Local Governance and Service Delivery in Bosnia and Herzegovina

Transcription

1 From Stability to Performance Local Governance and Service Delivery in Bosnia and Herzegovina Document of the World Bank Sustainable Development Department Europe and Central Asia Region (ECSSD)

2 From Stability to Performance Local Governance and Service Delivery in Bosnia and Herzegovina Sustainable Development Department Europe and Central Asia Region (ECSSD) Document of the World Bank January, 2009

3

4 ABBREVIATIONS AND ACRONYMS ALDI BFP BH CDPC CEPOS DFID DPA EBRD EDA EIU EU RED EU SAA EU FBH FBHMTC FIAS GAP GDP GIB HDZ HSP HSS IDPs IMF IMF GFS ITA KM LDP LSMS MA MC MOF MOPE MTCs MTDS MUCCE MZ Local Development Initiatives Agency Budget Framework Papers Bosnia and Herzegovina Community Development Planning Committee Center for Policy Studies UK Department for International Development Dayton Peace Agreement European Bank for Reconstruction and Development Economic Development Administration Economist Intelligence Unit European Union Regional Economic Development Project European Union Stabilization and Association Agreement European Union Federation of Bosnia and Herzegovina Federation of Bosnia and Herzegovina Ministry of Transport and Communications Foreign Investment Advisory Service Governance Accountability Project Gross Domestic Product Annual Budget Report (Godisnji Izvjestaj o Budzetu) Croatian Democratic Union Croatian Party of Rights Croat Peasant Party Internally Displaced Persons International Monetary Fund IMF Government Finance Statistics Indirect Taxation Authority Konvertible Mark Local Development Project Living Standards Measurement Study Municipal Administration Municipal Council Ministry of Finance Ministry of Urban Planning and Environment Ministries of Transport and Communication Medium Term Development Strategy Ministry of Urban Housing, Communal Services, Civil Engineering and Ecology Mjesna Zajednica

5 NGO O&M OBD OECD OHR OSCE PDP PEFA PFM PIER PRSP RAD RD RS RSMTC RzB SBH SDA SDP SDS SEED SIGMA SME SMOTC SNSD SOE SRS UNDP UNHCR USAID VAT VP YEPP Non-Governmental Organization Operations and Maintenance Original Budget Data The Organization for Economic Co-operation and Development Office of High representative in Bosnia and Herzegovina Organization for Security and Co-operation in Europe Party of Democratic Progress Public Expenditure and Financial Accountability Public Finance Management Public Expenditure and Institutional Review Poverty Reduction Strategy Paper Sarajevo Kanton Solid Waste Management Company Road Directorate Republika Srpska Republika Srpska Ministry of Transport and Communications Work for Improvement-Peoples Party Bosniak Party for Bosnia-Herzegovina Bosniak Party of Democratic Action Social Democratic Party Serb Democratic Party South East Europe Enterprise Development Support for Improvement in Governance and Management Small and Medium Enterprises State Ministry of Transport and Communications Union of Independent Social Democrats State Owned Enterprise Serb Radical Party United Nations Development Programme United Nations High Commission for Refugees United States Agency for International Development Value Added Tax Vodoprivredas Youth Empowerment Partnership Program

6 ACKNOWLEDGEMENTS The report was prepared as a background analysis to inform the ongoing policy dialogue and operational program of the World Bank with Bosnia and Herzegovina on municipal governance and service delivery. This study was developed under the guidance of Orsalia Kalantzopoulos and Jane Armitage (Country Directors), prepared under the direction of Lee Travers and Wael Zakout (Sector Managers), and has benefited from the continuous encouragement and guidance from Dirk Reinermann and Marco Mantovanelli. The study was led by Aniruddha Dasgupta (Task Team Leader) and prepared by Yan Zhang, Kai Kaiser, Bjorn Philipp, Paula Lytle and Vesna Francic. Research support was provided by Irina Smirnov, Emir Dervisevic, Halko Basaric, and Zehra Kacapor. Bonita Brindley edited the report to a much more enjoyable piece. The collaboration from Mary McNeil and Andre Herzog of the World Bank Institute (WBI) is gratefully acknowledged. The household survey was conducted by PRISM Research, with support from the Austrian Government through WBI. Senad Sacic helped with coordinated mission schedules, data requests and translation work. Delphine Hamilton formatted the report and Tijana Medi} translated the document in local language. The report has benefited tremendously from discussions with and comments from colleagues at different stages of preparation, in particular: Goran Tinjic, Orhan Niksic, Irina Smirnov and Jacques Bure, Richard Martin Humphreys, and Junghun Cho. The peer reviewers were Matthew D. Glasser, Larry Hannah, William Dillinger, and Barbara Nunberg. The team is indebted to officials from the BH at all levels, to government agencies and non-governmental organizations, for providing key inputs, data and help with the analysis. The team is also grateful for comments and discussions with colleagues from the IMF, OSCE, OHR, EC, EBRD, KfW, USAID, Sida, UNDP and other organizations. Special thanks go to the insightful comments from the Governance Accountability Program (GAP) supported by USAID, Sida, and EKN. The team is particularly grateful to Marc Ellingstad, Elisabet Tomasinec, and Erik Litver for their support. An early version of the report was presented and discussed with government representatives from BH at all levels, including mayors, at a meeting organized by the World Bank, held in Sarajevo on September 17, 2008.

7

8 CONTENTS EXECUTIVE SUMMARY i Chapter 1: Local Governance and Service Delivery in Bosnia and Herzegovina Municipalities and post-conflict decentralization Municipalities are critical for public service delivery in Bosnia and Herzegovina The State of Local Service Delivery Outcomes in Bosnia and Herzegovina Variations across municipalities Variations across sectors Key findings: What do service delivery outcomes in BH tell us? The new challenge ahead: improvement of local service delivery performance Chapter 2: Bosnia and Herzegovina s Decentralization Framework The Evolving Decentralization Framework Dimensions of Accountability for Decentralized Service Delivery Horizontal Accountability Vertical Accountability Conclusions Chapter 3: Water and Roads: How Sector Structures Influence Local Service Delivery Performance Municipalities and the Roads Sector General state and organization Road Sector Financing Municipalities and Water Utilities General state and organization Water Sector Financing Conclusions Chapter 4: Options to Expand the Envelope for Capital Investments in Municipalities Overview Municipal Public Finances Own-Source Revenues Fiscal Transfers Single Account Fiscal Transfer Allocations Expenditure Composition and Efficiency Municipal Borrowing Conclusions Chapter 5: Public Financial Management Performance across Municipalities The Why and What of Sub-national PFM Key Findings and Summary of PFM Assessment

9 5.2.1 Credibility of the Budget Budget comprehensiveness and transparency Policy-Based Budgeting Predictability and control in budget execution Accounting, Recording, and Reporting External Scrutiny and Audit Donor Practices Conclusions Chapter 6: Conclusions and Going Forward Strengthen the Enabling Environment Expanding fiscal space Increasing accountability Supply Side: Enhancing Horizontal Accountability Demand Side: Enhancing Vertical Accountability Annex 1.1: Municipal Sample Survey Selection Annex 1.2: Service Outcome Indicators and Ranking Methodology Annex 2: BH Expenditure Assignments Annex 3.1: GIB Revenue Classifications Annex 3.2: GIB Expenditure Classifications Annex 4: Cantonal-Municipal Revenue Sharing Arrangements Annex 5.1: Local Political Dynamics and Space for Change Annex 5.2: Diagrams on Municipal Structure Annex 6.1: Summary of PEFA Outcomes of BH Municipal Governments Annex 6.2: PEFA Scoring Spider Diagram Annex 6.3: Detailed Assessment of PFM (approximately 10 pages) References

10 FIGURES Figure 1: Overview of Political/Administrative System Figure 2: Population and Size Distribution of Municipalities Figure 3: Average Population of Municipalities in Selected European Countries Figure 4: Service outcomes vary widely across municipalities Figure 5: Service Delivery Outcome Score Figure 6: Income alone does not explain better outcomes Figure 7: Small municipalities can achieve excellent service outcomes Figure 8: High population density helps, but not a condition for good service outcomes Figure 9: Service outcomes vary across old and new municipalities Figure 10: but are more income-sensitive in newer municipalities Figure 11: Access to basic services is low and people are not satisfied Figure 12: Room for improvement, particularly for water and waste Figure 13: Quality ratings need balanced interpretation Figure 14: Citizens face harsh consequences if roads are inaccessible during winter Figure 15: Almost half of households with water connection experience problems Figure 16: Higher-income municipalities provide better waste collection Figure 17: that meets household needs Figure 18: Quality and reliability decline with lower per capita municipal revenue Figure 19: Most households operate their own septic tanks Figure 20: Predominantly rural households face problems with their sewage system Figure 21: Feedback to providers would improve service quality Figure 22: Major Municipal Expenditure Assignments by Entity Figure 23: Governance Structure for Municipal Service Delivery Figure 24: Participation or attendance at various local governance activities Figure 25: Unbundling Roles and Responsibilities for Decentralized Service Delivery Figure 26: More roads than water pipes in under-funded municipalities Figure 27: Municipalities can increase satisfaction in the roads sector Figure 28: Road Network in BH Figure 29: Institutional Responsibilities and Decentralization of the Road Sector in BH Figure 30: FBH Road Sector Fiscal Flows before Figure 31: FBH Road Sector Fiscal Flows after Figure 32: Type of community involvement Figure 33: Sources and Uses of Financing for roads in FBH Figure 34: FBH Water Sector Fiscal Flows before July 2007 (status after July 2007) Figure 35: Municipal Expenditures in BH (2007, projected) Figure 36: Aggregate Municipal Expenditures (2006, estimated) Figure 37 : Municipal Diversity

11 Figure 38: Non-Tax Revenues versus Per Capita Revenues Figure 39: Municipal Revenue Structure (RS percent) Figure 40: Vertical Shares from the Single Account Figure 41: Phasing in Ratios for the Single Account distribution formula (percent) in FBH Figure 42: Formula versus Legacy in Municipal Single Account Allocations (FBH) Figure 43: Formula versus Legacy in Municipal Single Account Allocations (RS) Figure 44: Formula versus Legacy in Municipal Single Account Allocations (FBH) Figure 45: Wage Share against Per Capita Revenues Figure 46: Selected Capital Expenditures by Municipalities (RS, 2005) Figure 47: Expenditures by Functional Classification (percent) in RS Figure 48: Key Regulations Concerning Municipal PFM Figure 49: Service Outcome Score and PFM Scores Figure 50: PFM Scores and Population Size Figure 51: PFM Scores, Service Outcome Score, and Per Capita Revenue Figure 52: PFM Scores, Service Outcome Score, and Legacy Figure 53: Average Absolute Deviation of Primary Budgets for 20 Municipalities Figure 54: Sample Instances of Municipal Borrowing Figure 55: Summary of Policy Recommendations, Institutions, Actions, and Instruments Figure 56: Directly elected mayors are generally from the majority party in each Municipal Council.97 BOXES Box 1: Municipalities in Bosnia and Herzegovina: Legacies of Conflict Box 2: Sarajevo: Cooperation brings benefits to citizens Box 3: The BEACON Scheme: learning from performing municipalities Box 4: Major Political Parties in BiH Box 5: Bugojno-moving forward together Box 6: Breza-Citizens participate in building Local Infrastructure Box 7: Small steps to attract small business-ljubinje deregulates local fees Box 8: Ugljevik: Budgeting based on citizens needs Box 9: Foca: Local Democracy at work Box 10: Certifying quality public services: commitment to customer satisfaction Box 11: MZS have lost their pre-war role in community development

12 From Stability to Performance Local Governance and Service Delivery in Bosnia and Herzegovina EXECUTIVE SUMMARY 1. A decade after the conflict, Bosnia and Herzegovina municipalities have stabilized and are free from the fear of violence. Today, thanks to a positive external environment, the economy is stable and enjoying strong growth. The successful introduction and administration of the Value Added Tax (VAT) has boosted local-level revenues and led to a positive fiscal outcome in 2006 and After months of paralysis, the Stabilization and Association Agreement has been signed with the European Commission, putting the country on track for eventual European Union membership. This combination of political stability and good economic performance provides an opportunity to tackle systemic weaknesses in local-level public service delivery that limit the country s long-term economic growth and living standards. 2. Of necessity, the Dayton Peace Agreement created post-conflict institutions that secured peace through compromise and power-sharing. But now that peace is secured, and donor financing is dwindling, these institutions are beginning to appear cumbersome, expensive, and inefficient. Policy functions are fragmented and duplicated across two entities and levels of government. Despite large public sector spending-more than 45 percent of GDP-the quality of municipal public service delivery is low and users are dissatisfied. 3. The research for this report revealed that the municipal level is an exceptionally promising platform to strengthen government responsiveness across a range of basic services that benefit both households and businesses because municipal governments enjoy strong political legitimacy among citizens. This report argues that local service delivery performance can be improved without the delays associated with the time-consuming and politically difficult path of constitutional reform. In fact, the passage of the Local Self Governance Laws, Revenue Allocation Laws, and the ITA reforms has established the basic framework needed. Therefore, this report suggests that much can be accomplished by introducing a range of measures to improve local service delivery performance without increasing overall public sector spending. 1. Methodology 4. Three sets of instruments were adopted. The enabling environment was explored through a desk review of laws and regulations affecting municipal performance; and in-depth interviews with key informants to discover implementation status. The demand side of municipal services was explored by conducting a two-thousand-household user survey on local service delivery across 20 municipalities. The survey covers 10 services, five of which are under municipal responsibility. The supply side of services was explored by collecting National Fiscal Data (across all 142 municipalities) from standardized ( GIB ) data to understand the fiscal positions of municipalities and to examine priorities for managing fiscal space that had potential for improving service delivery. In-depth fiscal and public financial management diagnostics were conducted for 20 municipalities, an instrument from the Public Expenditure and Financial Accountability (PEFA) framework adapted to the municipal level to reflect characteristics and realities of municipal government and BH s local context. In addition, in-depth interviews were conducted with mayors and local councilors to discover the primary drivers for reforms, and with key informants on road and water sectors to explore sector-specific challenges. Lastly, brief case studies of service delivery examples illustrate preconditions for these successes. i

13 2. Main Findings 5. This report anchors itself in service outcomes. It offers three main findings on local service delivery outcomes and five primary dimensions that were found to influence them. A. Service Delivery Outcomes 6. Local service delivery outcomes are extremely uneven across municipalities and sectors, access to services is poor, and overall user satisfaction is low. More than a decade after the Dayton Peace Agreement, it is evident that among the services provided at the municipal level-water and waste have the most room for improvement. Almost 1.4 million BH citizens have no running water or waste removal. Water coverage and quality is a problem for many households. Waste collection coverage is inadequate and its quality and reliability decline with lower per capita municipal revenue. The quality of local roads is good but reliability and access need improvement. 7. Municipalities lack resources and need to increase fiscal space to expand access to basic services. This is especially true for rural areas and among municipalities with low per capita revenues. Furthermore, road and water sectors both have capital-intensive network infrastructures, and service outcome improvements depend heavily on the financial arrangements in both sectors. Expanding and rehabilitating roads and water pipelines requires funding that municipalities alone cannot provide. 8. Service providers lack incentives to improve their performance. Municipal service providers operate in an information vacuum. The current accountability structure isolates them from the feedback they need to improve performance, including feedback provided by an incentive system that rewards better performers. To date, no publicly available information exists on normative performance benchmarks, so service users and citizens have limited ways to hold delivery providers accountable, and without channels to receive feedback, the delivery providers have scant information on what they need to do to improve user satisfaction. B. Factors Influencing Service Delivery Outcomes 9. The report examines service delivery outcomes along five essential dimensions: (a) enabling environment; (b) sector structure; (c) municipal fiscal space and resources; (d) local public financial management capacity; and (e) accountability structure. 10. Enabling Environment. The basic enabling environment for BH municipalities to provide better public services has been in place since the passage of a series of laws such as the Local Self- Governance law in both entities, the law on Revenue Allocations, the introduction of VAT, and the clarifications regarding rules for municipal borrowing. The ITA reforms and the Revenue Allocation laws in both entities have institutionalized a more stable and autonomous source of revenues for municipalities. The recent Local Self-Governance Laws are a step toward strengthening local government discretionary space and accountability. But many challenges remain to fully implement the Local Self-Governance law and to harmonize sectoral legislation with the Local Self-Governance law. Although decentralization principles have been laid out, the degree of progress varies widely among sectoral ministries (including the cantonal ministries). Municipalities have been lacking national level champions to advance municipal sector. 11. Sector Structure. The overarching framework for decentralized service delivery is in place, but wide variations in road and water sectors functional and financial arrangements indicate that decentralization manifests differently by sector. Water has been largely decentralized to the municipal level, whereas in the road sector, municipalities and higher-level governments in BH share responsibility for local roads provision. As a result, improving municipal performance requires different actions for different sectors. In some sectors, such as roads, multi-level responsibilities create fragmentation that exacerbates reduced capacity, leading to poor performance outcomes. Therefore, streamlining or realigning the functional assignments is necessary to capture economies of scale and to improve coordination across jurisdictions. Some local roads service outcome improvements are at the mercy of decisions made at higher levels of government, but municipalities can improve customer satisfaction by offering incentives (performance-based management contract, for instance) for local road ii

14 maintenance companies to provide more reliable services. In contrast, in the water sector, cost and benefit are locally determined and costs should be covered locally. Municipal performance improvements will benefit from further commercialization of water utility companies, and strengthening municipal financial capabilities through appropriate tariff decisions made in the municipal councils. Both sectors also have common concerns-finance and regulatory oversight of the public enterprises from the municipal council and upper-level government agencies to ensure public services access, quality, reliability, and affordability, and to mitigate contingent liabilities. 12. Municipal fiscal space and resources. Bosnia and Herzegovina municipalities have diverse fiscal positions in current and prospective transfers from higher-level governments, own-source revenues, borrowing levels and potential, prevailing expenditure structures, and the related scope for efficiency gains and/or improved prioritization of expenditures. Similarly, municipalities have widely diverse revenue capacity-own-source revenues (user fees and charges) represent close to onethird of the total. The ITA/VAT transfers replaced derivation-based sales tax and reduce the disparities of per capita revenue across jurisdictions. Depending on the growth of ITA pool, VAT allocations could fall short of expenditure pressures for the special needs of urban or peri-urban municipalities with large returnees. All municipalities could increase fiscal space by improving their expenditure efficiency, for example by reducing gap-filling expenditures for unfulfilled higher-level government responsibilities and stop-gap transfers to affiliated public sector enterprises, by increasing efficiency in wage bills, and by better prioritizing projects. Recently, municipal borrowing has been allowed in both entities, which increases the importance of a quality regulatory environment and enhanced monitoring of municipal fiscal well-being, therefore a cross-cutting issue is the lack of an effective financial reporting system and process on which to base quality analysis of fiscal space and the potential for its expansion. 13. Local public financial management capacity. The fundamental legal framework for good financial management is largely in place in BH. Recent financial management reforms have achieved some success: Bosnia and Herzegovina has established fundamental public financial management systems for managing macro-fiscal developments, some elements for enabling strategic allocation of resources, and a few tools for improving operational/technical efficiency. However, a systematic assessment of PFM outcomes in 20 municipalities shows that the fragmented and evolving policy and legal/regulatory environment has handicapped local-level policy design and implementation. In general, better PFM performance is associated with better overall service outcomes across 20 municipalities. PFM performance varies greatly across municipalities, requiring intensive customized capacity building at all levels, particularly on medium-term (multi-annual) budgetary planning to ensure strategic resource allocation and prioritized sectoral spending, general oversight of utilities and public enterprises, budget execution, and accounting records and practices. Benchmarking municipal public financial management is a useful practice to highlight weaknesses, allow governments to develop remedial plans, and provide the basis for reform dialogue among development partners and governments. 14. Accountability Structure. Detailed analysis of the existing and emerging accountability mechanisms reveals strengths and weaknesses in the checks and balances to ensure good service outcomes. Municipalities are subjected to vertical accountability to non-state actors-citizens and service users-via periodic elections and informal processes through which citizens demand good performance from officials and service providers. The introduction of jurisdiction-based direct election of Mayors provides the opportunity to build a broad-based policy platform and enhance performance accountability to citizens. Elected Mayors have the desire to create well-run municipalities, and their peer-to-peer relationships create both the competitiveness and the collegial support that appear to drive reforms. 15. Horizontal accountability includes state-to-state relationships either at the same level-municipal council and municipal administration-or across tiers of government. Horizontal accountability is characterized by weak oversight on the part of the Municipal Council on the service and PFM performance of the Municipal Administration, including the governance and financial relationship between municipal companies and municipal government. Horizontal accountability is further complicated by a four-level split and de facto overlapping of responsibilities. Higher-governments (Ministry of Finance and line/sectoral ministries) monitoring and oversight roles are constrained by lack of service performance standards, weak infrastructure for financial reporting, and lack of monitoring and evaluation capacity. Overall, the current accountability structure is weakened by lack of adequate information on service delivery outcomes and effective channels for feedback. iii

15 3. Recommendations Going Forward 16. In this evolving post-war period, Bosnia and Herzegovina can begin to pursue policy agendas to orient its systems for higher performance without the time consuming and politically difficult path of constitutional change. This report points out that the achievements of the last few years-significant changes in legislation, regulations, and financing-can now make a consolidated contribution to the transition from municipal stability to municipal performance. 17. To improve municipal services delivery performance the government needs to tackle three main areas. First BH needs to keep strengthening the enabling environment. Second, municipalities must expand their fiscal space to enhance local service delivery. Each municipality should examine the four dimensions of fiscal space-transfers, own-source revenues, expenditure efficiency, and borrowing to maximize the dimensions with the most potential to increase local-level fiscal capacity to improve service outcomes. Third, municipalities must be held accountable to citizens for service delivery. This will require implementing a performance-based system that includes publicly available performance benchmarks and indicators, and offers incentives to providers to improve their service delivery. The following key recommendations will help realize this crucial transition. 18. Strengthen enabling environment. BH must fully implement the Laws on Local Self- Governance and harmonize sectoral legislation with Laws on Local Self-Governance. Development partners need to support a coherent agenda for the municipal sector, including encouraging stateand entity -level champions to effectively implement decentralization and municipal sector operation. Sector ministries should work with municipalities to develop policies and strategies and set up frameworks for sectoral standards and institutional arrangements to monitor, evaluate, and enforce sector performance standards for municipal service delivery. 19. Expand fiscal space. First, the State and two entities should complete the phasing in period of revenue sharing formula to enable a predictable and transparent revenue stream to the municipalities. Second, affluent municipalities must increase own-source revenues by gradually raising tariffs to cost-recovery levels, adopting business-friendly fees and charges, and developing more buoyant tax bases, including property taxes. Third, the Ministry of Finance at all levels should strengthen its ability to monitor fiscal risks and help municipalities to be on a sounder financial footing. The international community can provide technical assistance to municipalities to strengthen their public financial management as well as capital investment planning and project appraisal capacities, which would enable them to take advantage of borrowing. Fourth, municipalities must increase expenditure efficiency, which will require detailed data and clear functional classifications to align resource allocation with policy. 20. Increase accountability. Both vertical and horizontal mechanisms require strengthening through a framework for service delivery standards, monitoring and evaluation, and oversight bodies to ensure compliance across constituent jurisdictions. Higher-level governments should work with municipalities to establish minimum service standards and estimate fiscal resources needed. Next Ministry of Finance and sector ministries need to strengthen their financial and technical oversight functions and design performance-based grants to complement the existing intergovernmental transfers. Municipal councils must improve oversight of service providers, monitor the decisionmaking of municipal executives, and municipal administrations must enhance their PFM capacity. Citizens and local officials could play a more significant role in securing benefits of efficient local government, and increase civil society voice and participation. Development partners can promote systematic and independent benchmarking of municipal service delivery and PFM performanceover temporal and spatial dimensions-because accurate and comprehensive data are essential to increase both supply and demand-side accountability. iv

16 Chapter 1: Local Governance and Service Delivery in Bosnia and Herzegovina 1.1. Municipalities and post-conflict decentralization 1. The current political context is a legacy of conflict. In November 1995, the Dayton Peace Agreement (DPA) ended the war that began in April 1992, following a referendum on independence from Yugoslavia when Bosnian Serb leaders refused to be incorporated into the new state and declared a separate state of their own. In the ensuing conflict some 250,000 people in Bosnia and Herzegovina died, the scale of internal population displacement was massive, and one million people became refugees-primarily in other European countries. 2. Local fighting varied in intensity and alliances shifted during the war. Sometimes Croat and Bosniak forces fought each other and sometimes they were allied against the Serbs. On March 1994, the Washington Agreement negotiated a ceasefire between Bosniak and Croat forces, which resulted in a federation between them. The two groups were largely concentrated but still living in close proximity along informal but well-understood demarcation lines that were determined by local conflicts. The 1995 DPA ended remaining conflict between the Serbs and the allied Croat-Bosniak forces. 3. BH s post-conflict institutional structure sought to ensure peace. The decentralized state structure carefully balances political/ethnic interests. The national government remains weak and most power is vested in two entities: the Federation of Bosnia and Herzegovina (FBH) and Republika Srpska (RS), plus the special district of Brcko (see Figure 1). The Federation is divided into ten cantons; Sarajevo Canton has a special status that primarily gives it greater control over its constituent municipalities. 4. The DPA established two entities-fbh and RS. FBH has two tiers of sub-entity governments, primarily to accommodate both Bosniak and Bosnian Croats interests. FBH s structure of ten cantons and 79 underlying municipalities is more complex and decentralized than in RS, where relations between the entity and municipalities are direct. 1

