Report No. 39953-ECA From Social Funds to Local Governance and Social Inclusion Programs A Prospective Review From the ECA Region (In Two Volumes) Volume I: Main Report May 2007 Human Development Sector Unit Europe and Central Asia Region Document of the World Bank TABLE OF CONTENTS VOLUME I: MainReport Acknowledgments iii ExecutiveSummary iv Introduction 1 Chapter 1 The Natureand Relevanceof SocialFunds - 3 1.1. What I s a Social Fund? 1.2. Why I s a Social FundRelevant? Chapter2 Overview of SocialFundsinECA - 5 1.3. Context 1.4. Basic Facts about Social FundsinECA 1.5. Types of Social Funds and Trajectories Chapter3 OrganizationalStructureof ECA Social Funds - 13 1.6. Policy Autonomy 1.7. Structural Autonomy 1.8. Operational and Managerial Autonomy 1.9. Budgetary and Government Accountability 1.10. Composition and Sustainability of Social FundFinancing 1.11. Conclusions Chapter 4 LocalGovernanceandProvisionof Infrastructure - 22 1.12. Context 1.13. Rationales for SF Intervention 1.14. SFs as Instruments of Local Governance Innovation andReform 1.15. Implications for the Future Chapter 5 Vulnerability and Social Exclusion - 35 1.16. Introduction andContext 1.17. Rationale and Objectives 1.18. Lessons from ECA SP Experience of Social Care 1.19. Implications for the Future Chapter6 Options for FutureEngagementwith SocialFunds - 46 1.20. Are Social Funds Relevant for ECA? 1.21. FutureTrajectories for Local Governance Reform 1.22. Future Roles for SFs withinthe Social Inclusion and Protection Agenda 1.23. Issues Specific to EU Integration 1.24. Developing a Coherent Vision Bibliography 60 i VOLUME 11: TechnicalAnnexes Annex 1:Portfolio of SF/CDD ProjectsFundedby the WB -Financing Arrangements Annex 2: Project Development Objective of Social FundProjects fundedby the WB Annex 3: CDD and SF Annex 4: Summary ofAvailable Dataon SF Performance Annex 5: Quality ofthe Social Fund/CDD Portfolio inECA Annex 6: Description of Oversight Bodies of SF Annex 7. Bosnia andHerzegovina-Intensityof CDP Capital Funding Republika Serpska Annex 8: Description of Institutional Arrangements of ECA SFs InvolvedinLocal Infrastructure Annex 9:Informationon Impact and ProcessEvaluations on Social Funds inECA .. 11 ACKNOWLEDGMENTS This report was prepared by a team led by Rodrigo Serrano-Berthet (TTL, HDNSP) and composed by the following consultants: Jan Barrett, Mariana Felicio, Gagik Khachatryan, Ivana Rossi, Sandra Schnellert, and Julie Van Domelen. Written feedback on an earlier version of this report was provided by hush Bezhanyan (Social Funds Program Team Leader), Maniza Naqvi, Caroline Mascarell, Ana Maria Sandi, and John Innes, (ECSHD); Sandor Sipos (HDNSP); and Radhika Srinivasan and Mark Woodward (ECSSD). Louis Helling provided advice on study design. Peer reviewers included Professor Rob Laking (School of Government, Victoria University o f Wellington), KeithMcLean(ECSSD), Sandor Sipos (HDNSP), and Mark Woodward (ECSSD). ... 111 Executive Summarv Context and Objectivesof the Study I. The role andrelevance ofSocialFunddCommunity-Driven Development (SF/CDD) have been highly debated in the international development community. Some conceive these programs only as parallel and temporary arrangements that can ensure short-term delivery o f development benefits in contexts of dysfunctional public sectors. Others emphasize the flexibility o f the SF/CDD instrument in adopting different institutional forms depending on the country context, and their contributions to long-term development challenges. 11. Nevertheless, since the first ECA social fund was created in 1993 in Albania, governments and donors have actively supported Social Funds in the Region as instruments to deal with a range o f issues, from inadequate access to services, to local infrastructure provision, to social exclusion. The World Bank i s currently supporting SF operations in 14 ECA countries and preparing new operations infive more. 111. The circumstances under which most Social Funds were created have changed significantly. For example, there i s less conflict; institutions shaken up during transitions have become more stable and capacity has grown; and local elections have taken place in many countries. Many o f these changes have taken place in the context of either incorporation into the European Union, or the expectation (held by most ECA countries) o f future accession to the European Community. IV. The main objective o f this report is to provide guidance to ECA management and client governments on two questions: i. Is the social fund instrument still relevant in the ECA region? Under what conditions? ii. Whatarethefutureimplicationsintermsofcontinuation, transformation, andexit options under different country scenarios? V. The study gathered information from a universe o f 16 SFs. Main data sources included bank documents (Project Appraisal Documents [PADS], Implementation Completion Reports [ICRs], aide-memoires, mid-term reviews, evaluations); literature review; a questionnaire o f social funds; short field trips to several countries (Georgia, Bosnia and Herzegovina, Romania, Turkey, Ukraine); and interviews with Task Team Leaders (TTLs). VI. The study does not provide country-specific recommendations but rather looks at broad trends across the Region. iv Relevance and Typology of Social Funds VII. This i s the basic question that policy makers must answer to justify the relevance of an SF instrument: What in the country context or policy objective makes a discretionary fund for multi-sectoral, demand-driven, pro-poor investments in local infrastructure and services a more appropriate choice than using regular government systems programs? Here are three possible rationales: i. Stop-gap: Government structures and systems are dysfunctional and the Fund .. provides a temporary means o f channeling resources to the local level. 11. Compensatory: There are exceptional problems that the regular transfer systems or sectoral programming are not well designed to address, such as natural disasters, discrimination o f minorities, or deep pockets o f poverty. iii. Innovation and reform: SFs can promote innovation inservice delivery and in systems development. VIII. Assessing social funds requires some sort o f typology that classifies the policy objectives to which they correspond. Based on the experiences o f SFs in ECA, it is possible to distinguish two overarching policy orientations: i. Improving local service provision for the poor through increased access to infrastructure. SF intervention i sjustified inresponse to shortcomings inthe regular system for local capital investment programming. This type of SF intervention should be discussed in the context o f a diagnosis o f the country's decentralization .. framework andpolicy. 11. Protecting poor and vulnerable groups from shocks and promoting their social inclusion. SF intervention i s justified in response to exceptional circumstances o f deprivation. The SF intervention should be discussed in the context o f a country's strategy for social protection, which identifies vulnerabilities and social risk management strategies to addressthem. A second distinction within each policy orientation relates to its approach to local institutional development-Le., the importance assigned to developing and strengthening inclusive, accountable, and/or cohesive local institutions, which can range from limitedto substantial or significant. IX. Combining these approaches and orientations yields four types of SFs: i. InfrastructureManagementFunds,whosedistinctivefunctionistoactascontract managers of small-scale infrastructureprojects. ii. LocalGovernanceFunds,whichseekto developthecapacity oflocalinstitutions (Community-Based Organizations [CBOs] and Local Governments [LG]) to manage local service provision ina participatory, accountable, and effective way. iii. Emerrrencv Fundsas a copingmechanismto post-conflict environments, natural disasters, and economic shocks. iv. Social Inclusion Funds, which deal with social exclusion and marginalization caused either by discrimination o f ethnic minorities, or by membership in a vulnerable group. ECA SFs are mapped into these four types o f SFs (see figure). V SigniJicant Moderate Local institutional development Limited OrganizationalStructureof ECA SocialFunds X. Looking at institutionalarrangementsamong the ECA SFs reveals that the stereotype of SFs as enjoying full autonomy is not accurate. In general, there i s more autonomy in operational and managerial procedures than in budgetary and accountability issues, but this may be a result of being donor-financed, as many of the same arrangements are found in Project Implementation Units (PIUs). There is more structural autonomy than policy autonomy because most SFs are either under a ministry or governed by ministry representatives through the Board of Directors. XI, There is not one blueprint or public management model for SFs in ECA, with some closer to fully autonomous agencies and others more integrated into line ministries. For every issue analyzed, there i s variation across countries. While the semi-autonomous organizational structure of an SF has tendedto be one of its comparative advantages, it is possible to have a fully integrated SF. Comparing across SFs, Social Inclusion Funds have tended be more integrated at the policy level and less autonomous at the structural and operational levelthan Local Infrastructure Funds. XII. To align SFs autonomy arrangements with their specific typology calls for greater integrationfor the local governance and social inclusion funds (see figure below). vi SFAutonomy ArrangementsUnderDifferentPolicy Objectives Sign8cant Local institutional development Limited Nature of intervention Local Governance and Infrastructure Funds XIII. The design o f most social funds gives primacy to infrastructure, even though their justification contains strong elements o f a reforming rationale. These are social funds that started under the "stop-gap" rationale and gradually included elements o f local institutional development into their design. As they did, the "local governance reform" rationale crept in; however, the fund still maintained its original logic o f being an implementing agency that actively manages projects. Most funds retained the semi- autonomous agency form, with a project implementation rather than policy-oriented mandate. While they have gradually devolved greater responsibilities to the local level, these agencies have still retained substantial control over basic decision- making processes. XIV. All infrastructure and local governance SFs inECA have contributed to more responsive investment allocation processes by providing relatively "untied" investment resources. Social funds have introduced new local institutions for LG-CBO interfaces, but have been limited because o f their project nature. While some social funds have allowed communities to manage resources and procure goods and services, and a few have granted procurement powers to local governments, none has transferred resources directly to be managed by local governments. Although these practices have increased local expenditure responsiveness, a more strategic design would have a more lasting impact. Even more, it would allow tapping into the many unexploited opportunities that vii exist to advance local governments and local service delivery units' "downward accountability" to citizens and communities. I organizations Iemphasis on CBO emphasison LGI xv. Although Local Infrastructure SFs have played a positive role in the transition period, most of them need to reform their institutional setting to be mainstreamed into a country's decentralization strategy, or exit. However, there is a risk that in the movement towards mainstreaming, the value-add of citizen voice and control o f local governments, which is the comparative advantage of SF/CDD, gets lost. This is what seems to have happened in Georgia, where the Georgian Social Investment Fund (GSIF) was subsumed under the Municipal Development Fund (MDF). Management should learn from this and other examples of social fund transitions inother regions. Vulnerability and Social Exclusion XVI. A growing number of ECA social funds have as their core mission protecting poor and vulnerable groups from shocks, marginalization, and social exclusion. ECA SFs that primarily address vulnerability and social inclusionhave developedfor a variety of reasons, including weak institutional capacity; exceptional problems like social exclusion of particular groups such as people with disabilities or minorities; de-institutionalization; deep pockets o f poverty; and problems that cut across institutional boundaries, such as homelessnessor drug abuse; and the need to promote innovation inthe models, range, and organization of social care service delivery, particularly in countries emerging from the former Soviet Union (FSU) experience. These can be divided between Emergency SFs created in fragile, often post-conflict environments (originally Bosnia and Herzegovina, Albania, Armenia, Tajikistan, and Kosovo) and Social Inclusion SFs created in order to provide innovative solutions to social protection issues (Slovakia, Bulgaria, Romania, Serbia, and Turkey). XVII. SFs that played a critical role in addressingemergency post-conflictand transition needs were often part of the initial Bank entry into FSU countries and were able to effectively work in weak and chaotic institutional environments. In addition, there have viii been situations where the rapid response capability o f SFs was used to respond to natural disasters, as in the case of the Armenia Social Investment Fund (SIF) earthquake reconstruction efforts. In general, emergency response is no longer the relevant institutional goal o f ECA SFs. Most, if not all, SFs have evolved away from crisis. XVIII. Social inclusion SFs, have three main areas of emphasis: (i) care services, (ii) social social inclusion o f marginal groups, and (iii)employment and training: 1. Social care has beenby far the most significant area o f SF involvement inthe social protection agenda. The main policy objective o f SFs involved in social care has been to support the development o f a community-based system o f social care services. SFs have shown the flexibility to act as niche players targeting groups and regions that mainstream social care services find difficult to reach. SFs have promoted local government/community partnerships and service user involvement. 11. e . In contrast to the former centrally planned, universal service perspective characteristic o f the Communist system, SFs have introduced innovative strategies to address the needs o f marginalized populations. SF operations directed towards other marginalized groups such as older people and people with disabilities have so far tended to follow the more traditional approach of increasing access to social ... assistance services. 111. Although job creation i s frequently cited as one o f the benefits that accrue from the SF, it i s generally a secondary benefit o f improving local infrastructure. Beyond temporary employment, several E C A SFs have as one of their core objectives to improve employment opportunities via training, information, and placement services, or via their access to income generation activities. Optionsfor FutureEngagementwith Social Funds XIX. For the future, there are really only two overarching and distinct policy objectives that social funds inECA may pursue: i. improvinglocalgovernanceandinfrastructureprovision,and ii. addressingvulnerabilityandsocialinclusion. XX. A social fund can be a relevant instrument within any o f these two policy domains as long as its design fits within one o f the following two rationales, and the rationales fit the country context, namely: ..i. introducing reforms and innovations for local service provision, and 11. providing incentives to compensate for shortcomings inthe regular system o f public provision, including equity objectives. XXI . The strategic objective for social funds trying to improve local governance i s to generate incremental changes that support the shift towards responsive and accountable local governments, community-based organizations that provide opportunities for citizens to identify and address community concerns, and an enabling environment that facilitates relations o f co-production and mutual accountability between LG-CBOs. XXII. It is possible to identify a few entry points that would be appropriate depending on country context, including ix 0 pilotingperformance-basedgrants to complement the fiscal transfer system, 0 compensatingfor unfundedmandates, 0 linking local governments andcommunity groups, 0 channeling co-financing to link municipalities and/ or national sectoral objectives, 0 exitingto a municipal development fund, and/or mainstreaming into line ministry support to local governments. XXIII. The same rationale of supporting innovations and reforms, or compensating for inadequateprovisionof services, can be applied to social inclusion policies. Potential key roles for SFs inthis context: Increasing specialization as niche players, focusing on groups that are not well served by mainstream provision, or as mechanisms for targeting localities/regions that are lagging behind. Inemployment and training, using the comparative advantageof the SF's capacity to incorporate employment and training initiatives targeted at disadvantaged and marginalized groups withinbroaderprograms of support. Addressing social exclusion and the needs of marginalized households and communities with mechanisms designed to support a cross-cutting approach. SFs are particularly well placed to fulfill this role, especially in environments with limited experience of multi-agency cooperation and limitedNGO capacity. Serving as appropriate vehicles for experimenting with broader innovations, like conditional cash transfers. In social care, SFs may have an important role to play in developing innovative approaches, so long as they are envisioned as core programs and not microprojects. Innovations in social care must extend beyond social assistance to include access to labor markets, healthcare, training, and public services. The flexibility and multi-sectoral experience of the ECA SFs can help create integrated responses to social care needs of a different nature than the multi-sectoral project typologies of the previous generation of SFs. XXIV. SFs should be advertised more strongly as instruments that can facilitate the objectives of the EU accession process. ECA countries that are joining or hope to join the European Union must bring their policy frameworks into alignment with EU goals. The EU accession process represents an important opportunity to mainstream social funds expertise into national systems for local governance and social inclusion, and for the EUprocessto benefit from SF experience. The approach and methods usedby social funds inECA are highly compatible with the ones favored by the EU, for example, local partnerships, participatory development, a competitive process for allocating investments, and coordinating and channeling external assistance. While the operational route will be different depending on whether the objective is local infrastructure or social inclusion, there exists some common advice for taking advantage o fthe SF instrument: i. Forpre-accessioncountriesthathaveanSF,itiscriticalthatSFsstartparticipating as early as possible as a pre-accession instrument (previous engagement with the EUis akey determinantofthe role that agenciesplay during post-accession). ii. SFs should try to aligntheir menuof investments with the menuof activities eligible under EU regulations (low absorption of EU funding in the early years of post-accessioni s mainly due to the lack of systems capable of financing EU-eligible activities). X iii. Inpostaccessioncountries,ifSFsbecomepartoftheEUformalarchitecture,they must choose betweentheir role as a grant-making body with project approval power and their role as facilitator of local capacity building for presenting projects (EU rules state that playingbothroles represents a conflict of interest). This problem can be avoided if SFs become country-owned instruments (not part of the EU formal architecture) through which governments implement specific measures within an operational program. For instance, inBulgaria the SF will likely be the recipient of a Global Grant to implement specific measures of the Human Resources Operational Program. xxv. As summarized inthe table below, there are multiplepossiblefuture paths.Transitions will needto be carefully thought-out and a consensus formed for eachcountry. There is a great risk of losing the accumulation of experience and innovations of SFs if a clear institutional transition path is not spelled out, especially in the cases where exit is the preferred option. The transitions must be made with a clear understanding of the overriding policy objectives and inlight of the current country context and opportunities. Cofinancing for inter- Pilotingperformance- Mainstreaming municipal cooperation to based grant models CDD/local governance address fiscal imbalances Building participatory approach in agency and equity mechanisms responsible for Financingwindow to absorb Awarding community capacity building/local EU funds partnership grants for governance agenda civic engagement ProvidingMunicipal Development Funds or Regional Development Financing window for Innovations for social Mainstreaming inline pockets o f poverty, inclusion ministry vulnerable groups Community-based Supporting local social care services governments to take on Integration o f multiple social inclusion and agencies (sectoral, social care objectives local government, NGO ) xi Introduction 1. The role and relevance of Social FunddCommunity-DrivenDevelopment(SFICDD) have been highly debated in the international development community.' Some conceive these programs only as parallel and temporary arrangements that can ensure short-term delivery o f development benefits in contexts o f dysfunctional public sectors. They note that maintaining SFs beyond those very specific settings inevitably means a lack o f sectoral coordination and policy coherence, weakening o f public sector institutions, and investment sustainability. Others emphasize the flexibility o f the S F K D D instrument in adopting different institutional forms depending on the country context, and their contributions to long-term development challenges-- e.g., improving local governance and service delivery by strengthening accountability and co- production relations between communities and local governments, targeting vulnerable and excluded groups, and/or incubating bottom up innovations. Objectiveof the Report 2. The mainobjectiveof this reportis to provide guidance to ECA management and client governments on two questions: i. IsthesocialfundinstrumentstillrelevantintheECAregion?Underwhatconditions? 11. What are the future implications in terms o f continuation, transformation, and exit a . options under different country scenarios? Scope, Data,and Structureof the Report 3. Scope. The aim o f this study is to provide guidance on the question o f social fund relevance. This i s not an impact evaluation o f SFs, although the report summarizes some of the available information about SF performance to give a sense o f their effectiveness. Nor is this a country-specific diagnostic that will provide country-level recommendations. Instead, this report assesses broad trends across the Region. Substantial effort was invested in developing an analytical framework to approach the study questions, including a typology o f social funds. 4. Data. The study gathered information from 16 SFs. Main data sources included bank documents (PADS, ICRs, aide-memoires, mid-term reviews, and evaluations); literature review; a questionnaire o f social funds; short field trips to several countries (Georgia, Bosnia and Herzegovina, Romania, Turkey, andUkraine); and interviews with TTLs. 5. Structure. The report is organized into six chapters and a set o f annexes. Chapter 1 defines social funds and their main rationales. Chapter 2 provides an overview o f their origins in ECA, basic facts about the Bank operations and SF performance, and develops a typology based on policy objectives. Chapter 3 summarizes the institutional arrangements o f social funds inthe Region and then reviews them within the wider vision o f optimal public sector arrangements. 1For a summary see Bhatia (2005). Other texts that touch on this debate are Rawlings et al. (2004); Mansuri and Rao (2004); WE3 (2002, 2004a, 2005a); Van Domelen (2000); Helling, Warren, and Serrano (2005); McLeanet al. (2006); Parkerand Serrano (2000); andTendler with Serrano (1999). 1 Chapter 4 looks at local infrastructure and governance funds, evaluating their design against a set of good practice benchmarks for promoting local governance, and drawing implications for the future. Chapter 5 conducts a similar exercise but for social inclusion funds. The final chapter summarizes the main answers to the study questions and elaborates a set of options for future engagement with social funds, taking into account different country contexts. Annexes provide more detailed backgroundmaterial. 2 Chanter 1 The Nature and Relevance of SFs - 6. The nature of the social fund instrumenthas led to a wide diversity of social funds across the world. This diversity has often clouded the debate about their relevance, since different versions o f social funds exist. Sometimes, what i s taken as the defining features of an SF are particularities o f a specific country version. Other times, there are features that are commonly but not always present in a social fund, but they are not essential features of the model and may reflect more the influences o f Bank financing. At still other times people claim that their program i s not an SF, when indeed the core essential features that define this instrument are the same. 1.1What I s a SocialFund? 