99494 SOCIAL PROTECTION IN FRAGILE AND CONFLICT-AFFECTED COUNTRIES: TRENDS AND CHALLENGES SOCIAL PROTECTION IN FRAGILE AND CONFLICT- AFFECTED COUNTRIES: TRENDS AND CHALLENGES Mirey Ovadiya Adea Kryeziu Syeda Masood Eric Zapatero © 2015 International Bank for Reconstruction and Development / The World Bank 1818 H Street, NW Washington, DC 20433 Telephone: 202-473-1000 Internet: www.worldbank.org The findings, interpretations, and conclusions expressed here do not necessarily reflect the views of the Executive Directors of The World Bank or the governments they represent. The World Bank does not guarantee the accuracy of the data included in this work. The boundaries, colors, denominations, and other information shown on any map in this work do not imply any judgment on the part of The World Bank concerning the legal status of any territory or the endorsement or acceptance of such boundaries. Rights and Permissions The material in this work is subject to copyright. Because the World Bank encourages dissemination of its knowledge, this work may be reproduced, in whole or in part, for noncommercial purposes as long as full attribution to this work is given. Any queries on rights and licenses, including subsidiary rights, should be addressed to the Office of the Publisher, The World Bank, 1818 H Street NW, Washington, DC 20433, USA; fax: 202-522-2422; e-mail: pubrights@worldbank.org. Photos: Copyright © Andrea Borgarello/World Bank Cover design/layout and editing: Nita Congress CONTENTS Acknowledgments v Abbreviations vi 1 Introduction 1 2 Basic Concepts 3 Social Protection 4 Characteristics of Fragility 6 Fragile States 7 Fragility and Poverty 7 3 Arriving at a Typology 9 4 Trends and Findings 15 Objectives 16 Programming Choices 17 Coverage 20 Expenditures 24 Delivery Mechanisms 27 5 Highlights of Program Delivery 31 6 Conclusions 37 iii iv Social Protection in Fragile and Conflict- Affected Countries: Trends and Challenges Bibliography 41 Boxes 3.1: Classification of Fragile and Conflict-Affected Countries 11 4.1: Evaluation Results for the Yemeni Social Fund for Development 29 Figures 2.1: Moving from Fragility to Institutional Resilience, Security, Justice, and Jobs 5 2.2: Characteristics of Fragile and Conflict-Affected Situations 6 2.3: Share of the World’s Poor Living in 3.1: Typology of Fragile and Conflict-Affected States 10 4.1: Total Number of Social Assistance Programs in Fragile and Conflict-Affected Countries, by Program Type 18 4.2: Average coverage of Social Assistance Programs in Fragile and Conflict-Affected Countries, by Program Type 19 4.3: Social Assistance and Social Insurance Coverage in Fragile and Conflict-Affected Countries, as a Percentage of Total Population 22 4.4: Social Assistance and Social Insurance Coverage, as a Percentage of the Poorest Quintile 23 4.5: Social Insurance Expenditures in Selected Fragile and Conflict-Affected Countries as a Percentage of GDP 26 4.6: Safety Net Spending as a Percentage of GDP in Fragile and Conflict-Affected Countries 28 Tables 4.1: Benefit Incidence of Social Protection in Poorest Quintile in Selected Fragile and Conflict-Affected Countries (%) 21 ACKNOWLEDGMENTS T his paper has benefited from valu- Junko Onishi, Stefano Paternostro, Nicola able advice and guidance from Lucy Pontara, Lucian Pop, Anita Schwarz, Victoria Bassett, Carine Clert, Aylin Isik Dik- Strokova, John Van Dyck, Michael Weber, melik, Ugo Gentilini, Samira Ahmed Emily Weedon, Ruslan Yemtsov, Giuseppe Hillis, Matthew Hobson, Mira Hong, Kelly Zampaglione, and Yongmei Zhou. It was Johnson, Alex Kamurase, Cem Mete, Mont- developed under the overall guidance of serrat Pallares Miralles, Matthew Norton, Anush Bezhanyan. v Abbreviations ASPIRE Atlas of Social Protection—Indicators of Resilience and Equity CPIA Country Policy and Institutional Assessment GDP gross domestic product PMT proxy means testing 1 INTRODUCTION T his discussion paper is the first output objectives in these countries, reflects on under the umbrella of the program- their revealed objectives, and discusses pro- matic work on social protection in gramming options chosen to achieve those fragile and conflict-affected states. objectives as well as how several countries This work aims to develop operational guid- have overcome particular operational and ance to teams on the likely determinants of capacity constraints. effective social protection programming and policy making in fragile and conflict-affected It is important to understand how social pro- settings. tection works in these settings, which often feature a combination of circumstances such This paper elaborates on the role of social as the following: protection programming, and program- ming design and implementation features, (( Acute poverty either concentrated or that are prominent in fragile and con- widespread (areas affected by conflict flict-affected states. In particular, the paper have lagged behind) and vulnerability to describes the universe of social protection shocks 1 2 Social Protection in Fragile and Conflict- Affected Countries: Trends and Challenges (( Lack of social cohesion/weak social fabric Despite many common characteristics, fragile and conflict-affected states are actu- (( Weak or destroyed infrastructure (phys- ally quite diverse, particularly with regard to ical, financial, etc.) metrics of state capacity and the extent to which they have an enabling environment. need for conflict management (( Implicit This paper presents a methodology that has among special groups (e.g., war veterans) been devised to group countries based on income, capacity, and extent of enabling (( Implicitneed for developing citizen trust environment. Use of this methodology will in the state aid in understanding trends, patterns, and key factors in policy making and program- ming choices—good and bad. 2 BASIC CONCEPTS O ne and a half billion people (nearly gility” should be recognized as a dynamic 30 percent of the world’s poor) live and multidimensional concept. Fragility in areas affected by fragility, con- extends over a broad spectrum of circum- flict, or large-scale organized crim- stances that manifest in a range of coun- inal violence. To date, no low-income fragile tries, including Iraq, Malawi, Myanmar, Sierra or conflict-affected country has achieved a Leone, and Timor-Leste, among others. As single United Nations Millennium Develop- suggested in the literature, perhaps a better ment Goal. While much of the world has way of approaching fragility is to differen- made progress in reducing poverty in the tiate among contexts by considering an enti- last 60 years, areas affected by cycles of con- ty’s level of resilience. Resilience is defined flict have lagged economically and have not (e.g., in OECD 2008a and World Bank 2011h) advanced their human development indica- as a political and social system’s capacity to tors (World Bank 2011h). adapt to shocks. Unlike the more amorphous concept of fragility, this is a highly useful Poverty and fragility become mutually rein- concept, in that it is more aligned with the forcing in such settings. In this regard, “fra- process any entity—a person, a family, a 3 4 Social Protection in Fragile and Conflict- Affected Countries: Trends and Challenges community, a country—must go through employment, entrepreneurship support, when facing multiple challenges. and/or input or food distribution programs (e.g., as in Iraq, Liberia, Sierra Leone, South One of the main mechanisms to help build Sudan, and Timor-Leste), as they realize the resilience and protect the poor and vulner- political and social significance of these pro- able is context-specific, effective social pro- grams in building confidence, including dis- tection programming. However, it is often enfranchised groups, and reducing social the case that ‘‘the greater the need for social tensions. protection, the lower the capacity of the state to provide it” (Devereux 2000); this is The 2011 World Development Report (World particularly true in fragile contexts. Govern- Bank 2011h) argues that strengthening ment capacity is likely to be even weaker in legitimate institutions and the ability of a terms of social protection than for social ser- state to provide stability, justice, security, vices such as health and education, since line and jobs lessens the probability of conflict ministries often retain some capacity even and fragility. Social protection thus plays an in postconflict and fragile situations. These important role in restoring confidence; trans- realities highlight both the need for social forming the institutions that provide secu- protection in fragile and conflict-affected rity, justice, and jobs; addressing external states, as well as the difficulties in setting stresses; and mobilizing international sup- up programs with limited capacity, funding, port to overcome fragility, violence, and con- and—at times—political will. flict (figure 2.1). For instance, through the provision of short-term employment to dis- enfranchised individuals, public works pro- Social Protection grams have the potential to restore a sense of identity to individuals and confidence in By definition, social protection plays an the ability of the state to deliver services and important role in providing income support improve social inclusion and equity (Andrews and access to basic social services to pop- and Kryeziu 2013). This, in turn, contributes ulations most at risk of being affected by to the objective of state building. systemic shocks, such as cycles of conflict and violence. In fragile settings in partic- While evidence on the impact of social pro- ular, social protection often has a dual and tection programming and policies on social simultaneous role of contributing to state cohesion is scant, international experience building and to reducing social inequali- suggests that social protection can be an ties and exclusion. By design, social protec- important platform for tion policies can provide income security to individuals through income support, access (( Promoting voice and participation through to employment opportunities, and insur- program processes, ance mechanisms. These policies then lead to improved social cohesion and reduced (( Improving social inclusion and equality probabilities of social conflict and violence. through temporary labor market partic- Governments often rely on short-term youth ipation, and 2. Basic Concepts 5 Yet, particularly in emergency situations, FIGURE 2.1: MOVING FROM FRAGILITY TO INSTITUTIONAL immediate relief and RESILIENCE, SECURITY, JUSTICE, AND JOBS long-term social cohe- sion may be contra- CITIZEN dicted by the need SECURITY, to rebuild the state JUSTICE, AND JOBS quickly. Some inter- EXTERNAL ventions that may be STRESS necessary to provide RES assistance quickly may TO trigger unintended TRANSFORMING RIN G CO consequences, inequal- NFIDENCE RE S ities, and tensions. TO T RANS F ORMI NG RIN G Thus, while social CONFI RE INS S ITU TOR T TRANSFORMING DENCE TIO protection can make ING NS CON I NS ITU FIDEN T TIO N an important contri- CE S INS ITU T TIO NS VIOLENCE and bution toward state FRAGILITY EXTERNAL SUPPORT AND INCENTIVES building and conflict reduction, it must also be acknowledged that SOURCE: World Bank 2011h. not every type of social protection program or policy can or does con- tribute effectively, and in (( Smoothing social tensions and building a timely manner, to these objectives. trust in response to sudden shocks as well as longer-term fragility (Andrews and There is much to be learned about social pro- Kryeziu 2013). tection in fragile states, particularly in the move away from fragmented programs and A recent stocktaking exercise of World toward effective institutions and systems. Bank–supported community-driven devel- Many operational challenges remain. For opment operations implemented in fragile example, quite a number of fragile countries and conflict situations notes that, even may have long-standing, politically diffi- though many operations do not explicitly cult-to-revoke social protection policies that have peace-building objectives and empha- are ineffective, regressive, and benefit very size service delivery instead, the implicit small and/or fairly well-off populations; or theory of change is one of short-term service that may have spawned a plethora of small delivery outcomes improving voice; instilling and fragmented programs that do not inform trust, confidence, and cohesion; and leading or complement one another. Common logis- to improved governance and a compact tical issues include low coverage, high costs, between citizens, service providers, and the information gaps, and poor administrative state (de Regt, Majumdar, and Singh 2013). infrastructure and physical settings. 