17 Figure 1: Overview of Political/Administrative System 5. FBH cantons have their own diverse organizational structures. 1 Ten highly autonomous cantons-a legacy of protecting ethnic homogeneity-have fully fledged legislative and executive structures; the canton heads are elected by a majority of the cantonal Assembly from candidates nominated by legislators. The main organizational units of cantons are ministries. Each canton has its own legislation, regulation, decrees, administrative decisions that might affect the municipal level Cities are an additional structure in urban governance. In the FBH, municipalities with more than 30,000 inhabitants can establish city status to amalgamate and coordinate policies among urban municipalities. There are three cities in FBH-Sarajevo3, Mostar, and Tuzla, and two in RS-Banja Luka and Srpsko Sarajevo. Cities have their own budgets, financed by own revenues, shared revenues, and grants from cantons (in the Federation) or Entity (in RS). There have been cases where the administrative relationship between cities and municipalities was not always clear and remained to be settled by politics or the judiciary. However, Mostar City has been reformed and unified, and Tuzla and Mostar are now being treated as other municipalities in FBH; but Sarajevo has kept a complex structure city in FBH. 7. BH Municipalities had a history of self-governance. In the former Yugoslavia, municipal governments had significant functional autonomy. 4 The last pre-war municipal elections were held in 1990 and during the war some internally displaced municipal council members established municipalities-in-exile during the war. Since the war, 5 many initiatives to achieve political stability have focused on municipalities, including improving public administration, participation, and accountability. 1. Five cantons (Una-Sana, Tuzla, Zenica-Doboj, Bosnian Podrinje, and Sarajevo) have a Bosniak majority, three (Posavina, West Herzegovina, and West Bosnia) have Bosnian Croat majority, and two (Central Bosnia and Herzegovina-Neretva) are ethnically mixed, with special legislative procedures to protect constituent ethnic groups. 2. Before the Law on Local Self-Governance was passed, each canton had its own cantonal constitution governing the operations of municipalities. All cantons passed local self-government laws between 1995 and 1999, which vary to a considerable degree, particularly regarding the assignment of functions and resources. These differences in the legal framework were reinforced by different implementation practices. 3. The City of Sarajevo encompasses 4 municipalities (Stari Grad, Centar, Novi Grad and Novo Sarajevo). 4. Bogdan Denitch, Evolution of Yugoslav Federalism. Journal of Federalism, Municipalities in the former Yugolsavia controlled revenue sources through taxation and revenue from local economic enterprises and exerted authority through the control of appointments of the heads of socially owned enterprises. 5. These municipalities-in-exile were abolished after the first post-war municipal elections in

18 8. Conflict legacies are reflected in current municipal structure. More than forty new municipalities were created; some pre-war municipalities lost territory or were divided, many new municipalities were a Mjesna Zajednica (MZ) of the parent municipality that was split off. Most new municipal boundaries reflect changes in ethnic composition. Municipalities were divided between the FBH and RS, and MZs with ethnic minorities became new municipalities. Many new municipalities are small, rural, have minimal infrastructure, and are populated by refugees and internally displaced persons (IDPs) from the parent municipality. Other municipalities had well-established infrastructure before the war but sustained significant damage. The DPA also affected municipal governance; new municipalities struggled with administration, and some had minimal infrastructure. The parent municipalities remained with fragmented territory, documentation, and history. 9. Municipalities are now a safe place to live. Recently the pace of return has slowed; it appears that many people who intended to return to their municipality of origin have done so. A 1998 survey found that displaced people considered personal security as the main obstacle to returning; but during , personal security concerns were largely supplanted by issues such as progress with reconstruction and potential for employment. In 2007, 6 most household survey respondents considered their municipality a safe place to live, even in a municipality such as Mostar, almost half of respondents considered the municipality safe, and less than one-quarter considered it unsafe Municipalities vary widely in size, population, and resources. Population size ranges from a micro municipality-istocni Stari Grad with 62 inhabitants, to Banja Luka with over 224,000. However, 18 municipalities have over 50,000 people, and six of these are over 100,000. Over half of municipalities have more than 18,000 inhabitants and most of BH s population lives in reasonably sized municipalities. 8 Figure 2 depicts total population in each size classification. About one fifth (27) of BH municipalities had less than 5,000 inhabitants but these accounted for less than one percent of BH s population. Figure 2: Population and Size Distribution of Municipalities 6. A 1998 survey found that displaced people said fear for their personal security was the main obstacle to returning and most preferred to return as a group (See Bosnia Social Assessment, 1999). In early 2000, after four and half years of international reconstruction efforts, refugee return began to increase dramatically (See ICG, Bosnia s Refugee Logjam Breaks: Is the International Community Ready? May 2000). 7. In Odzak, Jablanica, Mostar, Stolac, Berkovici, and Novo Sarajevo, less than half of respondents said their municipality is safe; but more respondents said, neither safe nor unsafe, or refused to answer. 8. The most recent census in Bosnia-Herzegovina was population figures are estimates. 3

19 11. Small municipalities can be viable. 9 In response to the war, BH created many new municipalities, which increased the total number by almost half and fragmented many into sizes that may be too small for efficient service delivery. However, BH municipalities are still on the larger end of the European average, in line with Netherlands gemeenten (Table 1). But averages disguise size distributions, especially in small or rural municipalities. Citizens of smaller municipalities may benefit from higher levels of government accountability, but the drawback of smaller municipalities may be their inability to attract skilled workers or achieve economies of scale. 12. Poverty is geographically concentrated in some communities. Poverty studies revealed that combined economic dislocation and societal cleavages dramatically reduced the social capital of some communities, and these same communities frequently have a high concentration of IDPs or refugees and returnees. 10 Furthermore the rate of poverty is higher in rural areas than urban, and as of 2004, the RS had a higher poverty rate than FBH, although the gap has been narrowing. Municipalities also vary greatly in terms of physical resources. Certain areas are mountainous with little arable land. Others are heavily forested and rely on exploitation of timber resources, and others have considerable agricultural production Municipalities are critical for public service delivery in Bosnia and Herzegovina 13. Municipalities are the key providers of essential public services in BH. These services include heating, local roads, sewage, solid waste disposal and water; municipalities also deliver social services-culture and communal services, housing, pre-school, and sports; and help facilitate citizens access to health and education services by maintaining primary and secondary school buildings, providing school bus transportation for students and teachers, and delivering physical education, ambulance services and health care supplies. Local self-governance laws empower municipalities in both entities to establish companies, institutions, or other organizations to manage, finance, and improve local infrastructure and services. Municipalities may adopt rules on operational procedures, regulate fees, and take management decisions; they provide administrative services and regulate business, which affects local economic development. 9. A range of other factors determine the optimal size of BH municipalities. Consolidating municipalities may not produce better outcomes. A recent World Bank review suggests that consolidating or merging local governments does not in itself bring about better results in efficiency or service delivery outcomes. The process and context for reforms significantly affect whether and how the reforms offset perceived costs of smaller size and jurisdictional fragmentation. Sometimes consolidation may not be politically viable, in which case cooperative arrangements across several municipalities can address disadvantages of smallness. However, cooperative arrangements may not work well if municipal size disparities are too great or if there are lingering ethnic tensions. For example, the municipalities near Sarajevo are split across entity boundaries, and each has a full municipal administration despite being only a few kilometers apart Poverty Assessment. 4

20 14. Municipal governments exclusively provide local roads, water, sanitation, and waste removal. The municipalities own public utility companies that provide these services. The only exception is Sarajevo Canton, where those services are provided by Cantonal utilities. In most municipalities, a Communal Services Department in the administration oversees public works, and municipal councils set tariff structures. Typically in larger urban areas, a single utility supplies water and sewerage, but in smaller municipalities, these utilities also manage waste, maintain roads, clean streets, remove snow, maintain gardens and parks, and operate funeral services and parking lots. The utilities finance their operations through user fees, but depend heavily on local and higherlevel government co-financing for capital investments. 15. Municipalities help facilitate health and education services. But responsibility for providing health and education rests with higher levels of government. In the RS, for example, compulsory and tertiary education is vested at the entity level, with separate laws covering all four educational branches. In the Federation, however, the cantons have Figure 3: Average Population of Municipalities in Selected European Countries Country Average Population per Great Britain 126,128 Lithuania 66,300 Sweden 30,040 The Netherlands 27,559 Bosnia and Herzegovina 27,042 Denmark 18,760 Belgium 16,960 Poland 15,561 Finland 10,870 Norway 9,000 Italy 7,105 Estonia 5,713 Germany 5,575 Latvia 4,400 Spain 4,930 Austria 3,421 Hungary 3,242 Switzerland 2,468 Czech Republic 1,659 France 1,580 Source: PRISM and Steiner, R (2001) Inter-municipal Cooperation and Municipal Mergers in Switzerland ; Horwath, T. (2000) Local governments in Central and Eastern Europe Decentralization: Experiments and Reforms. primary responsibility. Each of the ten cantons has its own law on preschool, primary, and secondary education, and cantons with universities have laws on higher education. There are no school fees, so education is financed entirely out of public funds, but ministries at various levels finance and manage education. In the RS, the Ministry of Education and Culture plans, executes, and monitors policy, and in the FBH, cantonal ministries of education do this. Municipalities main responsibility is to maintain school buildings, manage water and sanitation facilities, and provide communal sports facilities. 16. Municipalities are entirely responsible for district heating. Despite its importance, district heating has been broadly neglected by the municipalities and the entity-level energy ministries, and is only available primarily to urban residents of multi-family buildings. The FBH operates 25 district heating systems, most of which are in municipalities with populations over 25,000. In both the FBH and the RS, district heating is handled by municipal-owned independent public companies that completely lack regulatory oversight The State of Local Service Delivery Outcomes in Bosnia and Herzegovina 17. Service delivery outcomes vary across municipalities and sectors. The following section will examine municipal characteristics that contribute to these variations-urban and rural; high, medium, and low per capita revenues; population size; and impact from the war (newly created or split municipalities). Second, it will compare sectoral outcomes in water, sanitation, waste, local roads, and education along five dimensions-access, reliability, quality, responsiveness and overall satisfaction, and look at health care, along two dimensions. 18. BH municipalities are quite diverse. Population size, area/density, fiscal capacity, and the legacy of war vary widely. The average municipal population is around 27,000 but ranges from 62 5

21 (Istocni Drvar in the RS) to 224,647 (Banja Luka). Estimated per capita revenues range from 62 KM per capita (Ravno in FBH) to 1,466 (Novo Sarajevo in FBH). The micro municipality of Petrovac (189 inhabitants) reports per capita revenues of 4,825 and Istocni Drvar (62 inhabitants) per capita revenues of 6,291. The smallest municipalities typically result from Dayton accords fragmentation of larger municipalities across the entities. 19. The twenty sample municipalities reflect this diversity. Fifteen municipalities were selected randomly across three per capita municipal revenue strata and according to population shares. Nine municipalities were chosen in the FBH and six municipalities in the RS. All municipalities under 2,500 inhabitants (12 out of 142) have been dropped. The third component selected two municipalities within the city of Sarajevo, which is composed of four municipalities within FBH and three within RS. One municipality was selected randomly for FBH and RS respectively. Finally, the study also selected three of the largest cities outside of Sarajevo, namely Banja Luka (RS), Mostar (FBH), and Tuzla (FBH) Thirty indicators measure outcomes in six sectors. Figure 4 shows the deviation from average of each indicator in eight selected municipalities. Ilidza, the best performing municipality, on the upper left, scores better than average for 27 indicators; Berkovici, on the right, can beat average outcome only in six. Indicators for each of the five dimensions were selected individually according to each sector. Access to local roads means an asphalted road that connects the household with other parts of the municipality. Access to water means running water delivered by public pipe to inside the household; access to sewage means connection to a septic tank. Waste access refers to households that have nearby garbage dumpsters or regular garbage collection. As for education, elementary school enrollment ratio serves as proxy. For health it is access to a clinic or medical office. Regular services (road cleaning, more than 21 hours of water, no need for maintenance, and physical conditions of schools) measure reliability and the level of meeting citizens demands or absence of problems represent the quality of services. Responsiveness reflects citizen s trust in public institutions, i.e. the capability of formal institutions (service providers, municipality or neighborhood council) to solve problems with service delivery. Citizen satisfaction means their level of overall satisfaction. Box 1: Municipalities in Bosnia and Herzegovina: Legacies of Conflict Half of the 20 municipalities that comprise the research sample for this study are divided or newly established. The sample includes the range of patterns by which municipalities were reshaped. Some were parent municipalities, some were divided and others were MZs of parent municipalities. Stolac and Berkovici reflect both: Stolac was the multiethnic parent municipality of which Berkovici was an MZ. Stolac is now Bosnian-Croat majority with a Bosniak returnee population and Berkovici is majority Bosnian-Serb. One of the MZs of Ilidza became a municipality in the RS. Zvornik and [amac also lost MZs. The municipalities of Odzak, Kalesija, and Istocni Stari Grad were divided. In Odzak, the town center became the municipality. Istocni Stari Grad was created out of Stari Grad and consists of geographically dispersed MZs. Two of the smallest municipalities were MZs of another municipality that was not in the sample. Doboj-Jug was a MZ of Doboj, and Ribnik of Kljuc. Some municipalities, although not formally split by the DPA, functioned during the war or immediate postwar period as divided municipalities with separate administrations (e.g. Jablanica). Most new municipal boundaries reflect changes in ethnic composition. Municipalities were divided between the FBH and RS, and MZs with ethnic minorities became new municipalities. For example, Zvornik was majority Bosniak and is now majority Bosnian-Serb. Pre-war Doboj was approximately 55 percent Bosniak, the Bosnian-Croat population joined the municipality of Usora, and now the municipality is nearly all Bosniak. Some municipalities that maintained pre-war boundaries changed ethnic composition. Some municipalities that maintained pre-war boundaries changed ethnic composition. Eight municipalities have maintained the same borders-banja Luka, Breza, Jablanica, Ljubinje, Novo Sarajevo, Š[iroki Brijeg, Tuzla, Visegrad. Of these, Visegrad was Bosniak and Bosnian-Serb prior to the war and is now majority Bosnian-Serb. The Bosnian-Serb majority in Banja Luka increased from approximately half the population to over 90 percent. Mostar was initially divided into six municipalities (three Bosniak and three Bosnian-Croat) after the war, but subsequently OHR cancelled this division and established Mostar as a city with six areas. The OHR statute providing for the administrative unification has not been adopted by the city council. 11. For details on the Municipal Sample Survey Selection see Annex 1.1 6

22 Figure 4: Service outcomes vary widely across municipalities Percentage Point Deviation from Percentage Point Deviation from Average Sector Indicators ILIDZA (1) Average Sector Indicators BERKOVICI (20) 50% 50% 40% 40% 30% 30% 20% 20% 10% 10% 0% % % -10% -20% -20% -30% -30% -40% -40% -50% -50% -60% -60% -70% -70% Percentage Point Deviation from Average Sector Indicators 50% TUZLA (2) Percentage Point Deviation from Average Sector Indicators 50% STOLAC (19) 40% 40% 30% 30% 20% 20% 10% 10% 0% % % -10% -20% -20% -30% -30% -40% -40% -50% -50% -60% -60% -70% -70% Percentage Point Deviation from Average Sector Indicators 50% BANLA LUKA (3) Percentage Point Deviation from Average Sector Indicators 50% MOSTAR (18) 40% 40% 30% 30% 20% 20% 10% 10% 0% % % -10% -20% -20% -30% -30% -40% -40% -50% -50% -60% -60% -70% -70% Percentage Point Deviation from Average Sector Indicators 50% NOVO SARAJEVO (4) Percentage Point Deviation from Average Sector Indicators 50% KALESIJA (17) 40% 40% 30% 30% 20% 20% 10% 10% 0% % % -10% -20% -20% -30% -30% -40% -40% -50% -50% -60% -60% -70% -70% 7

23 1.3.1 Variations across municipalities 21. Service delivery outcomes vary significantly among municipalities. Using results from the household survey, a summary outcome indicator was generated that measures service outcomes for each municipality, comparing deviations from average sector outcomes in each of the 20 sample municipalities. Municipalities can then be ranked according to their scoring for each sector; though attempts to generate an overall performance ranking of municipalities should be approached with caution for several reasons (see Annex 1.2). The ranking in Figure 5 is based on results for all 30 indicators, the municipal values, the averages of those 20 municipalities and percentage points that the municipal outcome is above or below the score. An above-average municipality is assigned a leading score, and a municipality can score up to 30 if it is above average on all scores. Figure 5: Service Delivery Outcome Score 22. Higher per capita revenues alone cannot explain better performance. Municipalities with higher incomes have consistently better service delivery outcomes, but many low and middle income municipalities can deliver above-average services. Samac and Odzak have low-to-medium municipal revenues (KM 192 to KM 297) but both achieved high scores-samac is among the top five performers. However, none of the top three municipalities generates revenues much below KM 500 per capita. All five municipalities with per capita revenues above KM 600 perform significantly better than average, except Mostar. Poor municipalities, on the other hand, cannot entirely overcome the effects of low income to perform well. None of the three poorest municipalities (revenues below KM 160 per capita) attains above-average outcome levels for more than 12 of the 30 indicators. They represent half of the six municipalities on the bottom line of service outcomes. Outliers in this subgroup are Berkovici and Mostar, which markedly underperform, given their high incomes. Mostar is still struggling with the consequences of war, which inflicted severe infrastructure damage, and Berkovici is a newly created municipality (1995), which was subdivided from the pre-war municipality of Stolac. Stolac s service outcomes are next lowest to Berkovici, but Stolac also has much lower per capita revenues Stolac s and Berkovici s size and total revenues are comparable, but Stolac has ca. 13,000 inhabitants and Berkovici has only ca. 3,000. 8

24 Figure 6: Income alone does not explain better outcomes 23. Smaller municipalities can perform well despite higher fixed-cost ratios. Of the six sample municipalities with populations of around 50,000, four are among the top five performers. 13 However, below this threshold no clear pattern emerges. Doboj-Jug, a municipality with roughly 4,500 inhabitants scores very well and leads in 21 indicators. Kalesija, on the other hand, has a population of almost 35,000, but does only slightly better than the bottom three. The same is true for population density. The most densely populated municipalities achieve the highest service outcomes, but low density does not correlate with poor service outcomes. Municipal income appears to be a stronger determinant of performance outcomes-among the top ten performers, only Samac generates less than KM 200 per capita, whereas among the bottom ten performers, almost half belongs to the lower revenue tercile. 13. Mostar and Zvornik are not among them. 9

25 Figure 7: Small municipalities can achieve excellent service outcomes Figure 8: High population density helps, but not a condition for good service outcomes 24. Generally, DPA-created municipalities do not perform worse. Among the new municipalities, the level of per capita revenues has more significant impact on service outcomes than for their parent municipalities. Although new municipalities are also generally smaller, it is their limited experience rather than their limited size that inhibits performance outcomes compared to those of their more established peers. However, since each newly created municipality began with unique conditions, classifying and evaluating them as a group is misleading because some municipalities inherited functioning administrations and service providers and others started from zero. For example, Š[amac collects KM 192 per capita and achieves a high service outcome score of 22. Istocni Stari Grad, on the other hand, has per capita revenues of KM 203, but has low service outcome scores that place it in the bottom five municipalities. The difference between [amac and Istocni Stari Grad is that [amac was newly created through the loss of several MZs, but it kept functioning as an integrated municipality, whereas Istocni Stari Grad consists of geographically dispersed MZs-some of which are 50 kilometers apart with no direct roads connecting them. 10

26 Figure 9: Service outcomes vary across old and new municipalities Figure 10: but are more income-sensitive in newer municipalities Variations across sectors 25. Almost 1.4 million BH citizens have no running water or waste removal. More than a decade after the DPA, access to services is disappointing and access to critical basic services such as water supply and waste collection is unacceptably low. About one million citizens cannot reach other parts of their village or town via an asphalted road and half a million people lack any sanitation service. In a country with four million inhabitants there are some 130 municipal water companies and a comparable number of community service enterprises for waste collection and other services. However, despite this vast number of providers and remarkable levels of investment, only 65 percent of households receive their water from the public system. Almost 1.4 million people in Bosnia and Herzegovina don t have a running water connection in their household and lack regular waste removal services. Rural households fare even worse-one-third of rural households must use water from unsafe sources such as rivers, streams, springs, or wells because they lack municipal water connections. The rapidly growing volume of solid waste outstrips municipal utilities ability to keep pace. Every year BH citizens generate two to three million of tons of solid waste but only 36 percent of households have regular waste removal services. As a result, people burn or dump their waste illegally, polluting land, roadsides, and rivers. 11

27 Figure 11: Access to basic services is low and people are not satisfied 26 Overall satisfaction rates with services are low. Even the two highest-rated sectors-water and health-achieve service levels that satisfy only about 60 percent of the population. Less than half of the sample households were satisfied with waste collection, local roads, and sewage, and only one-third with the education system, but these results contrast sharply with ratings for reliability, quality, and responsiveness. 27. Health and education services perform poorly. Despite broad access, these two sectors are a major concern for citizens. However, municipalities have only minor responsibilities in these sectors, so their potential contributions to improvements are limited. Furthermore, from the vantage point of this study, evaluating quality is difficult for education and nearly impossible for health. 14 Two categories (physical and non-physical criteria) enable us to distinguish between municipal and higher-level government responsibilities. Education scores low on non-physical quality criteria (ca. 30 percent unsatisfied) that are mainly predetermined by cantonal (FBH) or entity-level (RS) authorities and out of reach of local influence. Nevertheless, municipalities are responsible for operation and maintenance of school buildings, therefore they can improve their contribution because only 74 percent of citizens were satisfied with the condition of the buildings, 70 percent with heating services, and only 62 percent were satisfied with sanitation facilities. Health services are rated higher than education, but dissatisfaction still appears widespread, results that correspond with the last Public Expenditure and Institutional Review (PEIR 2006), which found that the health sector was underperforming for the level of sectoral spending. Faced with low-quality public health care, citizens must seek private service providers, who are not covered by public health insurance. Nevertheless, private providers are a growing share of health provision in BH and add to citizens cost. 14 The survey dropped specific questions on health outcomes except for overall satisfaction, but asked about satisfaction levels with the quality of education. Results must be viewed with caution but citizen perceptions offer a general idea. 12

28 Figure 12: Room for improvement, particularly for water and waste 28. Water and waste have vast room for improvement. Among the four services provided exclusively at the municipal level-waste, water, roads, and sanitation-waste scores poorly along all five outcome dimensions (Figure 13) and no other service fails citizens quality and reliability expectations to such a great extent. Water supply, although needing significant improvement, appears to achieve more balanced outcomes across the five dimensions. Local roads score average on access and high on quality, but fall short on reliability due to poor road maintenance and inadequate winter street cleaning. The sewage system has surprisingly good ratings, especially compared to water supply, but this is likely because most BH households operate their own septic tanks and are not connected to public sewerage, so it is predominantly urban residents of multi-family buildings that use the public sanitation system. However rural sewerage connection rates are low-one out of five households relies on latrines or outhouses. 13

29 Figure 13: Quality ratings need balanced interpretation 29. Local roads-quality is perceived by local communities as good but reliability and access need improvement. Three-quarters of households have a paved road to connect them to other parts of the village or town because most roads are in good shape. 15 However, when municipal providers fall behind with winter snow removal there are serious consequences for almost half of affected households, particularly in rural areas where up to 40 percent have insecure access to food supply or medical treatment. 15. Note that the actual condition of the network is not as promising compared with the perceptions of the communities: A survey undertaken for the Secondary and Local Road Study in Albania shows that about 22% of the local network is in poor or very poor condition, 24% in fair condition, and only half the network in good condition. 14

30 Figure 14: Citizens face harsh consequences if roads are inaccessible during winter 30. Water coverage and quality is a problem for many households. Respondents pointed out that water coverage rates are low, and among households connected to the public water system, most experience frequent interruptions and problems with water supply. Agricultural producers said that summer water shortages constrained further development (particularly in the lower watershed). 16 In the middle watershed, some residents were concerned about water safety due to past disease outbreaks and others noted that cisterns were not emptied regularly, risking polluted water supplies. Reports from lower watersheds indicated that tapping into water pipes is common enough that water pressure is lower than levels agreed upon between municipalities. 16. Prism Research and Bosnia S., Forest and Mountain Protected Areas Project: Social Assessment, unpublished draft, Sarajevo: March 2006; Golder Associates and Bosnia S. Integrated Ecosystem Management of the Neretva and Trebisnjic River Basin Project: Study 4 Social and Rural Development Assessment, unpublished draft, Sarajevo: April