7. A social fund is basically a combination of two "technologies": on the one hand, the SF is a bundle o f procedures and methods that enabled it to act as a local investment programming and financing facility; on the other hand, it i s an instrument to introduce a community-driven development (CDD) approach into the way local investments are allocated and managed. A CDD approach is basically an approach that tries to improve the well-being o f poor people b increasingtheir control over the way investment resources are planned, executed, and managed. Since communities do not act in isolation but ina local space where they interact Y with local governments, sectoral service delivery units, the private sector, and civil society, the CDD approach has lately beenexpanded into a broader local governance appr~ach.~ 8. As a local investment programming and financing facility, a social fund performs three functions: i. It channels grants to local projects and is entrusted with investmentprogramming powers. There is a pool o f discretionary funds (a "Fund") managed by an organization *. with authority to approve and disburse grants for local capital investments. 11. It is responsible at the central government level for quality control and performance monitoring of the financed investments, and thus responsibility over procurement and financial management. The fund has ultimate fiduciary responsibility for the good use o f the funds; thus, its function is to ensure that proper quality control and performance ... monitoring systems o fthe investments are inplace (specified inan Operational Manual). 111. It finances multisectoral investments (typically local public goods), usually small-scale infrastructure but also social care services. The menu could be more open or closed, as long as there i s the possibility of inter-sectoralchoice. 9. As an instrument o f the CDD and local governance approach, a social fund has three defining features: i. Investmentsshouldhavebeenidentifiedanddemandedbytheirpotentialusersasatop priority for their development. While the demand-driven nature o f project selection i s the See Annex 3 and Jorgensen (2006) for a discussion o f the CDD approach and its internal variations. 3 See Helling, Serrano, and Warren (2005) for a fiamework that explains the interconnectedness that exists between institutions acting inthe local space, as well as between the approaches typically followed to strengthentheir role in service provision. See also McLean et al. (2006). 3 basic participatory feature that an SF must have to qualify as such, there i s a panoply of additional (optional) participatory mechanisms that SFs have promoted to increase the *. influence of poor people indecision-making processes. 11. Investments are targeted in that they benefit poor and vulnerable people, if not exclusively, at least to a larger extent than the better-off population. The degree and ... methods of targeting can vary significantly. 111. SFs are responsible for facilitating the development of local capacities associated with local investments to ensure the demand-responsiveness and sustainability of the investments it finances. This accommodates a wide range of approaches and institutional arrangements, 10. While these two technologies can be supported through separate instruments, the social fund specifically brings them together in one instrument. It i s important to acknowledge, though, that there 7s an underlying tension betweenthese two technologies, for the CDD approach ultimately favors local discretion and autonomy, while the programming and financing functions favor centralized controls. The multiple possible combinations of its six elements explain why this tension has resultedin a diversity of social funds that have not always been easy to catalog. Main sources of variation include whether grant recipients are local governments, community organizations, NGOs, or private firms, and the degree to which core functions are transferred to the local level. SFs can be more or less autonomous, a separate legal entity, or a ministerial unit. 1.2. Why I s a Social FundRelevant? 11. This is the basic questionthat policy makers must answer to justify the relevanceof an SF instrument: What inthe country context or policy objective makes a discretionary fund for multi-sectoral, demand-driven, pro-poor investments in local infrastructure and services a more appropriate choice than using regular government systems programs? 12. Here are three possible rationales: i. GovernmentstructuresandsystemsaredysfunctionalandtheFundprovidesatemporary means of channeling resources to the local level. For example, the country might lack a proper intergovernmental fiscal framework and functioning financial management system, or it could haveproblems of governanceand corruption. 11. There are exceptionalproblems that regular transfer systems or sectoral programming a . are not well designed to address. This includes everything from national catastrophes, to discrimination, to minorities, to deep pockets of poverty, but also to promoting inter- municipal cooperation among small, resource-strappedmunicipalities, or leveraging local ... resourcestowards national objectives through competition for national funds. 111. SFs are usedto promote innovationinservice delivery and insystems development. These options are further elaboratedinchapter 4 inthe context of local infrastructure funds. 4 ChaDter 2 -An Overview of Social Fundsin ECA 2.1 Context 13. Since the first ECA social fund was created in 1993 in Albania, governments and donors have actively supported Social Funds in the Region as instruments useful for dealing with a range of issues inherited by the transition, for example, from inadequate access to services, to local infrastructureprovision, to social exclusion. The World Bank is currently supporting SF operations in 14 ECA countries and preparing new operations in five more. The ECA SF/CDD portfolio grew rapidly from approximately $150 million in 1998 to over $1 billion in 2005 and attracted significant cofinancing by other donors (such as EU, DfID, Sida, KfW). The governments themselves are financing similar projects in Hungary, Serbia, Slovakia, and other countries4 14. The circumstancesunder which most SFs were created have changed significantly. When ECA SFs were created in the mid-1990s, most adopted a semi-autonomous institutional setup for project management. This was justified at the time, given that the central and local government capacity to plan, manage, and monitor services effectively was limited. Many societies were in the immediate aftermath of civil conflict. Free elections at the local level had not yet taken place, nor was legislation drafted to reconfigure the powers, rights, and responsibilities of various levels of government. There was widespread mistrust of government officials, and bureaucrats were for the most part perceived as inert, corrupt, inept, self-serving, unaccountable, and incapable of improving people's daily lives. The current situation, however, i s much more diverse, with some countries rapidly adopting decentralization policies, and significantly improving central and local government capacities. A particular aspect of the context in Eastern European countries is their incorporation to the European Union. The progressive replacement of Bank funds with EU funds increases the importance for the Bank of supporting institutions that are well aligned with the EUaccessionprocess. EUfocus on regional and local development and social inclusion makes them key issues of analysis inrelation to SFs. 2.2. BasicFactsabout Social FundsinECA 15. ECA SFs were mostly createdduringthe 1990s to dealwith the effects generatedby the transition from closed, totalitarian regimes to democratizing, market economies. These effects included collapsed infrastructure, post-conflict situations, and lack of social capital. Between 1993 and 2004, 17 SFs were created, with a new SF in almost every year during that 12-year span(Table 2.1). SFs in the Region have also created (with assistance from the Bank) a network for learning and exchange of experience and information among SFs and other local development programs (CDDs, rural development funds, etc.) See ECANET (www.ecanet.org). 5 Table 2.1 underpreparation;*=Not WB funded; C= Government decidedto close; M=Merged; A=Additional financing 16. The WB has played a critical role in the creation of the ECA SFs. The Bank has financed operations in 14 SFs and in all o f them a Bank loan has coincided with the creation of the SF.' In Turkey and Azerbaijan, the fund was created within pre-existing agenciedfunds. A total o f 26 operations and seven pilots were financed for about $550 million inWB funding (see Annex 1). Eleven SFs had their operations mapped to the Social Protection sector, and three to ESSD. 17. Outputs. As a multisectoral CDD program, social funds had two main outputs: (i) more and better physical assets for local public services and (ii) better local institutions (Jorgensen 2006). Tables 2.2 and 2.3 give an indication o f the physical assets created by ECA social funds. Eleven social funds invested about $350 million inabout 10,000 small-scale investments: 61% of the amount invested went to roads and bridges, small-scale water supply, and other utilities, and about 34% to social sector infrastructure 5 The year correspondsto the calendar year that appearsinthe cover ofthe ProjectAppraisal Document. 6 The SF creationdate is set on 1996, with the emergency Public Works and EmploymentProject (PWEP) since the current foundations were createdat that time, even ifinthe secondprojectthey were renamedandmerged. A WB PHRD grant supportedthe creationofthe Fund, which then operatedwith EUfunding. 8 The Bank was even presentinthose two that endedup not receivinga Bank loan. In Slovakiathe Bank provided a PHRD grant for setting up the Fund, and in Serbia the Bank provided informal technical assistance and advice during its creation (commenting on the Operational Manual, for instance) and is currently designing an operation that includesthem intheir design(LocalIntegratedService DeliveryProgram). 6 Table 2.2 Source: Social fund's MIS Table 2.3 Source: Social Fund's MIS 18. Satisfactory quality of social funds portfolio. All quality portfolio indicators for 25 social funds projects in ECA show good performance (see Annex 5). Almost 90 percent had satisfactory (80%) or highly satisfactory (8%) ratings for development objective and implementation progress. QAG ratings for Quality at Entry and Quality o f Supervision have all 7 been either satisfactory or highly satisfactory. ICRs and IEGReviews havejudged development outcomes to be satisfactory. Two additional messageshave come from the portfolio review: i. Theimportanceofembeddingthesocialfundinstrumentinanapproachfocusedon the strengthening of local institutions. The only two projects that got a highly satisfactory rating, for both Implementation Progress and Development Objective indicators, were Romania and Kyrgyz Republic, the two projects that made local institutional development their core objective. The only case in which IEG upgraded an ICR indicator was the Bosnia-Herzegovina project: IEG upgraded the project from "modest" to "substantial," understanding that the strengthening of municipal governments promoted by the project in B&H circumstances warranted a rating of "substantial." ii. The need to strengthenM&E and results systems. M&E had the lowest rating for implementation progress (76%), and IEG downgraded several ICRs due to poor M&E data. This i s consistent with the quick review o f evaluations gathered for the study. Even though many o f the study reports are titled "Impact Assessments," they are actually beneficiary assessments. Very few o f the actual impact assessments are properly designed-with both control groups and collection o f baseline information.' 2.3. Types of Social FundsandTrajectories 19. As mentioned in chapter 1, social funds are not meant to make policy but to be an instrument to implement policy. Assessing social funds, then, requires some sort o f typology that classifies the policy objectives to which they correspond. Answering this question i s not a straightforward exercise, for three reasons. First, they tend to have multiple objectives, which correspond to different policies. Second, their objectives are sometimes not clearly articulated within broader sectoral and multisectoral government policies. And, their objectives may shift over time. 20. Types. The typology proposed here looks at the broad policy objectives that SFs pursue, to identify the policy context under which the relevance and rationale of the SF instrument should be discussed. Based on the experiences o f SFs in ECA, it i s possible to distinguishtwo overarching policy orientations: lo i. Improving local service provision for the poor through increased access to infrastructure, quality o f services, and/or improved local governance o f basic services such as primary education, primary health, water and sanitation, local roads, etc. SF intervention is justiped in response to shortcomings in the regular systemfor local capital investment programming (its difficulty in reaching distant communities, its inefficiency, its lack o f responsiveness and accountability, etc.). This type o f SF 9See Annex 9 for the long list and a brief description, and Annex 4 for a summary table of these evaluations, plus some additional data on performance(cost effectiveness). loThe distinction between the two broad policy objectives parallels, in a way, the distinction made in the public economics literature between the allocation and the distribution function of the budget process (Musgrave 1959). While the local service provision objective is more aligned with the allocation function (Le., influencingspending decisions across sectors), the social protection objective is clearly aligned with the distribution function (Le., changingthe distribution o f income, wealth, or other indicators of economic well-being). 8 intervention should be discussed in the context o f a diagnosis o f the country's decentralizationframework andpolicy, which defines the intergovernmental relations for local service provision. ii. Protectingpoor and vulnerablegroupsfrom shocks andpromoting their social inclusion through targeting o f investmentresources, development o f local capacities, and innovative approaches to problems o f vulnerability and exclusion. SF intervention is justiJied in response to exceptional circumstances of deprivation that are created in the aftermath of shocks (such as in post-conflict countries, natural disasters, economic shocks, etc.) or due to the existence o f deep pockets o f poverty and social exclusion (most clearly inthe case o f discriminated ethnic minorities, such as the Roma).' The SF intervention should be discussed inthe context o f a country's * strategy for social protection, which identifies vulnerabilities and social risk management strategies to address them. 21. The main distinction between these two orientations,then, relates to the nature of the problemand the intervention.Inthe local service provision scenario, SFs are a response to shortcomings inthe regular system of public service provision. Because o f the local public good nature o f most investments, it i s very difficult to exclude the non-poor from benefiting. In the social protection focus, SFs are a response to exceptional circumstances, either caused by covariate risks (natural disasters) or idiosyncratic risks (different forms o f discrimination and social exclusion), and they benefit a clearly identifiable disenfranchised group o f people. A second distinction within each policy orientation relates to the approach to local institutional development-i.e., the importance assigned to developing and strengthening inclusive, accountable, and/or cohesive local institutions, which can range from limited to substantial or significant. 22. Combiningthese approachesand orientationsyieldsfour types of SF (see Figure 2.1): i. InfrastructureManagementFunds.Thedominant objectiveofthistypeofSFisto increase access to social and economic infrastructure. Their distinctive function i s to act as contract managers o f small-scale infrastructure projects. Differently from local governance funds, they do not transfer funds to local governments and communities, but rather contract directly and supervise private contractors who are responsible for implementation. In other words, SFs are not financial intermediaries, but more o f an implementing agency. Their main justification is the weaknesdinefficiency o f ministries and local governments. Thus, de facto or de jure, they act as surrogates o f sector ministries. Their comparative advantage i s to assist in project formulation, appraise and supervise small-scale infrastructure projects, and target resources to underserved areas. ii. LocalGovernanceFunds.Thistypeoffundshasasagoaldevelopingthecapacity o f local institutions (CBOs and LG) to manage local service provision in a participatory, accountable, and effective way. Local institutional development i s as or more important than the infrastructure that is built. The emphasis i s on improving or 11See Sipos (2004) for a brief diagnosis on the problems of social exclusion amongvulnerable and minority groups inECAcountries,andtherelevanceofsocialfunds andother socialprotection instrumentsto addressthem. 9 changing the systems, procedures, and incentives that guide local service provision, bringing a bottom-up, local governance perspective. These funds tend to devolve greater project cycle responsibilities to local institutions. Organizationally, also, they tend to be characterized by different, smaller central units, with staff that knows more ... about participatory governance methods. 111. Emergency Funds. This type o f SF has often advanced social protection goals, because it represents a coping mechanismto post-conflict situations, natural disasters, and economic shocks as well as infrastructure goals (because it often focuses on infrastructure reconstruction). The only exception to this is economic shocks, in which the response i s cash transfers and income support targeted to those who have experienced the shock. iv. SocialInclusionFunds. These funds are different from emergency funds inthat they deal with problems o f social exclusion and marginalization caused either by discrimination o f ethnic minorities, or by membership in a vulnerable group. They aim to develop a more institutionalized approach to deal with problems o f social exclusion and social assistance, an approach that relies to a large extent on local institutions. They emphasize systems innovation, local partnerships, and community development, and are more closely linked with social assistance and active labor market policies. 23. The funds described above are "stylized" types. This does not mean that each type captures all features o f a specific SF, or that a specific SF cannot have elements o f more than one type. It means that, in most cases, it should be possible to identify a dominant theme that will situate the SF in any o f these four types. For instance, while all social funds have a concern for including vulnerable groups in the decision-making process for investment selection, only some SFs develop a core program aimed to actively reach these groups with development interventions. Likewise, while all funds have a concern for building local capacities, only some funds make local institution buildingone of their main objectives. It is possible, though, to have a social fund agency managing separate funding windows (and even different staffing), which would correspond to different SF types. Each type o f SF has a distinct policy rationale and justification attached, which allows it to be compared with other instruments. 24. Mapping SF. Figure2.1 maps 15 SFs across the four categories o f identified SFs, as well as the movements within countries from one category to another. Here are some o f the general observations that can be derived from this illustration: There i s a significant diversity in the main objectives pursued by ECA SFs, which are spreaded in a balanced way throughout the four categories. In the following chapters, when we look at each o f these types o f funds, we will see that there i s also significant variation within each type. There is a certain dynamism in the design of each SF. Six o f the nine SFs with more than one operation shiftedfrom one type o f SF to a different one. One important caveat to this figure i s that there are certain social funds that are still transitioning from one type o f fund to another. This is the case, for instance, in Romania's RSDF. While RSDFhas focused more strongly duringthe last couple o f years 10 on the social inclusion approach, it still maintains strong elements o f the Local Governance model, for instance with the implementation o f the Social Development Scheme for MiningCommunities (2004-20 10). Figure 2.1 Mapping Types of SFs Signifjcant Moderate Local institutional development Limited (*)Ialso refer to these t 25. Overall,there seems to be a good correspondencebetweentype of countrycontext andthe type of SF, althoughthis is notalways the case (see chapter4): Incountry contexts characterized by post-conflict environments (Kosovo, Bosnia 1 after the civil wars), SFs have adopted Emergency Funds characteristics. In country contexts with more fragile and dysfunctional public sectors and intergovernmental systems, with no serious decentralization policies, we find Infrastructure Management Funds, where the objective of increasing access to infrastructure dominates. Tajikistan is an example. In country contexts with weak public sectors but undergoing efforts at reforming them, including intergovernmental relations, SFs tend to be Local Development/Governance Funds, where contributing to local institutional development is an important objective. Bosnia 2, Macedonia, Albania, and the Kyrgyz Republic are all examples. Incontexts with more developed public sectors and intergovernmental systems, we find Social Inclusion Funds (Eastern Europe). SFs do notjustify their intervention on the basis o f local infrastructure financing, which happens through the intergovernmental system, 11 but on fulfilling central government social inclusion objectives. All Social Inclusion Funds are in countries and tend to be formal agencies, integrated with the Ministry of Labor and Social Services. 12 Chapter 3 Organizational Structure of ECA Social Funds - 26. The public managementand governancedimensionof SFs has been at the center of the debate about SF relevanceand appropriateness." Critics have expressedconcern with the potentially distortionary effects of SFs, characterizing them as PIU-like enclaves with different remuneration, procurement, and auditing procedures, not integrated with the national budget process, with financing that overlaps with sectoral funding, and more accountable to donors than to government. Supporters, inturn, have disputed claims of distortions and have argued that the autonomy and performance management practices adopted by SFs have brought efficiency and accountability gains to the public sector and introduced effective public management approaches. 27. Much of the debate around social funds mirrors the broader debate about more autonomous agencies within the public sector.The term agency is shorthand for the variety of organizations that are outside core government and do not have vertically integrated hierarchical relationship with a Ministry. Looking at the experience with autonomous agencies both in developed and developing country settings, Box 3.1 summarize the risks and benefits of the agency approach. Autonomous agencies can be either beneficial or harmful to public sector management depending on the objective, the particular institutional and country context, the specific rules and legal frameworks around autonomy, and management effectiveness. A priori, agency is neither good nor bad, but may be appropriate or inappropriate depending on the setting.l3 28. SF relative autonomy has been the most controversial aspect, but the debate has taken place with little actual examination of the institutionalarrangementsunderpinning SFs. How autonomous are ECA SFs and in what ways are they more or less autonomous?This chapter summarizes the basic facts about ECA SF institutional arrangements and then reviews SF institutional issues within the wider vision of optimal public sector arrangements.The chapter looks at five dimensions of SF autonomy: Policy autonomyrefers to the ability to interpret or decidepolicy inspecific cases. Structural autonomy refers to the overall setup of the SF, focusing on its form of creation, legal status, the institutional location, and the governing bodies. Operationalautonomy includesthe extent to which social funds abide by regular public sector rules and regulations inpersonneland financial management. Budgetary and government accountability looks at integration in budget, and the auditing process, as well as dependenceon donor funding. Composition and sustainability looks at sources of financing and their consequences with respectto smooth functioning and continuing budget support. ''Fora review ofthe debate see Bhatia(2005), also WB (2001). l3Agency issuesare similar to those aroundproject implementationunits (PIUs) interms ofthe trade-offs between distortionary effects versus enhanced management and results. SFs may have attributes like those of enclave or super-PIUs using the classification developed in "Guidance Note for Project Management: Strengthening InstitutionalCapacity during Project Implementation," WorldBank, October2005. 