6 Social Protection in Fragile and Conflict- Affected Countries: Trends and Challenges Characteristics of Fragility tiple sources of shocks, and weak financial capacity (figure 2.2). They also often have non-inclusive political and economic institu- The World Bank categorizes countries as tions (Acemoglu and Robinson 2012). fragile if they have a harmonized average Country Policy and Institutional Assess- Although many non–fragile and conflict-af- ment (CPIA) rating of 3.2 or the presence fected countries may face most of these issues, of a United Nations and/or regional peace- what exacerbates the situation in fragile and keeping or peace-building mission during the conflict-affected settings is that deficiencies past three years. A total of 36 countries con- exist across multiple issues at the same time stituted the list of fragile states in 2014. They and the deficiencies are mutually reinforcing are generally characterized by weak institu- (Andrews et al. 2012). In these environments, tional capacity, weak governance structures, building social protection systems requires an fragmented societies and competing elites, a analysis of the above constraints, as well as tendency to repeated cycles of conflict, poor of context-specific challenges to identify suit- infrastructure, poor access to services, often able objectives and solutions. It is important very high poverty and vulnerability to mul- to analyze how a combination of various FIGURE 2.2: CHARACTERISTICS OF FRAGILE AND CONFLICT-AFFECTED SITUATIONS Weak Poor financial infrastructure capacity Weak Poor access governance to services Low Fragmented institutional societies capacity Non- High inclusive poverty and FRAGILE AND CONFLICT- political and vulnerability AFFECTED SETTING economic institutions SOURCE: Adapted from World Bank 2012a. 2. Basic Concepts 7 fragile and conflict-affected country charac- fragile states and nonfragile states: about 40 teristics affects the needs of the population, percent compared to 20 percent (OECD 2013; the universe of possible policy and program- Sumner 2012).1 This trend is likely to continue ming responses, and—ultimately—the tra- (figure 2.3). jectory to building social protection systems in different settings. FIGURE 2.3: SHARE OF THE WORLD’S POOR LIVING IN FRAGILE STATES Fragility and Poverty PERCENT 60 Recent analysis of 50 global poverty trends 40 suggests that pov- 30 erty is increasingly becoming an issue 20 closely associated with 10 fragility (Chandy and Gertz 2011; Sumner 0 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2012). High prevalence of poverty in fragile states is a by-product SOURCE: Chandy and Gertz 2011. of fragility, but it is also a cause of fra- gility. This conclusion is consistent with the 2011 World As expected, human development indicators Development Report’s finding that the gap in fragile contexts are some of the worst. “No in poverty is widening between those coun- fragile country has yet achieved a single MDG tries affected by violence and other countries [Millennium Development Goal], and fragile (World Bank 2011h). The poor are dispropor- states are home to half of all children not in tionately found in fragile states, regardless primary school and half of all children who of the list of fragile states used (i.e., the Fund die before reaching their fifth birthday” for Peace’s Fragile States Index or the World (Chandy and Gertz 2011). As the list of fragile Bank’s Harmonized List of Fragile Situations). and conflict-affected states and territories It is estimated that while less than one-fifth covers a broad spectrum from low income (about 18.5 percent) of the world’s popu- to middle income, the nature of poverty is lation lived in fragile states in 2010, these quite varied. However, vulnerabilities are countries hosted about one-third of the world’s poor (400 million out of 1.2 billion). 1 These calculations are based on a list of fragile This makes for a more than twofold differ- countries in the “Alert” category of the Fragile ence in the prevalence of poverty between States Index. 8 Social Protection in Fragile and Conflict- Affected Countries: Trends and Challenges more acute for children and women in most third group that is most affected is youth, a of these settings. Child poverty and malnutri- demographic comprising large numbers of tion are very high, often resulting in increased working-age populations in many fragile and morbidity and poor educational outcomes. A conflict-affected countries. 3 ARRIVING AT A TYPOLOGY T o more precisely characterize social (( CAPACITY, which is here approximated by protection programming in fragile (1) the CPIA index, which rates countries and conflict-affected settings, the uni- against a set of 16 criteria in four clus- verse of fragile and conflict-affected ters: economic management, structural states, as listed by the World Bank in its 2014 policies, policies for social inclusion and Harmonized List of Fragile Situations, was equity, and public sector management divided into five different clusters.1 Figure 3.1 and institutions; and (2) country per capita illustrates these groupings. The two key vari- income, grouped according to 2012 gross ables used to map the clusters are national income per capita, calculated using the World Bank Atlas method;2 and 1 The classification was arrived at based on (1) cluster analysis of the fragile and conflict-af- fected country data set (using STATA), (2) updates based on recent political and economic develop- 2 The groups are low income, $1,035 or less; lower ments from the literature, and (3) the team’s middle income, $1,036–$4,085; upper middle judgment on countries’ overall positions vis-à-vis income, $4,086–$12,615; and high income, their capacity and enabling environment. $12,616 or more. 9 10 Social Protection in Fragile and Conflict- Affected Countries: Trends and Challenges FIGURE 3.1: TYPOLOGY OF FRAGILE AND CONFLICT-AFFECTED STATES CLUSTER A CLUSTER D CLUSTER C low  G  CAPACITY  g  high CLUSTER B CLUSTER E l = resource rich  l = resource poor low  G  ENABLING ENVIRONMENT  g  high NOTE: The typology represents the authors’ rendition based on (1) a cluster analysis of the fragile and conflict-affected country data set (using STATA), (2) updates based on recent political and economic developments from the literature, and (3) the team’s judgment on countries’ overall positions vis-à-vis their capacity (using the CPIA index as a proxy) and the enabling environment. Resource-rich countries are determined based on wealth in hydrocarbons and/or minerals. (( ENABLING ENVIRONMENT, approximated by cial development assistance) as a percentage the Corruption Perception Index. of gross national income, and the Organisa- tion for Economic Co-operation and Devel- Other measures, including resource wealth opment/Development Assistance Committee and net developmental assistance (i.e., offi- classification of fragile and conflict-affected 3. Arriving at a Typology 11 countries (box 3.1), have also been factored CLUSTER B. This group consists of countries into the cluster analysis. with low capacity, low income, and a weak enabling environment. Countries may still CLUSTER A. The first group comprises countries be in conflict and/or move in and out of con- with relatively high administrative capacity, flict with certain regions/territories outside lower/upper-middle-income status, and a government control or weak rule of law, weak enabling environment. Countries are and highly fragile due to institutions’ weak- characterized by recent or ongoing political ness. This group includes Afghanistan, the conflict, violence, or instability. Iraq, Syrian Central African Republic, Chad, and South Arab Republic (resource rich), and West Bank Sudan (resource rich); and Haiti and Somalia and Gaza are included in Cluster A. (resource poor). BOX 3.1: CLASSIFICATION OF FRAGILE AND CONFLICT-AFFECTED COUNTRIES The Organisation for Economic Co-operation and Development/Development Assistance Committee clas- sification of fragile and conflict-affected countries is as follows: 1. Deterioration (deteriorating governance environment/ongoing conflict). States where the ability (or willingness) of the state to perform its functions is in decline. This poor performance frequently springs from chronic low capacity and is often associated with very weak rule of law and territory beyond the control of government. These countries are often experiencing conflict or are highly vulnerable to conflict. 2. Prolonged crises or impasse (arrested development). States that fail to use their authority for pro-poor outcomes. The state’s ability to exert its will might be very weak, or very strong donors are typically unwilling to deal with the state directly. 3. Postconflict/crisis or political transition. The states offer a window of opportunity for stakeholders to work together with government on a program of reform. However, the transition is fragile, with the prospect of return to conflict remaining high. 4. Early recovery/gradual improvement. Countries where some effort is being made to improve performance but where performance is patchy. These countries might be postconflict or countries where conflict is not the primary driver. Often, there is no strong leadership championing reform within government, and capacity to implement reforms is weak. Frequently, countries can move from one category to another in a short amount of time. SOURCE: OECD 2008b. 12 Social Protection in Fragile and Conflict- Affected Countries: Trends and Challenges CLUSTER C. The third group is a set of coun- and occupied lands, skewed population dis- tries with medium capacity and average (or tributions, natural disasters, epidemics, and medium) enabling environment. This is a environmental hazards), most fragile and mixed group with countries that may have conflict-affected countries have a ranking of faced sporadic violence due to prolonged 8 or above in the Fragile States Index—e.g., political impasse, or sporadic conflict. This Chad, 9.5; Haiti, 9.6; and Sierra Leone, 9. By group includes Côte d’Ivoire, Mali, Sudan, way of contrast, the nonfragile states of the and the Republic of Yemen (resource rich); United Kingdom and the United States have and Nepal and Zimbabwe (resource poor). rankings of 2 and 3, respectively. Not surpris- ingly, most fragile and conflict-affected coun- CLUSTER D. The fourth group comprises tries suffer from uneven development (with relatively higher-income countries (i.e., real or perceived economic and social dispar- lower-middle income), with relatively ities), weak human rights records, highly (or high administrative capacity and a strong completely) illegitimate governments, and enabling environment. Among the countries weak and divided political ruling elites. in this cluster are Bosnia and Herzegovina and Kosovo. Using the typology, this paper intends to contribute to the discussion on social protec- CLUSTER E. The final group includes two tion policy making in fragile and conflict-af- subsets of countries based on their resource fected countries by presenting some of the status, all of which are low capacity and trends and challenges. Going forward, some have a relatively strong enabling environ- of these experiences will be further studied ment. Countries may have been out of con- and developed in-depth to identify key pro- flict over an extended period yet still face cesses and factors that can lead to successful a variety of governance challenges. The social protection programming and policies. countries in the resource-rich group include Burundi, Republic of Congo, Madagascar, The analysis in section 4 is based on a review Sierra Leone, Timor-Leste, and Togo; the of experiences of a representative sample of resource-poor group includes Comoros, fragile and conflict-affected countries. The Malawi, Myanmar, and the Solomon Islands. sample was chosen from data available in the World Bank’s ASPIRE (Atlas of Social Pro- Countries classified as having a weaker tection: Indicators of Resilience and Equity) enabling environment in the typology database. The analysis describes the mix of (mainly Clusters A and B) also have a higher social protection policies and programming, total ranking in the 2013 Fragile States Index. delivery mechanisms, and likely determinants Fragile and conflict-affected countries are of success in delivering or scaling up social generally characterized as facing numerous protection programming in fragile and con- demographic, economic, and social pres- flict-affected countries. The trends presented sures. In terms of demographic pressures are based on analysis done using available (including high population density relative data in ASPIRE for a group of fragile and to the availability of food, settlements that conflict-affected states and supplemental restrict human freedoms, border disputes research on seven countries. 