31 Figure 15: Almost half of households with water connection experience problems Respondents with access to public water system in their household, n=1325 (Number of missing cases: 0) Poor quality Physical conditions of pipes Unexpected interruptions No problems ALL RESPONDENTS 26% 19% 8% 57% Settlement Type Rural (429) Urban (896) 26% 27% 19% 19% 7% 8% 59% 56% Entity FBiH (884) RS (441) 33% 13% 26% 15% 7% 8% 57% 59% Population 2,000-12,500 (388) 12,501-25,000 (349) 25,001 and more (588) 18% 32% 29% 20% 3% 29% 12% 7% 62% 15% 57% 53% PC Municipal Revenue KM (239) KM (428) 301 KM and more (658) 40% 22% 24% 13% 10% 6% 33% 11% 7% 62% 49% 56% Type of the dwelling Multifamily residential building (341) Individual house (952) 30% 25% 21% 18% 7% 9% 53% 60% 31. Waste collection coverage is inadequate but service is good in high-income municipalities. Citizens suggested improvements to waste management services-install more bins, increase the frequency of pick-up, and maintain a regular schedule-appear to relate more to poorer municipalities that receive irregular services. Roughly one-third of households in poorer municipalities say that their garbage is not picked up frequently enough. Hence, poor quality in the sector is primarily linked to low coverage rates, which needs urgent attention from municipal authorities. On the other hand, citizens who have access to waste collection report good service-92 percent have service once a week or more, and almost 80 percent agree that collection schedules meet their needs. However, declining municipal revenues depress quality rating dramatically and the sector as a whole fails markedly because some 40 percent of citizens lack coverage, so these results appear to correlate strongly with access. 16

32 Figure 16: Higher-income municipalities provide better waste collection Respondents using waste collection service, n=1191 (Number of missing cases: 153) More often than once a week Once a week Less often ALL RESPONDENTS 40% 52% 7% Settlement Type Rural (395) Urban (796) 25% 48% 63% 47% 12% 5% Entity FBiH (785) RS (406) 45% 32% 48% 61% 7% 7% Population 2,000-12,500 (291) 12,501-25,000 (334) 25,001 and more (566) 12% 27% 64% 80% 69% 28% 4% 8% 8% PC Municipal Revenue KM (234) KM (356) 301 KM and more (601) 15% 28% 58% 65% 66% 39% 20% 6% 3% Figure 17: that meets household needs 17

33 Figure 18: Quality and reliability decline with lower per capita municipal revenue Quality of Waste Collection and Per Capita Revenue Reliability of Waste Collection and Per Capita Revenue 100% 100% 90% 80% R 2 = % 80% R 2 = % 70% 60% 60% 50% 50% 40% 40% 30% 30% 20% 20% 10% 10% 0% Per capita revenue (KM 2005) 0% Per capita revenue (KM 2005) 32. Sanitation remains problematic for rural municipalities. High quality ratings for sewage are surprising, particularly when compared to the low ratings of the water network. However, many households in BH operate their own septic tanks and public sanitation systems predominate only in urban households and multi-family buildings. Only 19 percent of rural populations are connected to public sewerage systems. One in five rural households relies on latrines or outhouses, and among those with a septic tank or connection to public sewers, some 21 percent have problems with the sewage system, most commonly obstructed pipes, stench, or complete system failure. Similar to waste collection, quality ratings diminish rapidly with declining municipal revenues and access to services. Figure 19: Most households operate their own septic tanks All respondents, n=1976 Number of missing cases: 21 Public sewers Septic tank Latrine/outhouse only Other ALL RESPONDENTS 42% 45% 9% 4% Settlement Type Rural (958) Urban (1018) 19% 64% 62% 29% 17% 2% 2% 5% FBiH (1191) 50% 43% 5% 2% Entity RS (785) 31% 48% 15% 6% Population 2,000-12,500 (588) 12,501-25,000 (597) 25,001 and more (791) 19% 45% 58% 62% 43% 34% 13% 6% 7% 5% 7% 1% PC Municipal Revenue KM (498) KM (591) 301 KM and more (887) 27% 46% 49% 59% 42% 39% 8% 6% 8% 4% 10% 2% Type of the dwelling Multifamily residential building (359) Individual house (1572) 31% 93% 54% 6% 1% 0% 11% 4% 18

34 Figure 20: Predominantly rural households face problems with their sewage system Respondents with septic tank or connection to public sewage network, n=1681 Number of missing cases: 43 ALL RESPONDENTS 0% 20% 16% Settlement Type Entity Rural (752) Urban (929) FBiH (1068) RS (613) 13% 17% 15% 21% Population PC Municipal Revenue 2,000-12,500 (473) 12,501-25,000 (495) 25,001 and more (713) KM (413) KM (502) 301 KM and more (766) 13% 19% 19% 15% 16% 18% Type of the dwelling Multifamily residential building (347) Individual house (1301) 11% 18% 33. Lack of responsiveness to users creates low rates of satisfaction. User ratings will continue to be low unless municipalities improve responsiveness and accountability among service providers. Surveyed citizens report poor problem-solving capacity among formal institutions-only about six percent of client complaints ever get solved when clients report problems to the service provider, municipality, or neighborhood council (MZ). This creates a downward spiral-few problems get solved so fewer citizens report problems and therefore few providers ever get any feedback about their failures. Among people surveyed, about two-thirds (67 percent) claimed that if they had problems with a service they would contact the institutional service provider, but in reality only 20 percent actually report problems (Figure 21), which means that providers lack the information they need to improve services. Figure 21: Feedback to providers would improve service quality. Feedback loops and problem solving capacity Road Waste Water Sewage Clients with problems that complained Cases where problem has been solved 19

35 Box 2: Sarajevo: Cooperation brings benefits to citizens Although cooperation requires extra coordination among stakeholders, experience from Sarajevo Kanton proves that economies of scale in the public sector translate into better service outcomes and higher client satisfaction. Sarajevo Kanton, the most populous canton in the FBH, comprises nine municipalities with around 400,000 inhabitants. It spans 1,277 sq. km and includes the capital of Bosnia and Herzegovina. As early as the mid eighties, municipal officials realized that serving a broader area required a higher degree of coordination. But the results were mixed and all too often, efforts failed. In 1984, Sarajevo hosted the Winter Olympics, and in the same year, the eleven independent utility companies could not agree on an integrated cleaning and waste collection plan. 17 That motivated Sarajevo city assembly to take a decision that would ensure efficient and reliable services throughout the capital regionthey founded the public utility RAD, a single company for waste collection and disposal, road maintenance, public parking management, and winter services. Today, RAD functions as an integrated service company under the Sarajevo Kanton and contracted by the municipalities. RAD finances its operations mainly from user fees for waste collection and disposal, traffic area reconstruction, and parking lot maintenance. Cleaning public areas, winter services, and other shared communal expenditures are financed from the cantonal budget, which also funds capital investments. Providing services for almost half a million citizens is still a challenge but RAD has adopted an organizational structure that combines functional and spatial decentralization that enables it to handle requests from different stakeholders. Under the umbrella of the cantonal company, regional teams in each spatial unit develop and implement operational plans in cooperation with each municipality, which creates smooth and efficient coordination since negotiations don t take place among eleven independent companies. Integrated plans and bundled operational knowledge boost performance, and create economies of scale. RAD s combined resources give it the leverage to exceed regular service delivery levels in BH, and its size enables it to assemble investment projects with larger-scale capital expenditure that would otherwise not be feasible. For example, the Sarajevo solid waste landfill is a precedent in BH and meets international operational standards. Only a few years ago, frequent incidents and contaminated soils threatened the sustainability of the landfill. Today, a water processing facility allows redirection of liquid filtrates into the Miljacka River without pollution. A self-sustaining recycling facility, the first of its kind in BH, functions at EU-level standards and will eventually turn a profit. Landfill gas emissions used to cause frequent incidents, but now a methanol energy production facility generates 0.3 Mega-Watt and has the potential to produce up to 1.5 MW Key findings: What do service delivery outcomes in BH tell us? 34. More systematic capital investment is needed. Municipalities need more capital investment to expand access to basic services, especially in rural areas and among municipalities with low per capita revenues where coverage is low. To expand water and sewage networks, build new roads, pave existing roads, and upgrade sanitation facilities in schools, municipalities will need significant levels of investment funding that cannot be generated at the local level alone. However, increased investment must be combined with systematic sectoral approaches. For example, water has poor quality ratings even among consumers who are connected to the system-water supply shortages, frequent interruptions, and low pressure. These problems with the network are directly related to lack of investment funding. Even for high per capita municipalities, the level of investment required for complete rehabilitation of the system is impossible to cover by municipalities alone. But in less capital-intensive services such as waste collection, improvements in quality and coverage can be achieved by a strategic sectoral approach, better performance and raising municipal per capita revenues. Consumer satisfaction increases significantly with increased coverage, which is within the scope of local authorities. However, other waste management activities, such as developing regional landfills, will need coordinated investments and additional sources of finance. 35. Municipalities with a sector strategy and clear priorities do better. Conditions for better performance differ among municipalities, but even at given levels of income, municipalities with an established game plan can make a big difference. Certain levels of income are required for municipalities to perform better but increased revenue alone will not lead to improved outcomes for citi- 17. At the time eleven municipalities. Today, Lukavica and Kasindol are part of the RS entity City of East Sarajevo. 20

36 zens. Among the many middle-income municipalities, there is a remarkable group of strong performers to follow suit. But their challenges differ from each other so that a performance improvement blueprint is not at hand. Ljubinje has made tremendous efforts to increase own-source revenues by promoting local economic development. Jablanica successfully reformed its internal organization and achieves outstanding efficiency gains from an improved municipal administration (see boxes). In the Canton Sarajevo, integrated utilities deliver better services to citizens in nine municipalities and the City of Sarajevo. Along with Banja Luka, they have integrated strategies for waste and water, and perform better than big municipalities that lack strategies. Often, low outcomes result from extensive fragmentation across municipalities and levels of government. As the separation of Berkovici as new municipality from pre-war Stolac illustrates, several newer municipalities had to start from zero to build an effective administrations and develop service providers. Many small-scale utilities bear high units costs and cannot realize economies of scale. 36. Responsibilities overlap and accountability is weak. Horizontally and vertically overlapping responsibilities result in unsystematic, fragmented investments, underfunded services and weak accountability to citizens. Service providers rely mainly on municipalities to mobilize capital investment funds, often from higher levels of government. Management objectives focus on operation and maintenance, but when tariffs are below production costs and collection rates are low, revenues barely cover recurrent costs. Local authorities are responsible for setting the tariff structure, but municipal councils hesitate to pass unpopular fee increases, so assets decline and service quality suffers. Dissatisfied clients search in vain for someone to hold accountable for poor services because responsibilities are unclear. Some 28 percent of respondents said they have no idea who to call when they experience problems with public services, which erodes their confidence in public utilities and service providers. Many households haven t had good experiences when lodging complaints, therefore fewer people complain because they think it is a waste of time. These accountability gaps translate into higher costs since providers often lack adequate information on quality shortcomings and quality shortcomings reduce clients willingness to pay. 21

37 Box 3: The BEACON Scheme: learning from performing municipalities Municipalities in BH are in a permanent cycle of incremental improvements and steady reforms to enhance local capacity and provide more efficient, effective, and responsive services to citizens. Innovations take place at many levels and within a broad range of themes but few of these initiatives were broadly recognized, scaled up, or applied elsewhere. Here is where the BEACON Scheme comes in. It was designed to raise standards in municipal governance by recognizing and rewarding excellence and providing a platform for municipalities to systematically seek and spread innovative practices. After extensive consultations with municipalities, the BEACON Advisory Panel selects three themes that reflect municipal priorities. Municipalities then can apply to the scheme, which is the first learning opportunity in the process, since they have to reexamine internal practices and policies. Short-listed applicants get assessed, must give a presentation, and host a verification visit to evaluate whether they demonstrate excellence against the criteria developed for each theme, and the willingness to share their best practices. This process offers municipalities insights into how their current performance compares to others. They receive constructive feedback and recommendations. Excellent municipalities are awarded BEACON status. In the first round of the 2006 Beacon Scheme in Bosnia and Herzegovina, six municipalities won awards. These municipalities now receive widespread recognition and their achievements are more valued by their own population. Others are eager to learn from this experience. Representatives from over 40 municipalities throughout Bosnia and Herzegovina have attended the Open Days, during which the award-winning six municipalities shared their knowledge and provided an opportunity for exchange. All municipalities collaborated on developing sustainable networks of reciprocal learning and create opportunities to enhance future cooperation. The first two years of the BEACON competitive reward system in BH encouraged reforms, built up an array of approaches to change, and supported peers looking for ways to leverage their own efforts. The scheme was launched in September 2005 as a joint initiative of the OSCE Mission to Bosnia and Herzegovina, the Council of Europe and the FBH and RS Associations of Towns and Municipalities. In August 2007, BH has congratulated its award-winners of the second round The new challenge ahead: improvement of local service delivery performance 37. All municipalities suffered the effects of conflict. Infrastructure destroyed during the war included water supply, electricity, sewerage, public buildings, and roads. The 1996 donor conference estimated that more than half of municipalities had damaged infrastructure. Housing stock was destroyed, and waves of refugees and IDP moved through municipalities during and after the war. Nearly half of all households were split and family members emigrated, moved to other Yugoslav successor states or to the other entity Local governments had to cope directly with the consequences of the conflict. Municipalities have faced the particular challenge of reintegrating returnees and internally displaced persons. At the end of the war, there were an estimated 2.1 million refugees and internally displaced persons (IDP) and today, these two groups account for a high proportion of municipal populations in both Entities-somewhere between 12 and 46 percent in the Federation and 22 to 39 percent in the RS. 19 Since the war, rates of population return to the municipalities have varied. In most cases the relative numbers of returnees are proportionate to municipal population size, but some have absorbed greater numbers of returnees. In any event, local authorities are in the front line of overcoming ethnic divisions, facilitating cooperation between conflicting interests, and rebuilding trust. Reintegration can rarely be achieved by higher levels of government, since they lack the proximity that local governments still have. 18. See Bosnia Social Assessment, UNHCR estimates that half of the population was internally displaced or refugees at some point during the conflict. 19. World Bank (2002), Bosnia and Herzegovina: From Aid Dependency to Fiscal Self-Reliance, A Public Expenditure and Institutional Review 22

38 39. Municipalities have achieved to deliver stability at the local level. Recently the pace of return has slowed and it appears that many people who intended to return to their municipality of origin have done so. 20 In the early years, returnees experienced threats to personal security. 21 Refugee return began to increase dramatically in 2000, after four and half years of international reconstruction efforts. 22 During , personal security concerns were largely supplanted by other considerations for returnees, such as progress with reconstruction and potential for employment.23 In 2007, most household survey respondents considered their municipality a safe place to live, even in a ethnically divided municipality such as Mostar, almost half of respondents (46 percent) considered the municipality safe and less than a quarter (23 percent) considered it unsafe The municipal challenge of today is to deliver better services. Today, more than a decade after the DPA, citizens are less concerned about stability and more concerned about getting better municipal services for their money and asking municipalities to promote a business environment that fosters economic growth and generates employment. Local governments and service providers will have to meet this demand in the face of dwindling donor funding. 41. Better municipal performance is crucial to meet citizens demands. In both entities, the important role of municipalities has been recognized and new laws emphasize local self-governance and independent competencies for providing basic public services. Local election held in 2004 and the recent VAT reform have strengthened municipalities role in facilitating human well being and economic growth. Many donors, including the World Bank, have carried out extensive investment and capacity building programs to support municipal development, ranging from utility management reforms, budgeting, planning, social inclusion, and accountability. However, despite the substantial international support for municipal-level capacity building, little information exists on municipal service delivery performance. 42. Improving performance requires adequate data to understand service outcomes across municipalities and sectors. However, until this study, there has not been systematic collection and dissemination of municipal performance data. The study present results based on a survey of 2000 households in 20 municipalities, and on an investigation of service delivery outcomes in ten sectorseducation, electricity, health care and housing, for which responsibility is shared or under higher level government; and heating, local roads, sewage, waste management, transport, and water, for which municipalities are responsible. 20. A 1998 survey found that displaced people said fear for their personal security was the main obstacle to returning and most preferred to return as a group. See Bosnia Social Assessment, Attacks on returnees included grenade attacks, mines planted close to returnee homes, and threatening actions such as destroying a bakery patronized by returnees. See UNHCR, Return Monitoring Survey: Minority Returnees to Republika Srpska, June See ICG, Bosnia s Refugee Logjam Breaks: Is the International Community Ready? May During , sustained efforts to ensure personal security, to enforce rule of law, and to guarantee personal property developed a virtous circle. 24. In Odzak, Jablanica, Mostar, Stolac, Berkovici, and Novo Sarajevo, less than half of respondents said their municipality is safe; but more respondents said neither safe nor unsafe, or refused to answer. 23

39

40 Chapter 2: Bosnia and Herzegovina s Decentralization Framework 1. The prevailing decentralization framework, shaped by Bosnia and Herzegovina s postconflict legacy, shapes how individual municipalities perform in managing public resources and delivering public services.25 The past years have also seen significant changes to laws, regulations, and ultimately, financing, that can contribute to the transition from municipal stability to stability plus performance. This framework chapter first identifies some key entry points to promote indepth municipal performance, then sets out key aspects of the enabling environment for municipal performance for managing public resources and delivering services. The chapter identifies formal and informal responsibilities for delivering municipal services and how they are financed from higher and local-level government sources The Evolving Decentralization Framework 2. Local performance depends on the intergovernmental enabling environment. Municipalities have important roles in basic service delivery, underpinned by good public financial management, but concerns have been raised about the intergovernmental fiscal, administrative, and political system (see World Bank 2006, Public Expenditure Review). Sometimes these roles and responsibilities are unclear, financing does not correspond to responsibilities (including transfers and own-source revenue assignments), municipal capacity is low, and the entity-cantonalmunicipal relations in FBH, and the bottom-up and top-down accountability for service delivery are overly complex. 3. Figure 22 summarizes the main sectors for which municipalities have significant roles and responsibilities, including financing; and Annex 1 describes expenditure assignments. While many of these services are typically decentralized, some expenditure assignments merit particular comment because they differ across entities. For example, no RS municipalities appear to have responsibilities for primary education, while some FBH municipalities do. Municipalities must fund expenditure responsibilities from their own revenues and intergovernmental transfers (e.g., shared revenues, formula-based transfers, or other support). In some cases, if higher or lower levels are not providing key services demanded by local citizens, municipalities have had to step in with their own resources for these unfunded mandates. 25. William Fox and Christine Wallich, Fiscal Federalism in Bosnia and Herzegovina: Subsidiarity and Solidarity in a Three Nation State, In Robert Eberl and Richard Bird, (2007), Fiscal Fragmentation in Decentralized Countries: Subsidiarity, Solidarity and Asymmetry, Edward Elgar Publishers. 25

41 Figure 22: Major Municipal Expenditure Assignments by Entity Local Utility/infrastructure FBH Water & Sewage Solid Waste Local Roads & Public Transportation (canton regional roads) Heating 26 RS Water & Sewage Solid Waste Local Roads & Public Transportation (entity: regional and trunk roads) Heating Education Pre-school (some Cantons took over this responsibility) Primary (municipalities in three Cantons are responsibility for maintenance and capital investments; all Cantons are responsible for teachers) Pre-school Secondary (capital improvement only) Health Administrative Services Social Welfare Ambulance Services Birth & Death Certificates Building Permits/cadastre services Business licensing (partially) Culture, sport, leisure Communal Inspections Social welfare (jointly with Cantons) Housing Shared responsibility with cantons, as per cantonal and BH laws Birth & Death Certificates Building Permits Business licensing Culture, sport, leisure Communal Inspections Social welfare Housing Municipalities finance some programs 4. Clarity and transparency in assigning roles and responsibilities is crucial to promoting performance. Without clear roles and responsibilities, municipal or other levels of government may have functional gaps, or overlaps and duplication that waste resources and effort. Ultimately, accountability is undermined if functions and the corresponding capacity and financing to deliver them are unclear. This refers to accountability directly to citizens or to higher levels of government. Similarly, excessively complex assignments can undermine sectoral service delivery, including investments, maintenance, and ongoing operations. 5. Reforms seek to strengthen the local service delivery-enabling environment. Recent legal changes have stabilized revenues and municipal government status. First, for revenues, since 2006, municipal governments have a more stable source of revenue from Indirect Taxation Authority (ITA)- collected state-level taxes. Second, in the political arena, since 2004, mayors have been directly elected by municipal citizens, which promise to increase mayoral accountability and stature with higher levels of government. 27 Both FBH and RS have adopted Local Government laws that clarify functional assignments to municipalities and their revenue sources, and in line with the European Charter on Local Self Governance, strengthen the role of local communities (MZ), and provide a legal framework for the Association of Cities and Municipalities to participate in legislative and policy-making processes in FBH. 28 However, many of the ten cantonal laws have yet to be aligned to these new laws. 6. Local self-governance laws are a major step forward. Two major developments in FBH local self-governance are that a recent federal law has laid out a set of principles to harmonize and clarify a myriad of self-governance laws in the ten cantons, and strengthens the role of Associations of Cities and Municipalities. Although the new law expands assigned municipal competencies, most cantonal laws regulating municipal competencies already contained these provisions. Expanded competencies include the explicit right to set local taxes and fees, to organize referenda, and to incur debt. The Associations of Cities and Municipalities now play a greater role in developing policy and legislation by submitting proposals and amendments. 29 The local community (mjesna zajednica-mz) is strengthened, and now citizens must participate in local self-government through referenda and citizens meetings. In RS, the new law also contains an exhaustive list of competencies at the local government level. 26. The big exception here is Kanto Sarajevo, where integrated Cantonal utilities provide water, sewage, solid waste and heating services. 27. In April 2004, the Law on Direct Election of Mayors was adopted. 28. In FBH, these include the Law on the Principle of Local Self-Government (Official Gazette 49/06), and new Law on Distribution of Public Revenues in FBH (Official Gazette 22/06). In RS, this includes the Law on the Local Self-Government (Official Gazette 101/04) and Amendments to the Law on Budgetary System of RS (Official Gazette 34/06). 29. Most interviewed officials stated that the role of the Association was especially positive and visible in drafting laws on local self-governance and regulations on budget transfers and relations between government levels. Smaller communities see the Association as a useful citizen advocacy mechanism. 26

42 7. In FBH the local government law strengthens mechanisms for local accountability. The law provides for local level mjesna zajednica (MZ), referenda, citizens meetings, and citizens initiatives, but organization of the MZ is not obligatory. The laws provide for transparency and public scrutiny of the work of local self-government units, and regulate administrative supervision of unit work, in accordance with European Charter on Local Self-Government. Cantons must adjust their laws to the new Federal Law within six months and municipalities will have to adjust their local statutes within one year. The effects of the new legislation are as yet unclear as implementation has just begun. Still, it appears that the new Law in FBH has improved the status of municipalities as the authority closest to citizens, but stops short of assigning significant competencies to local selfgovernance. 8. In both entities, the new laws provide for cooperation, supervision, and monitoring among higher-level authorities and local self-government. In FBH for the first time federal and cantonal authorities are obliged to consider all proposals, suggestions, and initiatives that come from the local authorities and respond in writing within 30 days. The process of supervision and monitoring of local self-government units, as prescribed in the new Laws, fully respects the principle from European Charter on Local Self-Government. 9. Both Laws regulate in detail local authorities scope of influence to develop local economic, social, and urban plans. Cities and municipalities have exclusive competencies to develop plans and programs for the development of local self government units, and to create conditions for economic development. Cities and municipalities are responsible for preparing urban and spatial plans, organizing the work of local administration to suit local conditions, or taking any other measure to improve local quality of life. 10. The new laws have institutionalized important aspects municipal government. Beginning in July 2006, the new law in FBH clarified existing municipal responsibilities but assign no additional functions. The new law details the scope of self-governance and relations between council and mayor. The role and financing of MZs are strengthened, including their role in the municipal affairs. Two categories of municipal and city revenues are clearly defined-own source and shared revenues, stipulating that no additional responsibilities can be transferred without proportional revenue transfers to finance them. What stands out about this law is the strong role of the Association of Cities and Municipalities in FBH, and their newly established position in the intergovernmental processes. Although in FBH an important challenge ahead will be harmonizing cantonal and federal law. 11. Constitutional reform represents ongoing political challenges in both entities. For example, the Constitutional Reforms in FBH demand that in municipalities with a significant majority (over 20 percent) of either ethnicity, the mayor and speaker of council must originate from a different ethnicity. Recently, no attempts have been made to change functional assignments across levels of government (including cantons) in reform efforts. The FBH Law on Revenue Allocation was subjected to constitutional challenge on the grounds that it infringed on cantonal interests but this challenge was not sustained. Overall however, the dialogue concerning constitutional change has taken place at an aggregated political level, and ultimately has been difficult to reach consensus for specific changes across political parties and groupings. 12. BH state level lacks a coherent strategy to follow-up reform local self-governance. The Partnership Group for Development of Local Governance in BH ( Partnerska grupa za razvoj lokalne uprave u BH )-a working group supported by the Associations of Cities and Municipalities of FBH and RS-prepared a Strategy for Development of Local Governance in BH, which has yet to make the agenda of relevant BH institutions. Such a strategy would need to be adopted by the two entities; and the Public Administration Reform Unit, a state-level agency charged with strategy and oversight of public administration reform in BH, would be in a position to track its implementation. 13. Donors influenced the political process of passing recent reforms. The introduction of ITA/VAT transfers for municipalities gained momentum with significant support from USAID/Sida supported GAP. In general, donor engagement in BH has been extensive but fragmented and has evolved in the decade since the conflict ended and included OHR, OSCE, UNDP, USAID, and the World Bank. These development partners have worked on community development, the enabling environment, and strengthening municipal structures, particularly capacity building. 27

