Social Funds 33805 Innovations Notes Integrating Social Funds into Local Development Strategies FIVE STORIES FROM LATIN AMERICA RODRIGO SERRANO The steady movement towards decentralization that Latin America has experienced in the last clientelism. The new government is seeking ways decade, often referred to as the "quiet revolution" to revert this. Finally, many governments in the has led governments and donors to rethink the region are reforming SFs into instruments that role that Social Funds (SFs) should play in promot- support local development. ing local development. While SFs had been rela- tively successful in building local infrastructure, This Note gives a quick overview of how these insufficient integration with public sector systems reforms are unfolding in five SFs in Latin America (both national and local) had raised well founded (see Table 1) and highlights some features of the concerns about institutional and investment sus- emerging models. It shows that many SFs are tainability. The challenge for governments going working closely with local governments. For through the process of decentralization was to these SFs the challenge is no longer whether decide whether to institutionalize SFs as a long- they undermine local governments or not term instrument for promoting local development but rather how they can become an effective or to disband these institutions. instrument of the country's decentralization ernments, communities, and sectoral agencies policy--i.e., how their interactions with local gov- A quick survey of the regional experience reveals three broad kinds of responses to this challenge. advance the decentralization policy objectives and a Some governments have opted to close down SFs, more balanced approach to local development. transforming them into pure fiscal transfers. Mexico, for instance, abandoned PRONASOL and While the enabling environment for decentraliza- put in place formula-driven, earmarked transfers to tion in these five countries has significant room for municipalities known as Ramo 33. Others have improvement (e.g., clarifying responsibilities across opted to ignore this challenge. Until recently, for levels of government), when compared to most instance, Guatemala had multiple overlapping SFs countries in Africa and Asia we see that basic that coexisted in an environment of patronage and building blocks are already in place. In all five coun- tries local governments have elected authorities, Table 1. Five SFsFondo in Latin America Working with Local Governments Nicaragua de Inversión Social de Emergencia (FISE) Honduras Fondo Hondureño de Inversión Social El Salvador Fondo de Inversión Social para el Desarrollo Local (FISDL) Bolivia Fondo Nacional de Inversión Productiva y Social (FPS) Peru Fondo Nacional de Compensación y Desarrollo Social (FONCODES) T H E WMO R L D B A N K SVEOPLTUEM B E R 2 0 0 5 E 3 N O . 1 Box 1. Bolivia: Social Funds as an Instrument for Managing Conditional Transfers to Local Governments In 2000, the government of Bolivia merged the earlier social investment funds into the FPS which became the single win- dow for channeling matching grant resources to municipal governments. FPS was the main instrument to implement the central government's anti-poverty policy called National Compensation Policy (NCP). The NCP aimed to rationalize fis- cal transfers to municipalities from government and donors, to introduce progressive criteria and a pro-poor bias to intergovernmental finances and to establish a mechanism to reconcile local demands with national priorities through a matching grant system of "relative prices." Government created the Unified Directorate of Funds (DUF) to approve and oversee the plans and budgets of the FPS and the municipal credit fund (FNDR), and to promote synergies and coordina- tion among these two funds and with the rest of the public administration. While these reforms suffered from a number of design and implementation problems, the overall assessment is that both the creation of FPS and DUF moved the system in the right direction in terms of rationalizing the government's invest- ment policy. One lesson that comes out of this experience is the need to plan the merger processes better--while FPS was able to go back recently to the funds' previous level of administrative efficiency, during the transition the "internal processes used by the FPS became so complicated that the number of days that it took from the moment of submitting a subproject to the start of its execution grew to 490 days." (see PAD Bolivia-PSAC for Decentralization II). and access to a local tax revenue base as well as to aernment. purpose transfers from the central gov- national priorities. Bolivia and El Salvador reforms toInempower citizens in local governance general all countries there are ongoing were the first examples of this trend, and the other three countries are moving in a similar and service provision, as well as awareness that direction. (see Box 1) decentralization should proceed in a context of fis- · Second, to be the implementation arm of a cal responsibility. local investment national program for poverty reduction, focused on strengthening local gov- The increased commitment to decentralization in ernment and community capacities