3. Arriving at a Typology 13 The analysis should be caveated by noting sourced nature of social protection in fragile the scarcity of data and empirical evidence on and conflict-affected countries. Finally, and social protection across fragile and conflict-af- perhaps most importantly, it should be noted fected countries, including labor force surveys that social protection programming in fragile and poverty and demographic data. Accurate and conflict-affected countries could greatly or up-to-date data on total national expendi- benefit from rigorous impact evaluations on tures on social protection are also unavailable, those interventions that have been successful mostly due to the fragmented and externally in reaching their objectives. 4 TRENDS AND FINDINGS S ocial protection programs and policies to strategic visioning, often entailing a lack in fragile and conflict-affected set- of coordination mechanisms and high depen- tings occupy an especially broad spec- dency on donor funding (Honorati and Rodri- trum, balancing short-term emergency guez forthcoming). In addition, the majority needs (i.e., in response to conflict or natural of programs are fragmented by design, with disaster) with longer-term needs of reducing poor administrative linkages between them. chronic poverty and inequality. Across all 36 fragile states, there is a notice- able trend in social protection programming Though interest exists in building systems, toward cash transfers, public works, and most social protection interventions in fragile skills development programs/self-employ- and conflict-affected countries remain some- ment support; maintained support for com- what ad hoc and opportunistic. The World munity-based services; and a shift away from Bank’s social protection assessment review in-kind interventions. The balance between reveals that, in most low-income countries, social assistance and social insurance varies, there are serious challenges at the policy level but social insurance is much lower on average. 15 16 Social Protection in Fragile and Conflict- Affected Countries: Trends and Challenges Research suggests that the composition of protection has a more limited scope and social protection is not static, and the mix mostly serves to help the poor and vulner- of programming and the emphasis of social able. Similarly, in resource-rich countries with protection changes with shifts in the degree low capacity and low income per capita, of fragility and the distance from conflict. social protection policies often seem to be The typology captures a snapshot in time; used to maintain social stability. transitions continually occur, and there is steady movement between the various In Cluster D countries (high capacity and strong clusters as emergencies recur. This fluidity is enabling environment), social protection has a most apparent in Clusters A and B, but is also strong legal basis and is viewed as a right. For observable in the other clusters in response example, Kosovo provides a noncontributory to multiple and arising challenges such as pension to all its citizens over the age of 65 Ebola, weather events, political upheavals, regardless of whether they are rich or poor. and others. Cluster E countries (low capacity and strong enabling environment) mostly use social pro- tection for supporting the most vulnerable. Objectives Togo, for example, spends 59 percent of its social protection expenditures on food and SOCIAL PROTECTION OBJECTIVES IN FRAGILE in-kind transfers. In resource-rich Cluster E AND CONFLICT-AFFECTED SETTINGS ARE MUL- countries, social protection is used to promote TIFACETED. Although the primary objective social cohesion.1 In Timor-Leste, most of the of social protection is to protect the indi- vidual against life events/risks and economic and other shocks, in fragile and conflict-af- 1 The 2010 International Institute of Social Studies’ Indices of Social Development, which fected settings, social protection interven- are measures of social cohesion and fragility, tions are used—often following a conflict illustrate that fragile and conflict-affected period—to reduce social tensions, promote countries have lower civic activism scores (and social cohesion and social inclusion, and in turn, lower awareness and information on political processes) than nonfragile developed serve as a peace dividend for certain parts states, among other measures (http://www. of the population. Social cohesion in this con- indsocdev.org/). The range of civic activism text is associated with sustaining peace and scores is between 0.41 in Chad (a low-capacity social stability in a country with a history of country with a weak enabling environment) to 0.51 in Bosnia and Herzegovina (a high-capacity conflict; social inclusion means supporting country with a strong enabling environment). groups that have historically been margin- This finding reflects the fact that civil societies alized. There are noticeable differences in in fragile and conflict-affected countries have the social protection objectives of high- and a lower level of engagement in civic activities, such as peaceful protests. In terms of intergroup low-capacity countries (as per the typology cohesion, fragile and conflict-affected countries developed above) as well. In high-capacity are very heterogeneous. This is because the countries (Clusters A, D, and to some degree index measures the probability of violence and C), the scope of social protection is broader terrorism in the country; thus, a fragile and con- flict-affected country with recent or ongoing and used as a platform to develop a rights- violence such as Iraq has a very low score (0.18), based social contract between the citizens while Togo has a rating (0.72) on a par with the and the state. In low-capacity states, social United Kingdom and the United States. 4. Trends and Findings 17 social protection spending goes to provide cash or the unemployed. Within social assistance, transfers to volatile groups such as the veterans most of the fragile and conflict-affected of the freedom struggle. In Cluster C coun- countries analyzed in this programmatic tries (moderate capacity and average enabling work have unconditional cash transfer pro- environment), social protection is mostly used grams or in-kind assistance programs with for social inclusion. For instance, Nepal pro- categorical benefits being prominent. A few vides social assistance to low-caste individuals countries also have cash transfer programs and widows, both historically marginalized with soft conditionalities in place. Labor pro- groups. The Republic of Yemen, a resource-rich grams are scarce and serve very few people. Cluster C country, has used social protection They have a narrow focus on public works, to maintain social and political stability. The employment, or entrepreneurship support government has been reluctant to abolish the programs; and a very limited focus on job financially unsustainable petroleum subsidies, brokerage or formal employment services. because any indication of doing so triggers protests. In addition, it has substantially scaled As shown in figure 4.1, food, in-kind, or near- up the Social Welfare Fund, a cash transfer cash programs are most prominent in fragile program, in the aftermath of the anti-govern- and conflict-affected countries, followed by ment protests and crisis in 2011/12. In countries public works and cash transfers. with weak enabling environments (Clusters A and B) which face sudden bursts of violence or Further insights on the dynamics of social political turmoil, social protection is still fairly protection programming emerge on ana- strongly associated with emergency humani- lyzing coverage by program type: tarian aid; in Cluster A countries with higher capacity, there is a more noticeable transition (( Cash transfers are most prevalent among or attempts toward systematic social protec- Cluster D countries, which have high tion approaches. For example, West Bank and capacity/strong enabling environments; Gaza have been progressively transitioning they are also common in Cluster A coun- from in-kind to cash transfers and to better tries, which have challenging environ- delivery systems such as registries, targeting, ments of current or partial conflict. and payment systems. (( Inhigh-capacity countries with strong enabling environments (Cluster D), social Programming Choices protection is largely systematic and deliv- ered through a few social assistance pro- OVERALL, FRAGILE AND CONFLICT-AFFECTED grams and social insurance. For example, STATES TEND TO HAVE A STRONGER FOCUS ON in Kosovo, the basic pensions program SOCIAL ASSISTANCE THAN ANY OTHER TYPE OF covers almost all of the targeted popula- SOCIAL PROTECTION PROGRAMMING, THOUGH tion (over the age of 65), in line with the THERE ARE NOTABLE CASES—SUCH AS KOSOVO, concept of the program being a citizen- NEPAL, AND WEST BANK AND GAZA—WHERE ship right provided by the state. MORE IS SPENT ON SOCIAL INSURANCE THAN ON SOCIAL ASSISTANCE. Social insurance coverage (( Other social assistance (primarily subsi- is typically limited to public sector employees dies for fuel and food) is most prevalent 18 Social Protection in Fragile and Conflict- Affected Countries: Trends and Challenges FIGURE 4.1: TOTAL NUMBER OF SOCIAL ASSISTANCE PROGRAMS IN FRAGILE AND CONFLICT-AFFECTED COUNTRIES, BY PROGRAM TYPE NUMBER 40 Afghanistan Bosnia & Herzegovina Burundi 35 Central African Republic Chad Comoros 30 Congo, Rep. Côte d’Ivoire Haiti 25 Iraq Kosovo 20 Madagascar Malawi Mali 15 Nepal Sierra Leone Solomon Islands 10 South Sudan Syrian Arab Republic Timor-Leste 5 Togo West Bank & Gaza 0 Yemen, Rep. Food and near cash Public Unconditional Conditional Zimbabwe (conditional and Works cash transfer cash transfer unconditional) SOURCES: World Bank 2014e, annex 5; and readily available information on ASPIRE for fragile and conflict-affected countries. 4. Trends and Findings 19 among resource-rich countries in Cluster C considerably expanded its Social Welfare (e.g., Sudan and the Republic of Yemen). Fund—which provides unconditional cash Nevertheless, In Cluster C (moderate transfers to the elderly, orphans, and capacity and average enabling environ- other vulnerable groups—from 100,000 ment), a number of countries appear to beneficiaries in 1996 to 7 million in 2011. be moving toward cash transfers as the On the other hand, coverage of subsi- preferred method of social assistance dies appears to be three times higher despite a continued reliance on subsi- than any other social assistance program dies. The Republic of Yemen recently and (figure 4.2). FIGURE 4.2: AVERAGE COVERAGE OF SOCIAL ASSISTANCE PROGRAMS IN FRAGILE AND CONFLICT-AFFECTED COUNTRIES, BY PROGRAM TYPE a. Cash transfer b. Public works/cash for work PERCENT PERCENT 20 20 15 15 10 10 5 5 0 0 Cluster D Cluster A Cluster C Cluster E Cluster B Cluster B Cluster C Cluster E c. School feeding/in kind d. Other social assistance PERCENT PERCENT 20 20 15 15 10 10 5 5 0 0 Cluster E Cluster A Cluster C Cluster B Cluster C Cluster E Cluster B SOURCE: ASPIRE data retrieved January 2015. NOTE: Other social assistance (figure 4.2d) refers primarily to subsidies. 