43 14. Recent reforms on municipal financing should result in more equitable levels of local service delivery. The ITA reforms and the Laws on Revenue Allocations in both entities have provided municipalities with a source of more stable and autonomous revenues. Phasing in formulabased allocations of ITA/VAT resources to municipalities in each of the entities should help establish more equitable levels of service delivery across municipalities. 30 (See Chapter 4). Over the medium term, ITA/VAT allocation will benefit poorer municipalities more. However, wealthier and better managed municipalities are more capable of borrowing to increase their capital expenditure. The significant increase in the ITA/VAT (untied) transfers raises questions about prevailing accountabilities surrounding the use of these funds by municipalities. A framework to address these issues is introduced in the subsequent section Dimensions of Accountability for Decentralized Service Delivery 15. In BH, the prevailing legal and regulatory environment has increased local government discretion but there is also a need for accountability checks and balances against abuse. In the local context, accountability involves relationships among central governments, executive and legislative branches of local governments, citizens, and private firms. 16. Horizontal accountability is the control that state institutions exert on the performance of other public agencies. Two dimensions of horizontal accountability exist in the context of decentralized service delivery: (i) lower-level government entities (including government-owned enterprises/utilities) are accountable to government oversight agencies; (ii) the executive branch is accountable to the legislature (state institutions must be able to check abuses by government branches within the same level of government). 17. Vertical accountability is the means by which citizens, mass media, and civil society seek to enforce standards of good performance on officials and service providers. 18. A brief review of how accountability mechanisms are performing can indicate potential entry points, and instruments to address weaknesses in the existing system Horizontal Accountability 19. Horizontal accountability is complicated by a four-level split of responsibilities. Governing authority, functions, and budgets are divided among the state, the two entities, ten cantons within FBH, and the municipalities. The DPA reinforced the cantonal structure created by the Washington Agreement. These arrangements complicate FBH municipalities governing context. Under the FBH constitution, cantons provide a means for both Bosnian-Croats and Bosniaks to have governing authority. Six canton populations are essentially Bosniak, two are essentially Bosnian-Croat, and two are mixed. Cantons have similar checks on ethnic dominance. In the two ethnically mixed cantons, the constitutions have veto structures to prevent dominance by one ethnic group. Although Bosniaks might control a municipality, Bosnian-Croats might control the cantonal government. For example, if a municipality s ethnic majority differs from that of the canton, the municipality assumes certain governance functions that would otherwise have been exercised by the canton If higher-level government resource transfers to municipalities were accompanied by monitoring and accountability, it would ensure that no municipalities fall below agreed minimum standards. Monitoring, which would include standards for public financial management, could help assess the state of the enabling environment for local performance. However, this would require higher-level government to have the capacity and motivation to promote local performance. 30. Law on the Principle of Local Self-Government (Official Gazette 49/06), and new Law on Distribution of Public Revenues in FBH Official Gazette 22/06). A new Law on the Local Self-Government (Official Gazette 101/04) and Amendments to the Law on Budgetary System of RS (Official Gazette 34/06) were adopted/ 31. In FBH, if the dominant ethnic group in the municipality differs from the canton majority, education, culture, housing, public services, land use planning, etc., must be allocated to the municipal level to protect the minority within the canton (Jokay, 2000). 28

44 21. Few structures and procedures exist to link the three governments. Intergovernmental interaction and cooperation is minimal and formal (e.g., through ad hoc visits and exchange of letters). The situation in FBH is compounded by an additional level of intermediate government, the canton, and federal responsibilities for addressing municipal governments that are unclear. Cooperation between canton and municipality occurs primarily along party/ethnic lines 32 according to qualitative interviews The agency with primary responsibility for local government differs by entity. In RS, the Ministry of Administration and Self-Government performs administrative supervision of local self-government activities. The sixty-four municipalities are in direct contact with the entity s relevant ministries and the entity Assembly. 34 RS mayors can exercise political pressure and negotiate for better revenue allocation through the RS Municipal Association. Even though formal responsibility is with the FBH Ministry of Justice, FBH has no ministry exclusively responsible for local government. Federal and cantonal bodies within their corresponding areas of authority perform administrative supervision of activities of local self-governance units. It is said that the canton governments are in daily contact with their respective municipalities, but the nature of canton-municipal relationship varies (see 3.2.2). 35 The federal government s laissez faire approach to municipalities is changing only slowly; and means that FBH government has little contact with municipalities. 23. Higher-government line/sectoral ministries coordination and technical guidance roles are constrained. As in the case of road and water (Chapter 3), sectoral laws regulate planning, financing, management, and set service standards and mechanisms for sectoral oversight. Unlike other countries, the state government in BH does not provide coordination and technical guidance in most sectors. Instead, two entity governments, and cantonal governments in the FBH, via line ministries, are responsible for monitoring and ensuring quality compliance across constituent jurisdictions and enabling a wide variety of service providers across sectors. In reality, however, due to intrinsic sensitivity to cross-jurisdictional subsidies (often across ethnic groups), higher-level governments have rarely fulfilled such functions, resulting in vastly different access to services across jurisdictions and under provision of those services that have geographic spillovers. 24. Higher-level sector regulation produces a multiple-institutional framework. For example, institutions responsible for water and sanitation include the Ministry of Agriculture, Water Management and Forestry (in the FBH and the RS), and the ten cantonal ministries in the FBH. Four special water agencies (Vodopriviedas 36, two in the FBH and RS each) monitor compliance with the Law on Water, which covers environmental protection and water pollution. For waste management and disposal, both entities have regulatory structures that deviate. In the Federation, the Ministry of Urban Planning and Environment (MOPE) is responsible for policy and legislation The FBH Ministry of Health regulates clinical waste, and the FBH Ministry for Energy, Mining & Industry regulates industrial waste. At the cantonal level, ten ministries have a similar structure. In the RS, waste management policy and legislation is under the Ministry of Urban Housing, Communal Services, Civil Engineering and Ecology (MUCCE). Finally, in the road sector, policymaking and standard setting is under the Ministry of Communication and Transport (RS and FBH respectively) and their respective road directorates. 25. Financial accountability mechanisms remain weak and lack infrastructure. The Treasury system allows the entity and cantonal Finance Ministries to track government expenditures at the entity and canton level and ensure conformity to the adopted budget. Municipalities, on the other hand, have begun introducing locally developed financial management information system (FMIS) Kalesija has a political profile reflecting the cantonal level; officials noted that the municipality was given more grant support for capital investments. In addition, five cantonal assembly representatives are from the municipality, and one was described as lobbying for a budget increase. In Stolac, another municipality with canton-municipality party alignment, most of the municipal budget came from canton authorities. 33. In interviews, some municipal officials reported that relationships with higher levels of government were determined by municipal population size. Interviewed officials from smaller municipalities felt that their voice was not heard. They noted an advantage of the Association of Cities and Municipalities is that each municipality has an equal voice. 34. The RS Constitution stipulates that the electoral system must ensure that all municipalities are appropriately represented in the RS National Assembly (Art.71, al.2). 35. In reality, some cantonal officials rarely convene meetings with municipal officials, perhaps once a year, and otherwise the two levels communicate little. (Interviews in Tuzla) 36. In 2007 Vodoprivredas were transformed to Agency for watershed of Adriatic Sea and Agency for watershed of Sava River. 37. About 50 municipalities have functioning FMIS and additional 15 municipalities are expected to implement FMIS by February

45 but are not fully linked with the higher-level treasury systems and the cantons are not part of FBH s entity-level treasury system. The FBH Finance Ministry and cantons use the GIB data (brief execution reports) as the sole source of information on sub-entity-level finances, chiefly for consolidation. So far, the FBH Finance Ministry and cantons do not collect the adopted budgets or analyze budget execution, or monitor the evolution of the intergovernmental finance system over time to inform decisions on expenditure responsibilities across levels of government, or future changes in fiscal policies. 38 It would be worthy considering linking all three tiers financial management information systems to exchange information more timely and accurately while each tier of government keeps spending authorities as granted by relevant laws and maintain the current software to manage their finance internally. Figure 23: Governance Structure for Municipal Service Delivery On Budget Off Budget MUNICIPAL ADMINISTRATION MUNICIPAL COUNCIL Mayor General Administration Office Line Department Public Utility Water Sewerage District heating etc. Private Enterprises Lower Spending Unit Primary school Public library Culture & sports center etc. Citizens Approve budgets Report expenditures Regulate 38. Ministries of Finance for both entities (in FBH through cantons) are collecting local government revenue and expenditure data, and the Macroeconomic Analysis Unit of the Indirect Tax Authority is collecting data for the entire BH. However, these data lack sufficient detail on local governments own revenues; in the FBH these data are only available consolidated at the cantonal level. 30

46 26. Mayors play a central executive role in municipal organization. 39 Until 2004, mayors in both entities were chosen by a directly elected Local Council or Assembly, which exerted ultimate power within the municipal jurisdiction (following the first post-war elections in 1997, and subsequent elections in 2000). Since 2004, mayors have been directly elected and this executive position has a four-year mandate in all municipalities except Mostar and Sarajevo, where the mayor is still elected from the city council. Mayors represent the municipality, lead and manage enforcement of municipal council/assembly decisions and act, propose decisions and manage the work of the municipal administration, propose and dismiss department heads, and exercise rights and duties defined by municipal statute and other regulations. 27. Municipal administration has line departments responsible for direct or indirect service provision. 40 Lower spending units such as schools, day care, welfare, ambulance services or other healthcare, culture, and sports centers, or libraries provide direct services and directly report their expenditures and revenues as part of the overall consolidated municipal budget. For indirect provision, municipalities often have affiliated structures-a general utility company and/or specialized companies responsible for services such as water and sanitation; some municipalities have created their own for-profit enterprises. These public companies are off-budget. Line-departments provide administrative oversight and, together with finance department, oversee financial aspects. 28. Municipal councils/assemblies supervise municipal administration. The Local Self- Governance Laws in both entities have established an oversight system for municipal councils for finance, legal, public outreach, and citizen participation for the municipal administration. In general, the municipal council/assembly sets policy, passes local legislation, regulations and other municipal-level acts, coordinates and monitors the work of communal public institutions and enterprises, monitors municipal administration in general, and performs other duties stipulated by municipal statute and law. The municipal council/assembly comprises members elected locally by secret ballot. According to the FBH and RS laws on Local Self-Governance, councilors cannot be dismissed from duty during their mandate of four years. 29. In BH during the past decade, regulatory changes and incomplete privatization processes have led to some ambiguities in the governance and financial relationship between municipal structures and municipal government. Local council monitors and oversees the governance and financial relationship between municipal companies and municipal government, including the transfer of subsidies and how contingent liabilities from these entities are incorporated in municipal books. Although their governance structure and relationship with government are defined in the Laws on Public Companies, (see Chapter 3), in practice, it is unclear whether such governance structure and accountability mechanisms have been implemented effectively. Not surprisingly, tariff setting is more a function of municipal councilors political considerations than sustainability in service delivery; councilors are reluctant to enforce revenue collection Vertical Accountability 30. Vertical accountability refers to the state being held accountable by non-state actors. The channels of vertical accountability are political elections, mass media, and informal processes through which citizens organize themselves into associations that demand good performance from officials and service providers. 31. In post-dayton BH, political parties and the party system are based on ethnicity. Three major ethnic groups are proportionally represented in parliament according to their numbers in the community as a whole. Each major nationalist party continues to maintain support almost exclusively from their ethnic group. A range of reforms to the electoral system have attempted to undercut nationalist parties and promote multi-ethnic parties but progress has been moderate and the major parties continue to form along ethnic lines by appealing to ethnic identity There are two names for Council/Assembly in Bosnia and Herzegovina. In the FBH, the official name is Municipal/City Council; the RS retained the name used in the SFRJ-Assembly. 40. In the 20 municipalities surveyed, most municipalities have a general utility company for street cleaning, street lighting, garbage collection, public park maintenance, among others services. About half of municipalities have specialized companies responsible for water supply and wastewater collection and treatment. A few municipalities have created their own for-profit enterprises. 41. The October 2006 general elections appeared to rekindle inter-ethnic tensions. 31

47 Box 4. Major Political Parties in BH* This excludes parties represented at the entity, state, or cantonal level but inactive in the municipalities. Bosniak - SDA (Party of Democratic Action) and SBH (Party for Bosnia-Herzegovina) Bosnian-Croat - HDZ (Croatian Democratic Union), HSP (Croatian Party of Rights), HDZ-1990 (reflects a split in HDZ), and HSS (Croat Peasant Party) Bosnian-Serb- SNSD (Union of Independent Social Democrats, SDS (Serb Democratic Party), SRS (Serbian Radical Party) and PDP (Party of Democratic Progress) Multi-ethnic- SDP (Social Democrats), RzB (Work for Improvement- People's Party). 32. Most political parties are silent on the issue of municipal governance. Dayton Peace Agreement emphasized ethnic, rather than civic, identities as the basis for representation in key institutions. This approach has empowered Nationalist parties, although none of the parties has articulated a position on local governance, except SDP-a multi-ethnic party. Parties mention support for local governance reform in their programs, but in general, political programs are characterized by appeals to ethnic or national identity Jurisdiction-based direct election of Mayors could help to build broad-based policy platforms and enhance performance accountability to citizenry. According to a recent BH governance structure study, public debate about the performance and output of the public administration seems to occur only during municipal elections, when politicians seek to win votes by promising better service delivery. Respondents, including mayors, noted that citizens view directly elected mayors in more of a leadership role, which encouraged them to seek out mayors directly to solve problems. Mayors expressed a sense of accountability now that citizens of all ethnic backgrounds hold them responsible for municipal results. The mayor of Buzim had been elected as a member of the SDA; in 2004, he ran and won as an independent-a case of shifting pan-ethnic concerns in an ethnically neutral way. In fact, the creation of Brcko-a multi-ethnic district, set a precedent in BH because ethnicity is not a factor in determining Assembly members, the police force is multi-ethnic, and the school system is integrated. 34. Local communities (MZs) are an intermediary between municipal self-government bodies and citizens, although their presence and roles differ across municipalities. The Law on Local Self Governance defines MZs as legal and political community bodies that liaise and work closely with citizens. The MZs are constituted largely on spatial, historical, economic and cultural bases. Operating under municipality statutes and supervised under the Local Council,43 MZ representatives are not directly elected. In some municipalities, the local council nominates candidates and the mayor makes the final decision. In some cases, MZ representatives are volunteers whose level of involvement and activity depends on personal motivation. The MZs are not fully self-governed and lack public mandates to perform administrative and other services. Their fiscal operations are reflected in municipal budgets. Large variations exist in the number of MZs set up in each municipality-ranging from one in Ljubinje to 72 in Zvornik, and averaging 18 per municipality. No MZs exist in Stolac, and in Mostar, MZs exist only on paper. The roles of MZs differ across municipalities; some are active in local affairs (In Breza, the municipality has channeled investment funds directly through the MZs); others are less active or marginalized. Budget hearings, for example, are held through MZs. 35. Local Self Governance Law established mechanisms for citizen participation.44 Municipalities have established public hearings, feedback lines, ombudsmen, designated open office hours/days for municipal council/assembly members and the mayor, and open MZ sessions for citizens to observe and participate. Municipal service centers have been established in 11 out of the 20 surveyed municipalities. 45 Formal voluntary associations in post-war BH-citizens associations and non-governmental organizations (NGOs)46 provide channels of participation This analysis focuses on major political parties represented at state and local levels. 43. In FBH, MZs are regulated by the Law on Principles of Local Self-Governance, Articles of Association, and Rules of Local Communities. In RS, MZs are regulated by the Law on local Self Governance. Organization, operations and financing of the MZ are established in the MZ Statute. 44. Local Self Governance laws in both entities require municipal governments to facilitate citizen participation in community life and decisionmaking so that citizens can influence decisionmaking on issues such as local infrastructure. 45. Mostar, Samac, Tuzla, Kalesija, Zvornik, Buzim, Banja Luka, Novo Sarajevo, Ilidza, Breza, Jablanica. 46. Citizen associations are usually devoted to the interests of a specific ethnic or social group or to leisure activities and were often subsidized by municipalities. NGOs are generally value-oriented or focus on service delivery and were funded by the international community. 47. Donors see citizen participation as key to good governance. Many donor activities during the reconstruction period focused on increasing citizen participation and ensuring transparent municipal governance (See page 85 on donor activities). In all municipalities, increased participation and transparency are seen to be good outcomes on their own but they also improve planning, counter corruption, and help avoid state capture (ICG). After the war, increased access to formal mechanisms ensured minority ethnicities could have input to municipal government. 32

48 36. MZs, NGOs, and civil society are deemed important and municipalities and NGOs have good relationships. 48 Most NGOs representatives interviewed concurred with the municipal officials opinion that the relationship was good, although the NGO-municipality relationship is less clear in smaller municipalities. 49 Box 5: Bugojno-moving forward together Bugojno is in central BH in the Vrbas river valley and belongs to the Central Bosnia Canton (CBC). The prewar population was 46,889, of which 19,697 or 42 percent were Bosniaks; 16,031 (34.2 percent) were Bosnian-Croats; or 18.5 percent were Bosnian-Serbs; and 2,488 (5.3 percent) were other ethnicities. Today, Bugojno remains quite diverse. The population is about 39,000 of which 24,400 are Bosniaks (62.5 percent), 11,000 are Bosnian-Croats (28.2 percent), 3,500 are Bosnian-Serbs (8.9 percent) and there are100 others. The Bosniak Party of Democratic Action (SDA) has 13 of 25 seats on the Municipal Council, but despite its majority, governs in a very constructive and inclusive way, which makes the work in the Municipality proactive and forward looking. In fact, Bugojno has been recognized as a best practice example of cooperation within the municipal administration, and externally, with citizens. The Mayor organizes briefings every Monday morning with department heads so they can keep one another informed and improve inter-departmental communication. Similarly, the Municipality has supported NGO work, especially youth groups, and recognized the importance of civil society contributions to the community. The 2006 Municipality budget included 20,000 KM for NGO projects and 25,000 KM in The Municipality established a Commission comprising municipal and civil society representatives to approve grants to NGOs, and supported establishment of a Local Advisory Group, comprising community actors and civil society representatives who provide input, develop initiatives, and design proposals to improve services to citizens. The Municipality also established a Youth Empowerment Partnership Program (YEPP) group in charge of local development that lobbied for 25,000 KM for youth entrepreneurship for Participation in formal local governance mechanisms remains relatively low. Citizens demonstrated awareness of the various sources of information available in the municipality: a routine has been established for responses to information requests, and many municipalities have established procedures to handle standard transactions and publicized these procedures. However, most survey respondents have not participated in or attended any community activities in the past 12 months 50 ; most citizens interviewed had never contacted local government representatives. 51 In contrast, between percent of respondents in ten municipalities had contacted civil servants. 52 Although municipal officials interviewed say that MZs are a primary resource for information on community needs and citizen inclusion, citizens do not take advantage of MZs as mechanisms for participation, and do not see them as a meaningful resource. Fewer respondents than in the past consider MZs a trustworthy local institution Participants in local governance activities considered the experience useful. Of those that have participated local governance activities, public meetings on the municipal budget are the most well attended; municipality assembly sessions are the least popular. Respondents describe MZ council meetings as the most useful (81 percent), and consider least useful the activities related to NGOs other than meetings (70 percent). 48. Two mayors said that initial problems with NGOs included distrust, competition for donor funding, poorly coordinated activities, and lack of communication with the municipal government. However, now mayors noted that some NGOs receive funding from municipalities and their roles are more generally accepted. In Banja Luka, brochures prepared by the Helsinki Citizens Assembly on municipal functions were well-received by the municipal government. In Berkovici, the mayor stated that funding for NGOs had been increased to three percent of the total budget. 49. NGO representatives cited funding made available to NGOs, and in one case, the use of municipal premises as the evidence of cooperation. Two new mayors were singled out as having improved relationships with NGO, Zvornik and Visegrad. In two other municipalities, NGO representatives disagreed with the perspective of municipal officials, stating that the municipality is non-transparent in its dealings with NGOs, and municipal fund allocation criteria are lacking. 50. Earning a livelihood was cited as the main constraint to civic participation. Municipal officials and NGO representatives interviewed thought most citizens were struggling to earn a living and had no time to participate. Many people work more` than one job to make ends meet. Other qualitative research in BH cited travel distance to municipal meetings as a barrier to rural respondents participation. 51. Exceptions are Buzim and Odzak. In Buzim, 56 percent of respondents had contacted their MZs in the past 12 months; in Odzak, 70 percent had. Both municipalities also had higher rates of contact with other local government representatives. MZ representatives in Odzak are invited to municipal council meetings and meet with the mayor as well. 52. Odzak, Buzim, Samac, Banja Luka, Kalesija, Breza, Novo Sarajevo, Zvornik, and Ljubinje. 53. In Buzim, 92 percent of respondents could name their MZ representatives, 72 percent in Odzak. In Kalesija and Breza, by contrast, only slightly more than 10 percent of respondents could name the representative. 33

49 Figure 24: Participation or attendance at various local governance activities Public meetings on municipal budget 20% Local council sessions 13% Any unpaid communal activities Public hearings other than on municipal budget NGO activities other then meetings Municipality assembly sessions 9% 8% 6% 5% 39. Overall, client feedback loops remain weak. Part of vertical accountability is the relationship of frontline service providers to citizens. As service provider, municipal companies should be able to understand and evaluate citizen demand for public services and get direct feedback on performance outcomes. However, this vertical accountability chain appears to be broken, as evidenced by low responsiveness. Among people surveyed, about two-thirds (67 percent) claimed that if they had problems with a service they would contact the institutional service provider, but in reality only 20 percent report problems, and of the problems reported, only about six percent are ever solved. Lack of feedback channels is only partly to blame, since an increasing number of municipal service companies have introduced citizen complaint centers or service hotlines. 2.3 Conclusions 40. The legal and regulatory framework for decentralized service delivery has been established in BH. The local government finance reforms brought by the ITA reforms and the Laws on Revenue Allocations in both entities, have institutionalized a more stable and autonomous source of revenues for municipalities. The recent Local Self-Governance Laws in both entities are a step toward strengthening local government discretionary space and accountability. 41. Full implementation of legislation and reforms are key to maximizing benefits of decentralized service delivery. The accountability framework (vertical and horizontal) maps the actors and their interrelationships, which can influence service delivery (and the quality of intermediate processes such as public financial management). Detailed analysis of the existing and emerging accountability mechanisms reveals strengths and weaknesses in the checks and balances to ensure good local governance. The next chapter examines how the decentralization framework and accountability mechanisms manifest in two sectors-water and local roads. Following that, the report will identify effective entry points for strengthening accountability. 34

50 Chapter 3: Water and Roads: How Sector Structures Influence Local Service Delivery Performance 1. Adequate financing is only one dimension of effective local-level service delivery. Figure 25 highlights seven key dimensions of managing service delivery across levels of government that shape the performance of each municipal government. Performance can be seen as cross-cutting (i.e., How well do municipals appear to manage public funds?) or sector-specific (i.e., How responsive and efficient is the provision of local roads to local needs?). Effective local-level municipal service delivery depends on much more than simply assigning a service to a municipal government. 54 Decentralized service delivery provision must be unbundled to understand whether capacities and accountabilities across levels of government are assigned effectively. Figure 25: Unbundling Roles and Responsibilities for Decentralized Service Delivery Policymaking Standards Setting Administration/contracting Financing Service Delivery Regulation Who sets the main policy guidelines for a service (e.g., free primary education as a national policy)? Who sets the standards (e.g., nation-wide testing, local building standards)? Which government authority administers day-to-day services? Who finances the services? Capital investments, wages, O&M? Does one level determine the allocative mix? What area of sectoral financing passes through the Sub-national budgets? Who produces/delivers the service (e.g., this can often be private)? Who regulates a service (e.g., professional accreditation in health)? Monitoring & Evaluation Who monitors and evaluates a service (e.g., how can local citizens provide feedback)? Source: Public Sector Group, Fiscal Decentralization Diagnostic Guidelines, Washington, DC (November 2007) 2. Decentralized service delivery differs by sector. To optimize functional and financial arrangements for decentralized service delivery, each sector s different characteristics must be analyzed. This chapter examines how sectoral organization and financial arrangements influence service delivery outcomes using the road and water sectors two major service provision functions for which local governments are responsible. 55 Figure 26: More roads than water pipes in under-funded municipalities Roads Access and Per Capita Revenues Water Access and Per Capita Revenues Road connects household to village 100% 90% 80% 70% R 2 = % 50% 40% 30% 20% 10% 0% ,000 1,500 2,000 PC Reve nue s in KM Running water within household 100% 90% 80% 70% R 2 = % 50% 40% 30% 20% 10% 0% ,000 1,500 2,000 PC Revenues in KM 54. International experience provides broad guidance on the appropriate level of assigning a service delivery function. Services that are expected to have significant externalities across local jurisdictions can be assigned to higher levels of government and/or coordinated across levels of government (e.g., inter-regional roads networks). Moreover, local jurisdictions may under-provide a service if its citizens can access the service without tapping into their own budget. In turn, other jurisdictions may also under-provide if they feel burdened by free-riders. Ultimately, appropriate assignments depend on country context and local conditions. The subsidiarity principle argues that the lowest and presumably most responsive level of assignment is most appropriate. 55. In fact, in no other sector than water do municipal per capita revenues correlate as strongly with access. 35