for plan- these five countries has signaled the end of SFs as ning, financing, and managing local infrastruc- ad-hoc structures for financing and managing local ture services in a participatory and infrastructure investments. Through different accountable manner. (see Table 2) reform strategies governments are integrating SFs into the institutions supporting the governance This Note highlights selected aspects of the expe- and provision of services at the local level. These rience of these five SFs in supporting the planning, reforms--which involved either mergers of funds, financing, and implementation of local investments absorption by line ministries, or the institutional- (see Table 2 for a summary). ization of SFs as permanent agencies with clearer and permanent mandates--are ongoing and unfin- PARTICIPATORY MUNICIPAL PLANNING ished.that to rolesFirst, aregradually However, there is a convergence in the All five countries have developed statutory regula- · being assigned to SFs: tions mandating Participatory Municipal Planning conditional, matching grant mechanisms that transform these funds into (PMP) processes, and in all countries Social Funds leverage municipal investments toward only finance investments that come out of such plans. This is a major change from the previous sit- T H E WMO R L D B A N K SVEOPLTUEM B E R 2 0 0 5 E 3 N O . 1 Table 2. Social Funds Support to Local Development in Five Countries in Latin America Planning Financing Implementation Who has the main Do projects responsibility for Are LG and come from Statutory Do LG implementation communities Participatory Guidelines manage Allocation (contracting, involved in Country Municipal Plans? for PMP the funds? Mechanism supervising, etc)? maintenance? Honduras Yes Municipal Yes. accredited LG Only Formula based, Community Maintenance Plans, Yes. Municipal Associations) Municipal Accredited (LG, FHIS Development (Poverty Map) Strategic Plans transfer to Some LG also Organizations Maintenance Comm groups Non-accredited (SF) Nicaragua Yes Municipal Yes. LG Only accredited (Poverty Map). Formula based Accredited (LG). Preventive Yes. FISE Planning System Some LG also Transitioning to Non accredited Organizations Maintenance Community Maintenance Fund transfer to reflecting matching grants (SF) Comm groups sectoral priorities. Peru Yes Yes. Participatory Only accredited (Poverty Map) Formula-based Community-LG LG or Joint Yes FONCODES Budgeting Law LG Teams El Salvador Yes Minimum Criteria Yes. All LG that win Biddings Competitive LG Yes FISDL guidelines Bolivia Yes Popular Yes. payment and FPS LG approves indicative Formula based, LG 25% reduction in Yes. Also there is a FPS Participation Law pays contractors allocation municipal been used yet) O&M (it has not reward for good counterpart as a (poverty map). Matching grant reflecting sectoral priorities LG: Local Governments uation in which SF financed isolated projects pre- sented mostly by individual communities and national regulations were largely influenced by the approved centrally by the Social Fund. This previ- experience of the Social Funds with PMP ous approach was criticized because of its poten- processes. The almost decade-long experiment of tial to undermine local governments, and its lack of thesemethodologiesFunds with participatory plan- ning Inthree Social transparency and downward accountability. · contexts where local governments are suggests the following lessons: While in Bolivia and Peru Social Funds changed elected, autonomous entities, social funds their project selection in response to the passing should strive to finance projects that come out of new national laws regulating municipal planning of participatory municipal plans rather than (e.g. the Law of Participatory Budgeting in Peru), in isolated community demands. While initial the three Central American countries the new conditions in some countries may be adverse to institutionalizing PMP, its potential benefits T HSVEOPLTO R L D B A N K E W U M E 3 N O . 1 E M B E R 2 0 0 5 Graph 1. The Municipal Planning System in Nicaragua1 MDP Municipal Strategic Planning Strategic Guiding Instrument, MUNICIPAL long term (8 years) and DEVELOPMENT continually updated PLAN Programming of MIP instrument, rolling, annually Four year programming Municipal Investments Municipal Investment Program (mediumupdated and short term) PMP Annual instrument to manage Annual Operational Plan-- Municipal and control local public Municipal Budget Budget by (Short term) spending Program (in terms of allocative efficiency, inclusiveness, and accountability) make it worthwhile to take Salvador, have guidelines establishing a set of minimum criteria that municipal plans should · gradual steps towards its adoption. Ifnomunicipalitiesby meet but without specifying a unique method- canstatutory instruments on PMP, social funds lack such plans, and there are ology for fulfilling them. FISDL provides simple giescontribute and comparative information on municipal and jumpstarting the interest on the issue. developing PMP methodolo- performance for each criteria as benchmarks The more these instruments are designed for municipalities to identify where they need based on the "municipal" rather than the to improve. "social fund" perspective, the