13 Box 3.1: The Agency Approach: Risksand Benefits Most Social Funds in ECA are part of the larger movementtowards new public "agencies." The term agency is shorthandfor the variety of organizations that are outside core government and do not havevertically integratedhierarchicalrelationshipwith a Ministry or Department. Agencies have a long history in OECD countries. The two most common reasons for setting up agencies are to improvethe performance of the public sector or to make public decision makingmore credible by separating it from direct political intervention. In most OECD countries, agencies generally have worked well and meet important needs of good governance. There is some evidence that in the appropriate roles they have improved both the quality and credibility of public sector performance. Duringthe last 50 years, it seems that only a few autonomous bodies have been drawn backunderthe traditionallyvertical integratedhierarchy. On the other hand, the jury is still out on agency experiments in the developing world. Such agencies include parastatals, AGETPs, some social funds, independent revenue authorities, and extrabudgetary funds (pension funds, health insurance funds, road funds, etc). Some agencies appear to have created benefits; others have caused significant problems for public governance. In a wide range of countries where ministries do not pay market wages or have no culture of performance, shifting functions into agencies with greater autonomy may be seen as the only way to respond to major failuresof public sector management. Based on internationalexperience, agenciesmay or may not be the right instrument.There are some questionsthat shouldbe addressedwhen consideringthe creationof agencies: Be clear why the changeis beingmade-and why it will bebetter thanthe statusquo. Ask what a specific change proposal implies for the quality of governance in the core public sector and what must be done to improve overall public management (financial management, public employment practices, performance managementbas well as or instead of trying to avoidthe problemby ring-fencingfunctions inagencies. Understand the environment and the risks it may present to a new agency. Where are the enemies of change? Who gains power and who loses it? Inside the public sector, look to powerful bureaucratic forces like existing line ministries or ministries of finance for significant opposition. The staff of existing organizations and their unions as well as the managementmay oppose change. Follow principles of good governance: Generally, what is appropriate in government departments to safeguardthe basic interestsofthe State is appropriate for agencies. Ask whether the mission of an agency really requires that it be constitutedas a separate legal entity, with its own board and governing law, or whether some lesser degree of autonomy, perhapsunder a parent ministry,will equally serve the purpose. Make sure that central ministries have the incentive and capability to manage the new relationships;that the politiciansand legislatorsare clear on their powers and responsibilities of policy setting and surveillance; and that, in the agencies themselves, staff have the necessarymanagement skills. Expect a first burst of enthusiasm and motivation for the mission of the new organization, maybe a charismatic leader with the strong support of a powerfulpolitician, followedby a re- grouping of the forces of opposition and re-emergence of old habits and cultures of performance. How is improvement going to be embedded and thus survive beyond the pioneeringphase? Source: Laking(2005). 14 3.1. PolicyAutonomy 29. ECA SFs are typically subject to national and sectoral policy oversight. SFs are not usually viewed as independent policy-setting agencies, as would be the case with autonomous regulatory agencies, for example. In most cases, SFs are one of a set of poverty reduction instruments under a national poverty reduction strategy that i s determined at the central level. Sectoral agencies typically serve on SF Boards of Directors to ensure policy coherence, and subprojectsmay be subject to the approval of local sectoralrepresentatives. 30. The degree of effective policy coordination that the Fund has with government priorities and sectoral policies varies by the type of SF and the country context. Despite an oversight role on paper, there may be unequal power relations, such as when the SFs have more resources and strong political backing that give them more latitude in setting or not observing sectoral policies. In addition, there are often policy vacuums. For example, many SFs have had to develop sectoral norms in areas where there were none available, particularly in the introduction of the more community-based service arrangementsthat were not provided through the central ministries. 3 1. Some SFs have a better fit as instruments of national policy than others. Referring to the typology of SFs reviewed inchapter 1, Social InclusionFunds have a relatively easy fit and justification within the overall social protection policy. Social Inclusion Funds have a clearly defined mandate (Bulgaria, Slovakia, Serbia, Turkey) within Ministerial policy. This is strengthenedby the fact that the parent Ministry allocates the budget to them. This is the case even in Bulgaria where the SF is a separate legal entity, but it i s established as a second-tier budget holder entity within the Ministry of Labor and Social Policy (MoLSP). Local Infrastructure Funds have a more contested and precarious status; their original justification usually comes as atemporary instrumentthat substitutes for a dysfunctional public sector. 32. However, many Local Infrastructure Funds increasingly have a broader policy goal of introducing different, more community-driven approaches to improve local governance in the transitional economic environment. While the region has a long history of social assistance and social welfare interventions and institutions, the decentralization agenda i s relatively new. Rooted deeply in the long tradition of central decision making during Soviet times, public administration still lacks an institutional culture of power sharing, delegation of responsibilities, and local autonomy. While adequate provisions of non-interference and autonomy are guiding principles in the new laws on local self-government, these formal provisions have yet to be fully internalized and enforced, and fiscal decentralization has lagged. In this policy environment, SFs have introduced important innovations in bottom-up and participatory local planning and implementation of investments (see Box 4.1 on the Kyrgyz Republic program inchapter 4). 3.2. Structural Autonomy 33. ECA SFs have a relatively high degree of structural autonomy, although most are nominally located under a government ministry. Eight of the 14 ECA SFs are located or affiliated with a Ministry, mostly the Ministry of Labor and Social Policy (Bulgaria, Serbia, Slovakia, Ukraine) or Economy and Finance (Macedonia, etc.). Four SFs are affiliated with the office of the President, Prime Minister, or Vice Prime Minister. Intwo (Albania14 and Kyrgyz Republic) we could not identi@ a direct affiliationwith a core government office. 34. Most are formed through a legaldecreethat providesattributes different than a line ministry program, often with separate governance arrangements. Intotal, six Social Funds were createdby Law (Bulgaria, Romania, Turkey, Ukraine, Bosnia, and Herzegovina); six were created by Decree (Albania, Armenia, Georgia, Kyrgyz Republic, Moldova, Tajikistan); two by Ministerial Resolution (Macedonia, Slovakia); and one (Kosovo) o f them falls into the "other" category since it i s an independent NGO that underwent the formal process of NGO creation. Most Local Infrastructure Funds are separate legal entities governed by a Steering Board, while Social Inclusion Funds tend to be units or programs within a ministry (Serbia, Slovakia). Of those that are legal entities, most are semi-autonomous governmental agencies or institutions, either a government foundation or a nonprofit created by government. One of them, in Kosovo, became a non-governmental organization. And only one of them was specified to be a unit within the Prime Minister's Office createdto implementthe project. 35. All SFs that are legal entities have an oversight body. This varies from Board of Directors to Steering Committee, Advisory Committee, and Supervisory Board, among others, but they all tend to play the same role, and have a somewhat similar composition and structure (see Annex 6 for detailed description of composition), Usually they have one or more of the following four functions: (i) approval of policies and proceduresfor the SF (operational manuals and guidelines; (ii)assurance of transparency and compliance with rules/objectives; (iii)inter- ministerial coordination; and (iv) provision of strategic guidance to the SF. Most boards have about 10 members, but this number can be as high as 21 (Kyrgyz Republic). The majority o f central government members come from line ministries. Half of the SFs encourage participation of NGO representatives. Only three SFs include members of local governments (Albania, Armenia, and Kyrgyz Republic). 36. The chair of the oversight body tends to be assigned to a high executive authority. This authority can be the Prime Minister or President of the country (Albania, Armenia, Moldova, Tajikistan' 5, and Romania); the Vice-Prime Minister (Ukraine); the Minister of Economy andor Finance (Macedonia, Kosovo); or the Minister of State (Turkey). An outstanding question is "How effective have these oversight bodies been in fulfilling their functions?" That the highest responsibility of the oversight body often resides at the highest level of the executive body could be a sign of the degree of commitment from the government to the SF objectives, allowing this political capital to function more effectively. At the same time, it raises the concern about politicization of the body and therefore o f the SF's activities. Also, given how busy these top officials are, the question is, "To what extent do they have time to get involvedinthe oversight body?" 3.3. Operational and Managerial Autonomy 37. As a group, the majority of ECA SFs enjoy relative operational and managerial autonomy. Table 3.1 below presents the degree of autonomy inthese aspects. Of note: 14 *'ADF'scase BoardofTrustees is chairedby the DeputyPrimeMinister. Inthe of Tajikistan, the PrimeMinisterappointsthe Chair ofthe TASIFBoard. 16 Inpersonnelpolicies, 11 of 14 are free to hire and fire staff, even though only seven of 14 are able to set different staff terms and conditions. Infinancial management, most SFs enjoy some form of autonomy from government procurement and contracting procedures (nine o f 13); from government disbursements procedures(10 of 13); although less from tax laws and other regulations (six of 13). 38. The Social Inclusion Funds have much lower autonomy levels than the Local Infrastructure Funds. The social funds in Serbia and Slovakia have no exceptions granted in personnel, procurement, disbursement, or other operational aspects. This is consistent with their establishment as a unit or program within a social welfare ministry. The more structurally autonomous SFs also tend to have higher degrees of operational and managerial autonomy. Table 3.1 (*)Accordingto the agreements betweengovernmentanddonors 3.4. Budgetaryand GovernmentAccountability 39. SF budgets are usually includedin the nationalbudget. So far, only Albania, Kyrgyz Republic, and Turkey SF budgets are not captured in the budget (see Table 3.2). Turkey is the only extra-budgetary fund (EBF), partly financed through earmarkedtaxes. 40. There are differences in the way they appear in the budget. For instance, Armenia, Bosnia Herzegovina, Bulgaria, and Georgia show the SF as a line item within a certain part of the budget. For example, Bulgaria specifically shows the SF as a separate line inthe budget of the Ministry of Labor and Social Policy, and Bosnia shows it as a separate line in the 17 consolidated annual budget of the Ministry of Finance. On the other hand, Moldova records its social fund as several budget lines, depending on the type of investment (water supply and sanitation, education, culture, gas supply, roads, grants to local public administration, etc.). In Romania, the SF appears bothas external sources of funding and as a line item when the funding i s devoted to local contributions. 41, When the governmentbudget capturesthe social fund expenditure, it might include all of it or only the part that is locally financed.InArmenia, Georgia, Moldova, and Romania, the government budget captures all social funds expenditures. In Albania, it only captures the part that is locally financed. And in Ukraine, the government budget captures the social fund expenditures according to the sources of funding, as it reflects it in the budget. The way SF budgets are captured in the budget is typically the same way all donor-financed projects are treated. Table 3.2 < r * Differentreporting, I s the social fund budget auditing requirements ~ includedinthe Whatdoes it for SF Country governmentbudget? capture? budgeffexknditure Source: Informationrequestedfrom TTLs and socialfunds directly. (*) Accountingandreportingrulesapplicableto not for profit organizations. 42. Recurrentcost associated with social funds investmentstends not to be included in the bud et (Georgia is the only country in which recurrent costs appeared in the central budget)'. This is usually because SFs are not the responsible party for the recurrent costs associated with the microprojects. These are usually to be absorbed either within a sectoral or local government budget or by community associations and use groups. When looking at local budgets the situation changes: at least four of the cases for which there is information on this l6 Albania, Armenia, KyrgyzRepublic, Tajikistan, andUkraine respondedthat it was not included.Moldovaand Romaniarespondedthat they didnot know. The informationis missingfor Bosnia, Bulgaria, KOSOVO, Macedonia, Slovakia, andTurkey. 18 matter responded affirmatively and only two negatively. Expenditures of Armenia's, Georgia's, Tajikistan's, and Ukraine's social funds were budgeted in the local budgets, whereas Albania's and Kyrgyz Republic's were not. In other cases, like Moldova, there i s no new construction and therefore all facilities already have allocated operating costs. In the few cases where there are new facilities, recurrent costs are included inthe local government budget. 43. In addition to typically beingincludedwithin the national budget, nine of 13 ECA SFs are also subject to government procedureswith respect to reportingand auditing. In addition, SFs are subject to external auditing. Infact, most social funds (nine) are subjectto both internal and external auditing. At least nine social funds-Albania, Armenia, Azerbaijan, Georgia, Kyrgyz Republic, Moldova, Romania, Tajikistan, and Ukraine-have both national and international auditors. Turkey's and Bulgaria's social funds seem to only have national auditors, and Bosnia and Herzegovina and Kosovo only international ones. External auditors, whether national or international, are appointed through a bidding process. Macedonia has an external auditor but it is not clear ifit is national or international.l7 3.5. Compositionand Sustainabilityof Social FundFinancing 44. All Social Funds have multiple sources of funding (international donors, central governments,local governments, and beneficiariesof projects). The sources of funding have important consequences for the smooth functioning, sustainability, and ownership of the Fund. All cases receive funding from international donors, the central government, and the local communities. 45. Donor fundingtends to predominate,consistent with the functionmany SFs have of helpinggovernmentsmobilizeand coordinatedonor funding.The importance of World Bank funding varies across countries. In recent years it has been less than 50% in countries like Albania and Moldova, and more than 50% in countries like Armenia, Kyrgyz Republic, and Romania. EUfunding is important mostly inthose countries that have gone or are going through the accession process (Slovakia, Romania) but also among some of those that are further away from this process. Albania stands out inthis regard, with almost 40% of its funding from the EU as well as from more than eight donors funding the SF. 46. Overall, SFs receive relatively little government financing. In terms of domestic financing, the central government contribution is almost always less than 20 percent of the total budget, and usually centers around 10 percent. In general, local government and community contribution does not exceed 10 percent, although there are a few cases in which it is quite significant. For instance, Bosnia and Herzegovina achieved a high level of cofinancing from municipalities and citizens: cofinancing from citizens inthe Federation of BHreached 17percent of total financing of projects and inthe Republic Srpska accounted for 9 percent. Municipal and In those cases where there are national auditors, seven social funds-Armenia, Georgia, Moldova, Romania, Tajikistan, Turkey, and Ukraine-responded that they are appointedby the country's supreme audit institution. Only Albania and Kyrgyzstan's national auditors do not follow this appointment procedure. In five cases (Albania, Armenia, Georgia, Moldova, and Turkey), the social fund is covered in the Auditor General's report to the Parliament. And, for instance, Ukraine's is not, since the Auditor's report is only submitted to the Supervisory Board, not the Parliament.Finally, only four social finds-Armenia, Moldova, Tajikistan, and Ukraine-responded that the Public Accounts Committee (or equivalent) extends its monitoring to the social fund when scrutinizing governmentaccounts. 19 community financing together reached47.8 percent of project financing inthe Federation and 41 percent in RS. In general, given ECA income status, EUmoney to come, and the World Bank's increasing shift to budget support, a central issue for the future will be for SFs to shift to receiving an increasing share of their financing from government resources. 3.6. Conclusions 47. The stereotype of SFs as enjoying full autonomy is not accurate. In general, there is more autonomy in operational and managerial procedures than in budgetary and accountability issues, but this may be a result of being donor-financed, as many of the same arrangements are found inPIUs. There is more structural autonomy than policy autonomy, because most are either under aministry or governedby ministry representativesthrough the Board ofDirectors. 48. There is significantflexibility in the publicsector characteristicsthat SFs can adopt. There is not one blueprint or public management model that SFs have followed in ECA, with some closer to fully autonomous agencies and others more integrated into line ministries. For every issue analyzed (from legal structure and institutional location, to managerial and operational autonomy, to budgets and accountability), there was variation across countries, and always at least one case of full integration. While the semi-autonomous organizational structure of an SF has tended to be one of its comparative advantages, it i s possible to have a fully integrated SF. 49. The semi-autonomous agency approach adopted by most first-generation SFs in ECA was a clear response to both emergencies (includingpost-conflictsituations) and to the weak institutionalenvironmentinthe transitioncountries.Usingthe matrix oftypologies of SFs previously developed, with respect to Emergency Funds and Infrastructure Management Funds, the adoption of the agency approach has allowed for results in weak institutional environments. This approach may still be appropriate under such conditions. However, most ECA SFs that began with this perspective have eventually migrated to a vision of greater institutional development impact and more permanency, particularly as the institutional context has stabilized. 50. Lookingto the matrix presentedin Table 3.3, with respectto SFs that have greater institutional development goals, there will be a necessary migration away from the fully autonomous models. Comparing across SFs, Social Inclusion Funds have tended to follow an organizational model that is more integrated at the policy level and less autonomous at the structural and operational levels than Local Infrastructure Funds. Local Infrastructure Fundsthat are seen as too autonomous and not sufficiently adaptive as a decentralized and improved local governance instrument risk being shut down. For example, the Georgia SF has been absorbed into a municipaldevelopment fund. 20 Figure3.1 SF AutonomyArr Significant Local institutional development Limited 21 Chapter 4 -Local Governanceand Provisionof Infrastructure 4.1. Context 51. Improving local infrastructure provision and local governance has been and still remains a high priority across ECA societies. ECA as a Region has been undergoing rapid decentralization. While countries are at different stages o f this process, the direction o f the trend i s clearly towards deepening the path towards decentralization, with the EU as the major driver o f this change. '*All the countries included inour study except those inCentral Asia have signed the EUCharter o f Local Self-Government, which commits governments to a devolutionary form o f decentralization. l The challenges are significant, and include limited political and fiscal 9 autonomy, fragmentation o f local administrative units, and lack o f stable, predictable, and transparent fiscal transfers for local service provision. The quality o f local democratic governance is usually poor, given a culture o f hierarchical, top-down local governments that are often capable but have no culture o f downward accountability (partly due to the lack o f discretionary resources to respond to local preferences). This picture i s compounded by an atomized civil society accustomed to viewing collective action with suspicion.20 52. The ProjectDevelopmentObjective of all Local InfrastructureFunds21has included improvementsin bothlocalinfrastructureand localgovernance(see PDOs inAnnex 2). One objective involves financing increased access o f poor people to local infrastructure; the other involves improving local governance, namely, improving the capacities o f local governments and/or community organizations to plan, implement, and manage local investments in a responsive, accountable, and effective way. As mentioned in chapter 2, the relative weight o f these two objectives varies according to the fund subtype, with some emphasizing improvements 18 See World Bank (2001) for a detailed analysis o f decentralization challenges facing ECA countries at the turn o f the millennium. 19 Signing the Charter is a political gesture that indicates the direction in which the country aims to move, even if at a slower pace than others. The essential source o f municipal law for European Countries is the EU Charter of Local Se[f-Government.The Charter is an international treaty opened to signature by the members of the Council of Europe. The Charter states that (a) the instituteo f local self-government secures the rights o f citizens to participate in the conduct of local affairs; (b) it is at the local level that this right can be most directly exercised; (c) the existence o f local authorities with real responsibilities can provide an administration that is both effective and close to citizens; (d) the safeguarding and reinforcement o f local self-government secures the democratic decentralization o f power; and (e) self-government entails the existence o f local authorities endowed with democratically constituted local bodies and possessing a wide degree o f autonomy with regard to their responsibilities, the ways and means by which those responsibilities are exercised, and the resources requiredfor their fulfillment. *'"Thepost-socialist development o f civil society has been uneven, and reflects the experience o f people in ECA, for whom the need to rely on informal ways o f coping with shortages and other daily problems taught them to survive as individuals and view collective action with suspicion. As Smolar suggests, `Far from creating "a new socialist man" free o f egoism and greed, communism actually bred atomized, amoral cynics good at doubletalk and "working the system". ..' (Smolar 1996). This atomization, reluctance to act collectively, and state intolerance for autonomous organizations left people with extremely limited opportunities to create legitimate alternatives outside rimary circles of family and friends" (WB 2002, p.12) ` I Commentary about SF inthis section refers only to the 10 social funds called Local Infrastructure Funds.Annex 8 has a table describing their institutional arrangements. When referring to social fimds in this section, it should be understood as limitedto the local infrastructure type. 22 on the infrastructure side (what Ihave called Infrastructure Management Funds), and others on the way local actors interact andmakecollective decisions (Local GovernanceFunds). 