20 Social Protection in Fragile and Conflict- Affected Countries: Trends and Challenges (( School feeding/in-kind assistance is tive action, thereby contributing to the most prevalent among Cluster E coun- country’s stability, peace, and sustainable tries comprised of primarily low-income, growth. In the Republic of Yemen, under low-capacity countries with low human the Social Fund for Development, 4.5 mil- development indicators. These countries lion poor and vulnerable individuals (rep- have also successfully used community resenting 65 percent of the population structures to provide social assistance to living in extreme poverty) benefited from target populations. The village chiefs in over 2,000 community projects. Among Timor-Leste are instrumental in helping the positive impacts were an increase in with beneficiary selection; in Togo, an girls’ schools, an increase in enrollment already existing social institution of rates, improved access to rural roads, and femmes-mamans has been used to deliver reduced travel costs and times. Through food for the school feeding program. Both Afghanistan’s National Solidarity Pro- Sierra Leone and Togo have active commu- gram, community development councils nity-driven development programs. allocated funds to rebuild infrastructure, repair schools, and install water pumps to (( Incountries with weak enabling environ- the benefit of over 13 million people coun- ments (Clusters A and B) and those prone trywide, promoting the state’s credibility to frequent emergencies, infrastructure and local governance. breakdown, and dysfunctional markets, social protection is largely in the form of (( Regarding labor market interventions, in-kind social assistance. Haiti has a Tar- about half of all countries and territories geted Nutrition Program in place, which in fragile situations (19 out of 36 on the provides food-based assistance to the vul- 2015 list) have some type of youth employ- nerable. Numerous countries in Clusters B, ment programs in place. Globally, 62 youth C, and E (e.g., Afghanistan, Madagascar, employment programs out of 733 (8 per- Sierra Leone, South Sudan, Togo, and the cent) are in fragile and conflict-affected Republic of Yemen), which have lower states. These interventions are most com- capacity settings, make use of community monly used in Africa and the Middle East structures to deliver social protection—in and North Africa, and have a strong focus particular, to facilitate access to services on skills training and entrepreneurship and to implement public works, livelihood promotion. support, or school feeding programs. (( Such approaches have supported signifi- Coverage cant postdisaster reconstruction efforts, and have helped shape new dynamics Typically, fragile and conflict-affected coun- at the community level by working to tries suffer from low coverage of social pro- restore social cohesion. Sierra Leone’s tection programs; this is particularly true for community-driven development project social insurance and labor market programs. assisted fragile and vulnerable war-af- A high percentage of the population in these fected communities in reducing poverty countries generally lives below the poverty and building local capacity for collec- line, and access to services is often limited. 4. Trends and Findings 21 Moreover, usually due to less-than-optimal ever, that this rate can be misleading, as targeting mechanisms, a high number of there is large variation within and between benefits accrue to the nonpoor. This last is social insurance and social assistance pro- borne out by ASPIRE data: the global average grams. Overall, coverage of social insurance for benefit incidence for the poorest quin- lags that of social assistance (figure 4.3), tile for all social protection is calculated at and the poor are more likely not covered about 23 percent; among fragile and con- by either social assistance or social insur- flict-affected countries, it is only about 14 ance schemes. Bosnia and Herzegovina percent (table 4.1). The remainder of this sec- and Kosovo are the main exceptions in this tion examines trends in coverage of social regard, primarily due to their inheritance assistance and social insurance programs in of the Yugoslav pension schemes. In Iraq, fragile and conflict-affected countries, using which has social assistance coverage of 80 data from ASPIRE where available. percent, universal food vouchers account for the high coverage rate. On average, coverage of social protection programs in fragile and conflict-affected Figure 4.4 goes one step further in explaining settings is around 27 percent, compared to the low coverage of the poor in fragile and a global average of 43 percent. Note, how- conflict-affected countries. Looking at the TABLE 4.1: BENEFIT INCIDENCE OF SOCIAL PROTECTION IN POOREST QUINTILE IN SELECTED FRAGILE AND CONFLICT-AFFECTED COUNTRIES (%) All social All social Country/economy assistance All social insurance protection Afghanistan 8.24 3.76 7.02 Bosnia and Herzegovina 30.58 14.13 14.63 Iraq 17.93 10.18 12.22 Kosovo 43.38 19.02 25.35 Malawi 6.44 0.57 2.88 Mali — 0.47 0.47 Nepal 15.76 0.94 6.24 Timor-Leste 1.39 — 1.32 West Bank and Gaza 63.65 45.30 62.57 Yemen, Rep. 19.08 17.15 10.96 Average 22.90 12.30 14.30 SOURCE: ASPIRE data retrieved January 2015. NOTE: — = not available. 22 Social Protection in Fragile and Conflict- Affected Countries: Trends and Challenges FIGURE 4.3: SOCIAL ASSISTANCE AND SOCIAL INSURANCE COVERAGE IN FRAGILE AND CONFLICT-AFFECTED COUNTRIES, AS A PERCENTAGE OF TOTAL POPULATION PERCENTAGE 0F TOTAL POPULATION 90 Coverage of social assistance Coverage of social insurance 80 70 60 50 40 30 20 10 0 Afghanistan Bosnia & Herzegovina Congo, Rep. Haiti Iraq Kosovo Malawi Madagascar Nepal Sierra Leone Timor-Leste West Bank &Gaza Yemen, Rep. Côte d’Ivoire SOURCE: ASPIRE data retrieved January 2015. NOTE: Coverage data corresponds to different years for each country/economy, as follows: Afghanistan, 2007; Bosnia and Herzegovina, 2007; Republic of Congo, 2005; Côte d’Ivoire, 2002; Haiti, 2001; Iraq, 2006; Kosovo, 2006; Madagascar, 2010; Malawi, 2010; Nepal, 2010; Sierra Leone, 2011; Timor-Leste, 2007; West Bank and Gaza, 2007; and Republic of Yemen, 2005. poorest quintile of the population, of the tributory social protection schemes beyond 14 countries included in the figure, 10 have the civil service (which accounts for a small social assistance coverage of less than 30 per- percentage of the total population). Also, cent. The range of social insurance spending there are often large information and is from 0 to 48 percent, with nine countries capacity gaps mitigating effective mainte- spending less than 10 percent, for an average nance of such systems; moreover, records are of 3 percent. lost during conflict/fragile situations, and the organizations and institutional settings tend For fragile and conflict-affected states, it is to be damaged. Especially in Middle East and not uncommon to see such low coverage of North African fragile and conflict-affected social insurance, given the relatively small countries, where social insurance schemes percentage of the population in formal exist, they impose an enormous financial employment and the unavailability of con- burden. Consequently, the schemes tend to 4. Trends and Findings 23 FIGURE 4.4: SOCIAL ASSISTANCE AND SOCIAL INSURANCE COVERAGE, AS A PERCENTAGE OF THE POOREST QUINTILE PERCENTAGE 0F TOTAL POPULATION 90 Coverage of social assistance Coverage of social insurance 80 70 60 50 40 30 20 10 0 Afghanistan Bosnia & Herzegovina Congo, Rep. Côte d’Ivoire Haiti Kosovo Malawi Madagascar Nepal Sierra Leone Timor-Leste West Bank & Gaza Yemen, Rep. Iraq SOURCE: ASPIRE data retrieved January 2015. NOTE: Coverage data corresponds to different years for each country/economy, as follows: Afghanistan, 2007; Bosnia and Herzegovina, 2007; Republic of Congo, 2005; Côte d’Ivoire, 2002; Haiti, 2001; Iraq, 2006; Kosovo, 2006; Madagascar, 2010; Malawi, 2010; Nepal, 2010; Sierra Leone, 2011; Timor-Leste, 2007; West Bank and Gaza, 2007; and Republic of Yemen, 2005. have low coverage, be fragmented, yet still enabling environment, the government account for a high level of spending—for established a series of social assistance pro- example, in West Bank and Gaza, pension grams to address the most urgent needs of schemes equal 4 percent of gross domestic the population following the decade-long product (GDP), and cover only 15 percent of civil war. The programs, however, suffer the population. from low coverage, high leakage, and inef- ficient administration. Programs remain Among the countries analyzed, key deter- small, underfunded, and highly dependent minants of social protection coverage are on donors. Similar experiences are evident in level of administrative capacity and govern- Nepal, where the government has not been ment policy/vision. For instance, in Sierra able to consolidate programming and pre- Leone (Cluster E), a country classified as a vent the inclusion of the nonpoor in social low-capacity setting with a relatively strong assistance programs. 24 Social Protection in Fragile and Conflict- Affected Countries: Trends and Challenges In contrast, in West Bank and Gaza (Cluster A), Leste, the three major social assistance pro- with a high capacity and a very weak enabling grams fail to reach 60 percent of the bottom environment, 16 existing social safety net two quintiles. Thus, inefficient targeting (to programs cover more than half of the poorest undeserving or unintended groups) in social quintile—a rate above the world average. protection programming remains a serious The West Bank and Gaza’s cash transfer issue, even in fragile and conflict-affected program, which uses a proxy means testing countries where coverage is relatively high. (PMT) formula to ensure greater inclusion of the poor, reached 65 percent of the poor population in 2013 (97,000 households). The Expenditures World Food Programme also distributes food aid to about 800,000 beneficiaries, with the TRENDS. Most fragile and conflict-affected help of the Ministry of Social Assistance and countries— especially those with low other nongovernmental organizations; this capacity and those that lack natural resource accounts for nearly a third of all safety net wealth—rely on external funding for social spending. In Kosovo, another high-capacity protection. This dependency raises concerns case, coverage is relatively high for social regarding the financial sustainability of social insurance/pensions, with the basic pension protection in such countries. Some high- reaching 35 percent of the population. How- and medium-capacity countries like Kosovo ever, the country lags with regard to social (Cluster D) and Nepal (Cluster C) or resource- assistance and labor programs: social assis- rich countries like Timor-Leste (Cluster E) are tance reaches 11 percent of the entire popu- the exceptions to this generalization. lation, and 30 percent of the poorest quintile (Gassmann and Roelen 2009). The majority of fragile and conflict-affected countries also direct a large proportion of Timor-Leste (Cluster E) and the Republic social protection expenditures to social of Yemen (Cluster C) (low capacity–strong insurance programs and categorical bene- enabling environment, and medium fits, though the coverage of these programs capacity–average enabling environment, remains low with poor benefit incidence, as respectively) have also managed to cover shown above. Social insurance programs, in larger portions of their populations. In the particular, can be regressive in many fragile Republic of Yemen, the Social Welfare Fund and conflict-affected countries because they reaches close to 7 million beneficiaries or cover civil servants (sometimes solely) and, 1.5 million households, which represents in high-capacity countries, formal sector 7 percent of the population. In Timor-Leste, employees; these are typically nonvulner- 53 percent of the population receives some able groups. In fragile and conflict-affected sort of social assistance. However, in both countries with an aging covered population, countries, the benefits tend to accrue more the cost of social insurance will rise quickly to the nonpoor. In the Republic of Yemen, and soon. The coverage of pension systems 70 percent of the beneficiary population was in most fragile and conflict-affected coun- not in the intended target group; of these tries remains overwhelmingly low, and the untargeted beneficiaries, 75 percent were populations are predominantly young. This not classified as poor. Similarly, in Timor- combination leads to pension system depen- 4. Trends and Findings 25 dency ratios (beneficiaries/contributors) that important source of income for many fragile are fast increasing. Although the rationale and conflict-affected households, and that behind social assistance programs and sub- the impact of changes in remittances due sidies in fragile and conflict-affected coun- to shocks (such as a conflict) can be large, at tries—as in many other countries—is that least for those households that benefit from they would support vulnerable groups, the them (World Bank, n.d.). Remittances tend to nonpoor benefit disproportionately, making serve as insurance against risks and help mit- these programs and expenditures highly igate vulnerability (as in Haiti and Pakistan in regressive and ineffective. Nonetheless, they response to natural disasters). Unlike foreign fulfill political and social objectives. aid, remittances tend to have a countercy- clical nature, as they increase during down- SOURCES OF FINANCING. In Kosovo, a Cluster D turns in the recipient countries. They can country (high capacity and strong enabling constitute a high level of household income; environment), the government budget however, they cannot be a substitute for finances social protection schemes apart from domestic growth and employment-genera- the social insurance program, which is funded tion efforts. In Mali, remittances are saved by employers (Including the government for for unexpected events and serve as a private public sector employees) and employee con- safety net (Ratha 2013). tributions. These schemes constitute 5.8 per- cent of GDP, which is almost the same as the While the