Paper No. 96/ June 2006 Exploring Partnerships between Communities and Local Governments in Community Driven Development: A Framework Keith McLean Louis Helling Jana Orac Rodrigo Serrano SOCIAL DEVELOPMENT PAPERS Community Driven Development Paper No. 96/ June 2006 Exploring Partnerships between Communities and Local Governments in Community Driven Development: A Framework Keith McLean Louis Helling Jana Orac Rodrigo Serrano Copyright © 2006 The International Bank for Reconstruction and Development 1818 H Street, N.W. Washington, D.C., 20433 USA All rights reserved Manufactured in the United States of America First Printing June 2006 Documents in the Social Development Department's the Working Papers Series disseminate the findings of work in progress to encourage discussion and the exchange of ideas on social development issues. Papers in this series are not formal publications of the World Bank. The findings, interpretations and conclusions herein are those of the author(s) and do not necessarily reflect the views of the International Bank for Reconstruction and Development/The World Bank and its affiliated organizations, or its Executive Directors, or the governments they represent. The papers carry the names of the authors and should be cited accordingly. The series is edited by the Social Development Department of the Environmentally and Socially Sustainable Development Network of the World Bank. To request copies of the paper or for more information on the series, please contact the Social Development Department. Papers are also available on the Social Development Department website: Web : http://www.worldbank.org/socialdevelopment (see "Publications" in the navigation menu) E-mail: sdcommunications@worldbank.org Social Development Department The World Bank 1818 H Street, NW Washington, DC 20433 Fax: 202-522-3247 Printed on Recycled Paper CONTENTS Acronyms and Abbreviations................................................................................................................................. iv Acknowledgements................................................................................................................................................... v Executive Summary.................................................................................................................................................. vi Study Objectives................................................................................................................................................... vi Methodology ........................................................................................................................................................ vi The Analytical Framework................................................................................................................................vii Lessons from Four Countries............................................................................................................................vii Conclusions.........................................................................................................................................................viii Operational Implications ..................................................................................................................................viii The Research Agenda.......................................................................................................................................... ix Next Steps: The Road Ahead...............................................................................................................................x Understanding the Decentralization Context ....................................................................................................1 The State of the Art................................................................................................................................................2 Getting Past the Hesitation...................................................................................................................................3 II. Objectives of the Report........................................................................................................................................4 III. Methodology .........................................................................................................................................................5 IV. Some Key Concepts..............................................................................................................................................6 Decentralization .....................................................................................................................................................6 Local Government .................................................................................................................................................6 Community-Based Organization.........................................................................................................................7 Coproduction..........................................................................................................................................................8 Accountability ........................................................................................................................................................9 Opportunity Space...............................................................................................................................................10 V. Strengthening Accountability and Coproduction Relationships in Different Local Government Contexts..............................................................................................................12 VI. The Analytical Framework................................................................................................................................14 Characterize the Enabling Environment for LGs and CBOs..........................................................................14 Identify the Opportunity Space .........................................................................................................................15 Evaluate the Scope for Accountability and Coproduction.............................................................................17 VII. Synthesis of Case Study Lessons.....................................................................................................................22 Variations in Opportunity Space.......................................................................................................................22 i Mixed Results Using CDD to Alter the Enabling Environment....................................................................23 Coproduction and Accountability Relationships in Practice: Examples from the Philippines, Zambia, and Tanzania.................................................................................23 Interplay of Shifting Opportunity Space and New Partnership Opportunities: Examples from Nicaragua ........................................................................25 INIFOM 2 and FISE 4 ...........................................................................................................................................27 VIII. Conclusions ......................................................................................................................................................29 IX. Operational Implications...................................................................................................................................31 X. The Research Agenda..........................................................................................................................................33 XI. Next Steps: The Road Ahead............................................................................................................................34 XII. Framework Appendix.......................................................................................................................................35 Step 1: Assessing the Enabling and Constraining Factors for Empowering CBOs and LGs....................36 Step 2: Mapping the Opportunity Space for LG-CBO Accountability and Coproduction Arrangements...........................................................................................38 Step 3: Identifying the Operational Implications of the Opportunity Space for Improving Local Accountability and Coproduction via CBOs and LGs .....................................................39 XIII. Case Study Appendixes ..................................................................................................................................41 Appendix 1. The Philippines: The Challenges of Implementing Devolution ................................................41 Local Government Context.................................................................................................................................41 CBO Context.........................................................................................................................................................43 Mapping the Opportunity Space ......................................................................................................................44 CDD Projects and LG-CBO Partnerships .........................................................................................................44 Mindanao Rural Development Project (MRDP)..............................................................................................44 Kalahi-Community Integrated Delivery of Social Services (CIDSS).............................................................46 Filling the Opportunity Space............................................................................................................................48 Appendix 2. Zambia: Pushing for Change under a Stalled Decentralization Framework...........................50 Local Government Context.................................................................................................................................50 CBO Context.........................................................................................................................................................52 Mapping the Opportunity Space.......................................................................................................................53 CDD and LG-CBO Partnerships: The Zambia Social Investment Fund (ZAMSIF)....................................54 Appendix 3. Nicaragua: Expanding the Opportunity Space for Local Partnerships through CDD Operations..................................................................................................57 Period 1: Missed Opportunities (1990­1994)...................................................................................................57 Period 2: Piloting Innovations that Push the Boundaries of the Policy Environment (1996­2000) ....................60 Period 3: Using Projects for Policy Reform (2001­2004).................................................................................66 ii Appendix 4. Tanzania: Testing Partnerships within a Decentralization Patchwork.....................................71 CBO Context.........................................................................................................................................................73 Mapping the Opportunity Space.......................................................................................................................74 CDD and LG-CBO Partnerships: The Second Tanzania Social Action Fund (TASAF II) .........................75 Bibliography..............................................................................................................................................................78 Boxes and Tables: Box 1: Community-Based Organizations ...............................................................................................................8 Box 2: Bolivia's Law of Popular Participation .....................................................................................................11 Box 3: Inclusive Decision Making to Improve Local Governance in Macedonia............................................30 Box 4: Municipal Transfers Law............................................................................................................................68 Table 1: Coproduction between LGs and CBOs with Respect to Devolved Functions.....................................9 Table 2: Characterizing the Enabling Environment............................................................................................14 Table 3: Opportunity Space (by dimension) .........................................................................................................15 Table 4: Accountability Relationships....................................................................................................................17 Table 5: Coproduction Relationships.....................................................................................................................19 iii ACRONYMS AND ABBREVIATIONS APL adaptable program loan CBO community-based organization CDD community-driven development CFAD community funds for agricultural development CIAC Community Investment Advisory Committee CMU country management unit CSC Community Supervision Committee ESW economic and sector work FISE Nicaragua Emergency Social Investment Fund FONDEM Municipal Development Fund (Nicaragua) IA Intercommunity Assembly INIFOM Nicaraguan Institute for Municipal Development KALAHI-CIDSS Kapitbisig Laban Sa Kahirapan­Comprehensive and Integrated Delivery of Social Services LG local government LGU local government unit LMU local monitoring unit M&E monitoring and evaluation MDG millennium development goals MIP municipal investment plan MIVF municipal intervillage forum MPS municipal planning system MRDP Mindanao Rural Development Program MTU municipal technical units NGO nongovernmental organizations O&M operation and maintenance PAD project appraisal document SRP social recovery projects TA technical assistance TASAF Tanzania Social Action Fund WDR World Development Report ZAMSIF Zambia Social Investment Vice President Ian Johnson Sector Director Steen Lau Jorgensen Sector Manager Caroline M. Kende-Robb Task Team Leader Keith McLean iv ACKNOWLEDGEMENTS This report contributes to the ongoing process of strengthening the linkages between CDD and decentralization. It represents an additional effort to strengthen the knowledge base and tools on how to maximize the community interface with local governments within the context of local development. It builds on past or ongoing work by many regions on the challenges of scaling up CDD; on partnerships between community-based organizations (CBOs) and local governments (LG); and on mechanisms that strengthen overall local governance and community empowerment. It presents an analytical framework for assessing entry points, opportunities, and constraints to strengthening LG-CBO partnerships around coproduction and accountability. Examples are drawn from case materials in four countries: Zambia (the Zambia Social Investment Fund, or ZAMSIF), Tanzania (the Tanzania Social Action Fund, or TASAF I and II), the Philippines (the Mindanao Rural Development Project, or MRDP, and the Kalahi­CIDSS Project), and Nicaragua (the Nicaragua Emergency Social Investment Fund, or FISE I-IV, and the Rural Municipalities Project, or PROTIERRA I and II). This report was prepared by a team comprising Keith McLean (SDV, Task Team Leader), Rodrigo Serrano (HDNSP), Louis Helling (consultant), and Jana Orac (consultant). The authors would like to thank all Bank staff, country project managers, and other national stakeholders who took the time to share their insights and experiences with us. This contributed significantly to the quality of the final report. All remaining errors are ours. Finally we would like to acknowledge the excellent information and advice provided by Danielle Christophe, Amer Hasan, Reiner Forster, Jeffrey Ramin, Andre Herzogg, Daniel Owen, Susanna Shapiro, and other colleagues in SDV. Social Development Economist, Community Driven Development Social Development Department, ESSD The World Bank 1818 H St, NW, Washington DC 20433 Telephone (202) 473-5656, Fax (202) 522-3247 v EXECUTIVE SUMMARY This report presents a framework with which to analyze institutional options for partnering community- based organizations (CBOs) with local governments (LGs) in community-driven development (CDD), illustrated through four country cases: Zambia, Tanzania, Nicaragua, and the Philippines. The report also addresses operational implications, a research agenda, and next steps. In recent years, CDD interventions have increasingly sought to integrate local governments more extensively into the basic design. CDD is an approach that supports collective action, community empowerment, and demand-driven local service delivery. It empowers the poor and vulnerable by placing them in the driver's seat with respect to decisions on a range of development interventions intended to improve their economic and social conditions. Importantly, the approach continues to evolve and mature, and efforts have now turned to addressing fundamental questions of fiscal and institutional sustainability and scaling up. Lessons from internal and external review of efforts toward scaling up have underscored the newly emergent consensus that integrating rather than bypassing local governments is important for long-term sustainability and the ability to scale up CDD. This new awareness has important implications for the evolution of CDD and controls how CDD is designed in different decentralization contexts. Study Objectives The intended audience for this report consists of task and country teams, other donors, and especially external clients (governments and project managers) with responsibility for designing CDD programs, reforming intergovernmental systems and decentralization efforts, local government support operations, social accountability initiatives, and empowerment programs. Its conclusions emerged as a result of ongoing dialogue between country clients and donor staff working on related types of operations on how best to move forward on a common agenda. It was motivated by increasing numbers of client country requests to support their efforts to incorporate local governments (LG) more systematically into national CDD strategies. This report helps readers to understand local government structures and incentives and to identify the key local government reforms needed to strengthen coproduction1 and accountability2 partnerships with community-based organizations (CBOs). The report suggests ways to diagnose adequately the scope for partnerships (the opportunity space),3 and it draws lessons from four country cases4--Zambia, Tanzania, Nicaragua, and the Philippines--on how the scope can be assessed in different contexts. Methodology A three-phase approach was adopted for assessing the enabling environment, opportunities, and constraints to deepening LG-CBO partnerships in CDD: 1. A framework is presented that characterizes the enabling environment for LGs and CBOs. 2. The framework is applied to four countries--Zambia, Tanzania, Nicaragua, and the Philippines-- each of which had at least one CDD operation attempting LG-CBO partnerships. It is used to characterize the country environment and assess (i) the opportunities for local partnerships and (ii) the types of feasible accountability and coproduction relationships. 3. Finally, the report draws lessons, identifies research gaps and operational implications, and suggests next steps. 1. Coproduction refers to shared responsibility among multiple stakeholders for delivery of certain goods and services. It usually involves both joint financing and implementation. 2. Accountability exists when an agent can be made to answer to the intended beneficiaries for a set of deliverables. 3. Opportunity space refers to the range of possibilities offered by the enabling environment without altering the fundamental institutional structures pertaining in a given context. 4. These cases were chosen because of ongoing client efforts to strengthen such partnerships. vi The Analytical Framework The analytical framework designed to assess institutional options and opportunities to link CBOs and LGs consists of three elements. Characterize the Enabling Environment for LGs and CBOs Are local governments and CBOs enabled or constrained? In a given country context, this assessment reviews the enabling environment for CBOs and LGs across each of the three dimensions--political, fiscal, and administrative--reviews the legal and regulatory context, and attempts to provide an overall picture of initial conditions. The terms enabled and constrained refer to the extent to which local institutions can exercise their envisaged roles as citizens' groups and governments without substantial constraints on their basic functioning. Identify the Opportunity Space What is the initial scope for partnering LGs and CBOs? Assessment of the enabling environment for LGs and CBOs establishes the opportunity space.. In other words, by looking at the degree to which the institutional environment is enabling or constraining along the administrative, political, and fiscal dimensions, it is possible to identify the, strengths, weaknesses, and comparative advantages of LGs and CBOs in a specific context. On this basis, institutional arrangements for effective local partnerships can be identified. Evaluate the Scope for Coproduction and Accountability What does the opportunity space assessment imply about coproduction and accountability? The enabling environment for CDD maps to a set of implications for coproduction and accountability arrangements. Here the relative strengths and weaknesses of CBOs and LGs along the three dimensions become quite important. Strongly democratic and participatory local governments operating in highly decentralized fiscal systems, for example, may be well positioned both to articulate and to respond to citizen demands. In the context of well institutionalized community empowerment schemes, CBOs may similarly possess both supply- and demand-side capacity. In many countries, however, the constraints are more pronounced, as are the potential advantages of LG-CBO partnerships. Lessons from Four Countries The opportunity space and operational responses varied across the four country contexts. Even within the same country, projects responded differently to the opportunity space. This variation both revealed inconsistencies in the approach and yielded lessons for other contexts. · The Zambia Social Investment Fund (ZAMSIF) tried to exceed the limits of the opportunity space with its District Investment Fund (DIF). This component ran into substantial implementation difficulties, however, due to the lack within the country of a sustained decentralization dialogue supporting the decentralization reform agenda needed to strengthen the capacities of district governments. · Nicaragua succeeded by recognizing and addressing LG-strengthening issues at the macrolevel and by piloting sequenced innovations within two types of projects: a social fund (the Fund for Emergency Social Investment, or FISE) and a demand-driven rural investment fund (the Rural Municipalities Project, or PROTIERRA). These two project types evolved and adapted over time in response to the changing enabling environment and opportunity space. · In the Philippines, which has a good local government structure, the opportunity space was largely exploited by the Mindanao Rural Development Project, but with inadequate emphasis on vii the community-driven aspects. The second project reviewed, the Kalahi-CIDSS, largely underexploited the scope for coproduction synergies with local governments but substantially strengthened the citizen interface with LGs through social accountability mechanisms. · In Tanzania, a local government support project aimed at strengthening district LGs is being implemented in parallel with the second Tanzania Social Action Fund (TASAF II) in complementary but somewhat incomplete ways. The country dialogue is underdeveloped with respect to rationalizing the structure of fiscal transfers from districts to villages. Important reforms are needed that that could substantially strengthen the scope for coproduction. Conclusions The opportunity space concept provides a useful means for exploring the potential scope for LG-CBO partnerships. By applying the framework and the concept of opportunity space, it is possible to assess the enabling environment for the kinds of LG-CBO partnerships with which this study is concerned. Application of the concept must be flexible, however, as countries don't always fit neatly into the enabled or constrained categories but sometimes fall into the grey area between. Opportunity space should be used as the starting point, but this foundation can be viewed as dynamic, expanding with the right mix of macro- and microinterventions or contracting in response to intransigent constraints. Nicaragua and Zambia provide clear examples of dynamic expansion and enduring constraint, respectively. Detailed analysis of the macroconstraints on local governments, an active dialogue among key stakeholders, competition, and opportunities to tackle structural constraints seem to matter most when building from the opportunity space. Coproduction and accountability reinforce each other. The four country cases examined reveal both coproduction and accountability relationships, which seem to be well linked. In fact, it appears that the stronger the coproduction linkages, the more involved are the accountability relationships specified under the partnership arrangements. The weaker the decentralization framework, especially fiscal decentralization, the weaker the partnership possibilities. This was clearly demonstrated across the four cases: In Zambia and Tanzania, where the decentralization frameworks are weak, coproduction and accountability relationships tended to be weaker than in the Philippines and Nicaragua. Embedding the partnership approach into a decentralization dialogue is critical. The four cases show that, where decentralization frameworks are weak or nonexistent, establishing coproduction relationships that go beyond the natural (and limited) opportunity space, open up the enabling environment, and partner LGs and CBOs requires instituting a broad decentralization dialogue. Learning is not yet complete. CDD operations increasingly try to build strong relationships between communities and local governments. While the framework outlined here helps in understanding the main considerations and opportunities for this process, more lessons will evolve as a greater number of operations supporting CDD and LGs explore and exploit interface opportunities in earnest. Operational Implications Integration at the local level requires technical integration within client and donor teams. Whether the community level or LG is a black box depends on the perspective of the practitioner. To most CDD practitioners, the LG is an unaccountable, elite-captured black box. To most public sector management experts and decentralization practitioners, the community is a fuzzy, amorphous concept that does not really address the fundamentals of how governments and service delivery work. In reality, to improve viii local governance and services, partnerships are needed that coalesce these perspectives for mutual support and information sharing and that engage the private sector and other local actors. Country clients should coordinate donors to avoid duplication and contradictions. Lessons derived from situations in which donor coordination was lacking suggest that donors often put in place procedures that complicate, contradict, and duplicate one another's efforts. This can be very confusing and costly to local communities and LGs whose capacities are already limited. Given the weak track record on such coordination, however, countries may need to take the lead wherever they have a coherent vision of how best to organize local development partnerships. LG-CBO partnerships should not be designed without detailed analysis of the decentralization framework, including the intergovernmental fiscal system, and of LG capacities and constraints. This requires marshaling the relevant decentralization expertise to support the preparation of CDD operations and to offer suggestions on the scope for linkages. Local government support operations need to strengthen social accountability relationships to help realize allocative efficiency. Harmonizing the community-LG interface in client countries' and donors' portfolios provides some scope for harmonizing participatory methodologies for accountability arrangements. Social accountability tools such as citizens' report cards and participatory planning and budgeting help to complement formal public consultations (public meetings, for example) and deepen the quality of local decision making. This helps to better match citizens' preferences with budget allocations. Social accountability should be mainstreamed where local governments have discretion and adequate resources to respond to local communities. Stakeholder consultations and participatory planning methodologies should, in principle, be built into all local development operations and should also be consistent in type across a country portfolio of CDD operations. If local governments are legally and fiscally constrained, however, and therefore unable to respond to citizen demands, standardized social accountability is likely to induce significant frustration among the citizens. Cost-sharing arrangements between various stakeholders should be consistent among projects within a country. Except for pilot programs, communities and local governments should not be subjected to an array of cofinancing arrangements that vary from project to project. Where variability occurs, it should reflect the financial capacity/level of poverty of the specific communities and governments so that poorer target groups face lower financing burdens, based on defined subproject types or sectors. Careful and dispassionate monitoring and evaluation and impact assessments could help reconcile approaches. Project promoters tend to make competing claims about their plan's overall impact and effectiveness, in terms both of benefits accruing to communities and of integration with local governments. Where a portfolio contains multiple projects and harmonization is an issue, M&E could be treated as a public good and coordinated by the quality assurance team or the CMU. This would achieve the additional benefit of ensuring a set of common benchmarks against which to evaluate the effectiveness of and learning about the different operational modalities. Task teams could help to define the monitoring indicators with the quality assurance team, based on the development objectives of each project. The Research Agenda In the course of writing this report, three questions emerged that might usefully guide possible future research into the issues covered, thus helping to strengthen the overall agenda. · Under what conditions can CDD operations catalyze fundamental changes in decentralization policy, significantly strengthening the enabling environment for local governments? ix · Where the government remains highly centralized, what do-no-harm incremental steps could be incorporated into the CDD approach consistent with strengthening local partnerships? · ShouldgovernmentsandtheWorldBanktacklestructuralandcapacity flawsupfront,inadvanceof project preparation or as part of it, or should they proceed in parallel with complementary dialogues and sequenced remedial efforts? Next Steps: The Road Ahead Learn by doing through country pilots. Building on this report, country pilots are envisaged in Zambia, Angola, and the Philippines to support both clients and country teams and to further refine the overall analytical framework. Translate the framework into an assessment tool. To be used by clients as the basis for conducting the analysis of institutional options, this tool, planned for FY06, will overlap with the country pilots and will build on the lessons derived from the detailed country level assessments. Follow up economic and sector work and other research. Additional analytical work is planned for FY06 aimed at addressing issues of convergence between CDD and LG support operations. This ESW will also support the process of portfolio harmonization, which has emerged as an active challenge for many clients. Maintain a multidisciplinary, cross-sectoral approach and dialogue. This report was motivated by the interests of country clients, donors, and staff working on social funds, CDD, decentralization, social accountability, and empowerment issues. These efforts require a range of skills and perspectives, and it is important to maintain the dialogue among these varied specializations both to sustain momentum and to expand the possibilities for increased learning about and stronger integration of decentralization and community-driven development. x I. BACKGROUND In recent years, efforts at community-driven development (CDD) have increasingly sought to integrate local governments more extensively into the basic design of the interventions.5 CDD is an approach that supports collective action, community empowerment, and demand-driven local service delivery. It empowers the poor and vulnerable by placing them in the driver's seat with respect to decisions on a range of development interventions intended to improve their economic and social conditions. At approximately $2 billion in annual lending (FY2000­2005) to operations using the CDD approach (henceforth CDD operations), CDD represents an important line of business that has been mainstreamed across a range of single and multiple sector interventions. Importantly, the approach continues to evolve and mature, and efforts have now turned to addressing fundamental questions of fiscal and institutional sustainability and scaling up. Lessons from internal and external reviews of CDD scale-up operations have underscored the newly emergent consensus that integrating rather than bypassing local governments yields advantages for long-term success and sustainability.6 This new awareness has important implications for the evolution of CDD and its design for different decentralization contexts. This report aims to contribute to the body of knowledge available to clients by introducing a framework for assessing the kinds of coproduction and accountability partnerships possible between local governments and communities. Its flexibility suits it to use within different decentralization structures and CBO enabling environments, as illustrated by its application here to four country contexts in which clients sought to build sustainable local partnerships for service delivery between citizens groups and local governments. Understanding the Decentralization Context Not all decentralization contexts offer easy opportunities to work with local governments. In fact, in many client countries, little or no decentralization of government has occurred, beyond the traditional deconcentration7 of central government units. The variety of contexts ranging from deconcentration and weak civil society or CBOs to strong decentralization (substantial devolution) and strong civil society or CBOs creates a number of important strategic challenges for CDD practitioners and governments seeking to deepen links with local governments (LGs). Issues of sequencing and implementation emerge, raising such choices and trade-offs as whether to (i) strengthen communities first, then local governments; (ii) strengthen communities to create bottom-up pressure on local governments; (iii) empower citizens who would then create the bottom-up demand for greater democratization of subnational decision making (decentralization); (iv) create strong, accountable local governments that could sustain a participatory culture and strong citizen engagement beyond the life of a specific project(s); and (v) strengthen community support and local governments simultaneously by emphasizing mutual reinforcement and complementarities. LG design matters greatly in terms of results. Not all local governments are created equal, and often this is by design. The decentralization structure of the local government, particularly its administrative, fiscal, and political dimensions, determine its scope for action; level of responsiveness, participation, and democratic decision making; and capacity to deliver services and meet citizen expectations. CDD operations can support some of these efforts--participatory planning and social accountability, for example--but other areas will require macrolevel engagement with respect to the overall design of the 5. The term local government (LG) refers only to an elected local government, statutorily or constitutionally defined and corresponding to formal levels of subnational administration; it does not refer to deconcentrated units of the central or provincial government. 