calculations. Notes: Based on total expenditure in 2010/11 estimated M. 236 million (or US$31.5 million). 105 Motseng Logistics Services (2011). 106 Estimations from the WFP using 2010 local prices and 2010 ration composition. 58 124. There are indications that the School Feeding Program has had some positive impact on schools, but it is difficult to determine with certainty the direct impact of the program. The 2011 review of the School Feeding Program107 indicated that the program seems to have increased enrollment, stabilized attendance, and reduced dropout rates. An evaluation in 2009108 indicated that the program effectively supports the enrollment, retention, and concentration of students, but since the program covers all schools, there is no control group that would make it possible to compare the performance of schools with and without school feeding. 125. There are operational aspects of the program that could be made more efficient by simplifying implementation arrangements and modifying the food menu. The current program is micro-managed, food does not always reach the schools, and the arrangements with the individual cooks who are responsible for purchasing food are problematic. In addition, this is not cost-efficient since food is bought individually in small quantities and often at high prices. In addition to making adjustments to the implementing arrangements (see further suggestions from the recent review in the next paragraph), there is scope for increasing efficiency if program managers were to consider adjusting the current menu of food to decrease the overall cost of the program. Both assessments of the SPF109 highlight insufficient implementation capacity as one of the major challenges of the program. Other weaknesses that the assessments identified included a lack of adequate or necessary equipment for preparing food in schools, a weak tracking and monitoring system, a lack of variety in the menu, and delays in making payments to cooks. 126. The Ministry of Education and Training (MoET) recently commissioned a review of the SFP110 with the aim of developing a framework for a unified version of the program after the current support from the WFP ends. The proposal suggested outsourcing that 107 Motseng Logistics Services (2011). 108 Haag et al (2009). 109 Motseng Logistics Services (2011) and Haag et al (2009). 110 Motseng Logistics Services (2011). 59 management of the SFP to a managing agent, while the MoET would retain its oversight role and still pay the cooks employed in the scheme. The unified SFP would offer a standardized menu across all regions. Cooks would no longer be responsible for purchasing the ingredients, which would be the responsibility of the managing agent. Local farmers would be identified to supply perishables (such as maize meal and pulses), while bulk items (such as vegetable oil) would be purchased from international markets. The proposed model was estimated to cost an average of M.580 per child per year (US$83), which would be slightly lower than the current cost of the government’s part of the program (M.637 /US$85) but significantly higher than the WFP-run part of the program (M.495/US$66). The review envisioned the outsourced SFP as having additional benefits beyond providing a consumption transfer and education outcomes in that it could potentially also yield benefits to the local economy through the creation of markets and the provision of income support to local farmers. 127. The average cost per child is high compared to equivalent programs in other countries, and from a poverty-reduction perspective, the SFP is not an efficient way of providing consumption support to households. As indicated earlier, the average cost per child is high for both the government-led and the WFP-led parts of the current School Feeding Program. The costs indicated by the 2011 review of the program by Motseng Logistics Services is high (US$85 and US$66 respectively) compared to averages of about US$40 elsewhere (see Box 3). Even the cost estimates for the proposed outsourced model are high. It is worth noting that, provided unit costs can be contained, over time the overall fiscal cost of the program in Lesotho will inevitably decline as the primary-school-aged population shrinks.111 111 Primary enrollment peaked at 420,000 students about five years ago and is now 389,000. 60 128. School feeding programs present policymakers with a quandary. On the one hand, they have the dual benefit of providing a consumption transfer and enhancing educational outcomes. On the other hand, the evidence worldwide has shown that their impact on educational performance is difficult to measure and is often marginal (greater impact could be achieved, for example, by spending the same money on improving educational quality). At the same time, as a safety net, school feeding is not a particularly cost-effective way of transferring resources to the poor, because: (i) it generally costs more to deliver the food than the value of the transfer to the household; (ii) if neither schools nor students are targeted, a lot of the transfer goes to non-poor households; and (iii) there is some evidence112that the impact on children’s total food intake may be marginal because there is a substitution effect, with families providing less food at home because they know the child will be eating at school. Given that in Lesotho the actual estimated value to the household of the meal provided (US$24.15) is much less than the cost to deliver the meal (US$66 to $85), purely from the point of view of a safety net, it would be more efficient to deliver cash directly to the beneficiaries than to provide them with food. 129. The School Feeding Program in Lesotho is well-established and has strong social and political support, so there is no question that it will not continue. However, it is worth noting that many countries (including most high-income ones) do not provide universal school feeding. Given that in Lesotho over 60 percent of the benefits go to non-poor households, it is worth considering over time withdrawing school meals from better-off schools and/or better-off areas. The other obvious challenge is to reduce unit costs. 112 Grosh et al (2008). 61 Box 3: Cost and Effectiveness of School Feeding Programs School feeding programs exist in one form or another in almost all countries. Some cover millions of students each year. In India and Brazil, the programs reach 130 and 37 million children respectively, and both programs, which were originally funded by external support, have transitioned into sustainable national programs that are now providing technical support to others. In other (often low-income) countries, the coverage of school feeding programs tends to be lower and to be more dependent on external support (particularly from the WFP). In Africa, programs typically reach less than 10 percent of the school population (ranging from 50,000 in Ghana to more than 500,000 in Ethiopia, Tanzania, and Uganda). In Lesotho, the program covers 390,000 children. The effectiveness of school feeding programs depends on several factors, including the type of transfer (take- home rations, in-school meals, fortified biscuits, or a combination of these) and the targeting mechanism. Take-home rations (which have an average annual per capita cost of US$50) involve significant administrative costs but can be more narrowly targeted than in-school meals and, therefore, can cover more of the very poor for the same expenditure. In-school meals (which have an average annual per capita cost of US$40) are a relatively expensive way of feeding the poor since these meals are usually provided to all children in certain schools or grades regardless of their poverty status. Fortified biscuits (which have an average annual per capita cost of US$13) generally involve lower administrative costs since they are cheaper and easier to distribute than meals, but they are less substantial and they are normally less nutritious than the meals. A recent paper concluded that it is difficult to assess the effectiveness of school feeding programs since their impact is partially on education and partially on health. It points out that school feeding is not among the best investments to make in terms of nutrition since it generally does not focus on the period during which children are most vulnerable to malnutrition (in utero through to two years of age). It further argued that school feeding is not the best way to use funds for education, but it did emphasize that, while school feeding may not be the best educational intervention, it can be and an important element in achieving an effective education system. Sources: Bundy et al (2009) and Alderman and Bundy (2011) Nutrition Support Program for Malnourished Children and other Vulnerable Groups 130. The Nutrition Support Program aims to reduce malnutrition among vulnerable groups in four districts where malnutrition is particularly high. It consists of three components: (i) a preventive one, which provides which provides “blanket” supplemental feeding to all children under the age of 2 and to pregnant and lactating mothers during the lean season in the covered districts; (ii) a curative one, which provides additional targeted feeding to children and others suffering from acute malnutrition; and (iii) universal feeding at early childhood development (ECD) centers (similar to the School Feeding Program but for pre-school children). 62 131. Coverage. The program is providing food support to a total of about 85,000 beneficiaries in the four districts most affected by chronic malnutrition ‒ Mokhotlong, Qacha’s Nek, Thaba-Tseka, and Berea. To reduce the incidence of acute malnutrition, the program supports supplemental feeding for over 40,000 women and children suffering from acute malnutrition, and it is hoped that this level of coverage will be maintained for the next five years. To improve child growth and development indicators, about 35,000 children aged 6 to 23 months and pregnant women receive regular food supplements, this number is expected to drop to about 20,000 in the near future. Finally, about 7,500 children aged 2 to 5 years old who are enrolled in ECD centers managed by the communities that are receiving support (see Table 16: Nutrition Support Program, 2012 Coverage) Table 16: Nutrition Support Program, 2012 Coverage Current Number of Beneficiaries Reached (2012) (1) Curative – Acute Malnutrition Supplementary Feeding Supplementary Feeding for Children under the age of 5 and for 8,590 and pregnant and lactating women 34,204 household members for ART/TB (2) Preventive Feeding – To Improve Child Growth and Development Blanket food supplements for all children under the age of 2, 35,271 pregnant and lactating women, and PLWH (generally just in Jan- March) (3) Early Childhood Development Center School Meals Early childhood nutrition for children between 2-5 years old 7,546 Total 85,611 Source: Author’s discussions with WFP Lesotho, March 2012. 132. Benefits. Beneficiaries generally receive food as shown in Annex 3. Those suffering from acute malnutrition receive a supplementary feeding mix of sugar, oil, and a corn-soya blend (CSB). In addition, if there is an indication that a family is chronically food-insecure, they may also receive a take-home food ration consisting of maize, vegetable oil, and beans. The blanket feeding of children aged 6 to 23 months of age and pregnant and lactating women consists of essentially the same food package of corn-soy blend (CBS), oil, and sugar provided on a monthly basis. Children in the ECD centers receive a micronutrient-rich meal of porridge prepared with fortified maize meal twice a day for 180 days per year. 63 133. Targeting and Implementation. The program is targeted in the first instance geographically to four districts that represent about 47 percent of the population. Within these districts, the curative component is targeted to children identified as being acutely malnourished on the basis of their height and weight for age or, in the case of adults suffering from HIV/AIDS, to anyone with a body mass index below a defined cut-off point. These beneficiaries are identified by health center staff, but they reportedly have had problems in measuring child malnutrition indicators accurately. The blanket feeding is targeted in time during the January to March lean season. The ECD component is universal for all children enrolled in early childhood education. The program is new, and as yet there is no monitoring data that would make it possible to verify its targeting accuracy. 134. The WFP partners with the Ministry of Health and UNICEF to implement the program. The WFP provides food through clinics and health centers, and UNICEF supports the treatment of severely malnourished children. Once the food has reached the distribution points, cooperating partners organize beneficiary lists and manage the distribution of the food supplements to beneficiaries. The WFP also works with the Ministry of Agriculture and Food Security, the Ministry of Forestry and Land Reclamation, and the Food and Agricultural Organization (FAO) to establish community gardens. 135. Expenditure, Financing, and Cost-effectiveness. This program is fully financed by the WPF, and the total cost of the two-year program is estimated to be around US$6.3 million (M.47 million). Table 17 indicates the estimated total cost for the planned beneficiaries as well as the cost per meal. With no data available on the retail value of the meal as a transfer to the household, it is not possible to assess the cost-effectiveness of the program. 64 Table 17: Estimated Cost for Planned Beneficiaries and Cost per Meal Estimated total cost for Feeding category planned beneficiaries (US$) Cost per meal (US$) SFP 133,596 0.022 ECD meal 558,780 0.009 HIV/TB/PMTCT client 605,787 0.022 HIV/TB/PMTCT family meal 2,897,244 0.021 Blanket feeding meal 926,182 0.005 Total 5,121,589 Source: Data provided to the authors by WFP Lesotho. 136. Impact and Commentary. Since this is a very new program, it is too early to assess its impact. However, this program could be a good way of using transfers to tackle the problem of malnutrition – a high priority aspect of poverty in Lesotho – and in some form should be part of the national safety net system for the poorest. Therefore, it will be very important to evaluate the implementation of this program and assess its potential for being scaled up in the future if efficient. One option may be to link its nutritional elements with the expanding cash grants program. 137. In other countries, cash transfer programs have successfully linked the receipt of benefits to nutrition interventions, including growth monitoring, regular health center visits by pregnant women and children under the age of 5, the provision of micronutrients, and nutrition education. An example where a country has attempted to use a cash transfer program to achieve better nutrition outcomes is the Juntos program in Peru (see Box 4). Another example is in Senegal, where the government has scaled up the Nutrition Enhancement Program that provides cash transfers to vulnerable mothers and children under the age of 5 to reduce their risk of nutrition insecurity following the increase in the food prices.113 Cash or food transfers could be used as an incentive for extremely poor families that contain pregnant and lactating women and/or infants to ensure that they have access to a set of interventions to prevent and treat under-nutrition, including: (i) promoting good nutritional practices (breastfeeding, complementary feeding after six 113 World Bank (2009). 65 months, and improved hygiene); (ii) the provision of micronutrients (sprinkles, therapeutic zinc complements for diarrhea treatment, de-worming drugs, or iron-folic acid supplements for pregnant women); and (iii) therapeutic feeding for severely malnourished children with special ready-to-use therapeutic foods. Box 4: Using Cash Transfers to Achieve Better Nutrition Outcomes in Peru Juntos is a conditional cash transfer program in Peru that was created in 2005. The program serves nearly half a million poor rural households with children under 14 years old. The program provides a cash transfer equivalent to approximately US$33 per month, which represents 15 percent of household consumption, to the mother in each beneficiary household. The conditions attached to receiving the payment include the expectation that pregnant women and children under the age of 5 will make regular visits to health clinics and that at least 85 percent of elementary school-age children will attend school. In 2007, the government made nutrition a priority for its social policy and committed itself to reducing chronic malnutrition in children under the age of 5 by 9 percentage points in five years. As a result, the government modified the Juntos program to enable it to achieve better nutrition outcomes by providing a basic package of interventions in health and nutrition. An innovative aspect to the modification of Juntos is that the changes were pilot-tested in one district of Peru (San Jerónimo in the Apurimac Region). This involved the establishment of a multi-sectoral inter- agency working group (known as the Grupo Apurimac) that turned out to be critical to the program’s success, since the group is an important instrument for coordinating the delivery of transfers between the different sectoral ministries and particularly between Juntos and the Ministry of Health. Source: Vargas Silvana (2011). Emergency Relief Program 138. In response to the current (2012) crisis caused by two successive bad harvests, the government has launched an immediate relief program to provide vulnerable people with food during the lean season. Although the intended target is 725,000 people who are estimated to be vulnerable on the basis of crop forecasts, actual coverage is being constrained by a lack of resources so will be much lower. The program aims to provide a basic package of maize, pulses, and oil to the elderly and children in acutely affected areas. It also aims to expand the coverage of the reforestation program operated by the Ministry of Forestry (discussed below) that provides cash-for-work employment to the poor. 66 139. The program is administered by the Disaster Management Authority. The food distribution component uses community targeting by means of village committees that include representatives of local women, the disabled, and children as well as the local MP, chief, and members of the Community Council. The eligibility of potential beneficiaries is verified by questionnaires administered by the program on household composition, assets, and income sources. There is no evidence available on the accuracy of the targeting system. OVC Bursary Scheme 140. The OVC bursary scheme was introduced in 2001 and is implemented by the Ministry of Education and Training with financing from both the government and the Global Fund. 141. Coverage and Benefits. The program provides bursaries to about 20,000 secondary students who are identified as OVCs who live in all 10 districts of the country. In 2011, about 14,700 were supported directly by the government, and 5,300 were financed by the Global Fund. The support provided under the government part of the program covers tuition fees114 and boarding fees (for schools with such facilities). The 5,200 students financed under the Global Fund component also receive uniforms and hygiene kits. The average transfer per beneficiary is about M.2,381 (US$317) per student per year .115 142. Targeting and Implementation. At the beginning of year school year, district bursary administrators are given a quota on how many beneficiaries can be catered for in the coming financial year. This quota is normally determined by the available budget. The bursary administrators are responsible for identifying beneficiaries and submitting beneficiary lists to the program’s headquarters so that they can then be paid. Prior to the payments being processed, a verification exercise is supposed to be carried out to ascertain 114 Including stationery, book fees, feeding fees, and special subject fees. 115 Based on a communication from the OVCs bursary section of the MoET. 67 that the beneficiaries on the list are indeed enrolled in school. Based on the beneficiary list, the payments are made to the bank accounts of the schools attended by the bursary beneficiaries. Targeting is supposed to be based on the following criteria: The criteria for applicants include: The application consists of the presentation of the following documents, which are considered evidence of the status of the applicant as an orphaned and vulnerable child: • Double orphans having lost both parents • Applicant’s letter of motivation • Single orphans whose remaining parent is • Applicant’s birth certificate chronically ill • Chief’s letter confirming need • Children whose parents have severe • Proof of school admission disabilities that hinder them from earning a • Copies of parents’ death certificates living • Medical letters confirming parent’s unfitness • Children living with needy to earn a living