51 3. Municipalities can increase satisfaction levels in the road sector. In municipal service delivery outcomes, functional assignments in the sector make a difference. For example, higher client satisfaction rates for roads correlate strongly with steady maintenance, regular cleaning, and reliable winter services directly provided by the municipality (Figure 27). By contrast, in the water sector, increasing service reliability (24/7 running water, sufficient pressure) requires capital-intensive rehabilitation. Figure 27: Municipalities can increase satisfaction in the roads sector Roads Reliability and Satisfaction Water Reliability and Satisfaction Citizen Satisfaction 100% 90% 80% 70% 60% R 2 = % 40% 30% 20% 10% 0% 0% 20% 40% 60% 80% 100% Reliability (Regular cleaning) Citizen Satisfaction 100% 90% 80% 70% 60% 50% R 2 = % 30% 20% 10% 0% 0% 20% 40% 60% 80% 100% Reliability (more than 21h running water & no need for repair) 4. Sectoral organization and financial arrangements influence outcomes. Sectoral variations in functional and financial arrangements for decentralized service delivery offer different entry points for future reforms and capacity building. The following section will discuss the two sectors in terms of the following: (i) functional assignments across levels of government; (ii) financial management arrangements and characteristics of the main service providers-the Road Directorates in both entities, and utilities in water, and relevant municipal units; and (iii) how sector organization affects outcomes in municipal service delivery. 3.1 Municipalities and the Roads Sector General state and organization 1. Overall condition of the network remains poor. Despite substantial road reconstruction in post-war Bosnia and Herzegovina, half of the main road network, 57 percent of the regional network and most of the local road network are not in good condition (PEIR 2006). The road sector remains in need of substantial capital expenditure to clear the maintenance backlog. Despite increased domestic revenue, recurrent needs in the sector are not being fully met. 2. Main, regional, and local road classifications are linked to functional responsibility for maintenance and development. In FBH, the entity government is responsible for trunk roads and the cantonal government for regional roads. In RS, the entity government is responsible for both trunk and regional roads, and municipal governments in both entities are responsible for local roads. Figure 28: Road Network in BH Road infrastructure in BH (km) Entity Trunk roads Regional roads Local roads Total RS 1,781 2,183 FBH 1,958 2,724 13,730 22,376 Total in BH 3,739 4,907 Source: Consultant Report (in prep for PEIR 2006) 36

52 3. The transport sector institutional structure is overly complex. Multiple layers of administration exist at the state, entity, cantonal, and municipal levels. The Dayton Peace Agreement gave the state limited power-regulatory coordination for international and inter-entity transportation, although the state has continued to assume more responsibilities. 56 In 2003, BH established the State Ministry of Transport and Communications (SMOTC) as part of the process of national-level institution building. 57 Transportation infrastructure policymaking, development, management, and financing responsibilities rest largely with the entities. Both the Federation (FBHMTC) and the RS (RSMTC) have separate Ministries of Transport and Communications. Four Directorates operate in the road sector, two in each entity, responsible for motorways and primary roads. Within the Federation, ten Cantonal Ministries of Transportation and Communications (MTCs) and four Cantonal Road Directorates are responsible for regional and local road network planning and management. The Roads Directorate/Company plans and manages trunk roads, and regional roads in RS, rather than physical implementation. Road maintenance and construction is contracted out to public/private/mixed contractors, normally through competitive bidding. 4. Municipalities are responsible for local roads. In both entities, municipalities are responsible for participating in the management and financing of local roads, primarily maintenance. Similar to arrangement at higher levels, the actual service delivery-maintenance and operations-are often contracted out through competitive bidding. 5. Strategic sectoral planning and implementation are difficult and costly. The road sector is plagued by a complex organizational structure that defies coherent outcomes and creates a patchwork of domestically and internationally funded programs. Ten cantonal MTCs, plus four cantonal Road Directorates create a layer of administration that consumes disproportionate amounts of the sectoral budget. Network size and budget can be too small to take advantage of good planning, construction economies-of-scale, and technological progress. The fragmentation of responsibilities in the sector exacerbates reduced capacity in the respective bodies. Figure 29: Institutional Responsibilities and Decentralization of the Road Sector in BH State Federation RS Entity Canton Municipality Entity Municipality Policy making State MTC on international collaboration FBHMTC Cantonal MTC (strategy for regional and local roads) RSMTC Standard setting FBHMTC for all levels RSMTC for all levels Regulation FBHMTC RSMTC Administration /contracting (planning) Entity Road Directorate Cantonal MTC (through Cantonal Road Directorates in 4 cantons), including planning of local roads Department of Urbanism or equivalents Entity Road Directorate Department of Urbanism or equivalents Financing Trunk road, regional roads and local roads Regional roads and local roads Local roads (primarily for maintenance) Trunk, regional and local road Local roads (primarily for maintenance) Service Delivery Private sector provider Public and/or private sector provider Public and/or private sector provider Private sector provider Public and/or private sector provider Monitoring & Evaluation of outputs Entity MTC road inspector for trunk road Cantonal MTC road inspector for regional and local road Entity Road Inspectors for trunk and regional roads Authorized municipalities for local roads Source: Laws on Roads in FBH and RS 56. A BH-wide Transport Master Plan was drafted but never officially adopted by the Parliament. The World Bank is supporting the drafting of a National Transport Strategy led by the State MOTC. The State Law on Road Safety is under discussion and efforts are under way to develop a BH Law on Roads. 57. The Law on Ministries and other Authorities of Bosnia and Herzegovina. As published in the Official Gazette of Bosnia and Herzegovina No. 5/

53 6. A draft Road Law aims to realign responsibilities within the Federation. The law will centralize some road functions: the entity Ministry of Transport and Communications will expand its mandate to main roads and regional roads; 58 the cantonal Road Directorate and authorities responsibilities will be reduced to local roads Road Sector Financing Major revenue sources are fuel duties and motor vehicle registration fees. 60 Prior to 2006, 40 percent of total revenues-fuel duties 61 plus full proceeds from annual motor vehicle registration fees-would go to the Federal Road Directorate and 60 percent to the cantons (as revenue of four Cantonal Road Directorates or, in the other six cantons as revenue of MTC). 62 The cantons had to allocate 25 percent of total funds to municipalities in their jurisdiction.63 Today, a share of the revenues is coming through the single account for both entities road funds. Details on the distribution are described in the revenue allocation law, while the distribution among cantons has been specified in the FBH rulebook on revenue shares. 8. Not all cantons have fulfilled their obligations: Tuzla Canton, for instance, did not transfer those revenues to the municipalities in Sarajevo Canton shares road sector revenues only between municipalities and the Cantonal Road Directorate. 9. Road financing comes mainly from higher-level government. The funds municipal governments receive are horizontally allocated among municipalities according to geographic area, population, and length of local roads (Pacific Consultants International 2007). Driving license fees and various local road fees (e.g., for permitting billboards along the local roads) are collected for the expenditure needs of local roads-but these are of minor value and a large share of the total local road financing comes from higher-level government transfers. 10. Dedicated revenues from motor vehicle registration remain the same, but the proportion of fuel duties that went to the Road Directorates (including the six cantons without RDs) are now channeled through the Single Account allocation, with shares determined by the Law on Single Account (3.9 percent in FBH, 3.5 percent in RS). Consequently, the amount available for the Road Directorates/Road Company now depends on the overall single account pool, reducing predictability and amounts received in sub-entity governments responsible for the road network at all levels 64 This led some Road Directorates to lobby for extended earmarking and increases in fuel duty, or a larger share from the Single Account Allocation Excluding motorways, which are under the Motorway Directorate. 59. Cantons are another layer of sub-entity government so FB structure is more complex than that of RS. Therefore, this section focuses on the flow of funds in FBH for illustration purposes, with occasional reference to the RS side when appropriate. 60. In the FBH, these revenues are collected by cantonal government and then shared between the entity (upwards) and the municipalities (downwards). In 2005, fuel duties and motor vehicle registration fees contributed KM34 million and 60 percent for the RDs in FBH, KM36 million and 38 percent for RS RD. But share was smaller in latter case, as they received KM24 million from GSM licenses that year Pfenings dedicated, with remaining 3 Pfenings going to the Entity general budget. 62. The actual proportions each Canton received in 2006 are detailed in Annex Entity Road Laws define road sector revenue sources and vertical allocation to the entity, canton, and municipality levels. 64. Some maintain that the fuel tax share of total revenues for roads in BH is very low (8.0 percent) compared with EU standard of percent (even lower than Albania). Some think that gas station profits are too high (30 percent), suggesting room to charge more fuel tax. (Interview with Sarajevo Canton Road Directorate, March 28, 2007.) 65. Interview with the Entity Road Directorate in the Federation. 38

54 Figure 30: FBH Road Sector Fiscal Flows before Pf/l o Tolls (highway, bridges, tunnels, etc.) o Foreign vehicles fee o Other fees applicable to Federal levels 40 Entity Road Directorate Federation Ministry of Transport Fuel Taxes 15 Pf/l Motor Vehicle Registration Fee 12Pf/l Cantonal Ministries of Transport (4 Road Directorate) 25 percent total Road Driver License Fee Municipality (79) Other Local Road Fees Local Communities Transfers (both capital and current) Transfers (both capital and current) Figure 31: FBH Road Sector Fiscal Flows after 2006 Other Indirect Taxes, including VAT o Tolls (highway, bridges, tunnels, etc.) o Foreign vehicles fee o Other fees applicable to Federal levels Single Account Fuel Taxes 15 Pf/l 3.9 percent 40 Entity Road Directorate Federation Ministry of Transport Motor Vehicle Registration Fee Cantonal Ministries of Transport (4 Road Directorate) 25 percent total Road Driver License Fee Municipality (79) Other Local Road Fees Local Communities Transfers (both capital and current) 39

55 11. Most expenditures are for maintenance but existing funds are insufficient. 66 The bulk of funds go to maintain existing roads and preserve existing investments, but funds are insufficient even for trunk and regional road networks (PEIR 2006). Not surprisingly therefore, local roads (total 13,730 km) are in poorer shape than main and regional roads. 67 The 20 municipality survey results also show that the lack of satisfaction with local roads is largely correlated with service reliability a function requiring maintenance funds which depend on the municipalities. 12. Allocation for capital investment has been largely ad hoc. This is particularly true at subentity levels. Investment plans tend to ignore overall fiscal envelope, thereby offering little guidance for investment prioritization-capital improvements versus operations and maintenance, or among capital projects (PEIR 2006). In fact, most capital investment for the entire sector has been provided by the international community (PEIR 2006). 13. For local roads, capital investments depend on transfers from higher tiers of government. Municipalities have struggled to meet maintenance requirements with limited financing so it is common for local communities to provide capital contributions to urgently needed road upgrading and construction projects. Community involvement in local roads-capital contributions and labor-is highest in the poorest one-third of municipalities measured by per capita revenue (Figure 32). Moreover, the availability of community contributions increasingly is a criterion for prioritizing investment decisions, for example in Sarajevo and Zenica-Doboj Canton. Figure 32: Type of community involvement Only those who reported that there has been some community involvement, n=487 Number of missing cases: 37 Settlement Type Entity Finansially Manual work We provided construction material or food to workers ALL RESPONDENTS Rural (288) Urban (199) FBiH (294) RS (193) 47% 70% 66% 76% 85% 79% 55% 60% 49% 39% 6% 7% 5% 8% 3% n.s. P <. 01 r =. 41 Population 2,000-12,500 (174) 12,501-25,000 (187) 25,001 and more (126) 68% 74% 67% 56% 49% 63% 2% 7% 10% P <. 01 r =. 14 PC Municipal Revenue KM (186) KM (131) 301 KM and more (170) 83% 64% 61% 46% 52% 68% 1% 12% 6% P <. 01 r = Including periodic, routine, and winter maintenance. 67. Road sector expenditure needs are large: in 2005, the FBH allocated KM39 million to maintain primary and secondary networks; RS allocated KM36 million for routine and winter maintenance (BH Ministry of Communications and Transport, Dec 2005). By contrast, over the next five years, annual maintenance needs are estimated at KM137 million on the trunk and regional road network, and KM24 million for the local road network (excluding maintenance backlog outlays). 40

56 Box 6: Breza-Citizens participate in building Local Infrastructure In the immediate post-war recovery period, transparency and public involvement at the local level was given little attention in Bosnia and Herzegovina. But now increasingly municipalities have realized the importance of close cooperation among citizens, civil society, and the private sector for implementing infrastructure projects. The capital planning process is one of the most important functions of local governments; it must not only reflect local community priorities but also contribute to developing a strategic plan for the municipalities. A successful capital planning process must be transparent and establish channels for communication with citizens. In 2005, Breza Municipality in Northern Bosnia and Herzegovina began to implement a program for citizen participation on community development. Local authorities joined the Governance Accountability Project (GAP), which incorporates a method of Community Development Planning Committees (CDPC) to select local priority projects and develop local community strategic goals. Breza Municipality, led by the Mayor and Municipal Council, created administrative conditions for CDPC work. The CDPC in Breza comprised the MZ representatives, local businesses, non-governmental sector representatives and has successfully identified priorities for developing the local community. The Breza CPDC provides local communities with a tool to prioritize and select capital investments-each year it calls for proposals for capital projects financing. The local administration distributes forms to the MZs and other interested organizations, holds public hearings, and hosts radio shows and through this transparent and open procedure, capital investments are selected. Breza, similar to most BH municipalities, has very little budget for capital investments. During the post-war period, donors and favorable loans financed most infrastructure reconstruction. But now the international presence is shrinking, higher-government level funding for capital investments is insufficient, so municipalities are developing methods to maintain existing infrastructure and build new projects. In Breza, the municipality offers to finance local road construction with matching funds of 50 percent of total investment, if local citizens contribute the remaining 50 percent. This is a citizens initiative that come through their MZ, and the local council has passed a resolution that obligates the mayor and administration to match citizens funds budgetary funds. This practice is now widespread throughout BH and has proven its utility in road local-level construction, sewerage and water supply projects. This process has significantly improved transparency, citizen participation and oversight. Ultimately the CDPC commission continues as part of the annual budgeting process to select priority local community projects, and in the future, this process of citizen participation will qualify local authorities to apply for EU funds available to local governments. Figure 33: Sources and Uses of Financing for roads in FBH Construction and Upgrading Maintenance and Operation Trunk Road Entity capital budget Designated fund for the RD + entity current transfers Regional Road Cantonal capital budget + entity capital transfer Designated fund for 4 cantonal RDs (and 6 cantonal transport ministries) + entity current transfers Local Road Cantonal capital transfer + entity capital transfer (i.e., from donors) + community contribution Municipal budget + community contribution 41

57 3.2 Municipalities and Water Utilities General state and organization 14. Water utilities are numerous but coverage remains incomplete. Rural access is low and about one third of rural households (40 percent in rural RS) use water from unsafe sources. Water utilities and their services are public companies, therefore independent under the law, but their legal status has not prevented political meddling, mainly on tariff-setting practices. Tariff levels are often too low to cover operation and maintenance (O&M), let alone provide sufficient funds for capital investments. 15. In both entities, water provision is assigned exclusively to the local level. Article 12 in RS Law on Local Self-Governance assigns independent competencies in providing many services to municipalities, which includes regulating and securing public utility services and establishing companies, institutions, and other organizations to provide the services. In the Federation, Article 8 of the Law on Principles of Local Self-Governance assigns the management, finances, and facilities improvement of local utility infrastructure to local self-government, including water supply. Article 13 further empowers the municipal or city council to found enterprises and institutions for the execution, to adopt Rules of operational procedures and regulations of fees, and make decisions on the management. Although specified in less detail, similar rights and obligations are assigned to the Municipal Assembly in RS Law Article 14 and Water utilities operate mainly as public companies. Their governance structure and relationship with government are defined in the Laws on Public Companies. Therefore, by law in both entities, the following bodies are mandatory: Assembly, Supervisory Board, Management bodies and Audit Committee. The Supervisory Board has a minimum of three members, selected according to public appointment laws 68, and it supervises the Management body and appoints or dismisses its members. Individuals holding political positions cannot be appointed to the Management body. Management reports to the Supervisory Board upon request, normally once a year. The Supervisory Board reports to the Assembly (the municipal council in the case of municipal-owned enterprises), which represents company stakeholders. The Assembly decides on operating procedures, code of conduct, and the plan of activities, which are prepared by the Supervisory Board. Both laws oblige public companies to establish an Audit Board or Audit Committee, which appoints an external auditor and the director of the internal audit department 69. In practice, it is unclear whether such governance structure and accountability mechanisms have been implemented effectively. 17. In 2006 some water sector responsibilities transferred to the State level. This transfer rarely affects regulation and policymaking in the sector, but it ushered in the Ministry for Foreign Trade and Economic Relations as a player in BH water sector, even though its role in inter-regional matters is limited to coordination. 18. Environmental protection and water pollution regulation is defined at the Entity level and specified in the Law on Water. Two special agencies, Vodopriviedas, (VPs) operate in FBH and the Directorate for Water in RS to monitor compliance with this law. New laws on water, effective July 2007, will transform the structure of VPs in FBH from public companies to special agencies that report to the Ministries of Water. RS will introduce an additional directorate, matching VPs regional responsibilities in FBH. 19. Some FBH cantons have introduced Laws on Public Services. These laws regulate beneficiaries and providers roles and set service standards; some laws even specify how municipalities should organize service provision. In Sarajevo Canton, the law specifies that tariff rates must be set at cost-recovery levels and if that cannot be done, the cantonal budget must make up the short fall. In practice, however, tariffs are still well below cost-recovery and deficits are not covered due to lack of funds. 68. Law on Ministerial, Governmental and Other Appointments in the Federation of Bosnia and Herzegovina ( Official Gazette of the Federation of Bosnia and Herzegovina No. 34/03) and Law on Conflict of Interest in Government Institutions in Bosnia and Herzegovina ( Official Gazette of Bosnia and Herzegovina No. 16/02) 69. In FBH, companies with fewer than 50 employees are exempt from having internal control departments. 42

58 20. Municipal size affects choice of service provider for water and sewerage. In bigger urban areas a separate company usually provides both water and sewerage but in small municipalities, one general utility company is typically in charge of waste collection, local road maintenance, park services, among others. All municipalities operate their own (separate or integrated) water companies; except Sarajevo Canton, where nine municipalities have transferred the municipal services provision to the cantonal level. As a result, the water sector is fragmented, with roughly 130 municipal water companies serving a population of around 4 million people. Some 67 water utility companies have created the Bosnia and Herzegovina Water Works Association, to promote integrated national and regional policies in the water sector Water Sector Financing 21. Sector tariffs are too low to cover basic costs. Typically operation and maintenance costs of water utilities are financed by user charges, municipal councils set user fees and charges, and water companies have limited influence on the tariff structure. In BH, local authorities tend to set fees well below production cost to avoid losing local political support. In addition, low collection rates and low willingness to pay among users, particularly large public sector consumers such as military facilities, hospitals and schools, add to under funding. High outstanding bills create arrears on both sides of the equation, apparently common practice within the public sector. 22. Municipalities are responsible for capital expenditures. Since water utilities are severely underfunded, they rely entirely on municipalities to provide capital investment funds. Municipalities also depend on funds, grants, and subsidies, which sends them to seek co-financing from higher levels of government-entity or cantonal budgets, or both. Several international agencies, including the World Bank, provide investment funding for water infrastructure improvements, and municipalities can provide guarantees that enable water companies to borrow on the commercial market. However, borrowing is restricted and funds are limited so municipalities rarely manage to meet their enormous water sector investment needs. Water companies interviewed for this study consistently cited lack of investment funds as the cause of their low quality scores and their biggest constraint to rehabilitation and expansion of the networks, but low collection rates also contributed to shortage of capital funds. 23. Higher levels of government co-finance capital investments. Entity and cantonal governments finance their contribution to water sector investments from the general budget and funds generated by water fees that are in addition to user charges-a fee of 0.09 KM per m3 (special water management and pollution fee), and all for-profit companies must pay 0.5 percent of salaries. Effective July 2007, new water laws specify that 40 percent of these fees will go to the special water agency (former Vodoprivieda), 45 percent go to the cantonal budget, and 15 percent to the entity budget. 70 These funds are intended to co-finance municipal capital investments. Under the old law, the Vodopriviedas received 70 percent of these revenues (20 percent went to the cantonal and 10 to the entity budget), but they had co-finance up to 30 percent of municipal investments. Projects were selected by the advisory board and approved by the ministry of Agriculture, Water Management and Forestry. 70. Federation law amendments state that these fees are shared according to formulae defined by the Laws on Water. 43

59 Figure 34: FBH Water Sector Fiscal Flows before July 2007 (status after July 2007) Vodopriviedas (Special Agency) Entity Canton Special water management and pollution fee Percentage of salaries from for-profit companies 0.09 KM/m percent of salaries o General budget o Water fees o User charges 70 percent (40) t 10 percent (15) t Municipality 20 percent (45) t Co-finance upto 30 percent (no mini. /max. limit) Set Tariff Capital investments User Charges Water Utility (Public Company) Guarantee Borrow for capital Investments Donors/ Commercial Market 3.3 Conclusions 24. Sectoral variations in decentralization affect service outcomes. Wide variations in road and water sectors functional and financial arrangements indicate that decentralization manifests differently by sector. Water has been largely decentralized to the municipal level, but in the road sector, municipalities and higher-level governments in BH share responsibility for local roads provision. In both cases, major service provision agencies have been corporatized to ensure professional operation and relative financial autonomy. However, in the road sector, RDs operate mainly at the entity and cantonal level to capture economies of scale, to allocate resources strategically, and to improve coordination across jurisdictions. But in the water sector, cost and benefit are locally determined and costs should be covered locally. The concerns common to both sectors however are financial and regulatory oversight of the public enterprises to ensure public services access, quality, reliability, and affordability, and to mitigate contingent liabilities. 25. Allocations for capital investment need clear criteria. Road financing comes mainly from shared revenues that are based on fuel duties and registration fees. Funds are used for both maintenance of existing roads and investments in new roads, under regional coordination. The present sharing formula in the road sector helps to ensure fairly equal distribution of available funds, although at a low level and generally insufficient to significantly improve local roads-especially when compared to the regional and trunk road network. Municipalities are expected to contribute to financing road maintenance, but not road extension. However, in the water sector, municipalities must cover the major share of capital expenditures, so that underfunded municipalities suffer disproportionately from lack of investment funding. They cannot afford to extend water networks or to invest in urgently needed rehabilitation measures, and as a result, fewer citizens of poorer munici- 44

60 palities have access to safe drinking water. Although the Vodopriviedas now Agency for watershed of Adriatic Sea and Agency for watershed of Sava River do co-finance selected investment projects from revenues generated by water fees, prioritization criteria are unclear and selection of investment projects appears to be ad-hoc. 26. The water sector needs urgent measures to become cost-effective. Municipal water company operations are financially unsustainable. Quality and reliability suffer from steadily deteriorating conditions of the water network, eroding consumer willingness to pay and thereby weakening the income base and eroding funds required to finance desperately needed network extensions. Improving water sector service delivery outcomes will require water company operations to become more efficient and cost-effective. Most water utilities, despite a commercially oriented organizational structure, still have plenty of room for more efficient operation but efficiency gains alone will be insufficient to mobilize the resources missing in the sector. The pricing schemes must be adjusted to cost-recovery levels by municipal councils because prevailing water tariffs are too low to even cover current operation and maintenance costs, which accelerates asset decline. Water tariffs frequently succumb to local political capture and are set well below production prices. The justification given is the need to maintain affordable prices for vulnerable groups, but ironically the socially vulnerable suffer disproportionately from current practices because they are frequently the very people who are excluded from the network, and endure bad water quality and frequent service interruptions. In effect, low tariffs constitute untargeted subsidies that benefit higher income groups to the detriment of the whole water system. Instead, to protect the access of vulnerable groups to a reliable safe water supply, a targeted subsidy with clear eligibility criteria would be preferable. Otherwise, water should be supplied on a user-pays, cost-recovery basis. 27. Improving municipal performance requires different actions for different sectors. Whereas water utilities must focus on more efficient and cost-recovery operation, local road companies must increase service provision reliability. In the water sector, municipal performance improvements will benefit from further commercialization of water utility companies, and strengthening municipal financial capabilities through appropriate tariff decisions made in the municipal councils. By contrast, in the roads sector, some local roads service outcome improvements are at the mercy of decisions made at higher levels of government, but municipalities can improve customer satisfaction by providing more reliable services. 45