6. See Gillespie 2003; Binswanger and Aiyar 2003; and Binswanger and Tuu-Van 2005. 7. This refers to local offices of central or provincial governments. 1 decentralization structure. The resultant reforms can in turn support and deepen linkages and further the CDD agenda. To evaluate the best options for LG-CBO partnerships and to be strategic requires integrated CDD and decentralization analysis. Such analysis should cover (a) the types, roles, rights, and capacities of local institutions; and (b) the intergovernmental decentralization framework, the quality of decentralization already instituted, and the roles, powers, and capacities of the local governments. To achieve such analytical synergy, two important sets of perspectives and skills must converge: (i) a CDD practitioner who understands the local institutions, civil society organizations, and collective action, and (ii) a decentralization or local government practitioner who understands the intergovernmental fiscal framework; the local government structures, capacities, and incentives; and the ways in which the LGs link to the wider public sector. Lacking either of these perspectives, strategic decision making on the scope for meaningful local partnerships is likely to be biased in ways that ultimately miss important opportunities and constraints. Such errors of omission would undermine the likelihood that the design choices will maximize the opportunity space for scaling up, deepening, and enhancing the sustainability of CDD.. The State of the Art About one-third of CDD operations involve participatory local governments in some way.8 In addition, about 60 percent of CDD operations work to address policy and institutional reforms oriented toward CDD--including decentralization.9 Five broad tendencies can be seen in how LGs are involved in CDD operations: · Local government may be involved in decision making but not in direct management of resources; communities and project management units continue to manage most of the resources. · In a somewhat ceremonial role, the LG acts as a clearinghouse to vet community proposals, endorsing some before sending them to higher levels of the formal project management structure for approval. · Local governments sometimes cofinance subprojects, mainly by matching community and project funds but also, at times, by contributing all or part of the community contribution on behalf of the community. · CDD operations may attempt to introduce LGs to participatory planning (and budgeting) methodologies directly or through the inclusion of some elected local representatives on participatory local multistakeholder committees. · Changes in enabling environments and policy and institutional reforms are sometimes tackled through dialogue about the demonstration effects of the CDD operations' good practices. Often, however, dialogue does not extend to structural issues, and investments may be lacking in the types of technical assistance (TA) standard for policy changes promoted under adjustment and sector operations. Local government support projects place relatively little emphasis on community-driven processes and LG-CBO partnerships. Often, however, they will place some emphasis on formal mechanisms for bottom- up accountability, such as public meetings and consultations or the extension of freedom of information acts (FOIA) to LGs.10 In practice, these measures have often proven to be inadequate to engage citizens, especially where the LG has little flexibility or autonomy within the overall decentralization framework. 8. World Bank CDD portfolio monitoring database. 9. Data on percentage of total lending was not available, however. 10. A notable exception is the increasing focus on participatory municipal planning for slum upgrading in the urban context. 2 Dialogue between CDD and decentralization practitioners and social accountability specialists has begun in earnest, and a number of joint training activities have been held that have helped integrate CDD approaches with decentralization and social accountability. Getting Past the Hesitation Some CDD practitioners remain unsure about whether and how best to integrate local governments into CDD. The local government is often viewed as a black box: nontransparent and unaccountable. The common view among CDD practitioners is that "local governments are often susceptible to elite capture, with the result that public decision-making reflects disproportionate influence by well-off and well- connected groups. ... Standard mechanisms for accountability, such as elections, audits, or performance benchmarking, often do not work well in environments where information is scarce, open elections are unfamiliar or rare, and there is lack of clarity about expected performance standards."11 Where elite capture, opaque decision making, citizen marginalization, and inadequate performance benchmarks and monitoring arise in local governments, they tend to be crude reflections of broader societal patterns permeating all levels of society, from the communities to the central government. Given the strategic importance of LGs to CDD and service delivery, these lacunae (where they exist) should not be allowed to become the basis for bypassing local governments. Reluctance retards progress on effective interface. Skepticism about LGs, often corroborated by partial attempts to link communities to local governments, militates against exploiting real opportunities that may exist to strengthen the citizen-LG interface. CDD practitioners often use skepticism about LG functioning as a justification to perpetuate pure direct support to communities through parallel structures. Yet several actions can set local governments on an appropriate path and create an appropriate platform for CDD and broader civic engagement interventions, including: · Strengthening electoral systems, especially local electoral systems, and ensuring regular elections; · Improving access to information through formal and informal mechanisms; · Clarifying functional assignments of local governments; · Matching financing to functions and ensuring adequate local fiscal discretion, combined with fiduciary oversight and good practices; and · Developing performance standards and outcomes with adequate monitoring systems. Tackling such issues, however, requires wide-ranging policy dialogue to strengthen the decentralization framework. A CDD operation is not an appropriate instrument with which to address such mostly structural policy issues. Rather, CDD operations need to coordinate with macro decentralization policy reform processes in order to strengthen the structural incentives affecting local government performance and accountability. While seeking to partner CBOs with LGs, it is important to understand the fundamentals of a supportive decentralization framework and how to sequence possible decentralization reforms with CDD preparation or redesign to build citizen interface and strengthen social accountability to local governance. 11. Wong and Guggenheim 2004. 3 II. OBJECTIVES OF THE REPORT Introduce a framework for exploring LG-CBO partnerships in different country contexts. The main objective of this report is to provide client government policymakers and program managers, CDD and social accountability practitioners, and country teams with a framework for analyzing institutional choices and establishing entry points for CDD designs applicable to different decentralized and quasi- decentralized contexts. The framework will enhance understanding of local government structures and incentives and help identify the key local government reforms needed to strengthen coproduction partnerships and social accountability to community groups. The report suggests ways to adequately diagnose the actual scope for LG-CBO partnerships and accountability relationships and draws lessons from four countries Zambia, Tanzania, Nicaragua, and the Philippines chosen because of their ongoing efforts to strengthen such partnerships. Fundamental findings stress the benefit of the following steps: · Build on decentralization reforms; · Create additional space for partnerships within existing intergovernmental arrangements and local government capacities; and · Exploit the existing opportunity space to integrate LG-CBO partnerships into program design. Strengthen the diagnostics by focusing on LG-CBO interface options. This ESW report does not tell a client or country team whether to integrate local governments into CDD; rather, where the decision to do so has been made, it provides a framework for more closely analyzing potential entry points. The approach used suggests the need, when framing initial discussions, to combine from the outset technical expertise on both decentralization and strengthening local governments. Focus on one important pillar of the local development interface. LG-CBO partnerships are a subset of the local interactions that must be strengthened, sustained, or expanded to facilitate holistic local development. Other factors include links to the private sector, NGOs, indigenous organizations, and deconcentrated units of higher level governments (for example, sector ministries). The decision to focus on this particular subset is strategic, helping to keep the scope manageable and also complementing ongoing work supporting local development frameworks for local governance, empowerment, and social accountability.12 Finally, the cases presented in the appendixes (and partially summarized in the text) illustrate the approach. Two of the four country cases will also be subjected to detailed analysis in the context of pilot activities planned for FY2006­07 (the Philippines and Zambia). This further analysis will be done in close collaboration with the clients concerned and will offer a good opportunity to combine and reconcile decentralization and community perspectives in support of CDD. 12. Work on local governance is also being undertaken within the context of the empowerment agenda (PRMPR), as well as under social accountability and participation (SDV) programs. Such initiatives are also linked to country programs and client support. 4 III. METHODOLOGY To assess the enabling environment and opportunities for, and constraints to deepening LG-CBO partnerships in CDD, a three-phase approach has been adopted. Phase I: A preliminary indicator-based framework was developed to characterize the enabling environment for LGs and CBOs with respect to coproduction and accountability-based partnerships, including the opportunity space within which a CDD operation might work to develop such partnerships.13 Phase II: The framework was applied to cases in four countries, each chosen because it had at least one CDD operation attempting to foster LG-CBO partnerships. The framework was used (i) to characterize broadly the countries' decentralization frameworks and to try to assess where each country would likely fall in terms of opportunity space for local partnerships; and ,(ii) to predict the types of accountability and coproduction relationships that would be possible (a priori). The study then reviewed the existing LG-CBO partnerships in CDD operations in these four countries to map the nature and structure of the partnerships to different levels of opportunity afforded by the enabling environments. Phase III: The experience provided by the four cases was analyzed and lessons were drawn about the usefulness of the framework, identifying critical steps to be taken to fill knowledge gaps, and devising an operational and research agenda. 13. The preliminary framework, described briefly here, is more fully detailed in the Framework Appendix. 5 IV. SOME KEY CONCEPTS This section establishes working definitions for some of the main concepts employed in this paper and used to guide the analysis. The key terms include the following: · Decentralization o political decentralization o administrative decentralization o fiscal decentralization · Local government · Community-based organization · Coproduction · Accountability · Opportunity space Decentralization Decentralization14 can be defined as the process by which authority and responsibility for some substantial government functions are transferred from the central government to intermediate and local governments and often as well to communities and the private sector. This study examines decentralization alogn three dimensions: political, administrative, and fiscal. Political decentralization means the transfer of policy and legislative powers from central governments to autonomous, lower-level assemblies and local councils that have been democratically elected by their constituencies. Administrative decentralization involves placing substantial planning and implementation responsibilities in the hands of locally situated civil servants who fall under the jurisdiction of elected local governments. Fiscal decentralization entails according substantial authority for revenue and expenditure to intermediate and local governments. Other terms relevant to the decentralization context include devolution and self-government. Devolution encompasses the three dimensions of decentralization, adding to the defining process the concept of empowerment. Where functions, responsibilities, and means are devolved, local governments are empowered to make administrative, political, and fiscal decisions with respect to their assigned functions. Self-government refers to the condition pertaining where a level of government has dominion over substantial, clearly defined functions and can pass and enact laws with regard to these functions within its area of jurisdiction ­ including state, district, village, and so on. Before self-government can be said to exist, unambiguous political, fiscal, and administrative devolution of assigned subjects must have taken place. Local Government As used here, the term local government refers to an elected subnational government below the central or intermediate level that has been enshrined in the constitution or established by high-order statutory law to govern and serve a specific territorial jurisdiction within specified parameters. In practice, a local government can be an incorporated or unincorporated jurisdiction below the state or provincial level of government, including cities, municipalities, towns, townships, boroughs, districts, special purpose districts, authorities, counties, or similar local government entities. 14. The definitions in this section are based on work by Keith McLean; see McLean 1998. 6 Generally, staff of all levels of government operate within these local units, and, in less decentralized contexts, most local units are populated with civil servants and other staff of the central (or provincial) state appointed to implement and govern programs and services on behalf of the state. The definition of local government precludes such staff and related institutions, which often coexist with elected local within the same space. Partnerships referred to in this study focus on those between citizens and their elected representatives and may include private-sector partnerships. Community-Based Organization Throughout history, communities have organized themselves to address collective and individual needs. CBOs are normally membership organizations made up of a group of individuals in a self-defined community who have joined together to further common interests. Members are often people living near one another in a given urban neighborhood or rural village, but they may also be people united by a common interest rather than geographic community. The common interest might relate to the production, consumption, and use of pooled resources or the delivery of services; examples include women's groups, credit circles, youth clubs, farmers' cooperatives and associations, irrigation associations, forest and watershed management groups, artisan groups, fishery associations, and parent associations. CBOs can be stand-alone groups, or they can be linked to federations of groups at the regional, national, or international level. CBOs can be informal or formal. Informal organizations, such as women's and men's clubs and neighborhood groups, pursue joint interests and often appear more accessible to the poor than do formal organizations, which have legal status, formally stated rights and responsibilities, and a legally binding governance structure for recruiting members, selecting leaders, and conducting affairs. In theory, this notion of community could be disaggregated into citizens, service users, and CBOs, where the CBO provides the interface between citizens or service users and a project, program, or government entity or a private or public service provider. In practice, in most sectors in which CDD operates (including micro credit),15 collective action is used as the organizing principle, with the CBO structure serving as the locus of citizen or user representation and participation. The extent to which CBOs are internally democratic, participatory, and accountable plays an important role in determining transparency, deepening overall local governance, reducing elite capture, and enhancing allocative efficiency throughout the range of related local services. 15. This involves the use of group credit and joint liability contracts. 7 Box 1: Community-Based Organizations How do CBOs, NGOs, and elected local governments differ? This report considers CBOs to be membership organizations aimed at furthering the interests of its own members and NGOs, or nongovernmental organizations, to have broader scopes of activities such that they might assist CBOs and pursue commitments not directly benefitting NGO members. CBOs differ from elected local governments in that they are voluntary and choose their own objectives. In contrast, local governments are mandated to be responsible for revenue collection and the delivery of a variety of infrastructure and services. CBOs may interact closely with local government, with other levels of government such as local representatives of central ministries, with the private sector, and with NGOs. How can CBOs be made pro-poor? CBOs do not always represent the interests of poor people. To ensure that CDD has an impact on poverty reduction, CBOs must include poor people as members and represent their needs and interests. That does not mean that CBO membership should always be limited to poor men and women, but it does mean that the functioning and leadership of the CBO should clearly represent the interests of poor people along with those of the less poor. On what basis should the decision be made whether to work with new or existing CBOs? It is frequently advisable to work through existing organizations. At times, however, no good match can be found between the project and an existing organization; for example, a local organization may have very limited membership, rendering it unsuitable for association with projects requiring the involvement of several villages or of an entirely different group of people. Also, where communities have highly inequitable social organizations, creation of new groups may be called for to achieve program objectives or to promote the participation of disadvantaged people. Both new and special-purpose organizations are more effective when they build on the community's positive organizational traditions. The Moldova Social Investment Fund provides as example of this, with traditional decision-making mechanisms being used to establish community priorities; the Zambia Social Recovery Project, too, used project committees concerned with school investments by drawing on the country's strong tradition of parent-teacher associations. Source: Dongier et al., 2001. Coproduction Coproduction refers to the production of goods and services with the assistance of individuals not formally part of the producing organization. Production of goods traditionally produced by the state, such as health, education, and infrastructure, can potentially be facilitated by inputs from the beneficiaries or end users.16 The collective effort involved in coproduction develops new production functions, alters the availability of information and the nature of incentives, and creates social capital among citizens. In CDD common patterns of coproduction are built around service delivery. Service users or intended beneficiaries (target groups) contribute inputs--time and effort, money, or materials--as direct contributions to the production of collective or public goods and services. These create partnerships between service beneficiaries and service providers in which each agrees to a set of outputs and contributes to its production. Although income from local taxes would not normally be considered coproduction, fees for service or cost recovery arrangements and levels jointly defined by CBOs, LGs, and service providers in the context of designing long-term operation and maintenance (O&M) plans and affecting the technical choices for development interventions could be considered as part of coproduction. Coproduction in CDD emphasizes collective action. Public sector or private service providers normally partner with organized communities rather than with individuals or households. Community coproduction arrangements usually rely on CBOs to engage effectively on behalf of communities. These CBOs must therefore be enabled by adequate legal status, internal governance arrangements and resources, and administrative and technical capacity. Such capacities must often be built through 16. Ostrom 1996. 8 facilitated community support in the context of project preparation or execution. Examples of CDD coproduction arrangements include community built and maintained schools staffed by public sector teachers; publicly supported infrastructure improvement projects relying on community contributions of cash, materials, or labor; urban sanitation systems linking municipal solid waste removal with community-based waste collection; and rural forest management arrangements linking government-issued licenses with community monitoring of resource use. In more decentralized contexts, coproduction relationships are set up between LGs and community groups and seem to involve communities through collective action or as contractors for public-sector functions devolved to local governments.17 See Table 1. Table 1:. Coproduction between LGs and CBOs with Respect to Devolved Functions Stretching Role Deepening Role Nature of Tasks Performing a service for a fee Organizing citizens and mobilizing resources, often on a voluntary basis. Length of Usually short term Usually long term Involvement Critical CBO Technical Capacity for mobilization Capacity Type of CBO Well qualified technically, even if Preferably long standing; technical strengths are relatively new secondary Task Assignment Contractual; considerably well Flexible and based on mutuality and consultation defined in advance The pattern outlined in Table 1, derived from an analysis of partnership experiences in eleven cases,18 shows that, even in developed countries with substantial decentralization, applying the subsidiarity principle to local development may still yield LG-CBO partnerships, depending on the local context, culture, and history. Accountability Generally, accountability refers to the proactive process whereby public officials inform their constituencies of their plans of action, justifying their behavior and results and receiving sanction accordingly.19 As shown in the WDR 2004, "Making Services Work for the Poor," accountability relationships can be extended to the private service providers. The WDR 2004 defines four broad sets of accountability relationships and roles for service delivery: · citizens or clients who exercise voice over politicians; · policymakers who enter into compacts with organizational providers; · organizations who manage frontline providers; and · clients who exercise power over frontline providers. 17. Krishna 2004. 18. The cases, under World Bank auspices, were Northeast Brazil Rural Poverty Alleviation and Rural Poverty Reduction Projects (NRDP); Ghana Community Water Supply and Sanitation; Nicaragua Rural Municipalities Program (PROTIERRA); Zambia Social Investment Fund (ZAMSIF); and Karalla Rural Water Supply and Sanitation. Other cases were Wellington New Zealand: Partnership between City Council, Its Contractors and Community Organizations; Chicago Alternative Policing Strategy; Porto Alegre and the Participatory Budget; Denmark: Local Governments and Community-Based Organizations; Norway: Models of Local Government and Community Involvement; and West Virginia Water Supply. 19. Ackerman 2005. 9 In CDD, CBOs leadership usually intermediates between the citizen or client and the frontline service provider(s). This study emphasizes accountability relationships in which the frontline provision is (or by subsidiarity should be) local and production is jointly shared between CBOs, local governments, and, to some extent, local deconcentrated units of higher level governments. Accountability relationships are examined in the context of coproduction relationships, to assess the extent to which citizens are able to hold service providers to account for the quality of services (or lack thereof), hold local governments and local bureaucrats accountable for the types of services provided, and influence the local priorities subsequently articulated in the planning and execution of annual budgets. Collective action is a main source of client power. The emphasis on CBOs in the CDD approach reflects an understanding of a deep constraints on individual voices flagged by WDR 2004: "Even when there is an opportunity to redress complaints, monitoring and follow-through are public goods--the benefits accrue to the entire group while the costs are borne by a few. This is true for communities as well as individuals, but groups of people generally find it easier to elicit support from members than from individuals going it alone. So client power expressed outside market transactions will almost always be expressed through collective action."20 Accountability should be multidirectional. This study also acknowledges that, over time, accountability deficits may emerge within CBOs as well; for example, community leaders may become complacent or corrupt and lose sight of their primary roles as agents and intermediaries of broader citizen preferences and priorities. In some cases, CBO executives become too close to contractors and other private providers or become dominated by extended families. Three-way accountability in the context of LG- CBO partnerships give citizens an avenue for redress of grievances, as they are able to use the local government as a referee or accountability check on CBO executives and to catalyze elections of new CBO leaders as needed. Thus, strengthening the accountability interface between these two actors can ensure proper checks and balances over time; CBO leaders monitor LG actions, decisions, and performance, and vice versa. Social accountability strengthens the demand side of governance. Social accountability mechanisms refer to a broad range of actions (beyond voting) that citizens, communities, and civil society organizations can use to hold government officials and bureaucrats accountable. These include citizen participation in public policy making, participatory budgeting, public expenditure tracking, citizen monitoring of public service delivery, citizen advisory boards, and lobbying and advocacy campaigns. Opportunity Space Opportunity space refers to the maximum achievable outcome given initial conditions. It refers to the range of possibilities offered by the enabling environment, absent any efforts to alter fundamental social structures or relevant institutions. In the context of this study, opportunity space refers to the most extensive coproduction and accountability partnerships between LGs and CBOs feasible within the existing enabling environment, including the decentralization framework, the local government structure, and the enabling environment of the community-based organizations. The opportunity space can be dynamic, changing over time. Change in the opportunity space usually requires a set of structural alterations in the enabling environment's initial conditions. Possibilities include a radical new law (for example, Bolivia's Law on Popular Participation, outlined in Box 3) or a new fiscal devolution (for example, the decision of the government of Kerala to devolve 30 percent of its 20. World Bank 2003b, 71. 10 capital budget to LG21 or Nicaragua's Municipal Transfers Law, described in Box 6 in Appendix 3) capable of significantly widening the scope for local development partnerships and service delivery. Box 2: Bolivia's Law of Popular Participation The Law of Popular Participation (1994) aimed to decentralize political, fiscal, and administrative responsibilities and resources from the central to the municipal governments. This law introduced three innovations: (a) municipal governments' share of the national budget increased from 10 to 20 percent; (b) territorially structured community organizations became legally able and required to elaborate local development plans that became the ingredients of a municipal plan; and (c) members of the community organizations driving the oversight committees were given the power to monitor, audit, and veto municipal budgets. Source: Dongier et al. 2001. 21. The government of Kerala, India, was one of the few to take actions fully consistent with the spirit of the seventy-third and seventy-fourth Amendments to the Constitution of India (1993­94), which formally provided for three tiers of local self- government. 11 V. STRENGTHENING ACCOUNTABILITY AND COPRODUCTION RELATIONSHIPS IN DIFFERENT LOCAL GOVERNMENT CONTEXTS Coproduction arrangements nearly always include mutual accountability mechanisms. This is natural. It would be unreasonable to require communities and citizens to cofinance services over which they had no influence. In fact, the principle rationales for community contribution are ownership (reflecting genuine demand) and empowerment of CBOs to hold other stakeholders accountable. As a rule, CBOs should be involved in the planning and monitoring of any services to which they contribute. Without a fairly deep level of community participation, potential coproduction is normally reduced to beneficiary mobilization: communities contribute labor, materials, or money without any say regarding the uses for which these resources are employed. Local governments are public governance bodies and service providers with a defined territory, legally specified powers, resource bases, and functional responsibilities. Thus LGs are a special case of public organizations that can engage in the sorts of coproduction and accountability relations described above. Because of their relatively small scale, their physical and social proximity to their constituent communities, and their statutory or constitutional standing, LGs are particularly amenable to such local partnerships. In establishing coproduction with an LG, it is important to ensure its ownership and stake in CDD. CDD typically requires a contribution from even the poorest communities, usually upwards of 5 percent of subproject costs in cash or kind. CDD good practice usually requires that a community-owned operation and maintenance plan be put in place to ensure the sustainability of new physical assets and benefit streams. Increasingly, with more serious efforts to integrate local governments through partnerships, local governments are being asked to provide counterpart contributions as well. On the community side, such counterpart contributions have long been recognized both as empowering and as creating a sense of ownership. CBOs are much more likely to take participatory decision making seriously, to challenge providers and policymakers, and to hold providers and policymakers accountable for decisions where community contributions figure into the financing equation. It has also allowed Bank or donor resources to reach a greater number of potential beneficiaries, because 10 percent community contributions free up 10 percent of donor funds that can then be used to support other communities that would otherwise be left out. So counterpart contributions facilitate ownership, empowerment, and scale. Yet this principle is often overlooked when LG partnerships are forged, making it a significant missed opportunity. A local government that participates in a community prioritization process and endorses individual and collective plans from CBOs within its jurisdiction is not likely to attach much importance to the process unless it has a financial stake. As WDR 2004 points out, in order to be a stakeholder, one must have a clear stake. Treating counterpart contributions as fungible can blur lines of coproduction and accountability. In practice, many projects commingle counterpart contributions from various local stakeholders. In the Northeast Rural Development Project, counterpart contributions from the state and municipal governments tend to be treated flexibly, and this varies across states. In the Romania Rural Development Project, a 10 percent contribution is expected from the commune government and communities combined, of which at least 2.5 percent must be in cash. Such flexibility is often needed to ensure that local cofinancing is achieved, and, if managed well, it can help maintain critical accountability links. In contrast, if one entity such as the LG or any other coproducer finances the entire local contribution on behalf of all "coproducers," that entity is likely to try to assert itself over time, and the communities will likely permit this. Without a financial stake--however small--partners are likely to take the participatory processes less seriously and to be more relaxed about weakening accountability over time. 12 Ideally, coproduction should be an entry point for citizens' influence on the overall practices of local governments and other service delivery actors by providing a general opening up of the decision-making processes. Far too often, however, accountability remains restricted to the specific subset of activities being jointly produced, which is treated as distinct from the overall budget. 