guardians/caregivers • Abandoned children • Out of school children such as e.g. herd boys, servant girls • Children with special needs 143. Exactly how appropriate applicants are selected is not clear, and clarifying this is critical to the outcome of the program since it seeks first and foremost to allow OVCs to access and stay in secondary school who would not usually have such an opportunity. The main determining factor for eligibility is currently orphanhood, which the program equates to some extent with vulnerability. As pointed out earlier in this document, however, the correlation between orphanhood and poverty is not strong, and, although the list of selection criteria include other criteria such as abandonment, in reality access to the bursary fund is based largely on the child being either a single or a double orphan. Furthermore, the necessity to submit a host of documentation including official documents such as birth and death certificates, which many children who are in difficult circumstances may not be able to access easily, further constrains the program’s ability to reach the poorest. While these documents do lend credibility to an applicant’s claim to be orphans 68 and vulnerable, in themselves they are not sufficient evidence of this. There ought to be an opportunity for each applicant to be screened and to explain their situations in more depth as this would help the decision-makers to have a greater understanding of the complexities of vulnerability and need. 144. Expenditure, Financing, and Cost-effectiveness. The average annual cost of the program is around M.48 million (US$6.9 million). The program is being funded by the government with some support from the Global Fund, which provides financing for around 5,200 students at a cost of M.13 million (about US$1.8 million) per year.116 145. Impact and Commentary. The bursary scheme is an important effort to enable deserving children to go to secondary school, but it is not necessarily targeting the poorest. The poverty status of the beneficiaries is not known, but it seems as if the program is not particularly pro-poor. If the targeting criteria were adequately defined and implemented, then the program could reach the neediest more efficiently. As discussed in Chapter 2, poverty among orphans does not appear to be very much higher than among children as a whole, and there is an element of regressivity built into the program, in that those attending higher-cost schools ‒ who are presumably generally from better-off families ‒ receive larger bursaries. In addition, children who go to secondary school tend to come from the better-off segment of population; for example, only 5 percent are from poorest quintile while 40 percent are from the richest quintile, and 85 percent of secondary students are from the top 3 quintiles. However, thanks to free primary education, more poor students have been given the opportunity to access the secondary level and therefore could benefit from the program. The government provides funds to cover the school fees, but parents or guardians are still expected to cover other costs related to school attendance (such as those for school materials and uniforms), which may be a challenge for the worst- 116 These costs do cover tuition costs but no costs for the administration of the program, which are subsumed in MoET’s budget. 69 off families. In addition, the number of beneficiaries varies each year depending on the available budget, which naturally affects the continuity of support to the students. 146. Two assessments of the scheme have been undertaken that have mostly focused on its operational aspects and that highlighted targeting as one of the main areas that needs to be improved. An assessment undertaken in 2010117 highlighted the potential for reaching the students that are most in need by using school- and community-based targeting and community selection committees to identify beneficiaries based on need instead of using the OVC criteria. The report also highlighted the need to develop clear guidelines for selecting and supporting beneficiaries and stressed the importance of establishing an adequate system for tracking, monitoring, and reporting information on the recipients of the bursaries. Another review118 found that the OVC definition of vulnerability is broad and does not make it possible to accurately target the most vulnerable and that there was scope for improving the identification and registration process, including through the use of NISSA. The report also highlighted a shortage of resources, data collection challenges, and incomplete links with other OVC programs as other challenges that are undermining the impact of the scheme. An evaluation of the program is currently being planned, and a more rigorous assessment of targeting effectiveness should clearly be one of the areas explored in the evaluation. Public Works Transfers: Cash-for-Work and Food-for-Work Schemes 147. Lesotho is unusual in not having any large-scale public works programs (PWP) to provide temporary employ opportunities to the poor. There have been a number of food- for-work programs in the past that were used to distribute food aid, mostly financed by the WFP and mostly in particular areas in response to drought or floods. 117 Mwansa (2010). 118 Ayala Consulting (2011 b). 70 148. The Ministry of Forestry operates the Integrated Watershed Management Program that employs about 9,000 people at a time (for a total of about 96,000 annually) to plant trees and do other environmental conservation work at the village level. Although not explicitly poverty-targeted, the program nonetheless provides an injection of funds into rural, and sometimes remote, villages, which tend to be poorer than average. Villagers are employed on a first-come first-served basis, but each person can only work on the scheme for one month. There is no analytical evidence on who benefits, but the wage rate – at M.48 per day – is well above the effective daily wage rate for unskilled agricultural labor (about M.20 per day – see Chapter 2), so it is likely that the program is attracting people who are not the poorest. The lack of clear operational guidelines puts pressure on program implementers to expand the program too rapidly to more geographic areas, which would jeopardize the potential environmental benefits that the program could yield if it were to be implemented in a more systematic manner. 149. In addition, as part of its new country program for Lesotho, the WFP is currently formulating a works-based program intended to promote soil conservation. The intention is that the program will have the dual benefit of providing an immediate transfer to the poor while at the same time increasing the production of food in the area where they live. It is expected to build on the Ministry of Forestry program and to employ about 10,000 people annually during six months in the lean season. 150. In many developing countries, employment on public works at low wage rates has been used as an effective way to make transfers to the poor. It has the benefit of potentially being self-targeting (because only those who really need the transfer will show up and work) and of being more politically and socially acceptable than other forms of transfers (because it is not seen as a handout and because it is perceived as financing a productive investments). Furthermore, public works can be scaled up and down in response to cyclical needs or can be re-targeted geographically, and, if well-designed, they create physical assets that contribute to rural development and raise longer-term incomes in the areas (see 71 Box 5). However, the implementation of public works can be challenging, and their labor- intensity can vary greatly. This means that public works can be an expensive way of making transfers, costing $2 to $3 for every $1 of benefit transferred to the poor if a concerted effort is not made to channel a large share of expenditure to wages and to design the scheme to target the poor.119 Box 5: Public Works to Support Rural Development - An Example from Ethiopia Ethiopia’s Productive Safety Net Program (PSNP) was designed to replace the emergency humanitarian appeal system as the main instrument in the country’s safety net. The PSNP was launched in 2005, and one of its goals is to heighten the impact of safety net transfers through public works investments. These investments are mainly in soil and water conservation and the creation and maintenance of roads and other types of infrastructure and are expected over time to help households to graduate out of food insecurity by increasing agricultural productivity and supporting the development of the rural economy. From a menu of eligible investments, the PSNP adopted an integrated approach to watershed planning that aimed to reverse severe environmental degradation. Because of the protective function of the PSNP, public works needed to be labor-intensive to transfer as much income as possible to the recipients. A target labour- intensity of 80 percent was adopted at the district level but could be lower or higher at the individual project level. PSNP public works are operated on a large scale and generated more than 172 million person days of labor in 2007. Sub-projects are chosen by community committees, and training is provided on Community Participatory Watershed Development to the development agents. Once the community develops a shortlist of desired public works, the development agent designs the sub-projects with assistance as required from district technical staff. The design plans are then reviewed and approved based on the Community Participatory Watershed Management Guidelines. An impact assessment conducted in 2008 concluded that the main positive outcomes of the PSNP’s public works have been: environmental regeneration (10 sampled communities have seen increases in the water table, vegetation cover, and species diversity); increased access to water supply (the PSNP doubled the number of water supplies in the sampled villages, and households reported traveling shorter distances to access water supplies); expanded use of small-scale irrigation (this helped to increase livestock for 4 to 12 percent of households);119 additional farmer training centers; and time savings from road investments. The public work program also served as a complementary source of financing for routine maintenance. Sources: Grosh et al (2008) and Van Domelen and Coll-Black. 119 This is because of the costs of complementary materials, skilled labor, and supervision, as well as the costs of planning and organization of the works. The proportions vary greatly from country to country – the best programs have paid out up to 80 percent of their costs in wages to unskilled labor, while the worst have paid less than a third. 72 151. A public works program seems to make sense for Lesotho, but it would be advisable for policymakers to introduce it in pilot form initially and to carefully consider the wage level. Given the fairly uniform spread of poverty in rural areas, a self-selecting employment program for the able-bodied poor may make sense in the context of Lesotho. However, no established institution currently exists that could be expanded to run a large- scale public works program so it would be advisable to implement a pilot scheme to learn lessons about what works and does not work in the context of Lesotho before introducing a full-scale program. Also, Lesotho has a fairly high administered minimum wage (M.48 per day), and, if the public works program were to be required to strictly adhere to this minimum wage, then it is unlikely that it would be possible to self-target the poor. The total transfer from working for 20 days, for example, would be M.960 per month, which is substantially higher than the poverty line income, and which, therefore, is likely to attract many non-poor applicants. Therefore, policymakers might wish to consider exempting public works from the minimum wage legislation, thereby making it possible to set the wage level for the program low enough to encourage self-targeting by the poorest. 152. Two key lessons from successful PWPs in other countries are that: (i) the transfer needs to be sustained and predictable to reduce poverty and (ii) the works need to be labor-intensive enough that a large share of spending goes on the wages paid to unskilled workers. 153. Given the importance of labor-intensity in public works and the fact that there is a large problem of under-funded road maintenance in Lesotho, one option to consider might be to increase the budget for the routine maintenance of rural roads and to use this money to employ large numbers of poor people on such work at the local level. A recent study120 examined the very large backlog of under-funded road maintenance. Given Lesotho’s mountainous, unstable terrain and concentrated rainfall during the rainy season, earth 120 Africon Consulting (2010). 73 roads deteriorate quickly. In the absence of regular maintenance, investments made in road construction are often lost, but spending on routine maintenance is a low priority and any available funds tend to be allocated to urban and paved roads. The study estimated that about M.73 million (US$9.7 million) should be spent annually on maintaining low-level earth roads. Although not targeted specifically to the poor, such spending could create a large number of regular income opportunities in remote areas. For example, if the M.73 million were spent on earth road maintenance and 50 percent of that sum were paid in wages, then it could provide employment transfers to 30,000 poor households with each of them receiving about M.1,200 annually (US$171).121 Experiments in South Africa and Malawi have shown that such maintenance programs can be used to employ village women and provide them with a predictable source of funds each year. Agricultural-Related Transfer Programs Agricultural Input Fairs: Input Vouchers for the Poor 154. Recognizing the shortcomings of the universal subsidy (see the section on the National Fertilizer and Input Subsidy below), in 2007 the government adopted a targeted scheme that has been operated for the past several years with support from FAO/EU. It targets input subsidies to 22,500 smaller farmers. The scheme was launched partly in response to the global rise in prices of both food and inputs. The estimated cost of planting a hectare of maize, for example, had risen from $60 in 2006/07 to US$275 by December 2008.122 It was also partly intended to address the fact that small farmers were not getting access to fertilizer sold under the general subsidy scheme and that private traders were not operating in the remote areas where many poor farmers live. 121 Based on 50 percent of total expenditure (M.36 million) going to wages at an effective compensation rate of M.30 per day, and average employment of 40 das per household, yielding a total transfer of M.1,200 per household – which could be concentrated at one time or spread across the year. 122 Mokotimi (2011), p.1. 74 155. Coverage and Benefits. The program reaches about 22,500 farmers annually, which implies coverage of about 112,500 total beneficiaries or about 6 percent of the population. At the agricultural input fairs, selected farmers are provided with input vouchers that are worth about half the price of the inputs for a half-hectare of maize, beans, or other basic crops. The fairs focus on providing seeds as much, or more, than fertilizer. The beneficiaries are supposed to be small but viable farmers, with preference in theory given to those who are particularly vulnerable or food-insecure (such as female or child-headed households or those with poor food security). 156. The value of the vouchers provided was M.800 per household in 2009/10 and M.630 (about US$84) in 2010/11.123 The actual value of benefits to the household depends on what they are able to produce from the fertilizer and seeds that they have been given and, thus, on the value of their additional crop output in terms of the food that they would otherwise have to buy from the market. In the absence of any rigorous analysis of output effects, it is not possible to quantify this value, but it could range from zero (if there is no supply response) to as much as a whole year’s worth of maize and beans. If it were 500 kilograms of maize, it might have a market value of M.2,000 (about US$500) or more.124 157. Implementation and Targeting. The fairs are scheduled at the beginning of the planting season and are held at 68 agricultural resource centers spread throughout the country. The fairs are run by the Ministry of Agriculture and Food Security (MFAS) with support from FAO officers. They make inputs available to farmers in areas where private traders might not otherwise operate and thus reach smaller farmers who generally do not get access to the subsidized inputs provided under the general subsidy scheme (which tend to be available only in urban centers). 123 The difference reportedly being due to changes in the Maloti exchange rate and input costs. 124 Based very roughly on having to purchase maize at M.80 per 12.5 kilogram bag. Actual benefits could be much higher based on the additional value of beans, and the highest levels of incremental output reported in post-harvest surveys (see Table 18 below). However, the incremental value of production becomes much lower as farmers move into surplus production and sell the excess produced at farm-gate prices. 75 158. The FAO and MAFS organize private traders to participate in the fairs and provide vouchers to the selected beneficiaries to redeem on the day of the fair. The traders can then take the vouchers back to the FAO to be redeemed. Prices at the fairs are actually slightly higher than regular retail prices for the same seed, in part because traders have to provide small quantities and incur the expense of traveling to remote places 125 (see Table 18). 159. Local MAFS staff, chiefs, and councilors draw up a list of qualifying beneficiaries for the area covered by each fair, and these identified beneficiaries are invited to attend (anyone can attend and buy inputs at the fairs, but only selected beneficiaries receive the vouchers). On the day of the fair, the names of those selected are read out, and they collect vouchers in various denominations, which they can then spend on the seeds, fertilizer, or tools being sold by the participating dealers. Table 18: Agricultural Fair, Subsidized, and Retail Prices for Selected Agricultural Inputs a/ Input Trade Fair Price Retail Price Subsidized Gov’t Price Maize Seed (AFG4410) 10kg. 290-310 250 100 Wheat Seed (Gariep) 10kg. 300-400 290 200 Bean Seed (PAN 148) 25kg. 280-330 250 260 b/ Fertilizer (2:3:2) 115 115 115 Source: Mokitimi (2011), Appendix B. Notes: a/ Note input fair prices vary by location with prices in the mountains being about 10 to 20 percent higher than in the lowlands. b/ Our understanding is that only government-subsidized fertilizer was sold at the 2010/11 AIFs. 160. Targeting. Beneficiaries are selected by MFAS staff in consultation with local chiefs and community councilors. Beneficiary households are supposed to have the following characteristics: (i) access to at least 0.5 hectares (1.2 acres) of arable land and a history of successful farming; (ii) have at least one able-bodied member; and (iii) have left at least 50 percent of their available land fallow the previous season because they were unable to afford inputs. In addition they should exhibit vulnerable household characteristics, such as: 125 In 2010/11, MAFS also participated in the fairs as a supplier, selling fertilizer at the general subsidy price. 76 - Having no access to formal income sources or remittances. - Being headed by females or children or having members who are orphans. - Lacking livestock. - Having to purchase or otherwise obtain a high proportion of food consumed due to their own low production. - Having chronically ill members. 161. Clearly there are some inherent contradictions between the two sets of criteria. For example, child- and female-headed households are less likely to meet the second requirement (to have at least one able-bodied member), and few households with access to at least 0.5 hectare would own no livestock. Nevertheless, the criteria, if rigorously enforced, should result in a legitimate target group for safety net support that constitutes the poorest farming households. In reality, there is no easy way of assessing factors such as land ownership, the amount of land left fallow, or reliance on purchased food, and it seems unlikely that MAFS staff and community leaders would be able to apply these criteria rigorously in the short time available to them to select beneficiaries. 