61

62 Chapter 4: Options to Expand the Envelope for Capital Investments in Municipalities 4.1 Overview 1. Bosnia and Herzegovina municipalities have diverse fiscal positions. The expenditure and revenue positions of the 106 municipalities suggest that each differs in own-source revenues, current and prospective transfers from higher-level governments, borrowing levels and potential, prevailing expenditure structures, and the related scope for efficiency gains and/or improved prioritization of expenditures. In managing its finances to provide optimal public services to citizens, each municipality must assess how much fiscal space it has in each of these dimensions. 71 Some dimensions depend on municipal-level administration (for example, managing existing own-revenue bases to gain efficiencies in existing expenditures). But constraints and prospects, such as the evolution of ITA/VAT transfers, depend on entity governments. Each municipality must assess choices and options for good intergovernmental fiscal practice This chapter analyzes levels, structure, and trends in municipal revenues and expenditures to examine priorities for managing fiscal space that could contribute to improved service delivery. 73 The analysis, which draws on comprehensive data on public finances in FBH and RS, will identify municipal groups that appear to share public financial management challenges and priorities, and pay special attention to how the allocations from the Single Account, introduced in 2006, are likely to impact municipalities. This chapter provides an overview of the evolution of aggregate revenues in the municipal sector as a whole and comments on the available data. It provides a detailed analysis of differences in revenue structure across individual municipalities with a focus on own-source revenues, and an overview of likely impacts from the reform of transfers. It analyzes expenditure patterns (notably wages) relative to revenues. The final section addresses sub-entity borrowing. Annex 1 and 2 provide the revenue and expenditure categories collected in the GIB data. 71. Alternative perspectives have focused on facets of increasing fiscal space for priority expenditures for local governments. See Billand, Charles J. (2006), Expanding Local Government Resources for Capital Projects through Municipal Borrowing and Other Market-Based Financing, Global Urban Development, Vol. 2, Issue 1, focuses on market-based borrowing and own-source revenue mobilization, rather than traditional efforts to increase grant-based funding from central governments or donors. 72. Most of the literature cites Sub-national governments being subject to hard-budget constraints, i.e., that they will not over-borrow and/or spend inefficiently, expecting higher-level government bail outs. 73. The analysis relies on standardized municipal data reported to entity level governments ( GIB data ) for budgets and realizations. These data have limitations, especially in functional classifications (i.e., reporting by sector). At the time of writing, only partial data were available for 2006 realizations, and the analysis of the effect of ITA introduction transition year is based on these. 47

63 4.2 Municipal Public Financ Municipal budgets total over one billion KM of spending (5.7 percent of GDP). Figure 35 presents current municipal expenditures compared to other components of public expenditures in BH.74 Overall expenditures are at 45.2 percent of GDP (2007). These are accounted for by levels of government (state, entities, FBH cantons, and municipalities) and extra-budgetary funds. Municipalities in turn accounted for about 21.7 percent (RS) and 10.4 percent (FBH) budgets. Figure 36 summarizes available data for the municipal sector in 2006 by entity. While RS accounts for just over a third of estimated GDP in BH, total revenues for its municipalities are about half of the total. The larger relative size of the municipal sector is in part explained by the presence of the additional cantonal level in FBH, which is not the case in RS. Figure 35: Municipal Expenditures in BH (2007, projected) RS ExBudgetary Fund RS Road Fund RS Municipal (est) State Brcko Federation Budget Federation Road Fund Fed Catonal & Municipal Budgets RS Budget Fed ExBudgetary Funds Source: IMF (2007) 3. Fiscal reporting by municipalities is subject to limitations. All municipalities are supposed to report their revenues (own collections or transfers) and expenditures to higher levels (through the so-called GIB forms), but consolidated data on municipal finances are problematic. Data problems are even more pronounced in FBH, where reporting is intermediated by the cantonal level. 75 Historically, the GIB process has reported expenditures by both functional classifications-education, communal affairs (see Annex 3.2), and economic classifications such as wages. However, since municipalities appeared to be reporting largely by administrative units with rather imprecise mappings to the functional classifications (see also Chapter 5), this reporting was dropped in Consequently the data pertaining to municipal expenditure prioritization by sector provide only an approximate picture of the main sectors municipalities are spending on. 76 Specific analysis in this chapter utilizes largely the RS data (given data problems for FBH) and where possible, introduces comparisons to FBH. 74. Expenditures for higher levels are net of onward transfers. 75. Cantons consolidate their respective municipalities for reporting to the FBH level. Efforts to directly collect municipal data for FBH yielded between observations out of Could be problematic if fourth quarter spending differed significantly from first three. 48

64 Figure 36: Aggregate Municipal Expenditures (2006, estimated) GDP 2006 Proj Municipal Revenues Municipal Revenues (Million KM) (Million KM) (% GDP) FH 11, % RS 5, % Total 17,533 1, % Source: GIB Data, IMF(2006) 4. Municipalities exhibit wide diversity in fiscal/revenue capacity. Figure 37 summarizes the averages and ranges of 2005 reported per capita revenues (budgets and realizations) and 2006 budgets, plus population figures. The lowest-revenue municipalities for 2005 budgets were Kalesija and Velika Kladusa, both FBH municipalities with under KM 77 per capita. The lowest revenue municipality in RS was Vukosavlje, with KM 88. Municipalities in RS reported the highest per capita revenue for the 2006 budget, likely related to their status as micro-municipalities. Istocni Drvar with 62 inhabitants had projected per capita revenues of over KM 10 and Petrovac/Drinic with 189 inhabitants, 3.5 thousand. The RS municipality of Bosanski Brod with 20,424 inhabitants reported per capita revenues of KM Figure 37: Municipal Diversity Mean Min Max Sd Per Capita Revenues (2005 Budget, KM) Per Capita Revenues (2005 Realizations, KM) , , Per Capita Revenues (2005 Budgets, KM) , ,080.8 Population 26, ,647 30,104 Area (km2) The wide range of per capita revenues suggests that some poor municipalities may be under funded in meeting minimum expenditure responsibilities, so their priority should be to enhance own-revenue collections or transfers, and to maximize expenditure efficiencies. Among the fiscally poorest 20 percent, per capita municipal incomes are below KM 150 (see Annex 3) and the impact of the new ITA equalization component on their revenue position remains to be seen. 4.3 Own-Source Revenues 6. Municipal own-source revenues comprise fees and charges. Figure 2 summarizes average municipal revenue structures. The so-called tax revenues (see Annex 1, #2-5) are taxes and entity transfers, and in FBH, cantonal transfers. Non-tax revenues (Annex 1, #17-30) are municipal own-source revenues, under municipal rate setting control, which are dominated by fees and charges. 49

65 7. Figure 38 plots the share of non-tax revenues as overall revenues for all RS municipalities against per capita total revenues. The graph highlights how this revenue base differs in relative importance across municipalities. 77 Figure 38: Non-Tax Revenues versus Per Capita Revenues Bosnia Municipal Own Revenue Mobilization 2006 Budgets Share of Non-Tax Revenues (%) Per Capita Total Revenues (KM) FBH Fitted values RS Fitted values Figure 39 summarizes municipal revenues in RS. Almost half of RS municipality revenues derived from sales/wage taxes, and since 2006, the single account transfers. Personal income tax sharing in 2005 accounted for just under 8.0 percent of realized revenues, higher-level government transfers, just over 70 percent, and the balance was own-source revenues such as fees and a small amount of capital gains. Given the more variable quality of the FBH data, a more direct comparison of differences in municipal revenues structure across the two entities was not possible, although the general patterns in revenue structure were similar. 77. Some municipalities projected revenues of over 2,000 KM per capita in 2006 budgets but these were omitted from the graph. They included PETROVAC/DRINIC (pop, 189, KM ), BOSANSKI BROD (pop , KM , and ISTOCNI DRVAR, pop 62, KM ). 50

66 Figure 39: Municipal Revenue Structure (RS percent) Type Budget Actuals Budget Actuals Budget Actuals Budget Actuals A. "Tax Revenues" Transfers Company income and profit taxes (4+5+6) Transfers Social security contributions Transfers Personal income tax Transfers Property tax Transfers Sales tax (VAT from Jan 2006) ( ) Transfers Taxes on international trade and transactions Transfers Other taxes B. "Non-Tax Revenues" Own Revenues from business activities, property and positive exchange rate differences (18 to 22) Own Fees and charges (24 to 28) Own Fines Own Other non-tax revenue Own C. Capital gains (32) C. Grants Transfers Current grants (39+40) Transfers Capital grants (42+43) Own E. Revenue from budgets of lower tier budgetary units TOTAL Recommendations 8. Short-term: Enhance own-source revenues. The sole own-source revenue base of municipalities is user fees, but enhancing revenues from the existing consumer base will require municipalities to upgrade service quality to improve user willingness to pay, increase revenue collection rates and require municipal legislatures to raise existing tariff rates to cost-recovery levels. Fee increases and better collection will be more politically palatable if the local leadership can offer examples of where this has been successful. But municipalities should be cautioned against raising revenues by levying nuisance taxes (hefty permits levies, for example) because this short-term gain is a disincentive for longer-term economic growth. Awareness campaigns and higher-level supervision are required to eliminate illegal taxes. 9. Medium term: develop more buoyant tax bases. All municipalities have benefited from the new allocation system, but richer urban municipalities may benefit less. However, they are well placed to raise user fees for urban services, an important medium-term agenda-for example, Brcko is promoting the introduction of local property tax. This requires attention to policy through legislation and implementation through administration, and done skillfully, will enhance both municipal revenue and accountability. 51

67 Box 7: Small steps to attract small business-ljubinje deregulates local fees Ljubinje Municipality is a small, predominantly rural municipality in Southern BH, in Republika Srpska Entity. Local administration is determined to use existing resources efficiently by creating conditions favorable to small- and medium-size enterprises (SMEs) and to attract investors. Last year the Municipality formed and funded an off-budget Agency for SME development that aims to build capacity and skills among local entrepreneurs and potential entrepreneurs, who will learn how to apply for funding, run a business, and form locallevel associations. Furthermore, the municipal administration in Ljubinje has simplified and shortened internal administrative procedures for business registration so that the applicant need only submit the documents and civil servants complete the process. In less than one week, the local administration issues a registration number, contacts the central statistical office in Banja Luka for a statistical ID number, and registers the business with the Tax Administration. To attract investors, the local administration lowered local communal fees that businesses have to pay, and decided to completely exempt start-up businesses from local taxes and fees. The Mayor said: We have decided that we need to attract businesses and investments here. This means all Ljubinje Municipality resources-land and everything else, should be used to attract capital and create employment. In addition, local government earmarks funds in the annual budget to subsidize agricultural producers and small businesses. As a result, the share of local tax revenues has risen to high levels in Ljubinje. 4.4 Fiscal Transfers 10. The newly implemented Single Account-based transfer is the single most important revenue source in municipalities (57 percent in FBH and 70 percent in RS in 2007). 78 Distribution is based on historical sales tax collections, and an equalization-based formula that will be phased in over several years (six years in FBH and nine years in RS). Prior to Single Account allocations, derivation-based sales tax sharing was the main revenue transfer component for municipalities. 79 For horizontal allocations, 2006 is a transitional year; formal ITA Single Account allocations were introduced mid-year and before that earlier versions of sales tax collection were used. In aggregate terms, RS municipalities enjoy a higher sharing rate of the entity ITA pool than FBH (24 versus 8.4 percent as per Figure 40). The difference is due to cantonal shares-over half of the Single Account transfers are in FBH Single Account Fiscal Transfer Allocations 11. Single Account allocations are assigned directly to municipalities. The annual Single Account pool dictates the amount of the vertical tranche for the municipal sector and from that each municipality receives their relative share. In FBH, the cantons also receive a share but revenues from the Single Account are allocated first across the state, FBH, and RS. 80 In 2006, several delays in ITA distribution arose from disputes about relative shares between entities, and the vertical distribution for municipalities and cantons (see Figure 40) In 2006, municipalities appear to have received 2006 VAT and 2005 VAT back payments. (GAP) 79. RS revenue sharing (income tax) is conditional on an under-development index. 80. Law on the System of Indirect Taxation stipulates the order of payments made from the Single Account: First payments go to the State Budget (in the amount of the adopted state budget), remaining funds are allocated to the entities and Brcko District (omitted in Figure 1). Before allocations go to the entities and Brcko District, foreign debt service is deducted and transferred directly to State Budget. State Budget has priority in financing, which is significant and may consider potential functional transfers from entity state level (as with defense, e.g., police structures). 81. This section is based on Basaric (2006) and Levitas, Tony, Jasmina Djikic, and Brankica Lenic (2006). Figure 40 percentages are from the December 2006 amendments to the RS Law on the Budgetary System, which were valid as of January March 2006 amendments allocated the RS share from ITA in proportion 73,5-23:3-5, which means that the December 2006 amendments increased municipal share by 1.0 percent; it also extended the transition period from six years to ten. 52

68 Figure 40: Vertical Shares from the Single Account Brcko District (3.6 %) Single Account ~550 State FBH (64.6 %) 3.9 RS (31.9 %) Cantons Municip alities Road Fund Municip alities Road Fund 12. Allocations have two components: legacy and new formula. Basic relative allocations across municipalities have a legacy component based on 2005 realized collections of wage and sales taxes, 82 a share that is decreasing over time, and a new formula component that follows: FBH = 0.68*Population *Area *Primary Pupils *Development Index RS = 0.75*Population *Area * Secondary Pupils 13. The FBH development index measures fiscal capacity, based on historical collection of sales and wage taxes, 83 and the FBH and RS formulae modify the final allocation. For example, FBH municipalities with over 60 thousand inhabitants receive more allocations (a 20 percent additional weight on their population). Previously, the RS had also imposed several hold harmless clauses. If a municipality were to receive less in absolute terms than before, its amount would be adjusted to ensure that it received just as much as previously in nominal terms. 84 However, the RS implementation module has changed. This clause was removed and it is now a 10 year implementation period with 10 percent changes every year. Three cantons in FBH receive an additional factor. Figure 41 presents the weight of the formula versus the historical sales tax collections as it is phased in over six years in FBH. The transition is slightly different for RS. In 2008 the formula and historical shares tax each account for half of the allocations, whereas by 2011 the allocations would be fully formula based. Figure 41: Phasing in Ratios for the Single Account distribution formula (percent) in FBH Formula Historical Sales Tax Collections In FBH, the new Law on Revenue Sharing was adopted in May 2006, and came into force in January In the interim legal vacuum, FBH adopted two Decisions that regulated vertical and horizontal distribution of indirect revenues and prescribed the amount to be shared was not to exceed 2005 indirect taxes, and obligated cantons to allocate to their municipalities an amount that matched 2005 indirect taxes. 83. The development index captures the amount a municipalities average per capita sales tax and wage tax deviates from average collection of these taxes. The less a municipality has of the FBH average (i.e., low development index), the larger coefficients it attached. 84. These provisions were deleted in the December 2007 amendments. 53

69 14. The new formula. The formulae are driven by population size, with a smaller relative weight given to geographic area and education enrollments, except for the development index in the FBH formula. This benefits larger municipalities because it assumes they need more services. However, the formula does not incorporate the notional fiscal capacity of municipalities-since most poor rural municipalities have a lower fiscal capacity, richer urban municipalities have an advantage. But this difference may not merit changes because, in the first place, proxy data for fiscal capacity-such as regional income-may be unavailable; second, in per capita terms, rural municipalities come out ahead, because they benefit from the land area factor; and third, urban municipalities expenditure needs may be larger because they need more infrastructure, for example. 15. Legacy and new formula components create more equitable distribution. Figure 42 and Figure 43 plot municipal receipts based on the ratio of the legacy share of total transfers versus formula transfers of the overall pool. A municipality with a ratio greater than one will be receiving more transfers from the increasing weight in the formula. Municipalities that received substantial sales and wage tax sharing will became less dependant on the transfers under the new formula, such as large municipalities including Mostar, Tuzla, and Banja Luka, because their relative share of historical wage and sales tax allocations is greater than what they would receive based on their shares under revenue allocation laws. Figure 43 suggests that only 10 of 63 RS municipalities are becoming less dependant by this measure but these 10 accounted for over 46 percent of total Single Account transfers in 2006 based largely on the legacy sales allocations. Banja Luka alone accounted for 26 percent of these transfers. However, after the full phase-in for RS over 10 years, its relative share is anticipated to decline to 14 percent. Similarly, in FBH, 22 of the 71 municipalities are becoming less dependant on the Single Account transfers (Figure 44). Figure 42: Formula versus Legacy in Municipal Single Account Allocations (FBH) Figure 43: Formula versus Legacy in Municipal Single Account Allocations (RS) 54

70 Figure 44: Formula versus Legacy in Municipal Single Account Allocations (FBH) 16. The growth of the Single Account pool, and the transition formula allocations. In 2006 buoyant ITA collections created windfalls which exceeded even the most optimistic expectations. Similarly, revenue growth was substantial in 2007 and But this trend may not continue. The transition over the next five years creates dynamic annual allocations. For example, a municipality such as Mostar would receive a small share of a growing overall ITA. A stagnant ITA, plus a declining ITA allocation might spark a lobby for a more favorable formula, a delayed new formula phasein, or discretionary allocations. If ITA shares fall short of expenditure pressures, for instance, for the special needs of urban or peri-urban municipalities with large displaced populations, or poor municipal financial management such as overstaffing, municipalities may compensate with near-sighted, short-term, own-source revenue seeking to the detriment of local economic growth-for example, licenses fees for new housing, and hefty permits for business startups. Recommendations 17. The legislative framework that determines transfers from the Single Account has been in place and should be respected. Single Account transfers are introduced directly from entity to municipality in both FBH and RS, ensuring that sub-entity governments have a more predictable and transparent revenue stream. Overall, the formula design seems to promise greater equity in financing basic municipal service delivery. The Ministries of Finance of both entities should respect within-entity distribution formula and monitor the dynamics during the transition period. 55

71 4.5 Expenditure Composition and Efficiency 18. Larger municipalities stand to lose from Single Account transfers. Over time, large urban municipalities that enjoyed the bulk of sales/wage tax will lose out in relative terms and, given the overall buoyancy of VAT revenues, may also lose out in absolute terms, even when adjusted for inflation. To compensate in the medium term, these municipalities will have to increase own-source revenues or increase yields from other bases; and all municipalities could realize gains from assessing their current expenditure prioritization and efficiency. 19. Municipalities have options to enhance fiscal space on the expenditure side. This study revealed several options to reduce unnecessary expenditures. For example, a quarter of case study municipalities reported having to fill expenditure gaps left by higher levels of government, notably in health and education. Clarifying roles and responsibilities would yield savings but the potential magnitude of these savings as a share of total municipal expenditures was not quantified. Furthermore, the prevalence of stop-gap transfers to affiliated public sector enterprises such as utilities reveal that some municipalities subsidize inefficient service providers operations that do not cover costs. Over the longer term this creates investment gaps for existing infrastructure, and could also signal poor capital expenditure prioritization. Strengthened cost recovery will increase own-source revenue and open space for alternative expenditures. 20. On average, wages consume significant portion of municipal expenditures (29% in FBH and 19% in RS). Most municipalities wages are under 40 percent of total annual expenditures but some spend between 40 and 60 percent, due to staff size and wage levels. Figure 45 suggests that the wage share is more in poorer municipalities. Anecdotal evidence suggests that public sector wages can differ among municipal staff, entity staff, and federal staff but the federal level appears to be the reference point. Figure 45: Wage Share against Per Capita Revenues Share of Wages in Total Revenues (%) Bosnia Municipal Wage Expenditure Pressures 2006 Budgets Per Capita Total Revenues (KM) FBH Fitted values RS Fitted values 56

72 21. Recurrent and capital expenditures, including transfers. On average, RS municipalities realized 25 percent of total expenditures on materials and service expenses (which includes travel, energy, utilities/communications, materials, fuels). In 2005, 36 percent of this category was accounted for by contracted services, and grants and transfers accounted for an average of 22 percent of total realizations, of which one-third constituted transfers to individuals, suggesting that a significant part of municipal expenditures are used for social safety nets. Municipalities reported that half of transfers went to non-profit organizations, presumably for public facilities such as sports venues and libraries. On average, transfers to private companies accounted for only 3.0 percent and public companies, 2.0 percent, although well over half of all municipalities reported transfers to public companies. However, in notable outliers such as Ribnik and Srbac in RS and Vogosca, more than 50 percent of 2005 transfers were spent on public enterprises. Transfers to private companies were more limited, only about one-third (20 of the 62 RS municipalities) reporting these. 22. Municipalities deployed direct capital expenditures or transfers to MZ and other enterprises. Figure 46 presents reported per capita capital expenditure realizations across all RS municipalities. 85 In 2005, all municipalities reported some form of direct capital expenditures, mostly to acquire equipment and buildings (56 and 43 of 62 municipalities in RS) respectively. As expected, the largest per capita magnitude was for buildings (79 KM per capita). Just over half reported expenditures for reconstruction and maintenance. Imprecise data make it difficult to establish whether other municipalities are neglecting maintenance at the expense of new capital projects. Just over half of municipalities reported capital transfers, albeit limited (20 KM per capita). Only one-quarter reported capital grants to lower tiers of government (i.e., MZs), while just under half reported making grants to other organizations-possibly utilities and other municipal bodies. Figure 46: Selected Capital Expenditures by Municipalities (RS, 2005) Average Highest Lowest Reporting (KM PC) (KM PC) (KM PC) # Fixed asset acquisition expenses Acquisition of land, forest and plantations Acquisition of buildings Acquisition of equipment Acquisition of other fixed assets Acquisition of long-term assets in the form of rights Reconstruction and maintenance Non-material investment Capital grants to other tiers of government Capital grants to other tiers of government Capital grants to individuals and non-profit organizations Total Revenues Reporting by sectoral/functional classification is inadequate. Figure 47 summarizes RS expenditure categories by major functional classification. The GIB data omit detailed expenditure breakdown reporting, and no 2006 functional classifications were yet reported at the time of writing; reporting problems result from prevailing administrative and accounting procedures. About half of expenditures are assigned to general services, with a slight downward trend over time; housing and communal affairs, and economic affairs appear to increase over time; and notably, about five percent of expenditures are assigned to education. 85. FBH data were incomplete, but the partial results presented generally consistent findings. 57

73 Figure 47: Expenditures by Functional Classification (percent) in RS Budget Actuals Budget Actuals Budget Actuals Budget Actuals General public services na na na Defense na na na Public order and safety na na na Economic affairs na na na Environment protection na na na Housing and communal services na na na Healthcare na na na Recreation, culture and religion na na na Education na na na Social security na na na Total Effective municipal expenditure requires a well-functioning PFM process. Aggregate data provide only the most basic sense of where municipalities could be enhancing the service delivery impact of their expenditures. More refined expenditure prioritization and implementation emerges from a high-quality planning and budgeting process that ensures value-for-money procurement (Chapter 5). Recommendations 25. Increasing public resources available to municipalities (see above) will require consistent reporting on allocations and implementation of these resources-this remains a priority. Without basic efforts to strengthen this information, municipal managers and national stakeholders (e.g., Ministry of Finance, local government) will be unable to examine evolving patterns, much less link these to differing municipal outcome measures. Support is needed, similar to initial efforts by USAID GAP, to introduce basic financial management information systems at the municipal level, which can facilitate consistent reporting across municipalities. 4.6 Municipal Borrowing 27. Until recently, municipal borrowing was restricted in both entities. Through 2006, borrowing in FBH was restricted by annual budget laws. The most recent RS law imposes limits on municipalities based on debt servicing ratios. The Law on Indebtedness, Debt and Guarantees of the Republika Srpska (passed in May 2007) allows municipalities to borrow if servicing does not exceed 18 percent (increased from ten percent) of the previous year s own-source revenues plus transfers. Furthermore, municipalities can issue guarantees (e.g., for associated public enterprises) up to 30 percent of the previous year s revenues. The FBH Law on Indebtedness, Debt and Guarantees allows municipalities to take long term debt only under the conditions that yearly debt service of all loans, including loans incurred by their enterprises for which municipalities issued guarantees, does not exceed 10 % of revenues collected in previous fiscal year. However, projected Single Account transfers to FBH cantons will likely increase their borrowing room in aggregate volume over that of municipalities. 28. Reported municipal debt stocks in RS are an estimated 2.5 percent of RS s GDP or 27 percent of RS revenues in The RS Ministry of Finance reports total municipal debt balances just under KM 150 million, excluding guarantees. However, only 33 million of this debt appears to have been actually incurred. Moreover, the Ministry appears to have approved a further KM 22 million for guarantees. Detailed data of debt servicing (interest plus principal) are not available. Assuming a overall debt service ratio of 20 percent, this would allow for a debt stock of RS 2.8 billion, suggesting that on aggregate, based on reported approvals, RS municipalities are well within sanctioned limits, and hence have a potential to enhance their fiscal space within these paramaters. 58