Organisation for Economic Co-op- global average at 6 percent. Cluster E (low eration and Development’s list of fragile capacity and strong enabling environment) states differs slightly from the World Bank’s, consists of both resource-rich and resource- the trends it portrays regarding remittances poor countries. Timor-Leste, a resource-rich in a recent report mirror those noted above country, funds all social protection program- (OECD 2014). It finds that in many fragile ming through its $7 billion petroleum fund. and conflict-affected countries, particularly Togo a resource-poor country, funds only middle-income ones, remittances have out- 25 percent of its social protection program- paced aid as the major source of develop- ming through its own revenues. In Cluster D ment funding; this is in contrast to nonfragile (low capacity and weak enabling environ- developing countries, where foreign direct ment) countries, most social protection investment constitutes the larger share. programming is externally funded and off budget, as in Haiti. In Cluster  A countries LEVEL OF EXPENDITURE. In many fragile and (high capacity and weak enabling environ- conflict-affected countries, expenditures ment), social protection financing is mixed: on social protection (including government ranging from, e.g., West Bank and Gaza with and fairly substantial external sources) are a high dependence on external sources to close to the global average as a percentage Iraq, which has less dependency. Neverthe- of GDP. However, a large majority of these less, spending is within a budget framework expenditures are skewed toward categor- and concentrated around national programs. ical benefits and social insurance. Within social protection spending, there is consid- REMITTANCES. Overall, empirical evidence erable heterogeneity with regard to expen- suggests that remittances represent an diture by category of social protection. For 26 Social Protection in Fragile and Conflict- Affected Countries: Trends and Challenges example, social insurance spending ranges Bosnia and Herzegovina is an outlier with from 0.4 percent to 9.4 percent of GDP, with respect to its spending on social insurance, the average pension spending among the as figure 4.5 shows. While its spending is fragile and conflict-affected countries at nearly 10 percent of its GDP, its coverage is about 2 percent of GDP (of the representa- also quite high at 30 percent (total old-age tive sample shown in figure 4.5). This level beneficiaries/population over 65 years). of expenditure is unsustainable, considering the low coverage and young demographic In Cluster C countries (moderate capacity profile in most of these countries, where an and average enabling environment), there average of less than 5 percent of the popu- has been a noticeable increase over the last lation is over 65 years of age. Comparisons few years in social protection spending. For to other countries with similar proportions example, Nepal increased its social protection of an elderly population confirm the heavy spending from 1.5 percent of GDP in 2008 to burden of pension expenditures in many 2.5 percent of GDP in 2012 following its period fragile and conflict-affected countries. of conflict; 95 percent of the expenditures FIGURE 4.5: SOCIAL INSURANCE EXPENDITURES IN SELECTED FRAGILE AND CONFLICT- AFFECTED COUNTRIES AS A PERCENTAGE OF GDP Percentage of GDP 10 9 8 7 6 5 4 3 2 1 0 Bosnia & Herzegovina Iraq West Banka & Gaza Kosovo Mali Yemen Malawi Syria Nepal Zimbabwe CAR Togo Burundi Cote d'Ivoire Myanmar Afghanistan Sierra Leone SOURCE: Data retrieved from HDNSP (Human Development Network Social Protection) Global Pensions Database (http://www. worldbank.org/en/topic/socialprotectionlabor/brief/pensions-data), January 2015. 4. Trends and Findings 27 were financed by tax revenues. Mozambique Gaza Cash Transfer Program uses a single and Rwanda, although not formally classi- registry and fully functioning PMT system. fied as fragile and conflict-affected countries, Nepal and Timor-Leste have less administra- have also increased allocations to social pro- tive capacity; they continue to rely primarily tection in the aftermath of conflict. Mozam- on categorical and community targeting, bique’s expenditures rose from 0.2 percent of respectively. Overall, it is evident that lon- GDP in 2010 to 0.5 percent in 2014. ger-term and better-established programs, in attempting to target a higher percentage In many fragile and conflict-affected coun- of the poor, are turning to PMT methods tries, social protection programming is used to minimize inclusion/exclusion errors and to benefit former “freedom” fighters/war establish more sophisticated mechanisms veterans and their survivors. Due to low cov- for targeting (e.g., West Bank and Gaza, the erage, these schemes and benefits are highly Republic of Yemen). regressive and absorb significant portions of social protection spending. PAYMENT SYSTEMS. As with other adminis- trative processes, capacity and financial The recent State of Social Safety Nets 2014 infrastructure are needed in order to use report (World Bank 2014e) illustrates the effective methods of payment. However, the range of spending patterns across fragile and availability of mobile technology has allowed conflict-affected countries, with most relying many low-capacity countries with relatively on external sources of funding for their social weak enabling environments to overcome safety net expenditures. Figure 4.6 shows the the challenge of low administrative capacity. variation in this spending as a percentage Although sophisticated payment methods of GDP across 14 countries with comparable are available in Cluster B countries—such data available from 2008 onwards. as Haiti and Somalia—coverage remains quite low. Higher-capacity countries, such as Kosovo, rely on financial institutions and pri- Delivery Mechanisms marily electronic payments to transfer social insurance benefits. West Bank and Gaza has TARGETING. Administrative capacity plays a recently switched from post offices to banks crucial role in the ability of a government to in delivering cash benefits, as has Nepal. At target the poor or other deserving popula- the other extreme, Mozambique and Timor- tions. Several countries—including Mozam- Leste still use trucks or person-to-person bique, Sierra Leone, and the Republic of methods to deliver cash benefits. Yemen—use a combination of targeting methods including categorical, geograph- MONITORING AND EVALUATION. While building ical, community, and self-selection to select an evaluation base can take years and sub- beneficiaries for social protection programs. stantial resources, few impact evaluations In Sierra Leone and the Republic of Yemen, associated with the countries discussed here there has been a move away from simply have been completed. Evaluations do exist of using categorical, community, or self-tar- the Republic of Yemen Social Welfare Fund geting toward adding PMT so as to mini- (see box 4.1), the West Bank and Gaza Cash mize inclusion errors. The West Bank and Transfer Program, the Togo School Feeding 28 Social Protection in Fragile and Conflict- Affected Countries: Trends and Challenges FIGURE 4.6: SAFETY NET SPENDING AS A PERCENTAGE OF GDP IN FRAGILE AND CONFLICT- AFFECTED COUNTRIES PERCENT 6 5 4 3 2 1 0 Timor-Leste Sierra Leone Bosnia & Herzogivina Kosovo Yemen Iraq Nepal Madagascar Syria West Bank & Gaza Mali Togo Solomon Islands Afghanistan Country/economy Percentage Data as of Timor-Leste 5.91 2009 Sierra Leone 3.50 2011 Bosnia & Herzegovina 3.33 2010 Kosovo 1.47 2012 Yemen 1.44 2008 Iraq 1.22 2009 Nepal 1.20 2009 Madagascar 1.10 2010 Syria 1.00 2010 West Bank and Gaza 0.81 2010 Mali 0.50 2009 Togo 0.50 2009 Solomon Islands 0.26 2009 Afghanistan 0.02 2009 SOURCE: World Bank 2014e. 4. Trends and Findings 29 BOX 4.1: EVALUATION RESULTS FOR THE YEMENI SOCIAL FUND FOR DEVELOPMENT The Republic of Yemen’s Social Fund for Development focuses on pro-poor long-term development through Community and Local Development, Capacity-Building, Small- and Micro-Enterprise Development, and Labor-Intensive Public Works Programs. The Labor-Intensive Public Works Program played a crucial role in cushioning beneficiary communities from the economic shock of 2010–11, and averted potential longer-term consequences related to selling off assets and incurring debt. The public works program was scaled up to increase relief and humanitarian responses. By employing community members in public works activities, the program effectively reached a large number of rural households, increased average wages, shifted the structure of the workforce away from work in the lowest-paid sectors and caused a significant increase in the probability of female employment. Similarly, the Rain-fed Agriculture and Livestock Program has helped improve agricultural productivity by fostering the formation and training of small producer groups in the country’s rural regions. The program allowed villagers to self-select into participation in project groups where they would make investments in agriculture, receive training in organization and agricultural best practices, and receive subsidies for the purchase of livestock. The program had strong impacts on increasing community solidarity and—to some degree—female empowerment. Program, and the Sierra Leone Public Works many Latin American countries with mature Program. The impact evaluations, by nature, social protection programs—including Brazil, are not systematic; thus, it remains difficult Colombia, and Mexico—beneficiaries have to build evidence on the effectiveness of the option of utilizing different channels for programs in fragile and conflict-affected complaints, both at the national and pro- countries. When they are available and show gram levels. Because such systems require positive results (as in West Bank and Gaza, dedicated administrative capacity, they tend Sierra Leone, and the Republic of Yemen), to be less available in fragile and conflict-af- they provide a strong basis for scaling up fected states. A few of these countries with and systematizing programming. long-standing programs are attempting to incorporate grievance mechanisms to the GRIEVANCE AND REDRESS MECHANISMS. These extent possible. In the Republic of Yemen, mechanisms seek to capture complaints the public works program under the Social related to program functions (e.g., tar- Welfare Fund offers a complaint box for any geting of program beneficiaries) and provide issues raised by program beneficiaries. The redress. Such mechanisms are available in issues are compiled and discussed, and mea- three forms: government agencies, indepen- sures are taken each month to respond to dent redress institutions, and the courts. In the beneficiaries. 5 HIGHLIGHTS OF PROGRAM DELIVERY T his section presents lessons learned tection programming in a number of fragile from selected policies and programs and conflict-affected countries, regardless of applied in various fragile and con- income or capacity levels. Often, these pro- flict-affected settings and illustrating grams are deemed critical for maintaining different social protection approaches. The social balance in the aftermath of a conflict, intent is to provide a sense of what is hap- and providing a dividend to those who took pening on the ground to bring to life the data part in liberation wars. Timor-Leste, an oil- presented in section 4. It is also intended to rich East Asian state, has such a program. provide an understanding of the challenges The Timorese fought a war of independence and issues confronting social protection in from Indonesia between 1975 and 1999. At fragile and conflict-affected settings. that time, the leaders of the movement promised the fighters jobs and a better life CATEGORICAL BENEFITS MAY BE CRITICAL FOR after independence. This commitment was SOCIAL COHESION AND POLITICAL STABILITY, also reflected in the constitution when the BUT ARE UNSUSTAINABLE. Veterans’ pension country was formed in 2002. However, after programs are a common part of social pro- independence, there was a feeling that these 31 32 Social Protection in Fragile and Conflict- Affected Countries: Trends and Challenges promises were not being fulfilled. Mass riots example of how—with moderate capacity in 2007 and an assassination attempt on the and an average enabling environment— prime minister in 2008 are largely attributed rapid scale-up is possible, provided there is to this sense of discontent. political will. In response, the government expanded its Sound technical assistance from the World social protection efforts. The increasing size Bank also helped improve targeting. Tech- of the Petroleum Fund played a role in this nical assistance in 2009 funded by the Bank