162. There is only limited evidence of the effectiveness of targeting or of whether, and how, these criteria have been applied. A 2010 post-planting survey of beneficiaries (FAO (2010)) found a significant share (ranging from 36 percent to 58 percent) were either female, elderly, or child-headed households, and the vast majority were not employed, suggesting that the vulnerability criteria were generally being applied. About two-thirds of beneficiary households had run out of food from their own crops, implying they are among the poor, if not the poorest, in their communities. However, there are also reports of favoritism, of chiefs selecting their own relatives, politicians selecting party supporters, and MFAS staff selecting prominent farmers who will produce quality crops for agricultural 77 shows,126 but, in the absence of any rigorous evaluation of beneficiaries, it is not possible to say whether these examples represent a significant distortion of the targeting process. 163. Expenditure, Financing and Cost-effectiveness. In 2010/11 386 tons of seeds were distributed and 201 Mt. of fertilizer. Based on the program’s average of about 18,000 beneficiaries per year over the past two years,127 the face value of the vouchers alone would be about M.13 million128 or US$1.7 million per annum. The total costs are not known, but the program is labor-intensive, with substantial involvement by FAO staff (for example, in printing and distributing the vouchers) and of ministry staff (about 10 staff members are required at each fair). Program budget documents for 2011 suggest that the program’s management and delivery costs represented between 23 percent and 31 percent of total program costs,129 with a total annual cost of about US$2.5 million.130 164. Impact and Commentary. The 2010 post-harvest survey of beneficiaries131 showed that beneficiaries had had significantly higher yields and food production than non- beneficiaries (Table 18), suggesting that the scheme has had a major positive impact on household food consumption. The survey also contained a qualitative assessment 132 that concluded that most beneficiaries’ food security situation had improved and that the area under cultivation had increased. Other benefits cited by the assessment were the savings in transport costs to small farmers and the fact that, without the program, private traders would not have operated as they were undermined by the general subsidy provided by the government. 126 Mokitimi (2011). 127 The program reached 22,550 households in 2009/10 and 13,185 in 2010/11. The reason for the decrease is not clear but was presumably due to program budget constraints. 128 There were 22,551 vouchers issued at M.800 each in 2009/10 and 13,184 vouchers issued at M.630 each in 2010/11. 129 FAO and Irish Aid program budget data provided to the authors, September, 2011 130 Estimate based on voucher costs for 22,000 beneficiaries plus overheads. However, the annual cost of the program varies per year depending on the projects to be funded and the number of beneficiaries to be supported. 131 FAO (2010) 132 FAO (2010) p.9. 78 Table 19: Impact of the AIF Program on Household Maize Production Zone Maize Production Maize Yield Food Availability (Mt.) (t/ha.) (months) Lowlands Beneficiaries 2.49 1.04 27 Non-beneficiaries 0.40 0.33 5 Foothills Beneficiaries 2.70 2.25 25 Non-beneficiaries 1.09 0.91 15 Senqu River Valley Beneficiaries 1.34 0.73 20 Non-beneficiaries 0.47 0.53 7 Mountains Beneficiaries 1.00 0.95 11 Non-beneficiaries 0.81 0.95 6 Source: FAO (2010). 165. The program clearly has some concrete benefits. It makes inputs available to small farmers in areas where private traders might not otherwise operate, and it saves farmers the cost of travelling long distances to buy inputs. Above all, it potentially provides a highly leveraged form of transfer to food-insecure smallholder farmers. It also avoids the negative problems associated with the general subsidy, which undermines the development of private traders and constitutes a regressive transfer because it increases instead of decreases with increasing land ownership and the ability to buy inputs. Restricting beneficiaries to a relatively small amount of inputs is potentially self-targeting (or at least de-selects the better-off) but may not target the poorest of smallholder farmers. 166. Clearly it is not rational to provide a larger proportional subsidy to better-off farmers under the general subsidy scheme and a smaller one to poorer farmers under this more poverty-targeted program. If the program can be effectively targeted to the intended beneficiary group and if it can consistently generate the kind of supply response that can potentially be achieved, then this program, or something like it, can potentially be a cost- effective productive safety net program. A more rigorous assessment should be carried out of the poverty status of the households that the program is currently reaching based on data from a sample survey of existing beneficiaries before policymakers decide whether to expand it further. Finally, the program suffers from the obvious problem of being dependent on financing and implementation support from donors, and therefore, if the 79 evaluation determines that this would be useful, it should be more integrated into the routine operations and financing of the government. Other Transfer Programs 167. In addition to the safety net programs described above, the government also spends a significant amount on two other transfer programs: the National Fertilizer and Input Subsidy program and the Tertiary Bursary program. While neither of these is a direct safety net programs, they are worth examining in the context of reforming the existing safety net in Lesotho, since spending on these programs could potentially be redirected to more targeted programs to protect the poor. National Fertilizer and Input Subsidy 168. The government currently provides a general, untargeted subsidy on fertilizer and other agricultural inputs. 169. Coverage and Benefits. The subsidy is provided across the country and has varied between 30 percent and 50 percent of the retail cost. At present, the subsidized fertilizer is sold at M.110 per 50 kilogram bag compared to a market price of about M.250 per bag. So farmers who can get access to the subsidized fertilizer enjoy a subsidy of M.140, or about US$20, for every bag that they buy. The actual benefit is the net additional value of the food that a farmer can grow with the subsidized fertilizer and seeds, but we have no way of knowing what that is without detailed tracking of how farmers use the inputs and the yields that they achieve. 170. Expenditure, Financing, and Cost-effectiveness. The program is fully funded by the government. Spending on the program varies from year to year depending on the budget allocation. In 2009/10, it was M.59 million (about US$8 million), but in 2010/11 it was only 80 M.18 million ($2.6 million), although this drop was partly because stocks of subsidized fertilizer were carried over from the previous year. 171. Implementation and Targeting. The subsidy program is implemented by the Ministry of Agriculture. The government procures inputs from South Africa and then distributes them to private traders, with the stipulation that they must sell at fixed retail prices. At present fertilizer is sold to traders at M.80 per bag and they must retail it at M.110. As a general subsidy, the program is untargeted. In theory, farmers must confirm the average amount of land that they cultivate before accessing subsidized fertilizer – in order to prevent farmers from stockpiling it – but there is no evidence on whether or not this is enforced. 172. Impact and Commentary. The total amount allocated to the program is not adequate to cover the total volume of inputs, so the subsidy must be de facto rationed. However, there is no targeting mechanism, so it is highly unlikely that it is benefitting poorer farmers. There are no data on who benefits from the subsidy, but to the extent that larger farmers have more land, and therefore use more fertilizer, and have more cash, and therefore are more likely to be able to purchase the subsidized fertilizer, it seems evident that the subsidy is regressive. Table 20 illustrates the proportion of the subsidy that would accrue to farmers with different landholding sizes if it were distributed randomly. Table 20: Distributional Impact of the Fertilizer Subsidy Scheme, Some Illustrative Examples Landholding Size 0 0.5 1.0 5.0 20.0 (ha.) (Very Poor) (Poor) (Moderately Poor) (Middle) (Wealthy) Fertilizer Use (bags) 0 4 8 40 160 Value of Subsidy 0 560 1,120 5,600 22,400 (US$) 0 $80 $160 $800 $3,200 Source: Author’s calculations. Notes: Fertilizer use is illustrative based on four 50 kilogram bags per 0.5 hectare and a subsidy of M.140 per bag. 81 173. The general subsidy scheme is not intended to be only a poverty reduction program but also to boost agricultural production and promote fertilizer use, as well as to insulate farmers from the impact of large international increases in fertilizer prices in recent years. As a poverty reduction transfer, fertilizer subsidies can potentially have a large consumption effect – generating up to $3 worth of benefits in the form of additional maize production of the poor for each $1 worth of fertilizer provided (this is assuming that the poor household is a net buyer of maize and that the additional production saves him or her from having to buy maize during the lean season). However, the current general subsidy is not having this effect, in part because the scheme does not appear to be resulting in the hoped for supply response (probably because of sub-optimal application rates), in part because it is geographically concentrated in non-poor areas (see Annex 4), and in part because the incentives for abuse are high. Since the scheme cannot provide all of the fertilizer needed, many farmers must buy inputs at the full commercial price, so any farmer who gets subsidized fertilizer can re-sell it at the commercial price within Lesotho or in South Africa. 174. There are additional problems associated with the National Fertilizer and Input Subsidy. It undermines the development of a network of private traders since retailers and importers are unwilling to invest in stocks of fertilizer and seed because of the uncertainty regarding government sales and pricing. Furthermore, it runs against the government’s stated policy of facilitating the development of agricultural markets rather than directly intervening in them. It also consumes the administrative capacity of the Ministry of Agriculture and Food Security staff who spend time implementing the subsidy scheme at the expense of providing extension support, and it is open to corruption and to being captured by those who are well-connected. Tertiary Bursary Scheme 175. The largest single transfer program operating in Lesotho in terms of its budget is the higher education loan bursary scheme, which pays the fees, living expenses, and costs of 82 books for students who attend universities and other tertiary education institutions. Although in theory a loan scheme, there is almost no repayment, so it is effectively a pure transfer program. 176. Coverage and Benefits. The bursary scheme was established in the 1970s to ensure an adequate supply of local graduates to staff the civil service. However, it has since been transformed into a general program that provides a bursary to anyone who attends university whether in Lesotho or outside the country. The total sums paid annually in respect of each student in Lesotho institutions range from M.7,000 to M.60,000, while for students in South Africa the sums are between M.40,000 and M.60,000 per student per year. Graduates are supposed to repay the loans on a sliding scale (50 percent if they are employed by the government, 55 percent if they get a job in the private sector in Lesotho, and 100 percent if they are employed outside of Lesotho), but actual repayments represent only about 4 percent of the annual outlays,133 without even allowing for administrative costs. In addition, there is no interest charged, so in real terms the program is costing the government even more than the estimated M.575 million. 177. Expenditure, Financing, and Cost-effectiveness. The program is fully funded by the government, and its total cost per year is about M.575 million (US$76.7 million). Financially, the program has grown to such a scale that it now represents a very significant share of education expenditure (in 2008 the tertiary sector in Lesotho received 36 percent of sector resources), costing 3.5 percent of GDP, and absorbing about 6 percent of the government’s total expenditure in 2010/11. 178. Impact and Commentary. Benefit incidence analysis shows that only 1 percent of students who go to university come from the poorest quintile and that 85 percent come from the two richest income quintiles. One of the stated objectives of the program is to 133 Actual expenditure M.596 million in 2010/11 with collections amounting to M.19 million. Tables provided by the National Manpower Development Secretariat (NMDS). 83 help needy students to attend university, but most such children never get to university because their families cannot afford the costs of secondary schooling. At the moment, the program is completely untargeted, with no means testing for household income. Funds are thus provided to many students, whose families are in a position to pay tuition fees and living expenses, thereby crowding out the few low-income students who do manage to qualify for higher education but at lower qualification levels. Furthermore, students who attend university outside Lesotho are supposed to repay only an amount equivalent to the fees they would have paid at the University of Lesotho, not the actual amount that they receive, resulting in a higher subsidy for those who study outside the country who presumably also tend to be among the better-off. In 2010/11, such-out-of-country payments accounted for about 20 percent of disbursements from the bursary fund (mostly for students studying in South Africa). 179. While promoting tertiary education may well be a rational objective of the national development strategy (and even of a poverty reduction strategy), the program as it is currently structured suffers from a number of fundamental flaws: it is costly, it mostly benefits households who do not need a public transfer, and it needs to be made more financially sustainable. 180. Several studies have concluded that the design of the program is no longer appropriate or sustainable, and there is substantial room to increase its cost- effectiveness.134 The government recognizes this and is considering how to restructure the program over time.135 Options for reform include better targeting and means-testing and making the program more commercially oriented and more truly a loan scheme. 134 See, for example, Restructuring of the National Manpower Development Secretariat and Revolving Fund, 2005 (MFDP (2005). 135 The World Bank is currently supporting the government to prepare a technical note on the program. 84 Observations on the Overall Package of Transfers 181. In conclusion, a number of observations can be made on the current overall package of transfer programs:  Total spending on all transfers is high. Based on the analysis in this report, total spending on transfer programs in Lesotho amounts to about US$197 million annually or roughly US$104 per person. If this amount were targeted to the poorest 25 percent of the population, it would provide a transfer of about $319 per person, almost equal to the poverty line income.136  Only a small share of the poor is being reached despite almost M.1.5 billion or 16 percent of national revenue being spent on various transfers. Most existing programs reach only a few percent of the poor (see Table 24 and Figure 7).  The benefit levels provided by most programs appear to be high relative both to the poverty line income and to those provided in other countries (examples include the benefits provided by the School Feeding Program and the Old Age Pension).  The package seems to be missing some relevant aspects of poverty. Although the package of safety nets is addressing some of the relevant aspects of poverty discussed in Chapter 2, it is missing some others, for example, seasonal poverty and malnutrition.  Benefits do not seem to be well-targeted to the poorest. In the absence of any data on the incidence of any of the programs, it is not possible to say how well targeted they are to the poorest. However, a priori it would appear that the existing package of benefits is not well-targeted to the poorest (especially the Old Age Pension, the agricultural subsidies, and the School Feeding Program). 136 Based very roughly on 525,000 people, a rough poverty line of US$340 (based on a recent UNDP estimate of M.200 per month), and program operating costs of about 15 percent. 85  There is no overall framework for coordinating transfers or programs. As a result, there are almost certainly overlaps and duplication of beneficiaries, with some of the poor receiving benefits from multiple programs and some receiving none.  The existing package of programs is not counter-cyclical, in other words, it is not capable of being scaled up or down in response to changing economic conditions or to shocks such as drought.  None of the existing programs have been rigorously evaluated either with respect to whom they are reaching or to their impact on poverty or consumption. This makes it difficult for policymakers to know with any certainty which programs make the most effective use of resources in reducing poverty. While this report is a first attempt to evaluate costs and effectiveness, more rigorous examination of the poverty levels of beneficiaries and of the impact of safey net programs on their consumption should be a priority.  There is scope for increasing the promotive impact of programs. While some of the existing programs are productive in that they aim to raise the poor out of poverty in the longer term (for example, the OVC Bursary Scheme and Agricultural Input Fairs), there may be some scope for increasing the promotive impact of programs either by adding conditions to transfers related to investments in human capital or by using public works programs to create productive assets.  The predictability of the transfers is a challenge. When budget allocations and donor funding for programs vary from year to year, this can make the payment of transfers unpredictable, which means that recipient households cannot rely on having that support when they need it. Evidence has shown that predictability is a fundamental requirement to ensure that transfer programs have the maximum possible impact.  The weakness of data, both on household poverty and on the coverage and impact of programs, makes it difficult for policy-makers to know which programs are effective at reducing poverty, and to determine if they represent value for money. 86  The lack of a single agency to coordinate expenditures and programs has led to problems of duplication and overlap in which some needy households benefit from multiple programs while others who are equally in need benefit from none. (Issues of coordination and institutional capacity are further discussed in Chapter 4). 182. The implications of these findings for moving forward with a national safety net strategy are discussed in Chapter 5. Coverage, Costs, and Benefit Levels of Current Programs Coverage 183. As Table 21 shows, the vast majority of existing transfer programs covers no more than a few percent of the population – or even of the very poor. Based on these rough assumptions regarding the incidence of benefits, only the School Feeding Program reaches a significant share of the poorest (estimated at slightly less than a quarter), while the Old Age Pension, the Child Grants Program, and the Nutrition Support Program each reach about 5 percent (one-twentieth) of the poorest. The other programs reach only a very small proportion of the poorest – typically around 1 percent. Figure 6 shows the same information graphically. 87 Table 21: Coverage of Existing Transfer Programs Approximate Approx. % Approx % of Program Coverage of Population the Very Poor (2010, Beneficiaries) Old Age Pension 83,000 4.4% 4.4% a/ Child Grants Program 27,200 1.5% 3.9% Public Assistance 9,500 0.5% 1.3% b/ School Feeding Program 389,000 20.5% 22.5% c/ OVC Bursary Scheme 20,000 1.1% 1.5% d/ Nutrition Support Program (WFP) Child Nutrition 37,000 1.9 % 3.2% Blanket Feeding 32,000 1.7% 2.2% ECD Feeding 30,000 1.6% 1.6% e/ Agricultural Input Fairs 21,600 1.1% 0.8% f/ National Fertilizer and Input Subsidy n.a. n.a. n.a. g/ Tertiary Bursary Scheme 16,200 1% < 0.1% Integrated Watershed Management Public Works 58,000-115,000 3-6% 2-4% h/ Program Notes: This table based on an estimated population size of 1.9 million, an estimated extreme (food) poverty rate of 30 percent, and rough assumptions regarding proportions of beneficiaries that are among very poor under each program. a/ This is the number of beneficiary children in approx. 