13 VI. THE ANALYTICAL FRAMEWORK An intuitive analytical framework has been designed for evaluating the institutional options and opportunities for linking CBOs and LGs. The framework follows three main stages: · Characterize the enabling environments for the LGs and CBOs using indicators in three dimensions: political, fiscal, and administrative. The functions and legal or regulatory framework are also reviewed. · Assess the overall opportunity space for LG-CBO partnerships, given their initial conditions along the enabling environment dimensions. · Map the diagnosed opportunity space to the set of corresponding coproduction and accountability relationships possible even without significant policy reforms affecting program preparation and review possible project, program, or country system interventions to estimate the extent to which they exploit or transcend available opportunities. Some of the main considerations in applying the analytical model are discussed below. Characterize the Enabling Environment for LGs and CBOs Are local governments and CBOs enabled or constrained? In a given country context, this assessment reviews the enabling environment for CBOs and LGs across each of the three dimensions--political, fiscal or financial, and administrative capacity--reviews the legal and regulatory context, and attempts to provide an overall picture of initial conditions. Table 2 below provides a summary of the structure of the enabling environment characterization, showing the assessment made in terms of both the de jure (by law) and de facto (in reality) situation along the three dimensions. Step 1 in the framework appendix, provides a more detailed look at the main indicators used to characterize each dimension. Table 2: Characterizing the Enabling Environment LG Environment CBO Environment de jure and de facto de jure and de facto LEGAL FRAMEWORK: FUNCTIONAL AND REGULATORY CONTEXT (INDICATORS) Political Dimension Political indicators Political indicators Fiscal/Financial Dimension Fiscal indicators Fiscal indicators Administrative/ Capacity Administrative factors Administrative factors Dimension By comparing the environment for CBOs and LGs along each dimension, one can characterize the extent to which a given country's context conditions are relatively more or less favorable for CBOs or for LGs to demonstrate initiative and proactively enter into local partnerships. One might find, for example, that the local governments in a given country are functionally and administratively enabled by a public administration law or policy but that their political and fiscal base is extremely weak due to the absence of fiscal or political devolution. Simultaneously, the local environment in the country might also have community organizations that are politically very engaged, even considered participatory, yet are financially very dependent on NGOs and project financing. By 14 looking jointly at these factors conditioning LG and CBO activity, one can assess the extent to which they are likely to be capable of entering into effective collaborative relationships and determine the types of interfaces a CDD approach could offer or introduce to strengthen the roles of LGs and CBOs in local service delivery. Identify the Opportunity Space What is the initial scope for coproduction and accountability relationships? Assessment of the enabling environment for LGs and CBOs characterizes the opportunity space. In other words, by looking at the degree to which the institutional environment is enabling or constraining along the three dimensions, it is possible to identify likely mutual strengths and weaknesses and the comparative advantages of LGs and CBOs in a specific context. On this basis, strategies for developing effective coproduction and accountability relations between LGs and CBOs--strategies that take into account the implications of the institutional environment enabling or constraining these relationships--can be identified. A schematic approach to characterizing the opportunity space for LG-CBO partnerships is summarized in Table 3.22 Table 3: Opportunity Space (by dimension) A B C D LGs and CBOs LGs Constrained, CBOs LGs Enabled, LGs Enabled, Constrained Enabled CBOs Constrained CBOs Enabled Legal, LGs and CBOs have LGs with few significant LGs have significant, LGs have Functional, few significant responsibilities; CBOs well-defined significant, well- Regulatory functions and permitted can implement functions; CBOs able defined functions; Context domains in many sectors to operate in few CBOs able to act domains in many sectors LGs lack popular LGs credible and LGs and CBOs Political LGs and CBOs lack legitimacy and credibility; legitimate; CBOs lack credible, Dimension popular legitimacy and credibility CBOs representative, popular legitimacy legitimate, credible, and accountable and credibility independent LGs have few and tightly LGs well funded and LGs well funded LGs and CBOs have constrained resources; fiscally autonomous and fiscally Fiscal few and tightly CBOs well funded, with for local services; autonomous; Dimension constrained discretion to deploy funds CBOs financially CBOs financially resources according to local constrained well funded priorities LGs lack staff (or LGs adequately LGs adequately skilled staff) and LGs lack staff (or skilled skilled and staffed; skilled and staffed; have weak staff) and have weak CBOs have little CBOs have skilled Administrative organizational and organizational and implementation and staffs and / Capacity implementation implementation capacity; collective action experience with Dimension capacity; CBOs CBOs have skilled staff experience collective action have little and experience with implementation collective action experience l 22. A more complete version of this information is available in the Framework Appendix. 15 The ease and range of linkage options varies with the opportunity space. The most favorable institutional environment for developing LG-CBO partnerships is represented in the table by column D; both LGs and CBOs are enabled along each of the three dimensions of the institutional environment, and a project planner would have maximum flexibility to link these two sets of local institutions according to the nature of the highest priority local development issues. The CDD planner would be unencumbered by restrictions in terms of LG functional, fiscal, political, and administrative autonomy, as the local government would in all likelihood have a high degree of capacity and discretion. In addition, with active CBOs and a civil society with a track record of activism and results, this enabled-enabled scenario would be optimal, allowing the program designer to focus coproduction relationships according to the subsidiarity principle and the deepening versus stretching rationale. Client contexts commonly exhibit constraints for CBOs and LGs along one or more dimensions, and display a mix of features, as shown in columns B and C of Table 3. In such situations, it is advisable to forge partnerships that benefit from the comparative advantages and mitigate any relative weakness. In a context where local governments are adequately funded but lack accountability or a history of participation, for example, a useful approach could be to engage CBOs as key institutions for representation and accountability and to link them to the LG budget process to maximize the allocative efficiency of LG expenditures.23 Several client countries display disabling environments unfavorable both to CBOs and LGs, as profiled in category A of the table. Such contexts are characterized by high degrees of centralization and weak civil society. Such challenging contexts represent real obstacles to effective LG-CBO partnerships of any sort. Yet a strategic choice can still be made. Generally, CDD practitioners opt for programs of direct support to communities and the capacity support usually places emphasis on social mobilization, building social capital, and CBO formation. In the same context, a local government or decentralization planner would opt for decentralization reform with local government strengthening, technical capacity building in administration, resource mobilization, budgeting, and so on, possibly even including performance-based management. The planner would generally pay little attention to the community interface. Proponents of each perspective would therefore take the view that the issues associated with strengthening the other side of the accountability and governance equation are too complex and intractable. Yet, to a neutral observer, it may not be obvious why a local development strategy could not be built at the onset that tries to strengthen both communities and local governments to play complementary and mutually reinforcing roles.24 The opportunity space can be dynamic, expanding over time through policy reforms and pilots that tackle critical constraints. Development programs can then invest their efforts, through policy analysis and dialogue or through piloting and demonstration, in promoting a more conducive enabling environment and thereby expanding the boundaries of the opportunity space along specific dimensions judged to represent critical constraints. The enabling environment diagnostic therefore identifies a current equilibrium that can be changed over time to be more conducive to local governance and local development. Both de facto and de jure constraints offer opportunities for forward movement. A de facto constraint implies that legal provisions and rules are in place but are not being implemented. In such a context, the project could exploit existing legislation and focus on implementation and compliance. Where the 23. Lack of a participatory culture is typically a feature of Eastern European countries, even where local governments are reasonably well articulated and possess good frameworks. This participatory link has been emphasized in the Romania Rural Development Project, with the aim of making rural infrastructure provision more community driven. 24. LICUS contexts present different considerations. It is difficult to strengthen local governance where the central government is weak or nonexistent. In some cases, however, local governance is the only entry point for peace-building and national reconstruction. 16 constraint is de jure, it means that practice has sidestepped law, and an argument could be made that the law should be aligned with practice. Evaluate the Scope for Accountability and Coproduction The enabling environment for CDD implies a set of possible coproduction and accountability arrangements. In assessing these, the relative strengths and weaknesses of CBOs and LGs along the three dimensions become important. Strongly democratic and participatory local governments operating in highly decentralized fiscal systems, for example, may be well positioned both to articulate and to respond to citizen demands. In the context of well institutionalized community empowerment schemes, CBOs may similarly possess both supply and demand side capacity. In many countries, however, the constraints are more pronounced, as are the potential advantages of LG-CBO partnerships. In countries where CBOs are relatively enabled along the three dimensions but LGs are neither demand responsive nor democratic and participatory institutions, CBOs may provide an accessible avenue for strengthening LG accountability. In these cases, the relatively enabled political legitimacy of CBOs may favor them as venues for demand articulation and accountability leading to strengthened responsiveness in LG planning, management, and service provision. Table 4: Accountability Relationships A B C D Constrained LGs, Constrained Constrained LGs, Enabled LGs, Enabled LGs, CBOs Enabled CBOs Constrained CBOs Enabled CBOs Both CBOs and LGs play a minor role in LGs can play a major Both CBOs and LGs Legal / LGs likely to focus service provision; role as service providers, can provide services Functional / accountability CBOs likely to focus while CBOs play a (individually or Regulatory upward, at best their advocacy on limited role; CBOs may coproduced); each Framework pressuring pressuring the focus their efforts on can provide a venue deconcentrated deconcentrated state pressuring LGs to for citizen influence state services rather than on LGs improve services over service providers CBOs and LGs CBOs legitimately Empowered and Both CBOs and LGs cannot legitimately represent citizen responsive LGs provide legitimately reflect represent citizen interests and priorities; a venue for aggregating citizen priorities; Political priorities and LGs less legitimate or citizen priorities; CBOs electoral and other Dimension interests vis-à-vis responsive to unlikely to be credible representative service providers, community advocacy channels for transmitting mechanisms for CBO resulting in limited citizen concerns and LG levels downward improve accountability responsiveness Both CBOs and LGs allocate or LGs allocate and Both CBOs and LGs LGs have few manage few resources manage significant can allocate and resources, at best and so are not likely to resources, a principle manage resources; able to advocate to face accountability venue for social participatory planning Fiscal state bodies pressure; CBOs can be accountability via and budgeting and Dimension regarding budget held accountable by participatory planning expenditure allocations and to citizens for resources and budgeting and monitoring may monitor state they manage expenditure monitoring; increase the expenditures at resource-poor CBOs responsiveness and local level probably marginal efficiency 17 Table 4: Accountability Relationships A B C D Constrained LGs, Constrained Constrained LGs, Enabled LGs, Enabled LGs, CBOs Enabled CBOs Constrained CBOs Enabled CBOs CBOs and LGs LGs have limited LGs can implement local Both CBOs and LGs have limited capacity to collect, decisions and provide capable of capacity to collect, analyze, or transmit detailed information to implementing local analyze, or information to citizens; citizens on resource use decisions and Administrative transmit CBOs may play a and services; likely to be providing information Dimension information; likely significant role in a greater focus for to citizens regarding limited in informing citizens and accountability than resource use and accountability for transmitting their views generally weak CBOs services, creating governance and to local state bodies channels of local service provision accountability In other countries, LGs may be constrained in their ability to respond to local demand for particular services due to administrative capacity constraints. Where the physical distance between LG offices and communities is relatively great and communication relatively difficult, for example, LGs may not effectively deliver services through administrative channels even when they possess adequate financial resources. This is a common situation in rural areas throughout the developing world: public agencies, including LGs, are often logistically unable to reach their clientele with the required frequency and intensity. In such cases, CBO coproduction arrangements can complement the capacities of LGs by providing them with community-level partnerships through which services can be delivered more effectively (thus stretching function).25 25. See Step 3 in the Framework Appendix for possible accountability and coproduction arrangements. 18 Table 5: Coproduction Relationships A B C D Constrained LGs, Constrained LGs, Enabled LGs, Enabled LGs, Enabled CBOs Enabled CBOs Constrained CBOs Constrained CBOs Legal / Limited opportunity for LGs only authorized or enabled in a LGs have significant Both CBOs and LGs have authorized Functional / coproduction of services by few sectors to effectively enter into responsibility but are unlikely to roles in service provision; defining Regulatory LGs or CBOs service delivery partnerships, even engage CBOs partnerships for complementary roles and appropriate Framework when CBOs take the initiative service provision linkages can produce effective partnerships Limited ability for CBOs and CBOs can represent citizen LGs are legitimately able to Both CBOs and LGs are able to Political LGs to legitimately influence interests and priorities but are likely aggregate citizen interests and legitimately represent popular interests; Dimension service mix and quality to focus their efforts on priorities, while CBOs are less mechanisms for coordination and partnerships with local state bodies representative and legitimate; negotiation of multiple CBO priorities at or NGOs that provide services; LGs service regime likely to be LG level may produce coproduction have few incentives to respond to dominated by LG plans, budgets, responsive to citizens citizen or CBO initiatives and management Limited opportunity for CBOs able to contribute to LGs have discretionary resources Both CBOs and LGs have discretionary Fiscal allocation of local resources achievement of their priorities, but for priority services, but CBOs are resources available for services; systems Dimension (LG or CBO) to finance LGs likely to be weak since they unable to contribute to services; that integrate and account for their services have few discretionary resources; thus LGs are likely to act as contributions can promote effective both may need to rely on local state suppliers, and CBOs at best may coproduction bodies to finance partnerships represent service consumers (not coproducers) Administra- Limited organizational basis CBOs can develop capacity to LGs may become capable of Both CBOs and LGs have capacity to tive and capacity for CBOs or LGs pursue their priorities, but LG entering into partnerships to contribute to production of services; Dimension to enter into partnerships implementation capacity is often deliver services, but CBOs are definition of roles and relationships can dependent on the central state; rarely capable of effectively be based on the comparative advantages often capacity enhancement is fulfilling their potential role in of each supply driven and not matched to service coproduction local needs 19 Electoral systems matter for citizen engagement and social accountability. The four main options for electoral systems are · First-past-the-post, or constituency (ward)­based elections; · Proportional representation, or party list, elections; · Mixed approach elections (combining elements of the proportional and direct approaches); and · Single nontransferable elections. In practice, local elections tend to be either by party list or ward based, even though general experience shows that first-past-the-post is the preferred method. The party list variants obscure accountability and encourage urban bias, as most of those elected tend to come from the urban core of the concerned area (including the main village).26 In more progressive localities, the winning party (or the local political culture) may decide ex-post to assign councilors to represent different sections of the locality, but often with no requirement that they come from or reside in that section of the locality. Thus, these representatives are responsible for but not necessarily accountable to the citizens. Where the party list electoral system is in place, it is important that the CDD or LG support intervention by introducing participatory planning and implementation methodologies that strengthen the accountability link between citizens and local governments with respect to resource allocation and local development decisions. Fiscal autonomy affects accountability and joint production possibilities. This is another relationship often overlooked or misunderstood in assessing the behavior and coproduction possibilities of local governments. The more fiscally constrained an LG, the less willing and able it will be to enter into coproduction relationships with citizen groups (or to be accountable downwards). The more dependent a local government is on transfers from the center, the more likely it will be to focus its accountability upwards as it tries to satisfy transfer criteria. This is especially the case where transfers are ad-hoc, rather than formula-based, with a mix of general and specific purposes, as would be recommended by specialists. Ad-hoc transfers are by nature opaque and often respond more to political considerations than to local need. While very few LGs finance all expenditure needs from their own revenues, the higher this percentage in terms of total local revenues the more willing the LG will be to enter into local compacts with citizens. This represents a virtuous circle in which, in exchange for specific commitments with respect to service levels, citizens become more willing to increase tax compliance, resulting in overall revenue increases. In some cases, citizens agree to a local tax increase tied to a specific service, such as rehabilitation of the local schools, or to specific fee-for-service regimes beyond general taxes. 26. This is confirmed by global experience. Specific examples include local governments in Nicaragua and Romania where the bias is observed, tied to the party-list system. 20 Box 4: Coproduction and Accountability in the Romania Rural Development Project The Romania Rural Development Project is a two-phase US$100 million APL currently implementing phase 1 ($40 million). The project aims to strengthen local governance and the capacity of commune governments and community groups to undertake socially, economically, and technically viable investment in roads and water supplies. It combines CDD and local government strengthening in its approach by increasing the interface between citizens and their lowest tier local government via a community driven and participatory approach to these two devolved but almost completely unfunded functions. Through a facilitated process, Community Investment Advisory Committees (CIACs) composed of community members and local councilors have been constituted in all one hundred pilot communes. A three-person multidisciplinary team provides CIAC members with a series of technical training and capacity building modules, ranging from community consultation to the financial and economic aspects of project preparation. The CIAC identifies subproject priorities and technical specifications, articulates the project proposal, and submits the application to the project management units. Pilot communes receive two project cycles totaling up to $300,000 of project funds, with a maximum of $200,000 for any one subproject. They also receive two cycles of training. The second cycle is supplementary and tailored to local conditions as revealed in CIAC feedback; commune level development planning for budgeting and local public finance was the most requested new module. Coproduction: A 10 percent community contribution towards capital cost is required, of which up to 3 percent can be kind. In practice, the local council has tended to provide some of the 7 percent cash contribution from its own resources and has mobilized the required community contribution. The CIAC is required to develop the O&M plan as evidence of sustainability, and together with the local administration it assigns or hires the appropriate expertise to maintain the infrastructure. The financial application and contract execution is done by the mayor or vice-mayor or someone designated on behalf of the council. Accountability: The CIAC supports the commune government in supervising subproject works. A three-person local monitoring unit (LMU), its members selected from among stakeholders, is responsible for monitoring project implementation and the work of contractors and is empowered to report any poor practices to the project management. (The LMU receives capacity building training with the CIAC.) In practice, mayors report that most citizens have been keeping an eye on implementation, as they feel a sense of ownership of the process. Source: World Bank, 2002b. 21 VII. SYNTHESIS OF CASE STUDY LESSONS This section relies primarily on document reviews, supplemented by interviews with project staff and other experts, to synthesize the basic lessons derived from the application of the LG-CBO analytical framework to contexts in Zambia, the Philippines, Tanzania, and Nicaragua. These cases show how the framework can be used in the context of country pilots and detailed field research to identify issues and questions requiring further in-depth exploration.27 These case studies were chosen because they were in countries with at least one CDD operation that was attempting to build LG-CBO partnerships. Each case review began with an examination of the country's formal de jure legal, functional, and regulatory context for LGs and CBOs, as well as of its real world (de facto) practices. The opportunity space for each country was then assessed along the three key dimensions discussed above ­ administrative, political, and fiscal ­ and the implications of the opportunity space for coproduction and accountability relationships were identified and evaluated. Actual CDD operations were chosen to illustrate the roles assigned to LGs and CBOs in project design and implementation. Comparison of the framework assessments with emerging realities highlighted any questions about the effectiveness of the utilization of the opportunity space and explored any divergence between the anticipated and the actual opportunity space revealed. This section provides a brief overview of the enabling environment and opportunity space in the four countries; the coproduction and accountability partnerships in the Philippines, Zambia, and Tanzania; and the dynamic interplay of these elements in Nicaragua. Variations in Opportunity Space Tanzania and Zambia were characterized by constrained LGs and largely constrained CBOs, while Nicaragua had constrained LGs but largely enabled CBOs. The Philippines, despite some problems in its LG framework, had a largely enabled LG-CBO context. In addition to these countrywide variations, projects unfolded differently within the countries' opportunity spaces. The consequent inconsistent results of CDD efforts yielded lessons for other contexts. · The Zambia Social Investment Fund (ZAMSIF) tried to exceed the limits of the opportunity space with its District Investment Fund (DIF). This component ran into substantial implementation difficulties, however, due to the country's lack of a sustained decentralization dialogue to support the decentralization reform agenda needed to strengthen the capacities of district governments. · Nicaragua succeeded by recognizing and addressing LG strengthening issues at the macro level and by piloting sequenced innovations within two types of projects--a social fund (the Fund for Emergency Social Investment, or FISE) and a demand-driven rural investment fund (the Rural Municipalities Project, or PROTIERRA). These two project types evolved and adapted over time in response to the changing enabling environment and opportunity space. · In the Philippines, with its good local government structure, the opportunity space was largely exploited by the Mindanao Rural Development Project, but with inadequate emphasis on the community-driven aspects. The second project reviewed, the Kalahi-CIDSS, largely underexploited the scope for coproduction synergies with local governments; it did, however, substantially strengthen citizen interface with LGs using social accountability mechanisms. · In Tanzania, a local government support project to strengthen district LGs was implemented in parallel with the second Tanzania Social Action Fund (TASAF II) in a complementary but somewhat incomplete manner. The country dialogue was underdeveloped with respect to rationalizing the structure of fiscal transfers from districts to villages. The need remained for 27. Detailed case write-ups appear in Case Appendixes 1 through 4. 22 important reforms that could substantially strengthen the scope for coproduction. (The two teams have held extensive discussions on harmonizing their projects and synchronizing their approaches.) Mixed Results Using CDD to Alter the Enabling Environment In Nicaragua, CDD operations have helped to improve the opportunity space over time. Four social funds (FISE phases I to IV) and two demand-driven rural investment funds (INIFOM I and II, also known as PROTIERRA) were involved in this gradual evolution. The first two FISE projects did not actively seek partnership with the municipalities. PROTIERRA I was the first of the CDD projects to directly engage with and support the municipalities. It also tried to broaden participation by working with and strengthening indigenous community consultative mechanisms, known as the InterCarmacas Assemblies. This created some competition among the approaches, and FISE III then started to work more with municipalities. All of this really took place within the existing intergovernmental framework, however. PROTIERRA II went a major step forward. In the context of project preparation, the team commissioned detailed assessments of the decentralization framework and the intergovernmental fiscal system. This led to a substantial push under for the government to expand the fiscal resource base of municipalities. The project piloted an increased fiscal transfer. This innovation created momentum that led to the enactment of a new Fiscal Transfer Law, which will increase transfers gradually from 1 percent to at least 10 percent by 2010. The new law had a major impact because it expanded Nicaragua's opportunity space for LG-CBO partnerships and made substantial resources available to underpin participatory local development. In Zambia, efforts to push the boundaries of the opportunity space were undermined by lack of a macro decentralization progress. While Zambia's LG environment was observed to be constrained across the board, its CBOs were assessed as being partly enabled (they remained fiscally constrained). Despite a highly constrained decentralization context and similarly constrained LGs, the ZAMSIF team took a bold step and calculated risk by attempting to integrate LGs into the project formally via a District Investment Fund (DIF). This was done using a special capacity building component through which LGs assumed increasing project responsibilities in five graduated phases. The major assumption underlying this plan ­ that the government would implement a decentralization policy ­ proved wrong, however, and since decentralization was not a high enough priority within the context of the Bank-Government macro dialogue, inadequate pressure was placed on the system to reform the decentralization framework. Ultimately, the DIF component suffered significantly from the lack of a supportive environment, and it was scaled back at midterm. Yet, it is not clear that the new decentralization policy would have sufficiently strengthened LGs to perform their envisioned roles. Coproduction and Accountability Relationships in Practice: Examples from the Philippines, Zambia, and Tanzania The Philippines The Philippines Opportunity Space Dimension LGs CBOs Political Partly enabled Enabled Functional Partly enabled Enabled Fiscal Partly enabled Partly enabled Administrative Partly enabled Partly enabled 23 The two projects examined in the Philippines ­ the Kalahi-CIDSS and the Mindanao Rural Development Project ­ engage similar sets of stakeholders, but do so differently. Assessment of the enabling environment suggests that Philippine LGs are entering into partnerships involving staff, money, or in-kind resources and technical support to communities. MRDP's midterm turnaround experience appears to indicate that, in the Philippines, municipal LGs already have the space to be reasonable development partners in all four dimensions. At village LG (barangay) level, Kalahi-CIDSS has given new life to legally mandated governance structures by making them responsible for subprojects. Support and guidance, of course, have played a critical role in these positive developments. Among the factors that appear most important when building and maintaining LG partnership potential are (i) providing LGs with real opportunities and incentives to perform (including access to funds); (ii) creating oversight mechanisms and checks to keep LGs on track; and (iii) establishing other forms of support, ranging from capacity building to guidance, from the sectoral body and other institutional partners. Zambia Zambia Opportunity Space Dimension LGs CBOs Political Constrained Partly enabled Functional Constrained Partly enabled Fiscal Constrained Constrained Administrative Constrained Partly enabled The Zambia Social Investment Fund (ZAMSIF) was conceived as a ten-year, adaptable program loan, with a first phase running from 2000 to the end 2005. It followed two Social Recover Projects (SRP I and II), designed in the late 1980s as the country was emerging from one-party rule, that worked directly with communities (and the deconcentrated technical staff of the central ministries) to deliver quick impacts and to mitigate the social dimensions of future macro adjustment reforms. Ten years later, it can be seen that one of the challenges of preparing ZAMSIF was the need to strengthen local governments in the continued absence of a robust decentralization framework; strong LGs would foster more sustainable impact and create a project exit strategy that would leave capable local councils in charge. Indeed, the design team hoped that the project would spark bottom-up demand for decentralization. To support local government performance, the project devised a process involving five graduation levels, progress through which would align a local government's demonstrated capabilities with its responsibilities in each of two ZAMSIF components: a community investment fund (CIF) financing subprojects in individual communities, and a direct investment fund (DIF) financing capacity building as well as funding district- level subprojects serving multiple communities under local government management. As expected, at project launch most of the local governments were in the lowest levels (1 and 2) and had little real power. 