decision as well. Timor-Leste uses its natural introduced PMT as a targeting method and resource wealth for social protection funding. applied it to the 2008 Social Welfare Fund It spends 15 percent of non-oil GDP and 5.9 beneficiary and applicant survey. PMT helped percent of total GDP on social protection, a the country identify existing nonpoor bene- majority of which goes to veterans’ pensions. ficiaries and new poor beneficiaries, thereby The current level of funding appears finan- improving targeting; although there are still cially unstable. A 2011 fiscal sustainability problems with inclusion errors. analysis undertaken by the National Direc- torate of Budget of the Ministry of Finance PROGRAMS CAN BE SUCCESSFUL IF BASED ON shows that the fund will reach zero in 2026 PRE-EXISTING SOCIAL INSTITUTIONS. Togo has if growth in expenditure is not constrained used an existing informal social mechanism for and non-oil economic growth is below its school feeding program, which has proven expectations. Out of the total social assis- to be successful in a low-capacity setting and tance spending in Timor-Leste, 60 percent has reached a large number of beneficiaries. goes to veterans’ benefits, covering only 1 The program benefits approximately 20,000 percent of the population. Despite concerns children annually in the most deprived, geo- regarding financial sustainability, veterans’ graphically isolated, and disaster-prone areas pensions are used to contain a volatile group of the country. Its aim is to provide adequate and maintain political stability. calories to children vulnerable to malnourish- ment. The meals are prepared by selected vil- RAPID SCALE-UP IS POSSIBLE IF THERE IS POLIT- lage women, or femmes-mamans (Andrews et ICAL WILL, RELATIVE CAPACITY, AND FINANCING. al. 2011). These femmes-mamans are a familiar The Republic of Yemen has been very suc- feature in Togolese villages preparing and cessful in rapidly scaling up its Social Welfare selling food in the market or on the street. By Fund, which today provides cash benefits to employing these women to cook the school close to 7 million people (28 percent of the meals in the beneficiary schools, the program population) in all 21 governorates. The fund also provides income-generation opportuni- has expanded its coverage from 100,000 ben- ties for the poor. eficiaries at its start, to almost 1 million poor and vulnerable Yemeni households over a The program has made an important contri- 10-year period; it expanded to 1.5 million bution toward attracting and retaining ben- households during the political crisis in 2011. eficiary children in school, providing school The fund’s budget has grown from $4 million access to children who are older and have not at the outset to $200 million in 2008/09 and yet enrolled, and—increasingly—attracting around $300 million in 2012. This is a clear young girls. Results show increases in enrol- 5. Highlights of Program Delivery 33 ment, decreases in dropout and absentee errors are lower than other programs that rates, and a reduction of the age at entry are widely considered to be successful (e.g., in primary school in all regions; the findings Bolsa Familia in Brazil and Oportunidades in have been particularly positive among girls. Mexico). As such, the Cash Transfer Program is An increase in new enrolments in beneficiary highly efficient, with cost-benefit ratio anal- schools in 2009–10 was 16 points higher than yses showing that for each unit of currency in the group of control schools. The dropout spent on transfers by the Ministry of Social rate was 0.9 percent in beneficiary schools Affairs, 0.66 units goes toward reducing compared to 1.4 percent in control schools. the extreme poverty gap. Many partners of The retention rate expressed in terms of per- social protection programs in West Bank and centage of children attending school every Gaza—including the United Nations Relief day is 2 percent higher than that of the con- and Works Agency for Palestine Refugees in trol schools. Further, the meals served pro- the Near East, the World Food Programme, vide necessary calories to the children. A and UNICEF—have had access to, and many nutritional assessment of the food served at rely on, the PMT database managed by the beneficiary schools shows that school meals Ministry of Social Affairs for designing and are providing between 60 and 90 percent of targeting their own programs. the daily caloric intake needed for primary school–age children. The West Bank and Gaza Cash Transfer Pro- gram benefits close to 100,000 households PROXY MEANS TESTING CAN IMPROVE TAR- and 600,000 individuals (about 14 percent of GETING EFFICIENCIES OF EXISTING SOCIAL the total population). It is considered one of PROTECTION PROGRAMS. The West Bank and the most advanced cash assistance programs Gaza’s dramatic shift in social safety net in the Middle East and North Africa region, policy in 2004 emphasized the need to pro- using a sophisticated management informa- vide assistance to extremely poor households tion system and a uniform payment modality. and the importance of using a PMT mecha- Moreover, West Bank and Gaza provides a nism to verify the eligibility of beneficiaries. best practice example regarding the creation With the help of the World Bank, the Cash and use of a unified registry of beneficia- Transfer Program in West Bank and Gaza has ries across social safety net programs which been able to effectively—and in a relatively has significantly improved targeting accuracy short time—develop, implement, and secure and crisis response capacity. In normal times, donor and Palestinian Authority buy-in for unified registries can reduce costs and facil- an effective and objective poverty-based tar- itate coherence and convergence because geting system. The system accurately iden- all agents work with the same database. In tifies and provides regular and timely cash times of crisis, unified registries can be used transfers to poor households, using benefit to quickly disburse additional benefits to the levels that are tailored to household com- target population or quickly expand coverage position and poverty levels. The use of PMT by adjusting eligibility criteria. has been found to be highly accurate, having identified almost 70 percent of target cases Note that PMT is not used in postconflict correctly; a large majority of the beneficiaries or emergency situations, as in such cases, are extremely poor. The inclusion/exclusion the extreme poor are not the only targeted 34 Social Protection in Fragile and Conflict- Affected Countries: Trends and Challenges households and most PMT indicators (i.e., administrative capacity and number of qual- household assets) are difficult to calculate ified staff. Also, distance from conflict has and may have been affected by the conflict. been an important factor in building and maintaining delivery systems over time. It PMT AND UNIQUE IDENTIFIERS ARE EFFEC- would be difficult to administer PMT or TIVE TOOLS TO CURTAIL INCLUSION ERRORS. In establish a reliable registry in an ongoing fragile and conflict-affected countries with or recent conflict situation in which many high capacity and a strong enabling envi- households may be displaced or difficult to ronment such as Kosovo, best practice pro- locate or identify; or that may lack the assets cesses can be employed to reduce inclusion needed to build a PMT mechanism and mea- errors. The Kosovo Social Assistance Pro- sure relative welfare. gram covers 11 percent of the total popula- tion and 13 percent of all children; coverage MOBILE TECHNOLOGY CAN OVERCOME IMPLE- rates are higher for the poorest quintile (30 MENTATION CHALLENGES AND IMPROVE COM- percent) and 35 percent of the poorest chil- MUNITY-BASED PROGRAMS IN LOW-CAPACITY dren (Gassmann and Roelen 2009). Using a SETTINGS. Sierra Leone is characterized as combination of eligibility criteria and proxy having low capacity and a strong enabling means and means testing to target benefi- environment, with resource wealth. Despite ciaries, Kosovo has only a very small inclu- its low administrative capacity, the country sion error in the program. The majority of provides an example of how mobile tech- benefits are awarded to families in the two nology can help improve community-based poorest quintiles, and only a small share of programs and make implementation pro- benefits is awarded to nonpoor families. cesses more efficient. Over 70 percent of spending on the social assistance scheme went to households in the Mobile technology is being used in the Sierra bottom quintile, while less than 3 percent Leone Youth Employment Support Project. in the top quintile benefited. This compares The project—which seeks to provide capacity favorably with regional averages of 62 and building to youth institutions, to finance 4 percent, respectively. The combination of policy studies and analysis, and to promote eligibility criteria, PMT (asset), and means the effective national coordination of all testing (income) works well in differenti- youth employment support initiatives in Sierra ating between the poor and the nonpoor. Leone—has benefited about 10,000 youths Moreover, the unique identification system each year over the last three years and has that has been put in place allows rapid and helped them become more employable and/ accurate verification of beneficiaries with or to transition into the labor market. Ben- little room for fraud and misuse of benefits. eficiaries of the Sierra Leone Youth Employ- ment Support Project have adapted easily to Certainly, the reasons for the relatively suc- the new mobile technology. As in many other cessful delivery of social assistance programs countries, e-payments have proved to work through PMT and a common registry are well, despite low capacity and a fragile set- the institutional legacy of social protection ting. Communities have played a pivotal role in Kosovo and the resulting relatively high in ensuring these successes. 5. Highlights of Program Delivery 35 Under the Youth Employment Support data and payment flows. Upon registration, Project, and its Cash for Work component in all beneficiary information is added to an particular, several implementing challenges electronic timesheet, wherein the payment were identified, including registry, payment amount is directly computed and beneficia- system, and monitoring and evaluation. In ries receive their SIM cards which are regis- tackling these implementation issues, the tered to be used for electronic payments. use of smart phones/mobile technology proved to be a very cost-efficient solution. SOCIAL PROTECTION CAN BE USED AS A PEACE- Smart phones were used to register a com- BUILDING TOOL. Following the end of its prehensive range of information/inputs. The armed conflict and the signing of a peace phones were also operated on- and offline agreement, Nepal has expanded its social and used to upload data in real time, pro- protection coverage considerably from 2006 vided the beneficiaries had a SIM card and on to forge social unity and ease tensions, network coverage. targeting social protection programs in par- ticular conflict-affected regions. Spending Given the low capacity and absence of effi- has increased from 0.5 percent of GDP in cient beneficiary targeting and registry 2004/05 to about 2.5 percent of GDP in 2012. mechanisms, there was a general lack of iden- Social protection investments have concen- tification documents, and the existing paper trated on various categorical cash transfer documentation suffered from errors and was programs, school feeding, and public works. difficult to access (Rosas and Martin 2014). Mobile technology was introduced in order The commitment of the Nepali state to to find a solution to the lack of documenta- social protection and social inclusion can tion. Staff members were quickly trained to be inferred from the interim constitution use smart phones to collect information on drafted in 2007. The document views employ- potential beneficiaries and to take photos ment and social security as fundamental for the beneficiary IDs. Each subproject rights of every citizen. The state especially registration with mobile technology lasted aims to ensure the socioeconomic security of one day. Thus far, more than 6,600 benefi- marginalized and vulnerable groups through ciaries have been registered in over 86 sub- such measures as child grants, scholarships, project sites. Where paper documentation and senior citizen and disability allowances. existed, smart phones were used to digitize The current programs aim to support vulner- the information, which resulted in a digital able groups including the elderly, women, beneficiary database. The database allows and children, especially in remote parts of for enhanced coordination among different Nepal, as a way of reducing structural ineq- social protection players and institutions, by uities in the country. allowing for data sharing and comparison of information (Rosas and Martin 2014). Despite the numerous programs and poli- cies in place, the breadth and depth of social The use of mobile technology in improving protection remains quite low. Existing social beneficiary registration resulted in a better assistance programs reach less than a quarter payment system as well, through better of the most vulnerable populations; and, in 36 Social Protection in Fragile and Conflict- Affected Countries: Trends and Challenges particular, cash transfer programs do not The majority of the country’s social protec- appear to be as efficient or effective as they tion schemes are in the fiscal budget and are could be. Total coverage of the cash transfer funded through tax revenues (rather than programs does not exceed 3 percent of the through external aid). They could thus be Nepalese population, while poverty rates are interpreted as elements in a nascent social 31 percent nationally. contract between citizens and the state. The government has demonstrated a greater In the context of transition and peace commitment to social protection than many building, the visibility of the state in social other, richer, surrounding countries—an protection policy and programming in Nepal attempt at addressing socioeconomic secu- is something needs to be viewed positively. rity in a systemic manner. 