53. The relevance of local infrastructure funds should be discussed in the context of the intergovernmental relations system. Local infrastructure funds finance local public goods. A widely held view, both in the academic literature as well as among practitioners, is that local public goods are most appropriately deliveredlocally, through decentralized financing, planning, and delivery systems, and that local government bodies should play a key role.22The literature on fiscal federalism provides a solid argument for why the provision of local infrastructure should be financed through an intergovernmental fiscal system that combines local government revenues, fiscal transfers, and subnational borrowing, This system should be predictable, transparent, and have a smooth and continuous functioning, andthe main instrumentto reform it i s an intergovernmental reform program. Social funds are not an obvious part of that system since they are flexible, discretionary funds controlled by the central government. The question then is this: "why have a socialfund; whatfunction can itplay in this system?" 4.2. Rationales for SF Intervention 54. The short answer is flexibility. Flexibility for three purposes: (i)to deal with dysfunctional systems, (ii)to introduce local governance reforms and innovations, and (iii) to adjust the intergovernmental system without making structural changes. These purposes are describedbelow. i. "Patching-up'' the system: a temporary stop-gap measure while government systems are developed. In countries where systems for channeling, planning, and managing funds were broken down, a special fund was an effective stop-gap measure to provide development assistance. The argument here i s that the costs of waiting for the systemto be fixed are too hightojustify doing nothing. The addedvalue of an SF comes from its "implementing agency" features-i.e., its know-how in appraising, financing, and monitoring the implementation of projects identified by local actors. This fits the Infrastructure Management SF type. .. 11. Reforming local governance systems. Simply increasing resources for local infrastructure without improving the governance of local service provision will not lead to improved welfare outcomes.23 Improved local governance is about the formulation and execution of collective action at the local The recent 22 Local governments are best placed to make allocative decisions about local public goods (with the scale depending on the nature o f the goodservice). Preferences about these goods vary across individuals and groups, and local governments are more closely situated than central government to the individuals and groups they represent and for whom they make decisions. Local governments are thus better equippedto elicit information regarding local preferences, and citizens are better able to communicate their interests to local governmental decision-makers. This exchange o f information regarding preferences, when coupled with some discretionaryjnancial resources to meet priority preferences, and strong relationships ofaccountabili@, ensures that locally determined priorities are better addressed (compared to going through more centralized mechanisms for resource allocation and investment selection). 23 The WDR 2004 provided a compelling case for this argument (WE3 2004c) 24 Shah (2006, pl). Although it could be argued that local governments are the preeminent local governance institution, the value o f the concept o f governance is that it transcends narrower confines o f local governmental institutions to include a process o f social interaction based on informal norms, networks, community organizations, 23 Governance and Anticorruption (GAC) Strategy paper points out that strengthening local governance i s one o f the key entry points for governance reform and explicitly recognizes the role o f CDD approaches: Even when opportunities for governance reform at the national [level/ are limited, there may be entry points at the local level. Local politics vary considerably, so that there are sometimes very striking differences in local governments' interest and willingness to engage in governance reform. In some settings, top-down reforms can recognize, support, and rewardprogress at the local level. In other settings, the entry point might be bottom-up participatory reform, such as community-driven development (CDD), especially when it also supports the development of local government capacity and accountability. In a small number of countries where prospects for government reform are limited, CDD operations may be the only feasible entrypointfor reform (WB 2006, p.46). The potential comparative advantage o f a social fund is twofold: (a) to promote the development o f capacities for local accountability, responsiveness, and co-production at the micro or local level, and (b) to pilot reforms o f the macro environment for local governance in countries with weak governance systems. This fits the Local Governance SF type. iii. Adjusting the intergovernmental system. A flexible fund can be useful to compensate systemic failures without having to change the structure o f the system. For instance, in countries where local governments are too small and fragmented (a common problem in several countries in the Region) and there i s no political will for structural reforms, providing incentives for alliances between municipalities for more efficient service provision could be a second-best solution. In countries where local governments have significant transfers but unclear responsibilities, a cofinancing fund can be a way to leverage local resources, aligning local andnational objectives. 55. Aligning rationales and objectives. While all social funds combine both infrastructure and institutional development objectives, depending on the main rationale driving social fund intervention, one o f these two objectives will be its main comparative advantage and will have primacy inits design (Table 4.1). and informal neighborhood organizations which facilitate and promote collective decision making in the context o f citizen-to-citizen and citizen to state interactions (Shah, 2006, p.2; for a discussion on the concept of local democratic governance see Bell 2007). Approaches to improve local governance have emphasized either community empowerment, local government strengthening, or service delivery units governance (Helling, Serrano, and Warren 2005, Ahmad et al. 2005). 24 Table 4.1 Comparative Advantages of SFs Acc 56. The design of most social funds gives primacy to infrastructure, even though their justification contains strong elements of a reformingrationale. These are social funds that startedunder the patching-up rationale and gradually incorporated elements of local institutional development elements into their design. As they did, the "reforming rationale" crept in, but they still maintained their original logic of an implementing agency that actively manages projects. Most of them retained the semi-autonomous agency form, with a project implementation rather thanpolicy-oriented mandate. While they have gradually devolved greater responsibilities to the local level, they have still retained substantialcontrol over basic decision-making processes. 4.3. SFs As Instrumentsfor Local GovernanceInnovationand Reform 57. While the institutional development focus of some SFs was on community-based organizations,for others itwas the CBO-LG interface-and within these, some emphasized the LG side and others the CBO side.25The degree to which control for fund allocation and managementwas delegated to local actors as a proxy for local institutional development is where we see variations across social funds (see Table 4.2). Only a few social funds have been clearly designedto give primacy to local institutional development. From their inception, both Romania and Kyrgyz Republic defined local institutional change as their main rationale. Romania Social Development Fund, for instance, was the first social fund in ECA that made social capital development its main objective. 25For some time there has been a debate about whether SF/CDD programs should focus on communities or local governments. There is an emerging consensus now that the main comparative advantage of social funds is on CBO- LG partnerships, in particular looking at relations of accountability and co-production (Parker and Serrano 2000, Krishna 2004, McLean et al. 2006). 25 Table 4.2 Mapping 1 organizationsI with emphasis on CBO I emphasis on LG] Main institution being strengthened 58. Menu. All social funds contributed to more responsive investment allocation processesby providingrelativelyuntiedinvestmentresources.Ina context where earmarking and lack of autonomy for allocation decisions have beenthe n o m for local institutions, having some resources to exercise deliberative and problem-solving skills is positive in and of itself. Even though most SFs have positive or closed menus (which define a set of eligible investment expenditures), these menus generally are broad enough to cover the basic local public infrastructure that citizens demand. As Tables 2.1 and 2.2 in chapter 2 show, primary education, water, and roads and bridges are the three investments that account for almost 75% of all total investment. The most important restriction is that they do not tend to finance productive initiatives. Only Kyrgyz Republic and Moldova have a negative or opened menu (which allows any type of expenditure to be financedexcept a short list of items). 59. Fund Ownership. However, most funds have retained a relatively high degree of controland discretionover the allocationof funds, drivenby their objectiveof ensuringthe targeting of poor communities. Most funds follow a two-stage targeting process: first, they allocate funds to the regional or municipal level, following a formula based on population size and poverty indicators; second, within those regions or municipalities, the social fund leads or supports a local process of profiling and ranking communities according to need, and stimulates the presentation of project proposals that fit below a certain budget ceiling. Two features of this allocation procedure reduced the degree of local autonomy and responsibility. First, the social fund retained ultimate authority for approving projects, introducing possible biases in the local prioritization process-i.e., CBOs and LGs commonly try to guess what kind of projects are more likely to be approved. Second, allocating funding on the basis of individual projects rather than a budget envelope creates rigidities and inefficiencies in the local allocation process that reduce autonomy (e.g., it makes the planning process about presenting projects rather than about local problem solving, inwhich funding i s one of several elements at their disposal). 26 60. Prioritization process, The high degree of centralization of fund ownership was compensated to a certain extent by having a participatory approach to investment selection. Even though the depth and degree of institutionalization of citizen participation has varied across cases, all social funds requiredthat project initiatives were identified through open community meetings where deliberations about preferences and trade-offs were encouraged. This approach stands in stark contrast to the purely government-driven, top-down approach to decision making that was pervasive inECA during the transition. 61. Social funds contributed even more to responsive and accountable allocations when they followed a more strategic approach in which investments were not only identified in a participatory manner but they (a) came from a local or district level plan, (b) that was comprehensive,and (c) prepared within a hard budget constraint,26and when financing was not only untied but also allocated through (c) a multi-annual, (d) block grant, that (e) mobilized significant cofinancingfrom the users of those investments. Most funds meet one or more of these criteria. The only fund that meets all six criteria is the Village Investment Project (VIP) inthe Kyrgyz Republic (see Box 4.1 and cofinancing, which is discussed inchapter 6). 26 The local or municipal space offers a more adequate unit for territorial planning than the village or community level, for several reasons: except for very small-scale investments, community scale is often smaller than the territorial catchments associated with public services and infrasrncture. When decisions are made at this smaller scale, it is difficult to ensure the coherence o f planning and governance decisions. Prioritizing at the local level allows some degree o f coordination and strategic direction to the stream o f resources flowing into the local space and provides a venue for prioritizing across sectors. A comprehensive plan means that most funding sources for local investments are captured in the planning exercise rather than having a fkagmented planning process with multipleplans, each drivenby the requirements o f a different funding source. A hard budget constraintforces local decision makers to confront the trade-offs between priority uses o f available financial resources. 27 Box 4.1 A Strategic Designfor LocalGovernance Reform: Kyrgyzstan'sVillage InvestmentProject The Government of Kyrgyzstanis engagedin an ongoing, but still fragile, processof administrative and fiscal decentralization.Kyrgyzstanhas a three-tieredsubnational administration in which only the third tier (called rural communities or village communities) is an elected local self-government (LSG) body. There are 473 LSGs, each with an elected mayor, an elected council (vi1 kenesh), and a local executive body (aiyl okmotu)that is legally subordinatedto the kenesh. Two recent legislativereforms are (i) adoptionoftheLawonLocalSelf-GovernmentandtheLocalStateAdministrationin2002,which the strengthens community self-government by formalizing the role of community-based organizations as functional bodies of self-government, and (ii)the enactment of the Law on the Financialand Economic Basis of Local Self-Governments in September 2003, which establishes (limited) financial autonomy for local self-governments, including provisions on the collection, administration, and distribution of local taxes and fees. The strategic objectiveofthe projectwas to support the ongoingdecentralizationpolicyby strengthening the capacity of citizens and local governments to engage injoint solutionsto local common problems. Therefore, there was a strong emphasis on institution-buildingat the local level, which was reflected in the institutionalarrangements: Strengthening CBOs. The residents of each participatingvillage (aiyl), which constitutes a Village Investment Union, establish a representative Village Investment Union Executive Committee (VIC). The VIC's role is to identify key problems and aspirations as well as microprojectsthat addressthese. VICs consist of six or moremembers andmust includeat leasttwo women, and representativesofthe poor andyouth.VIC membersare electedduringvillage meetings. Strengthening LG-CBO interface.A Local Investment Union Executive Committee (LIC) for the entireaiyl kenesh area functionsas a coordinatingbody amongthe VICs from all aiylswithin the aiyl kenesh area. It includes representatives from each of the VICs and the members of the aiyl kenesh, includingthe mayor ofthe aiyl okmotu. Local comprehensive plan. The Local Investment Council prepares an Aiyl Okmotu Investment Strategy (AOIS) and the Aiyl Okmotu Investment Plan (AOIP) that outlines priority investment projects to be funded. These localdevelopment plans are forwardedto the aiyl kenesh for its approval and adoption as the local governments' Social and Economic Development Program Development Plans. Multi-annual block grant. The projectprovides each ofthe 473 LICs with a four-year blockgrant for these investment plans. The size of the grant was based on the local government population, amounting to about $15 per capita for the four years. This grant is given in four annual allocations that are knowninadvance, and impose a hardbudgetconstraint. Some results: 226 LSGs formulated and formally approved social and economic ab1 okmotu development plans; all 226 LSGs conducted multiple public budget and planning meetings and hearings; more than 1,500 public information boards have been set up by the communities in public places where all essential details on meetings held and decisions reached, community investment plans, budgets, community contributions, expenditures planned and incurred, procurements made, as well as microprojectimplementationtimetables and progress achieved, are made public.Inaddition, in a context where Local Councils have been integrated by the same local clan leaders, the VIP seems to be facilitating the emergence of a new breed of local political leadership. The composition of the local councilhas changed, and about 30% ofthe new electedmayorswere VIP activists. 28 62. In half of the 10 SFs, the selection of projects for submission to the social fund is made through a prioritizationprocess that takes place at the municipal level, with varying degrees of community participation (Albania, Bosnia, Moldova Small Towns, Macedonia, Georgia, inaddition to Kyrgyz Republic). Any municipal level prioritizationprocessesfaces two key challenges: - First, how to balance the empowerment of local governments as coordinators of theplanning process with the responsiveness of that process to citizens and communities inputs and preferences. SFs in ECA align at different points along this continuum. Some emphasize more local governments' role-in Albania and Bosnia, for instance, the local council makes the final decision in the selection proce~s.~'Others emphasize the community role-in Macedonia and Georgia, a special body was formed with community and (one) local government representativefor planning and implementation. Second, how to balance a planning process that it is neither an ad-hoc list of priorities prepared just for capturing money fiom a donor nor an overly long and sophisticated process that ends up being of little practical use because there are no resources tofund it, and even creates problems because of raised expectations that go unmet. In Macedonia, Bosnia, and Albania28, projects come from a list that is prepared for the SF. In Moldova Small Towns component (see Box 4.2), Georgia, and Kyrgyzstan projects come from a comprehensive plan aimed at providing some strategic direction to local investments beyond social fund financing. In some cases, like Georgia, preparing these comprehensive plans is risky since the fund provides financing for only one project, and local governments were very small andre~ource-less.~~ is why it is important to provide a block grant that allows This more flexibility to fund the plan. Infact, when this happens (as inKyrgyzstan), the profile of projects is quite different interms of size and variety. 63. In the other half of the SF, projects are prioritized only at the community level and then submitted directly to the social fund (Armenia, KOSOVO, Romania, Tajikistan, and Moldova rural areas). To mitigate the high risk that these investments will be disconnected from or undermine local government, several of these funds require that local government's formal approval (Romania). In other cases, like Armenia, a member of the local government is part of the community group that identifies andpresents the proposal, so inpractice there is a link. *'In Albania, the role of the council appears to be stronger in urban than in rural areas. In urban areas, municipal authorities propose subprojects in a public meeting o f the municipal council with participation o f local residents. The municipal council then selects a first and a second priority subproject. In rural areas, ADF provides project information during a commune level meeting with commune council and village representatives (at least one village representative per village must be present). Each village then must submit one priority subproject to the head of the commune council. At a public meeting at the commune level each village chief is required to present hisiher village's priority subproject to the commune council, which then selects one. 28 In Albania, ADF had a pilot to test a more participatory Local Development Planning approach (Cooley et al. 2004) andwith TFESSD the Bank also supported a Participatory Budgeting Toolkit (Srinivasan 2006). 29 This has changed recently with the consolidation of about 1,100 LSGs into 69 units at the raion level. 29 Box 4.2 IntroducingParticipationand Downward AccountabilityPracticesin Local Governments-Moldova Social InvestmentFund2 The objective of this subcomponent was to introducethe local government to the participatory planning and implementation practices, specifically to help them involve citizens in the problem solving and decisionmaking, andto be transparent and accountableto its constituency.Fundingunder this component has been provided to the local governments of the 42 small towns (with populations less than 20,000 inhabitants) where the poverty (accordingto the recent poverty assessment) has dramatically increasedin the last few years. USAID and UNDP facilitated the development o f strategic local plans for 21 of the eligible small towns. The SIF facilitated the development of strategic action plans for the remaining 21 smalltowns andhas beenworkingwith the other 21 to updatethe plans. SIF is also helping to build capacity of small town primarias (local government institutions) to do participatory budgeting, increase local resource mobilization, and be more responsive to citizens with respect to service delivery.A number of capacity and governance elements are financedin additionto the subproject investments, such as public hearings on budgets, citizen reportcards, and mechanisms for local resource mobilization. The MSIF works directly through the local government institutions to develop their capacity for participatoryplanning, accountability,andtransparency. 64. Most social funds have mobilized significant local revenues as counterpart funding. Local revenue mobilization encourages accountability and partnerships. The more investment financing sources come from own source revenues (taxes, fees), the better the chances of responsive and accountable service provision. Several funds exceeded the usual "minimum requirement" of 10percent. For instance, Moldova averaged 19percent, and Kosovo 17percent. Social Funds that devolved more responsibility to the local level tended to mobilize more local resources. For example: The Bosnia CDP (where municipalities have significant autonomy to allocate funding for capital investments, conduct the bidding process, supervise project implementation, and be responsible for maintenance) has the highest level of municipal counterpart funding. Municipal and community financing together reached 48 percent of project financing in the Federation and 41 percent in RS. Cofinancing from citizens in the Federation reached 17 percent oftotal financing of projects and inthe RS accountedfor 9 percent. In Kyrgyzstan's VIP community contributions to microproject financing has considerably exceeded the minimum requirement of 25 percent of the community grant (including 3.75 percent in cash), and in many communities it i s rising over time as they gain confidence in the VIP approach and in their own ability to manage. In 2005 the average community contribution was almost 64percent of the community grants comparedto 48 percent in2004. By the end of 2005, community contributions to the investment costs of 1,303 completed microprojects totaled almost $3 million, including$0.6 million incash. 65. Procurement and financial management. Some social funds have allowed 'communitiesto manage resources and procure goods and services, and a few have granted procurement powers to local governments; none of the 10 funds, however, has transferred resources directly to be managed by local governments. This has been a missed opportunity, because by now there is a solid international experience showing that local actors (either communities, mixed bodies, or local governments) can take over these responsibilities if proper mechanisms are designed to train and accompany them, regardless of country context, and furthermore that this is the best way to develop capacities through learning by doing. Three SFs invery different country environments inECA (Moldova, Romania, Kyrgyzstan) have delegated procurement to community groups with positive results. A few social funds, like Bosnia and Albania, actively involved local governments inprocurement, having a learning-by-doing effect. 30 In fieldwork conducted inBosnia and Herzegovina, mayors acknowledged that the procedures the SF required them to follow in terms o f procurement have had an impact in the way they procure their own resources-i.e., incorporating "penalty points" into the contracts with private contractors for delays inconstruction. 