24 Tanzania Tanzania Opportunity Space Dimension LGs CBOs Political Partly enabled Partly enabled Functional Partly constrained Enabled Fiscal Constrained Constrained Administrative Constrained Partly enabled The Second Tanzania Social Action Fund (TASAF II), built on the experiences of the first TASAF, became effective in May 2005 and is expected to run until 2009. Since the project is in its very early days, this examination highlights LG-CBO roles and partnerships in the design of TASAF II as shaped by the lessons learned from the first TASAF. TASAF II focuses on enabling communities to "request, implement and monitor sub-projects that contribute to improved livelihoods linked to Millennium Development Goal (MDG) indicator targets in the Tanzania Poverty Reduction Strategy."2 The range of subprojects thus spans several sectors (among them health, education, infrastructure, and agriculture) and may include items such as construction of bridges and community roads, construction and equipping of schools, food production projects, vocational training for vulnerable populations, and improved maternity care. In general, TASAF II intends to provide a greater role for both local governments and CBOs than did the first TASAF. Under TASAF I, communities were generally able to identify, prioritize, and implement subprojects, provided that support and guidelines on access to funds were available. Without management above the community level, however, opportunities for successful mainstreaming of CDD approaches appeared limited. TASAF II thus transferred some project management responsibilities from the project management unit to the village and district local governments, the better to align with the existing decentralization framework and to give these bodies practical opportunities for developing better working relationships and for applying their skills. The specification of a clear role for village councils was also expected to alleviate some of the conflicts between subproject management committees and district LGs that surfaced under TASAF I. TASAF I, for example, had trouble with a separate public works program component for which district LGs handled both implementation and funds; as a result, public works were folded into the National Village Fund component, with implementation to be handled by the community management committee.3 Importantly, implementation of TASAF II coincides with a complementary project, the Local Government Support Project (LGSP), which supports the Local Government Reform Program by helping to "strengthen fiscal decentralization, improve accountability in the use of local government resources, and improve management of intergovernmental transfers."28 TASAF II thus focuses predominantly on the space between local governments and communities, while LGSP addresses relevant institutional and fiscal decentralization issues from the central government through to local government levels. This complementarity explains why TASAF II, while addressing local government performance and involvement in community development, emphasizes communities and especially that most "local" of local governments, the village council. Interplay of Shifting Opportunity Space and New Partnership Opportunities: Examples from Nicaragua 28. LGSP news release, 30 November 2004. 25 The development of LG-CBO partnerships in Nicaragua began slowly with a period (1990­1995) during which participants in FISE 1 and FISE 2, focusing on short-term gains, failed to seize the opportunity to build local partnerships offered by the country's decentralization framework. Two subsequent phases, detailed fully in Case Appendix 3, demonstrate how later projects tried to influence and adjust to new opportunities. The period between 1996 and 2000 was characterized by innovative local partnership programs. Constitutional reform and a new decentralization law improved the enabling environment somewhat, and INIFOM 1 introduced a number of pilots that further addressed the shortcomings of the enabling environment and promoted accountability and coproduction partnerships. FISE 3, confronted by the shortcomings of its previous strategy and stimulated by the example of INIFOM 1, began a set of pilots aimed in a similar direction. The period from 2001 to 2005 saw programs scaled up and policy reforms enacted. Operating in an environment similar to that of the previous period, the new operations, FISE 4 and INIFOM 2, moved to institutionalize the pilots and promote needed reforms. These efforts (examples include the Municipal Planning System and the Fiscal Transfers Law) were largely successful and led to a significantly improved enabling environment for local partnerships. INIFOM 1 and FISE 3 The World Bank supported both INIFOM 1 in 1996 and FISE 3 in 1999, both of which took advantage of the opportunity space and designed innovative pilots aimed at improving the policy environment and the practice of local partnerships. Both operations tried to improve the fiscal and capacity dimensions of the enabling environment while promoting local partnerships to deepen coproduction and accountability between LG and CBOs. INIFOM 1 fostered the Rural Municipalities Development Project, or PROTIERRA, which evolved significantly between its identification in mid-1994 and its approval in September 1996.29 Its initial rationale was to assist the Government of Nicaragua (GON) in implementing the actions recommended in the 1993 National Environmental Action Plan (completed with World Bank assistance) and the National Forestry Action Plan. Detailed analyses completed at the time suggested that sound natural resources management could help create employment and that natural resources management, through a demand- driven and participatory development approach, was the most sustainable means of doing so. The focus was later shifted to a broader conception of local development. To empower local governments to engage in partnerships with communities, INIFOM 1 addressed some of the weaknesses in the enabling environment, particularly the fiscal and the capacity dimensions. The thirty-two targeted municipal governments received a yearly allocation of program funds based on needs indicators, such as population and poverty, and were given wide freedom in deciding, through municipal technical units (MTUs), how to employ these funds. Projects submitted by community associations were appraised by MTU staff, with final approval coming from the municipal council. Subprojects (with the exception of capacity building subprojects) required matching funds of two types: (i) funds from communities (in kind or labor) and (ii) funds from municipal governments. Municipal councils were given the authority to accept or reject communities' subproject proposals, but they had to abide by both the local priorities set by their counterpart Intercommunity Assembly and also by the overall guidelines established for the country by INIFOM. This decision-making process balanced central, local, and community concerns. FISE 3 planners, seeing that INIFOM had moved forward in terms of local partnerships, and noting also the shortcomings (particularly in terms of sustainability) of its initial centralist strategy, designed it along similar lines: (i) providing essential small-scale infrastructure, mainly in education, health, and water supply and sanitation (in close coordination with the responsible agencies), (ii) focusing on the poorest communities, and (iii) "strengthening, on a pilot basis, municipal management which, in turn, should lead 29. World Bank 2002c. 26 to more sustainable subprojects at the local level."30 Later statements were more explicit that the credit would support FISE's "new focus on [...] municipal/ community involvement."31 FISE's program of strengthening municipal governments also involved components that compensated for LG fiscal and capacity weaknesses. FISE began a pilot with nine municipalities to which it delegated the management of investments up to $100,000. It also established MTUs responsible for carrying out the full subproject cycle, with the exception of the ex ante evaluation. FISE Managua still had to check whether sector approval had been obtained and approve the project before the municipality could contract it to a firm. In terms of financing, FISE's main difference from INIFOM was that resource transfers were tied to sectors considered priorities by the central government, thus complementing the INIFOM approach nicely. While INIFOM simulates an unconditional fiscal transfer, FISE 3 followed the principal-agent model in which local governments act as central government agents for the delivery of certain goods and services. To address the challenge of low sustainability of local investments (problems in FISE 1 and FISE 2), FISE 3 created the Preventive Maintenance Fund (FMP), basically a partnership between the central government, municipal governments, and community organizations to ensure investment sustainability. Through the FMP, the national government provides funding to municipalities for preventive maintenance of primary schools and health centers. This cofinancing is progressive: extremely poor municipalities receive a higher percentage than do others. Community maintenance committees supported by local government officials manage the work. These committees have a strong incentive to perform well since they can receive funds each year only if they complete the maintenance agreed on the year before. INIFOM 2 and FISE 4 INIFOM 2 and FISE 4 were highly coordinated operations, making it useful to discuss them together. In the fiscal dimension, INIFOM 2 tried to improve transfers by creating a new system, the Municipal Development Fund (FONDEM), the constitution for which emphasized the rationale of transfers as instruments of fiscal equalization (art. 177). Nicaragua's transfers were allocated mainly in proportion to population, a criterion that did not help close horizontal imbalances and thus did not advance the principle of fiscal equalization. Designed as a "general-purpose grant" mechanism, FONDEM aimed to complement existing municipal own-source revenue and various ongoing programs with negotiated, purpose-specific transfers. To this end, the project assisted INIFOM in revising the legal framework of local government finances; drafting appropriate policy, legal, and regulatory texts; and developing and extending to all municipalities participating in the project (i) improved systems and manuals of local fiscal administration, (ii) standards and procedures for financially sustainable delivery of municipal services, and (iii) the system for land and property registration, or Cadastre (Sistema de Cadastro, or SISCAT). The design and implementation of FONDEM strongly influenced the national debate on fiscal transfers and the development of a related legal framework, thus changing the enabling environment: in July 2003, the GON approved the Municipal Transfers Law, which substantially increases the funds available to municipalities from about 1 percent of tax revenues to at least 10 percent by 2010. 30. World Bank 1998, 1. 31. World Bank 1998, 2. 27 Both the World Bank and INIFOM have found it difficult to overcome the "project" logic of INIFOM 2, however, and the implementation of FONDEM clearly suffered from this. On the Bank side, the integrity of the FONDEM concept would have been better served through some form of "budget support" rather than through an "investment loan." INIFOM also substantially resisted loss of its role as "fund manager." Previously, INIFOM and FISE had developed separate systems for managing investment projects, adding to the systems required by other donor or government programs and causing confusion and waste in capacity-building efforts. To solve this problem, INIFOM 2 and FISE 4 divided the capacity-building component into different modules, and each agency developed, in consultation with municipalities, generic modules that would not be project specific but were applicable to all municipal clients. Based on comparative strengths, FISE and INIFOM developed modules for investment projects and financial management, respectively. The planning and accountability structures and processes set up by INIFOM 1 and FISE 3 were project specific, and a sustainability risk was associated with the end of the projects. One of the most significant impacts of the operations with INIFOM and FISE was that the pilots each undertook with regard to Participatory Municipal Planning became institutionalized as the Municipal Planning System (MPS) and were adopted as statutes for local planning rather than as project specific methodologies. (See the diagrams in the Case Study Appendix 3.) This MPS built upon the momentum and lessons provided by the previous pilots; the Territorial Committee, for instance, plays a role similar that of the Intercommunity Assembly. While FISE and INIFOM showed that local governments could successfully manage local funds and investments, one of the remaining challenges was to teach local governments how to empower community associations to implement their own subprojects and manage subproject funds. To address this, FISE began a pilot call "Proyectos Guiados por la Comunidad."32 FISE transfers resources to local governments, which in turn, for projects identified in the Local Municipal Plan and particularly suitable for community implementation, transfer them to CBOs. Local governments and community associations cofinance projects, and both are responsible for the projects' sustainability. 32. See Emergency Social Investment Fund Operational Manual: http://www.fise.gob.ni/pgc. 28 VIII. CONCLUSIONS The opportunity space provides a useful concept for exploring the potential scope for LG-CBO partnerships. Applying the framework and the concept of opportunity space is a useful way to assess the enabling environment for the kinds of LG-CBO partnerships with which this study is concerned. Application of the concept does, however, require flexibility, as countries don't always fit neatly into enabled or constrained categories but sometimes fall into the grey area between. The opportunity space is an important starting point that can be dynamic, either expanding with the right mix of macro and micro intervention, or becoming an increased constraint. The Nicaragua and Zambia cases present clear examples of each course: Nicaragua underwent dynamic expansion; Zambia experienced enduring constraint. In determining one or the other outcome, detailed analysis of the macro constraints for local governments, an active dialogue among key stakeholders, competition, and opportunities to tackle the structural constraints seem to matter most. A CDD project in isolation is unlikely to be able to strengthen the local government incentive structure. The opportunity space should be examined periodically to capture potential shifts and to identify new opportunities and constraints. This is very important given the possible shifts discussed above. Most CDD projects that foster partnerships also try to influence local governance more broadly. In order to understand whether these outcomes are being achieved, the opportunity space should be reviewed periodically, for example, as an input into the midterm review. Coproduction and accountability reinforce each other. In all the examples cited, both coproduction and accountability relationships seem to be well linked. In fact, it appears that the stronger the coproduction linkages, the more involved are the accountability relationships specified under the partnership arrangements. This is consistent with the framework's prediction that the stronger the collective stake (investment of resources and time), the greater the sense of empowerment and the greater the sense within the community of being entitled to results. The weaker the decentralization framework, especially fiscal decentralization, the weaker the partnership possibilities. This was clearly demonstrated across the four cases, and it was striking that in the Zambia and Tanzania cases, where the decentralization frameworks were weak, the coproduction and accountability relationships tended to be weaker than in the Philippines and Nicaragua. Embedding the partnership approach into a decentralization dialogue is critical. The four cases show that, whenever the decentralization framework is weak or nonexistent, attempts to partner LGs and CBOs in effective coproduction relationships beyond the natural (and limited) opportunity space can occur only if the enabling environment is opened up through an overall dialogue on decentralization. 29 Box 3:. Inclusive Decision Making to Improve Local Governance in Macedonia By promoting collaboration between municipalities and local communities in planning and management, the Macedonia Community Development Project (MCDP) has increased the transparency and dynamism of local development. Prior to MCDP's intervention, community development committees 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 linking municipalities 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--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 consist of community representatives, including both local notables and members of 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 of ideas aimed at mitigating 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 of 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 of 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, I am more relaxed performing my functions as mayor because the community understands the constraints and there is satisfaction of people in the 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 LG-community partnerships to support implementation of national decentralization policy. As a result, more inclusive local governance has become a key institutional foundation for more effective local development in Macedonia. Source: World Bank 2005. Our learning is not yet complete. CDD operations increasingly try to build strong relationships between communities and local governments. While we have a framework for understanding the main considerations and opportunities for linking, more lessons will evolve as a greater number of operations (CDD and LG support) attempt to explore and exploit interface opportunities. Additional research on how and when the two approaches converge within countries would also help to strengthen learning for local governance and local development. 30 IX. OPERATIONAL IMPLICATIONS Integration at the local level requires technical integration within client and donor teams. Whether the community level or the LG is a black box depends on the perspective of the practitioner. To most CDD practitioners, the LG is an unaccountable, elite-captured black box. To most public-sector management experts and decentralization practitioners, the community is a fuzzy, amorphous concept that doesn't really address the fundamentals of how governments and service delivery work. The reality is that to attain local governance and improve local services, partnerships are needed that coalesce these perspectives to improve mutual support and information exchange and that engage the private sector and other local actors. Country clients should coordinate donors to avoid duplication and contradictions. Lessons from contexts in which donor coordination was lacking suggest that donors often put in place procedures that complicate, contradict, and duplicate those of other donors. This can be very confusing and costly to local communities and LGs, whose capacities are already limited. Given the donors' weak track record on such coordination, countries may need to take the lead wherever they have a coherent vision of how best to organize local development partnerships. LG-CBO partnerships should not be designed without detailed analysis of the decentralization framework, including the intergovernmental fiscal system, and LG capacities and constraints. This requires that the marshalling of relevant decentralization expertise to support the preparation of CDD operations and to offer suggestion on scope for linkages. Local government support operations must strengthen social accountability relationships to help realize allocative efficiency. Harmonizing the community-LG interface in client countries' and donors' portfolios provides some scope for harmonizing participatory methodologies for accountability arrangements. Social accountability tools such as citizens' report cards and participatory planning and budgeting help to complement formal public consultations (for example, public meetings) and deepen the quality of local decision making. This helps to better match citizens' preferences with budget allocations. Social Accountability should be mainstreamed where local governments have discretion and adequate resources to be responsive to local communities. Stakeholder consultations and participatory planning methodologies should, in principle, be built into all local development operations and should also be consistent in type across a country portfolio of CDD operations. If local governments are legally and fiscally constrained, however, and therefore unable to respond to citizen demands, standardized social accountability is likely to induce more frustration than would accountability. Cost-sharing arrangements between various stakeholders should be consistent within a country. Again, except for pilots, it is not clear why, moving forward, communities and local governments should be subjected to an array of cofinancing arrangements depending on the project. If any variability exists, it should be on the basis of the financial capacity or poverty level of the community or local government, with poorer target groups facing a lower financing burden, based on subproject type or sector.33 Cofinancing projects with overlapping objectives, however, should not require significantly different community contribution percentages, without a clear logic. Careful and dispassionate monitoring and evaluation and impact assessments could help reconcile approaches. Project planners tend to make competing claims about their plans' overall impact and 33. Another possibility could be same percentage counterpart contribution at reduced subproject size. 31 effectiveness, both in terms of benefits accruing to communities and of integration with local governments. Where a portfolio contains multiple projects and harmonization is an issue, M&E could be treated as a public good and coordinated by the quality assurance team or the CMU. This would achieve the additional benefit of ensuring a set of common benchmarks against which learning and the effectiveness of different operational modalities could be evaluated. Task teams would help define the monitoring indicators, based on the stated development objectives, and would agree to these with the concerned monitoring teams.34 34. Since projects do have internal monitoring needs, perhaps the governments and the Bank could agree on a common core set of indicators, allowing additional indictors to be monitored according to the needs of the various stakeholders. 32 X. THE RESEARCH AGENDA This report has tried to explain the interrelatedness of the enabling environment for CBOs and LGs in the context of designing effective LG-CBO partnerships for local development. It has explored how initial conditions--local governance structures, the decentralization framework, and, to some extent, collective action--coalesce to sustain and deepen CDD. Yet several questions remain unanswered, some of which are detailed below as future research questions the resolution of which could help strengthen the overall agenda. Under what conditions can CDD operations catalyze fundamental changes in decentralization policy, significantly strengthening the enabling environment for local governments? The emerging lesson seems to be that it is difficult to expect a CDD operation to handle fundamental changes in decentralization policy. Yet, in the context of Nicaragua, some important changes were promoted that strengthened the overall framework. Efforts in Zambia were far less successful, while efforts in Tanzania have been mixed. In the Philippines one of the projects has tried to facilitate greater sector decentralization. More research is needed to analyze just how far a CDD operation can push structural changes and under what conditions good results are likely to be obtained. How do these changes affect the overall capacity of local governments to deliver on citizen expectations? Where the government remains highly centralized, can a series of do-no-harm incremental steps be incorporated into the CDD approach that would be consistent with strengthening local partnerships? The need exists for a more extensive review of approaches to local partnerships undertaken in the context of weak decentralization frameworks (for example, contexts affected by conflict). It is difficult to tell on the basis of a few cases whether any useful and possibly less ambitions steps can be taken to strengthen the overall scope for LG-CBO partnerships. An important related issue is whether, as an interim measure while awaiting the strengthening of the LGs, the interface between CBOs and deconcentrated units of the center-state should be strengthened, or whether this would in fact help to delay fundamental subnational democratization. In such contexts, would a clear demonstration of government commitment to decentralizing be important? And what would constitute adequate evidence of genuine commitment? Should governments and the Bank tackle structural and capacity flaws in advance of project preparation or as part of it, or should redress of these flaws proceed in parallel with a sequenced, complementary dialogue? Sequencing of reforms, policy, and project interventions is always a challenge, even within decentralization reform itself. How best to implement CDD based on partnerships and the sequencing of approaches across the range of contexts and types of opportunity spaces identified above is major challenge that warrants further research. 33 XI. NEXT STEPS: THE ROAD AHEAD Learn by doing through country pilots. Three country pilots are planned that will build on this ESW and support both clients and country teams and the further refinement of the overall analytical framework. The participating countries are Zambia, Angola, and the Philippines. The main objective of these pilots will be to build local governance platforms35 by strengthening the learning, capacity, and mechanisms needed to integrate the CDD approach with local governments (LG) and social accountability. Subsidiary objectives include the following: · Work with three country clients and Bank teams to develop a framework for making institutional choices that simultaneously support CDD and local governments and decentralization reforms (Zambia, Angola, and the Philippines).36 · Introduce a more effective methodology for analyzing institutional options, capacities, and macro and micro constraints within different decentralization contexts that affect partnerships between communities and local governments. · Adapt social accountability (SAc)37 tools to contexts in which LGs are central to local governance platforms; use SAc as the foundation for citizen monitoring of and feedback on project impacts and service quality and as a tool to help empower citizens to demand improved local services and transparent decision making Translate the framework into an assessment tool. An important follow-up activity is to translate the analytical framework into an assessment tool that can be used by clients as the basis for conducting the analysis of institutional options. This follow-up work is planned for FY06 and will overlap with the country pilots to build on the lessons from the detailed country level assessments. Conduct follow-up research. Additional analytical work is planned for FY06 aimed at addressing issues of convergence between CDD and LG support operations. This work will also support the process of portfolio harmonization, which has emerged as an active challenge for many clients (for more information, see Section X above). As the pilots, regional analytical work, and programmatic experiences evolve in client countries, additional important empirical and analytical questions are likely to surface. Maintain a multidisciplinary, cross-sectoral approach and dialogue. This report was motivated by the interests of country clients and staff working on social funds, CDD, and decentralization, social accountability, and empowerment. These concerns represent a range of different skills and perspectives, and it is important to maintain a dialogue among these varied specializations to sustain momentum and expand the possibilities for more learning and stronger integration of decentralization and community- driven development. Design teams from both donors and governments should therefore strive to reflect a multidisciplinary approach. 35. Local governance platforms integrate decentralization to elected local governments, with CDD approaches and social accountability tools to strengthen good governance at the local level and increase citizen voice, including that of the poor and vulnerable, with respect to local service delivery and decision-making transparency. 36. In addition to MRDP and Kalahi, the ARMM Social Fund will be included to harmonize approaches. 37. Social Accountability (SAc) is defined as an approach toward building accountability that relies on civic engagement, that is,, ordinary citizens and/or civil society organizations participate directly or indirectly in extracting accountability. In a public sector context, SAc refers to a broad range of actions and mechanisms that citizens, communities, independent media, and civil society organizations can use to hold public officials and public services accountable. 34 XII. FRAMEWORK APPENDIX 35 Step 1: Assessing the Enabling and Constraining Factors for Empowering CBOs and LGs Focus on relevance to coproduction and accountability of LG-CBO relationships.. LG Environment CBO Environment de jure de facto de jure de facto Legal / Functional / For what service delivery functions is the LG responsible? In what sectors and for what services do CBOs have a service delivery Regulatory role? Do laws specify how LG does this, i.e., direct provision, Framework contract-in, contract-out, community role? Does policy/practice For which sectors/services are they direct producers? Coproducers? define this? For which of these services are they linked to LGs? Is the LG's legal status defined? How and where? Can LGs Is the BO's legal status defined? How and where (e.g., under NGO law, hold money (e.g., in bank accounts) or enter into legal association law, company act, or other)? Can they hold money (e.g., in agreements? bank accounts) or enter into legal agreements? Does the LG have power to legislate (enact by-laws, etc.)? How or through what processes are CBOs created? What legislative powers do they have with regard to the assigned functions? How are CBO executives selected, and to whom are they accountable? In many cases, actual membership is not clearly defined (definitions Which LG officials are elected; i.e., is the mayor or other might be something like "all water users," for example); only executive Political executive elected directly or indirectly? and/or board identities are made clear. Factors What is the electoral system? How are officials elected: by Do CBOs have a formal role in local governance? CDD planners should party list, ward, first past the post, etc. usually expect the answer to this to be "no." What are the specific powers of the mayor and of the council? What additional mechanisms exist for LG accountability to citizens? Are they consultative, participatory, or other? Who participates? Do any special reservations ensure participation by minorities, the disadvantaged, women, etc.? Fiscal What is total LG expenditure as % of public expenditure? What is the CBO's resource base? Factors Of this, what % is from the LG's own source revenues and What are the sources of CBO funding? Assessment requires what % is transfers? How do funds flow? Do they come directly from state bodies or via LGs? qualitative overview of Of its own source revenues, what portion comes from taxes where resources What is the CBO's ability to mobilize resources (e.g., obligatory "taxes" 36 LG Environment CBO Environment de jure de facto de jure de facto come from and of vs. fees (cost recovery)? and fess vs. voluntary accumulations)? what discretion the LG has to allocate What is the nature of transfers (e.g., general purpose vs. What are the LG-CBO and CBO-sector resource flows? resources. specific purpose, formula vs. discretionary/ad hoc), and what is any assessment of LG performance is factored into the size What is the correspondence between the CBO resource base and its of the transfer? functional roles? Are transfers predictable? That is, is the size of the transfer What degree of clarity and openness exist in the CBO planning/budget reasonably stable from year to year? Are transfers made on process, including both formulation and execution? time (de facto) or stipulated to be made at reasonable point in FY (de jure)? Expenditure autonomy: Does the LG have flexibility to allocate/reallocate between recurrent and capital expenditures? What is the share of capital vs. recurrent expenditures in the overall budget? (If no discretion is allowed and recurrent expenditures consume all funds, the LG will have no scope for partnership.) How clear are the expenditure assignments and their correspondence with functional assignments? What degree of clarity and openness exists in the LG planning/budget process, in both formulation and execution? How are political executive vs. professional administrative What authority does the CBO have to manage funds and sign contracts? roles balanced in the LG? What degree of specialization/"professionalization" exists in CBO What is the LG's human resource base and system, including administrative functions? staffing structure, managerial autonomy, staff capacity, career mobility, staff development, etc.? What are the governance vs. executive roles pertaining to CBO Administrative/ structure? Capacity Does the LG have the flexibility for contract-in, contract-out, or joint contracts with other LGs or CBOs? Does the CBO have the flexibility for contract-in, contract-out, or joint Factors contracts w/ LGs or other CBOs? What flexibility exists to alter LG organization to meet local priorities? What is the nature, relevance, and timeliness of TA provided by the state and/or LGs to CBOs? What is the nature, relevance, and timeliness of TA provided by the state to LGs? Is the CBO service delivery system performance focused? Is the LG service delivery system performance focused? 