6 CONCLUSIONS SOCIAL PROTECTION PLAYS A SIGNIFICANT social inclusion, and more equitable access ROLE IN FRAGILE AND CONFLICT-AFFECTED SET- to services and benefits. Nepal has instituted TINGS IN MAINTAINING OR REGAINING SOCIAL several categorical programs that increase BALANCE, BUT THE COST CAN BE SIGNIFICANT. coverage dramatically, even though they do Although there is weak empirical evidence not always benefit the poorest. on the impact of social protection in pro- moting and improving social cohesion, mea- HISTORY, SOCIAL DYNAMICS, AND IN-PLACE sures such as subsidies and categorical cash INSTITUTIONAL STRUCTURES PREDETERMINE benefits are widely used in fragile and con- SOCIAL PROTECTION POLICY AND PROGRAM- flict-affected countries to ease political and MING CHOICES. Countries often make policy social tensions and as rewards to certain and programming choices as a response to population groups following an episode of the situation on the ground. Social protec- conflict. In some countries, such as Mozam- tion objectives and interventions are influ- bique and Rwanda, qualitative studies and enced by a country’s social needs and the anecdotal evidence point to social protec- relationship it has had with its citizenry (i.e., tion including measures that promote voice, in Bosnia and Herzegovina and Kosovo, 37 38 Social Protection in Fragile and Conflict- Affected Countries: Trends and Challenges long-standing programs, particularly social FINANCING SOURCES ARE CRITICAL DETERMI- pensions, have created strong ties with the NANTS FOR THE DIRECTION OF SOCIAL PROTEC- citizens. Similarly, in resource-rich countries, TION POLICY AND PROGRAMMING IN FRAGILE subsidies have helped shape relationships AND CONFLICT-AFFECTED SETTINGS. A large and set long-term trends. To a large extent, majority of fragile and conflict-affected existing institutions, informal networks, countries are low income and resource poor, and cultural understanding of government and face serious budget constraints for social responsibility to social welfare determine the programming. External financing is a major social protection agenda. source for social protection expenditures for the medium term. In many cases, external PROGRAMMING CHOICES AND BALANCE BETWEEN financing steers and shapes the medium- SOCIAL ASSISTANCE, SOCIAL INSURANCE, AND to long-term policy discourse. Although LABOR VARY WIDELY IN FRAGILE AND CON- with different income and capacity levels, FLICT-AFFECTED COUNTRIES, AND DEPEND ON THE Kosovo, Sierra Leone, West Bank and Gaza, CAPACITY, INCOME, AND EXTENT OF AN ENABLING and the Republic of Yemen have all forged, ENVIRONMENT IN THE COUNTRY. Level of admin- to varying degrees, a national social pro- istrative capacity, together with the relative tection vision with strong technical support strength of institutions and rule of law in the and financing from external partners. All country, determines the type of interventions of them have reformed, rationalized, and that are possible to implement in fragile and scaled up their social protection program- conflict-affected countries. Low-capacity ming following emergency situations and countries with weak enabling environments long-standing conflict. have a concentration of emergency-type and fragmented policies and programs with low POLITICAL LEADERSHIP IS A CRITICAL DETERMI- coverage and coherence, which are largely NANT OF THE DIRECTION OF SOCIAL PROTEC- in kind. High-capacity countries with strong TION POLICY AND PROGRAMMING IN FRAGILE or moderately strong enabling environments AND CONFLICT-AFFECTED SETTINGS. Reforms have been able to establish longer-term and rapid scale-up in program coverage in social assistance, social insurance, and some fragile and conflict-affected settings have labor programs. Additionally, such countries been possible with enabling institutional are working on first- and second-genera- environments. More important, however, is tion design and implementation issues such political leadership that is willing to bear the as increasing coverage, reducing targeting financial, social, and institutional costs and errors, and establishing more effective man- to mobilize public opinion. West Bank and agement information systems, as well as ratio- Gaza and the Republic of Yemen have moved nalizing policies and programs. Frequently in forward in reforming their targeting mech- fragile and conflict-affected countries, the anisms and cash transfer programs, while populist measures undertaken to support the Bosnia and Herzegovina and Timor-Leste poor and vulnerable tend to have regressive have resisted rethinking categorical bene- outcomes and to suffer from major errors of fits for war veterans and their families—pro- exclusion (e.g., categorical benefits, subsidies). grams that are highly regressive. 6. Conclusions 39 DESIGN AND IMPLEMENTATION CHOICES, AND mechanisms, an institutional enabling envi- PARTICULARLY THE DELIVERY OF SOCIAL PRO- ronment and modest administrative capacity TECTION PROGRAMS, VARY WIDELY DEPENDING are necessary. ON COUNTRY CAPACITY AND THE ENABLING ENVIRONMENT. It is possible to use modern SOCIAL PROTECTION IN FRAGILE AND CON- technology successfully (e.g., for payments) FLICT-AFFECTED SETTINGS IS DYNAMIC, AND even in very high-risk and low-capacity envi- ITS COMPOSITION SHIFTS WITH CHANGES IN ronments such as Haiti, Somalia, and South COUNTRY CIRCUMSTANCES. As emergencies Sudan. Such interventions remain small scale, occur and recur, or as countries stabilize, however, and scale-up is difficult in the the composition of social protection shifts absence of an enabling institutional environ- to meet needs. For example, periodic surges ment. To put long-term delivery platforms in in in-kind assistance are common as scale-up place, such as identification and targeting of cash transfer programs. BIBLIOGRAPHY Acemoglu, Darren, and James Robinson. 2012. Bagash, Thabet, Paola Pereznieto, and Khalid Why Nations Fail: The Origins of Power, Pros- Dubai. 2012. “Transforming Cash Transfers: perity, and Poverty. New York: Crown Pub- Beneficiary and Community Perspectives on lishers. the Social Welfare Fund in Yemen.” Project report, Overseas Development Institute, Andrews, Colin, Maitreyi Das, John Elder, Mirey London. Ovadiya, and Giuseppe Zampaglione. 2012. “Social Protection in Low Income Countries Barma, Naazneen H., Elisabeth Huybens, and and Fragile Situations: Challenges and Future Lorena Viñuela, eds. Institutions Taking Root: Directions.” Social Protection and Labor Discus- Building State Capacity in Challenging Con- sion Paper 1209, World Bank, Washington, DC. texts. Washington, DC: World Bank. Andrews, Colin, Elena Galliano, Carolyn Turk, Braithwaite, Jeanine, and Daniel Mont. 2008. and Giuseppe Zampaglione. 2011. “Social “Disability and Poverty: A Survey of World Safety Nets in Fragile States: A Communi- Bank Poverty Assessments and Implications.” ty-Based School Feeding Program in Togo.” SP Discussion Paper 0805, World Bank, Wash- SP Discussion Paper 1117, World Bank, Wash- ington, DC. ington, DC. Breisinger, Clemens, Wilfried Engelke, and Andrews, Colin, and Adea Kryeziu. 2013. Olivier Ecker. 2011. “Petroleum Subsidies in “Public Works and the Jobs Agenda: Pathways Yemen: Leveraging Reform for Development.” for Social Cohesion?” Background paper for IFPRI Discussion Paper 01071, International World Development Report 2013, World Bank, Food Policy Research Institute, Washington, Washington, DC. DC. Asian Development Bank. 2012. “Nepal: Brisson, Nathalie Lamaute. 2013. “Systèmes de Updating and Improving the Social Protec- protection sociale en Amérique latine et dans tion Index.” Technical Assistance Consultant’s les Caraïbes: Haïti.” Project document, Eco- Report, Asian Development Bank. nomic Commission for Latin America and the Caribbean, Santiago. Assaf, Nabila. 2013. “Stimulating Business and Employment in Yemen.” MENA Knowledge Camnahas, Lourenco. 2012. “Democratic and Learning Quick Note 91, World Bank, Republic of Timor-Leste: Updating and Washington, DC. Improving the Social Protection Index.” Tech- nical Assistance Consultant’s Report for TA-REG 7601, Asian Development Bank. 41 42 Social Protection in Fragile and Conflict- Affected Countries: Trends and Challenges Chandy, Laurence, and Geoffrey Gertz. 2011. European Commission. 2008. “Social Protection “Poverty in Numbers: The Changing State of and Social Inclusion in Kosovo: Under UN Secu- Global Poverty from 2005 to 2015.” Policy Brief rity Council Resolution 1244—Executive Sum- 2011-01, Brookings Institution, Washington, mary.” European Communities, Brusells. DC. Gassmann, Franziska, and Keetie Roelen. Cho, Yoonyoung. 2011. “Informality and Pro- 2009. “Impact of Social Assistance Cash Ben- tection from Health Shocks: Lessons from efit Scheme on Children in Kosovo—Report to Yemen.” Policy Research Working Paper 5746, UNICEF Kosovo.” Maastricht Graduate School World Bank, Washington, DC. of Governance, Maastricht, the Netherlands. Commission of the European Communities. Gouvernement de la République d’Haïti. 2014. 2009. “Kosovo Under UNSCR 1244/99 2009 “Plan d’action pour l’accélération de la réduc- Progress Report.” Commission Staff Working tion de la pauvreté extrême.” Port-au-Prince, Document, Commission of the European Com- Haiti. munities, Brussels. Gubbels, John, David Snelbecker, and Lena de Regt, Jacomina, Shruti Majumdar, and Zezulin. 2007. “The Kosovo Pension Reform: Janmejay Singh. 2013. Designing Communi- Achievements and Lessons.” SP Discussion ty-Driven Development Operations in Fragile Paper 0707, World Bank, Washington, DC. and Conflict-Affected Situations: Lessons from a Stocktaking. Washington, DC: World Bank. Hackstein, Katharina, Candace Miller, and Esam Mahdi. 2008. “Impact Evaluation of Devereux, S. 2000. “Social Safety Nets for Pov- the EU Contribution to the Palestinian Cash erty Alleviation in Southern Africa.” ESCOR Transfer Programme (CTP) in the Framework of Report R7017, Department for International the PEGASE Programme of Support to “Vulner- Development, London. able Palestinian Families (VPF).” Final Report Volume 1. European Union. Elgazzar, Heba, Firas Raad, Chokri Arfa, Awad Mataria, Nisreen Salti, Jad Chaaban, Djavad Hamad, Bassam Abu, and Sara Pavanello. 2012. Salehi-Isfahani, Sanaz Fesharaki, and Mehdi “Transforming Cash Transfers: Beneficiary and Majbouri. 2010. “Who Pays? Out of Pocket community perspectives of the Palestinian Health Spending and Equity Implications in the National Cash Transfer Programme. Part 1: The Middle East and North Africa.” HNP Discussion Case of the Gaza Strip.” Project report, Over- Paper, World Bank, Washington, DC. seas Development Institute, London. Elgazzar, Heba. 2011. “Raising Returns: The Dis- High Level Panel of Experts on Food Security tribution of Health Financing and Outcomes and Nutrition. 2012. Social Protection for Food in Yemen.” HNP Discussion Paper, World Bank, Security. Rome: Food and Agriculture Organi- Washington, DC. zation of the United Nations. Engelke, Wilfried. 2012. “Joint Social and Eco- Hillis, Samira, Phillippe Leite, Laura McDonald, nomic Assessment Republic of Yemen.” MENA and Lillian Frost. 