6,906 households (as of September 2011 payments). b/ Includes both government and WFP-funded schools. c/ Includes both government and Global Fund-funded OVC bursaries within government system. d/ Coverage of individuals - based on discussions with WFP Lesotho. e/ This is the number of households, with the number of individual beneficiaries estimated at 108,000 (based on estimated five people per household). f/ Based on beneficiaries as reported in World Bank (2010a), a review of the Tertiary Bursary Scheme. g/ Estimates vary depending on how long it is assumed household members are employed. The program employs 9,600 people at any one time – estimates here correspond to one or two months average employment and refer to number of households benefitting. 88 Figure 7: Existing Programs, Estimated Coverage of the Very Poor 25.0% 20.0% 15.0% 10.0% 5.0% 0.0% Poverty Relevance – Benefit Levels 184. Table 22 shows the benefit levels for the existing programs and how they compare with the estimated food poverty line of M.185 per month per adult. The Old Age Pension delivers a significant share of the poverty line income as do the Agricultural Input Fairs (which can generate consumption gains well above the poverty line). The Child Grants Program provides a transfer that is a bit less than one-fifth of the poverty line – this may be acceptable to the extent that it represents a top-up to households who have other viable income sources but is inadequate for truly destitute families. Similarly the Integrated Watershed Management Public Works Program probably provides too low a transfer to significantly increase consumption among the poor. The two bursary programs provide very large transfers relative to the other programs. While it is recognized that they represent a different form of investment, the very high level of benefits under the Tertiary Bursary Program, which tend to go to the non-poor, highlight the need to revisit these transfers. 89 Table 22: Benefit Levels of Existing Programs Benefit Level As % of Food Assumed Unit Program (2010Maloti) Poverty Line of Coverage Income Old Age Pension 300 monthly 47% Household a/ Child Grants Program 120 monthly 17% Household b/ Public Assistance 100 monthly 54% Individual c/ School Feeding Program t.b.q. To be quantified Individual d/ OVC Bursary Program (MoET) 2,000 annual 90% Individual e/ Nutrition Support Program (WFP) n.a. n.a. Individual f/ Agricultural Input Fairs 630+ annually 7%-50% + Household g/ Tertiary Bursary Scheme <37,800 annually 1,700% Individual Integrated Watershed Management Public 960 annually 11% Household h/ Works Program Notes: Based roughly on estimated food poverty line of M.185 per adult equivalent per month and household size of four adult equivalents in cases where benefits are assumed to be shared with other household members (Child Grant Program, OAP, Agricultural Input Fairs, and Integrated Watershed Management Public Works). a/ Child grants benefit averaged over household size of four adult equivalents. b/ Public assistance based on single individual – proportion of poverty line income lower to the extent that benefits are shared among family members. c/ School feeding based on standard school meal under government program provided for 180 days per year (averaged over the year, implies equivalent transfer of 15 meals per month annually). d/ OVC bursary based on total expenditure averaged over number of beneficiaries, but is likely to be lower in reality due to administrative costs. e/ Nutrition Support Program: based on standard monthly ration of 0.45 kg of sugar, 0.6 kg. of vegetable oil, and 6 kg. of corn-soya blend; retail value of this transfer to households has not been quantified. Low estimate based on three months per year (blanket feeding) and high estimate on 12 months (continuous support for acutely malnourished). f/ Agricultural Input Fairs – benefit depends entirely on how much is produced using inputs. Minimum 7 percent of food poverty line is based on retail value of voucher alone, though highest levels of production can exceed poverty line income. g/ Tertiary bursary based on total expenditure of M.575 million and estimated 16,200 beneficiaries in 2011, h/ Integrated Watershed Management Public Works Program based on employment of one month annually per household (20 days) at M.48 per day. 90 Table 23: Costs and Cost-effectiveness of Current Programs Total Cost Approx Annual Cost Beneficiary Unit a/ Program (Annual) per Beneficiary Old Age Pension M.371 million $589 Individuals (US$49.5 million) Child Grants Program M 16.5 $220 Household (US$2.2 million) Public Assistance Grant M.13 million $181 Individual (US$1.7 million) School Feeding – Government M.236 million $81 Individual ($US31.5 million) OVC Bursary Scheme M.48 million $320 Individual (US$6.4 million) Nutrition Support Program (WFP) M.47 million $74 Individuals (US$6.3 million) Agricultural Input Fairs M.17 million $122 Household (US$2.3 million) National Fertilizer and Input Subsidy M.44 million -- Beneficiaries data (US$5.9 million) n.a. Tertiary Bursary Scheme M.575 million $4,733 Individuals (US$76.7 million) Integrated Watershed Management M.112 million $130-257 Households Public Works Program (US$14.9 million) Notes: a/ 2011 or most recent year; based on exchange rate of 7.5 Maloti/US$, averages based on expenditure as per estimates in text and Table 14. 185. The Old Age Pension and the bursary programs are fairly expensive in per capita terms. The agricultural input fairs, subject to further investigation of costs and benefits, appear to be a very cost-effective way of transferring resources to the poor. The feeding programs, although they achieve other purposes, are not particularly cost-effective as a transfer mechanism as greater consumption increases could be achieved just by giving households an equivalent or a smaller amount of cash. This is particularly true of the School Feeding Program. The cost-effectiveness of the Integrated Watershed Management Public Works Program cannot be accurately assessed until detailed documentation is available on the composition of costs under the scheme, but it appears to be a cost-effective means of transferring resources. Finally, the Child Grants Program should be a low-cost way of making transfers to the poor, but its cost-effectiveness will not become clear until the initial investment period is over, and it is seen whether the current level of expensive technical assistance can be scaled back (Table 23). 91 CHAPTER IV: INSTITUTIONAL ISSUES, PROGRAM DELIVERY, AND TARGETING 1. Current National Social Protection Policy and Institutions 186. The government is in the process of finalizing a National Social Development Policy that emphasizes the critical role played by safety nets in combatting poverty. The draft National Social Development Policy (NSDP) indicates that the goal of the government’s policy is to prevent and reduce poverty with a view to ensuring that all of the Basotho enjoy an acceptable minimum standard of living. In order to achieve this objective, the draft NSDP emphasizes the need to develop and implement a comprehensive social protection system that includes safety nets. 187. Building on the NSDP, the government intends to develop a Social Protection Strategy to articulate the overall objective of its social protection sector (including social safety nets) and guide reforms. The government recognizes the need to develop a Social Protection Strategy to increase institutional coordination as well as harmonize legal frameworks and consolidate existing safety nets. The creation of the new Ministry of Social Development demonstrates the growing emphasis on social protection in Lesotho, but the effective coordination of all of these efforts will be one of the main determinants of success. One way to increase coordination and the efficiency of safety net programs would be to create a central agency responsible for coordinating them. The establishment of such an agency (and the building of its capacity) would not only increase coordination, thereby decreasing the risk of overlap of beneficiaries between programs, but would also make it easier to monitor programs. It is expected that the newly established Ministry of Social Development will take the lead on the social protection agenda moving forward. 137 137 The Department of Social Welfare in the Ministry of Health and Social Welfare was the leading the social protection agenda in the country until the creation a separate Ministry of Social Development in June 2012. The Child Grants Program and Public Assistance are being implemented by the new Ministry. Most of the larger transfer programs are implemented by other ministries. For example, the Old Age Pension is implemented by the Pensions Unit in the Ministry of Finance and Development Planning, the School Feeding 92 188. Development partners play a significant role in some of Lesotho’s safety net interventions by providing both technical and financial support. One such intervention is the National Information System for Social Assistance (NISSA), which was initiated as part of the Child Grants Program (CGP) with financial support from the European Union. The first phase of the pilot was finalized with significant implementation support from UNICEF and World Vision. In the second phase, NISSA will be expanded to cover more beneficiary households within the CGP and possibly three other programs, building on the lessons learned from the first phase. Another program that receives significant outside support is the Nutrition Support Program in which the WFP is responsible for delivering food to most of the remote areas served by this program. 2. Program Delivery and Capacity 189. The goal of the government is to build in-house and local capacity to improve the delivery of social safety net programs. The pilot implementation of NISSA is an important first step towards achieving this goal. In the first phase of NISSA, the government relied on significant amounts of outside support for both NISSA’s administration and financing. The goal of the government in the next two years is to build in-house and local capacity, in particular in the recently established Ministry of Social Development, to implement the pilot NISSA without having to rely on outside support. Towards this goal, in 2013, the government will pay the benefits under the CGP out of its own resources while 50 percent of the program’s administration costs will continue to be donor-financed. In 2014, the government plans to finance benefits as well as the administration costs of NISSA as reflected in its Medium-term Expenditure Framework. The Ministry of Social Development is in the process of finalizing the implementation plan for this expansion for the next two Program and the OVC Bursary Scheme are both implemented by the Ministry of Education and Training, and the National Fertilizer and Input Subsidy and the AIFs are operated by the Ministry of Agriculture and Food Security. Other safety net programs are implemented by the Ministries of Public Works and Transport, Forestry and Land Reclamation, and Local Government and Chieftainship. 93 years. This increased capacity will certainly contribute to improving the delivery of safety net programs in general. 190. Cash transfers need to be provided in a regular and predictable manner to have a long-lasting impact in terms of reducing poverty. While the technical and financial support of development partners is valuable and critical particularly in the initial phases of a program to supplement and build in-house institutional capacity, this support is typically designed to be short-term and unpredictable, which means that most interventions tend to be ad hoc and short-term. Therefore, the central agency for social protection would have a crucial role to play in coordinating and planning the support provided by development partners to increase the effectiveness and sustainability of social safety net programs. It is also critical to ensure that the support provided by the donors is adequately linked with institutional, organizational, and human capacity developments. 191. Feasibility and cost-efficiency are other important considerations in ensuring the effective implementation and sustainability of transfer programs. Programs that are initiated with support from development partners are sometimes very costly. Although these are often necessary to establish required systems, they can be overly sophisticated and difficult to manage. It is therefore important to focus on systems that are advanced enough to meet the needs of an efficient program but at the same time are manageable to implement at the national, district, and local levels. 192. The government recognizes the importance of gradual transition to implementation of NISSA relying only on local capacity.138 The Ministry of Social Development is reviewing the implementation arrangements of NISSA with technical assistance from the World Bank to ensure that any necessary modifications and 138 As noted in Chapter 3, the financing for the program is gradually being taken over by the government, which will be a major achievement but will present transitional challenges as the shift is made from donor disbursement systems to those of the government. 94 improvements are made before NISSA is scaled up nationwide. To use a PMT effectively, timely and accurate data on household consumption and assets are needed, which will need concerted support from the Bureau of Statistics and the Ministry of Development Planning. 193. The establishment of the pilot NISSA is an opportunity to create a national registry. NISSA currently contains a database of about 30,000 households that are receiving benefits under the CGP. The aim of the government is to eventually use NISSA as a national registry for all social assistance programs. This has worked successfully in many other countries (including Colombia and Brazil) 139 and has the potential to reduce the costs and increase the efficiency and coordination of safety net interventions. However, it will be crucial to ensure that there is adequate capacity to manage and maintain this database, including the capacity to collect data at the local level 194. Clear objectives and operational features are necessary to ensure that there is a broad common understanding and support for safety net programs. Broad community support for safety net programs helps programs to run smoothly. Also, programs that are too complicated or complex are difficult to explain to the public, and this can create confusion and in some cases mistrust. It is therefore important to make sure that the objectives and operational features (such as the eligibility criteria) of safety net programs are commonly understood and accepted. As was stated in some of the reviews of the OVC Bursary Scheme, there is a need for clear guidelines and a clear definition of vulnerability (for example, within the OVC definition) to ensure that there is a consensus about who should benefit. This will not only facilitate the targeting process but also lead to a better understanding within communities about who should benefit from the program. 139 Ghana is one of the African countries that has introduced a common targeting mechanism for some of the main safety net programs, and this is expected eventually to lead to a single registry. 95 195. Adequate payment mechanisms for cash transfer programs can not only increase the security of the transactions but also significantly reduce program administration costs and opportunity costs for beneficiaries. The government is already using outsourced payment mechanisms for some of the country’s safety nets. For example, Post Offices are being used to pay the Old Age Pension. In addition, the government is considering introducing electronic payment mechanisms to increase efficiency and security. Although there may be logistical challenges involved with certain payment mechanisms, it is worth exploring the most appropriate and most efficient ways to deliver transfers in the context of Lesotho. 196. Technology has the potential to improve the delivery and monitoring of safety net programs. There are many successful examples of how technology has been used to improve the delivery of safety net programs across the world, including in the Africa region. For example, beneficiaries can be identified and registered using biometric devices, GPS- enabled cameras can be used to monitor the outcomes of public works programs on the ground, and, in many African countries, governments have been increasing their use of electronic cash payments. One such country is Kenya where the government has introduced the agent banking model in its Hunger Safety Net Program and is also beginning to use this model in its CT-OVC program (Box 8). 96 Box 8: The Agency Banking Model for Paying Benefits in Kenya The agency banking model is currently used in the Hunger Safety Net Program (HSNP) in Kenya and on a pilot basis by a WFP-supported program in Northern Kenya. It is also being introduced in parts of the Cash Transfer for Orphans and Vulnerable Children (CT-OVC) program. In this model, banks are authorized to appoint agents to provide limited deposit-taking and withdrawal services. These agents are equipped with a combination of point of sale machines, card-readers, mobile telephones, and bar-code scanners for bill payment transactions. They may also have personal identification number (PIN) pads and personal computers that connect with the bank’s server using some kind of internet connection. Clients of the agent (including recipients of the transfer programs) use a magnetic strip card to access their bank account or use their mobile phone to access it through e-wallet. Customers are typically identified by a PIN but can also be identified using biometric information such as their fingerprints or photographs. The banking agents verify, authorize, and settle disputes over transactions as well as synchronizing and updating the transactions remotely with the bank’s central database. Source: World Bank (2012). 197. Safety nets can encourage households to invest in their own human capital if conditions are imposed on the receipt of transfers. In these conditional cash transfer (CCT) programs, beneficiary households are required to fulfill certain conditions, such as ensuring that their child is enrolled and attends school or making sure that their children are immunized or that their births have been registered. Evaluations of CCT programs have shown that they have often had a positive impact in terms of developing human capital. However, monitoring the actual implementation of conditions can sometimes be very costly and operationally challenging, and the success of the conditions depends on there being an adequate supply of school places or health facilities in the areas where the program is being implemented (see Box 9). 97 Box 9: Conditional Cash Transfer Programs Conditional cash transfers (CCTs) generally have two objectives: to alleviate short-term poverty by providing households with a cash transfer and to reduce intergenerational poverty by requiring families to invest in the human capital of their children. CCTs are programs that transfer cash, generally to poor households, on the condition that the recipient households make investments in the human capital of their children by fulfilling certain “conditions” set by the program. Such conditions are often associated with health and education outcomes, including regular health visits and vaccinations for children under the age of 5 and school enrollment and school attendance for older children. Recently, some governments have also started using these programs to achieve better nutritional outcomes by including conditions such as growth monitoring, prenatal care for mothers, and the provision of micronutrients and nutrition education (Peru’s Juntos was recently redesigned to achieve better nutrition outcomes). There have also been some pilot experiences (including in Tanzania and Lesotho) where CCTs were designed to encourage the prevention of sexually transmitted infections. CCT programs have spread rapidly throughout the world in the last decade, and the largest ones are in Latin America (for example, Mexico’s Oportunidades, Brazil’s Bolsa Familia, and Colombia’s Familias en Accion), but they are getting increasingly common in Africa as well (for example in Malawi and Kenya). Households that benefit from these programs are selected using a variety of targeting methods, often including community involvement to identify the neediest households. CCTs often target children from birth to some point in their teen years. The value of the benefit is usually equivalent to around 15 to 25 percent of baseline household consumption, and the cost of nationwide programs such as those in Brazil and Mexico represent around 0.4 percent of GDP per year. Evaluations have yielded evidence that CCTs have had a positive impact on human capital development by improving health, education, and nutrition-related outcomes, while increasing the consumption of the poor in both middle- and low-income countries. For instance, in Brazil, between 2003 and 2008, the CCTs explained 35 percent of the reduction in extreme poverty and 13 percent of the fall of inequality. In Colombia during the first two years of the program, beneficiaries increased their total consumption by 15 percentage points more than non-beneficiary households. In Malawi enrollment and attendance rates among beneficiary girls grew significantly more than the rates for those not receiving transfers. In Cambodia, enrollment rates among children in the program are significantly higher than those outside the program. In Nicaragua the CCT program reduced chronic malnutrition by 7 percentage points and reduced child labor in program areas by 2.5 percent in 2001 and 4.9 percent in 2002.In Mexico, the CCT program reduced poverty by 17 percent, and households consumed 7 percent more calories than previously. Finally, in Jamaica, the CCT program resulted in an increase in the number of preventive health visits by children of between 17 and 31 percent. Adequate design and implementation capacity are essential elements of any successful CCT program, and policymakers need to ensure that health and educational services are adequately equipped for the conditions applied to these programs. A CCT program needs to be carefully designed, with several key factors needing to be in place, including appropriate targeting mechanisms, a management information system, and clear definitions of the level and type of the benefit and of the conditions. The procedures for monitoring beneficiaries’ compliance with the conditions also need to be clear (which can often be complicated and administratively complex). It is also essential that accessible and effective health and education facilities are available in the areas where beneficiaries live. The administrative costs of CCTs include the cost of delivering the cash to the beneficiaries, the cost of targeting, and the cost of verifying the compliance of beneficiaries. These costs vary depending on the country’s administrative capacity and whether a national registry of beneficiaries is already in place. Sources: Barros et al (2010), Grosh et al (2008), and Bastagli (2011). 