37 Step 2: Mapping the Opportunity Space for LG-CBO Accountability and Coproduction Arrangements Characterizing the Opportunity Space (by dimension) A B C D LGs and CBOs LGs Constrained, CBOs LGs Enabled, LGs Enabled, Constrained Enabled CBOs Constrained CBOs Enabled Legal / LGs and CBOs have few LGs with few significant LGs have significant, well- Functional / significant functions and responsibilities; CBOs permitted LGs functions significant, well defined functions; CBOs can Regulatory domains and can implement in many defined; CBOs can operate in act in many sectors Context sectors few domains Political LGs and CBOs lack LGs lack popular legitimacy and LGs credible and legitimate; LGs and CBOs credible, Dimension popular legitimacy and credibility; CBOs representative, CBOs lack popular legitimacy legitimate, independent. credibility credible, and accountable and credibility LGs with few and tightly LG well resourced fiscally and LG well resourced and Fiscal LGs and CBOs with few constrained resources; CBOs autonomous for local services; fiscally autonomous; CBOs Dimension and tightly constrained well resourced and with CBOs financially constrained financially well resourced resources discretion to deploy resources according to local priorities LGs lack staff (or skilled LGs lack staff (or skilled staff) LG adequately skilled and LG adequately skilled and staff) and have weak staffed; CBOs have little staffed; CBOs skilled and Administrative organizational or and have weak organizational or implementation and collective experienced w/ collective Dimension implementation capacity; implementation capacity; CBOs skilled, experienced w/ collective action experience action CBOs implementation experience small action 38 Step 3: Identifying the Operational Implications of the Opportunity Space for Improving Local Accountability and Coproduction via CBOs and LGs Coproduction Relationships (by dimension) A B C D Constrained LGs, Constrained LGs, Enabled CBOs Enabled LGs, Constrained CBOs Enabled LGs, Enabled CBOs Constrained CBOs 1 Limited opportunity LGs only authorized/authorizable in a LGs have significant responsibility Both CBOs and LGs have for coproduction of few sectors to effectively enter into but are unlikely to engage CBOs authorized roles in service provision; Legal / Functional services by LGs service delivery partnerships, even partnerships for service provision defining complementary roles and / Regulatory and/or CBOs when CBOs take the initiative appropriate linkages can produce Framework effective partnerships 2 Limited ability for CBOs can represent citizen interests LGs able to legitimately aggregate Both CBOs and LGs able to CBOs and LGs to and priorities but are likely to focus their citizen interests and priorities; legitimately represent popular legitimately influence efforts on partnerships with local state CBOs are less representative and interests; mechanisms for Political service mix and bodies or NGOs that provide services; legitimate, with service regime coordination and negotiation of Dimension quality LGs have few incentives to respond to likely to be dominated by LG plans, multiple CBO priorities at LG level citizen/CBO initiatives budgets, and management may produce citizen responsive coproduction 3 Limited opportunity CBOs able to contribute to achievement LGs have discretionary resources Both CBOs and LGs have for allocation of local of their priorities, but LGs likely to be for priority services but CBOs are discretionary resources available for resources (LG or weak since they have few discretionary unable to contribute to services; services; systems that integrate and Fiscal Dimension CBO) to finance resources; both may need to rely on thus LGs likely to act as suppliers account for their contributions can services local state bodies to finance and CBOs at best may represent promote effective coproduction partnerships service consumers (not coproducers) 4 Limited organizational CBOs can develop capacity to pursue LGs may become capable of Both CBOs and LGs have capacity basis and capacity for their priorities, but LG implementation entering into partnerships to deliver to contribute to production of Administrative CBOs or LGs to enter capacity is often dependent on the services, but CBOs rarely capable services; definition of roles and Dimension into partnerships central state; often capacity of effectively fulfilling their potential relationships can be based on enhancement is supply driven and not role in service coproduction comparative advantages of each matched to local needs 39 Accountability Relationships (by Dimension) A B C D Constrained LGs, Constrained Constrained LGs, Enabled CBOs Enabled LGs, Constrained CBOs Enabled LGs, Enabled CBOs CBOs 1 Both CBOs and LGs likely to LGs play a minor role in service LGs can play a major role as Both CBOs and LGs can provide focus accountability upward, at provision; CBOs likely to focus service providers, while CBOs services (individually and jointly best pressuring deconcentrated their advocacy on pressuring play a limited role; thus CBOs via coproduction); thus each can Legal / state service providers deconcentrated state service may focus their efforts on provide a venue for citizen Functional / providers rather than LGs pressuring LGs to improve influence over service providers Regulatory services Framework 2 Limited ability for CBOs and LGs CBOs can legitimately represent Empowered and responsive LGs Both CBOs and LGs can to legitimately represent citizen citizen interests and priorities, can provide a venue for legitimately reflect citizen priorities and interests vis-à-vis while LGs are often less aggregating citizen priorities, but priorities; electoral and other Political service providers; likely to result legitimate and less responsive to CBOs are unlikely to provide a representative mechanisms at Dimension in limited downward community advocacy legitimate channel for both CBO and LG levels may accountability transmitting citizen concerns improve responsiveness 3 Both CBO and LG have few LGs allocate or manage few LGs allocate and manage Both CBOs and LGs can allocate resources; at best they may resources and so are not likely to significant resources, providing a and manage resources, so advocate to state bodies re be the focus of accountability; principle venue for social participatory planning and Fiscal budget allocations and monitor CBOs can be held accountable accountability via participatory budgeting and expenditure Dimension state expenditures at local level by citizens for resources they planning and budgeting and monitoring may increase the allocate or manage expenditure monitoring; responsiveness and efficiency of resource-poor CBOs probably resource use at both levels marginal 4 CBOs and LGs have limited LGs may have limited capacity to LGs may be capable of Both CBOs and LGs may be capacity to collect, analyze, or collect, analyze, or transmit implementing local decisions and capable of implementing local transmit information; result may information to citizens; CBOs providing information to citizens decisions and of providing Administrative be likely limits in accountability may play a significant role in regarding resource use and information to citizens regarding Dimension of governance and service informing citizens and services delivered; thus they are resource use and services provision transmitting their views to local likely to have a greater focus on delivered, thus creating potential state bodies accountability than the generally venues for accountability at both weak CBOs levels 40 XIII. CASE STUDY APPENDIXES APPENDIX 1. THE PHILIPPINES: THE CHALLENGES OF IMPLEMENTING DEVOLUTION38 The Philippines has put into place a laudable decentralization policy and accompanying legislation, built around devolution of responsibilities and finances, that are widely hailed as being at the forefront of this effort in the region and among developing countries generally. At the same time, more than a decade after passage of the landmark Local Government Code, implementation of decentralization policy remains a struggle, with the national government continuing to dominate or work in parallel with local governments in many sectors and activities. The increasing importance of CDD operations in the Bank's Philippines portfolio makes it worthwhile to explore how operational approaches fit with the decentralization framework (and indeed with each other). An initial look at the design of two projects ­ the Mindanao Rural Development Project (MRDP) and the Kalahi-Comprehensive Integrated Delivery of Social Services (CIDSS) project ­ shows somewhat different choices regarding the role of local governments. With a government counterpart that expressed a strong commitment to implementation of devolution, MRDP has plunged into the challenge of forging transition mechanisms. It has emphasized participatory mechanisms and complementary roles for local governments and CBOs, channeling about 80 percent of project funds through local governments from the start of devolution operations. Kalahi-CIDSS has focused above all on community empowerment and participation and channels funds directly to community-held subproject accounts while involving municipal and village local governments in project decision making, implementation, and monitoring. Local Government Context Political Dimension The Philippines has three levels of local government units (LGs): province, city and municipality, and village (barangay). The 1987 Constitution specifies that local government shall enjoy local autonomy, revenue raising authority, entitlement to transfers from taxes and "national wealth," and a three-year term of office for all elected officials;39 it also notes the existence of legislative bodies. The President has general supervisory authority, which is delegated to the Department of Interior and Local Government (DILG). The fundamental piece of legislation is the 1991 Local Government Code, which has been praised as being highly advanced. Among other things, it provides for the creation of "local special bodies" (such as local health and school boards and the local development council) to provide policy recommendations and participate in plan development; it requires consultations and public hearings for planning and 38. Sources for this case study include Project Appraisal Documents for MRDP and Kalahi-; the Philippines Public CIDSS Expenditure Review (2003); an untitled draft decentralization report prepared for the World Bank by Julie Slok; interviews with project staff (Carolina Figueroa-Geron of MRDP TTL, Andrew Parker of Kalahi-CIDSS TTL, and Bhuvan Bhatnagar, formerly of Kalahi-CIDSS TTL); Mr. Magistrado Mendoza, Director, Kaisahan; and Ms. Lorena Navallasca of Process Foundation- Panay, on the status of CBOs. 39. From 1982 to 2002, the term of office for village officials was changed several times, finally returning to three years. 41 implementation of projects and for key LG decisions; and it calls on LGs to promote partnerships with people's organizations (the term in the Philippines for CBOs), NGOs, and the private sector. After almost fifteen years, implementation of the local government code remains a challenge. Actual devolution lags far behind the policy vision and, looking specifically at LG-CBO interaction, a number of measures have been implemented inadequately or not at all. The effectiveness of the Local Special Bodies ­ including the local development council ­ has been criticized as weak, for example, and that of the mandated public consultations (which often do not materialize) is likewise questioned. On the positive side, the elections process has become established, and while political elites do continue to dominate, evidence indicates real electoral contests are taking place. Functional Assignments The local government code appears to give LGs clear authority over certain functions, but, at the same time, both LGs and the national government have authority to launch activities in those areas.40 Since national bodies are often still accountable for outcomes, they reportedly try to control LG actions. Section 17 of the code sets out devolution of health, social welfare, agriculture, environmental protection, and local public works and highways to LGs, with responsibilities varying by level of LG. Some of these functions have not been devolved at all, however, and, in other cases, difficulties with devolution attempts (notably in health) have seen the central authorities take back power (and staff). Fiscal Dimension Most LG revenues (approximately 70 to 80 percent) are from central transfers. Opinions differ as to whether transfers adequately cover the cost of devolved functions and staff, and in practice transfers can be delayed or reduced in amount, reflecting national-level fiscal problems.41 The most important transfer, the Internal Revenue Allotment (IRA),42 is largely untied except for a requirement that 20 percent be spent on development activities, but in practice much of this transfer is consumed by large LG wage bills, which tend to crowd out other spending.43 Own revenue sources, which account for the remaining 20 to 30 percent of LG revenues, include a large number of taxes, fees, and charges. Assessment and actual collection are reported to be weak due to pitfalls that include poor data, complex design, noncompliance by payers, and staffing weaknesses. Analyses have noted weaknesses in financial management capacities, which, as might be expected, appear to be more pronounced in lower tier LGs. Problems include lack of compliance with regulations as well as shortages of qualified staff. In spite of the weaknesses and fiscal difficulties, LGs do have at least some funds to spend on development activities, although decisions about these may be imperfect and may not necessarily reflect community preferences.44 40. Concurrent functions create significant overlap and ambiguity. 41. Such delays adversely affect resource predictability and hence undermine local planning. 42. IRA is assigned by a formula based on LG land area, population, and equalization. 43. The lingering central government influence on salaries of devolved staff set off demands for pay increases by other LG staff, with resulting wage-bill growth. LGs also hire temporary or contract staff to circumvent government influence, reportedly engaging in some nontransparent salary practices. The full extent of personnel costs may be hidden, since some are classified under other budget line items. 44. Provincial LGs, for example, appear to prefer funding construction of concrete roads (less so maintenance of those roads or construction of gravel farm-to-market roads and their maintenance) and rural water-supply systems, but providing little or nothing to communal irrigation. World Bank 1999, 82. 42 Administrative and Capacity Dimension The devolution of staff has added to LGs' pool of technical staff, which include engineers, planners, accountants, procurement staff, social welfare officers, and sectoral specialists (such as agricultural and fisheries technicians).45 Since local governments are perceived as offering poor career prospects, some devolved staff have sought to leave, and recruiting qualified new staff can be difficult. In an environment where patronage and elite capture are deemed to be high, the three-year term of office has been criticized as yet another factor that encourages short-term, visible infrastructure projects over more substantive medium-term policies with less immediate or less visible payoffs. The local government code allows for joint ventures and other cooperative arrangements with CBOs (referred to as people's organizations), NGOs, and the private sector for service delivery, capacity building, and livelihood projects. The code calls for a planning body, the Local Development Council, at each level of LG, to be chaired by the LG chief executive. Only half or fewer of LGs have created councils, however, and of these only about a third have involvement with NGOs or CBOs. Village level planning is usually the weakest, with development plans tending to be lists of projects rather than strategic documents. By default, higher level LGs tend to develop plans based on their assumptions of what the lower levels need. CBO Context46 Political Dimension The Philippines has, in general, a dynamic civil society. Together with NGOs, CBOs have been active in taking advantage of the Local Government Code's provisions to become involved in development issues, for example, via the local special bodies and local councils. Over the past fifteen years, CBOs have become increasingly vocal and confident enough to approach government agencies themselves (whereas in the early days they may have required NGO support). At least three options are available for registering (sector-specific options may exist as well): with the Securities and Exchange Commission, with the Department of Labor and Employment, or with the Cooperative Development Authority. CBOs may hold bank accounts, as long as they have established themselves as legal entities via registration. Functional Assignments No external prohibitions bar CBO involvement in functions or sectors, and in practice CBOs provide services in practically all areas of community life. Examples of issues around which CBOs organize include environmental protection and management, reproductive rights, maternal health, farming and fishing, governance and citizenship, and access to justice. Cooperatives also are involved in credit, enterprises, and training. CBOs are direct producers of services and are also allowed by law to partner with local governments, with or without contracts. Fiscal Dimension 45. Some 70,000 staff had been devolved by 1997. The local chief executive has the power to set the staffing pattern and organizational chart, subject to approval by the Sangguniang (local council). The chief executive also appoints the municipal administrator, whose term of office is coterminus with chief executive's. Other LG staff are career professionals who have passed the civil service exam. 46. The information in this section is based on an interview with Ms. Lorena Navallasca of Process Foundation-Panay and e-mail communication from Mr. Magistrado Mendoza of KAISAHAN. 43 CBOs generally have very small resource bases. Major sources of funding include foreign organizations (for example, UNDP has a small grants program), private organizations, service contracts, members' contributions, and fund-raising events. Some limited funds may be available from government bodies (for example, under contracts or from IRA transfers by village councils). In general, CBOs adopt participatory processes for planning, budgeting, implementation, project monitoring, and their other activities. Administrative and Capacity Dimension Because of the dynamism of the NGO sector, CBOs have access to training and assistance in areas such as community organizing and leadership and identifying, prioritizing, and planning for community concerns. CBOs might be expected to have at least twenty-five members and a leadership, selected in accordance with their articles of incorporation, constitution, and by-laws. In some instances, NGOs help organize CBOs and facilitate discussions, prior to elections, on questions such as the characteristics of a good leader, the meaning of leadership, and so on. In rural areas, educational attainment is fairly low. (Members typically have elementary school educations, but particularly in mountainous areas some may be illiterate.) People are well aware of their rights in each sector, however, and appeal to the government on this basis. Mapping the Opportunity Space * Considering the above, the opportunity space for LG-CBO accountability and coproduction arrangements appears to be the following: The Philippines Opportunity Space Dimension LGs CBOs Political Partly enabled Enabled Functional Partly enabled Enabled Fiscal Partly enabled Partly enabled Administrative Partly enabled Partly enabled For the above opportunity space, one would expect to find a wide range of coproduction and accountability partnerships covering provision of a variety of services, financing, and implementation between local governments and CBOs. This would generally be possible, albeit requiring adequate support and monitoring, including oversight by the main project counterpart and donor. CDD Projects and LG-CBO Partnerships The two CDD operations reviewed in this case study both emphasize participatory approaches and integrate into project design coproduction and accountability relationships between local governments and CBOs (as well as other partners). They nevertheless differ in emphasis, with one leaning more toward local governments and the other toward communities and CBOs. The particular nature of each program is further explained below. Mindanao Rural Development Project (MRDP) Project Overview MRDP was APL approved in late 1999 and became effective in March 2000, with four phases expected to span some fifteen years. The project aims to reduce poverty among indigenous peoples by improving income and food security through "implementation of better targeted agricultural and fisheries-related 44 rural development and biological diversity conservation programs, and improved LG institutional, management and financial systems."47 Its four components are (1) rural infrastructure, including roads, community irrigation, rural water supply, and capacity building in infrastructure management; (2) community funds for agricultural development (CFAD),,a municipal-level fund financing diverse, community-driven subprojects; (3) rural development planning and resource allocation and agriculture and fisheries productivity enhancement, which, among other things, supports improved LG capacities for development planning, resource allocation, financial management, and monitoring and evaluation;48 and (4) coastal/marine biodiversity conservation. The government counterpart, the Department of Agriculture (DA), stands out for its strong desire to devolve responsibilities so that LGs rather than the DA handle front-line service delivery and to ensure sustainability and mainstreaming of institutional reform achievements under MRDP.49 MRDP's design fits tightly with this objective and aims to create a model transition mechanism that the DA will be able to apply elsewhere in the Philippines. The project works directly with all three tiers of LGs, and from its very beginnings it has channeled funds for the rural infrastructure and CFAD components via LGs. Initially, in year 1, funds flowed to the provinces possessing adequate financial management systems; thereafter the flow shifted to individual municipalities as their financial management improved to meet acceptable standards, with monitoring by provincial LGs and the DA's Project Coordination Office.50 (Many municipalities had adequate capacity at the onset of MRDP1.) In addition to providing learning-by-doing and other support to develop abilities at all LG tiers, by facilitating the DA's transition from implementer to facilitator of local-level planning and management of development, MRDP is also creating the necessary space above LGs to increase the chances that their new skills can be applied in a functioning devolved framework, thus making them better partners for communities (for example, by encouraging a more robust bottom-up planning process). Coproduction and Accountability Relationships While a detailed analysis of LG-CBO partnerships is beyond the scope of this case study, a review of project design shows coproduction and accountability arrangements throughout the subproject cycle, from identification (and earlier) stages through operations and maintenance. The following are illustrative. Coproduction · Both LGs and communities contribute in cash and in kind to CFAD.51 · Together with LGs and NGOs, CBO members are trained in sustainable marine and fisheries management, so that they can educate residents and community leaders, thereby enhancing community participation in identifying, planning, establishing, and operating protected coastal areas. · CBOs participate in operations and maintenance of water supply projects, with Rural Waterworks and Sanitation Associations and Village Waterworks and Sanitation Associations levying water charges for the operation and maintenance of LG- 47. World Bank 1999, 2. 48. LG institutional capacities are deemed a major bottleneck in rural development. World Bank 1999, 13. 49. Consistent with this, the DA created a Project Coordinating Office staffed by its own employees and staff from participating LGs, rather than use a consultant-staffed Project Management Unit. 50. The potential risks of channeling CFAD funds via municipalities are balanced by the relatively small sums involved as well as by the benefits from enhanced municipal capacity through learning-by-doing. 51. Municipal LGs contribute 25 percent of the initial overall fund, and communities contribute a minimum of 10 percent in kind, depending on subproject type. 45 constructed projects, such as rehabilitation and construction of point-sources and communal faucets. For communal irrigation, both LGs and the Irrigators' Associations contribute toward operations and maintenance,52 with LG staff receiving training in communal irrigation management (previously handled by a central agency) so that they can better support the Irrigator's Associations. Accountability · The Municipal-level Multi-sectoral Committee brings together municipalities, representatives from rural communities, NGOs, and private sector representatives to manage the subproject selection process in CFAD and to prioritize municipal roads. Communities are involved in selecting types of interventions and locations under CFAD, as well as, under the rural infrastructure, subprojects concerning rural roads and the water supply.. · Farmers are involved in preparing proposals for rehabilitating community-owned run-of-river communal irrigation schemes. Municipalities prepare monthly balance statements and progress reports on CFAD, which they provide to village LGs and CBOs. Village working groups in turn post subproject information in public places. · Moving up through the LG tiers, provincial and municipal Agricultural Development Plans must reflect consultations with communities (as well as with technical agencies). From Design to Real Life: Experience of Phase 1 Bringing about institutional change and skill enhancements in LGs and communities has not been easy ­ but it has proved possible. The midterm review, while confirming that MRDP's concept and design were valid, rated the project unsatisfactory for reasons including the following: LGs did not follow-up to ensure activities were done on time and to project standards; LGs were overly involved in procurement and management of CFAD subprojects, contrary to guidelines putting beneficiaries in the lead; difficulties arose in integrating community priorities into the development plans of municipal and provincial LGs and the DA's regional branch; weaknesses emerged in the organization of beneficiary groups and their coordination with support groups at the municipal and provincial levels; and disbursement levels were low. Yet within eighteen months, the project had achieved a strong turnaround, and when phase 1 closed at the end of 2004, it was rated satisfactory. Building on the fundamental commitment of all parties (and the reinforcing involvement of oversight bodies including the Departments of Finance, Budget, and National Planning) and recognizing the need to do better led to a number of improvements. All LG tiers increasingly incorporated community priorities into their development plans and investment programs and improved their ability to prioritize their resources according to community needs; the project coordinating office recruited staff with needed skills (for example, financial management); new systems were established to track LG compliance in submitting audit reports and financial statements; additional community organizers went to the field to help CBOs better organize their communities; LGs developed greater appreciation for participatory processes, which they began to apply in their non-MRDP activities; and CBOs and LGs developed stronger relationships. Numerous institutional challenges remain for subsequent project phases, but substantial progress appears to have been made. Kalahi-Community Integrated Delivery of Social Services (CIDSS) 52. They are expected to contribute larger shares over time, reflecting increased income to communities from higher agricultural production and increased LG revenues from levies on agricultural products. 46 Project Overview Kalahi-CIDSS objectives are to "to strengthen community participation in local governance, and develop local capacity to design, implement and manage development activities."53 Three components support these objectives: (1) community grants, largely supporting simple community infrastructure, such as access roads, clean water, schools, and health facilities; (2) implementation support to formal and informal local institutions; and (3) monitoring and evaluation. The government counterpart is the Department of Social Welfare and Development (DSWD). The project became effective in December 2002, and participants expect a six-year implementation period. This project places a strong emphasis on involving communities (in their entirety, rather than targeting subgroups) in bottom-up local development and on facilitating implementation of the decentralization framework. By design, it did not create any parallel structure but rather has taken steps to bring to life village- level entities that, while mandated under the Local Government Code, in practice have often not been formed or have been ineffective. Important among these are the village assembly (which often is not convened or is poorly attended) and the village council, the lowest unit of local government. While the village council is best connected to community wants and needs, its bare bones core structure (a handful of elected councilors and one or two staff) means that its operational capacities are quite limited without a set of local government code-mandated committees (composed of villagers and councilors) that, in Kalahi-CIDSS communities, have been revitalized through their role in the project. As part of strengthening community participation in local development, and because of issues of capture and collusion in the municipal LGs through which funds to villages usually flow, Kalahi-CIDSS transfers money directly to village-level subproject accounts that are jointly managed by the community and village LG.54 LGs (principally municipal and village tiers) are involved throughout project phases and subproject activities and also receive training. In this manner, the project hopes to improve LG capacities, to encourage LGs to adopt Kalahi- type processes in their non-Kalahi activities, and to prepare them to take over Kalahi programs in their territory when the project ends its involvement (after three annual funding cycles in each participating municipality).55 The fact that some subprojects identified but not funded under Kalahi have subsequently been funded by LGs appears to indicate some success in integrating community preferences into regular LG operations.56 Coproduction and Accountability Relationships The project emphasizes working with entire communities, rather than with subgroups, and revitalizing communities, meaning it builds into the subproject cycle collaboration among CBOs, village committees attached to village councils, and LGs.. The following points are illustrative. Coproduction · Both communities and LGs must contribute to subproject costs.57 53. Bhatnagar and Burkley 2004. 54. The three signatories to the account are the village treasurer (who is legally responsible for the funds), a Kalahi staff member, and a villager (who may often be the village council head). The option of transferring some funds to villages LGs as a block grant, rather than for specific subprojects, has also been discussed for possible future implementation, though no decision has been taken. 55. Bhatnagar and Burkley 2004. 56. Ibid. 57. Though no minimum is set, the size of counterpart contributions is a criterion in project selection, and in practice LG and community cost-sharing has added up to an average 40 percent. Village contributions include cash and in-kind as, apparently, do LG contributions (staff time and land are both counted). 47 · Subproject preparation involves cooperation among project preparation teams, village representative teams, a municipal inter-agency council, and municipal technical staff, as well as NGOs.58 · Implementation is handled by village teams, who can receive assistance from municipal and village LG staff, project staff, or hired expert consultants, as needed. This cooperation is facilitated by training, available to both CBOs and LG staff, on technical aspects of implementation (construction, reporting, procurement, financial management, operations, and maintenance). · CBOs are responsible for operations and maintenance (funded from fees collected by user associations); municipal staff and area coordination teams monitor progress and offer help, as needed. · A memorandum of understanding identifies the permanent LG staff assigned to the project. In at least one case, a municipality has hired (and paid) engineers, facilitators, and financial analysts to take on project-related work.59 Accountability · The mayor chairs (but does not vote in) the municipal inter-village forum (MIVF), which selects proposals according to rules and criteria agreed to by elected village representation teams and makes indicative fund allocations.60 Prior to and following these meetings, village assemblies meet several times to discuss and approve progress, approve the results of the MIVF, elect members to a village subproject management committee, and validate detailed proposals.61 · The mayor also chairs the municipal inter-agency committee that brings together LG department heads, deconcentrated national agencies, and NGOs to ensure that Kalahi infrastructure projects involving multiple departments or affecting as yet undevolved sectors receive the necessary operating supplies (for example, textbooks, medicines, and education and health staff). · Kalahi subprojects are included in municipal development plans to ensure sustainability. The project also helps municipal LGs to prepare plans for incorporating Kalahi-style participatory processes for planning and resource allocation with the intention of spreading institutional benefits to nonproject activities. The project has now completed two annual cycles in a number of communities. Anecdotal evidence indicates some success in effecting behavioral changes, notably the adoption by some LGs of Kalahi-CIDSS approaches for their general, nonproject operations. Impressions from municipal elections in 2004 suggest that mayors supportive of Kalahi generally did better than who did not support the project. Filling the Opportunity Space What, then, do these two projects indicate about where LGs and CBOs end up in the real world? Are their actual partnerships consistent with the predictions that arise from assessments of whether their contexts are constrained or enabled? 58. The municipal inter-agency council is a coordinating body of sectoral representatives. Depending on the sector, members may be deconcentrated central agency staff or municipal employees. 59. Bhatnagar and Burkley 2004. 60. In two rounds, first reviewing initial proposals and subsequently reviewing detailed versions. 61. Village assemblies are mandated under the local government code, but often they are not called or do not reach the required 50 percent plus one majority. Village assemblies under Kalahi appear to have no problems meeting and exceeding the attendance requirement. 