20133. “Designing an Knowledge and Learning Quick Note 70, World Advanced Cash Transfer Program in the Pal- Bank, Washington, DC. estinian Territories.” MENA Knowledge and Learning Quick Note 98, World Bank, Wash- Engineering and Management Consulting ington, DC. Center. 2012. “Assessment of Mosa Targeting System under the Cash Transfer Program CTP Holmes, Rebecca, and Shizu Aphadiya. 2009. (P129124): Gaza Strip.” “The Role of Cash Transfers in Post-Conflict Nepal.” Cash Transfers Series, Overseas Devel- opment Institute, London. Bibliography 43 Holmes, Rebecca, and Adam Jackson. 2007. Koehler, Gabriele. 2011. “Social Protection and “Cash Transfers in Sierra Leone: Appro- Socioeconomic Security in Nepal.” IDS Working priate, Affordable and Feasible?” Cash Trans- Paper 370, Institute of Development Studies, fers Series, Overseas Development Institute, Brighton, UK. London. Lombardo, Annalisa. 2012. “Mapping of Social Honorati, Maddalena and Rodriguez, Luz. Protection Programmes in Haiti.” UNICEF/ Forthcoming. “Social Protection Assessment Oxfam, UK. Review.” World Bank, Washington, DC. Ministerio da Planificacao e Desenvolvimento Instituto Nacional de Accao Social. 2013. “Plano de Mocambique. 2010. “Programa Estrategico Economico e Social 2013.” Mocambique. de Reduccao da Pobreza Urbana.” International Crisis Group. 2011. “Timor-Les- Ministry of Social Affairs, Palestinian National te's Veterans: An Unfinished Struggle.” Asia Authority. 2010. “The Palestinian National Briefing 129, International Crisis Group, Brus- Program for Social Protection Cash Transfer sels. Strategy.” Palestinian National Authority. International Labour Organization. 2004. —. 2010. “Social Protection Sector “Social Protection for People in the Informal Strategy.” Palestinian National Authority. Economy of Nepal.” International Labour Office, Kathmandu. —. 2011. “Business Strategy.” Palestinian National Authority. IMF (International Monetary Fund). 2009. “IMF Staff Visit to Kosovo: Concluding Statement.” Monchuk, Victoria. 2013. Reducing Poverty and IMF, Washington, DC. Investing in People: The New Role of Safety Nets in Africa. Washington, DC: World Bank. —. 2011a. “Republic of Kosovo: Staff Report for the First Assessment Under the Mountfield, Ben. 2012. “Cash Voucher Pro- Staff-Monitoring Program.” IMF, Washington, gramme: Review of Voucher Assistance as DC. a Safety-Net Transfer Modality in the Gaza Strip.” Study conducted for Oxfam GB and UN —. 2011b. “Republic of Mozambique: Pov- World Food Programme. erty Reduction Strategy Paper.” IMF, Wash- ington, DC. OECD (Organisation for Economic Co-opera- tion and Development). 2008a. “Concepts and —. 2011c. “Statement by IMF Mission to Dilemmas of State Building in Fragile Situa- Kosovo.” IMF, Washington, DC. tions: From Fragility to Resilience.” OECD-DAC Discussion Paper, OECD, Paris. Jones, Nicola, and Mohammed Shaheen. 2012. “Transforming Cash Transfers: Beneficiary and —. 2008b. “Service Delivery in Fragile Sit- community perspectives of the Palestinian uations: Key Concepts, Findings and Lessons.” National Cash Transfer Programme. Part 2: The OECD-DAC Discussion Paper, OECD, Paris. Case of the West Bank.” Project report, Over- seas Development Institute, London. —. 2013. Fragile States 2013: Resource Flows and Trends in a Shifting World. Paris: Khanal, Dilli Raj. 2012. “Social Protection/Social OECD. Security in Nepal.” Situational analysis, Interna- tional Labour Organization. —. 2014. Fragile States 2014: Domestic Revenue Mobilization in Fragile States. Paris: OECD. 44 Social Protection in Fragile and Conflict- Affected Countries: Trends and Challenges Ratha, Dilip. 2013. “The Impact of Remittances United Nations Relief and Works Agency for on Economic Growth and Poverty Reduction.” Palestine Refugees. “UNRWA Medium Term Policy Brief 8, World Bank, Washington, DC. Strategy 2010–2015.” República de Moçambique. 2010. “Estrategia Upreti, Bishnu Raj, Sony KC, Richard Mallett, Nacional de Seguranca Social Basica 2010– and Babken Babajanian. 2012. “Livelihoods, 2014.” Ministerio da Mulher e Accao Social. Basic Services and Social Protection in Nepal.” Working Paper 7, Secure Livelihoods Research Rosas, Nina, and Andrea Martin. 2014. “Using Consortium, London. Smartphones to Improve Delivery of Safety Net Programs in Sierra Leone.” Social Protection World Bank. 2007. “Timor-Leste Youth Devel- Note, World Bank, Washington, DC. opment and the Labour Market.” World Bank, Washington, DC. Sawada, Naotaka, and Jian Zhang. 2012. “Promoting the Rural Farm and Non-Farm —. 2008a. “Food Price Crisis Response Businesses: Evidence from the Yemen Rural Trust Fund Project Paper on a Proposed Addi- Investment Climate.” World Bank, Washington, tional Financing Grant under the Global Food DC. Crisis Response Program in the Amount of US$7 Million to the Republic of Togo for the Schwartz, Anita, and Aylin Isik-Dikmelik. 2014. Community Development Project (CDP).” “Chapter 7: Kosovo Public Finance Review.” World Bank, Washington, DC. In “Republic of Kosovo Kosovo Public Finance Review: Fiscal Policies for a Young Nation.” —. 2008b. “Integrated Safeguards Data- Report ACS9351, World Bank, Washington, DC. sheet: Appraisal Stage.” World Bank, Wash- ington, DC. Shaipi, Kushtrim. 2005. “Report on the Present State and Future of Social Security in Kosovo.” —. 2008d. “International Development Kosovar Research and Documentation Insti- Association Emergency Project Paper for a Pro- tute. Pristina. posed Grant in the Amount of SDR 10.6 Million (US$17.2 Million Equivalent) to the Republic of Silva, J., V. Levin, and M. Morgandi. 2012. Togo for a Community Development Project.” “Inclusion and Resilience: The Way Forward for World Bank, Washington, DC. Social Safety Nets in the Middle East and North Africa.” MENA Development Report 72975, —. 2009. “Integrated Safeguards Data- World Bank, Washington, DC. sheet Appraisal Stage: Togo—Financial Sector and Governance Project.” World Bank, Wash- Social Security Administration. 2013a. “Social ington, DC. Security Programs Throughout the World: Africa, 2013.” SSA Publication No. 13-11803, —. 2010a. “Coping Strategies in Rural Social Security Administration, Washington, Yemen and Policy Implications.” Report 51927- DC. YE, World Bank, Washington, DC. —. 2013b. “Social Security Programs —. 2010b. “Project Paper on a Pro- Throughout the World: Asia and the Pacific, posed Additional Grant Under the Pilot Crisis 2012.” SSA Publication No. 13-11802, Social Response Window in the Amount of SDR 5.8 Security Administration, Washington, DC. Million (US$8.7 Million Equivalent) to the Republic of Togo for the Community Develop- Sumner, A. 2012. “Where Do the World’s Poor ment Project (CDP).” World Bank, Washington, Live? A New Update.” IDS Working Paper 393, DC. Institute of Development Studies, University of Sussex, Sussex, UK. Bibliography 45 —. 2010c. “US$11.0 Million Equivalent —. 2012b. “Implementation Completion and Grant in the Amount of SDR 6.2 Million and Results Report on a Grant in the Amount (US$9.0 Million Equivalent) to the Republic of of SDR 45 Million to the Republic of Yemen Sierra Leone for a Youth Employment Support for a Private Sector Growth and Social Protec- Project.” Report 54419-SL, World Bank, Wash- tion Policy Development Policy Grant.” Report ington, DC. ICR00002250, World Bank, Washington, DC. —. 2011a. “Can Social Safety Nets Help —. 2012c. “Implementation Status & Peace Building in Nepal? Results from a Social Results: Sierra Leone Youth Employment Sup- Protection Technical Assistance Program.” Pre- port (P121052).” Report ISR6586, World Bank, sentation, World Bank, Washington, DC. Washington, DC. —. 2011b. “Human Development—Social —. 2012d. “Kosovo: Country Partnership Protection Pilot Project. Project Information Strategy FYs12-15 Concept Note.” World Bank, Document (PID).” World Bank, Washington, Washington, DC. DC. —. 2012e. “Mozambique: Social Protection —. 2011c. “Palestinian National Plan Assessment.” World Bank, Washington, DC. 2011–2013. Social Protection Sector Strategic Plan Summary. Part 1: The Social Protection —. 2012f. “Project Appraisal Document System.” World Bank, Washington, DC. on a Proposed Grant in the Amount of SDR 9.2 Million (US$14 Million Equivalent) to the —. 2011d. “Project Appraisal Document Republic of Togo for a Community Develop- on a Proposed Grant in the Amount of US$10 ment and Safety Nets Project.” World Bank, Million to the Palestine Liberation Organiza- Washington, DC. tion (PLO) for the Benefit of the Palestinian Authority (PA) for the West Bank and Gaza —. 2012g. “Social Safety Net Reform Cash Transfer Project.” Report 59698-GZ, Project: Implementation Completion and World Bank, Washington, DC. Results Report (ICR).” Report ICR2488, World Bank, Washington, DC. —. 2011e. “Project Appraisal Document on a Rapid Social Response MDTF Grant.” World —. 2012h. “Togo: Towards a National Bank, Washington, DC. Social Protection Policy and Strategy.” World Bank, Washington, DC. —. 2011f. “Project Information Document: Appraisal Stage—TL Social Protection Adminis- —. 2012i. “West Bank and Gaza Targeting tration Project.” World Bank, Washington, DC. Assessment of the Cash Transfer Program.” MENA Report ACS890, World Bank, Wash- —. 2011g. “Project Information Docu- ington, DC. ment (PID) Concept Stage Togo Community Development and Safety Net Project (PDSplus) —. 2012j. “Yemen: Joint Social and Eco- (P127200).” World Bank, Washington, DC. nomic Assessment for the Republic of Yemen.” World Bank, Washington, DC. —. 2011h. World Development Report 2011: Conflict, Security, and Development. —. 2013a. “Aide-mémoire République Washington, DC: World Bank. du Togo Mission d’Appui à la mise en œuvre du Projet Secteur Financier et Gouvernance —. 2012a. Gender Equality and Develop- (PSFG).” World Bank, Washington, DC. ment. Washington, DC: World Bank. —. 2013b. “Country Partnership Strategy for the Democratic Republic of Timor-Leste for the Period FY2013-FY2017.” World Bank, Wash- ington, DC. 46 Social Protection in Fragile and Conflict- Affected Countries: Trends and Challenges —. 2013c. “Labor Market Issues in Timor- —. 2014c. “Poverty and Inclusion in Haiti: Leste: Current State, Prospects and Chal- Social Gains at Timid Pace.” Report, World lenges.” World Bank, Washington, DC. Bank, Washington, DC. —. 2013d. “Project Information Document —. 2014c. “Project Paper on a Proposed (PID) Additional Financing. Community Devel- Restructuring and Additional Credit in the opment and Safety Nets Project Additional Amount of ADR 7.9 Million (US$12.1 Million Financing.” World Bank, Washington, DC. Equivalent) to the Republic of Togo for the Community Development and Safety Nets —. 2013e. “Restructuring Paper on a Pro- Project.” World Bank, Washington, DC. posed Project Restructuring of the Commu- nity Development Project (p110943) IDA Grant —. 2014d. Seeing Is Believing: Poverty in H602-TG June 29, 2010 to the Republic of the Palestinian Territories. Washington, DC: Togo.” World Bank, Washington, DC. World Bank. —. 2013f. “Restructuring Paper on a Pro- —. 2014e. The State of Social Safety Nets posed Project Restructuring of Togo-Financial 2014. Washington, DC: World Bank. Sector and Governance Project Grant H454-TG Approval Date 03/21/2009 to the Republic of —. n.d. “Beyond ODA: Philanthropy, Togo.” World Bank, Washington, DC. Remittances, and Other Transfers.” Develop- ment Dialogue on Values and Ethics website, —. 2013g. “Sierra Leone Social Protection World Bank. Assessment.” Report 78195-SL, World Bank, Washington, DC. World Food Programme. 2014. “Secondary Impact of WFP Voucher Programme in Pales- —. 2013h. “Timor-Leste Social Assistance tine.” Public Expenditure and Program Performance Report.” Report 73484-TP, World Bank, Wash- Zarazua, Miguel Nino, Armando Barrientos, ington, DC. David Hulme, and Sam Hickey. 2010. “Social Protection in Sub-Saharan Africa: Will the —. 2014a. “Implementation Comple- Green Shoots Blossom?” Working Paper 116, tion and Results Report (IDA-H4090 IDA- Brooks World Poverty Institute, University of H6020TF-93124) on IDA Grants in the Amount Manchester, Manchester, UK. of SDR 16.4 Million (US$25.9 Million Equiva- lent) and on a Co-Financing Grant from the Zimmerman, Jamie M., and Kristy Bohling. Food Price Crisis Response Trust Fund in the 2013. “Helping Ti Manman Cheri in Haiti: Amount of US$7.0 Million to the Republic of Offering Mobile Money-Based Govern- Togo for a Community Development Project.” ment-to-Person Payments in Haiti.” Wash- World Bank, Washington, DC. ington, DC: Consultative Group to Assist the Poor (CGAP)/World Bank. —. 2014b. “International Development Association Project Appraisal Document on a Proposed Grant in the Amount of SDR 9.1 Mil- lion (US$14.0 Million equivalent) to the Repub- lique of Togo for a Maternal and Child Health and Nutrition Services Support Project.” World Bank, Washington, DC.