98 3. Targeting Issues and Choices 198. Deciding how widely or narrowly to target the benefits of transfer programs has a major impact on their costs, affordability, and potentially on their effectiveness in reducing poverty. Lesotho has opted for universal coverage of several of its major programs (the Old Age Pension, School Feeding, and the National Fertilizer Subsidy), while others have been more narrowly targeted (the Child Grants Program and the OVC Bursary Scheme, for example). Table 2424 shows the trade-offs between the universal and targeted approaches. In moving forward, the government will need to carefully consider which groups to prioritize. Table 24: Benefits and Drawbacks of Universal and Targeted Programs Universal Programs Targeted Programs - Easy to administer, no targeting required. - Potentially much less expensive if benefits - Expensive since benefits are provided to the whole can be targeted only to those in need. population. - Administratively difficult and the targeting - Politically and socially acceptable since everyone itself may be expensive. benefits. - May be socially and politically divisive as - By default results in large errors of inclusion (giving some members of communities are included benefits to people who do not need them) since it and others excluded. benefits everyone in a certain group. - Open to errors of inclusion (giving benefits to people who do not need them) or exclusion (missing out those who are entitled). 199. The challenge for policymakers is to find targeting mechanisms that can work within the existing constraints and to do so at reasonable cost . Targeting is particularly difficult in a country like Lesotho where: (i) a large share of the population is poor and there are few obvious differences between the conditions of the ultra-poor and the poor; (ii) there is no information on household income; and (iii) administrative capacity at the local level is limited, making it difficult to operate complex targeting systems. This section explores some targeting possibilities and summarizes the efforts that have been made in Lesotho to move towards a universal registration system. Box 10 presents the different targeting mechanisms that are available. Which targeting method is the most appropriate depends not only on the technical strengths and weaknesses of each method but also on 99 the available data, budget constraints, implementation capacity, and political preferences. The rest of this section discusses the pros and cons of each method in the Lesotho context. 100 Box 10: Different Poverty Targeting Methods Geographical targeting is easy to implement and monitor, and it typically involves much lower administrative costs than many other targeting methods. The rationale behind it is that, in many developing countries, the differences in living standards between communities or regions are often larger than the differences within communities or regions. Geographical targeting alone, however, implies that the non-poor who live in the targeted poor areas will inevitably benefit from the program, while poor people who live in the non-targeted areas will not be covered. In Lesotho, only the Child Grants Program (as described in Chapter 3, the CGP is scheduled to be rolled out to 22 out of 65 community councils by 2015) and the Nutrition Support Program currently use geographical targeting. Self-targeting has the benefit that the poor select themselves to participate in a program. This is usually done by offering employment on public works at a wage rate sufficiently low that only those who are extremely poor will take it or by providing very low levels of benefits such as very small fertilizer packs or inferior goods such as broken rice or yellow maize that would not appeal to any households that could afford to buy food. The difficulties with self-targeting are that policymakers often have difficulty keeping to the low benefit levels, thus defeating the objective of the system. Also, if extreme poverty is widespread, a large share of the population may self-select even at very low benefit levels, in which case some other form of filter is needed. In Lesotho at the moment, there is no truly self-targeted program. Community targeting involves using community groups, usually an elected committee or leadership group, to identify those families or individuals in a village who are eligible for support in accordance with certain criteria. Community targeting is often the only way of identifying which individuals fill very specific criteria (for example, not all orphans or disabled people in a village are destitute, and only local people can distinguish those that are in need and those that are not). The problem is that building the capacity to do community targeting well needs a lot of support and continuous monitoring to keep the system working. Committees need to be trained and their selections double-checked. There can be instances of favoritism and/or committees may be unwilling to make divisive decisions about who benefits so they may decide to spread benefits equally among all of the poor, thus again defeating the purpose of the system. Finally, experience shows that, while villagers can manage categorical distinctions (such as the destitute elderly or orphans), they often have difficulty in distinguishing the very poor from the poor in general on poverty grounds alone. In Lesotho, the Agricultural Input Fairs use a form of community targeting (consulting with Chiefs and councilors) and the Child Grants Program uses Village Assistance Committees. Means testing involves correctly measuring income and is the best way to determine the eligibility. However, it is very difficult to obtain correct earnings data in countries such as Lesotho. Proxy means testing was developed in order to address the weaknesses of means testing and has increasingly been used in social safety nets in developing countries. The method uses easy-to-verify indicators (such as the type of house that people live in) as proxies for their poverty status. However, proxy means testing has its own weaknesses. It is still labor-intensive to collect the data required and, due to the restrictions on the indicators used (the need for them to be easy to obtain and to observe), the final score is not perfectly correlated with consumption, which can lead to large inclusion or exclusion errors. In Lesotho, the Child Grants Program uses a PMT model, and an initial evaluation by Oxford Policy Management (OPM) noted challenges related to underlying data weaknesses and limited effectiveness in selecting the very poorest, although it also found the program had been fairly successful in reaching poor households (see Chapter 3). More generally, low levels of asset ownership in Lesotho especially among the rural poor implies that instruments such as proxy means testing that rely on differences in characteristics between poor households to target beneficiaries may be of limited use in distinguishing the poorest from the poor in general. Categorical targeting involves providing benefits to people with particular characteristics (for example, elderly people, families with children, or orphans). It is not strictly a poverty targeting mechanism because the policy objective may be to support all families with children or all old people rather than primarily to reduce income poverty. Categorical targeting can be used in conjunction with community targeting or proxy means testing to select particularly poor beneficiaries with the specified characteristic (as in the case of the Child Grants Program) or can be applied universally (as it is in the Old Age Pension and the School Feeding Program), in which case everyone who fulfills the criteria receives the benefit whether they are poor or not. 101 Which Method Makes Most Sense for Lesotho? 200. Lesotho is implementing both universal and strictly targeted programs. Lesotho has opted for – at one extreme – almost universal programs (such as the Old Age Pension, the National Fertilizer Subsidy, and the Tertiary Bursary Scheme) and at the other extreme a very tightly targeted approach (the Child Grants Program). The drawbacks of the first approach are the very high total program costs, the fact that they deliver a lot of benefits to people who do not really need them, and, as we have seen, the fact that few of the poorest are reached. This makes this approach an expensive way to reduce poverty. At the other end of the spectrum, the Child Grants Program attempts to identify the poorest households using a complex combination of categorical and community targeting along with proxy means test. The evaluation of the Child Grant Program140 showed that the households that were identified as eligible through the CGP targeting process were significantly more likely to be poor (74 percent) than those who were deemed to be ineligible (43 percent). However, the same evaluation also showed that only 51 percent of beneficiaries were among the poorest and that there were significant (67 percent) exclusion errors, meaning that a large number of very poor households were not being included in the program. These results indicate that there is substantial room for improvement in the targeting of the CGP program. 201. The lack of information on the targeting performance of safety nets programs in Lesotho makes it difficult to draw conclusions on the performance of individual safety net programs. In the absence of good data on where in the welfare distribution the current beneficiaries of Lesotho’s safety net programs fall, it is difficult to say how well the various targeting instruments have performed in Lesotho.141 It appears that community targeting has not been particularly successful (for example, in the Agricultural Input Fairs program), 140 Pellerano et al (2012) 141 The Ministry of Social Development and Bureau of Statistics will fill in this knowledge gap by adding a social safety net module to the CMS, with technical assistance from the World Bank. 102 but this may be because not enough effort has been invested in the community targeting process. A recent report142 made the following criticisms of the PMT formula: (i) the PMT model was estimated using nationally representative data and is not designed to reflect local differences in the poverty profile; (ii) the dataset used for the estimation of the model (from the 2002/03 HIES) was outdated and the quality of the data was reported as poor, with inconsistencies, outliers, and difficulties in data processing; (iii) partly as a consequence of poor data quality, a series of discrete models rather than a continuous model was implemented to predict correlations with consumption expenditure, leading to a significant loss of information and precision in the estimation; and (iv) the micro dataset used to estimate the model did not contain any information about key assets or income flows (such as pensions). Based on these findings, the government attempted to revise the existing PMT formula using the 2010/11 HIES data. However, some technical constraints related to the 2010/11 HIES data prevented the government from undertaking this revision. 202. Categorical targeting seems a rational approach to take in the Lesotho context, especially for some very specific groups (such as grandparent-headed households and double orphans), perhaps in conjunction with community targeting in some cases, to identify which of the elderly or disabled are destitute within a village as opposed to those who live with families that are capable of looking after them. 203. The fairly even geographical distribution of poverty across districts and ecological zones (see Table 5 in Chapter 2 and Annex 1) suggests that geographical targeting may not be very effective in the sense that large numbers of the poor would be left out. It is possible that a more fine-tuned approach at the sub-district level could target resources to pockets of the poor, but at the moment there is no sufficiently disaggregated database that would make it possible to identify or calibrate food poverty at this level. 142 Pellerano et al (2011b). 103 The Move towards a National Targeting System 204. To increase the efficiency of the safety net system, one approach that could be explored would be to have a single set of eligibility criteria for all (or a majority of the) social safety net programs in the country. If implemented effectively, the establishment of a common targeting mechanism could enhance objectivity and transparency, thereby minimizing the scope for interference and rent seeking in the selection process and making targeting more efficient. However, this would require a careful analysis of existing programs and their targeting criteria to ensure that the unified criteria served the purpose of each individual program. Furthermore, the operational aspects of such a single targeting mechanism would need thorough analysis to ensure that it is feasible and affordable to implement on the ground. 205. Lesotho is a small enough country (1.9 million people) that a universal registration system is feasible. At a minimum, it could form a unified database that could be used to determine who is benefitting from which program and to coordinate coverage. Whether it can be an effective targeting tool will depend on the reliability of the data generated for distinguishing levels of consumption. Progress on this will need to be tightly monitored and evaluated using independent surveys to determine if in fact the NISSA system is accurately identifying the households with the lowest consumption. If it is successful at doing this, then it might be able to distinguish pockets of poverty at the sub-district level, and geographic targeting would then be feasible in Lesotho. 104 Box 11: The Child Grant Program and the National Information System for Social Assistance (NISSA) in Lesotho In an effort to increase the coordination and improve the targeting of safety nets in Lesotho, the government has embarked on the establishment of the National Information System for Social Assistance (NISSA). The system has been launched and is currently being used for the Child Grants Program (CGP) described in Chapter 3. The CGP provides a regular transfer to poor households with children that are selected through a combination of an objective proxy means test (PMT) and community validation. The targeting for the CGP follows several steps. Following the formation of Village Assistance Committees (VAC) at the community level, a door-to-door census was conducted to collect information that would inform the first stage of selection ‒ using a PMT to discern five different poverty groups (from poorest to richest, called NISSA1 to NISSA5). Some categorical filters were also added at this stage, including the requirement that only households with children aged 0 to 18 be included in subsequent targeting steps. The next phase involved sharing the lists of all households that were registered in the door-to-door census with the VAC and asking them to validate the households that had been categorized as the poorest. Those households that both the PMT and the VAC considered to be the poorest were included in the final list of selected households. A recent report reviewed the targeting of the CGP, including the NISSA, and found that the households that were found to be eligible through the CGP targeting process were significantly more likely to be poor (74 percent) than those who were not eligible (43 percent). This was an encouraging finding, but the report also highlighted several issues that needed to be resolved in order to improve the system. Given that NISSA starts with a door-to-door census that collects data on all households at all income levels, the system has the potential to be used for all social assistance programs. Accordingly, the government plans to expand NISSA into a national system for all social assistance programs in Lesotho and is currently in the process of developing a road map for expansion (taking into considerations the findings of the recent report). The government aims to carefully analyze the current NISSA and to ensure that its design is in line with the government’s strategy for social protection and social assistance programs and to ensure that the design of the program allows for adequate data collection and system maintenance. Source: Pellerano et al (2012). 105 CHAPTER V: CONCLUSIONS AND RECOMMENDATIONS 206. The aim of this review has been to assess whether social safety nets or transfer programs can play a stronger role in reducing extreme poverty and inequality in Lesotho. The government is implementing a variety of such programs that are having an important impact on many people who are poor and vulnerable, but much of existing expenditure on transfers goes to the non-poor. Therefore, more could be achieved if these efforts were re- focused and if the targeted programs were expanded. In this chapter we: (i) assess the broad role that transfers might play in Lesotho’s national development strategy in the coming decade; (ii) suggest some options for a unified national safety net program that that could be developed over the medium term and give the associated cost implications; and (iii) make some specific recommendations for short-term measures to strengthen existing programs. 1. The Role of Safety Nets and Transfers in National Development Strategy 207. While achieving more inclusive growth is the ultimate solution to poverty in Lesotho, there are some strong arguments for providing targeted transfers to those in extreme poverty. The solution to rural poverty in the long run will almost certainly be the continued migration of rural residents to urban areas combined with increased agricultural productivity and greater connections to markets for those who continue in the rural sector. However, in the short to medium term, there is a strong case for providing targeted transfers to those most in need. The main argument in favor of this is that inequality of income remains high in Lesotho, and the country has relatively high per capita GDP, at least compared to most other Sub-Saharan African countries. A second argument is that shortages of cash income in rural areas constrain smallholder agricultural production and are limiting effective demand that hinders the development of markets and off-farm opportunities. A third argument is that there are a large number of people in some specific categories who are incapable of supporting their own minimum levels of consumption (for 106 example, child-headed and grandparent-headed households). All of these arguments point to the need to provide a combination of categorical support to some very specific groups of the ultra-poor in the poorest areas of the country, and of transitional productive support to those who are either being left behind by growth or who live in low-productivity areas. 208. The high level of inequality in Lesotho reinforces the case for transfers for two reasons. The first reason is purely arithmetic, in that taking a relatively small amount from the top end of the income distribution can provide a disproportionally large consumption boost to the very poor. The second reason is dynamic, in that the persistence of high inequality shows that the model of growth pursued so far has not been successfully raising the incomes of the poorest and, therefore that new approaches need to be considered, including providing direct transfers to the very poor linked with measures to increase their productivity and with conditions that encourage them to invest in their human capital development. 