74 The highest leveraging appeared to be Srbac (3.5 times debt stock to 2005 revenues). Eight municipalities reported debt stocks exceeding 2005 revenues: Oš{tra Luka (1.6), Novigrad (1.5), Nevesinje (1.34), Kne `evo (1.23), Petrovo (1.17), Prnjavor (1.15), Laktaši (1.05), Gradiška (1.03). 28. Long-term financing appears to dominate municipal borrowing. The GIB data distinguish between long- and short-term borrowings, although criteria for this distinction are unclear. The GIB data suggest that RS municipalities borrowed KM 5.6 million in 2005, and FBH municipalities, KM 4.3 million. In FBH municipalities less than 2.0 percent of this was short-term; in RS, 15 percent of municipalities reported short-term borrowing. Reported long-term borrowing was derived mainly from domestic sources. These data raise the question of whether FBH borrowing occurred inside or outside the legal framework, which until recently strongly restricted borrowing. 29. FBH fiscal data quality is especially poor but has seen significant improvement. The GIB forms pass through the FBH cantons and are then sent to the entity level where all municipalities in that canton are aggregated, but quality control for this reporting is poor. Cantons exhibit apathy about reporting, which may set the tone at the municipal level. The introduction of the VAT, and the phasing in of the revenue allocation formula from the Single Account, is likely to affect the borrowing position of different types of municipalities. In some cases, several municipalities may also build up initial surpluses given the windfall transfers in In relative terms, richer urban municipalities will increasingly have to rely on own fiscal capacity and cost-recovery. Borrowing can be effective to promote much needed capital investments at the local level. However, in many cases this presumes the ability of municipalities to adequately assess and prioritize potential projects. Recommendations 30. Ministries of Finance at all levels should strengthen its ability to monitor fiscal risks and help municipalities to be on a sounder financial footing. Until recently, borrowing has been frozen in FBH municipalities except through multi-laterals, e.g. World Bank, although the GIB data suggest that some municipalities have incurred liabilities. A priority is to clarify which FBH entity-level agency is responsible for monitoring and approving municipal borrowing, ideally linked to enhanced monitoring of municipal fiscal data, as currently introduced in the FBH MoF. In FBH, the cantons will also need to be monitored, given their significant fiscal space in the area of borrowing under these measures. Similar efforts could be directed at strengthening the already established system of monitoring at RS. Better periodic monitoring should also dispel concerns that municipalities may be too rapidly exploiting the fiscal space for borrowing as set out in the legal limits. 31. Identify capacity gaps and needs for municipal capital projects and borrowing. Particularly in FBH, the market for municipal borrowing has only recently opened. BH has a relatively well developed and competitive banking sector. Previous donor support (e.g., LDP) has already sought to strengthen this market, and several banks have already expressed more interest in this market (including through cross-marketing with other services). Therefore, it may be useful to better understand the project preparation needs across a range of municipalities, and how this demand can be best supported. At the same time, borrowing tenders are subject to prevailing procurement legislation. To achieve the best value for money, greater transparency and dissemination of these standards across both executive, legislative, and other municipal stakeholders could be supported. The international community can provide technical assistance to municipalities to strengthen their public financial management as well as capital investment planning and project appraisal capacities, which would enable them to take advantage of borrowing. 59

75 4.7 Conclusions 32. Municipal priorities will differ. When BH municipalities fiscal positions are analyzed using available data on municipal budget levels, structures, and trends, there are wide differences in finances and expenditures, which will affect priorities. 33. The impact of state/entity level reforms on municipalities will differ. For underfunded municipalities, the phasing in and targeting of the ITA transfers are crucial. Property tax reforms are a medium-term proposition but will have the largest impact on richer municipalities. For most municipalities, formal borrowing limits are acceptable. Given the recent relaxation on borrowing in FBH, significant effort will be needed to enhance monitoring of borrowing patterns for those municipalities. 34. Aggregate budget GIB data provide a limited view of fiscal space and options. In the functional area, GIB data reveal whether municipalities were spending on unassigned functions. The original budget data (OBD) for individual municipalities differ on reporting structures, and even by year, which prevents a systematic integration of the data. A more systematic analysis of the fiscal space of BH municipalities is significantly constrained by the absence of effective reporting systems and processes. Special efforts are needed to convince key stakeholders of the value of improving these reporting systems. First, municipalities themselves must accept the value of good fiscal reporting not only to their own executive and legislative, but also their ability to compare themselves to other municipalities. Second, international development partners will need better information on the fiscal capacity, and linked therewith performance, of individual municipalities to better target and monitor their assistance over time. Finally, entity and state level agencies, notably Ministries of Finance, must better appreciate the value of this type of reporting, given the increasing resources being channeled to the municipal sector and their oversight responsibilities with respect to these resources. 60

76 Chapter 5: Public Financial Management Performance across Municipalities 5.1 The Why and What of Sub-national PFM 1. Sub-national public financial management (PFM) practices-from budgeting, planning, execution, to external scrutiny and monitoring-expose a set of cross-cutting issues that contribute to variations of service delivery outcomes. Since service delivery is increasingly decentralized, weak PFM functions and local government institutions are likely to undermine benefits inherent in bringing public services closer to local communities, including broad-based accountability. This section summarizes a systematic assessment of PFM outcomes in 20 municipalities-the first attempt in BH, and discusses reforms that could improve the situation The enabling environment and local policymaker actions shape municipal PFM outcomes. Examining the main drivers for PFM outcomes in a Sub-national context is the best method of discovering where potential exists for improvements. For example, some outcomes are primarily driven by the overall intergovernmental enabling context-crucial budget transfers might become empty promises, or the entity government may fail to conduct external audits. Other outcomes are driven primarily by municipal behavior or capacity. For example, despite timely announcements on transfers from higher levels of government, significant delays in the annual budget approval occur due to gridlock among political parties that prevents timely passage of the municipal budget. In this context, the Municipal Administration (MA) and Municipal Council (MC) must be differentiated. 86. The Public Expenditure and Financial Management (PEFA) performance indicator framework guided the assessment of individual municipal Public Financial Management (PFM) outcomes; the framework covers 28 indicators, plus three for donor behavior (See Annex 6.1 for a summary of PEFA framework). Typically, the indicators have been applied to national contexts, but are increasingly used at the Sub-national level. The report adopts a streamlined assessment framework to assess PFM performance in 20 case municipalities. While the PEFA indicators are typically graded on an A-D scale (PEFA 2005), for this pilot phase, the team used some indicators on a yes/no basis, and where possible collected quantitative data (e.g., deviations from planned to realized budgets), and noted qualitative dynamics in each area (including factors/ issues that may not have been considered in advance of the field work). The objective of the PEFA indicators is to better assess why outcomes are occurring, not simply to score them. 61

77 Summary of PEFA The Public Expenditure and Financial Accountability (PEFA) Framework is intended to: (i) provide integrated assessment of performance of national PFM systems; (ii)demonstrate progress in PFM performance over time; (iii) enable regular, rigorous, evidence-based monitoring by domestic and international stakeholders; and (iv) facilitate harmonization through wide international acceptance. This framework identifies six critical dimensions of performance of an open and orderly PFM system as follows: 1. Credibility of the budget - The budget is realistic and is implemented as intended 2. Comprehensiveness and transparency - The budget and the fiscal risk oversight are comprehensive, and fiscal and budget information is accessible to the public. 3. Policy-based budgeting - The budget is prepared with due regard to government policy. 4. Predictability and control in budget execution - The budget is implemented in an orderly and predictable manner and there are arrangements for the exercise of control and stewardship in the use of public funds. 5. Accounting, recording and reporting - Adequate records and information are produced, maintained and disseminated to meet decision-making control, management and reporting purposes. 6. External scrutiny and audit - Arrangements for scrutiny of public finances and follow up by executive are operating. Against the six core dimensions of PFM performance, 28 high-level indicators of country PFM performance are structured into three categories: A. PFM system out-turns: the immediate results of the PFM system in terms of actual expenditures and revenues by comparing them to the original approved budget, as well as level of and changes in expenditure arrears. B. Cross-cutting features of the PFM system: the comprehensiveness and transparency of the PFM system across the whole of the budget cycle. C. Budget cycle: the performance of the key systems, processes and institutions within the budget cycle of the central government. In addition, this framework also includes 3 donor indicators D. Donor practices: elements of donor practices which impact the performance of country PFM system. C. Budget Cycle D. Donor Practices Policy Based budgeting A. PFM Out-turns External scrutiny and audit B. Cross-cutting features Comprehensiveness and Transparency Predictability and control in Budget Execution Budget credibility Accounting, Recording, Reporting 62

78 3. The enabling environment consists of legislation, regulations, guidance, and instructions at State, entity, and cantonal (in FBH) and municipal levels. Recent legislative and regulatory reforms have improved public financial management performance, and some areas, for example, Subnational borrowing, internal audit, have pending legislation or policy in development. Table 2 summarizes key legislation that governs BH public financial management for municipalities, and highlights circumstances where PFM arrangements differ between the Federation and RS. 87 Figure 48: Key Regulations Concerning Municipal PFM State FBH RS Credibility of the budget Comprehensiveness and transparency Policy-based budgeting Predictability and control in budget execution BH Law on Payments into a Single Account and Revenues (Dec 2004) Law on Public Procurement FBH Organic Budget Law 2006 FBH Law on Treasury FBH Law on Principles of Local Self-Governance FBH Laws on Revenue Allocation Cantonal Laws on Revenue Allocation Law on Debt, Indebtedness, and Guarantees of the Federation of Bosnia and Herzegovina RS Budget System Law Law on Treasury System in RS The RS Law on Local Self- Governance Relevant bylaws, rulebooks and instructions by MoFs Law on Indebtedness, Debt and Guarantees of Republika Srpska, Accounting, recording and reporting FBH Laws on Accounting and Auditing Law on Internal Audit in Pubulic Sector in FBH )(Official Gazette 48/08) RS Laws on Accounting and Auditing Law on Internal Audit in Public Sector in RS (Official Gazette 17/08) External scrutiny and audit Law on Audit of institutions of FBH (Official Gazette 22/06) Law on Audit in Public Sector in RS (Official Gazette 98/05) Donor practices Source: See Annex 3.3 for details. 5.2 Key Findings and Summary of PFM Assessment 4. Recent reforms on financial management have achieved some success. Bosnia and Herzegovina has established fundamental PFM systems for managing macro-fiscal developments, some elements for enabling strategic allocation of resources, and a few tools for improving operational/technical efficiency.88 However, the fragmented and evolving policy and legal/regulatory environment has handicapped policy design and implementation at the local level, which requires strengthening the link between legislation/regulation and implementation, and intensive but customized capacity building at all levels. 5. Better PFM performance is associated with better overall service outcomes across 20 municipalities (Figure 49). Exceptions are Odzak and Ljubinje, ranking 6 and 8 in service outcomes but only 16 and 18 on PFM scores. Similarly, Buzim is 2 in PFM measures but only 15 on service outcomes. PFM performance does not appear to correlate strongly with per capital revenues or population size (Figure 50 and Figure 51). In general, older municipalities have better PFM scores than newly established municipalities (Figure 52). 87. Full references to laws and background documents are found in the report bibliography. 88. See World Bank, 2006, A Fiduciary Update on Public Financial Management Bosnia and Herzegovina. 63

79 Figure 49: Service Outcome Score and PFM Scores BERKOVICI STOLAC MOSTAR KALESIJA ISTOCNI STARI BUZIM VISEGRAD ZVORNIK RIBNIK BREZA JABLANICA SIROKI BRIJEG LJUBINJE ODZAK DOBOJ-JUG SAMAC NOVO SARAJEVO TUZLA BANJA LUKA ILIDZA Service Outcome Score PFM Score (28+3 indicators) Figure 50: PFM Scores and Population Size PFM Score and Population Size PFM Score , , , , ,000 Population Size 64

80 Figure 51: PFM Scores, Service Outcome Score, and Per Capita Revenue PFM Score and Service Outcome Score PFM Score Service Outcome Score Low pcrev Midium pcrev High pcrev Figure 52: PFM Scores, Service Outcome Score, and Legacy PFM Score and Service Outcome Score PFM Score Service Outcome Score New Municipality Old Municipality 65

81 6. PFM performance varies greatly across municipalities (see Annex 6.3 for spider diagrams), indicating the need for a tailored PFM capacity-building strategy grounded in municipallevel analysis. Breza PFM Results Donor Practices External Scrutiny & Audit Budget Credibility 80% 40% 0% -40% -80% Comprehensiveness & Transparency Policy-Based Budgeting Accounting, Recording & Reporting Predictability & Control in Budget Execution Ilidza PFM Results Donor Practices External Scrutiny & Audit Budget Credibility 80% 40% 0% -40% -80% Comprehensiveness & Transparency Policy-Based Budgeting Accounting, Recording & Reporting Predictability & Control in Budget Execution 66

82 Mostar PFM Results Donor Practices External Scrutiny & Audit Budget Credibility 80% 40% 0% -40% -80% Comprehensiveness & Transparency Policy-Based Budgeting Accounting, Recording & Reporting Predictability & Control in Budget Execution Berkovici PFM Results Donor Practices External Scrutiny & Audit Budget Credibility 80% 40% 0% -40% -80% Comprehensiveness & Transparency Policy-Based Budgeting Accounting, Recording & Reporting Predictability & Control in Budget Execution Credibility of the Budget 7. Further reforms are urgent to enhance budget credibility so that the budget can be trusted by users who will implement it and deliver services, and by citizens and businesses. Reforms need a two-pronged approach: first, improve municipal technical capacity to carry out revenue forecasting (in particular for own-source revenues), and second, ensure predictability of shared revenues and transfers from higher level government. 8. For most municipalities, municipal budget credibility-measured as the difference between approved budgets and realizations-was weak and has not changed appreciably over the past three years. In at least one of the past three years ( ), three quarters of municipalities had either overspent, or failed to spend, their original budget-a variation of larger than five percent in absolute terms.89 Admittedly, by more lenient standards, which sets the reference threshold at ten and fifteen percent respectively (or B-D grades under the full PEFA framework), the number of below-par performers is more limited. Some 12 of the 20 municipalities exceeded a 15 percent deviation no more than one year over the three year period. 89. The average absolute deviations were 10.3 percent in 2004, 10.7 in 2005, and 18.5 in

83 9. Internal and external factors contribute to weak budget predictability. 90 The overall intergovernmental enabling context and weak municipal capacity in budget forecasting, execution, and scrutiny contribute to weak budget predictability. Since transfers and grants comprise two thirds of total municipal revenue, municipal budget credibility to a large degree depends on higher-level governments providing municipal governments with reliable and timely forward estimates of annual transfers and grants, and most importantly, to accept accountability for amounts promised. However, the transfers and grants have become the biggest source of unpredictability, with an average 54 percent absolute deviation during In 2006, the government envisioned the VAT would provide municipalities with more predicable transfers directly from the state level. However, the first year of implementing Single Account transfers eroded budget predictability. First, distribution coefficients were published too late, which meant that municipalities prepared their 2006 budget based on past revenue proceeds, and entity governments conservative guidelines suggesting expected revenue outturns for Later, disputes over the single account revenue share across the two entities and the Brcko district frequently delayed transfers from indirect taxation in 2006, resulting in a revenue windfall in the first quarter of the year, with revenues being transferred in bulk at the end of March. Actual municipal expenditures significantly exceeded budgeted expenditures, since by then municipalities had already significant arrears to be covered, such as wages for example. Figure 53: Average Absolute Deviation of Primary Budgets for 20 Municipalities % 2004 % 2005 Total Expenditure Operating Expenditure Capital Expenditure Total Revenue Tax-revenue Non-tax-revenue Transfers and Grants % Compounding the problem, most municipalities exhibit weak capacity for revenue forecasting, as demonstrated by aggregate revenue out-turns compared to original budgets. Most municipalities base their revenue forecasts on the previous year s revenue, often using nine month figures to meet requirements of the budget calendar. Few municipalities had forward-looking perspectives that considered fiscal capacity or local revenue mobilization potential. Compared to tax revenue forecasting, municipalities tend to be optimistic about non-tax revenues collection, 93 leading to large deviations in non-tax revenue outturns averaging 21 percent absolute deviation. Delays or poor information on higher-level transfers, 94 especially capital transfers, and unclear and everchanging revenue-sharing arrangements made municipal revenue protection problematic.95 Community contributions to capital investment are by nature less predicable, thus adding to poor budget projection. 11. Many factors contribute to expenditure deviations. Capital expenditure deviations resulted from (i) unclear entity-and cantonal-level rules of capital project selection; (ii) failure to transfer matching funds; (iii) unexpected/unplanned transfers to capital projects; (iv) donor programs; and (v) loan refinancing. Operating expenditure deviations resulted from (i) reductions to operational outlays to meet national and international standards; (ii) financing deficit and capital expenditure; (iii) increased salary base by higher-level government (from 86KM to 102KM in RS); and (iv) difficulty shedding employees during unsuccessful/unfinished local administration reforms, or political changes in government. 90. Please note that this analysis was based on data collected as part of the PFM assessment. Some experts suggest that in general the revenue projections by OMS are filtered down to municipalities in a timely manner. However, in FBH, as the communication from the Ministry of Finance to municipalities goes through the cantons, which are not quick enough with information sharing. In fact, the projections provided by OMA are often too conservative every year, hence loosing credibility. 91. Compared to average deviation of 16 percent on tax-revenue and 21 on non-tax revenue during the same period. 92. For example, the RS Entity Government suggested revenue increases of about 5.0 percent to its municipalities. 93. Non-tax revenues include fees from flooded land, utilities, construction land, business premises, compensations, and other fees and charges. 94. Such as Doboj jug and Ilidza. 95. For instance, Una-Sana Canton and Herzegovina Neretva Canton. 68

84 12. New responsibilities pressure municipalities to switch expenditure priorities. A fraction of municipalities received explicit new expenditure responsibilities in the past three years. For example, Mostar, Jablanica of Herzegovina-Neretva Stolac Canton and Buzim of Canton Una-Sana, and Ljubinje in the RS, adopted a Law on Social Protection in 2005, which delegates some social assistance functions to municipalities, but no additional resources were transferred. A quarter of municipalities reported having to fill the funding gap left by higher-level governments, primarily in education, health care, social protection, welfare, and in one case, arrears for the sport and cultural center operated by the cantonal government. 13. One-third of municipalities have accumulated expenditure payment arrears. 96 Three municipalities have incurred salary arrears. Of seven municipalities, most have records, but only three of them know the stock of their arrears (2.0, 3.0, and 21 percent) with an average of 8.7 percent of total expenditure. This diminishes municipal ability to invest locally and remain financially independent, which may reveal a need to expand the budget envelope for some. 14. Lawsuits against local governments are potentially large liabilities. Legal liabilities threaten budget credibility and disrupt municipal operations. About three-quarters of municipalities have been sued. 97,98 Some municipal accounts have been blocked; some revenues have been redirected to creditors or plaintiffs (e.g., Stolac). In 2006, the FBH government amended the Law on Executive Procedure, ruling that municipalities and cantons that are under court rulings must pay their debt/arrears and must set aside a minimum of five percent of revenues for this purpose. This amendment assures creditors and plaintiffs of recouping their money, and protects municipalities and cantons from potential bankruptcy or operational disruptions. 15. Case studies suggest some positive trends since the 2004 election. Qualitative interviews suggest that mayors in municipalities such as Ribnik, Stolac, Zvornik made budget credibility an important platform for their new governments after the 2004 elections. Before 2004, Ribnik, for example, had large accumulated arrears and no budget approved by the local council. The new mayor cleaned up the books and restored budget credibility by paying off arrears and working with local council on a realistic budget Budget comprehensiveness and transparency 16. Improved functional classification is essential. Budget classification lacks comprehensiveness and functional classification is particularly weak. As a result, local governments have been unable to assess the strategic allocation of budget resources among economic sectors. Municipalities and higher level governments cannot track spending efficiency and effectiveness of services. In the short run, the Ministries of Finance in both entities must clarify functional classification and provide further technical assistance to improve the policy focus in resource allocation. In the long run, budget classification and accounting systems should be structured and aligned to enable performance orientation, which will also require service outcomes to be monitored and measured. 17. Budget classification in BH, as defined in Budget Laws in both entities, 99 allows spending to be tracked at all levels by administrative and, economic classification but functional classification is yet to be introduced at the municipal level. Despite the introduction of the automated treasury system at all levels of government, there is no unified Chart of account and those used by three tiers of government are not fully compliant with international public sector accounting standard. This would raise a concern regarding compatibility of financial information and overall transparency of public spending Budget Laws in both entities regulate expenditure payment arrears. In the FBH, budget users shall not incur arrears without Ministry of Finance approval. In RS, legislation forbids arrears at all levels, although Berkovici has accumulated some arrears. 97. Some municipalities started to levy local taxes in an effort to raise local revenues. These taxes were later ruled illegal by constitutional court (e.g., Siroki Brijeg s turnover tax on a meat packing factory). 98. Most keep a record of all court suits, half of municipalities plan to repay the arrears and lost cases. 99. In the Federation, FBH Law on Accounting and Auditing legislation and government Decree (and subsequent Rulebook) also provide framework for Chart of Accounts OSCE has partnered with USAID and Sida funded Governance Accountability Project (GAP) and developed a Budget and Finance Guide as a reference document for Mayors, Heads of Finance, councilors and citizens in all BH municipalities. 69

85 18. Significant deviations from the Charts of Accounts exist in actual budget reporting at the municipal level in both entities. Economic classification is common practice but functional classification and administrative classification are less popular. Increasing numbers of municipalities use all three classifications (from 5 in 2004 to 11 in 2006). Weak local government compliance on functional classification in part reflects a lack of clear instruction and clarification from higher-level government. Both entities require GIB reporting, however some municipalities object to the narrowness of GIB forms, which has led to a reduction of reported items (about half of those reported to the municipal council). 19. Budget documentation information is increasingly comprehensive. Although the amount of budget information submitted to the local legislature is rising, the level of unreported extra-budgetary government operations is unclear. Half of the 20 sample municipalities have included five or six of the requested six elements of key information-prior year s budget outturn, budget proposal, fiscal deficit, deficit financing, and debt stock (when applicable, see Annex 3.2, PI 6). However, about half of municipalities failed to explain the budget implications of new policy initiatives, and about half have unreported government operations for which no data were available on composition or magnitude. Box 8: Ugljevik: Budgeting based on citizens needs. Ugljevik is situated in the Semberija region, North-East Republika Srpska. It has an estimated 17,000 citizens, down from the 1991 census, which recorded 25,645 citizens. The RS government categorizes Ugjevik as a developed municipality thanks to a coal mine and an electric plant that is the largest source of municipal revenue. Per a RS Government decree, these revenues from concessions must be used for capital investments and development projects. The local government understands that sustainable municipal development is enhanced through participatory planning. This philosophy, combined with the revenues for capital investments and development projects created favorable conditions in Ugljevik, where Municipal authorities have established a model of drafting the budget with citizen inputs. The first step in drawing up the budget is to consult MZ representatives and conduct a survey of citizen opinions. The Commission for Development within the Municipal administration evaluates the survey results and the MZ inputs using criteria set by MZ Councils and citizens to ensure all projects have equal status and priorities are selected democratically. This participatory approach to budget setting was recognized as a model for BH. 20. Over three-quarters of municipalities provide support to their MZs, but how this support is managed varies. MZs have been established in almost all municipalities in the sample. 101 The role of local communities (MZs), as subdivisions of municipalities, have recently been clarified and institutionalized through the Law on Principles of Local Self-Government in FBH and the RS Law on Local Self-Governance. 102 In practice, over half of municipalities have passed local legislation to regulate MZs activities. Although financing is largely secured for MZs operating costs, most transfers are ad hoc; only two municipalities transfer the funds according to objective formulae (Zvornik and Visegrad). Only half of MZs account for their spending, less than half have a system of auditing. 21. Municipal government subsidies and guarantees to their public enterprises may create fiscal risk. Only three municipalities have legislation that covers aspects of financial management and reporting for locally owned public enterprises. 103 Municipalities subsidize capital expenditures for their utility and for-profit-companies (SOEs), and the enterprises self-finance for recurrent expenditures through user tariff collection or sale of goods and services. All municipalities reported that enterprise subsidies are recorded but the detailed municipal budget is unclear about how much capital expenditure municipalities allocated to their enterprises and it is impossible to track. Municipalities have also issued guarantees to their public enterprises, creating potential fiscal risk. 22. Weak oversight of public enterprises poses fiscal risks to local governments. Few enterprises have internal audits but most municipalities have external audits on the annual financial statements of local enterprises/institutions. However, Municipal Finance Directors interviewed were often unaware of external audit results. Most local legislation on financial management excludes local enterprise reporting Except Stolac and Mostar Local Self-Governance Laws do not specify MZs revenue composition. In the FBH, some revenues include self-financing, service-provision fees, gifts, and other resources intended for use in MZ plans and programs Although three-quarters of municipalities have legislation on locally owned public enterprises. 70