66. New local institutions. Social funds have introduced new local institutions for the LG-CBO interface, but have been limited because of their project nature. Several SFs have created LG-CBO committees or councils to improve the transparency, responsiveness, and accountability o f local investment allocation and implementation. A limitation of these experiences has been their "project" nature. Inmost cases, these bodies have been circumscribed to SF resources, raising the question about the extent o f their impact on the regular government systems and procedures for local capital investment programming and execution. Despite their "project" nature, anecdotal evidence suggests that they have often achieved a "demonstration effect." In Macedonia, for instance, mayors were saying that they will continue using the Community Implementation Committees (CICs) because they are in line with EU requirements for civic participation (Box 4.3). Box 4.3 Inclusive Decision-Makingto Improve Local Governance in Macedonia Prior to MCDP's intervention, communities had no clear channel to constructively propose specific actions to their Municipal Councils, resulting in frustration and widespread disenchantment with local authorities. To create a venue for communication and deliberation linkingmunicipalities and their citizens, MCDP supported the creation of innovative Community Implementation Committees (CICs) at the municipal level-including both public sector and civil society members-in order to complement the role of the more formal and political Municipal Councils. These CICs assumed responsibility for outreach and promotion of local development at the community level and were given the decision-making authority over financing for community microprojects through a participatory planning process. The CICs are comprised of community representatives, including both local notables and members o f vulnerable groups; municipal representatives including elected councilors, mayors, and key officials; local representatives of central government agencies; and locally prominent NGOs and civil society organizations. CIC meetings employ a consensus-building process consisting of (a) the presentation of social needs by each community representative, (b) a discussion o f ideas aimed to mitigate community problems, (c) prioritization of problems and proposals by vote of all CIC members, and (d) allocation of MCDP grant funds to finance priority community microprojects. The model of inclusive problem identification and deliberation joining community, municipal, and local state representatives has proved an effective means o f building partnerships that transcend formal organizational relations. The collective decision-making process enables stakeholders to hear each other's concerns: local government representatives better understand community priorities and communities better understand that resources are limited and that setting priorities is a difficult but essential aspect of public management. The advantages o f expanding decision-making beyond the municipal council without undermining its formal role are expressed by the mayor of Orizari: "Ever since the creation of the CIC, Iam more relaxed performing my functions as mayor because the community understandsthe constraints and there is satisfaction of people inthe community." The CIC model of inclusive governance as introduced by the MCDP has been influential: national training programs for municipal officials have incorporated the multi-stakeholder approach to decision-making as well as the principle of local government-community partnerships to support implementationof national decentralization policy. Source: Helling, Serrano, and Warren (2005) 31 67. "Muscle memory." The institutionalizationand capacity developmenteffect of many social funds innovations has been limited by the one-time nature of their interventions. Learning requires repetition. Several SF finance community or LG projects are one-time only. This limits considerably the capacity development effect of SF intervention. For SF financing to have a significant CD effect, local actors must experience participatory systems and procedures multipletimes to develop "muscle memory." 68. Performance incentivesfor capacity building. While some funds tried to introduce performanceincentivesin their resourceallocationmethods,they can do it better. The field of local governance is littered with training courses for local governments and CBOs, the results of which have often been disappointing. While a few social funds have introduced performance incentives as a way to improve local capacity development, the approaches have often been ad- hoc. For instance, Bosnia's CDP "performance grants" consisted of assigning an indicative budget ceiling to all municipalities for the program duration of four years (150,000 euros), giving all municipalities funding for a first subproject, and making subsequent funding conditional on successful completion of the previous project. One risk of this approach is that those municipalities with greater institutional capacities (usually the better off) might end up receiving more investments than those with fewer capacities. Infact, there was a direct correlation between LG budget per capita investments and CDP per capita transfers, deepening horizontal fiscal imbalances-the range of variation of CDP transfers was very high, from 0.3 1euros to almost 19 euros per capita (see Annex 7). Another risk i s discretion and manipulation. Without an external, objective assessment of previously agreed benchmarks, the process is opened to manipulation; and it becomes very difficult for LGs to know which areas they need to improve. Chapter 6 discusses Performance-Based Grants (PG), which offer a more comprehensive and strategic model for introducing performance incentives. 69. Developing a marketfor training and technical assistance (TA) services. Several social funds developed curricula on local expenditure management topics for local governments and organized the delivery of training to local governments (for example, Armenia, Macedonia, Kyrgyzstan). While these activities are useful and definitely provide an added value to the country, it is not necessaryto have a social fund to conduct them. The service that a social fund provides inthis case comes from its expertise incontracting out technical services and deploying them at the local level. It would be useful to find out if any social fund pursued a strategy of stimulating the development of a market for training and TA services linked with local governance and local development, which is a longer-term, more sustainable strategy. 70. Country context. While a country's decentralization environment conditions the opportunity space for LG-CBO partnerships, the comparative advantage of a fund is to "push" the boundaries of this opportunity space. Social funds in ECA show a certain correspondencebetweenthe degree of LG integration in SF and the degree of decentralization in the country. The SF that is among the least integrated with LGs (Tajikistan) is the one with the least decentralization. Conversely, SFs that work more closely with LGs (Bosnia, Albania, Macedonia, Kyrgyzstan) correspond to countries undergoing rapid decentralization. However, this is far from deterministic. The decentralization environment in Armenia is more advanced thaninMoldova, butthe designof the SFhasbeenless integratedwith local governmentsthanin Moldova. 71. The existence of weak or fragile states with weak local governments does not mean that funding should be centralized or that it should bypass local governments and go directly to community organizations.These parallel systems can be designed in a way that is 32 consistent with and strengthens the possibility of building government systems over the long haul (see discussion on performance-basedgrants and UNCDF's experience inthe chapter 6). 4.4. Implicationsfor the Future 72. Although Local Infrastructure SFs have played a positive role in the transition period, most of them need to reform their institutionalsetting to be mainstreamedinto a country's decentralizationstrategy, or exit. Most funds that started under the emergency or "patching-up" rationale gradually incorporated elements of local institutional development into their design. As they did, the "local governance reform" rationale crept in, but they still maintained their original logic o f an implementing agency that actively manages projects. Most of them retained the semi-autonomous agency form, with a project implementation rather than policy-oriented mandate. While they have gradually devolved greater responsibilities to the local level, they have still retained substantial control over basic decision-making processes. Though they have increasedlocal expenditure responsiveness,a more strategic design would have amore lasting impact. Even more, it would allow tapping into the many unexploited opportunities to advance local governments' and local service delivery units' "downward accountability" to citizens and communities. 73. Currentjuncture of socialfunds. Most social funds are actively or passively facing the scenario where they haveto reformand mainstreamor exit. Ofthe 10local infrastructure funds, only two seem to have a relatively stable situation (Moldova and Kyrgyz Republic). Of the remaining eight: One is transitioning into a social inclusion fund (Romania). Two have either been closed (Macedonia) or temporarily absorbed by another agency to be finally closed down (Georgia). The two funds that were more closely linked with local governments are considering options to migrate into an MDF (Albania), or some kind of merger with Regional Development Agencies (Bosnia). Inthe two funds that are more linkedto the infrastructure managementlogic (Tajikistan and Armenia) there is a consensus that the current operation will be the last one underthe existing modus operandi. Kosovo might be an exception due to the conflict environment. 74. Managing transitions. There is a risk that in the movement towards mainstreaming, the value-add of citizen voice and control of local governments, which is the comparative advantage of SFKDD, gets lost. This i s what seems to have happened inGeorgia, where the Georgia Social Investment Fund (GSIF) was subsumedunder the MDF. Management should learn from this and other examples of social fundtransitions inother regions.30 30For example, there are cases of attempts to merge local government and community-based approaches (Senegal, Benin) andto mainstream a social fund into the decentralizatiodfiscal fi-amework (Zambia, Bolivia). To date, there has beenno overarching assessmentofthis ongoing experience. 33 Box 4.4: Mergingthe Georgia SocialFundinto the Municipal DevelopmentFund Changes in decentralizationpolicy. InNovember 2003, about a year into the implementationof GSIF 11, Georgia experienced a process of politicalturmoil and renovationknown as the Rose Revolution.The government has recently made more decisive moves towards decentralization.In October 2006, a new territorial administrative setting was made effective along with local elections. As provided in the new organic Law on Local Self-Government Units (December 2005), a two-level administrative framework was applied, consisting of central government and local self-government levels-1 ,100 LSGs were consolidated into 69 units (59 former raions, four former Districts, and five special status cities). New localcouncils were elected inthese 69 units.The new Law on Budgets ofLSGwas adoptedinJune 2006: providingfor equalizationof local budgetsbasedon an objectiveformula. In 2006, the Government of Georgia decided to merge the GSIF into the Georgia Municipal DevelopmentFund. This has been part of a broader reform inwhich, in additionto its originalrole as a financial intermediaryfor municipal lending, the MDF has become an implementationagency for donot projects inthe infrastructuresector. MDF is responsiblefor the management of other projects financed by the World Bank (Second Georgia SIF, Second Education Project, the Irrigation Project plus supplementary financing), the EBRD, and the Millennium Challenge Georgia Fund(MCG). Common to all these projects is their primary objective: to rehabilitate infrastructure. Because their procurement, financial, contracting, supervision and other procedures are more or less the same, the merger into one agency was seen as a way to improvemanagementandreduce administrativeexpenses. Government saw GSIF only as an instrumentto finance infrastructure, and did not value its local institutionaldevelopment model. The new administrationperceived that the GSIF acted with too much wtonomy and was not in line with the ideas of the administration. Government did not consider GSIF Dbjectives of empowering communitiesLSGs to be of great relevance or urgent need, but rather as something accepted by the previous weak government under WB pressure. While the institutionalmodel Df GSIF had several of the weaknesses that have been analyzed in this chapter, the position of the new administration seemed to under-appreciate the value of engaging citizens and local governments. An :xample of this is the decisionto stop requestingcommunities and LSGs as a counterpart for investment xojects in order to speed up the process of construction.The Ministry of Finance, instead, covered the eequiredcounterpart. The challenge for social funds and the broader local governance approach is to present clear analytical nodels of how improvedlocal governance leadsto improvedservice delivery, and show results that back It up. Source: Fieldwork 34 Chanter 5 Vulnerabilitv and Social Exclusion - 5.1. Context 75. ECA countries face significant social protectionchallenges. Key issues in the social protection agenda include the following: 0 Highlevels o funemployment and underemployment, and limitedtraining services. A social care system(for orphans, disabled, the elderly, and other vulnerable populations) rooted inhigh-cost, low-quality residential care. 0 Social care services not a highpriority for local authorities, as the groups they serve have little visibility or voice. Social exclusion o f individuals and communities, including ethnic tensions3'. Weak institutional capacity, coming out o f the dislocations related to the transition from communism. Poor targeting o f traditional social assistance and cash benefits. Unsustainable pension systems at various stages o f reform. Non-income dimensions o f poverty that further reduce the ability o f the poor to take advantage o f opportunities or benefit from "mainstream" development interventions, for example, illhealth, mobility problems, neglect, and substance abuse problems. 76. In environments where there is a weak culture of inter-agency cooperation, individualsand groups with the most complex problemsare the most likely to fall through the safety net. This is particularly true for ECA's most marginalized populations, including ethnic minorities. Public organizations generally find it difficult to operate and innovate at the boundaries o f their institutions. This i s particularly true in the ECA context, where the predominant organizational culture i s a role culture based on specialization, formalization, and hierarchy. The tendency towards risk aversion i s further heightened by a Soviet legacy o f fear o f mistakes. 77. To address these issues, a growing number of ECA social funds have as their core mission protecting poor and vulnerable groups from shocks, marginalization, and social exclusion.This chapter provides an overview o fthat experience, and connects lessons learned to their implications for the future. It considers the ECA and country context, the main rationales for usingthe SF instrument, what the SFs inthe region have contributed and implications for the future. Inparticular, it looks at the role ECA SFs have played in relation to the social protection agenda, how that agenda i s being reshaped and expanded by the social inclusion perspective, and how SFs might respond. 311nEurope and increasingly in other parts of the world, the concept of social exclusion rather than poverty has progressively gained favor. The term better encapsulates the structural, multi-dimensional, and dynamic characteristics o f social and economic well-being, which together have an impact on the living standards o f individuals and groups. The adoption o f this concept has led to placing greater emphasis on cross-cutting policies and multi-agency cooperation and partnerships between government, the private sector, civil society, the community, and those who are suffering exclusion. Developing approaches being developed look at the life cycle and inter-generational implications o f social inclusion. 35 5.2 Rationale and Objectives of ECA's Vulnerability and Social Inclusion SFs 78. ECA SFs that primarily address vulnerability and social inclusion have developed .. for a variety of reasons. The justifications for using the SF instrument to address social protectiodsocial inclusion issues are the same as the general rationale for SFs: Weak institutional capacity in relation to availability and deployment o f specialist expertise, central/local integration, and multi-agency cooperation. Exceptional problems emanating from emergencies or situations that the regular transfer system or sectoral programming is not well designed to address. Among these are social exclusion o f particular groups such as people with disabilities or minorities, de- . institutionalization, deep pockets o f poverty, and problems that cut across institutional boundaries like homelessness or drug abuse. The need to promote innovation in the models, range, and organization o f social care service delivery, particularly coming out o fthe FSUexperience. 79. SFs that address emergencies were created in fragile, often post-conflict environments. Five o f the ECA funds were created in a specific emergency environment: Bosnia and Herzegovina, Albania, Armenia, Tajikistan, and Kosovo. SFs were often among the first Bank operations inthese countries and sought quick stabilization inaccess to basic services. Governance goals were limitedto the post-conflict goals o f creation o f legitimacy and calming o f internal tensions by delivery o f immediate benefits rather than long-term institution building. 80. SFs created in order to provide innovative solutions to social protection issues have been based on a variety of rationales. The ECA SFs that have core social protection objectives are Slovakia, Bulgaria, Romania, Serbia, and Turkey. They emphasize systems innovation, local partnerships, and community development, and they tend to be more closely linked with social assistance and active labor marketpolicies. They were created to Improve the quality of traditional, residential social care services for vulnerable groups. This includes infrastructure and service investments to improve the quality o f . service delivery, diversify residential services, and/or transfer management from central to local authorities. Develop community-based models of social care to (a) create models o f integrated networks o f service provision based on local assessments o f need; (b) realign roles and responsibilities in relation to policy formulation, service delivery, financing, regulation, and inspection; (c) act as an interim solution to the problem o f dispersed responsibilities/absence o f a lead agency for social care services; and (d) address problems o f low-quality infrastructure. Connect with hard-to-reach target groups through mainstream service provision. SF engagement in the social inclusion agenda has been prompted primarily by the need for innovative responses to the problems faced by disadvantaged and marginalized ethnic groups, particularly the Roma community. SFs have often been at the forefront in promoting involvement o f service providers other than the central government, such as local government, NGOs, and service user groups. Participatory and inclusive processes used to identify and implement small-scale investments were seen as crucial to promoting social inclusion, particularly in a governance legacy characterized by authoritarianism and centralization. 8 Increase employmentprospects of vulnerablegroups. To a lesser degree, this was to be done through the benefits o f temporary employment from infrastructure investments largely during post-conflict periods. None o f the ECA SFs have subsequently evolved into a full-blown workfare-type program. Instead, employment objectives tend to focus 36 on improving "employability" of marginal groups through training and education. Alternatively, microfinance has been supported as a vehicle for self-employment and, in some cases, income generation. 81. Social Inclusion SFs may have more than one of these goals. For example, the objective of the Romania Social Development Fund was twofold: reaching marginalized communities and introducing innovative community-based arrangements for services to vulnerable groups. Many of those ECA SFs (originally identified as primarily oriented to local infrastructure management) may also have a relevant role in addressing vulnerability and promoting social inclusion by including social care infrastructure intheir menus. 82. As presentedin chapter 3, the institutionalframework for the social inclusionSFs tends to be more integrated into the state machinery. Policy, structural, and operational autonomy i s relatively curtailed, as most programs are under a sectoral ministry. Emergency funds have been more autonomous, consistent with their being developed in weaker and more chaotic institutional environments, whereas the justifications for the social inclusion-oriented funds are more long-term innature andmore closely tiedto social protection reform objectives. 5.3 Lessonsof Experienceof ECA's Vulnerabilityand Social InclusionSFs Lessonsfrom the EmergencyInterventions 83. SFs have played a critical role in addressingemergency post-conflictand transition needs. They were often part of the initial Bank entry into FSU countries and were able to effectively work in weak and chaotic institutional environments. Inaddition, the rapid response capability of SFs was used to respond to natural disasters, as in the case of the Armenia SIF earthquakereconstruction efforts. 84. In general, emergency response is no longer the relevant institutionalgoal of ECA SFs. Most, if not all, have evolved away from crisis. The transition can be challenging institutionally, as noted in the evaluation of the Kosovo Community Development Project (see Box 5.1). As the transition from FSU institutional and economic structures matures, issues of institutional fit and longer-term policy objectives have come to the fore inall cases. 37 Box 5.1 Kosovo Community DevelopmentProject As an EmergencyResponse The conflict inKosovo imposedwidespread damage to housing and community infrastructure and led to significant internal displacement and uprooting of communities. In an unsure institutional environment, with the UN acting as interim administrator, the Bank approved in December, 2000, a $5 million trust fund to support the Kosovo Community Development Project in addressing recovery and reconstruction needs. The ICR highlights the importance of CDP in the post-conflict setting: The experience of the project confirmed the eflciency of social funds mechanisms in improving living conditions of the poor through the construction of small-scale inpastructure, and in building social capital ... ,In addition it demonstratedthe benejts of a community-basedproject, serving as a catalystfor aprocess ofparticipation. At the same time, the ICR highlights the institutional challenges of moving from emergency to longer- term: In the Bank`s dealings with these agencies, there often appears to be some uncertainty about theirfutures. Are they supposed to have a relatively limited existence? Is this to be solelyfor the duration of the "emergency" conditions, after which morepermanent government agencies might discharge itsfunctions? Alternatively, are the institutions to be set up on a trial basisfor an undefinedperiod, perhaps to see ifthey perform better than older agencies? In the Kosovo case, the Project goal was defined as building the CDF to be "an effective local institution that canplay a role in community development." Thus, there was a substantial achievementscored in establishing and launching the CDF under the exigencies of post-conflict "emergency conditions. Priority was given to providing rapid and visible assistance at community levels in the interests of meeting the main aim of improving their conditions. Accordingly, theproject planfocused attention on seeking to ensure sustaining the impact of the investments to be made. There was therefore relatively little need for a decision at that time about what to do for the long-term. Moreover, there was then considerable uncertainty about the roles of different government institutions in Kosovo. However, to their Lessons from the ECA SF Experience of Social Care 85. T o date, social care has been by far the most significant area of SF involvement in the social protection agenda (Bulgaria, Moldova, Romania, Serbia, Slovakia, Tajikistan, Turkey, Ukraine.) In the majority of SFs, social care is a component o f a wider operation to increase access to social and communal services.32 The main policy objective o f SFs involved in social care has been to support the development o f a community-based system o f social care services. Other key objectives include strengthening capacity to formulate, monitor, and evaluate social care services policies; developing a mixed economy o f care; testing legal, institutional, and financial arrangements; and building support for reform. Deinstitutionalization was initially cited as an objective. But, experience indicated that the development of community-based services per se i s only part o f the solution, contributing more to prevention, fulfilling unmet needs, and raising awareness o f alternatives than to the reintegration o f those already institutionalized. As the comparative advantage of SFs became clearer, the deinstitutionalization objective gave way to a broader formulation of policy objectives. 33 32Inthe Bulgaria, Serbia, and Slovakia SFs, social care is an elementofan SF operation focusing exclusively on social protectiodsocial inclusion. 33 This i s not to suggest that the SF instrument could not be used to support deinstitutionalization. To do so it must be designed within a coherent deinstitutionalization strategy. The Albania Social Services Delivery Project serves as an example o f how an SF might be involved. In addition to supporting the development of community-based services ASSDP is also supporting the reform o f the residential sector. The project will finance a development plan for each o f the centers that includes improvement o f the physical environment, preparation of a reintegration plan, 38 86. The financing of infrastructure has been the most common entree into social care services and the closest to the core business of many of the ECA SFs. SFs have increased the stock o f social care infrastructure, set new standards o f design, involved service users and caregivers inidentifyingtheir needs, and raised awareness o f the value o f good quality facilities. There are now many examples of social care facilities developed by SFs that far exceed the standards o f existing services. These facilities significantly contribute to the self-esteem o f service providers and users and create public support for quality social care. However, experience also found that if rehabilitation o f the premises i s not accompanied by modernized standards o f service delivery, the benefit to service users is minimal. 