more likely it will be that they are adopted as part of the FINANCING · normative framework regulating PMP. Once the government has passed the statu- Until a couple of years ago most of the SF funding tory regulations for PMP social funds can play was managed centrally by the SF. Currently these an important role as the implementing agency five SF are transferring all or a substantial portion managing the process through which assis- of their resources directly to local governments. In tance is provided to LG and communities to Honduras and Nicaragua some LG are transferring funds even further down to community groups. · develop these plans. There are different strategies for developing a Different contexts and policy objectives have led normative PMP framework. Some countries, these five social funds to develop different rules like Nicaragua, chose to develop a framework for allocating and accessing these transfers. that details processes, instruments, and insti- tutions (see Graph 1). Other countries, like El Bolivia's National Compensation Policy, imple- mented through the FPS, has two objectives. To 1INIFOM y AMUNIC (2002). Sistema de Planificación Municipal. Guia Ilustrada. Managua: INIFOM. T H E WMO R L D B A N K SVEOPLTUEM B E R 2 0 0 5 E 3 N O . 1 Box 2. FISDL's Program to Mobilize Financial Support from Associations of Salvadoreans Abroad Over the last couple of years, FISDL has been working to stimulate the contribution of associations of Salvadorans living abroad for financing of local development projects. The goals have been to: · Recognize and support the social commitment of Salvadorans living abroad, strengthening the communication links · Develop a program that strengthens the contribution of Salvadoran communities abroad for local development works between them and the different private and public national actors. · Strengthen the institutional structure that works with Salvadoran communities abroad. in their communities of origin. Associations of Salvadorans can participate not only in the 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 mil- lion dollars per bidding to co-finance projects where these Associations participate. Since September of 2,000, about 30 Associations of Salvadorans abroad have mobilized US$2.13 million or about 19% of the US$11.45 million invested in 45 development projects. Another US$2.28 million was leveraged from municipal gov- ernments, US$ 6.9 million from FISDL, and US$115,000 from Ministries. compensate for differences in poverty levels across municipalities, FPS defines indicative munici- effective in influencing municipal priorities, pal allocations following a poverty and population- changes in sectoral investment allocations at based formula. To gear local investments towards the municipal and FPS levels show different national sectoral priorities for poverty reduction, directions (eg, education went from 51% to FPS finances projects following a system of "rela- 12% in 2002­2004 in FPS, and from 17% to tive prices" in which the co-financing rate is lower 23% in 1999­2003 in municipalities) (Frank for sectoral investments that are a central govern- 2004). Part of the problem appears to be the ment priority and for poorer municipalities. Some resistance of donors and departmental lessons on the performance of these two objec- governments (prefecturas) to honor the tivesWhile are: National Compensation Policy, establishing · larger share ofmunicipalities thanreceived parallel financing mechanisms that undermine counterparts, itFPS been less than what was so farpoorer the NCP. expected according to the indicative allocation hasresourceshave their richer a El Salvador's FISDL has a competitive bidding formula. The lower capacity to manage invest- allocation mechanism that uses the share of ments (and the absence of mechanisms to counterpart funding as the main criteria for pri- compensate for them) hindered poorer oritizing investments in each bidding round. municipalities from fully executing their indica- FISDL finances only municipal infrastructure pro- jects above US$50,000 as a way of complement- · While the analysis that is available makes it tive allocation2 ing the focus on small-scale projects that LG impossible to verify whether the FPS has been tended to finance with their own resources. Contrary to what some feared, the process did Municipales. Draft. La Paz. 2 Aguilar Perales, Javier and Fernando Ruiz Mier (2004). Hacia Una Nueva Politica de Transferencias Condicionadas a Gobiernos T HSVEOPLTO R L D B A N K E W U M E 3 N O . 