48 First, it must be acknowledged that various forms of capacity building help local governments and CBOs to improve their technical and administrative skills; problems are identified and tackled early on. These advantages mean that participating CBOs, in particular, enjoy a relatively more enabling environment than do CBOs trying to go it alone. Local governments also benefit, of course; but being more complex and, embedded as they are in the intergovernmental system, subject to more de jure and de facto regulation than are CBOs, LGs might be expected to have more difficulties rising above constraints. On the other hand, the greater absolute value of their resources in terms of staff and budget also suggests that, given the political will, a well-tuned system to guide LGs in the right dimension can spark significant benefits. LGs participating in both projects seem capable of entering into partnerships involving staff, money or in-kind resources, and technical support to communities. MRDP's midterm turnaround experience appears to indicate that, in the Philippines, municipal LGs already have the space to be reasonable development partners in all four dimensions. At the village LG level, Kalahi-CIDSS has given new life to legally mandated governance structures by giving them responsibilities for subprojects. Support and guidance have, of course, played a critical role in these positive developments. Among the factors that appear to be important in building and maintaining LGs' partnership potential are providing real opportunities and incentives to LGs to perform (including access to funds); creating oversight mechanisms and checks within existing institutional structures to help keep them on track; and various other forms of support ranging from capacity building to guidance from the sector ministries and other partners. 49 APPENDIX 2. ZAMBIA: PUSHING FOR CHANGE UNDER A STALLED DECENTRALIZATION FRAMEWORK62 Zambia exhibits a weak decentralization setting, with long-standing discussion and debate that for many years yielded little real progress while raising serious doubts about the extent to which the government was sincerely interested in decentralization. A decentralization policy (with devolution as the ultimate objective) was finally approved in late 2002, but it launched only in August 2004. Policy implementation is still in its early days, and it is not clear whether the real implications of what devolution really means ­ both in terms of yielding power and of implementing the practical processes and actions required ­ are well understood. Not surprisingly, Zambia's central government (including its deconcentrated arms) has continued to dominate local governments, the real authority, technical service-delivery capability,,fiscal standing, and administrative capabilities of which appear to be highly constrained. CBOs, though facing a tough environment marked by widespread poverty and inadequate resources, nevertheless seem to manage to function reasonably well, when given appropriate support. Faced with a complex setting with little progress on decentralization, the Zambia Social Investment Fund (ZAMSIF) took a calculated risk in shifting from working with communities and the deconcentrated government ­ the approach used in predecessor projects ­ toward providing a greater role for local governments. Among other considerations, ZAMSIF planners hoped that this might contribute to bottom- up demand for decentralization. While this strategy has yielded some advances in terms of developing local government capabilities, reality has lagged behind initial expectations. Local Government Context Political Dimension The authority of local governments under Zambian law is subject to significant central government control. The constitution calls for a local government system to be prescribed by parliament, with councils to be democratically elected. The 1991 Local Government Act enables councils to "establish by- laws, hire and fire their own employees, raise revenues, borrow, dispose of, or acquire assets."63 The Minister of Local Government and Housing must approve LG actions in most of these areas (such as by- laws and budgets) and also has the authority to revoke by-laws, suspend councilors and staff, or appoint an administrator for the councils when she or he deems it necessary.64 The act is now under review as a step towards aligning it with the recently launched decentralization policy. Each council65 is divided into wards, each electing one councilor for a five-year term based on a first- past-the-post electoral system. Mayors or chairs are elected by the councilors, though the recent decentralization policy calls for direct election by citizens. While any transition to democracy takes time 62. Sources for this case study include ZAMSIF PAD; World Bank 2003d; World Bank 2003e; Ministry of Local Government and Housing, Draft Report on Decentralization and Local Government, December 2002; World Bank 2003c; and The Local Government System in Zambia, Commonwealth Local Government Forum, 2005, posted at www.clgf.org.uk. In addition, interviews were conducted with individuals with in-depth knowledge of ZAMSIF: Wim Alberts (current task manager), Laura Frigenti (former task manager), Steen Jorgensen (task manager of predecessor projects SRP I and II and participant in the ZAMSIF design process), Helen Mbao (Social Development Officer, Zambia country office), and Ed Mwale (Operations Coordinator, ZAMSIF Management Unit). 63. ZAMSIF PAD. 64. ZAMSIF PAD, Annex 11, Structure of Local Governments. 65. As defined in the Decentralization Policy, a district "council is a body of democratically elected representatives responsible for policy formulation and delivery of services in a given geographically defined area." 50 and will be imperfect, the advent of the multiparty system in Zambia is apparently being felt in contested elections and in citizens' demands that councilors respond to their needs. Functional Assignments The 1991 Local Government Act is broad ranging and largely vague regarding responsibilities for functions. It assigns city, municipal, and district (rural) councils sixty-three different functions in nine basic areas: administration, advertisement, agriculture, community development, public amenities, education, public health, public order, and sanitation and drainage.66 The Act does not clearly state, however, which functions are the sole responsibility of either the local authorities or the central government and which are to be shared. This can result in a large number of parties being involved in service delivery, with high potential for confusion; for example, eleven different parties are involved in the development and maintenance of water-supply systems.67 All councils (city, municipal, and district) are assigned the same responsibilities, with a few additional tasks for city and municipal councils.68 Councils are able to contract-in, contract-out, or enter into agreement with other organizations to carry out their work. In practice, councils fulfill only a small number of their formal responsibilities under the Act; most responsibilities have remained in the hands of central ministries or have been returned to them allegedly because of the councils' poor performance. In particular, health, education, and agriculture services remain the domain of deconcentrated government. Councils appear to deliver only "traditional local service, such as markets, bus stops, street lighting, grass cutting and road maintenance, construction permits, water, sanitation and drainage, some public health (abattoirs, inspection, pest control, refuse removal, etc.), and some low-cost housing."69 Fiscal Dimension While the Local Government Act requires the central government to make grants to LGs for services,70 in practice grants have been decided on an ad hoc basis, passed via the Ministry of Local Government as funds are available rather than on a set schedule. The transfer amounts have also been criticized as inadequate and indeed have been declining over time. For smaller councils, grants may represent 30 percent of total revenues, whereas for larger councils they might represent zero or a negligible amount. In March 2005, a meeting titled "Stakeholders Workshop for Local Governments in Zambia," hosted by the Ministry of Local Government and Housing (MLGH), resolved that MLGH should use the formula that ZAMSIF had developed for allocating the project funds that would have been made available in the 2005 budget (the "indicative planning figure") as the basis for allocations to individual local governments.71 66. More specifically, these include "maintenance of law and order; control of the movement of livestock, slaughtering of animals, and sale of meat; storage, market and preservation of agricultural produce, conservation of natural resources and prevention of soil erosion; provision of and maintenance of clean water; construction and maintenance of public roads; refuse removal and establishment and maintenance of sanitation systems; registration of births, deaths, marriages, clubs, and enumeration of persons or property connected with the administration of council areas; community development; street lighting; firefighting; public amenities such as parks, zoos, gardens; establishment and maintenance of cemeteries; establishment and maintenance of colleges, schools and day nurseries; establishment and maintenance of hospitals, clinics, and environmental health centers." Zambia, government of, 2002. 67. Zambia, government of, 2002.. 68. City and municipal councils also are responsible for maintaining trunk roads and serve as planning authorities under the Town and Country Planning Act and the Lands Act. 69. ZAMSIF PAD. 70. Specifically, for health, fire, roads, police, primary education, and agriculture, though in practice health, education, and agriculture remain centrally controlled. 71. E-mail communication from Ed Mwale, 2 June 2005. 51 LG fiscal autonomy to levy fees and charges and to decide on expenditure allocation is constrained by the approval authority of the Minister of Local Government and Housing over most revenue-raising decisions as well as over budgets. The central government has also taken away some levies, leaving LGs to search for other sources of funds. Own-revenue generation is low, reflecting small revenue bases as well as collection and management difficulties. These include poor monitoring of services for which fees are charged, reluctance to increase tariffs for political reasons, weak planning and budgeting, and a lack of expenditure discipline.72 Administrative Dimension Councils are fully responsible for their wage bill and have the power to hire and fire staff, although the central government has some influence (for example, the Local Government Act specifies educational qualifications for managerial posts). Past analyses have pointed to significant overstaffing (probably more so in larger urban councils than in rural ones), which likely crowds out other spending. Overall, some 53 percent of expenditures are for personnel costs; for individual councils this can be as high as 60 percent.73 A large proportion of council staff consists of low-skilled laborers, the legacy of past policies aimed at providing employment and income. Even when a council wishes to trim its staff, retrenchment has been difficult due to high severance payments. Nevertheless, many councils have reportedly started gradually reducing staff numbers by not filling vacancies or by paying severance when their budget allows, and in 2005 the central government allocated funds to finance severance payments, in hopes of enabling councils to reduce payrolls. Recruitment of appropriately qualified staff appears to pose a problem, particularly in rural areas seen as too remote and low in prestige. In addition, many months of salary arrears or nonpayment reduce the overall appeal of council positions and certainly affect the willingness and enthusiasm of council staff to undertake challenging, time-intensive work on donor-supported projects like ZAMSIF. District Planning Officers ­ whose involvement is important to the planning process generally and to the LG's role in ZAMSIF ­ have been cited among those professional positions in which rapid turnover has been harmful. Much of the technical expertise for local development remains lodged within deconcentrated central ministries at the district level (for example, education officers, buildings' supervisors from the Ministry of Works and Supply, agriculture and forestry officers from the Forestry Department, and social welfare officers), and communities of necessity look to these for cooperation and assistance. Deconcentrated district staff appear to advise project committees on procurement matters. CBO Context Since little analysis of the situation of CBOs was available, the following is based on interviews and project literature.74 Political Dimension No overall legal framework per se exists for NGOs or for CBOs. While such entities are able to register in the Register of Societies, the process is lengthy and cumbersome and many if not most CBOs do not do 72. Zambia, government of, 2002. 73. Figures cited in The Local Government System in Zambia, Commonwealth Local Government Forum, 2005, posted at www.clgf.org.uk, and the ZAMSIF PAD, respectively. 74. Interviews with Helen Mbao, Social Development Officer, World Bank country office, and Ed Mwale, Operations Coordinator, ZAMSIF MU. 52 so. Even unregistered CBOs can open bank accounts in the CBO's name, however, and they can enter into contracts. CBO operation is regulated by accepted local practice. Discipline is imposed by peer pressure, risk of lawsuits for abuse of public funds and, under ZAMSIF, monitoring by line departments and councils. CBOs are more evident and active in urban areas, much less so in rural areas, although they do form around particular interests when needed. Functional Assignments In the absence of a legal framework, there are, correspondingly, no restrictions on the spheres of CBO involvement. Education is the one sector in which an explicit formal basis exists for CBO involvement, via school-based PTAs whose creation is encouraged by the government. In practice, CBOs are active in a large number of sectors. To illustrate, ZAMSIF works with CBOs in health, education, environment, HIV/AIDS, community roads, gender, water and sanitation, and initiatives for vulnerable groups. Fiscal Dimension Given poverty levels, raising money from sources other than donors is difficult. CBOs might be able to gather enough money to send a community leader to the district to lobby or to lodge a complaint, but normally operate with in-kind contributions (for example, materials, labor, foodstuffs, or meals during activities). Some communities were found to be too poor even to contribute materials to ZAMSIF projects. ZAMSIF has seen some success in encouraging CBOs to approach potential donors beyond the project, to make independent proposals to other donors, and otherwise to advance their cause through such active approaches. Such efforts are relatively new and unusual. Administrative Dimension Some CBOs (notably in the education and health sectors) are particularly well-organized and possess literate and active leadership; in practice, this may skew community subprojects toward these sectors, since they have the most capable advocates. CBOs provide materials and direct labor as well as procuring some materials and services and do costing and budgeting themselves. Mapping the Opportunity Space The opportunity space for LG-CBO accountability and coproduction arrangements can be summarized as follows: Zambia Opportunity Space Dimension LGs CBOs Political Constrained Partly enabled Functional Constrained Partly enabled Fiscal Constrained Constrained Administrative Constrained Partly enabled Given the above described opportunity space, one would expect councils to have difficulty entering into robust service delivery partnerships and both councils and CBOs to have difficulty providing funds for projects. Likewise, CBOs and communities would presumably seek accountability not so much from councils as from the deconcentrated departments with the technical expertise and resources to assist them. 53 CDD and LG-CBO Partnerships: The Zambia Social Investment Fund (ZAMSIF) Project Overview ZAMSIF was conceived as a ten-year adaptable program loan, with a first phase running from 2000 to the end of 2005. It followed on two Social Recover Projects (SRP I and II), which, designed in the late 1980s as the country was emerging from repressive one-party rule, worked directly with communities (and the deconcentrated technical staff of central ministries) to attempt to deliver quick results and to open up space for the tough macrolevel reforms that would be coming down the line. Some ten years later, one of the challenges of preparing ZAMSIF was to build up local governments in the continued absence of a robust decentralization framework, thereby fostering more sustainable impact and creating a project exit strategy that would leave capable local councils in charge.. Indeed, the hope was that the project might spark a bottom-up demand for decentralization. The Bank's preparation team drew on decentralization expertise, local knowledge, and the positive experience of SRP, under which deconcentrated officials had been able to work effectively with communities. The project directly supported one of three key strategic priorities in the Country Assistance Strategy, that of fostering local government and decentralization.75 Also, at the same time, the Bank was preparing a public service capacity-building project that included, among the five components for its first phase, supporting the preparation of an implementation strategy for the government's (as yet unadopted) decentralization policy. These factors seemed to bode well for an emphasis on local government- community partnership. To support local government performance, the project devised a process of graduation through five levels that would align a local government's demonstrated capabilities with its responsibilities in each of two ZAMSIF components: a community investment fund (CIF), financing subprojects in individual communities, and a Direct Investment Fund (DIF) to be managed by local governments, financing capacity building as well as providing funds for district-level subprojects serving multiple communities.,. As expected, at project launch the majority of local governments were in the lowest levels (1 and 2) and had little real power. Meanwhile, with decentralization efforts stagnating, sectoral technical expertise remained largely in deconcentrated units of central government bodies reporting to Lusaka (as is the case to this day) rather than to elected local councils. Since the government instituted the District Development Coordinating Committees (through Cabinet Circular No. 1 of 1995) to serve as advisory forums through which line departments, local governments, donors, and NGOs could coordinate development activities, the ZAMSIF design specifically required, as a means of creating an incentive for line ministry staff to cooperate with local governments, that DDCCs work with local governments (district councils) in implementing the Indicative Planning Figure to support community-type projects. District councils are also required to have a District Planning Officer on staff and should have commenced a district planning process to be eligible for project support under the District Investment Fund. This funding was available to districts that were at least at level 2 on the capacity ladder. In practice, ZAMSIF came up against a number of stumbling blocks. While decentralization continued to be the subject of much discussion, real advances were slow (a policy was finally adopted in 2002, but it was launched only in 2004); technical expertise and authority remained with central line ministries and deconcentrated ministries, reporting to Lusaka or the provincial level. Capacity building aimed at district councils produced only modest gains, and by mid-2003, no councils had moved beyond the two lowest levels in the five-tier graduation ladder. With lackluster performance on community subprojects and on DIF projects and lagging disbursements and commitments, the project was declared unsatisfactory at its midterm review in 2003. 75. It is the impression of some observers, however, that country management put more priority on the other two CAS pillars: improving transparency and accountability and strengthening public-sector efficiency and capacity. 54 A restructuring ensued. This included pooling uncommitted CIF funds and making them available on a first- come-first-served basis (rather than having separate budgets per LG); the presumed bias toward better- performing districts was mitigated by earmarking a portion of the total for particularly weak communities and districts. In addition, some DIF component funding and staff were shifted to the CIF, and capacity-building assistance to local governments in the lower levels of the graduation scale was reoriented toward more basic hand-holding and implementation-oriented support, rather than reserved for more abstract learning activities targeting more developed local governments. The project's development objective was modified to stress a more gradual transition towards accountability relations between local governments and communities. ZAMSIF subsequently regained a satisfactory rating, with commitments and disbursements increasing. The emphasis on working more with better-performing governments yielded results in project implementation. By mid-2005, a handful of district councils had reached level 5, with a slightly larger number (six) in level 4; almost all of them were carrying out their responsibilities with relatively little trouble. From the perspective of improving local governments capabilities more broadly and of an initial project exit strategy of leaving behind capable district councils, however, it appeared that practical difficulties ­ the state of decentralization among them ­ had necessitated a rethinking of the initial high expectations. Coproduction and Accountability Relationships The following are examples of ZAMSIF coproduction and accountability relationships involving CBOs and local governments or CBSs and the central government where local governments are not empowered. Coproduction · Communities are required to contribute around 15 percent towards subproject implementation; for some particularly poor communities, this has proved too high a hurdle. ZAMSIF has been flexible, however, and has funded subprojects with community contributions even below 10 percent as long as the communities could demonstrate and express a felt need through active participation in meetings to identify community needs. · While the PAD does not cite a specific cofinancing requirement for local governments, other project elements ­ such as the requirement that LGs have filled the district planning officer position ­ do engage LGs in coproduction. A field study found that LGs were generally willing to contribute to ZAMSIF costs with staff time and material support (such as vehicles, for example, when available);76 in practice, however, local governments can find themselves without adequate operating budgets to pay, for instance, for fuel for staff visits to communities. · CBOs are involved in operations and maintenance. Parent Teacher Associations, for example, raise funds for operations and maintenance costs for school projects. In the case of a health center, a community might form a committee to ensure that the central government provided supplies as promised or set up user fees and implement income-generating activities.77 Poverty levels and the vulnerability of certain beneficiary groups can, of course, strain a community's ability to contribute to operations and maintenance. · At desk appraisal, sector ministries (for example, health and education) commit to paying recurrent costs. In practice, however, actual disbursements have been lower than budgeted amounts. 76. World Bank 2003c. 77. World Bank 2003d, Annex III: Review of MTR Questions and Issues. 55 Accountability · District institutional structures ­ including the District Development Coordinating Committee and its District Planning Subcommittee, the full Council, and its Plans, Works and Development Subcommittee ­ all play various roles in ZAMSIF, including appraising projects, providing technical inputs to subproject budgets, and approving recommended subprojects. · Communities organize committees to advance their interests, participate in planning for subprojects, and manage subproject finances and operational needs. As noted in the midterm review, committee links to district and ward development planning bodies were still fragile. Filling the Opportunity Space Project experience appears to have borne out what one would expect from the review of the extent to which local governments and CBOs are enabled or constrained: while good will may be in place, practical constraints such as staff capacities and budgets, as well as the constrained role of local governments in a very weak decentralization context, appear to have inhibited LG involvement with CBOs and communities. With key sectoral ministries retaining control over service delivery, CBOs are better served by seeking technical and other support from deconcentrated bodies that can help them in the day-to-day implementation process and that control the funds needed to operate and maintain subprojects once they are completed. 56 APPENDIX 3. NICARAGUA: EXPANDING THE OPPORTUNITY SPACE FOR LOCAL PARTNERSHIPS THROUGH CDD OPERATIONS Over the last fifteen years, Nicaragua has gone through a gradual process of decentralization. During this period, the Bank supported six CDD operations with two agencies that played a critical role in the country's decentralization policy. It supported four operations with FISE, the Nicaraguan Social Fund, and two operations with INIFOM, the agency responsible for municipal strengthening. These operations illustrate (i) contrasting proposals for exploiting a specific decentralization environment to promote local partnerships, as well as (ii) the potential and limitations that CDD operations can have in reforming the environment in which they function. Furthermore, it shows how "healthy competition" between Bank operations also influences a task team's perceptions about the opportunity space for promoting local partnerships. This discussion is organized around three chronological periods, characterized by the nature of the Bank's operational response to the country's decentralization environment. In a period of mostly missed opportunities between 1990 and 1995, FISE 1 and FISE 278 focused on short-term gains and failed to seize the limited opportunites for building local partnerships offered by Nicaragua's decentralization framework, despites its systemic shortcomings. In the period between 1996 and 2000, first INIFOM and then FISE 3, in response to constitutional reforms and a new decentralization law that somewhat improved the enabling environment, introduced pilots addressing shortcomings in the enabling environment and promoting accountability and coproduction partnerships The third period, from 2001 to 2005, saw scaling up and some institutionalization of these programs and further policy reform. FISE 4 and INIFOM 2 succeeded to a large extent in fostering a significantly improved enabling environment for local partnerships (through, for example, the Municipal Planning System and the Fiscal Transfers Law), even though new challenges continue to arise. Each of these stages is treated in more detail below. Period 1: Missed Opportunities (1990­1994) The Environment The foundation for local government during this period was the Municipal Law of 1988 (Ley de Municipios: Ley No. 40). During the first four years of the Chamorro administration (1990­1996) the main advances in decentralization were the creation of INIFOM (Instituto para el Fomento a las Municipalidades) in 1990 and the establishment of 143 municipalities in 1991 (a number that over the years would grow to the current 152 municipalities). The role of INIFOM was to promote the institutional development and strengthening of Nicaragua's municipal governments. Political Dimension During this period all important decisions were made by the Municipal Council, which had ultimate authority. Municipal councilors were elected by the local population, and the municipal council then elected the mayor. This system had two major accountability problems: · First, without direct election of mayors, the possibilities for citizens to hold the executive to account were constrained. · Second, the electoral system had a strong urban bias. Councilors were elected proportionally from political party lists, which tended to place candidates from urban areas first, with the result that a disproportionate number of councilors were urban 78. While these are not the actual names of the loans, they are used for the sake of simplicity. The actual names of FISE are Social Investment Fund I, III, III, and Poverty Reduction and Local Development Project. The actual names of INIFOM are Rural Municipalities Project (also known as PROTIERRA), and Second Rural Municipalities Project. 57 residents even when rural and urban residents voted in similar proportions. Roughly, it was calculated that on average only 13 percent of councilors resided in rural areas, while rural populations contributed roughly 50 percent of all votes. Hence, municipal councils did not internalize the concerns of rural communities as well as they did those of urban ones, and they were not well positioned to initiate activities directed at sustainable rural development, rural poverty, the environment, and natural resource management. Legal, Functional, and Regulatory Context The 1988 law authorized local governments to provide services and exercise administrative authority over (1) environmental protection; (2) regulation of urban development and land use; (3) solid-waste management; (4) reforestation; (5) oversight of natural resource use and protection; and (6) creation of mechanisms for popular participation in local decision making. While the number of functions assigned to municipal governments was very limited, much scope remained for delegating certain functions through principal agent arrangements. This explains the multiple decentralization pilots in health and education79 that the Chamorro administration began during this period and demonstrates that a certain space existed for promoting local partnerships between CBOs and LGs. Prominent among these efforts were the following: · Local Systems for Comprehensive Health Assistance (SILAIS- Sistemas Locales de Asistencia Integral de Salud). This was an effort to deconcentrate the health system to the regional level. One of its more promising innovations was the involvement of local governments and citizens through the creation of Local Health Councils. These regional councils were formed by the mayors of the municipalities, community organizations, doctors, and health workers. The expectation was that these councils would have planning and budget execution responsibilities. While the program was piloted in some municipalities, such as Matagalpa, the change of government in 1997 discontinued this initiative. · School Autonomy and Delegated Administration of Education. Since 1992, the Ministry of Education (MED) began two pilots to bring the administration of education closer to citizens. One pilot was aimed at delegating school management responsibilities to municipal governments. The other one delegated operational autonomy to a school council composed of parent association members, school authorities, professors, and other local stakeholders. This last pilot in particular was very successful. Fiscal Dimension During this period, most municipal revenues came from various local taxes and fees. In 1992, 74 percent of municipal revenues originated from taxes and fees, with the bulk coming from the sales tax (37 percent), the property tax (6 percent), the vehicle tax (5 percent), the business registration tax (8 percent), and user charges and fees (16 percent). Sales taxes and business registration taxes, in particular, were predominantly collected in the trading centers in which businesses were located, usually cities and towns. A big revenue and expenditure gap existed between the larger and more urbanized municipalities and the smaller and rural ones. The fourteen more urbanized municipalities accounted for 75 percent of all municipal expenditures but included only 49 percent of the population. While an urban municipality spent on average more than US$20.00 per capita per year, the more rural ones invested less than US$0.50 per capita per year. 79. Ortega Egg 1997. 58 Fiscal transfers were discretionary, extremely low, and had no mechanism to compensate horizontal imbalances. The amount was discussed every year, and it was very small, for example, representing only 3 percent of total municipality revenues in 1992. Administrative Dimension With the exception of a few urban local governments, the majority of LG had very weak capacities. Only the larger or better funded municipal governments (for example, Leon, Chinandega, and El Castillo) were able to exercise their functions by establishing municipal environmental offices, employing technical specialists, and creating environmental and natural resource advisory committees. Interpreting the Opportunity Space The context of Nicaragua in the early 1990s offered some space for promoting local partnerships through CDD projects. Despite the important shortcomings of the decentralization framework, certain beneficial factors ­ notably the existence of a municipal law and of elected local governments with local budgets and a certain capacity for local revenue collection ­ provided a foundation for establishing projects promoting local partnerships. Pilots undertaken by the Ministries of Education and of Health, two areas difficult to decentralize, indicate the existence of some opportunity space. Even without a solid decentralization policy, decentralization was part of the public-sector reform process and discourse of the Chamorro administration. The Operational Response: FISE 1 and FISE 2 The two multisectoral CDD operations supported by the World Bank during this period ­ IDA credits for FISE, offered in 1993 and 1995 ­ did not seize the opportunity space provided by the decentralization environment. FISE, an agency created in 1990, had originally been part of INIFOM.80 The objective of these two loans was very similar: to "sustain poverty alleviation efforts, maintain social cohesion during economic adjustment until line ministries would strengthen their institutional capacities and complete policy reform programs."81 The principal aim of FISE in these early years was to generate employment and to build and rehabilitate social infrastructure (schools, health posts, and water projects) that had fallen into decay or had been destroyed during the war of the 1980s. Even though FISE supported the construction and rehabilitation of local infrastructure, local governments and communities did not participate as important actors. While projects were supposed to be identified in a demand- driven way, the whole process lacked transparency and was prone to clientelistic practices, and poor decision making. FISE engineers traveled around the country with the FISE project menu, establishing projects based on suggestions from local politicians, municipal officers, NGOs, and so on, and less often based on their own observations and determinations. Projects were managed by private contractors with minimal input from communities and local governments. Even though the constitutional reform of 1995 had a pro-decentralization emphasis, FISE 2 (designed in 1995) did little to take advantage of this and mostly continued with its centralized model. While it did have as one of its 80. The Fondo de Emergencia Social (FES, the original name of the FISE) was created on November 20, 1990, on the initiative of USAID. FES was an autonomous agency with full contracting and property-holding rights. It was created by Decree 59-90, originally for a period of five years. Institutionally, however, FES at first formed part of INIFOM. USAID began to finance FES projects in November 1990. According to USAID officials who were then already involved in decision making regarding the fund, the FES projects absorbed too much INIFOM capacity, and INIFOM neglected the strengthening of municipalities for which it was created. For that reason, it was decided in September 1991 to split the two institutions. At that time, the name FES was changed to FISE, for Fondo de Inversión Social de Emergencia. In 1991, IDB began lending to FISE (Dijkstra 2000). 