209. A lot is already being spent on transfer programs of various kinds in Lesotho, but most of it does not go to the very poor. Of the M.1,500 million (US$197million) spent annually on programs that provide transfers to individual Basotho households, we estimate that less than 25 percent of it goes to the extreme poor.143 In the current fiscal climate, this M.1.5 billion, equivalent to about 17 percent of the total revenue,144 represents a major expenditure commitment. If the government’s objective is to reduce extreme poverty, a much greater impact could be made by focusing spending on programs that narrowly target the poorest. There are several promising existing programs including the Child Grants Programs, public employment under the Integrated Watershed Management Program, Nutrition for Vulnerable Groups, and Agricultural Input Fairs. Consolidating and building on 143 Based on the rough assumptions underlying Table 21 at the end of Chapter 3 on the proportion of the very poor being reached by existing programs. On this basis, about M.300 million out of M.1.5 billion goes to the extreme poor. 144 Based on 2010/11 spending on transfer programs (Chapter 3) and 2010/11 total revenue of 8.9 billion (Chapter 1). 107 theses and unifying them into a system of two or three major programs that operate nationwide could have a much greater impact on reducing inequality and extreme poverty in Lesotho in the coming 10 years. Spending could also have a more permanent impact on poverty levels if the promotive aspect of programs was strengthened. 210. Promotive safety net programs contribute to human capital development and help the poor to participate more in the growth process and to increase their productivity. However, given Lesotho’s tight fiscal circumstances, policymakers will need to make trade - offs between different investments, so any promotive safety net programs will have to be cost-effective to justify their place in the national budget. For example, spending on school meals to improve educational outcomes would have to be more cost-effective than spending the same money on improving educational quality. Also, these interventions will have to demonstrate their effectiveness in achieving their objectives. For example, there would be no point in providing fertilizer vouchers or subsidies if the inputs were not generating the expected yield increases. Finally, these interventions should be used to tackle very high priority problems, such as early childhood malnutrition in Lesotho. 211. Transfers have a number of other potential benefits in the case of Lesotho : (i) they represent an injection of funds into the lowest level of the economy where they have immediate multiplier effects; (ii) they may be able to hasten the economic transition by kick-starting grassroots activity (for example, by providing households with the money to buy agricultural inputs or to start non-farm businesses); and (iii) they can allow families to break the inter-generational poverty cycle, especially associated with the effects of HIV/AIDS. So What Should Be The Primary Objective of a National Safety Net Strategy in Lesotho? 212. We would argue that the overarching objective of a safety net strategy should be to protect the consumption of all Basotho at about the food poverty line. This provides a 108 clear, unambiguous focus and limits the potential cost and scale of transfers. Given the competing claims, pure transfers should be reserved for those who would not otherwise be able to feed themselves. Depending on which estimates one believes, this ranges from 10 percent to 34 percent of the population. Aspects of Poverty that can be Addressed with Safety Net Transfers in Lesotho 213. Safety net spending must be in line with national priorities. While there is much general poverty in Lesotho (57 percent of the population in 2010 could be classified as “poor” using the national basic needs poverty line145), there is not the depth of absolute food deprivation that can be found in many other parts of Sub-Saharan Africa. As one humanitarian aid worker has said: “This is not a country where people are starving.” This is an argument in favor of taking a narrowly targeted approach to providing transfers in Lesotho. One way of doing this would be to establish a unified safety net program aimed at addressing some well-defined national priorities. Two of the most logical priorities would be (i) HIV/AIDS and the associated risk of inter-generational poverty and (ii) malnutrition. Framing interventions to address these specific issues would also increase their political and social acceptability. 214. National spending on transfers should play a counter-cyclical and seasonal role. They should be designed to be expanded when times are bad (for example, when mining remittances fell) and contracted when times are good (for example, as textiles employment expands) and in response to shocks (such as droughts) and in response to changing fiscal conditions. Finally, a national strategy should focus on addressing seasonal poverty during the four or five month pre-harvest lean season. 145 Preliminary government estimates based on the 2010/11 HIES. 109 2. The Way Forward – Towards a Unified National Program. 215. There are a number of cross-cutting recommendations that emerge from our analysis that could apply to any kind of national safety net program in Lesotho: - Increase the predictability of transfers. Experience worldwide has shown that, for transfers to have a permanent impact on a household’s poverty level, they have to be sustained and reliable. Households need to be certain that the sum of money that they are expecting will indeed be available every month as this enables them to plan ahead and take more risks and make more productive investments. Some existing programs are successfully providing predictable transfers (such as the OAP and the CGP), while others are less reliable (such as the AIF, the Nutrition Support Program, the OVC bursaries, and the Integrated Watershed Management Employment Program, due in part to variable annual budget allocations). - Improve targeting and coordination. Like most countries, Lesotho cannot afford to waste money by providing transfers to people who do not need them. One way of avoiding this is to strengthen targeting to ensure that benefits only reach those most in need. The recent efforts to target the CGP and to establish a national registry (NISSA) have the potential to improve targeting and coordination. - Enable programs to be scaled up and down in response to need. Programs should be designed to be expanded and contracted in response to shocks, cyclical trends (such as remittance flows and the annual lean agricultural season), and the government’s fiscal capacity to finance transfers. - Evaluate programs more rigorously. It is important to carry out rigorous evaluations of existing programs to enable policymakers to assess which programs represent the best value for money in reaching the poor. These evaluations should involve 110 measuring the poverty status of the beneficiaries and the impact of the programs on the consumption of the recipient households and should be conducted independently of the agency responsible for implementing the program. One option might be to commission the Bureau of Statistics to manage independent assessments of the coverage and impact of safety net programs, which could potentially be linked with a consumption module of the Continuous Welfare Monitoring Surveys. 146 - Increase the promotive aspect of programs. Wherever possible, safety net transfers should contribute to improving conditions of the poor over the long term. For example, conditional cash transfers require recipients to adhere to certain conditions that aim to build their human capital and productivity, such as school attendance, health care, or nutrition interventions. - Increased the productive aspect of programs. Spending on transfers can be made more productive, for example, by using public employment of the able-bodied poor to maintain village roads or expand reforestation or for water and soil management or by expanding agricultural productivity in other ways, perhaps in an expanded and refined AIF program. - Move towards a unified national program. It makes sense to move towards establishing a unified national program that may still consist of several distinct interventions but that will operate within an integrated approach to objectives and targeting. 146 The World Bank is currently providing technical assistance for the development of a social protection module to be included as a pilot in the CMS. 111 - Evaluate impact of programs and improve consumption data . Systematically and independently evaluate the impact of transfer programs, and improve the quality and frequency of household consumption data which is needed to analyze targeting efficiency. The Value of the Transfer and the Duration of Support 216. There is no way of saying what the right level of transfer should be. Obviously some households require more support than others. For example, child-headed households probably need a transfer that meets almost all of their consumption needs, whereas many other poor households may need only enough to raise them above the food poverty line or to tide them over the lean agricultural season. Similarly, a household consisting of a single elderly person generally needs a smaller transfer than a household containing six children. Table 25 illustrates some potential transfer amounts per adult equivalent. Table 25: Some Illustrative Transfer Amounts (2011 Maloti per adult equivalent per month) Estimated Monthly Transfer a/ Provide four-months (lean season) maize consumption M.32 (US$4.32) [M.128 per household] b/ 25% of current consumption M.50 (US$7) [M200 per household] c/ 50% of food Poverty Line M.93 (US$13) [M.372 per household] Source: Author’s calculations. Notes: a/ Based on 8 kg. of maize per month, and price of M.50 per 12.5 kg. bag. b/ Based on current estimated average rural consumption of approximately M.200 per adult equivalent per month (Chapter 2). c/ Based on 2010 food poverty line of M.184 per adult equivalent per month. 217. These preliminary calculations suggest a minimum transfer of about M.100 to 150 per month (about US$17) is likely to be appropriate for households suffering from only limited seasonal food deprivation (such as very low-income rural households who have some land and some able-bodied family members), whereas the absolutely destitute might need a transfer more in the order of M.350 to 400 (about US$50) per month per household. These amounts can be compared with the average value of the benefits currently provided 112 by the Child Grants Program (M.120 per household per month) and the Old Age Pension Scheme (M.350 per pensioner per month).147 218. Another key decision involves how long to make the transfer available for. Over time many households will graduate from safety net support, and others will only need support for short periods. However, some will require assistance for a prolonged period (for example, homeless orphans who will eventually grow up and join the workforce but need sustained support until then), and others will need it throughout their lives (for example, the destitute elderly, the mentally incompetent, and the severely disabled). Also, even as some households graduate from assistance, others will fall into poverty, which means that Lesotho needs to establish a permanent national safety net system that can assist those in need over the long term. Options and Costs of a National Program 219. Table 26 shows the possible costs of a national unified safety net program. It presents these costs for different levels of coverage and benefits, ranging from reaching the poorest 30 percent (the majority of those who are estimated to fall below the 2010 food poverty line) to supporting only the poorest 10 percent (roughly equivalent to the 200,000 people who are estimated to be chronically food-insecure). 220. The table gives three versions of this safety net. The first posits a low transfer of M.37.50 (US$5) per capita per month (equivalent to M.188 per month for a family of five). The second posits a medium-sized transfer of M.75 (US$10) per capita per month (equivalent to M.375 for a family of five), and the third consists of a high transfer that would provide M.150 (US$20) per capita monthly (equivalent to M.750 for a household of five). The medium-sized transfer level is probably the most realistic as the low version may 147 See examples of the generosity of selected programs in Africa in Annex 6. 113 not be enough to make a difference to poverty and consumption levels while the high version may not be fiscally affordable. Note that these costings assume that benefits would be provided for all 12 months of the year, whereas many rural agrarian households will only need transfers during the four-month lean season. Table 26: Possible Scale and Costs of a National Safety Net Beneficiary Group/Strategy No. of Beneficiaries Approximate Approximate Annual Cost Annual Cost (US$ millions) (million Maloti) High Transfer (M.150 (US$20) per capita per month) Poorest 30% 570,000 M.1,230 m. $165 m. Poorest 20% 380,000 M.821 m. $109 m. Poorest 10% 190,000 M.410 m. $55 m Medium-sized Transfer (M.75 (US$10) per capita per month) Poorest 30% 570,000 M.615 m. $81 m. Poorest 20% 380,000 M.411 m. $55 m. Poorest 10% 190,000 M.205 m. $27 m. Low Transfer (M.37.50 (US$5) per capita per month) Poorest 30% 570,000 M.309 m. $ 42m. Poorest 20% 380,000 M.206 m. $ 28m. Poorest 10% 190,000 M.103 m. $14m. Notes: Based very roughly on specified monthly transfer per capita for 12 months a year and assuming 20 percent overheads and delivery costs. Assumes a population of 1.9 million. 221. It should be stressed that achieving the coverage and costs shown in Table 26 will depend on good targeting and the effective implementation of the programs. Actual costs will vary quite a lot depending on the instruments that are used. For example, public works interventions will cost somewhat more to deliver, while pure unconditional cash transfers will cost less but will involve the greater complexities and costs involved in targeting beneficiaries. 222. The costs listed in Table 26 can also be reduced because: (i) some portions of the population will only need support during the agricultural lean season (usually four months of the year) and (ii) households receiving the Old Age Pension should not in general be 114 eligible for assistance under other components of the safety net scheme. These numbers are not known but would need to be further analyzed. Program Choices 223. In the current fiscal environment it does not appear either sensible or feasible to advocate for a large-scale expansion of cash or in-kind transfers. However, there is scope for improving the performance of existing programs, as well as for rationalizing spending in support of a national package of interventions as proposed below. 224. Universal programs such as the OAP and the School Feeding Program do not appear to be the most cost-effective way of addressing extreme poverty. On a purely arithmetical basis, they are clearly providing at least half of their benefits to people who do not need them.148 Given that they already exist and are backed by strong popular and political support, it seems likely that they will remain an integral part of the social transfer system in Lesotho, but any efforts to expand them (for example, by lowering the pension age or enhancing the school meals package) should be resisted. Also, over time policymakers should consider ways to improve the targeting of school feeding. Any new transfers: (i) should be introduced as part of a unified national framework and should contribute to the overall objective of the country’s safety net; (ii) should be well coordinated with other transfer programs; and (iii) should be targeted to those most in need. 225. Based on the analysis in this study, it is recommended that a national package of interventions might consist of: 148 Based very roughly on approximately half of the population living below the poverty line (Chapter 2). 115  A single cash grants program, focused primarily on categorical groups of the extreme poor such as child-headed households, grandparent-headed households with no working aged adults, and the infirm and disabled who are not living with relatives. This program could include conditions requiring recipients to invest in their human capital development and/or be combined with an intervention designed to reduce early childhood malnutrition among beneficiaries. The existing Child Grants Program could be a basis on which to build this proposed cash transfer scheme for the destitute, though its targeting would need to be refined and its donor dependency would have to be reduced.  Some form of self-targeted public works program for the able-bodied poor. Lesotho is unusual in not having any large-scale cash-for-work or food-for-work program. However, there is potential to scale up the existing Integrated Watershed Management Public Works Program and turn it into a public works program employing the very poor on re-forestation, environmental works, and rural road maintenance. It would make sense to expand only gradually in order to learn lessons about what works in the context of Lesotho. Existing spending on road maintenance and public works could be used to employ the rural poor in the agricultural slack season.  A more focused and effective version of the Agricultural Input Fares vouchers scheme for the rural agrarian poor, though this will depend on a more rigorous assessment of the costs and benefits of the existing program and the extent to which it has been successful in raising the consumption of the poor. 226. There needs to be a national debate on what the coverage of these interventions should be, and it may make sense to start piloting a public works program to learn lessons for a potential expansion. Based on the analysis in this report, we would recommend that the initial target group for a national safety net system should be those with insufficient 116 food consumption– perhaps 200,000 to 400,000 people (or 40,000 to 80,000 households).149 The cash grants program would be targeted primarily to households with no able-bodied members and complemented by the public works and agricultural input programs, which would cover those households with able-bodied members. Financing 227. The amounts involved in financing such a program are likely to be affordable. They might amount to about M.411 million a year (US$55 million), for the medium-sized version in Table 26 (providing a transfer of about M.75 per capita per month to the poorest 20 percent of the population), representing about 4 percent of total public expenditure or 2.5 percent of GDP.150 To put this in perspective, total spending on transfers in Lesotho in 2010/11 amounted to about M.1.5 billion (US$197 million) annually (16 percent of public expenditure or 9 percent of GDP). 228. Much of this current spending is accounted for by spending on two large universal programs ‒ the Old Age Pension and the School Feeding Program ‒ and by the high -cost tertiary bursary scheme. Some funds can be freed by reducing the costs of school feeding, and the costs of the two untargeted programs that mostly benefit the non-poor ‒ the National Fertilizer and Input Subsidy and the Tertiary Bursary Scheme. How much can be saved in this way will depend on any political decisions made to realign program spending, but at least half of the money needed to fund a national safety net program could easily be found from these sources.151 In addition, some money that the government is already spending could be used to fund transfers to the poor, for example, the road maintenance 149 Numbers will need to be refined once the government has completed an in-depth poverty analysis of the 2010/11 HIES data. 150 Based on public spending of M9.38 billion and GDP of M16.3 billion in 2010/11. 151 Based on spending of about US$114 million equivalent annually on the School Feeding Program, the National Fertilizer and Input Subsidy, and the Tertiary Bursary Scheme. Just 20 percent savings on these programs could provide US$23 million equivalent, or half the amount needed for the medium-sized version of the national safety net scheme shown in Table 26 (US$55 million). 