87. When the ECA SFs began to move beyond infrastructure rehabilitation, into the area of social care service delivery, the challenges were significant. Capacity building needs were greater than in any other microproject typology. Knowledge and experience within the SF and at the local level was inadequate to the task o f generating a sufficient number o f good- quality proposals. Neither was the impact o f a CDD approach on the wider policy framework fully appreciated, nor did management o fpolicy linkages receive sufficient attention. 88. Despiteinitialdifficulties,the ECA SFs havesucceeded in generatingnew models of community-basedservice delivery. The Romania SF has produced one o f the most diverse portfolios: information and counseling centers, day care, home care, night shelters, and a center for refugees. A key lesson from the early stages o f SF engagement: If the aim is to provide a range o f different models, the call for proposals needs to be managed to achieve this outcome. Both Ukraine SF and Moldova SF 1, for example, generated a narrow spectrum o f models, primarily inthe area o f day care for children and young people with disabilities. Indeed, the use o f a CDD model in an area o f service provision requiring a carefully balanced network o f services i s one o f the main tensions. Stand-alone projects have limited impact on whole system reform and can undermine the localhational planning function in terms o f determining overall funding priorities. The Moldova SF 2, however, has targeted Local Authorities rather than local communities and supporting the development o f a network o f services rather than one-off projects. To stimulate the development o f community-based services in Croatia, where the level o f service development is comparatively high, the call for proposals requires applicants to demonstrate how their service will complement and link to the existing network o f services in the local area. 89. SFs have shown the flexibility to act as niche players targeting groups and regions that mainstreamsocial care services find difficult to reach.Ina more developed environment this role i s often fulfilled by NGOs. In ECA, however, there are few mature NGOs with the capacity to engage in transferring expertise at the operational level or through national policy advice. In the absence o f a robust NGO sector, the SF instrument fulfills some o f its roles, including flexibility, responsiveness, specialization, and human change agent. The Romania SF developed an excellent system o f mentoring and evaluation involving personnel from more developed NGOs. and plans for the diversification o f the centers' activities. The reformed centers will be transferred from central to local management. Local partners therefore have a stake in determining the future direction o f their local residential center(s) . 39 90. The processes promoted by SFs-Le., local government/community partnerships and service user involvementin assessing needs and planningand monitoring responses- are at the core of social care services delivery in developed environments. The relatively autonomous arm's length position occupied by SFs enables them to bring together a range of actors on neutral territory. The ECA SFs have generally been successful in establishing NGOAocal government partnerships that have resulted in sustainable local services. Some SFs (such as Moldova) made local authority funding a condition o f project implementation from the outset. Others (such as the Romania SF) provided funding for six to 12 months before expecting the local authority to contribute. 91. To date, linkagesto social care policy developmenthavebeen more readily achieved in SFs that are sector focused. The majority o f countries inwhich social care was a component o f a multi-sector SF had weak policy frameworks and no strategic vision. The policy vacuum appears to have had one o f two effects. SFs in countries with a basic tradition o f social care services generated a narrow spectrum o f models and thus had a limited demonstration effect (Ukraine, Moldova 1). SFs in countries with little or no tradition o f social care services interpreted the remit very broadly, pointing to an absence o f a common understandingo f core social care service functions and a lack o f a common direction for development o f the system (Tajikistan, Turkey). Lessonsfrom the ECA SF Experienceof SocialInclusion 92. The social exclusion concept has proved especially useful in pinpointing the difficulties faced by disadvantagedethnic groups and the need for integrated responses to addressingthe factors that contribute to their social exclusion. It also provides a politically acceptable characterization o f groups who might otherwise be seen as the undeservingpoor. The majority o f instruments work at the level o f the individual or household, mediating access to financial and other resources. The SF instrument has the advantage o f being able to operate at the level o f socially excluded groups and communities, which is particularly important for groups that have become ghettoized in neighborhoods with poor infrastructure, services, and employment opportunities. 93. In contrast to the centrally planned, universal service perspective that is characteristic of the Communist system, the ECA SFs have introduced innovative strategies to address the needs of marginalizedpopulations.Although SFs have entered this new arena under the broad social inclusion banner, the shift inemphasis has been largely brought about by the need to address the difficulties faced by disadvantaged ethnic groups and, in particular, the needs o f the Roma community. The Serbia and Romania Social Fund experience provides examples o f ways SFs are reaching the Roma Community (Box 5.2). 40 Box 5.2 Outreach to Roma: Examples from Serbia and Romania Experiencein integratingRomacommunities usingSFs comes from Serbia and Romania. The Serbia SF supported a project implemented by an NGO in partnership with Centres of Social Work aimed at improving outreach of social benefits and services. A group of volunteers, mostly from Roma communities, was trainedas communityadvocates. They identifiedthe barriersRoma face in accessing services, especially the lack of information on eligibility and application procedures and the issues in obtaining basic documentation. The project facilitated close cooperation between the volunteers and social workers and alleviatedthe heavy workload ofthe socialworkers. Both groups workedtogether to ensure that the clients were able to obtain the personal documents (ID cards, health care cards, etc.) required for accessing services. The project also contributed to building positive attitudes toward the Roma population in general and ensured better understanding of their needs among different stakeholders. The project highlighted the importance of outreach and participatory approaches to problemsolving. These were the achievementsofthe project: - Documentationobtained for beneficiaries(138 birth certificates, 182 identity cards, 24 residency --- certificates, 14citizenshipcertificates, 136healthcards, 103 employmentcards). Partnerships establishedbetweenthe Centres of Social Work, NGOs, and Romavolunteers. Lower cost of obtainingdocuments(e.g., administrativefees andrewards for volunteers). A highdegreeofsatisfactionamongparticipantsand beneficiaries. The Romania SF has promoted innovations in social care services, having developed more than 140 servicescoveringa broad spectrum ofneeds.Under the new Social InclusionProject, however, the needs of more traditional users of social care services will be addressed by the Ministry of Labor, Social Solidarity, and Family, while the SF will implement a program targetingdisadvantaged Roma. The aim is to improve infrastructure; develop community-based social services in health, education, and advocacy; and train staff and community members in social services. The SF is working in partnership with the NationalAgency for Roma (NAR) to implement the program. NAR sees the SF as a valuable partner because it already has credibility and tried and tested legal and administrativeprocedures, and it can bridgethe gap betweentraditional institutionsand Romaorganizations and communities. The SIP is an instrument to implement the Romania Joint InclusionMemorandum, which wasjointly prepared and signed by the European Commission and the Romanian government, with a view to increasing the capacity ofRomato participatein localdecision-makingprocessesandto access EUfunds inthe future. Source: Social InnovationFund(extractedfrom Ringold2006). 94. SF operations directed towards other marginalized groups (such as older people, people with disabilities, etc.) have so far tended to follow the more traditional approach of increasing access to social assistance services. The social inclusion agenda calls for a broader conceptualization o f needs and responses and of the nature of the partnerships that are needed to support this agenda. SFs in Serbia, Slovakia, and Poland provide examples of programs evolving to meet these challenges: . The Serbia Delivery of Integrated Local Services Project seeks to broaden the mandate o f the SF to consider proposals for innovative approaches that use an integrated approach (Le., proposals that combine health, education, and social assistance services). The Social Inclusion Partnerships (SIPs) developed under the Slovakia Social Fund proved to be very important in mobilizing various community players and ensuring consensus on the local level around microprojects reflecting real poverty and inclusion needs. Institutionalizing the SIPs under the auspices o f a new or existing NGO was a 41 significant factor in determining their overall success-it placed the partnership at the center o f the activity and further validated it by financing the operation of each SIP as a . small project management unit with dedicated financing and coverage of small operational costs. The Poland Post Accession Project provides a further example o f a new generation of projects focusing on social inclusion rather than traditional social protection. The project includes a component that will focus on combating social exclusion inpoor rural districts (gminas). The component i s linked to the National Action Plan and will finance technical assistance. The assistance will be used to develop techniques for social assessments, social inclusion strategies and action plans, social services standards, and a Social Inclusion Program. The goal i s to enable underdeveloped rural gminas to purchase high- priority services for youth, adults, people with disabilities, and older people, as well as staff training and technical assistance services. 95. Turkey's Social Development and Employment Project has piloted a conditional cash transfer program aimed at the extreme poor. The program is one o f several interventions designed to address social inclusion implemented by the autonomous General Directorate o f Social Assistance and Solidarity (SYDGM) under the Prime Minister's Office. Small stipends are provided to households that agree to keep their children in school and attend health services. An evaluation o f the pilot found modest net impacts on enrollment in primary and secondary schools, with the biggestnet impact on girls enrolled inthe first five grades. After a pilot phase, the Bank i s now financing $500 million to the CCT program for some 2 million poor children 0-18 years old, representing the poorest 8 percent o f the population. While the experience i s still new and requires more extensive evaluation before final conclusions can be drawn, there may be a role for such programs inECA countries that still have systemic issues in coverage and utilization o f basic education and health services, particularly ifthey aspire to enter the EU. Lessonsfrom the ECA SF Experienceof EmploymentandTraining 96. Althoughjob creationis frequently cited as one of the benefitsthat accrue from the SF, it is generally a secondary benefit of improving local infrastructure. The social protection effects o f social funds interms o f employment creation have been more important for their microeffects on individuals and households than for any effect at the macro or national levels. On the microlevel, evidence from Moldova and others has shown that the income transfers from temporary employment can be a significant source o f cash income and poverty reduction, especially at the beginning o f the transition and particularly in rural areas. However, these programs differ from the classic, national workfare programs where only very high labor intensity works are eligible, lower-than-market wages are paid, and the scale i s large enough to match the objective of addressing national employment issues. Instead, they have been mixed programs, balancing short-term labor generation objectives with development priorities o f the cornmunitie~.~~ 34There are, however, examples of ECA SF activities that were specifically designedto promotetemporary employment.KfW supportedthe Albania SocialFundto implementa pilot programinwhich grants were conditionedon halfofthe investment fundsbeingusedto employunskilledlabor. The majorityofprojects focused on improvinggreen spaces. Althoughsuccessfulinfulfilling its immediateaims, the longer-termimpactofthe pilot was limitedinthat it was neither extendednor mainstreamed.Inthe followingfindinground, KfWelectedto focus on water projects. 42 97. Beyond temporary employment, several ECA SFs have as one of their core objectives to improve employmentopportunitiesvia training, information, and placement services, or via their access to incomegenerationactivities.The comparative advantage o fthe SF lies in the capacity to include employment and training initiatives within an integrated program o f support to local development and/or social inclusion: The Bulgaria SF is the only SF in which "active labor markets" was the primary focus. The SF provides a range o f active labor market services, including economic and labor market information, employment services, retraining, small business assistance, business incubators, and local economic development planning. The role o f the SF was to leverage improved performance from the responsible bodies, including local employment bureaus. These bureaus were responsible for contracting a range o f active labor market services in their locality. The SIF acted as a "business office" for procurement o f goods and services and the disbursement o f funds to local employment bureaus. The project is showing excellent sustainability prospects, including potential for financing under the European Social Fund. Other SFs have supported retraining. For example, the Slovakia SF has utilized European Social Fund resources to support active labor market measures, including a program to train social workers in poor communities-but the lack of complementary investment funds was a drawback. The Romania, Tajikistan, and Turkey SFs have supported the creation of micro/smallbusinesseswith an incomegenerationfocus. There has been considerable debate about the appropriateness and effectiveness o f SF involvement in this area. Concerns include the potential to undermine microcredit schemes with the provision o f income-generating grants and the lack o f SF expertise in micro and small business development. SFs that have pursued these initiatives can each point to case studies where real benefits were provided to sections o f the population that otherwise would have no opportunity to break the cycle o f poverty and exclusion. SFs may be an effective vehicle for development o f a nascent microfinance sector where there i s no national experience. However, from the beginning the focus should be on piloting and eventually handing over a broader microfinance intervention that can support the entire gamut from policy formulation and legislationto industry supervision. Integrationwith the EuropeanUnionSocial Protectionand Social InclusionPolicies 98. The ECA Region is increasingly correlating policy development with European Union integration. ECA countries that are joining or hope to join the European Union must bring their policy frameworks into alignment with EUgoals. The Open Method o f Coordination (OMC) process is moving beyond identifying broad principles to focus on the effectiveness of policies. The OMC i s also concentrating policy effort on creating more and better jobs and a high level o f social protection and social cohesion.35The EU's social protection and social inclusion policy framework will bring these issues to the fore. 35 In March 2006 a new set o f common objectives was adopted. These overarching objectives are (a) social cohesion, equality between men and women, and equal opportunities for all through adequate, accessible, financially sustainable, adaptable, and efficient social protection systems and social inclusion policies; (b) effective and mutual interaction between the Lisbon objectives o f greater economic growth, more and better jobs, and greater social cohesion; and (c) good governance, transparency and involvement o f stakeholders in the design, implementation, and monitoring o f policy. Specific social inclusion objectives include (a) access for all to resources, rights, and services needed for participating in society and preventing and addressing exclusion; (b) social inclusion o f all, both by promoting participation in the labor market and by fighting poverty and exclusion; (c) good coordination o f 43 Box 5.3 Channelling EUStructural Funds: The Slovak Development Fund Experience The Social DevelopmentFund(SDF) started with a World Bank grant in 2003 under the umbrellaofthe Ministry of Labor, Social Affairs, and Family of the Slovak Republic (SR) one year beforethe accession of the SR to the EU. The SDF targets poor communities and socially excluded groups, including geographically isolated poor communities, and particularly Roma settlements; the unemployed in disadvantagedregions; and other vulnerable groups (disabled, homeless, single elderly, and other groups requiring services traditionally provided in residential care institutions). Unlike other social assistance schemes in SR, the SDF involves communities (or target groups) in the preparation, implementation, monitoring,andevaluation ofthe projects fundedthrough SDF. Social Inclusion Partnerships (SIPs) are formed to bring in a broad range of stakeholders to ensure the social inclusion of marginalized groups and communities. SIP priorities are to stimulate economic activities, promote equal opportunities, and address social exclusion.They seek innovativeapproachesto engagingcommunities and contributingto better coordinationofpublic,private, andnot for profit sectors. SIPs develop a strategy for social inclusionfor their territories, assisting in preparation of projects so that they are in line with the strategy. Institutionalizing the SIPs under the auspices of an NGO was a significantfactor intheir overallsuccess The SDF has become a conduit for EU structural funds (namely the European Social Fund [ESF]). The SDF focused on regional aspects of employability, particularly in the poorer communities. These activities were eligible for ESF funding through the ESF's Operational Programme Human Resources. The operational program was negotiated and approved by the European Commission in 2004. The SDF startedto operate inJuly 2004 as a nationalproject under the OperationalProgramme HumanResources, which was implemented on a pilot basis in three regions. The original 15 Social InclusionPartnerships grew into 21. This growth shows the project's success, but it also created an administrativeburden.There were difficulties in the arrangements for the financial flow of ESF funds, which resulted in tremendous delays intransfers to local recipients. Lessonsof experiencefrom the Slovak Republic's Social DevelopmentFundpoint to five specific design features to ensure the potential future uptake of funds associated with the use of EU structural funds, including: (a) define the eligible activities, keepingin mindthe fit with prospective funding sources; (b) develop a framework for financialflows; (c) scale up progressively;(d) incorporate trainingand capacity building; and (e) diversify fundingsources. Source:PojarskiandRingold(2006) 5.4 Implications for the Future: Orientation towards Greater InstitutionalIntegration and Impact 99. The ECA countries have largely moved beyond the type of basic demonstration effect provided by stand-alone CDD microprojects. Evenwhere development levels are low, the process of reform is better served by modeling a whole systems approachfrom the outset. To this end, investments should be targetedto the development and implementation of local social care delivery plans. 100. There is a great deal of potential for SFs as regular programs of the social protection line ministries. In terms of institutional structure, the sector-focused SFs are more social inclusion policies, such that they involve all levels of government and relevant actors, including people experiencingpoverty,andthat they are efficient, effective, andmainstreamed. 44 firmly embedded within the national social protection institutional machinery than their multi- sectoral counterparts. A number of factors seem to contribute to their stronger strategic orientation, including clarity of roles, proximity to new policy and institutional developments, andbetter integration with the state apparatus. 101. However, the flexibility of the multi-sectoralECA SFs may be better able to address the cross-cutting issues that affect vulnerable populations. Mature SFs with experience of social care, employment promotion, infrastructure development, community participation, partnership, and strengthening of social capital could make a valuable contribution to innovations in this area. This i s very much the role that FOSIS has played in Chile, where the multi-sectoral agency has been at the forefront in creating multi-sectoral programs to address exclusion (Box 6.3). Some of the implications regarding EU and other aspects are developed in the next chapter. 45 Chapter 6. Options for FutureEngagementwith Social Funds 6.1. Are Social FundsRelevant for ECA? 102. Judging the relevance of a social fund requires assessing its policy objective and precise rationale. The relevance o f social funds as institutional models does not hinge upon their organizational structure, because social funds can be more or less autonomous. The debate about the role and relevance o f SFs has focused excessively on the autonomy features and organizational form o f SFs rather than on their policy objectives and purposes. SFs can adopt more or less autonomous forms to suit country and policy objectives. Interms of organizational structure, the main conclusion i s that as SFs align themselves with rationales that have greater institutional development goals, there will be a necessary migration away from fully autonomous models. 103. The emergency, or stop gap objectives, of many ECA SFs was a clear response to both the challenges of post-conflict settings and to the weak institutional environment in the transition countries. However, most ECA SFs that began with this perspective have eventually migrated to a vision o f greater institutional development impact and more permanency, particularly as the institutional context has stabilized. The emergency, short-term objectives are no longer relevant, though existing programs may from time to time respond to shocks, but not as specialized agencies to that end. 104. For the future, there are really only two overarching and distinct policy objectives that social funds in ECA may pursue: improving local governance and infrastructure provision, and addressing vulnerability and social inclusion. A social fund can be a relevant instrument within any o f these two policy domains as long as its design fits within one o f the following two rationales, and the rationales fit the country context, namely *.i. introducing reforms and innovations for local service provision, and 11. providing incentives to compensate for shortcomings in the regular system o f public provision, including equity objectives. 105. This concluding chapter assesses the potential future paths based on the experience to date and lessons and examples from other regions. Options are discussed, taking into account the different country contexts present inECA, and drawing from global experience from other Regions. Given the country differences, there i s no "one size fits all" common trajectory. The chapter assesses future options from four angles: i. Futuretrajectoriesforlocalgovernancereform. ii. FuturerolesforSFswithinthesocialinclusionandprotectionagenda. iii. IssuesspecifictoEUintegration. iv. Developing a coherent vision. 