81. World Bank.1997. 59 objectives supporting the government's decentralization strategy, in practice it continued with the same subproject cycle, which had hardly any involvement from beneficiary communities or local governments. FISE 2 began to acknowledge the problems of investment sustainability by requiring the establishment of local committees to maintain and help run schools and health centers. Subproject beneficiaries were also expected to make a contribution of five percent of the subproject cost in cash, land, materials, or labor, although this was not mandatory. Period 2: Piloting Innovations that Push the Boundaries of the Policy Environment (1996­2000) The Environment The 1995 Partial Reform of the Constitution brought important implications for LG, which were regulated through the 1998 Reformed Municipal Law (Law 261). Political Dimension Two important reforms that strengthened local government downward accountability were (i) the elimination of regions as administrative units,82 removing a level of government that often undermined local government territorial authority; and (ii) direct election of mayors, vice-mayors, and councilors for four-year terms. Still, despite these improvements, the urban bias of the electoral system continued as one of the major barriers for electoral accountability. Legal, Functional, and Regulatory Context83 Law 261 laid out the general responsibilities of municipalities and central government. It embraced the principle of subsidiarity, according to which expenditure responsibilities should be assigned to the lowest government level able to perform the task efficiently.84 It stated that municipalities had broad competence on all activities aimed at promoting local socioeconomic developments and preserving the environment,85 and it distinguished between exclusive, concurrent, and delegated functions. Exclusive functions were limited to basic services to the local population, for example, home garbage collection and civil registry. Concurrent functions were those for which municipalities and the central government were jointly responsible, for example, education, culture, and the environment. Delegated functions were those the central government could delegate to municipalities, for example, construction and maintenance of local aqueducts or roads. From an economic point of view, the Municipal Law has three main shortcomings: · First, the law does not specify which level of government should pay for concurrent responsibilities; in the absence of clear legal guidelines, the use and the availability of financing have delimited the relative responsibilities. In practice, the central government or its agencies have assumed full responsibility for concurrent activities, especially in poor municipalities. 82. With the exception of the RAAS and RAAN. 83. Largely based on Spilimbergo 2004. 84. Article 2, Item 3, of the Municipal Law states that "any matter that has implication for the municipal socio-economic development and any function that may be performed efficiently within its jurisdiction or require a relationship with the local community must be of municipal competence. Municipalities must develop the necessary technical, administrative, and financial capability to carry out the responsibilities assigned to them." 85. Article 7 gives a detailed list of the municipalities' responsibilities including (1) local health, which includes garbage collection, water disposal, sewage, and basic local health structures; (2) maintenance of a civil registry; (3) maintenance of cemeteries; (4) urban and rural planning; (5) building and maintenance of local roads; (6) provision of water and electricity to the local population; (7) development of the environment; (8) development and oversight of local transportation; (9) defense and promotion of the rights of women and children; and (10) development of local infrastructure and promote economic activity. The law, however, does not say anything about financing. 60 · Second, the Municipal Law has a surprising number of conflicts with other laws and regulations (particularly those covering electricity, water, sewage, health, and environmental regulation).86 · Third, the law fails to recognize the limited resources available at the municipal level. The law offers more a wish list of activities that municipalities should ideally perform than a realistic indication of priorities, and it raises an unrealistic local expectation of higher transfers to finance these mandates. At the same time, the list of mandates is so unrealistic that it does not provide a functional benchmark for evaluating the municipalities' efficiency. Expenditure Responsibilities Established by 1998 Reformed Municipal Law (Law 261) Central Government Municipal Governments Exclusive: Defense, national security, foreign Exclusive: Urban infrastructure development and relations, electoral system, justice system, maintenance; management of local public services; Expenditure school education, higher (university) education, construction, maintenance, and administration of Responsibilities central banking, regulation of financial system, libraries; promotion of handicrafts; construction and taxation and public borrowing, regulation of management of public cemeteries; construction and public services and public infrastructure, civil maintenance of sports, parks, and recreational facilities; security, and social security civil register; and local institutional capacity building Assigned to both: Roads, public health, electricity, environment and natural resources conservation, agricultural and forestry development, culture, tourism promotion and development, public transportation development and regulation, solid waste disposal, execution and monitoring of local public infrastructure, water and sewage management Source: Ballivián et al. 2004. Fiscal Dimension Probably the most important change brought about by the Municipal Law was the constitutional provision that a percentage of the national budget would be fiscally transferred to local governments with lower resource mobilization potential. Fiscal Responsibilities for Central and Local Government after the 1998 Municipal Law Central Government Local Governments Resources Exclusive: Income and profit tax, value-added Exclusive: Local sales tax, property tax, some minor local tax, excise taxes, custom duties, other small taxes taxes and fees and fees Shared: Minor taxes, taxes on sugar Transfers Discretionary transfers from central government to municipalities; the amount, discussed every year with the budget, usually amounted to 1 percent of fiscal revenues Source: Ballivian et al. 2004. Administrative Dimension GON and the World Bank agreed to use IDA financing to expand the range of municipal activities beyond natural resource management to a wider menu of poverty reduction activities, and implemented a US$ 35 municipal development component in 32 of 150 municipalities. This component aimed to 86. Contradictory provisions of the Municipal Law include the following: (a) the 1998 Municipal Law states that municipalities are responsible for providing electric services, which is at odds with the 1998 energy sector law stating that the Nicaraguan Electrical Institute (INE) is the only agency responsible for electricity; (b) the 1998 Municipal Law states that municipalities are responsible for aqueducts, which is at odds with the 1998 aqueducts law stating that the Nicaraguan Aqueduct and Sewage Institute (INAA) is the only agency responsible for aqueducts, sewage, and water treatment; and (c) the Municipal Law presents conflicts with ministry regulations regarding the environment, health, and development. 61 develop municipal capacities for service delivery and provision, by providing a range of training and small grants to finance activities which conformed to a municipal investment plan developed using a participatory model of governance. Interpreting the Opportunity Space While the legal reforms still left important gaps in the institutional framework, particularly in terms of the concrete regulations to implement the new norms, their strong message in favor of municipal strengthening increased the political salience of decentralization. The changed environment also encouraged several donors to begin a number of small pilots with INIFOM. The Operational Response The World Bank supported two operations during this period, first INIFOM 1, starting in 1996, followed by FISE 3, starting in 1999. These programs took advantage of the opportunity space and designed innovative pilots aimed at improving the policy environment and the practice of local partnerships. Both operations tried to improve the fiscal and capacity dimensions of the enabling environment and also promoted local partnerships to deepen coproduction and accountability between LG and CBOs. The approach taken by INIFOM 1, officially known as the Rural Municipalities Development Project (or PROTIERRA), contrasts both with traditional approaches, in which rural development programs are implemented through central line agencies, and with some community-drive approaches, which focus exclusively on community groups. INIFORM 1 was concerned with establishing a permanent capacity for designing and implementing locally relevant development policies at the municipal level. Strengthening municipal governments was, therefore, considered a critical aspect of the strategy for implementation. It was thought important to develop the capacity of municipal authorities not only for implementing poverty alleviation programs but also for designing programs and policies and developing a coherent long-term strategy for area development. It was important to decentralize authority for planning and implementing development programs, and the municipality was selected to be the key institution spearheading this effort. Interestingly, the concept for PROTIERRA evolved significantly between its identification in mid-1994 and its approval in September 1996.87 The focus on local development came as a result of the diagnostic work that went into design. Its initial rationale was to assist the Government of Nicaragua (GON) in implementing the actions recommended in a 1993 National Environmental Action Plan (completed with World Bank assistance) and a National Forestry Action Plan. These plans were supported with detailed technical analysis, along with consultations at the community, municipal, and national levels. Analyses completed at the time suggested that sound natural resources management could help create employment and that natural resources management through a demand-driven and participatory development approach was the most sustainable means of doing so. The government had also accepted the idea that rational management of natural resources at the local level could contribute to reducing poverty, and it was prepared to make grants available through municipalities for such management. This view reflected an ongoing trend in national policy in favor of decentralizing responsibility for a broad range of economic and social development functions and transferring those responsibilities ­ and the resources to meet them ­ to municipalities. The GON's strategy was to transfer a percentage of its current budget resources to municipalities through untied block grants according to a specific formula and to improve the municipalities' capacity to manage these resources. The GON believed that this would create a more sustained effort in economic development and 87. World Bank 2001a. 62 poverty reduction. The process of developing and approving a national legislative framework for such a system was slow and problematic, however, because of the government's lack of experience; this was compounded by distrust between political parties and their concerns about the outcome of such a plan. The legislative action required to achieve this goal was included (and accomplished) as part of the Second Rural Municipal Development Project. In response to these concerns, the GON and the World Bank agreed that IDA financing under this project would support a broader range of municipal activities (beyond the original scope of natural resources management) as part of a more general poverty reduction effort. Thus, the rationale for the project was significantly altered in favor of supporting poverty reduction through municipal development and the decentralization of financing for municipal activities. The municipal development component was implemented in 32 of the 150 municipalities. Its budget of US$ 35 million was divided into four subcomponents: · Development of municipalities as decentralized service planners and providers and promotion of a more open and participatory model of governance (US$5.8 million); · Support for improving the institutional capacity of INIFOM and its various organizational units, strengthening the legal and financial framework for decentralization of government functions to communities, and establishing a specialized unit to implement the component (US$3.3 million); · Specific support for informing, training, and encouraging communities and civil society to participate in local service provision and to recognize and exercise their right to local control, transparency, and accountability (US$2.8 million); and · Provision of small grants for eligible subprojects proposed by individuals, communities, or municipalities that conformed to a municipal investment plan and could be financed within a municipal financial ceiling (US$23.4 million). Compensating for Shortcomings in the Enabling Environment To empower local governments to engage in partnerships with communities, INIFOM 1 addressed some of the weaknesses in the enabling environment, particularly in the fiscal and the capacity dimensions. Fiscal Dimension The thirty-two targeted municipal governments received a yearly allocation of program funds, based on needs indicators such as population and poverty, and they had wide freedom in deciding how to employ these funds.88 Most of the assistance was for direct capital investments (community subprojects), but a small part was also available for financing recurring costs, including salary and similar expenses. Community subprojects were usually small ($50,000 was the maximum project size allowed, and the average project size was $20,000), and they were targeted to benefit rural poor communities. Projects were identified on the basis of a local planning process involving all community stakeholders (see below). Administrative Dimension In order to enhance the municipalities' technical capacity for undertaking this effort, program funds were made available to them for creating municipal technical units (MTUs). Each MTU reported to the mayor and consisted of three people: a local project coordinator, a participation specialist, and an accountant responsible for maintaining project accounts. MTUs coordinated with INIFOM and other central 88 The project has been extended to cover additional municipalities, and the formula for resource sharing has been revised and made more complex. We need not go into these complexities here. It is important to note, however, that resource transfer is formula based in this case. It is not ad-hoc, nor does it depend on the number of subprojects submitted. 63 agencies, prepared annual municipal plans, appraised community subprojects, supervised implementation, and helped prepare long-term plans for the local area. Accountability Partnerships As mentioned before, one of the shortcomings of municipal governments was the urban bias in the selection of local councilors. In addition, most local governments suffered from low levels of accountability due to the ineffectiveness of central government audits as well as the lack of civic oversight mechanisms. To address these problems, INIFOM 1 created a permanent infrastructure for community representation at the municipal level, the Intercommunity Assembly, and promoted participatory municipal planning processes. 89 The Intercommunity Assembly (or IA) is a permanent assembly constituted at the municipal level by representatives of all community associations located within a municipal area. Like a legislative assembly for the local area, they establish long-term priorities, approve overall plans, and provide oversight to municipal government staffs. IAs have three principal responsibilities. First, they arbitrate between different community associations in cases of conflicting needs. By accommodating and prioritizing the needs of different community groups, they help resolve conflicts among communities over project resources. Second, IAs are responsible for developing an overall strategy for rural development, and they do so with the active participation of community associations. Third, IAs provide oversight and supervision, ensuring that community associations as well the municipal government implement rural development plans faithfully and well. IAs thus function as representative and watchdog bodies, watching over the activities both of community associations and of municipal staff. A Community Supervision Committee (CSC), a small subcommittee elected by IA members from among themselves, assists each IA in discharging these functions. Members of the CSC are expected to play a key communication role between communities and the municipal government. They supervise project and subproject execution, and they keep the IA informed of project execution at the community level. Simultaneously, they keep communities informed about municipal plans and subprojects, and they ensure that municipal council decisions are fully disseminated among community members. The planning cycle involves community members carrying out participatory planning exercises to agree on needs and investment priorities and to elect community representatives to the IA. The MTU integrates the priorities of all communities within each municipality into a municipal investment plan (MIP), which is then submitted to the IA for discussion and prioritization of investments among communities. By law, some of the municipal council meetings, and in particular those involving key budgetary decisions, are open to all the inhabitants of the municipality. The final decision regarding approval of the MIP and its incorporation into the municipal budget, however, rests with the municipal council. Projects submitted by community associations are appraised by staff of the MTU, and the municipal council gives final approval. The IA and community committees provide oversight through all steps of this process, guiding community associations, setting longer-term development priorities, arbitrating conflicts between community groups, and ensuring that municipal government staffs respect the priorities established by the assembly. Municipal councils have the authority to accept or reject communities' subproject proposals, but they must abide by both the local priorities set by their counterpart IA and also by the overall guidelines established for the country by INIFOM. Central, local, and community concerns must all be brought together and balanced in this decision-making process. Coproduction Partnerships 89. This section is based on Krishna 2004. 64 All subprojects (with the exception of capacity-building subprojects) require matching funds of two types: (i) community funds in kind or labor, and (ii) municipal government funds. The table below shows the extent of cofinancing required in different cases. Maximum Matching Grant (%) Subproject Category Type A and B Community Type C Community Environment 75 90 Municipal Infrastructure 60 75 Community Infrastructure 75 75 Community Productive 75 75 Technical Assistance 100 100 Note: Type A: Revenue per capita greater than US$15 per year; Type B: Revenue per capita between $3 and $15 per year; Type C: Revenue per capita between $0 and $3 per year. FISE 3--Third Social Investment Fund Project Seeing that INIFOM had moved forward in terms of local partnerships and noting the shortcomings (particularly in terms of sustainability) of its initial centralist strategy, planners for FISE 3 designed the program along lines similar to those taken by INIFOM. The Third Social Investment Fund Credit, approved in 1998, had the following objectives: (i) providing essential small-scale infrastructure, mainly in education, health, and water supply and sanitation (in close coordination with the responsible agencies), (ii) focusing on the poorest communities, and (iii) "strengthening, on a pilot basis, municipal management which, in turn, should lead to more sustainable subprojects at the local level."90 More broadly, it was stated that the credit would support FISE's "new focus on ... municipal/community involvement."91 Addressing and Compensating for Shortcomings in the Enabling Environment FISE's program of strengthening municipal governments also involved components that compensated for LG fiscal and capacity weaknesses. FISE began a pilot with nine municipalities to which it delegated the management of investments up to $100,000. In a manner similar to INIFOM, it also established municipal technical units with responsibility for carrying out the full subproject cycle, other than the ex ante evaluation. FISE Managua still had to approve projects before the municipality could contract it to a firm. FISE also checked whether sector approval had been obtained. In terms of financing, the approaches take by INIFOM and FISE complemented one another nicely. FISE's transfers of resources were tied to certain sectors considered priorities by the central government. While INIFOM simulated an unconditional fiscal transfer, FISE followed the principal-agent model in which local governments act as agents of the central government for the delivery of certain goods and services. Accountability Partnerships FISE introduced two main innovations in terms of accountability partnerships: · Its Participatory Municipal Planning pilot, implemented in sixty municipalities, used a methodology corresponding to that of INIFOM. · In the nine municipalities in which FISE had delegated the project cycle, FISE encouraged beneficiary committees to participate in the LG bidding process and to control the way in which the contracting process was performed. Coproduction Partnerships 90. World Bank 1998, p.1. 91. World Bank 1998, p. 2. 65 As mentioned, one of the problems with the centralist operations of FISE 1 and FISE 2 was the low sustainability of local investments. To address this challenge, FISE created the Preventive Maintenance Fund (the acronym, FMP, derives from the Spanish), which is basically a partnership among the central government, municipal governments, and community organizations geared toward ensuring investment sustainability.92 Through the FMP, the national government provides funding to municipalities for preventive maintenance of primary schools and health centers. This cofinancing is progressive: extremely poor municipalities receive a higher percentage than do others. The FMP functions as a conditional intergovernmental transfer according to transparent rules that stipulate the following: · Local Education or Health Maintenance Committees comprised of community representatives and local government officials must devise a maintenance plan and budget for each facility; · Each municipality must present an Annual Maintenance Plan aggregating facility plans and budgets, including community and municipal government contributions; and · Local counterpart resources must be provided by municipal governments and community committees and deposited in specified bank accounts. Once these conditions are met, FISE disburses budgeted funds in tranches to municipal governments, which in turn distribute them among the selected facilities. Community maintenance committees supported by local government officials manage the work. These committees have a strong incentive to perform well since they can receive funds each year only if they have completed the maintenance agreed to the year before. Since 1997, the FMP has channeled $3.5 million to communities through municipal governments for preventive maintenance. As a result of the FMP's cofinancing requirements, for every dollar provided by the national government, communities and municipalities have contributed 27 cents to facility maintenance, about half contributed by each. By 2002, the FMP had funded preventive maintenance on almost three thousand infrastructure and service facilities, representing 97 percent of the facilities built, repaired, or expanded with FISE funds. In addition, while initially limited to financing maintenance of FISE investments, since 2001 the Government of Nicaragua has expanded the FMP to cover all primary schools and health centers in the country. Period 3: Using Projects for Policy Reform (2001­2004) The Environment and the Opportunity Space The main difference between this period and the previous one was the availability of the results of the pilots introduced by INIFOM 1 and FISE 3. While these pilots had demonstrated that local partnerships could provide adequate means for promoting sustainable local development, their limitations and challenges were also quite evident. These challenges had to do with institutionalizing the innovations introduced by the pilots and deepening the decentralization process by engaging communities more strongly in coproduction. INIFOM 2 and FISE 4 set out to address these challenges. The Operational Response: INIFOM 2 and FISE 4 INIFOM 2 and FISE 4 were highly coordinated, making it appropriate and useful to discuss them jointly. Addressing Shortcomings of the Enabling Environment 92. Serrano and Warren 2003. 66 In terms of the fiscal dimension, INIFOM 2 tried to improve the fiscal transfer system by demonstrating the virtues of a new one. To do this, it created the Municipal Development Fund (FONDEM). As mentioned before, the constitution (art. 177) emphasized the rationale of transfers as instruments of fiscal equalization. Nicaragua's transfers were allocated mainly in proportion to population, a criterion which does not contribute to closing horizontal imbalances and thus did not advance the principle of fiscal equalization. Designed as a "general-purpose grant" mechanism, FONDEM aimed to complement existing municipal own-source revenue and various ongoing programs with negotiated, purpose-specific transfers. In addition to the establishing FONDEM, the project aimed to strengthen municipal capacity to generate local, own-source revenue through better assessment and collection of local taxes and fees. To this effect the project assisted INIFOM in revising the legal framework of local government finances, drafting appropriate policy, legal, and regulatory texts, and developing and extending to all municipalities participating in the project (i) improved systems and manuals of local fiscal administration, (ii) standards and procedures for financially sustainable delivery of municipal services, and (iii) a system for land and property registration, or Cadastre (Sistema de Cadastro, SISCAT). The design and implementation of FONDEM strongly influenced the national debate on fiscal transfers and the development of a related legal framework. This ultimately changed the enabling environment: in July 2003, the GON approved the Municipal Transfers Law, which substantially increases the funds available to municipalities from about 1 percent of tax revenues to at least 10 percent by 2010. The goal is to fulfill the constitution's intent by transferring amounts from the central government to the municipalities sufficient to cover the gap between constitutional mandates and local resources. The new law does not specify expenditure responsibilities, however, and previous legislation assigning expenditure responsibilities among different levels of government was ambiguous. This has created an important problem for fiscal neutrality, which the Bank has engaged to help address.93 93. See Ballivián et al. 2004; and Frank 2004. 67 Box 4: Municipal Transfers Law The law covers the total amount of transfers, the distribution criteria, limits to the use of the funds, procedures of disbursement, and auditing. Amount of transfers. The law establishes that the central government must transfer at least 4 percent of total fiscal revenues to the regional municipalities, starting in 2004, and that this percentage must increase by at least half a percent annually ­ provided the GDP grows by at least one percent ­ until at least 10 percent of fiscal revenues are transferred. Distribution criterion. According to the formula, Managua will receive a fixed share equivalent to 2.5 percent of transfers while the rest will be split among the other 151 municipalities. The total amount for the 151 municipalities will be split into four equal components. Each component reflects a different criterion for the allocation of funds: fiscal equity, fiscal efficiency, population, and execution. The fiscal equity criterion gives proportionally more funds to municipalities with less revenue potential. The fiscal efficiency criterion gives more funds to municipalities that are efficient in collecting the property tax revenue (IBI). The population criterion gives an equal amount per capita. The execution criterion gives proportionally more to municipalities that executed more projects with the transfers. Limits to the use of the funds. The law determines that a share of the transfer should be spent on investment. The share is 90 percent for Managua and is progressively lower for other municipalities with less income. On average, the share for investment will be at least 70 percent. Procedures of disbursement. In 2004, municipalities will receive transfers in three equal installments; starting in 2005, they will receive twelve monthly installments. Among the conditions for receiving the disbursement is transmission of the year's municipal budget and investment plan. Auditing. The law mandates that municipalities audit all transfers and send a copy of the report to the Auditor General's Office (Contraloría General de la Republica). Source: Spilimbergo 2004. Lessons from FONDEM Both the World Bank and INIFOM have found it difficult to overcome the "project" logic established by INIFOM 2, and the implementation of the FONDEM experiment has clearly suffered from this. · On the World Bank side, the integrity of the FONDEM concept would have been better served if assistance had been channeled through some form of "budget support" rather than through the "investment loan" instrument. · On the INIFOM side, abandonment of the role of "Fund Manager" was strongly resisted.94 94. For greater elaboration of these arguments, see Romeo 2003. 68 Based on the success of its pilot,95 FISE decided to scale up the delegation of the project cycle. It created the Municipal Accreditation System, which defined the conditions that municipalities needed to fulfill to access increasing levels of responsibility. It also developed, jointly with INIFOM, a capacity-building module. One of the problems of the previous period had been that INIFOM and FISE had developed separate systems for managing investment projects, adding further management levels to those required by other donor or government programs and causing confusion and waste in capacity-building efforts. To solve this problem, INIFOM and FISE divided the capacity-building component into different modules, and each agency developed, in consultation with municipalities, generic, non-project-specific modules applicable to all municipal clients. Following their comparative strengths, FISE developed the module for investment projects and INIFOM developed those for financial management. Accountability Partnerships The planning and accountability structures and processes set up by INIFOM 1 and FISE 3 were project specific, with the resulting risk that once a project was finished, accountability structures would dissolve. One of the most significant impacts of the INIFOM and FISE operations was that the pilots each established for participatory municipal planning became institutionalized as the Municipal Planning System (MPS) and were adopted as the local planning statute rather than being seen as merely a project- specific methodology.96 This MPS built on the momentum and lessons provided by the previous pilots. The Territorial Committee (Comité Territorial), for instance, plays a similar role to that of the Intercommunity Assembly. Coproduction Partnerships While FISE and INIFOM showed that local governments could manage local funds and investments successfully, one of the remaining challenges was teaching local governments how to empower community associations to implement their own subprojects and manage subproject funds. To address this, FISE began a pilot called Proyectos Guiados por la Comunidad.97 In this program, FISE transfers resources to local governments, which in turn transfer them to CBOs managing projects identified in the Local Municipal Plan and deemed particularly suitable for community implementation. Projects are cofinanced by local governments and community associations, and both are responsible for the projects' sustainability. Conclusions and Challenges Ahead A look at the trajectories of FISE and INIFOM over the last fifteen years shows that both agencies have evolved and adapted to the decentralization environment, pushing efforts to address its shortcomings and promoting innovative local partnerships. As a result, the 2004 enabling environment for partnerships was much superior to that prevailing in 1990 (see the table below). At the same time, the new environment to 95. An evaluation of FISE's Pilot (Grun 2002) concluded that the decentralized approach to investment management performed better than did 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 as well as in comparable municipalities during the same period. · Contracting and procurement was 40 percent faster and 7 percent cheaper. The latter was a result of greater competition among contractors in the decentralized model. · While real costs in the decentralized model were 4 percent lower than estimated costs, in the centralized model real costs for similar projects were 6 percent higher than estimated ones. Construction times for similar projects were 40 percent faster in the decentralized as opposed to the centralized model. In general, the decentralized model compared favorably in terms of number of fines, renegotiation of costs, and timeframes for project completions. 