117 budget and environmental and reforestation activities could routinely employ the poor at a low wage rate. 229. Finally, donors are spending a significant amount on supporting various safety net programs. These include the School Feeding Program, the Child Grants Program, the Nutrition Support Program, and Agricultural Input Fairs, as well as numerous programs to support OVCs and people with HIV/AIDS. Some of this money will be fungible, and some may not, but it is worth trying to channel this support over time into funding a unified national safety net program. Among other things, this will avoid the problem of stop-start stand-alone programs that are totally reliant on donor funding. Over time the government should be able to negotiate with individual donors as they move into new program cycles to persuade them to direct their financing towards the unified national program (which may still include some existing activities, such as the AIFs and the Nutrition Support Program). This has been successfully done in Ethiopia this where donors now provide pooled support for a single, permanent transfer scheme. Institutional Recommendations 230. The following are a number of institutional recommendations that we made in Chapter 4, which would apply no matter which model is adopted: - Establish a national safety net strategy for the next 5 to 10 years that spells out the country’s broad poverty reduction objectives, desired target groups and coverage, and program choices. The government’s move to develop a National Social Protection Strategy over the coming year provides a welcome opportunity to do this. - Eliminate any duplication of coverage, for example between the OAP, the CGP, and PA, and create an integrated beneficiary register. 118 - Establish a central body with policy oversight and expenditure planning authority over all transfer programs. - Continue gradually expanding NISSA once a clear vision for the system has been developed and careful consideration has been given to how the expansion would/should take place. The system should also be continuously monitored and evaluated to determine if it is in fact an effective way of increasing coordination and improving targeting. - Establish the capacity to systematically and independently evaluate the impact of transfer programs, to enable policy-makers to know which programs are effective at reducing poverty, and represent good value for money. One way of doing this could be to give responsibility for measuring program impact to an independent agency (such as for example the Bureau of Statistics). It will also require strengthening the quality and frequency of collecting household consumption data, which is needed to analyze targeting efficiency. Adding a module to the next household consumption survey that tracks which programs households are receiving benefits from is a good way of doing this. 3. Short- to Medium-term Recommendations for Improving Existing Programs 231. Finally, Table 27 summarizes some specific recommendations for improving how existing programs operate and for putting Lesotho on the path toward establishing a unified national program. Although some of these programs were not designed explicitly to target the poorest, the table highlights ways in which they could increase their operational efficiency and/or more effectively reach poor people. 119 Table 27: Key Observations and Recommendations for Each Program Program Key Observations Recommendations (Short Term= 1-2 years and Medium Term= 3-5 years) Safety Net Programs Child Grants The CGP has well-defined operational features as well as the Short Term: (a) Rigorously assess the effectiveness of current targeting Program necessary operational systems in place to implement the and the impact of the program; (b) hold policy-level debate on program. It is also one of the few programs that make a intended national scale and willingness to finance long term; and (c) systematic effort to identify the most vulnerable. The consider alternative criteria and methodologies for targeting, based on effectiveness of its targeting has not yet been evaluated evidence from evaluation. rigorously, and it is heavily dependent on external support. Medium Term: (a) Expand to a general program to reach the most destitute and (b) consider introducing co-responsibilities for the receipt of the cash transfers to encourage human capital development; particularly related to malnutrition. Public Assistance Coverage is too small to affect poverty nationally, and Short Term: (a) Consider integrating PA with an expanded nationwide payments are too sporadic to affect poverty at the cash grants scheme based on the CGP model (above). household level. Old Age Pension The OAP benefits many deserving poor people and there is Short Term: (a) Do not lower pension age at this stage; (b) exclude some evidence that the consumption and education of people who receive pensions from the private sector that are higher recipients has increased as a result of the pension. than the OAP; and (c) continue to explore ways to make payments However, the program also provides benefits to many non- electronically. poor elderly. Medium Term: Analyze the sustainability of the program in the long run. School Feeding While successful in reaching all students nationwide, the Short Term: (a) Continue to rationalize the program into a unified Program SFP is an expensive way of making transfers, and many of home-grown feeding program but make every effort to reduce costs – the benefits go to households that do not need them. possibly by aiming to provide a more modest meal; (b) increase operational efficiency; and (c) strengthen implementation capacity Medium Term: Eliminate free meals in better-off schools. Nutrition Support This program could be used to link the safety net with Short Term: (a) Continue implementation of the program but with Program malnutrition. Given its early stage of implementation, it is close monitoring and evaluation and assess its costs and effectiveness. too early to assess its effectiveness or impact. Medium Term: (a) Consider introducing an expanded nutrition program; (b) review experience with any other previous programs and successful international experience; and (c) determine what scope there is for expanding it as part of national safety net in addition to the unified cash transfer and public works programs. 120 OVC Bursary The program is an important effort to help deserving Short Term: (a) Ensure predictability of support; (b) consider Scheme children to go to secondary school, but does not seem to be alternative targeting methods to improve targeting of worse-off particularly pro-poor and the amount provided seems students; and (c) develop clear targeting and operational guidelines. inadequate and unreliable. Medium Term: (a) Consider giving additional support to students from worse-off families and (b) explore better links with other programs or potentially merging targeting with a national cash transfer scheme. Agricultural Input The program has the benefit that it targets smaller and Short Term: (a) Rigorously evaluate the program’s impact on Fairs more vulnerable farmers, but there is limited evidence on consumption before deciding whether to continue it. who it is reaching or on the impact it is having on the Medium Term: (a) Subject to the evaluation either make the program consumption of beneficiaries. Its funding is unpredictable part of the overall safety net or discontinue it when donor funding because of its donor dependence but provides a possible ceases and (b) ensure that any expanded program is more integrated basis for expansion into a wider-coverage program of into the routine operations and financing of the government. agricultural input support for the poorest among rural farmers. Other Transfer Programs Integrated The program shows some promise as a works-based Short Term: (a) Expand spending on low-level road maintenance; (b) Watershed transfer scheme that also improves the environment. It maximize labor-intensity & employment of local women and men; (c) Management Public could employ more people if expanded beyond critically evaluate experience with the forestry program and assess Works Program reforestation to other land management and infrastructure scope for expanding works-based transfers; and (d) carefully pilot and maintenance works and if it were more explicitly targeted monitor the WFP cash-for-work/food-for-work program to the poor through a combination of a low wage rate and Medium Term: Expand labor-intensive works-based transfers based on geographical targeting. Continuing annual employment for experience and evidence from ongoing and previous programs and of the same poor villagers is needed rather than one-off public employment for the very poor in other countries. projects. National Fertilizer The program is untargeted, there is no evidence as to its Short Term: (a) Consider more intense targeting in poorer areas; (b) and Input Subsidy output impact, and it seems unlikely that it reaches the reduce the scope of the National Fertilizer and Input Subsidy and retain poor. only for poor small-scale farmers; and (c) consider ways to prevent recipients from re-selling the inputs at the commercial price. Medium Term: Consider eliminating the program entirely but strengthen the targeted subsidy under the Agricultural Input Fairs instead. Tertiary Bursary The program as it is currently structured suffers from a Short Term: Consider how to restructure the program over time. Scheme number of fundamental flaws. It is costly, it benefits mostly Options include better targeting (possibly some form of means testing) households who do not need a public transfer, and it needs and shifting to a more commercially oriented program that is a more to be more financially sustainable. genuine loan scheme. Medium Term: Restructure the program. 121 References Alderman H. and D. Bundy (2011). “School Feeding Programs and Development: Are We Framing the Question Correctly?” World Bank Research Observer, July 26, 2011 Africon Consulting (2010). “Review and Update of Study of Projected Road Maintenance Needs.” Prepared for the Road Fund Secretariat, Maseru. Africon Consulting, Maseru. March. Ayala Consulting (2012). “NISSA/CGP Statistics.” Prepared for the Government of Lesotho/UNICEF. Ayala Consulting, Quito, Ecuador, January. Ayala Consulting (2011a). “PMT Formula Development Paper.” Ayala Consulting, Quito, Ecuador, February. Ayala Consulting (2011b). “Economic and Operational Assessment of Social Assistance Programs and Sector.” Prepared for UNICEF by Ayala Consulting, Quito, Ecuador, April. Ayala Consulting (2011c). “Strengthening and Expansion of CGP pilot and Coordination of Social Assistance: Progress Report July-November 2010 and Way Forward.” Ayala Consulting, Quito, Ecuador, November. Ayala Consulting (2010). “Report on the Feasibility of a Single Targeting System and National Information System for Social Assistance.” Prepared for the Department of Social Welfare and UNICEF. Ayala Consulting, Quito, Ecuador, June. Bank of Lesotho (2011). “Bank of Lesotho 2010 Annual Report.” Maseru, Lesotho. Barros, R.P., M. de Carvalho, S. Franco, and R. Mendoça (2010). “Markets, the State, and the Dynamics of Inequality.” New York, UNDP, January. Bastagli, F. (2011) “Conditional Cash Transfers as a Tool of Social Policy.” Economic and Political Weekly, Vol. XLVI, No 21. Blank, L. (2008). “Revised Child Grant Pilot Operations Manual.” Ministry of Health and Social Welfare, Maseru, Lesotho, November. BoS (2012a). “Consumer Price Index, 2011 Statistical Report.” Bureau of Statistics, Maseru, Lesotho. BoS (2012b). “Lesotho Agricultural Census 2009/10, Vol. 1: Rural Households and Crop Statistics.” Bureau of Statistics, Maseru, Lesotho. 122 BoS (2011a). “Consumer Price Index, 2010 Statistical Report.” Bureau of Statistics, Maseru, Lesotho. BoS (2011b). “Registration of Orphans and Vulnerable Children.” Bureau of Statistics, Maseru, Lesotho. July. BoS (2011c). “Continuous Multi-Purpose Household Survey: 2011.” Bureau of Statistics, Maseru, Lesotho. BoS (2010a). “Continuous Multi-Purpose Household Survey: Fourth Quarter Report October, 2010.” Bureau of Statistics, Maseru, Lesotho. BoS (2010b). “Living Conditions in Lesotho, 2009. Results from the First Round of the Continuous Multipurpose Survey (CMS)”” Bureau of Statistics, Maseru, Lesotho. BoS (2010c). “Statistical Yearbook, 2010.” Bureau of Statistics, Maseru, Lesotho. BoS (2009a). “2006 Lesotho Population and Housing Census; Analytical Report; Vol. IIIB Socioeconomic Characteristics.” Bureau of Statistics, Maseru, Lesotho, December. BoS (2009b). “Population by Age and Sex; 2006 Lesotho Population and Housing Census. ” Bureau of Statistics, Maseru, Lesotho, October. BoS (2008). “2008 Integrated Labor Force Survey Report.” Bureau of Statistics, Maseru, Lesotho. BoS (2006). “2002/03 and 1994/95 Household Budget Survey: Analytical Report.” Bureau of Statistics, Maseru, Lesotho. November. Bundy,D.; Burbano,C., Grosh,M., Gelli,A., Jukes,M., and Drake,L. (2009) “Rethinking School Feeding, Social Safety Nets, Child Development and the Education Sector.” World Bank and The World Food Program 2009; Washington D.C. Croome, D., M. Molisana, and A. Nyanguru (2007). “Impact of the Old Age Pension on Hunger and Vulnerability: A Case Study from the Mountain Zone of Lesotho.” Institute for Southern African Studies, University of Lesotho, prepared for the Regional Hunger and Vulnerability Project (RHVP) August, 2007. DSW (2011). “Situation Analysis of Orphans and Other Vulnerable Children in Lesotho.” Draft. Department of Social Welfare/ UNICEF, Maseru, Lesotho. September. DSW (2005). “National OVC Strategic Plan 2006-2011.” Department of Social Welfare, Maseru, Lesotho. 123 DMA (2011a). “Lesotho Food Security and Vulnerability Monitoring Report.” Disaster Management Authority, Government of Lesotho, Maseru. June. DMA (2011b). “Early Warning Bulletin.” Disaster Management Authority, Government of Lesotho, Maseru. December. FAO (2010). “Agricultural Input and Trade Fairs: Post-Harvest Report, July 2010.” Emergency and Rehabilitation Division, Food and Agriculture Organization, Rome. FAO (2008). “Lesotho – Initiative on Soaring Food Prices (ISFP): Situation Analysis and Country Action Plan.” Food and Agriculture Organization, Rome. FNCO (2011a). “A Study of Dietary Patterns, Energy Intakes, and Micronutrient Adequacy among Children under 5 and their Caretakers in Thaba-Tseka district, Lesotho.” Powerpoint presentation of results, Food and Nutrition Coordination Office, Government of Lesotho, Maseru, and World Food Programme, Rome. FNCO (2011b). “Nutrition Baseline Survey in Four Districts.” Food and Nutrition Coordination Office, Government of Lesotho, Maseru, and World Food Programme, Rome. January. FNCO (2007). “Lesotho Nutrition Survey November-December 2007: Factsheet.” Food and Nutrition Coordination Office, Government of Lesotho, Maseru. GFCU (2010). “Annual Report for the Global Fund Support to Lesotho 2009.” Global Fund Coordinating Unit, Ministry of Finance and Development Planning, Maseru, Lesotho. GFCU (2009). “Global Fund Round 2 Grant Closure Report.” Global Fund Coordinating Unit, Ministry of Finance and Development Planning, Maseru, Lesotho, January. Government of Lesotho (n/d). “Interim National Development Framework; 2009/10- 2010/11.” Maseru, Lesotho. Government of Lesotho (2011). “Post Disaster Needs Assessment - Heavy Rains 2010/11.” Government of Lesotho/World Bank/EU/UN. July. Grosh, M., C. del Ninno, E. Tesluic, and A. Ouerghi (2008). For Protection and Promotion: The Design and Implementation of Effective Safety Nets. World Bank, Washington D.C. Haag, P., F. de Meulder, and D. Khama (2009). “Mid-term Evaluation of WFP Lesotho Development Project 10582: 2008-2010 Support Access to Primary Education Evaluation Report.” Report to the World Food Programme, Rome. November. 124 IMF (2013); Lesotho Fifth Review Under Extended Credit Facility; Washington; April. Kardan, A,., I. MacAusian, F. Merttens, and Pellerano, L (2011). “A Rapid Assessment of the Lesotho Child Grants Program.” Oxford Policy Management, Oxford, UK. April. Lawson, D. (2008).“Infrastructure and Time Poverty in Lesotho.” South African Journal of Economics. March. Leduka, R. and J. Gay (2010). “Lesotho Country Social Assessment.” Background paper for World Bank Poverty Assessment for 2010. Sechaba Consultants, Maseru, Lesotho. Leboela, T., and S.D. Turner (2003). “The Voice of the People: Report on Community Consultations for the National Vision and the Poverty Reduction Strategy Paper.” Ministry of Development Planning and United Nations Development Programme, Maseru. LVAC/DMA (2010). “Lesotho Food Security and Vulnerability Monitoring Report.” Lesotho Vulnerability Assessment Committee (LVAC) and Disaster Management Authority (DMA), Maseru, Lesotho. LVAC/DMA (2011). “Lesotho Food Security and Vulnerability Monitoring Report.” Lesotho Vulnerability Assessment Committee (LVAC) and Disaster Management Authority (DMA), Maseru, Lesotho. Management Sciences for Health (2011). “Building Local Capacity for Delivery of HIV Services in Southern Africa Project: Annual Lesotho Work Plan October 2011-September 2012.” Maseru. Lesotho. MAFS (2011). “Lesotho Agricultural Situation Report 2010/11.” Department of Planning and Policy Analysis, Ministry of Agriculture and Food Security and Bureau of Statistics, Maseru, Lesotho. MAFS (2010). “Lesotho Agricultural Situation Report 2008/09.” Ministry of Agriculture and Food Security and Bureau of Statistics, Maseru, Lesotho. Marsland, N., M. Metz, M. Phoofolo and M. Sejanamane (2004). Food and Nutrition Insecurity in Lesotho: Problems, Trends & Responses. Ministry of Education (2011). “Education Statistics Bulletin 2010.” Ministry of Education, Planning Unit, Maseru, Lesotho. MoFDP (2012). “Draft National Strategic Development Plan.” Ministry of Finance and Development Planning, Maseru, Lesotho. 125 MoFDP (2011a). “2011/12 Budget – Detailed Estimates.” Ministry of Finance and Development Planning , Maseru, Lesotho. MoFDP (2011b). “Background Paper to 20011/12 Budget.” Ministry of Finance and Development Planning, Maseru, Lesotho, February. MoFDP (2011c). “Budget Overview Document 2011.” Ministry of Finance and Development Planning, Maseru, Lesotho. MoFDP (2009). “Background Paper to 2009-10 Budget.” Ministry of Finance and Development Planning, Maseru, Lesotho. MoFDP(2005), Restructuring of the National Manpower Development, Secretariat and Revolving Fund, commissioned by the Ministry of Finance and Development Planning; Maseru, 2005. MoFDP (n/d). “Social Protection in Lesotho.” Department of Development Planning, Ministry of Finance and Development Planning , Maseru, Lesotho. MHSW (2011). “Situation Analysis of Orphans and Vulnerable Children in Lesotho.’ Department of Social Welfare, Ministry of Health and Social Welfare. November. MHSW (2010). “Lesotho Demographic and Health Survey 2009.” Ministry of Health and Social Welfare and ICF Macro. September. MHSW (2006). “Lesotho Demographic and Health Survey 2004.” Ministry of Health and Social Welfare, Maseru, Lesotho. Motseng Logistics Services (2011). “The Development of a Framework for the Implementation of an Outsourced, Revised, and Sustainable School Feeding Program for the Ministry of Education in Lesotho.” Produced for the Ministry of Education. September. Mokitimi, N. (2011). “Review of Agricultural Input Trade Fairs in Lesotho.” January. Mwansa, A. (2010). “Assessment of the Bursary Scheme for Orphans and Vulnerable Children (OVCs).” Study for the Ministry of Education and UNICEF. February. ODI (2002). “Lesotho Food Security Issues Paper.” Prepared for the Forum on Food Security in Southern Africa, Overseas Development Institute, London. Olivier, M. (2010). “Social Security in Lesotho: A Situation Analysis Report.” International Institute for Social Law and Policy. Prepared for the ILO. October. 