6.2. Future Trajectories for Local GovernanceReform 106. The main conclusion in the area of focal infrastructure provision and governance is that although SFs have played a positive role in the transition period, most of them need to 46 reform their institutional setting to be mainstreamed into a country's decentralization strategy, or exit. Most funds that started under the emergency or "patching-up" rationale gradually included local institutional development elements into their design. As they did, the "local governance reform" rationale crept in, but they still maintained their original logic of an implementing agency that actively manages projects. Most of them retained the semi- autonomous agency form, with a project implementation rather than policy-oriented mandate. While they have gradually devolved greater responsibilities to the local level, they have still retained substantial control over basic decision-making processes. And though they have increased local expenditure responsiveness, a more strategic design would make a more lasting impact. Even better, it would allow tapping into the many unexploited opportunities to advance local governments and local service delivery units' downward accountability to citizens and communitie~.~~ 107. In any case, the fundamental question is this: Why is a separate financing and programmingfacility needed?Once that need disappears, either because it has beenabsorbed by the intergovernmental transfer system or because the problem meant to be addressedhas been solved, there is no role for a social find. Exit options include transferring capacities to a line agency charged with the decentralization and local governance agenda or to a municipal or regional development fund, or embedding lessons learnedinto the public investment system. 108. The implication is that governments need to rethinkthe strategic focus and design of these social funds, aligningthem better either as drivers of reform and innovationor as compensatory mechanisms, or phasing them out. The strategic objective for social funds trying to improve local governance is to generate incremental changes37that support the shift towards (a) responsive and accountable local governments, (b) community-based organizations that provide opportunities for citizens to identify and address community concerns, and (c) an enabling environment that facilitates relations of co-production and mutual accountability betweenLG-CBOs. 109. It is possible to identify a few entry pointsthat would be appropriateaccordingto a broadcategorizationof countrycontext. 0 Contexts can be categorized according to (a) the strength of their public sector governance system, and (b) the "opportunity space" for LG-CBO partnerships, which is determined by the degree of political, fiscal, administrative, and functional decentralization, and the strength of CBO and civil society (more or less constrainedenabled), * 0 Entry points for reform can be broadly categorized according to whether the emphasis is on reforming (a) the macropolicy environment (e.g., the intergovernmental fiscal system, the legal and regulatory framework, the electoral system, etc.); (b) the micropolicy environment (systems, procedures, guidelines, practices for local public expenditure management, civic engagement, etc.); or (c) both.39Efforts that focused exclusively on reforming the macroenvironment are the domain of public sector reform programs, and 36For a menu of options for advancing downward accountability in local government and CDD approaches, see Serrano (2006). 37See World Bank's Global Monitoring Report 2006 for an argument in favor of a strategic incrementalist approach for CDD programs (WB2006, p.159). 38For the "opportunity space" concept and how to apply it to operations, see McLean et al. (2006). 39For a more nuanceddiscussionofpolicy domains and impact strategy, see UNCDF (2005, pp. 15-17). 47 definitely not the comparative advantage o f a social fund, which lies more inthe last two entry points. 110. Withinthis perspectivethere are several specificoptions, including (a) pilotingperformance-basedgrantsto complementthe fiscal transfer system, (b) compensatingfor unfundedmandates, (c) linkinglocalgovernmentsand communitygroups, (d) channelingcofinancingto link municipalitiesand/ or nationalsectoralobjectives, (e) exitingto a municipaldevelopmentfund, or ( f ) mainstreamingintolineministry supportto local governments. 111. (a) Pilotingperformance-basedgrants. Creating a separate financing andprogramming facility as a social fund can serve to pilot a Performance-Based Grant System (PG). Unlike grants where funds are distributed to LGs simply to give them the means to execute specific functional mandates, PGs create incentives for improvements in performance by linking the LGs' performance in pre-determined areas with the access to and the size o f funding. LGs can access the allocation only after fulfilling certain minimumconditions. Most systems operate with performance measures, which are used to adjust the grants up and downwards based on performance. Both access conditions and performance measures can incorporate many o f the partnership and civic engagement reforms o f the CDD approach, but in a sustainable manner, because they are anchored in permanent systems. Once the pilot i s successful, scaling them up should be done through the intergovernmental fiscal system, and staff absorbed by the agency responsible for managing performance assessments and the capacity development component. Piloting a PG through a social fund makes sense in contexts o f weak public sector governance systems (where the fiscal transfer system does not inspire much confidence) and/or where government is ambivalent about decentralized development and needs proof.40In countries that have stronger governance systems and that are engaged in decentralization, the pilot should be mainstreamed from the start through the treasury system. 112. (b) Compensatingfor unfunded mandates. Another entry point that combines macro and microreforms is illustrated by the Romania Rural Development Project, a two-phase $100 million APL currently under implementation o f phase 1 ($40 million). Rather than providing block grants, this APL provided grants for roads and water supply investments, two devolved by almost completely unfunded functions. While empowering local governments to deliver on these mandates (macro reform), the project aims to strengthen local governance and the capacity o f commune governments and community groups to jointly undertake socially, economically, and technically viable roads and water supply investments (microreforms). Its approach combines CDD and strengthening o f local government by strengthening the interface between citizens and their lowest-tier local government-by adopting a community-driven and participatory approach to these two devolved but almost completely unfunded functions. Because o f the innovations o f this approach, itwas first piloted in 100communes. 40A good example of piloting PGs comes from UNCDF (UnitedNations Capital Development Fund), which has tested and refinedits Local DevelopmentPrograms in about 25 Least Developed Countries from Asia and Africa. Most of these countries face similar or more constrained environments than Tajikistan, Armenia, and similar countries in ECA. The Bank has often built upon UNCDF's pilots, scaling them up through country systems (Cambodia's RuralInvestmentandLocalGovernanceproject, Bangladesh's LocalGovernanceProgram), or parallel systemsifgovernanceconditionsdo not allow it (e.g., Mozambique'sDPFP). 48 113. (c) Linking local governments and community groups by bringing in community- based organizationsWhile inthe previous two examples, empowering local governments was the main entry point to strengthening the LG-CBO interface, other examples use community empowerment as their entry point. An example o f the community empowerment emphasis i s Kyrgyzstan's VIP, discussed in chapter 4. While local governments were involved in decision making, allocations were made by a committee with majority representation o f civil society. Procurement and financial management occurred at the community level. In contexts where the opportunity space for partnerships i s enabling, a possible entry point i s to allocate "community partnership grants" to LGto deepen these partnerships. This i s the model that the US Community Partnerships Grants has, and that has spread to South Africa, and other countries. CPG programs provide small grants to citizens to help cover the costs o f citizen-initiated neighborhood projects. Their central element is that small groups o fcitizens are the prime movers ininitiating proposals, organizing work plans, competing for grants, and then carrying out projects that improve daily life within their communities. These programs are designed to foster cooperation and proactive relationships among citizens, and between citizens and their local governments, in matters of service delivery and improvements in neighborhood and community ~ e l l - b e i n g .In ~ ~Latin America, some SFs have also experimented with community empowerment grants to local governments as an incentive for local governments to subcontract communities to manage small infrastructure projects. A more mainstreamed approach that does not require an SF i s earmarking a percentage o f capital grants for community empowerment projects defined through simple and clear rules o f access. This approach was used in Zimbabwe, where 10 percent o f the annual allocation to the District Development Grants (DDGs) was earmarked for communities. 114. (d) Channeling cofinancing. A flexible fund can be useful for dealing with systemic failures without having to change the structure o f the system. In countries where local governments are too small and fragmented and there i s no political will for structural reforms, a temporary second-best solution can involve having a cofinancing fund that stimulates alliances between municipalities for more efficient service provision.42 In Armenia, for instance, the Bank's Public Sector Modernization Project supports the conceptual and legal framework for inter-community unions (ICUs). The Ministry o f Territorial Administration has expressed interest inusingthe Armenian Social Investment Fundas an instrument to strengthen these inter- community unions (Box 6.1). The Honduras Social Fundhas tried this approach. It has supported more than 20 inter-municipal associations for thejoint provision o f ~ervices.4~ 41 For a description see Adams, Bell, and Brown (2002 and 2003). 42 Inter-municipal agreements, however, entail significant costs. For a discussion o f the costs and benefits o f amalgamation and inter-municipal cooperation for the case o f Latvia, see Dillinger (2003). Besides economies o f scale, managing a common natural resource shared by several municipalities is another reason for inter-municipal coordination. This is the case o f integrated watershed development around a river basin shared by several municipalities. There are many experiences o f territorial approaches to rural development based on inter-municipal cooperation. 43 The SF can provide conditional cofinancing to local governments to leverage local finds towards national sectoral priorities. This is most appropriate in contexts where LGs have discretionary resources for capital investments but where LG responsibilities are not clearly established. A co-financing hnd aims to align local and national preferences through differential co-financing rates. Bolivia's FPS i s an example o f that type o f fund (Serrano 2005). While the problem o f most ECA countries has been the opposite (too little expenditure autonomy), the EU process might create a similar situation in which a massive inflow o f resources needs to be allocated at the local level, and there is a need for a signaling mechanism to align local-national preferences. 49 Box 6.1 Institutionalizationof Inter-CommunityUnions: The Case of Armenia Excessive fragmentation of local government units in Armenia is a key policy reform that must be addressed in a rationalization of local government and design of a decentralization strategy. For political reasons, the government has been reluctant to change the status quo. Given the absence of a provision in the Armenian constitution for an intermediate level of government, a Bank Policy Note has proposed "the regulation and institutionalization of Inter-Community Unions (ICUs)" as an alternative mechanism, facilitating "territorial-administrative amalgamation or association aiming at a more efficient and effectivejoint service delivery" (WB 2005). It is recognized universally that ICUs, properly constituted, would be an asset, particularly for rural communities, in achieving economies of scale and facilitating the establishment o f the critical mass of resources and management or technical capacity necessary to take on large or complex tasks, or address local "cross-boundary issues." These might include, for example, irrigation projects, inter-community roads, or the establishment of kindergartens to serve a number of small, neighboring communities. Moreover, as many observers have suggested, the focus on ICUs is also an initial step towards the consolidationof smaller entities into larger municipalities. Interms of current practice in cooperation among communities and participation of local communities in other broader networks, this process seems at a very early stage. In recent years, ICUs for general purposes have been formed in several marzes, many with the support o f GTZ. However, under current legislation, there seems little likelihood of a dramatic increase in their presence. For the moment, most such organizations exist only as vehicles for donor-funded projects. There are, however, some exceptions. Meghrashen is one of eight villages that havejoined together to form a water users association, sharing a water source and pipeline, and collaborating in sharing the water. The mayor argued that further cooperation with others is essential to solving common problems. The villages have also recently agreed to work together in addressing a common problem with infestation of agricultural land by field mice, since the failure of surrounding villages to take similar action will result in re-infestation in the case of any village acting alone. All villages have agreed to levy a small fee on households to purchase pesticides to enable them to get rid of the mice. Sources: Rawkins (2006). WE3 (2005. 2004) 115. (e) Exitinginto a MunicipalDevelopmentFund. One possible exit option for a social fund is to transfer its most relevant skills and capacities into a Municipal Development Fund. MDFs act as intermediaries to provide credit to local governments and to other institutions investing in local infrastructure. There are many different types o f MDFs (see Box 6.2). In general they disburse publicly provided credit to local government and monitor local investments. MDFs are designed to remedy some o f the typical constraints o f local governments (lack o f institutional capacity, limited local resources mobilization, and limited access to long- term financing for investment programs) while preparing them to borrow in capital markets. They are typically structured as parastatal institutions and envisioned as transitional instruments. Despite their significant differences, some o f the capacities developed by a social fund are similar to the ones needed in an MDF. If an MDF already exists it can absorb some o f the staff and technologies that would strengthen or diversify the existing institution, as was done in Georgia. If there i s no MDF in the country, it would be valid to assess whether some o f the capacities contained in a social fund would be worth keeping as part o f the creation o f an MDF. The Government of Albania, for instance, is conducting studies to assess this option. Although MDFs do not provide credits exclusively for poverty-alleviation programs, they can include 50 poverty reduction objectives, like basic infrastructure investments in low-income areas and development o f marketplaces, schools, and clinics for poor residents.44 116. More than 60 countries have set up specialized financial intermediaries or municipal development funds for raising capital to on-lend to subnational governments. However, very few MDFs have evolved into market-oriented suppliers o f credit capable of mobilizingprivate sector savings (Peterson 1996). Some common features o f many successful MDFs (e.g., MUFIS inthe Czech Republic and FINDETER in Colombia) include the transfer o f credit risk to the private sector, "unbundling" functions like payment collections and credit analysis to specialized private sector firms, separating subsidies from lending, and providing technical assistance and capacity building for project preparation. A successful-if unconventional-example o f a fund that focuses on capacity building in municipalities i s Paranacidade in Brazil (Fay and Morrison 2005). 44 Tamil NaduUrban Development Fund(TNUDF), for instance, has a separate grant window for poverty-oriented investments, such as slum upgrading and cost of resettlement, and providestechnical assistance to municipalities in preparing such investments and improvingtheir own financial management. TNUDF, existing municipalfunds in some other projects may evolve to draw the private sector into small-scale urban investments. TNUDF is now managed by an asset management company, a joint venture between the TN government and private investment companies. The asset management company FINDETER, in Colombia, is another example of a market-oriented MDF, with a poverty alleviationmandate.FINDETER operates as a rediscountfacility for commercialbank lending to the municipalsector. It supplementsthe banks' project appraisalcapacity andthus improvesthe technical quality of their lending, but the banks take the commercial risk. Unlike some other MDFs, FINDETER has a poverty alleviationmandatethat it has triedto fulfill by givingparticular attentionto institutionallyweak smalltowns andby favoring investmentsinessentialservices-mainly water andsanitation. 51 Box 6.2 Modelsof MunicipalDevelopment Funds The internationalexperience with MDFs has been mixed; thus, the decisionto transform into an MDF shouldbe assessedvery carefully.There are many differenttypes ofMDFs.Two models that reflectvery different orientationsare discussedhere. The first type of MDF, currentlymore commoninthe developingworld, operates primarilyas a substitute for government capitalgrantsto local authorities.These programs supply capitalthroughMDFs at below- market rates, often combining subsidized loans with grants. Typically, MDFs of this type have a monopoly in lending to the municipal sector. Such MDFs exploit the favorable terms of their loans to impose stricter standards of project preparation on localities and to incorporate central or state government investmentprioritiesin determiningwhich projectsshouldbe funded. CreditLocal de France and several other MDFs in Western Europe have evolved through financial deregulation from closed- circuit lending institutions,which obtained capital at below-marketrates (from state grants) and lent to municipalitiesat below-marketrates,to institutionsthat competefreelywith privatesector lenders. The secondtype of MDF is intendedto serve as a bridge to the private credit market. They prepare the municipal and financial sectors for private lendingto municipalities. MDFs of this type lend at market rates of interest, allocate capital through arm's-length decisions of commercialbanks or other private- sector lenders, requirethat private lenders assumethe creditrisk of municipalloans, andtry to establisha track record of municipal creditworthiness. One such market-orientedMDF was developed in the Czech Republic. There, the MDF borrows funds from abroad, with a national government guarantee, then on- lendsthe funds to domestic commercial banks, which inturn lendsto municipalities.The municipalities do all project selection and preparation. The commercialbanks performall credit analysis and accept all repayment risk. The parastatal MDF merely confirms the creditworthinessof the commercial banks to which it on-lendsandmakescapitalavailable to nine banksparticipatinginthe system, so as to strengthen competition. The two models of MDFs need not be strict alternatives to each other. MDFs of the first kind normally have been introducedin environments where there is virtually no private lendingto local governments and where public authorities believe that private credit markets cannot be developed in the short and mediumterm. ~ ~~~~ ~~ ~ ~ 117. (f) Mainstreaminginto line ministries. As mentioned in chapter 1, social funds can be defined partly as an investment programming and quality control facility, and are known for having simple but robust operational manuals for assisting the design, appraisal, and supervision of local investments. When closing a social fund, one way of trying to absorb the capacity created in that institution is to transfer relevant staff SFs who may have a wealth of technical expertise and accumulated experience in identifying, appraising and monitoring local investments. This experience can be harnessed by the relevant line ministry in charge of supporting decentralization and local governments. This may be done directly by transferring manuals, management information systems, or personnel. It can also be done indirectly by the eventual transformation of the SF PIU into an NGO/private firm that provides technical assistance. When the Macedonia's SF was closed, this was a key option that its small core staff was considering for the future. 6.3. FutureRoles for SFs withinthe Social InclusionandProtectionAgenda 118. The same rationale of supporting innovations and reforms or the compensation logic can be applied to social inclusion policies. The comparative advantage o f the SF comes from its flexibility. SFs have shown their capacity to mediate the involvement of a range of actors at national and local level, to provide incentives for innovation, and to demonstrate a 52 unique ability to address gaps without necessarily laying claim to the space. Unlike a statutory organization or NGO involved in developing new services, SFs generally do not take control of the projects and programs they develop. The space remains open, enabling the most appropriate state or non-state actors to take on responsibilities as the system evolves. Exit is less of a necessity in the Social Inclusion Funds in part because most are already closely aligned with social protection strategies and institutions. 119. The potentialkey roles for SFs inthis context are four: i. Increasingspecializationasnicheplayersfocusingongroupsthatarenotwell served by mainstream provision or as mechanisms for targeting localities/regionsthat are laggingbehind. As niche players SFs can either play an interim role, exiting when there is a critical mass o f mature nationalhegional NGOs (or local governments) able to target gaps and pilot innovations, or they may play a .. more permanent programmatic role. 11. In employmentand training, usingthe SF's capacityto incorporateemployment and training initiatives targeted at disadvantaged and marginalized groups within broader programs of support. In the less developed ECA countries, a natural entry point for further engagement in employment and training would be through the local development agenda. Large-scale workfare programs that concentrate employment in temporary construction jobs using an SF-like model, like Trabajar in Argentina, are not typically viewed as a priority policy response to ... unemployment inECA. 111. Addressing social exclusion and the needs of marginalized households and communities via mechanisms that support a cross-cutting approach. SFs are particularly well placed to fulfill this role, especially in environments with limited experience o f multi-agency cooperation and limited NGO capacity. Experience indicates that SFs can make a valuable contribution to tackling the problems faced by socially excluded ethnic minorities. iv. Servingas appropriatevehicles for experimentingwith broaderinnovations,like conditionalcash transfers.Insocial care, SFs may have an important role to play in developing innovative a proaches, so long as they are envisioned as core programs and not rnicroprojects?' Innovations in social care need to extend beyond social assistance to include access to labor markets, healthcare, training, and public services. The flexibility and multi-sectoral experience o f the ECA SFs can help create integrated responses to social care needs o f a different nature than the multi-sectoral project typologies o f the previous generation o f SFs. 45 Despite progress on deinstitutionalization, many adults and children continue to live in low-quality residential care. In a desire to promote community-based care, SFs have largely steered away from the residential sector. With the direction o f reform established in the majority o f ECA countries, it would be timely to revisit this issue. SFs would be well placed to provide much needed impetus to reform along the lines o f the Albania Social Services Delivery Project pilot. 53 Box 6.3 Addressing Social Inclusion: The Case of FOSIS in Chile El Puente, or "The Bridge," is a program of Chile's Solidarity and Investment Fund (FOSIS). The principal goals of El Puente are to support families living in extreme poverty so that they may receive an income above the extreme poverty line and to enable them to develop the personal skills to become autonomous, meaningthat they should be aware of their rights as citizens and able to access the subsidies and services available to them to empower them to resolve their problems independently. A Family Support Counselor is assigned to each family for a 24-month period. The Counselor comes weekly at first, then biweekly, and then-gradually-monthly. The Counselor guides the family in establishing a benchmarked processto improve their living conditions and reach their objectives. The program's efforts are directed towards supporting families so that they can work to construct seven "pillars of support" in their lives: personal identification, health, education, family dynamics, housing conditions, employment, and income. FOSIS has historically worked across these sectors with microprojectsand capacity buildingfor poor communities and can bringinthis cross-cutting perspective. In El Puente, FOSIS enters into partnerships with municipal governments and links beneficiary families with local private sector and civil society services. This creates local support networks that can reinforce and provide sustainable support to the efforts to have these vulnerable households exit from extreme poverty. It is FOSIS's unique position and flexibility that allows it to mediate between these various partners and tailor the services to the individual circumstance of the program's families. Source: ILO STEP (2006). 