1 E M B E R 2 0 0 5 not discriminate against poorer and smaller municipalities--89% of the funds were awarded form of accreditation system establishing the ex- to municipalities in the poorest three quintiles. ante conditions that LG need to fulfill in order to As an incentive for good municipal practices, LG receive the transfer (see below). have to fulfill a set of conditions (related to par- ticipatory planning, investment efficiency and Some social funds have experiencing with new transparency) to be eligible to participate in the instruments to bring greater transparency and effi- bidding. During the last few years, FISDL has ciency in their financial transactions with LG. In placed a special emphasis in stimulating associa- Honduras, for instance, FHIS is piloting a Trust tions of Salvadorans living abroad to partner with Fund (Fideicomiso) with a national private bank municipal governments in funding local develop- which is responsible for making all the financial ment projects (Box 2). transactions through its extensive branch network. Similarly to Bolivia, the other three SF also make INVESTMENT MANAGEMENT municipal allocations on the basis of a formula CAPACITIES incorporating population and poverty indicators. While so far they do not have clear mechanisms to While the rationale for decentralizing resources influence the sectoral allocations at the municipal through SF in countries like Bolivia or El Salvador level, this appears to be changing. In Nicaragua, for was to influence local spending decisions, in coun- instance, the gradual increase in general purpose tries like Nicaragua and Honduras the main ratio- transfer to municipal governments in a context of nale was to gain allocative and productive effi- unclear transfer of responsibilities and concern for ciency by delegating resources and decision-making fiscal neutrality is leading government to convert power to local actors. In the latter cases, SF FISE into a matching grant scheme that links local entered into temporary principal-agent arrange- spending decisions with the national poverty ments with local governments under the assump- reduction goals. All three SF have in place some tion that relying on local capacities would produce services in a cheaper, faster, and more responsive Box 3. Efficiency and Accountability Gains of the Decentralized Approach in Nicaragua An evaluation of FISE's Pilot3 concluded that the decentralized approach to investment management performed better than the centralized model in terms of efficiency, effectiveness, and transparency. · The pace and volume for infrastructure building was much higher than the historical record in the same municipalities · Project Beneficiary Committees were able to participate in local government's bidding process, something which was as well as in comparable municipalities during the same period. · Contracting and procurement was 40% faster and 7% cheaper. The latter was a result of greater competition among absent in the centralized model. · While real costs in the decentralized model were 4% lower than estimated costs, in the centralized model real costs contractors in the decentralized model. for similar projects were 6% higher than estimated ones. Construction times for similar projects were 40% faster in the decentralized than in the centralized model. In general, the decentralized model compares favorably in terms of number of fines, renegotiation of costs, and timeframes for project completions. T H E WMO R L D B A N K 3Inversión Grun, Stephen (2002). Evaluación del Programa de Descentralización de Gestión y Manejo del Ciclo de Proyectos del Fondo de Social de Emergencia (FISE). Managua. SVEOPLTUEM B E R 2 0 0 5 E 3 N O . 1 Box 4. Fostering "Mancomunidades" (Municipal Associations) to Reach Municipalities with Weak Capacities Frustrated by the difficulty to decentralize its project cycle to a larger number of municipalities due to their weak man- agerial capacities, in 2002 and in partnership with KfW and GTZ, FHIS began a pilot with the Municipal Association of CAFEG. CAFEG which groups 7 of the poorest municipalities in the country located in the Lempira Department, was an attempt by these municipalities to share the costs of technical assistance by jointly financing a technical unit serving all seven municipalities. After the first year, in which FHIS subsidized the unit's expenses, municipalities have honored their initial commitment to fully cover the Unit's costs out of their transfers. So far, CAFEG has successfully managed 31 investment projects for a total of approximately US$800,000. It has a lower overhead than FHIS (6.7% instead of 10%). The technical unit has 4 employees (1 civil engineer, 1 administrator, and 2 social promoters) in addition to equipment, an office, and transportation. Encouraged by the success with CAFEG, FHIS has already signed agreements with 19 Municipal Associations (known in Honduras as Mancomunidades). Similar experiences are taking place in many other countries in the Region. and accountable manner. Preliminary evidence of the experience in Nicaragua supports these argu- Not all countries, however, have a gradual transfer ments (Box 3). strategy. Both El Salvador and Bolivia have dele- gated the project cycle to all local governments Decentralizing the project cycle is not an easy irrespective of their capacities. While in El Salvador task given that most local governments have weak this approach does not seem to have penalized managerial capacities. This is why some SF (in poorer municipalities in Bolivia it does. Comparing Peru, Honduras, Nicaragua) have followed a grad- these approaches should provide useful lessons. ual approach where municipalities have to prove that they have the minimum capacities for manag- One of the key roles that SF play under the decen- ing investments before assuming their new tralized model is to develop local government and responsibilities. While in Peru the accreditation community capacities to manage local