126 Parliament of Lesotho (2013); Budget Speech to Parliament 2013/14 Fiscal Year; Maseru. February. Pellerano, L., A. Hurrel, V. Barca, F. Hove, R. Beazely, B. Modise, S. Dodd, I. MacAusian, L. Crawfurd, and A. Kardan (2012). ”CGP Impact Evaluation: Targeting and Baseline Evaluation Report.” Oxford Policy Management, Oxford, UK, for the Government of Lesotho, April. Southern Africa Migration Program (2010). “Migration, Remittances, and Development in Lesotho.” Migration Policy Series No.52, Idasa. Cape Town. Thematic Working Group on Social Protection (2011). “Social Protection Issues Paper: National Strategic Development Plan 2012/13-2016/17.” Thematic Working Group on Social Protection (consisting of government of Lesotho and donors). Maseru. May. UNAIDS (2010). “Report on the Global AIDS Epidemic.” (http://www.unaids.org/globalreport/documents/20101123_GlobalReport_full_en.pdf) UNDP (2010). “Human Development Report 2010: The Real Wealth of Nations.” Pathways to Human Development, United Nations Development Program, New York http://hdr.undp.org/en/media/HDR_2010_EN_Complete_reprint.pdf UNDP (2007). “Lesotho, National Human Development Report.” United Nations Development Program, Maseru, Lesotho. UNESCO (2012). “Cash Transfers and HIV Prevention.” United Nations Educational, Scientific, and Cultural Association, Paris, February. United Nations (2010). “Lesotho Integrated Action Against Malnutrition – A Joint UN Nutrition Programme.” Maseru. May. United Nations (n/d). “Lesotho: Achieving the MDGs: Bottlenecks, Challenges, and Constraints.” Van Domelen, J. and S. Coll-Black. “Designing and Implementing a Rural Safety Net in a Low- income Setting: Lessons Learned from Ethiopia’s Productive Safety Net Program 2005- 2009.” (http://siteresources.worldbank.org/SAFETYNETSANDTRANSFERS/Resources/EthiopiaPSNPL essonsLearnedLite.pdf) Vargas Silvana (2011) Improving the design and Implementation of “Juntos” Program 127 WFP (2010). “Nutrition Support for Malnourished Children and Vulnerable Groups: Development Project Lesotho 200169.” WFP Board Project document, World Food Programme, Rome, November. WFP (2008). “Standard Project Report 2008 Lesotho - Support Access to Primary Education. Annual Report on the School Feeding Program for 2008.” World Food Programme, Lesotho. WFP (undated). “Lesotho Development Project 200199 - School Meals Program (Project document for 2011-2012).” World Food Programme, Rome. WFP (undated). “Social Protection and Food Assistance for Vulnerable Groups in Lesotho (WFP PRRO 10599 2008-2010).” World Food Programme, Rome. WFP (undated). “Food Assistance for Vulnerable Groups Lesotho 2008-2010 PRRO 10599; Program Document.” World Food Programme, Rome. World Bank (ongoing). “Social Safety Nets in Africa – A Regional Synthesis.” World Bank, Washington D.C, draft. World Bank (2013a). “Using Social Safety Nets and Transfers to Accelerate Poverty Reduction.” World Bank, Washington D.C, January. World Bank (2013b); Program Document: Lesotho First Growth and Competitiveness Development Policy Grant; Washington; April, 2013 World Bank and Government of Swaziland (2012). “Swaziland: Using Public Transfers to Reduce Extreme Poverty.” World Bank and Government of Swaziland, Washington D.C. November. World Bank (2011a). “Lesotho: Public Expenditure Management Efficiency Review – Making the Most of Shrinking Fiscal Space.” Draft. World Bank, Washington D.C. October. World Bank (2011b). “Technical Note on Higher Education Loan Bursary Scheme.” Draft. World Bank, Washington D.C. December. World Bank (2011c). “Smallholder Agriculture Development Project: Project Appraisal Document.” World Bank, Washington D.C. October. World Bank (2011d). “Lesotho Economic Update July, 2011.” World Bank, Washington D.C. World Bank (2010a). “Review of Financing Tertiary Education in Lesotho.” Draft.World Bank, Washington D.C. August. 128 World Bank (2010b). “Lesotho - Sharing Growth by Reducing Inequality and Vulnerability.” World Bank, Washington D.C. June. World Bank (2010c). “Second Poverty Reduction Credit: Program Document.” World Bank, Washington D.C. March. World Bank (2010d). “Third Poverty Reduction Credit; Program Document.” World Bank, Washington D.C. March. World Bank (2009) “Emergency Project Paper under the Global Food Crisis Response Program for a Rapid Response Child-Focused Social Cash Transfers and Nutrition Project”, World Bank Washington D.C. 129 ANNEXES Annex 1: Geographical Distribution of Poverty and Assets of the Poor Annex Table 1.1: Some Indicators of Geographic Poverty Poor Food Adequacy Est. Facing food deficit Children Stunted (2007 Nutrition (2011 LVAC) (2009 DHS) survey) % Number % Number % Number Berea 10.9 27,250 12.1% 30,173 35.3 9,168 ButhaButhe 2.4 2,648 15.5% 17,050 32.5 3,862 Leribe 2.9 8,509 13.6% 39,932 39.6 12,203 Mafeteng 5.4 10,400 36.4% 70,126 37.6 7,442 Maseru 5.2 22,464 33.9% 146,257 31.7 13,664 Mohale’sHoek 6 10,620 37.2% 65,891 44.9 8,797 Mokhotlong 4.5 4,410 30.7% 30,043 48.2 6,458 Qacha’s Nek 7.9 5,510 33.0% 23,033 47 3,723 ThabaTseka 20.2 26,240 32.9% 42,786 51.8 8,690 Compiled by author from FNCO (2007), LVAC/DMA (2011), and MHSW (2010). Annex Table 1.2: Poverty Incidence and Depth by District District Poverty Incidence Depth Berea 61 33 ButhaButhe 68 36 Leribe 63 33 Mafeteng 58 28 Maseru 54 27 Mohale’sHoek 63 34 Mokhotlong 50 24 Qacha’s Nek 57 32 ThabaTseka 40 20 Source: 2002/03 HIES in BoS (2006). 130 Annex Table 1.3: Some Basic Assets of the Poor and Non-Poor (2002/03) (Percentage owning asset) Asset Very Poor Poor Non-Poor Fields/Land 53.7 49.1 35.2 Cattle 35.1 33.6 25.7 Sheep 12.5 15.9 12.0 Goats 16.8 16.5 10.8 Cart 9.7 9.8 6.7 Sewing Machine 9.3 13.8 16.0 HIES data as reported in National Human Development Report (UNDP (2007) Table 3.5.1, p35. Annex Table 1.4: Percentage of Households That Own Durable Goods – 2002/03 and 2010/11 Variable names 2002–03 2010–11 Electric or gas stove 0.305 0.560 Microwave 0.014 0.059 Microwave combined with stove 0.004 0.019 Generator 0.017 0.082 Solar system 0.035 0.141 Washing machine 0.006 0.007 Dryer, drying cabinet, or pressing machine 0.011 0.012 Irons 0.791 0.607 Dishwasher 0.016 0.049 Fan 0.016 0.014 Air conditioner, humidifier, ventilator, or extractor hose 0.004 0.002 Vacuum cleaner or other cleaning machine 0.009 0.005 Sewing and knitting machine 0.113 0.076 Television 0.143 0.287 Video player 0.037 0.132 Satellite dish 0.034 0.154 Radio, any type, or any other audio equipment 0.541 0.564 Camera 0.042 0.014 Video camera 0.003 0.003 Other photographic equipment 0.001 0.000 Personal computer or other information processing equipment 0.008 0.028 Cellular phone 0.128 0.712 Landline 0.057 0.056 Motorcycle or scooter 0.005 0.003 Bicycle 0.027 0.021 Source: Preliminary government estimates based on the 2010/11 HIES. Note: Percentages are calculated for all available households in each year. 131 Annex Table 1.5: Per Adult Equivalent Scale in 2010/11 HIES Gender Age Kcal per day Adult equivalent Both 0–6 (months) 650 0.224 6–12 (months) 850 0.293 1–3 1300 0.448 4–6 1800 0.621 7–10 2000 0.69 Male 11–14 2500 0.862 15–18 3000 1.034 19–24 2900 1 25–50 2900 1 51+ 2300 0.793 Female 11–14 2200 0.759 15–18 2200 0.759 19–24 2200 0.759 25–50 2200 0.759 51+ 1900 0.655 Source: Lesotho Bureau of Statistics. Annex Table 1.6: Per Adult Equivalent Scale in 2002/03 HIES Gender Age Adult equivalent Both 0–12 (months) 0.224 1–3 0.448 4–6 0.621 7–10 0.69 Male 11–12 1 13–18 1.04 19–22 1.07 23–50 1 51–75 0.89 76+ 0.76 Female 11–12 0.81 13–18 0.78 19–22 0.78 23–50 0.74 51–75 0.67 76+ 0.59 132 Annex 2: Health and Education Indicators Annex Figure 2.1: Life Expectancy at Birth (years) 70 60 50 40 30 20 10 0 1960 1962 1964 1966 1968 1970 1972 1974 1976 1978 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 2006 2008 2010 Annex Table 2.1: Educational Attendance Rates by Wealth, Location, and Gender in 2009 Primary net attendance rate Secondary net attendance rate Boys Girls Boys Girls Wealth quintiles st 1 87 94 7 15 nd 2 90 97 15 27 rd 3 94 97 25 38 th 4 95 98 37 52 th 5 98 97 60 62 Location Urban 97 97 59 56 Rural 91 96 21 35 Source: 2009 DHS in MHSW (2010) Table 2.6, p. 17. 133 Annex Table 2.2: Dropout and Repetition Rations by Quintile (%) Dropout rate Repetition rate Wealth quintiles Boys Girls Boys and girls Boys Girls Boys and girls st 1 5.1 3.4 4.2 17.5 12.2 14.6 nd 2 4.6 4.6 4.6 18.5 15.1 16.8 rd 3 4.8 5.0 4.9 17.2 13.2 15.3 th 4 2.8 2.6 2.7 15.7 10.4 13.3 th 5 0.8 0.6 0.7 10.7 10.6 10.6 Source: Source: World Bank (2010b), Table 6.13, p. 146. Annex Table 2.3: Benefit Incidence of Educational Expenditure, 2004-05 (%) Share of enrollment Overall share of Primary Secondary Tertiary education budget Wealth quintile st 1 22 5 1 9 nd 2 21 10 2 11 rd 3 20 20 12 16 th 4 19 25 19 20 th 5 18 40 66 42 Gender of students Male 50 44 43 45 Female 50 56 57 53 Average unit cost 884 2088 37459 % of educational budget 37 19 42 Source: World Bank (2010b) Table 6.15, p. 147. 134 Annex 3: Beneficiaries and Expenditure of Public Assistance Annex Table 3.1: Beneficiaries of the Public Assistance Program District OVC Destitute Disabled Total Adults Maseru 945 1,149 5 2,099 Thaba-Tseka 388 332 30 750 Mafeteng 184 288 126 598 Mohale`s Hoek 463 479 0 942 Quthing 184 401 0 585 Qacha`s Nek 800 300 0 1,100 Berea 412 408 24 844 Leribe 458 547 31 1,036 Butha-Buthe 366 464 0 830 Mokhotlong 601 199 0 800 Total 4,801 4,567 216 9,584 % of Total 50% 47.7% 2.3% Source: Ayala Consulting (2011c). Annex Table 3.2: Approved Expenditure of Public Assistance (in 2010) District Total Approved Expenditure for Approved Expenditure for (no of PAs) Cash In-kind Support Maseru 2,099 2,477,600 Thaba-Tseka 750 910,000 20,000 Mafeteng 598 989,000 33,200 Mohale`sHoek 942 1,140,000 Quthing 585 1,440,000 170,500 Qacha`s Nek 1,100 960,000 70,000 Berea 844 1,080,000 Leribe 1,036 1,440,000 220,000 Butha-Buthe 830 1,080,000 Mokhotlong 800 1,140,000 Total 9,584 12,656,600 513,700 % of Total 96% 4% Source: Ayala Consulting (2011c). 135 Annex 3.3 – Benefit Levels – Nutrition/Supplemental Feeding Program; 2012 No.of Ration Size per Person per Month (kg.) Months Supplementary Feeding Sugar Veg.Oil CSB Fortified Pulses Maize Meal - Children<5 12 0.45 0.6 6.0 -- -- - PLWH 12 0.45 0.6 6.0 -- -- ECCD Centres 12 0.10 0.015 0.06 0.12 0.025 HIV/AIDS clients -- -- 7.5 -- -- Take-home Family 12 -- 0.6 -- 6.0 1.8 rations – HIV Clients Blanket Feeding - Children < 2 3 0.45 0.6 6.0 -- -- - PLW 3 0.45 0.6 6.0 -- -- ECCD School Meal -- .0001 -- .00015 .00003 Note: CSB: Soya Blend; Source: WFP Lesotho. 136 Annex 4: Distribution of Fertilizer Subsidy Annex Table 4.1: Geographical Distribution of Fertilizer Subsidy District % of Subsidized Approximate Value Fertilizer) of Subsidy (M) Berea 13,709 (32%) 1.9 million ButhaButhe 7,156 (17%) 1 million Leribe 10,115 (24%) 1.4 million Mafeteng 5,733(13%) 803,000 Maseru 5,994(14%) 839,000 Mohale’sHoek 201(0.4%) 28,000 Mokhotlong 49(0.1%) 7,000 Qacha’s Nek 0 -- 0 Qutthing 250 (0.6%) 35,000 ThabaTseka 37 (0.1%) 5,000 Source: Author’s calculations using data provide by the Ministry of Agriculture and Food Security 137 Annex 5: School Meal Composition Annex Table 5.1: Food Ration per Day under Government-funded Program Day Carbohydrates Vegetable Protein Quantities per Child Monday Papa Moroho 150 g papa, 100 g vegetable Tuesday Bread Beans ¼ loaf, 250 ml soup Wednesday Papa Moroho Egg 150 g papa, 100 g vegetables, 1 egg Thursday Samp Beans 150 g samp and beans Friday Papa Milk 150 papa, 250 ml milk Source: Motseng Logistics Services (2011). 232. The WFP funds school feeding in the most remote areas of Lesotho by providing dry rations of food to be cooked on the school premises at an average cost of M.2.75 (US$0.39) per child per day. This results in an annual cost of M.495 (US$71). Under the WPF-funded part of the program, school children are provided with a mid-morning snack of maize meal and with a lunch of maize meal, pulses, and vegetable oil for a total of 180 days per year. These calculations are based on half-day school attendance in accordance with the majority of the school schedule. The government provides meal preparation support of M.1.50 (US$0.21) per day per child covering the cost of fuel, salt, sugar, and the salaries of the cooks. The government also supports the cultivation of school gardens to increase the variety of food in school meals. Annex Table 5.2: Food Ration per Day under WFP-funded Program Commodity Type School Meal (gram/person/day) Maize Meal 150 Pulses 30 Vegetable Oil 10 Total 190 Total kilocalories/day 729 Source: Motseng Logistics Services (2011). 138 Annex 6: Generosity of Cash Transfer Programs in Various African Countries Annex Table 6.1: Generosity of Cash Transfer Programs in Selected Countries (excluding social pensions) As a share of Benefit level Benefit level (US$ poverty Country Program (LCU) eq.) Payment schedule estimates Burkina Nahouri FCFA 1,000, US$2, 4, or 6 per child per . Faso province OVC 2,000, or 3,000 quarter based on cash transfer the age of the a/ pilot child Food vouchers FCFA 1,500 per US$3 per person per month 22% of to urban poor person (ceiling of with 18 as the consumption at 9,000 per ceiling the poverty line household) or 15-18 days of cereal needs Ghana LEAP GHC 8-12 US$4-8 per household per 20% of the month bottom quintile’s household consumption Kenya Combined: CT- KSH 1,500 US$15-26 per household per <20% of the OVC, HSDP, month absolute poverty Disability grant, line OPCT, and Urban Food Subsidy Liberia Bomi CT pilot LD 700, 1,050, US$10, 15, 20, or per household per . 1,400 or 1,750 25 (average total month depending (average total household on the size of the b/ household payment US$25) household payment LD 1,750) Mali Bourse Maman FCFA 5,000 US$8-12 per household per . month Togo WAO Cash FCFA US$44/150 per child per year . Transfer for 22,000/75,000 in education primary/secondary Lesotho OVC bursaries M.1,537 average US$220 per student per . year Child grants M.120 US$17 per household per . month Public assistance M.100 US$14 per household per . month Mauritius Social aid Rs. 1,008 (plus US$33 per household per 16% of extra for children) month consumption of the poor Rwanda VUP Direct RwF 7,500-21,000 US$12-35 per household per Largely benefits Support (based on month elderly who lack household size) sufficient means to cope 139 Swaziland Public assistance E 80 US$10 per person per 17% of month consumption at the poverty line, 37% at food poverty line Young heroes E 180 US$23 per child per 39% of (double orphan month consumption at grant) the poverty line, 84% at food poverty line Zambia Social cash ZMK 60,000 (ZMK US$12 per household per 14% of cost of transfers 50,000 if no month basic food children) basket, 20% of consumption of quintiles 1-2 SPLASH vouchers K 65,000 US$14 per household per 17% of month estimated food poverty line Source: World Bank (ongoing). Notes: FCFA exchange rate rounded to US$1 = FCFA500. a/ In 2009 the program added an addition payment of FSCF 1500 in cash per household for flood-affected households to prevent them from selling food vouchers to finance other household costs related to rebuilding the home or replacing assets lost to the flood. b/ In addition, the household can receive a top-up of LD 150-300 per child sent to primary or secondary school. 140 MAP 33434 141 Social Protection & Labor Discussion Paper Series Titles 2012-2014 No. Title 1415 Sudan Social Safety Net Assessment by Annika Kjellgren, Christina Jones-Pauly, Endashaw Tadesse and Andrea Vermehren, May 2014 1414 Tanzania Poverty, Growth, and Public Transfers: Options for a National Productive Safety Net Program by W. James Smith, September 2011 1413 Zambia: Using Social Safety Nets to Accelerate Poverty Reduction and Share Prosperity by Cornelia Tesliuc and W. James Smith, March 2013 1412 Mali Social Safety Nets by Cécile Cherrier, Carlo del Ninno and Setareh Razmara, January 2011 1411 Swaziland: Using Public Transfers to Reduce Extreme Poverty by Lorraine Blank, Emma Mistiaen and Jeanine Braithwaite, November 2012 1410 Togo: Towards a National Social Protection Policy and Strategy by Julie van Domelen, June 2012 1409 Lesotho: A Safety Net to End Extreme Poverty by W. James Smith, Emma Mistiaen, Melis Guven and Morabo Morojele, June 2013 1408 Mozambique Social Protection Assessment: Review of Social Assistance Programs and Social Protection Expenditures by Jose Silveiro Marques, October 2012 1407 Liberia: A Diagnostic of Social Protection by Andrea Borgarello, Laura Figazzolo and Emily Weedon, December 2011 1406 Sierra Leone Social Protection Assessment by José Silvério Marques, John Van Dyck, Suleiman Namara, Rita Costa and Sybil Bailor, June 2013 1405 Botswana Social Protection by Cornelia Tesliuc, José Silvério Marques, Lillian Mookodi, Jeanine Braithwaite, Siddarth Sharma and Dolly Ntseane, December 2013 1404 Cameroon Social Safety Nets by Carlo del Ninno and Kaleb Tamiru, June 2012 1403 Burkina Faso Social Safety Nets by Cécile Cherrier, Carlo del Ninno and Setareh Razmara, January 2011 1402 Social Insurance Reform in Jordan: Awareness and Perceptions of Employment Opportunities for Women by Stefanie Brodmann, Irene Jillson and Nahla Hassan, June 2014 1401 Social Assistance and Labor Market Programs in Latin America: Methodology and Key Findings from the Social Protection Database by Paula Cerutti, Anna Fruttero, Margaret Grosh, Silvana Kostenbaum, Maria Laura Oliveri, Claudia Rodriguez-Alas, Victoria Strokova, June 2014 1308 Youth Employment: A Human Development Agenda for the Next Decade by David Robalino, David Margolis, Friederike Rother, David Newhouse and Mattias Lundberg, June 2013 1307 Eligibility Thresholds for Minimum Living Guarantee Programs: International Practices and Implications for China by Nithin Umapathi, Dewen Wang and Philip O’Keefe, November 2013 1306 Tailoring Social Protection to Small Island Developing States: Lessons Learned from the Caribbean by Asha Williams, Timothy Cheston, Aline Coudouela and Ludovic Subran, August 2013 1305 Improving Payment Mechanisms in Cash-Based Safety Net Programs by Carlo del Ninno, Kalanidhi Subbarao, Annika Kjellgren and Rodrigo Quintana, August 2013 1304 The Nuts and Bolts of Designing and Implementing Training Programs in Developing Countries by Maddalena Honorati and Thomas P. McArdle, June 2013 1303 Designing and Implementing Unemployment Benefit Systems in Middle and Low Income Countries: Key Choices between Insurance and Savings Accounts by David A. Robalino and Michael Weber, May 2013 1302 Entrepreneurship Programs in Developing Countries: A Meta Regression Analysis by Yoonyoung Cho and Maddalena Honorati, April 2013 1301 Skilled Labor Flows: Lessons from the European Union by Martin Kahanec, February 2013 1220 Evaluating the Efficacy of Mass Media and Social Marketing Campaigns in Changing Consumer Financial Behavior by Florentina Mulaj and William Jack, November 2012 1219 Do Social Benefits Respond to Crises? Evidence from Europe & Central Asia During the Global Crisis by Aylin Isik-Dikmelik, November 2012 1218 Building Results Frameworks for Safety Nets Projects by Gloria M. Rubio, October 2012 1217 Pension Coverage in Latin America: Trends and Determinants by Rafael Rofman and Maria Laura Oliveri, June 2012 1216 Cash for Work in Sierra Leone: A Case Study on the Design and Implementation of a Safety Net in Response to a Crisis by Colin Andrews, Mirey Ovadiya, Christophe Ribes Ros and Quentin Wodon, November 2012 1215 Public Employment Services, and Activation Policies by Arvo Kuddo, May 2012 1214 Private Pension Systems: Cross-Country Investment Performance by Alberto R. Musalem and Ricardo Pasquini, May 2012 1213 Global Pension Systems and Their Reform: Worldwide Drivers, Trends, and Challenges by Robert Holzmann, May 2012 1212 Towards Smarter Worker Protection Systems: Improving Labor Regulations and Social Insurance Systems while Creating (Good) Jobs by David A. Robalino, Michael Weber, Arvo Kuddo, Friederike Rother, Aleksandra Posarac and Kwabena Otoo 1211 International Patterns of Pension Provision II: A Worldwide Overview of Facts and Figures by Montserrat Pallares-Miralles, Carolina Romero and Edward Whitehouse, June 2012 1210 Climate-Responsive Social Protection by Anne T. Kuriakose, Rasmus Heltberg, William Wiseman, Cecilia Costella, Rachel Cipryk and Sabine Cornelius, March 2012 1209 Social Protection in Low Income Countries and Fragile Situations: Challenges and Future Directions by Colin Andrews, Maitreyi Das, John Elder, Mirey Ovadiya and Giuseppe Zampaglione, March 2012 1208 World Bank Support for Pensions and Social Security by Mark Dorfman and Robert Palacios, March 2012 1207 Labor Markets in Middle and Low Income Countries: Trends and Implications for Social Protection and Labor Policies by Yoonyoung Cho, David Margolis, David Newhouse and David Robalino, March 2012 1206 Rules, Roles and Controls: Governance in Social Protection with an Application to Social Assistance by Lucy Bassett, Sara Giannozzi, Lucian Pop and Dena Ringold, March 2012 1205 Crisis Response in Social Protection by Federica Marzo and Hideki Mori, March 2012 1204 Improving Access to Jobs and Earnings Opportunities: The Role of Activation and Graduation Policies in Developing Countries by Rita Almeida, Juliana Arbelaez, Maddalena Honorati, Arvo Kuddo, Tanja Lohmann, Mirey Ovadiya, Lucian Pop, Maria Laura Sanchez Puerta and Michael Weber, March 2012 1203 Productive Role of Safety Nets by Harold Alderman and Ruslan Yemtsov, March 2012 1202 Building Social Protection and Labor Systems: Concepts and Operational Implications by David A. Robalino, Laura Rawlings and Ian Walker, March 2012 1201 MicroDeterminants of Informal Employment in the Middle East and North Africa Region by Diego F. Angel-Urdinola and Kimie Tanabe, January 2012 To view Social Protection Discussion papers published prior to 2012, please visit www.worldbank.org/sp Abstract This report shows that while more inclusive growth is the ultimate solution to poverty in Lesotho, the country can and should use selective social transfers to reduce poverty more rapidly among the extreme poor. But because the majority of the transfers are received by people who are not among the extreme poor there is room for increasing the efficiency and effectiveness of spending on safety nets which. These programs should be productive and concentrate on the extreme poor Basotho. In the long run, Lesotho should move towards a more consolidated safety net and strengthen existing programs, such as the Child Grants Program, that already provide some important and positive outcomes and enjoy strong popular and political support. About this series... Social Protection & Labor 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., Room MSN G8-803, Washington, D.C. 20433 USA. Telephone: (202) 458-5267, Fax: (202) 614-0471, E-mail: socialprotection@worldbank.org or visit the Social Protection website at www.worldbank.org/sp. Printed on recycled paper