6.4. Issues Specific to EUIntegration 120. The ECA region is increasingly correlating policy development with European Union integration. ECA countries that are joining or hope to join the European Union must bring their policy frameworks into alignment with EU goals. The Open Method of Coordination (OMC) process is moving beyond identifying broad principles to focus on the effectiveness of policies. This will play out slightly differently between the Local Governance Funds and the Social InclusionFunds. But in either case, EU integration will affect both the policy framework as well as provide a significant source of funds that ECA countries must be able to effectively channel.46 121. The EU accession process represents an important opportunity to mainstream social funds expertise into national systems for local governance and social inclusion, and for the EUprocess to benefit from social funds experience. The approachandmethodsusedby social funds in ECA are highly compatible with the ones favored by the EU: local partnerships, participatory development, a competitive process for allocating investments, and coordinating and channeling external as~istance.~~ So far, however, the links have been weak. Countries 46 In addition to EU funds, social funds can be a channel for other funds, such as remittances, towards local development. Relative to GDP, remittances are significant in many ECA countries. In 2004, officially recorded remittances to the ECA region totaled over $19 billion. Migrants' funds represent over 20 percent of GDP in Moldova and Bosnia and Herzegovina, and over 10 percent in Albania, Armenia, and Tajikistan (Mansoor and Quillin, 2007). The socialfind of El Salvador (FISDL) has placed a special emphasisin stimulatingassociations of Salvadorans living abroad to partner with municipal governments in finding local development projects. Associations of Salvadorans can participatenot only inthe general Competitive Biddings window that FISDL uses to finance projects, but more recently they can also participate in a specific financing window where FISDL allocates $1 million per bidding to co-financeprojects where these Associations participate.Since September2000, about 30 Associations of Salvadorans abroad have mobilized $2.13 million or about 19% of the $11.45 million invested in 45 development projects. Another $2.28 million was leveraged from municipal governments, $6.9 million from FISDL, and $115,000 from Ministries (Serrano 2005). 47 Similarto social funds, the EUfavors aprocess of participatory,bottom-upidentificationandprioritization 54 engaged in accession or pre-accession run the risk o f repeating the same learning process that social funds followed: creating artificial institutions with weak links, centralizing too much decision-making power, etc. For social funds, this process opens the opportunity to mainstream and scale up some o ftheir methods. 122. Social Funds should be advertised more strongly as instrumentsthat can facilitate the objectivesof the EU accession process. For pre-accession countries that have an SF, it is critical to position SF as one o f the pre-accession vehicles. Experience from Bulgaria and Romania shows that previous engagement with EU i s one o f the most important determinants of the role that agencies and programs play during post-accession. In addition, SFs should try to align their menu o f investments with the menu o f eligible activities under EUregulations. During the initial years o f the post-accession period, the lack o f systems capable o f financing activities eligible by the EUi s one o f the causes o f the lower levels o f absorption of EUfunding. Forpost- accession countries that have an SF, the main decision concerns what figure do they adopt inthe EU architecture (intermediate body or a grant recipient implementing specific measures within an operational program). 123. Social funds play a similar role to the IntermediateBodies in the EU process, and have a business model that can help local governments develop projects that will become eligiblefor financing from the EU. The management of EU structural funds places significant demands on the governments o f new member states. The operational implications o f effectively channeling these resources have not been well understood. SF capacities in this area are not necessarily recognized, but in practice they perform a similar role to Intermediate Bodies in terms o f stimulating, appraising, financing, and monitoring local investment projects and plans. For those who are unfamiliar with SFs, the relatively simple technology that i s visible on the ground belies the robust operating procedures that underpin it. 124. I f they become an IntermediateBody, however,SocialFundswill have to be careful in avoidingwhat EUdefines as conflict of interest roles. SFs usually play two roles: (a) they are a grant-making facility that approves projects and (b) they are facilitator o f local capacity building. If they become an Intermediate Body implementing an EU structural fund program, however, they will have to adopt an arm's-lengthapproach to beneficiaries andjust provide basic information, but stand only on the approval side. They will lose the local mobilization and capacity-building element, which is one o f their strengths. One way to avoid this problem is to do what Bulgaria i s planning to do-i.e., make the SF a Global Grant recipient. Under this arrangement, the Ministry o f Labor as Managing Authority allocates resources to the Social Fund, through a separate agreement, as a country-owned instrument to implement a specific measure within an operational program. The fund does not appear in the formal architecture o f the EU, which i s concerned only with Managing Authorities and Intermediate Bodies, but it i s one o f the instruments the government has decided to use to implement specific aspects o f the of local investments.For instance, for rural areas, the key EU methodology is LEADER+, and for urban areas it is the Local Economic Development(LED) approach.BothLEADER+ andLED approaches share importantelements and are consistentwith the participatory localplanningprocessesdevelopedby social funds, particularlywith those funds that supported comprehensivelocal investmentplans such as Moldova's small towns or Kyrgyzstan's VIP. In Romania, the experience of both SF/CDD operations-the Romania Social Development Fund (RSDF) and the Rural Development Project (RDPtare proving useful for the EU process. Experience from RDP in particular, which worked at a somewhat higher level of aggregation than did RSDF-communes and associations of communes-is likely to help those communities that benefited from RDP support to assume a strategic position from which to implement LEADER. Already approximately seven of the 100 or more applications to become LEADERgroupswere initiatedby RDP communes inpartnershipwith neighboringcommunes (McLean 2006). 55 program. It i s expected that the Bulgaria social fund will receive between 5 and 10 million euros a year to implement the EUprogram. LocalGovernanceFundsandthe EU 125. While the EU has been strongly promoting the role of the Regional Development Agencies as coordinatorsof the process, the relevance of this model needs to be discussed on a country-specificbasis.48Efforts to create new entities at the regional level runthe risk o f creating political orphans-ignored by both national sectoral ministries and the local governments whose efforts they are supposed to coordinate (Dillinger 2007). In Romania, for instance, the eight regions do not represent a formal administrative level, and their intergovernmental status, not clearly defined, remains somewhat ambiguous (McLean 2006). On the contrary, in Bosnia, the RDA approach might offer an institutional avenue to overcoming the dysfunctional nature o f the intergovernmental system, but it is too early to tell (see Box 6.4). Box 6.4 Mergingthe SFwith RDAs in Bosnia & Herzegovina: A Good Idea? A key obstacleto localdevelopment in Bosnia and Herzegovina (B&H) is the dysfunctional structureof government. B&H lacks a uniform national intergovernmentalframework. Each of the two entities imposes different intergovernmentalrules; the 10cantons inthe Federationand aweak State together lack the leverage to coordinatethem. Ethnic conflict has split the municipalterritory in artificial boundaries (e.g., cities split in two municipalities belongingto different entities and/or cantons) hinderingefficient provisionof services and promotionof economic development. Reformefforts needto strengthenthe State level so that it can harmonize anationalframework, and the municipalitiesso that they can improveaccountabilityinservice provision. The Regional Development Agencies model seems to hold some promise for fostering municipal cooperation across entities and cantons to take advantage of economiesof scale and synergies. Over the last couple of years, the EU has supported the creationof five RegionalDevelopmentAgencies (RDAs) inthe country, covering all municipalities. Municipalities are the founders of RDAs and the core members of the assembly. At least four of the five RDAs are formed by municipalities from both entities promoting inter- entity solidarity and making municipalitiestheir core client. There are plans to link the five RDAs through a national network, and connect them to a body at the state level, thus empowering the State to push for a harmonizationof frameworksacrossentities. Merging social funds into the RDA model would provide a longer-term institutionalstrategy to social funds while supportingRDAs,which are still very weak institutions. While the Foundationsthat have been implementingthe CDP made sense as atransitionarrangement and havebeenquitesuccessful, they are created along entity lines and thus have more difficulty in bridging the divide (even if they have supported a few projects benefiting municipalities from both sides). Also, they have been created mostly as mechanisms to implement Bank projects, in contrast to RDAs, which have in regional development a more stand-alone and long-term mission. Each one is a limited liability company formed by its founders, and financed by member dues (each municipality contributes 0.25 percent of its own fiscal revenues from that FY), and additional government and donor funding. Just as important, a good part of EUfinancing will flow through these RDAs (70 million euros per year for the 2007-2013 period), and EUhas beentrainingRDAs in its procedures. Source: Fieldwork 48Social funds have had a timid engagement with the governmentbodiesresponsible for regionaldevelopment. The responsibilityfor regional development brings significant resources to the designated bodies, making them very powerful and attracting other powerful interests. In this context, and differently than the Social Inclusion SF, the Local Infrastructure SF has not been seen as a significant actor. SFs need to promotetheir relevant capacities more aggressively and ideallybecome an early entrant in the developmentof regional structures and processes ifthey are to successfullyengagewith these bodies. 56 126. Social funds and RDAs provide complementary attributes for territorial and regional development. The SFs' comparative advantages are their participatory and social inclusion focus with respect to local investments, which is lacking in RDAs, but which is important for EU. RDAs' distinct advantage involves local economic development. SFs have the advantage of efficient procedures and experienced staff for assessinglocal investments and good working relations with local governments. RDAs have expertise intraining local governments to access EUfunds. Social funds could becomethe inclusive local development unit of RDAs. VulnerabilityandSocialInclusionSFs andEUIntegration 127. The influence of EU integrationpolicieswill shape social protection and inclusion programs in most ECA countries in the future. The OMC is concentrating policy effort on creating more and better jobs and a high level of social protection and social c~hesion.~'ECA countries are moving along distinct social protection trajectories broadly characterized as early, late, and nonadopters of the European Integration Agenda. There are distinct ramifications for social protection and social inclusion policies by grouping: ' Early Adopters: Since the mid-l990s, the majority of these countries (mainly the EU-8 countries) have significantly reformed and strengthened the institutional framework for social protection, decentralizing responsibility for service delivery, involving non-state actors, and engaging with the broader social inclusion agenda. While the legislative and policy framework is in place, decentralized responsibilities are often not supported by financial transfers. Social protection and social inclusion issues include targeting structural funds, regional disparities, and pockets of poverty, and developing capacity to implement EU social protectiodsocial inclusion policies through the OMC. The Slovak Republic's experienceinchanneling EUstructural funds is describedinBox 5.3. Late Adopters: With the introduction of the European Neighborhood Policy in 2003, a move towards European Integration has become increasingly evident in the majority of the Neighborhood Countries (e.g., Armenia, Georgia, Azerbaijan, Moldova, Ukraine, and some others). Some of these countries are making the transition from low income to middle income countries, bringingwith this transition the problems of regional disparity, deep pockets of poverty, and the social exclusion of particular groups and communities. Social protection reform has muchfurther to go and engagement with the social inclusion agenda is at best patchy, but the general direction i s increasingly influenced by the wider . EUpolicy dialogue. Non-Adopters:The non-adopter group, comprised largely of the Central Asia countries, remains at a lower level of institutional development with weak legislative and policy frameworks and less technical and financial capacity. Other constraints include few actors developing community-based models of service delivery, centralized 49 In March 2006 a new set o f common objectives was adopted. These overarching objectives are (a) social cohesion, equality between menand women, and equal opportunities for all through adequate, accessible, financially sustainable, adaptable, and efficient social protection systems and social inclusion policies; (b) effective and mutual interaction between the Lisbon objectives o f greater economic growth, more and better jobs, and greater social cohesion; and (c) good governance, transparency and involvement o f stakeholders in the design, implementation, and monitoring o f policy. Specific social inclusion objectives include (a) access for all to resources, rights, and services neededfor participating insociety, and preventing and addressing exclusion; (b) social inclusion o f all, both by promoting participation in the labor market and by fighting poverty and exclusion; and (c) social inclusion policies that are well coordinated and involve all levels o f government and relevant actors, including people experiencingpoverty, and that are efficient, effective, and mainstreamed. 57 responsibilities for service, and an unclear role for local actors. The majority of social care services are low-quality residential services. 50 128. On an operational level, in ECA countries eligible for European Social Fund support, there is like17 to be a role for an instrument that can effectively channel to an integrated approach.5 The European Social Fund is an important instrument for tackling unemployment, but it can be difficult to reach the most disadvantaged and socially excluded groups through mainstream programs due to the multidimensional nature o f the problems they face. ECA SFs may have a critical niche role to play. 6.5 Developinga CoherentVision 129. As summarizedinTable6.1, there are multiplepossiblefuture paths. Table 6.1 Reform and innovation Exitoptions Pilotingperformance- 0 Mainstreaming basedgrant models CDDAocalgovernance Buildingparticipatory approachinagency mechanisms responsiblefor Awarding community capacity building/local partnershipgrants for governanceagenda civic engagement ProvidingMunicipal DevelopmentFundsor RegionalDevelopment Agencies 0 Innovationsfor social Mainstreaminginline inclusion ministry 0 Community-based 0 Supportinglocal social care services governmentsto take on 0 Integrationofmultiple social inclusionand agencies (sectoral, socialcare objectives localgovernment, NGO 130. Transitionswill need to be carefully thought-out and a consensus formed for each country. There is a great risk o f losingthe accumulation o fexperience and innovations o f SFs if a clear institutional transition path i s not spelled out, especially in cases where exit i s the preferred option. The transitions must be made with a clear understanding o f the overriding policy objectives and inlight o f the current country context and opportunities. 50For quite different reasons, Russiacould also be seen as belonging to the non-adoptergroup. However, the level of development and convergencewith European approaches suggest that the nature of the institutional challenges will notbe markedly different from those facedinother parts of EasternEurope. 51The EuropeanSocial Fund(ESF) helpspeople improvetheir skills and, consequently, theirjob prospects.Created in 1957, the ESF is the EU's main source of financial support for efforts to develop employability and human resources.Ithelps Member States combatunemployment,preventpeoplefrom dropping out ofthe labor market, and promote training to make Europe's workforce and companies better equipped to face new, global challenges. The ESF is one of the EU's four Structural Funds, which were set up to reduce differences in prosperity and living standards and help areas of Europe which, for one reason or another, are suffering difficulties. This is usually referred to as "promoting economic and social cohesion." The ESF channels its money into strategic, long-term programs in Member States and regions across the EU, particularly those where economic development is less advanced. 58 Bibliography Adams, CharlesF., Michael E.Belland Trevor Brown. 2002. "Building Civic Infrastructure: Implementing Community Partnership Grant Programsin South Africa," Public Administration and Development, Volume 22, pp. 293-302. Adams, Charles F., MichaelE.Belland Trevor Brown. 2003. "Strengthening Civic Society in South Africa: The Community Partnership Grant Program," Proceedings of the 95IhAnnual Conference of the National TaxAssociation, Orlando, Florida, pp. 124-30. h a d , Junaid, Shantayanan Devarajan, Stuti Khemani and Shekhar Shah. 2005. "Decentralization and Service Delivery" World Bank Policy Research Working Paper 3603. Bell, Michael. 2007. "Local Democratic Governance. What Is It?And How Do We MeasureIt?" Draft, Social Development Department, The World Bank, Washington, DC. Bezhanyan, Anush et al. 2004. "Social Funds in ECA: An Effective Tool to Reach Pockets of Poverty." WorldBank Spectrum (Summer). Pp. 43-45. Bhatia, Mukhmeet. 2005. "Social Funds: A Review of Public Sector Management and Institutional Issues." SP DiscussionPaper 0508, World Bank, Washington, DC. Cooley, Sharon. 2004. "Albanian Development Fund Community Works Project 11. Local Development Plan and Participatory Monitoring and Evaluation: Structure and Methodology." Draft, UrbanInstitute,UrbanResearchInstitute, ADF. Dillinger, William. 2003. Latvia: Beyond Territorial Reform. Washington DC: World Bank. Dillinger, William. 2007. Poverty and Regional Development in Eastern Europe and Central Asia. The World Bank, Washington, DC. Fay, Marianne, and Mary Morrison. 2005. Inpastructure in Latin America and the Caribbean: Recent Developments and Key Challenges. Washington, DC: The World Bank. Helling, Louis, Rodrigo Serrano, and David Warren. 2005. "Linking Community Empowerment, Decentralized Governance, and Public Service Provision through a Local Development Framework." Social Protection DiscussionPaper 535, World Bank, Washington, DC. ILO STEP Programme. 2006. Social Protection and Inclusion: Experiences and Policy Issues, Geneva2006. WHO Regional Office for Europe. Jorgensen, Steen. 2006. "Framing the Logic of CDD Operations. Towards a Project Theory for CDD Design and Evaluation." Draft. World Bank, Washington, DC. 59 Krishna, Anirudh. 2004. Partnerships between Elected Local Governments and Community Based Organizations. Exploring the Scopefor Synergy. Social Development Paper No. 52. Washington DC: The World Bank. Laking, Rob. 2005. "Agencies: Their Benefits and Risks." OECDJournal of Budgeting 4 (4). Mansoor, Ali and Bryce Quillin, 2007. Migration and Remittances: Eastern Europe and the former Soviet Union. Washington DC: The World Bank. Mansuri,Ghazala, andVijayendra Rao. 2004. "Community-Based (andDriven)Development: A Critical Review." The WorldBank Research Observer 19 (1): 1-39. McLean, Keith, Louis Helling, Jana Orac, and Rodrigo Serrano. 2006. "Exploring Partnerships between Communities and Local Governments in Community Driven Development: A Framework." Social Development Paper No. 96, World Bank, Washington, DC. McLean, Keith (2006). "Romania: Implementation of Regional Development and Integration with Other EUOperational Programs. Issues and Options for Maximizing Absorption and Impact of EUFunds.A Policy Note." Draft, World Bank, Washington, DC, November. Musgrave, Robert A.,1959, The Theory of Public Finance: a Study in Political Economy. New York: Mcgraw-Hill. Naqvi, Maniza, Philip O'Keefe, and Christian Bodewig. 2004. "Social Protection in Post- Conflict Environments inECA." WorldBank Spectrum (Summer).Pp. 27-32. Parker, Andrew, and Rodrigo Serrano. 2000. "Promoting Good Local Governance through Social Funds and Decentralization." Social Protection Discussion Paper 0022, World Bank, Washington, DC. Peterson, George E. 1996.Using Municipal Develpment Funds to Build Municipal Credit Markets. Preparedfor the Government of Indiaand The World Bank. Washington, D.C. Rawkins, Phillip. 2006. "The Institutional Review of ASIF 2 and Recommendationsfor ASIF 3." Preparedfor Albania Social Investment Fund. April. Rawlings, Laura, Lynne Sherburne-Benz, and Julie Van Domelen. 2004. Evaluating Social Funds: A Cross-Country Analysis of Community Investments. Washington, DC: World Bank. Serrano, Rodrigo. 2005. "Integrating Social Funds into Local Development Strategies. Five Stories from Latin America." Social Funds Innovations Notes. Vo1.3. No.1. Washington DC: The World Bank. Serrano, Rodrigo. 2006. "Social Accountability and Local Governments." Social Accountability Sourcebook. Washington, DC: World Bank. Shah, Anwar. 2006. Local Governance in Developing Countries. Public Sector and Accountability Series. Washington DC: The World Bank. 60 Sipos, Sandor. 2004. "Integrating and Empowering the Poor and the Excluded: The Roma and Other `Pockets of Poverty/"' WorldBank Spectrum (Summer). Pp. 35-37. Srinivasan, Radhika. 2006. "Participatory Budgeting Toolkit for Local Governments inAlbania. Social Development Team." ECA, World Bank, Washington, DC. November. Steffensen, Jesper, and Henrik Larsen. 2005. "Conceptual Basis for Performance-Based Grant Systems and Selected International Experiences." Background Paper, National Stakeholder Workshop, Nepal, Kathmandu. Tendler, Judith, with Rodrigo Serrano. 1999. The Rise of Social Funds. WhatAre They a Model Op Cambridge, MA: Massachusetts Institute of Technology, Department of Urban Studies and Planning. UNCDF. 2005. Delivering the Goods. Building Local Government Capacity to Achieve the Milliennium Development Goals. A Practitioner 's Guide @om UNCDF Experience in Least Developed Countries. New York: UNCDF. Van Domelen, Julie. 2000. "Anchoring Programs with Autonomous Funds and Existing Government Agencies." Community-Driven Development Issue Paper, World Bank, Washington, DC. World Bank. 2001, Decentralization in the Transition Economies: Challenges and the Road Ahead. Washington DC: The World Bank, PREM, ECA. . 2002. Social Funds: Assessing Effectiveness. Washington, DC: World Bank, OED. . 2004a. Designing Social Funds: Key Questions and Options. Washington DC: The World Bank, OED. . 2004b. "Armenia Public Sector Modernization Document." Project Appraisal Document, World Bank, Washington, DC. ,2004~.WorldDevelopment Report 2004. Washington DC: The World Bank. . 2005a. The Effectiveness of World Bank Support for Community-Based and - Driven Development. Washington DC: World Bank, OED. . 2005b. "Armenia: (De)centralization of Public Service Delivery and Local Governance." Draft, World Bank, Washington, DC, September 2005. . 2006a. "Strengthening Bank Group Engagement on Governance and Corruption." IMF-WB, September 8. . 2006b. "Serbia Social Assistance and Child Protection Note." Human Development Sector, ECA Region, The World Bank, Washington, DC. .2006c. Global Monitoring Report, 2006. Washington DC: The World Bank, 61