investments. system is binomial (in or out) social funds in Two of the areas where social funds usually have a Honduras and Nicaragua have a multi-tier gradua- comparative advantage are, first, in developing and tion system that allows municipalities to take on assisting in the implementation of systems and pro- more or less investments responsibilities accord- cedures for local investment management. The ing to their managerial capacities. For municipali- main challenge here has been to design a system ties with low capacities, social funds are support- that is not social fund-specific but instead general ing the formation of Municipal Associations which and applicable to all municipal clients. In Nicaragua, allow several municipalities to afford the shared the pilot phase revealed that local governments provision of technical services (Box 4). In the had different project management systems for municipalities that have not been accredited, the each donor or public agency, and that FISE's sys- social fund manages investments in its traditional tem was only adding to that fragmentation. To fix way (either through contractors or community this problem, and at the request of the municipal groups). development agency, FISE has developed a generic "municipal" investment management system. T HSVEOPLTO R L D B A N K E W U M E 3 N O . 1 E M B E R 2 0 0 5 Another system-wide innovation that FISE has brought to managing local investments is the In cases of program mergers, what transition establishment of a Preventive Maintenance Fund4. strategies can minimize the risks of institu- Municipal Matching grants. How to design tional paralysis? A second innovation has to do with bringing CDD · practices in the repertoire of local government matching grant systems that ensure the management tools. SF are trying different variations poverty focus of municipal investments? What of the same idea of having local governments and kind of M&E system is needed to measure their communities be jointly responsible for the con- governments to reconcile local participatory Planning. What systems and procedures allow effectiveness in influencing municipal priorities? struction and operation and maintenance of local · infrastructure. In cases like Honduras this means that the local government subcontracts and super- planning with sectoral planning? And within vises the Community Project Committee which local planning, to resolve the tensions between manages the resources for project implementation. popular preferences and technical knowledge? In cases like Peru it means having a mixed project How can national governments ensure that committee formed by three representatives from local participatory processes include the poor- the community (president, secretary, treasurer) est and the most vulnerable groups? and one from the municipality (fiscal). · Accreditation and capacity building. What was the performance in those Social Funds that fol- lowed a gradual approach to delegation of A LEARNING AGENDA responsibilities versus those that delegated to As decentralization processes deepen in Latin all LG at the same time? Why, among the lat- America, governments are confronted with the ter, some instruments seem to penalize poorer need to either reform social funds or to shut them LG (Bolivia) while others do not (El Salvador)? down. The experience of the five social funds · LG-Community Partnerships in implementation. reviewed here shows that the trend has been With local governments assuming a more cen- towards reforming them to play a clearer role in tral role in the provision of services, the chal- the overall decentralization policy. Because more lenge becomes how to prevent them from and more countries are facing the same challenge repeating the centralist tendencies that have these ongoing reforms provide us with an oppor- plagued national agencies. Partly to address tunity to learn about what works and what does this risk, SFs are giving incentives for LG-com- not. Some of the pending questions are: munity partnerships for the implementation of · Institutionalization. What kind of reforms work small-scale investments. Do these partnerships better to integrate social funds into public sec- produce better outcomes than without them? tor systems? What lessons do mechanisms Do differences in their design matter? like the Directorate of SFs tried by Bolivia · capital created or strengthened by social funds Social Capital. How to prevent that the social (and now being emulated by Guatemala) leave about the effectiveness of such mechanisms? is not project-specific but can be used toward broader developmental purposes? 4Experience. For more information see Serrano and Warren (2003). Re-engineering SFs for Local Governance: The Central American Spectrum (Fall Issue). "Social Funds Innovation Notes" are published informally by the Social Funds thematic group of the Human Development Network ­ Social Protection. For additional copies, contact the Social Protection Advisory Service, The World Bank, 1818 H Street, NW, Washington, DC 20433, USA, E-mail: socialprotection@worldbank.org. Copies are also available on-line through the Social Funds website http://www.worldbank.org/socialfunds. T H E WMO R L D B A N K SVEOPLTUEM B E R 2 0 0 5 E 3 N O . 1