96. AMUNIC-INIFOM 2004. 97. See Emergency Social Investment Fund Operational Manual: http://www.fise.gob.ni/pgc. 69 which they jointly contributed, particularly with the deepening of the fiscal decentralization process, has challenged both agencies to rethink their roles and strategies. 1990 2004 Constrained LGs, Constrained Enabled LGs, CBOs Enabled CBOs Functional Ambiguous allocation of functions Ambiguity continues Dimension CBOs began assuming greater role in service Legal delivery (schools, water) Political Only local councilors elected Reforms Elected mayors Dimension and INIFOM Participatory Municipal Planning and Fiscal FISE First Fiscal Transfers Law: Dimension Discretionary transfers from Projects the central government central government to must transfer at least 4 municipalities; the amount was percent of fiscal revenues, discussed every year with the progressively increased by budget and usually amounted to 1 at least half percentage percent of fiscal revenues point till reaching 10 percent of fiscal revenues in 2010 Preventive Maintenance Fund Greater CBO project cofinancing Administrative LGs have significantly Dimension LGs have very limited capacities improved financial, to manage their affairs administrative, and project cycle management systems 70 APPENDIX 4. TANZANIA: TESTING PARTNERSHIPS WITHIN A DECENTRALIZATION PATCHWORK98 Tanzania's decentralization legislation outlines an ambitious approach, assigning local governments a role in a wide range of sectors, but it often falls short on clarity (for example, in delineating actual responsibilities for the local as opposed to the central government), in addition to retaining substantial influence for the center over local governments' institutional environment and powers. In practice, this framework has resulted in patchwork of overlapping or splintered responsibilities and service delivery. Two recently inaugurated World Bank projects have been designed to address these conditions: one, the Local Government Support Project, works primarily on institutional issues and mechanisms, and the other, the Second Tanzania Social Action Fund, takes the practical approach of implementing community-driven development subprojects with the involvement of local governments. This case study focuses predominantly on the latter project, as it most directly addresses LG-CBO relations. Local Government Context This summary focuses on the situation in mainland Tanzania, excluding Zanzibar, which has a different local government structure. Political Dimension Tanzania reintroduced local governments in 1982, after having abolished them in the early 1970s. The constitution includes a broad reference to local government at the regional, district, urban, and village levels, leaving further definition to parliament. In both rural and urban areas, local government councils have the power to legislate. Councilors are directly elected, with elections held every five years, except for village or neighborhood LGs (in urban areas), for which the Ministry of Local Government decides procedures. Functional Assignments The two Local Government Acts (for urban and rural LGs) make LGs responsible for service delivery in a broad set of areas, including primary education, health, and local roads. Responsibilities are poorly defined, however, and sectoral ministries continue to be involved at all levels, in addition to exerting heavy influence over budgets and personnel.99 While the legislation is somewhat more specific regarding responsibilities, for lower-level local governments (in villages or urban neighborhoods), it provides only general guidance about required responsibilities.100 98. Sources for this case study include Project Appraisal Documents for TASAF II and LGSP; Steffensen et al. 2004; The World Bank 2001c; N. Mungai Lenneiye 2005; and interviews with Ida Manjolo, Social Projection Specialist, World Bank, and Helen Kijo-Bisimba, Director, Legal and Human Rights Centre (Dar es Salaam). 99. Act no. 8 (Urban Authority) and no. 7 (District Authority). Responsibilities include "to maintain and facilitate the maintenance of peace, order and good government; to promote social welfare and economic well-being; to further rural and urban social economic development; control and improvement of agriculture, trade, commerce and industry; enhancement of health, education and the social, cultural and recreational life of their inhabitants; development, mobilization and application of productive forces in the war on poverty, disease and ignorance" (Steffensen et al. 2004). 100. "Do all such acts and things that are necessary or expedient for the economic and social development of the village; initiate and undertake any task, venture or enterprise designed to ensure the welfare and well-being of the residents of the village; plan and coordinate activities of and render assistance and advice to the residents of the village engaged in agriculture, horticulture, forestry or other activity or industry of any kind; encourage the residents of the village in undertaking and participating in communal enterprises; and participate by way of partnership or any other way in economic enterprises with other villages." Ibid. 71 Formal policy established in 1998 provides more specifics than does the legislation, stipulating that LGs should provide basic public services, especially those related to poverty reduction, including primary education, primary health, agriculture extension services, local water supply, and roads.101 Actual responsibility for service delivery illustrates this mixed central-local approach. Primary education is the domain principally LGs -- but teachers are managed by a commission independent of the LG. LGs are principally responsible for primary and preventive health, but decentralized central government facilities also operate, with a lack of clear definition between decentralized and LG roles. Sectoral legislation makes LGs formally responsible for hospitals up to the district level, but there is a parallel system of Health Boards.102 In rural areas (but not urban ones), LGs are also responsible for water supply, sewage, and sanitation with support from regional bodies; however, the central government implements almost all capital investments in both urban and rural areas.103 LGs are responsible for district and feeder roads and for municipal and urban roads, although financing arrangements and technical capacities are sometimes inadequate. LGs also handle agricultural extension, for which some seven thousand extension workers were transferred to their authority. Fiscal Dimension Formally (under the local government acts), LGs are directed to "establish and maintain reliable sources of revenue and other resources to enable them to perform their functions effectively and enhance financial accountability." LGs also have the authority to collect taxes, levies, fees, and charges, although by-laws to this effect must be approved by the President's Office for Regional Administration and Local Government. LGs rely on central transfers for some 85 to 90 percent of their budget. Most of these transfers are conditional, sector-specific grants, mostly for primary education and primary health care (and predominantly for personnel costs), while funding for other responsibilities is limited.104 Timing and predictability have improved in recent years to relatively acceptable levels, though some delays have been reported in the wage bill moneys that constitute the bulk of the transfer. LG own revenues, subject to central decisions mandating how they are spent, appear to go largely toward administrative expenses.105 The central government also has significant power to constrain the LGs own- revenue generation, and recently it exercised it by abolishing certain important taxes (estimated to account for about 90 percent of village and neighborhood LGs own-revenues) that were deemed distortionary. Although the central government has committed to providing compensation in the form of unconditional formula-based block grants, with a proportion allocated to village and neighborhood LGs, this action has put LGs ­ particularly rural ones ­ under added fiscal pressure. The planning process is designed to be bottom-up, starting at village level. It reportedly suffers from a lack of realism, due in part to shortcomings in information about revenues and budget ceilings. Administrative and Capacity Dimension Most LG staff are field-based (about two-thirds are teachers; others include agricultural extension workers). The public service act does not give LGs direct legal control over their staff. The Minister of 101. The Policy Paper on Local Government Reform of October 1998 (hereinafter policy on LG reform). 102. National Health Service Bill of 2004. 103. In rural areas, water user associations may also have responsibility for managing the water supply. 104. Primary education is reported to account for about 70 percent of recurrent grant transfers and primary health care for 17 percent (Steffensen et al. 2004). 105. For example, LGs must contribute to a Local Government Loans Board. 72 Local Government appoints, promotes, and disciplines LG directors, who in turn appoint and discipline staff, although central government policies and independent bodies play some role.106 This can cause some tension between technical staff and elected councilors.107 In contrast, the 1998 Policy on LG reform calls for "non-subordinated local administration...and normal employer-employee accountability structure between the local government administration and the local elected leadership."108 Pay of LG staff is set nationally, with the base pay of LG staff generally similar to that of central government workers; lower benefits, however, inhibit the ability of LGs to recruit specialist technical staff, although senior managerial positions appear, in general, to be filled. Some positions (including executive officers at the ward and village levels) appear subject to high turnover. In theory, LGs can introduce financial incentives for staff, but, in practice, constrained budget preclude them from doing so. Since the central government encourages LGs to engage in public-private partnerships, reform-minded councils outsource some activities to the private sector or other nongovernmental actors. Although this has reportedly provided a positive boost to LG operations, weak administrative capacities can also mean that LGs have trouble drawing up clear contracts, monitoring work, and managing contractors. Whereas district and municipal LGs have a role in staff management and selection, staff at village LGs and wards are seconded from district and municipal LGs; specifically, the district appoints a village executive officer who acts as secretary to the village council. CBO Context Political Dimension Tanzania has an NGO law in place, but it has no explicit national-level legislation addressing the status of CBOs. CBOs thus work informally. If necessary, however, they seek more official standing by registering as NGOs; by getting a permit from the District Commissioner (under district by-laws), in the case of service-delivery CBOs; by forming corporations or cooperatives; or by entering into other types of loose affiliations. Sectoral policies also include official avenues through which community members can be involved in LG activities; the most common method of involvement is through user committees. These include school committees and district-level education boards, mandatory LG-stakeholder health committees that guide service provision and priorities for each health facility, water committees, and, in rural areas, road maintenance committees. CBOs can open bank accounts using their articles of association or the minutes from an initial meeting as documentation. The relatively high initial deposit required can be prohibitive, however. CBOs are supposed to participate in the planning process via the Ward Development Committee, but in practice they are rarely called on to do so. In general, LGs seem to view service-delivery CBOs favorably while taking a less positive attitude toward advocacy CBOs that encourage citizens to assert their rights, which are seen as creating trouble. Of necessity, CBOs seek cooperation from LGs since they can't operate in government facilities such as schools and health centers without it. More generally, even a paralegals' CBO felt it necessary to get council buy-in, to facilitate its work. 106. For example, teachers' discipline and personnel records are the responsibility of Regional and District Teachers Committees, not of the LGs. Staff numbers and individual postings are decided on by the central government, which means LGs have little authority or incentive to develop efficient administrative structures. 107. Issues include elected officials in positions of authority but with lower educational attainment than the technical staff, who believe councilors interfere in their work, and perceptions that individual LGs offer little career advancement. 108. Steffensen et al. 2004. 73 Functional Assignments In the absence of a formal framework, CBOs appear to be active in a broad range of sectors and types of activities, such as planting trees; encouraging children to attend school; providing meals to school children; helping to build classrooms; and providing information about HIV/AIDS. Their involvement in advocating for citizen's rights extends to advising citizens on where to turn for help with their problems and informing citizens of governance processes through which they can hold local government to account.109 Fiscal Dimension While some CBOs have managed to apply for outside funding, they more commonly run on small amounts of sporadically collected money. CBOs that regularly raise money within their communities tend to be faith-based (usually church groups), although other types of CBOs might agree on an annual or monthly subscription from members. Members of other CBOs might undertake paid work (for example, on a farm) to generate income or raise funds relating to a specific project, such as establishing a school. In one CBO concerned with HIV/AIDS, the associated villages came together to discuss how to support widows and orphans, with each village proposing its own fund-raising activity. Some CBOs with loose affiliations to programs, such as a Village Legal Workers CBO working under the District Land Management Program, might occasionally receive a small sum toward expenses. Budgeting and financial management are basic, with possibly a literate villager keeping a simple set of books. The CBO may not keep strictly to budgets, spending money on unplanned but socially necessary expenses such as hospitality for a guest. Administrative and Capacity Dimension CBOs usually have a simple constitution that specifies executive positions; depending on needs, these usually include a chairperson, a secretary, and, in the case of CBOs that collect funds, a treasurer. Positions are filled by vote at all-member meetings. The number of active staff tends to be ten or fewer, and due to lack of funds, they are generally volunteers (for example, a literate villager might keep the books), although a small honorarium might be given when money is available. The broader membership convenes occasionally to discuss the bigger issues. Depending on the CBO's purpose, membership might be broad (the entire community) or focused (for example, a Village Legal Workers CBO might be composed of some thirty or so paralegals from the district). Some technical assistance is available from funders (for example, Oxfam, Save the Children, and CARE) or from national NGOs, such as the Legal and Human Rights Centre, which helps train the Village Legal Workers CBO (paralegals) to advise villagers and encourages CBOs to join together for greater leverage. In some instances, the government might provide assistance, for example, by helping CBOs understand HIV/AIDS issues so that they can, in turn, educate community members. Mapping the Opportunity Space A summary of the opportunity space for LG-CBO accountability and coproduction arrangements appears in the following table: 109. For example, in the case of a village chair who ran the LG without convening the council, as required, CBO activities resulted in the village assembly holding a vote of no-confidence and removing him from office. Where village councils have distributed land without calling a village assembly to decide who gets it, paralegals' CBOs have informed villagers of the required process so they could demand the councils adhere to it; some have threatened lawsuits to win back illicitly awarded land. 74 Tanzania Opportunity Space Dimension LGs CBOs Political Partly enabled Partly enabled Functional Partly constrained Enabled Fiscal Constrained Constrained Administrative Constrained Partly enabled For the above-described opportunity space, one would expect coproduction and accountability relationships to be rather fragile. While both LGs and CBOs could, in theory, reflect citizen priorities in a number of spheres with some degree of legitimacy, the practical application in terms of actual development activities is constrained by weak organizational capacities and very limited resources. CDD and LG-CBO Partnerships: The Second Tanzania Social Action Fund (TASAF II) Project Overview Building on the experiences of the first TASAF, TASAF II became effective in May 2005 and is expected to run until 2009. Since the project is in its very early days, this examination highlights LG-CBO roles and partnerships in the design of TASAF II and lessons from the first TASAF. TASAF II focuses on enabling communities to "request, implement and monitor sub-projects that contribute to improved livelihoods linked to Millennium Development Goal (MDG) indicator targets in the Tanzania Poverty Reduction Strategy."110 The range of subprojects thus spans several sectors (among them health, education, infrastructure, and agriculture) and may include items such as construction of bridges and community roads, school construction and equipment, food production projects, vocational training for vulnerable populations, and improved maternity care. In general, TASAF II intends to provide a greater role for both local governments and CBOs than did the first TASAF. Under TASAF I, communities were generally able to identify, prioritize, and implement subprojects, provided that guidelines (on access to funds) and support were available. Without management above the community level, however, opportunities for successful mainstreaming of CDD approaches appeared limited. TASAF II thus transferred some project management responsibilities from the project management unit to village and district local governments, to better align with the existing decentralization framework and give these bodies a practical opportunity to develop better working relationships and apply their skills. The specification of a clear role for village councils is also expected to alleviate the conflicts between subproject management committees and district LGs that surfaced under TASAF I. TASAF I's troubles with a separate public works program component for which district LGs handled both implementation and funds resulted in public works being folded into the National Village Fund component, with implementation to be handled by the community management committee.111 Under TASAF II, the bulk of project funds are channeled to a National Village Fund held by elected village councils in rural and urban areas -- the lowest level of local government112 -- with higher level local governments (district) involved in (and receiving some project funds for) facilitation, appraisal, and monitoring. A second component addresses capacity enhancement both for communities and for other 110. TASAF II PAD. 111. Problems included inadequate procurement and accountability, delays in paying wages, and the tendency of local authorities to focus on meeting their own priority public works goals rather than addressing beneficiaries' needs; see N. Mungai Lenneiye 2005. 112. These councils are called Kijiji in rural areas and Mtaa in urban ones, where they are roughly equivalent to a neighborhood. The local government law provides only general definition of their responsibilities. 75 involved parties, local governments as well as private bodies, which facilitate progress on the poverty reduction strategy targets. Importantly, TASAF II implementation coincides with a complementary project, the Local Government Support Project (LGSP), which supports the government's Local Government Reform Program by helping "strengthen fiscal decentralization, improve accountability in the use of local government resources, and improve management of intergovernmental transfers."113 TASAF II thus focuses predominantly on the space between local governments and communities, while LGSP addresses relevant institutional and fiscal decentralization issues in the space from central government through to local government levels. This complementarity explains why TASAF II, while addressing local government performance and involvement in community development, places the bulk of its emphasis on communities and the most "local" of local governments, the village council. Variation across LGs -- with some exhibiting acceptable participatory planning and budgeting processes, staff capacities, and reasonable management and others falling short -- also underlies the design of project funds flows. In all cases, 90 percent of project funds go to a subproject account managed by an elected community management committee, using one of three possible channels: direct transfer to the subproject account; transfer via a village council account; and transfer via the district LG. The channel chosen depends on the assessed capacity of the village and district LG. Of the remaining 10 percent earmarked for administrative costs, one quarter goes to the district LG to cover facilitation and reporting costs and one quarter to the village council for the cost of facilitation, monitoring, and reporting of the community management committee's activities. The remainder goes to either the district or the village LG to cover the cost of supervision, depending on which of the two is considered the most capable of fulfilling that role.114 Coproduction and Accountability Relationships Some examples of LG-CBO coproduction and accountability in project design, as described in the PAD, are cited below.115 Particularly for accountability arrangements, the project emphasizes the role of project- specific committees composed of village members and elected at a village assembly, which may reflect both the scarcity of CBOs in rural areas and the limited administrative support available to village councils. Coproduction · Communities make a minimum cofunding contribution in cash or in kind; local governments are not required to contribute funds, but the project considers the cost of their staff time to be an indirect cost contribution.116 · CBOs (as well as NGOs and private contractors) are eligible to be "local service providers" and are named explicitly for subprojects dedicated to improving the income opportunities of vulnerable persons. They enter into a contract with the LG 113. LGSP news release, 30 November 2004. 114. This includes both direct supervision by LG staff and contracting of outside technical skills, including CBOs or NGOs. 115. Since the text largely uses general terms such as "communities" or "local service providers" rather than referring specifically to CBOs, these examples may understate CBO involvement. 116. In the case of the "safety-nets for able-bodied poor" subproject type, which pay submarket wages, the difference between wage levels and market rates is also considered a contribution. While the PAD does not explicitly mention a CBO role in generating these contributions, it is plausible that CBOs might be involved, particularly given that TASAF II spans a number of sectors and issues -- health, HIV/AIDS, education, and water supply, among them -- in which CBOs operate. The project considers the cost of the staff time of LG technical staff who work on appraisal, approval, and supervision tasks in the subproject planning cycle and on the Technical Planning Committees as indirect cost contributions. 76 for this purpose, with responsibilities that include subproject implementation and technical supervision of implementation. · Relevant CBOs are trained to work on reducing maternal mortality, one of the Millennium Development Goals targeted under TASAF II. · Coproduction by elected committees includes the Community Management Committee, which, as its name suggests, is responsible for acting as the administrative arm of the village council in managing the subproject. Its tasks include managing the subproject bank account (that is, receiving disbursements of TASAF II funds), bookkeeping, handling procurement, and other ongoing implementation tasks for all types of subprojects. Accountability · LG staff members are involved in activities prior to the start of the subproject cycle to help ensure that community-managed subprojects are integrated into the district planning cycle. · LG staff members become involved early on in the process of extended Participatory Rural Appraisals that verify relevance of existing analyses of community needs and priorities. This exercise, which builds relationships and also helps LGs understand community-level dynamics, might presumably involve discussions with CBOs.117 · District Council finance committees approve subprojects, with input from other government and civil society entities. · Most of the other accountability arrangements cited in the PAD, including interactions among a variety of committees, involve project-specific elected committees (rather than organic CBOs), such as the Community Management Committee. This committee regularly reports on the use of resources to village and district LGs, the approval of which is required at various points. Filling the Opportunity Space As explained above regarding the Philippines, TASAF II has provided extensive capacity support for projects that has strengthened the capacities of and environment for both CBOs and, to some extent, local governments.118 With this in mind, we again consider that CBOs and management committees operating under TASAF II, while still facing considerable challenges, will be able to achieve more than similar structures outside the project's reach. TASAF support under the capacity enhancement component will mitigate some contextual constraints through direct training but also by helping provide staff (in some instances), vehicles, and equipment. 117. As previously identified in wide-ranging, open-ended Participatory Rural Appraisals. 118. For example, TASAF I helped overcome banks' high deposit requirements by using the first tranche of project funds to open the account, into which cost-sharing contributions can subsequently be deposited. It may also influence the district level by-laws that apparently deal with some elements of CBO activities, including registration (N. Mungai Lenneiye 2005), and enhances local government capacity when it provides vehicles, equipment, and (in some instances) staff. 77 BIBLIOGRAPHY Ackerman, John M. 2005. "Social Accountability in the Public Sector: A Conceptual Discussion." Social Development Papers in Participation and Civic Engagement,Paper 82, World Bank, Washington, DC. AMUNIC-INIFOM. 2004. Sistema de planificación municipal, guia ilustrada: II, Versión actualizada. Managua: AMUNIC-INIFOM. Ballivián, Amparo, Antonio Spilimbergo, Jonas Frank, and Armando Barrios. 2004. "Municipal Decentralization in Nicaragua: Fiscal and Institutional Issues." Policy Note, World Bank, Washington DC. Bhatnagar, Bhuvan, and Clifford Burkley. 2004. "Philippines: Kapitbisig Laban Sa Kahirapan (Kalahi) Project." Paper prepared for the Shanghai conference. Binswanger, Hans, and Swaminathan Aiyer. 2003. "Scaling Up Community Driven Development: Theoretical Underpinnings and Program Design Implications." Internal memo, World Bank, Washington, D.C. Binswanger, Hans, and Tuu-Van Nyugen. 2005. "Scaling Up Community Driven Development for Dummies." Internal memo, World Bank, Washington, DC. Coirolo, Luis, Keith McLean, Mondonga Mokoli, Andrea Ryan, and Melissa Williams. 2001. "Community Based Rural Development: Reducing Rural Poverty from the Ground Up." World Bank Rural Strategy Discussion Paper 6, World Bank, Washington, DC. Commonwealth Local Government Forum. 2005. "The Local Government System in Zambia." www.clgf.org.uk. Djikstra, Geske. 2000. "Nicaragua: Fondo de Inversión Social de Emergencia (FISE)." Social Fund Background Paper, OED. Draft. Dongier, Philip, Julie Van Domelen, Elinor Ostrom, Andrea Ryan, Wendy Wakeman, Anthony Bebbington, Sabina Alkire, Talib Esmail, and Margatet Polski. 2001. "Community Driven Development." PRSP Sourcebook, Vol. 1. Washington, DC: World Bank. Frank, Jonas. 2004. "Decentralization in Nicaragua: The Challenge of Fiscal Neutrality." World Bank: Washington DC. Second draft. Gillespie, Stuart. 2003. "Scaling Up Community Driven Development: A Synthesis of Experience." International Food Policy Research Institute, Washington, DC. Grun, Stephen. 2002. Evaluación del programa de descentralización de gestión y manejo del ciclo de proyectos: Fondo de Inversión Social de Emergencia (FISE), informe final. Managua. 78 Guadagni, Maurizio, et al. 2000. Local Governments and Sustainable Rural Development: Effects and Potential Sustainability of the Rural Municipalities Project (Protierra) in Nicaragua. World Bank: Washington DC. Joshi, Anuradha, and Mick Moore. 2002. "Organizations that Reach the Poor: Why Co-Production Matters?" Paper presented at the World Development Report 2003/04 Workshop, Oxford. Krishna, Anirudh. 2004. "Partnerships between Elected Local Governments and Community-Based Organizations: Exploring the Scope for Synergy." Social Development Paper 52, World Bank, Washington, DC. Mclean, Keith. 1998. "Characterizing Decentralization in Nineteen Countries." Internal memo, Rural Development Department, World Bank, Washington DC. Lenneiye, N. Mungai. 2005. "Who Has the Yam, and Who Has the Knife? Social Action Funds and Decentralization in Malawi, Tanzania, and Uganda." Social Protection Discussion Paper Series, No. 0518. The World Bank, Washington, D.C. Narayan, Deepa. 1995. "The Contribution of People's Participation: Evidence from 121 Rural Water Supply Projects." ESD Occasional Paper Series 1, World Bank, Washington, DC. Nicaragua, Government of. Emergency Social Investment Fund Operational Manual. http://www.fise.gob.ni/pgc. Ortega Egg, Manuel. 1997. "La descentralización en Nicaragua." In Politicas y propuestas de descentralizacion en Centroamérica, coord. Cordova Macias. El Salvador: FLASCO. Ostrom, Elinor. 1996. "Crossing the Great Divide: Co-production, Synergy, and Development." World Development 24 (6): 1073­87. Petersen, Karen Heleg. 1997. "Decentralization and Municipal Finances in Nicaragua." Central America Department, World Bank, Washington DC. Romeo, Leonardo. 2003. Anexo 8: Evalucion del FONDEM; Contribution to the midterm review of the SPDMR.. The World Bank, Managua. Serrano-Berthet, Rodrigo, and D. Warren. 2003. "Re-engineering Social Funds for Local Governance: The Central American Experience." The World Bank. Washington, DC. Spilimbergo, Antonio. 2004. "Fiscal Decentralization in Nicaragua." The World Bank, Washington DC. Steffensen, Jesper, Per Tidemand, and Eke Mwaimpopo. 2004. "A Comparative Analysis of Decentralization in Kenya, Tanzania and Uganda: Final Report." The World Bank, Washington, DC. Tanzania, Government of. 1998. "The Policy Paper on Local Government Reform." 79 Wong, Susan, and Scott Guggenheim. 2004. "Community-Driven Development and Decentralization in the East Asia Region." In Flagship Study: Managing Decentralization in East Asia. The World Bank, Washington DC. World Bank. 1996. Rural Municipalities Project: Staff Appraisal Report. Washington, DC: World Bank. ________. 1997. Implementation Completion Report: Nicaragua, Social Investment Fund Credit. Washington, DC: World Bank. ________. 1998. Project Appraisal Document: Third Social Investment Fund Project. Washington, DC: World Bank.. ________. 1999. Mindanao Rural Development Project: Project Appraisal Document. Washington, DC: World Bank. ________. 2000a. Local Governments and Sustainable Rural Development: Effects and Potential Sustainability of the Rural Municipalities Project (Protierra) in Nicaragua. Washington, DC: World Bank. ________. 2000b. Tanzania Social Action Fund Project Appraisal Document. Washington, DC: World Bank. ________. 2001a. Nicaragua: Second Rural Municipalities Development Project. Washington, DC: World Bank. ________. 2001b. Nicaragua Poverty Reduction and Local Development Project. Washington, DC: World Bank. ________. 2001c. The United Republic of Tanzania Public Expenditure Review. Washington, DC: World Bank. ________. 2002a. Kapitbisig Laban Sa Kahirapan-Comprehensive and Integrated Delivery of Social Services: Project Appraisal Document. Washington, DC: World Bank. ________. 2002b. Romania Rural Development Project: Project Appraisal Document. Washington, DC: World Bank. ________. 2002c. Rural Municipalities Development Project: Implementation Completion Report. Washington, DC: World Bank. ________. 2003a. Philippines Public Expenditure Review. Washington, DC: World Bank. ________. 2003b. World Development Report 2004: Making Services Work for the Poor. Washington DC: World Bank; New York: Oxford University Press. ________. 2003c. ZAMSIF Beneficiary Assessment Report. Washington, DC: World Bank. ________. 2003d. ZAMSIF Mid-Term Review Issues Paper, 14 March. Washington, DC: World Bank. 80 ________. 2003e. ZAMSIF Mid-Term Review Issues Paper II, 24 August. Washington, DC: World Bank. ________. 2004a. "State-Society Synergy for Accountability: Lessons for the World Bank." World Bank Working Paper 30, World Bank, Washington, DC. ________. 2004b. Third Social Investment Fund Project. World Bank: Washington, DC. ________. 2005. "Linking Community Empowerment, Decentralized Governance, and Public Service Provision through a Local Development Framework." Internal memo, World Bank, Washington, DC. Zambia, Government of, Ministry of Local Government and Housing. 2002. "Draft Report on Decentralization and Local Government." 81