86 23. Municipalities do not understand relevant entity-level laws. The Laws on Public Enterprises (2006) newly introduced in both entities regulate the ownership, governance, and financing of public sector companies. However, considerable confusion/ambiguity surrounds the financial relationship between the utilities and municipalities, which creates potential municipal liabilities and threatens to erode service delivery outcomes. Furthermore, legislation and practice are disconnected, in part due to a failure to disseminate and clarify the new laws 104. In RS, the 2007 update on the Law on Indebtedness, Debt and Guarantees provides the framework to oversee aggregate fiscal risk. In FBH, a similar law was adopted by the parliament in December However, considerable confusion still exists about municipal mandates to issue guarantees and borrow from the credit market. 24. BH municipalities provide reasonable public access to fiscal information. All municipalities hold public hearings, often in MZs, on the draft budget before it is submitted to the local council, and sometimes hearings are broadcast by local radio or television. All municipalities publish the finalized budget in official Gazettes; 19 out of 20 do so using their notice boards. A quarter of the municipalities published their annual budget on-line, although the level of detail varies widely. All 20 municipal Finance Directors prepare quarterly in-year reports for local councils on budget execution and these are made available to citizens by request Policy-Based Budgeting 25. Little potential for multi-year budgeting and strategic sectoral planning at the local level. Although the basic enabling environment and tools for policy-based budgeting are in place and the budget process is orderly, multi-year perspective and strategic sectoral planning are hampered, particularly at the municipal level, by unpredictability of transfers from upper-level governments, and lack of local capacity. 105 Capacity needs to be built at all levels to translate government policy priorities into annual budget preparation, which requires a longer-term perspective. 26. Both entities have clear annual budget calendars. Timelines for the budget preparation activities of local governments are defined in the FBH Organic Budget Law and RS Budget System Law. Over the past three years, most municipalities passed their budgets before the start of the fiscal year, although slippage has occurred. Qualitative interviews suggest that it is often difficult to set strategic objectives and get annual budget passed easily if the mayor belongs to a political party different from the majority party/coalition in the council Enabling elements exist for strategic allocation of resources but practice lags. In both entities the budget framework papers (BFPs) 107 cover a three-year period, set out key revenue projections, expenditure policies and strategies to guide budget preparation for the entity, cantons, and municipalities. Legislation requires municipalities to prepare their own BFPs. Close to three-quarters of municipalities claimed to have done so but only about half of them have documentation of medium-term priorities. Moreover, in only half of municipalities do final sector plans reflect development priorities from the medium-term development plan, despite participation of local councils to set sector priorities and programs in most municipalities Predictability and control in budget execution. 28. The cantonal and municipal governments lack capacity to bring discipline to the budget execution system. As a result, government policies and programs are not carried out efficiently or effectively and underdeveloped cash and debt management constrain municipal ability to carry out large-scale infrastructure projects and ensure their successful operation. 04. For example, municipalities interviewed in RS thought entity government still owned the utilities. In FBH, a privatization scheme is being piloted in Tuzla, where the municipality holds 51 percent of all communal enterprises, but the municipal council regulates and sets tariffs at social prices. A board of directors is established for each enterprise. In Tuzla, the deputy department head of utilities sits on the water company s board In 2004, a country-wide Medium Term Development Strategy (MTDS), also known as Poverty Reduction Strategy Paper (PRSP), was prepared jointly by the three governments. However, even at the national level, the link between policy and budget remains very weak Law on Principles of Local Self-Governance introduces checks and balances to prevent the council from blocking budget passage. The mayor can declare the budget effective if the council fails to adopt it within 90 days The BFPs contain economic and social development forecasts, main macroeconomic indicators, and revenue and expenditure forecasts covering the period for which the BFP is prepared. 71

87 29. Revenue administration is ineffective. Municipalities in both entities lack significant policy and administrative autonomy over taxation. 108 Under local legislation, municipalities regulate only municipal administrative fees, court fees, rent, and tariffs/user charges. Only half of municipalities produced accurate and punctual account notices for tax payers and service users, and few impose late fees. In 2005, three-quarters of municipalities had revenue arrears but only one-quarter had plans to collect them. Collection costs for each type of revenue have rarely been analyzed. Debt and guarantee management in municipalities remain controversial. This is particularly true in the FBH; although the adoption of the FBH Law on Indebtedness, Debt and Guarantees has reduced ambiguities in the legal framework. 109 In RS, municipalities are allowed to borrow up to 10 percent of total revenues to cover debt service payments, without approval from the Ministry of Finance. The Ministry of Finance has a debt unit that approves all loans exceeding this 10 percent threshold; and local councils must sign off on borrowing. In March 2007 the RS National Assembly passed a new Law on Indebtedness, Debt and Guarantees of Republika Srpska, which permitted long-term debt for the municipalities to finance capital investments, refinance long-term debt, and finance past liabilities, upon approval from the Finance Ministry. Total debt-proposed and outstanding-due for payment in any following year should not exceed 18 percent of the previous year s regular revenue. Municipalities can issue guarantees to public municipal companies and others according to the decision of the Municipal Assembly. Total exposure of municipal guarantees must not exceed 30 percent of total regular revenues of the previous fiscal year. However, guarantees are not subject to municipal debt limits. 30. From 2002 to 2007, the FBH Law on Annual Budget Law banned municipalities and cantons from borrowing, except during 2002, but this provision is absent from the 2007 Annual Budget Law. 110 The 2007 Annual Budget Law stipulated that cantons and municipalities could borrow domestically if municipal/cantonal assemblies approve. Debt repayment was the sole obligation of these tiers of government. Cantons and municipalities could borrow to bridge short-term cash deficits, 111 finance capital investments; refinance FBH public debt; and to issue and pay guarantees. Cantons and municipalities could issue guarantees for capital investments only to companies in which they are majority shareholders; cantons only for capital investments if they had Cantonal Finance Ministry approval. The FBH Government and cantons must set upper limits for overall fiscal year borrowing. Cantons and municipalities must report regularly to the FBH Government on their indebtedness and guarantees. The FBH Government reserved the right to ban cantonal and municipal borrowing through provisions in annual Budget Law. In December 2007, FBH Law on Indebtedness, Debt and Guarantees was passed. It stipulates that cities and municipalities can generate long term debt only under the condition that debt service for each consecutive year, including servicing of the new loan and all loans for which they issued guarantees, does not exceed 10 % of revenues collected in previous fiscal year. Unlike in RS, guarantees count toward the debt limit in full in FBH. All foreign debt to be incurred by cities and municipalities must obtain prior approval by the FBH Parliament as well as the BH Parliament. 108 In the BH, the State legislates only customs, excise, and VAT. Entities (and cantons in the FBH) legislate most direct and indirect taxes, define tax base, rate, and exemptions. In RS, the entity government regualtes all taxes. 109 Debt stock limits are impossible due to lack of information on total debt assumed at Sub-national level (entities, cantons, and municipalities) Stipulated under Article 43 of the FBH Law on Annual Budget for Borrowing cannot exceed 5.0 percent of annual recurring revenues. 72

88 Figure 54: Sample Instances of Municipal Borrowing Municipality Purpose Source Loan Amount (in KM) Breza Water supply network Investment Bank of FBH Stolac Arrears commercial bank Zvornik Water supply network commercial bank Samac Ljubinje Berkovici Deficit financing and capital investment Municipal building reconstruction Deficit financing and infrastructural projects. Infrastructure commercial Bank (Razvojna banka, Nova banka) World Bank and commercial bank (Hypo-Alpe-Adria) commercial bank (Nova Banjalucka Banka) commercial bank (Nova Banjalucka Banka) 2x x in total 1x x in total 200,000 Visegrad Water supply network commercial bank Istocni Stari Grad Municipal building reconstruction commercial bank (Raiffeisen bank) 50, Buzim Water supply system Investment bank of FBH Odzak Local roads and water supply network commercial bank (Unicredit Zagrebacka banka) 1x x in total Tuzla Water supply network, road infrastructure World Bank Siroki Brijeg Water supply network World Bank Mostar Water Supply, Bridge, Roads Old Bridge Reconstruction Water Supply Kuwait development foundation, on-lending through Federal Investment Bank. World Bank (grace period until 2009) World Bank 16,000,000 US$ 4,000,000 US$ 12,000, Quantitative procurement indicators show good practices across municipalities. Municipalities are bound by the legal framework for procurement stipulated in a single state-level Law on Public Procurement, as stipulated by the European Commission.112 Tendering is required for procuring banking services on municipal accounts management and loans, and these must be re-tendered every three years. Actual tendering practices vary-some commercial banks reported widespread irregularities during tendering, but almost all municipalities reported that at least 75 percent of all contracts awarded were carried out through open competition. Contracts were signed only after proposed expenditures were included in the most recent approved budget. Contractors performance is monitored and reported to the procurement committee for future reference. Guidance was provided to officials to minimize conflicts of interest. However, smaller municipalities interviewed tend to negotiate directly with suppliers, which is justified by the small amount involved. 32. Internal audit quality remains weak across municipalities.113 Both entities have adopted Internal Audit Laws that introduce the obligation to undertake Internal Audits at all levels of government However, implementation is lagging behind the tight deadlines. Only one-quarter of municipalities have adopted local legislation that defines internal audit roles and responsibilities. Internal audit mechanisms, where they exist, tend to function well Accounting, Recording, and Reporting 33. Municipal-level accounting, recording, and reporting remain poor. This undermines incentives for effective budget implementation and weakens transparency and accountability. Most municipalities report having basic accounting competencies in place, but about one-quarter lack sufficient staff. The FBH municipalities and several others in RS say this is due to weak finance department guidelines. Most municipalities have a procedure manual but it is unclear whether it is municipal- or entity-specific. About seven of the twenty municipalities report that cash management legislation is inconsistent with national guidelines The Law defines competitive and compulsory international bidding thresholds for goods and services An Internal Audit Law is being drafted Law on Internal Audit in Public Sector in FBH (Official Gazette 48/08); and Law on Internal Audit in Public Sector in RS (Official Gazette 17/08) 73

89 5.2.6 External Scrutiny and Audit 34. External scrutiny and audit at sub-national level is scant and constrained by capacity. This has weakened administrative accountability to fulfill mandates and fiscal accountability for use of public resources. The FBH external audit office is auditing only federal and cantonal institutions. Despite their legal obligation to audit all FBH governments, external auditors are constrained by weak capacity to extend this function down to municipal level. External audit of municipalities by the entity-level Supreme Audit Institution remains limited. The legislative council plays an important role in external oversight Effective expenditure-tracking mechanisms are essential. Consistent budget classification, adequate accounting, recording, reporting, and auditing are vital to move toward performancebased budgeting at the Sub-national level. Extending the treasury system to local-level governments to track and monitor financial flows and public spending, and good information systems to monitor physical targets and spending outcomes are crucial preconditions for a performance-oriented PFM system. Further efforts are needed for both vertical and horizontal administrative and financial accountability-higher-level governments and users/citizens, using good performance-outcome information Donor Practices 36. Most donor financial management practices have used parallel processes with limited reporting in line with budget formulation or execution cycles. Most municipalities receive donor financial support, but no municipality reported this as direct budget support because most donors support projects and project funds are managed through designated bank accounts. Although most municipalities document externally financed projects in their budget, 116 donor support is not integrated in the annual budget as revenue, with estimated loan receipts, or as expenditure, with estimated project expenditures. The annual budget reflects only counterpart funds made available for project implementation. These practices make it difficult to track and monitor public investments in various sectors. 37. Donors should clarify funding allocation guidelines and timing. Only about one-third of municipalities receiving externally financed support said that donor support and guidelines were clear. Successful projects implemented by local governments (including cantons in the FBH require donors to develop a common understanding of local-level PFM capacity and a careful assessment of the potential to use country systems. 5.3 Conclusions 38. Public financial management capacity varies substantially across municipalities. The PFM results vary according to the individual municipal enabling environment and characteristics. Achieving a paradigm shift from stability to stability plus performance in local public service delivery will require the following actions for PFM. 39. Inter-governmental enabling environment. The fundamental legal framework for good financial management has been largely in place in BH. However, legal acts and regulations must be fully implemented to give practical direction for policymaking and service delivery at the local level. Some legislation must be adapted to the municipal level (e.g., Budget Laws on multi-year Budget Framework Papers), and some legislation needs guidance and technical assistance to implement (e.g., Budget Laws and Chart of Accounts on budget classification system; Law on Indebtedness, Debt and Guarantees of Republika Srpska and FBH, et al) Local council must approve all borrowing and long-term investment plans Notable exceptions are Banja Luka and Zvornick in RS. 74

90 40. Crucial fine-tuning of municipal-cantonal and municipal-entity relations must be completed to reap the benefits of decentralization. The transfers from the single account have improved budget predictability and fine-tuning will clarify and enforce rules of the game for transfers and shared revenues by shifting the role of cantons and entities from excessive control to oversight. Expenditure responsibilities should rest with higher-level government, and municipalities and cantons should focus on oversight processes to ensure that municipalities meet performance standards and spend their resources in line with legislation and their mandates. This shift will require comprehensive and uniform budget classification and an integrated monitoring and evaluation system for municipal government performance. 41. Strengthen municipal FM capacity. Efficient and effective municipal service delivery, and improved self-governance and accountability will require municipalities to strengthen their financial management capacity, particularly on medium-term (multi-annual) budgetary planning to ensure strategic resource allocation and prioritized sectoral spending, general oversight of utilities and public enterprises, budget execution, and accounting records and practices. In general, local councils must strengthen and institutionalize their oversight function. 42. Benchmarking municipal PFM provides crucial information. Benchmarking PFM provides information for policymakers at all levels, development partners, and taxpayers and can highlight weakness, allow governments to develop remedial plans, and set the agenda for reform dialogue between development partners and governments. Benchmarking could be conducted by an independent body, begin with a few volunteer municipalities, and scaled up. The scope of information can be selected jointly by the independent body and participating municipalities. The PEFA assessment adapted for this report could serve as a comprehensive example. 75

91

92 Chapter 6: Conclusions and Going Forward 1. To improve municipal services delivery performance, Bosnia and Herzegovina need to tackle three main areas. First, it must rationalize and then strengthen the enabling environment by ensuring full implementation of laws and policies. Second, it must increase fiscal space by targeting dimensions with the greatest potential to enhance municipal resources for service delivery. And finally, it must strengthen horizontal and vertical accountability by implementing a performance management system for municipal public service delivery, including offering incentives for better performers. 2. This chapter maps out key policy instruments that should be considered, institutions and the actions they might take to implement policy recommendations, and suggests potential areas for the international community to provide support. 6.1 Strengthen the Enabling Environment 3. Of necessity, the Dayton Peace Agreement created post-conflict institutions that secured peace through compromise and power-sharing but also resulted in a complicated multi-level government structure. The international community has been heavily engaged in the state-building agenda and BH leaders have taken the first steps toward making necessary amendments to the Dayton constitution The basic enabling environment is in place. Local Self-Governance law in both entities, the introduction of a stable source of income for municipalities, and efforts to clarify the rules for municipal borrowing have now laid the groundwork for BH municipalities to provide decentralized public services. Several primary challenges remain. First, the Local Self-Governance laws must be fully implemented. Second, the phasing-in of the transfer formula should be completed. Third, sectoral ministries (including the cantonal ministries ) uneven progress on decentralization needs to be corrected. The Ministries of Finance should work with line ministries (and the Ministry of Local Self-Governance in the RS) to complete sector-specific harmonization of sector legislation with the new Local Self-Governance laws. 5. Sector ministries should lead on deepening sectoral reforms. Sector policies and strategies need to be developed and sector ministries need to work with municipalities to set up frameworks at the sub-national level that include sectoral standards and institutional arrangements to monitor and evaluate, and enforce sector performance standards for municipal service delivery. 6. As post-war conditions change, expectations on local service delivery rise. Much can be done at the local level without waiting for constitutional change. Therefore, the international community needs to support a coherent agenda that supports the implementation of the enabling legislation, and identification of potential champions to implement decentralization and municipal sector operation. 6.2 Expanding fiscal space 7. The sustainable expansion of fiscal revenue space for municipalities will require several actions. First, the Ministry of Finance should enhance monitoring of the Revenue Allocation formula phase-in to evaluate municipal fiscal outcomes. The Single Account transfers are an impressive start to provide municipalities with predicable and more equitably distributed resources they need to help finance basic municipal service delivery. 8. Second, the Municipal Council of more affluent municipalities must increase ownsource revenues by gradually raising tariffs to cost-recovery levels. However, this must be carried out in conjunction with the Municipal Administration to provide oversight of service quality and communication links between users and service providers to remedy any deficiencies, thereby improving users willingness to pay. In the longer term, affluent urban municipalities need to develop more buoyant tax bases, including property taxes Under Article III.5 of the Dayton constitution, the international community and BH authorities have gradually taken steps to cede some authority to central state institutions, such as a state border service; a unified intelligence service; security, defense, and justice ministries; a state-level Indirect Taxation Authority; a state court and prosecutor; criminal and criminal procedure codes; and a civil service commission. 77

93 9. Third, the Ministry of Finance at all levels should support responsible municipal borrowing and this will require finance ministries to strengthen monitoring municipal liabilities (including contingent liabilities) and give municipalities clear messages and guidance on borrowing. The international community can provide municipalities with technical assistance to become creditworthy and to prepare quality projects that will attract financing. 10. Fourth, to increase expenditure efficiency, governments need detailed data to evaluate current municipal sector spending against outcome indicators. In the short-to medium-term, the Ministries of Finance in both entities must clarify functional classification and provide sufficient technical assistance to align resource allocation with policy. In the longer term, budget classification and accounting systems should be structured and aligned with a performance orientation; therefore, service outcomes must be monitored and measured. 6.3 Increasing accountability 11. Municipalities must be held accountable for the performance of services that they provide to citizens, particularly now that more resources are being channeled to the municipal sector. But until both vertical and horizontal mechanisms are strengthened, there are few incentives for municipalities to improve service performance. Both supply side (horizontal accountability) and demand side (vertical accountability) should be deployed. Vertical accountability to citizens is beginning to be strengthened by the direct election of mayors. Horizontal accountability among state actors such as Municipal Councils, Municipal Administrations, higher-level government, service providers, and the municipality peer group, will require setting up a framework for service delivery standards and for monitoring and evaluation, to ensure compliance across constituent jurisdictions Supply Side: Enhancing Horizontal Accountability 12. Municipal councils (MCs) do not demand enough from municipal administrations (MAs), which undermines service delivery performance. Now that the municipal executive is directly elected, the local council should strengthen its legislative and oversight capacity to challenge the administration, demand information, debate strategies with the executive branch, monitor administration decision making regarding service delivery, and take a long-term view on tariff setting. 13. Higher-level government-ministry of Finance and sector ministries-need to oversee municipal performance in service delivery and fiscal outcomes, and provide strategic and technical guidance to strengthen compliance. In terms of service delivery, line ministries need to develop overall sector strategies to guide the technical and quality aspects of performance. Most importantly, line ministries should work closely with the Ministry of Finance to establish minimum service standards, and estimate the resources required to deliver all obligatory local functions and a prioritized subset of mandated functions for immediate action. Meanwhile, sectoral ministries must build capacity, provide guidance, and coordinate monitoring and oversight bodies to enforce minimum service standards among municipalities. 14. In terms of fiscal outcomes, Ministries of Finance in both entities should strengthen their financial oversight functions, train and support local officials, and monitor and analyze developments in sub-national finance. Ministries of Finance need to develop PFM guidelines and provide technical assistance and training on financial accounting and reporting functions. 15. Municipal bodies must strengthen their public financial management capacity if municipalities are to fulfill minimum standards for service delivery. Municipal councils and administrations need to improve PFM capacity to ensure that public funds yield value and that budgets reflect policy objectives. Local parliamentarians need substantial capacity building in public financial management to improve their financial oversight function. The international community has an important role to play in PFM capacity building among local administration and parliamentarians, as proven in OSCE and GAP s on-going efforts on this aspect. 16. Fostering performance orientation will require incentives for better performers. To encourage municipal service providers to exceed the minimum standard, Ministries of Finance and 78

94 sector ministries should work with municipalities to design performance-based grants that will complement the revenue allocation system, encourage better service delivery, and ease the difficulty of reform measures. 17. Benchmarking municipal services and PFM over spatial and temporal dimensions can showcase better service delivery performers and reform front runners, and shed light on weaknesses so that governments can develop remedial plans, and create a platform for reform dialogue among development partners and governments. To pursue reforms effectively, and strengthen municipal governance, institutions and individuals need data that show how well municipalities are performing over time and in relation to other municipalities. Therefore, Associations of Cities and Municipalities or independent bodies, need to set up a benchmarking system and widely disseminate performance outcomes on key dimensions of municipal performance, including mandated services and PFM Demand Side: Enhancing Vertical Accountability 18. Jurisdiction-based direct election of mayors promises to provide a broad-based policy platform and enhance performance accountability to citizens. Political parties help promote broader reforms in the BH decentralization framework but they could play a more significant role in securing the benefits of efficient local government and improved municipal service provision. Both NGOs and civil society are well positioned to advocate for raising municipal performance as a priority issue in political party platforms. Ethnicity-based politics and party identification remain pervasive, and there is an ongoing role for international development partners to remain engaged in strengthening the dialogue with entity-and state-level political parties. In the longer term, political parties need a greater appreciation of municipal performance in their overall platform. 19. MZs should clarify and fulfill their role of providing bottom-up community voice. In the absence of functioning government, international and domestic actors and organizations have promoted social accountability (demand-side). Although participation mechanisms have been in place but now that the government structure is gradually taking shape, it is essential to improve the effectiveness of existing channels to strengthen accountability from within. The MZs, as the nexus between the two approaches (public/supply and social/demand), should be strengthened to provide a stronger bottom-up voice. 20. Municipal performance benchmarking, monitoring, and widespread dissemination of performance outcomes should be key to the BH municipal development agenda. To be effective in pursuing reforms and changes in municipal governance, civil society and NGOs need to know how well municipalities are performing over time and in relation to other municipalities. Information on public finance (i.e., budgeting and expenditure reporting) is essential at municipal level to inform local constituents and to encourage public participation in the political process. Information on performance standards and service outcomes is critical for service user citizens to pressure government to enforce policy implementation, to hold service providers accountable; and to allow officials and citizens to evaluate the extent to which a municipality is well-run, since this is the main objective of many mayors surveyed for this report. Better benchmarking information on key dimensions of municipal performance can strengthen local voice, not only at election time but also in the longer term across election periods. 21. Today in BH, no systematic collection and dissemination of municipal performance data exists at any level. International development partners can help fill the gap by facilitating systematic and independent benchmarking of municipal service delivery and PFM performance for the municipal reform agenda to gather needed momentum. 79

95 Figure 55: Summary of Policy Recommendations, Institutions, Actions, and Instruments. Policy Recommendations Key Institutions and Actions Instruments Strengthen sectoral alignment with decentralization framework Strengthen Enabling Environment Ministry of Finance to coordinate with line ministries (and the Ministry of Local Self- Governance in the RS) to harmonize sector legislation with Local Self-Governance laws Sector ministries to deepen sector reforms by developing sector policies and strategies that set up framework for service provision at the Sub-national level, including sectoral standards and institutional arrangements for monitoring and evaluation Carry out policy dialogue and provide technical assistance (TA) on sector strategies to Ministry of Finance and line ministries Provide TA to sector ministries and municipalities in sector reforms Identify and sustain central government champions and support policy analysis capacity to effectively implement decentralization and operation of the municipal sector Technical assistance to Ministry of Finance and Ministry of Local Government (RS) on policy analysis and monitoring Increase Fiscal Space Promote sustainable expansion of fiscal space for municipalities To ensure implementation of enabling legislation transfers Ministry of Finance of two entities to systematically monitor the effect of the new fiscal arrangements (revenue sharing from the Single Account) on municipal fiscal outcomes and local borrowing on national fiscal stability Ministry of Finance of entity and cantonal levels to further clarification of roles and responsibilities to reduce gap filling by municipalities to finance functions of higher level of government Provide TA to municipalities and cantons to enhance consistent reporting on allocation and implementation own source revenues Municipal councilors (especially in affluent municipalities) to gradually increase tariffs to cost-recovery level, while collaborating with municipal administration on community outreach Support continued efforts (such as GAP) to introduce basic financial management information systems at the municipal level expenditure efficiency Affluent municipalities to explore options for broadening tax base, such as introducing property taxation Municipalities to enhance capital investment planning, project prioritization, and procurement to ensure value for money Provide TA to municipal finance and to introduce property taxation Provide TA to enhance planning and budgeting process that ensure value-for-money borrowing Ministry of Finance at all levels to support responsible municipal borrowing and monitor municipal liabilities (including contingent liabilities) and give municipalities clear messages and guidance on borrowing Strengthen Vertical and Horizontal Accountability Provide TA to municipalities to prepare quality projects and to become creditworthy Enhance supply-side (horizontal) accountability Line ministries to develop minimum standards for service delivery and work with the Ministry of Finance to estimate fiscal resources needed Line ministries to coordinate monitoring and oversight bodies to ensure compliance at the local level Municipal councils to strengthen oversight function on service producers, monitoring administration decision making regarding service delivery Advise Ministry of Finance and line ministries on analyzing fiscal implications Provide capacity building for monitoring and evaluation at the central level 80

96 Policy Recommendations Key Institutions and Actions Instruments Ministry of Finance to strengthen financial oversight function, train and support local officials, and monitor and analyze developments in subnational finance Ministry of Finance to develop guidelines and provide TA on PFM to municipalities Provide TA to Ministries of Finance on financial oversight and PFM guideline development Provide capacity building to local parliamentarians and administrations on PFM to strengthen PFM outcomes Ministry of Finance and Line Ministries to initiate dialogue with municipalities to review options for and design performance-based grants to complement the existing intergovernmental transfer arrangements Associations of Cities and Municipalities in both entities and/or independent bodies to disseminate best practices and leverage peer pressure among municipalities Provide TA and pilot program in selected sectors Promote systematic and independent benchmarking of municipal service delivery and PFM performance Support pilot program and scaling up Enhancing demand side (vertical) accountability MZs to clarify and strengthen their role in providing better bottom-up voice and improve client-power feedback loops NGOs and civil society to educate citizens on the benefits of improved municipal performance to elevate to a central issue in political party platforms Promote systematic and independent benchmarking of municipal service delivery and PFM performance Support pilot program and scaling up 81

97

98 Annex 1.1: Municipal Sample Survey Selection 1. Bosnia and Herzegovina s administration is divided into two entities, the Republika Srpska (RS) and the Federation of Bosnia and Herzegovina (FBH), plus the special district of Brcko. The RS is sub-divided into 62 municipalities. The FBH is divided into ten cantons, which in turn are subdivided into 79 municipal districts. Bosnia has a total of 141 municipalities, and Brcko is treated as special district. The FBH is estimated to have 60.5 percent of the population, RS 38.3 percent, and Brcko 1.1 percent. 2. Due to resource and time constraints, this study selected 20 municipalities throughout the country to represent as broad as possible variation in likely differentials in service delivery and local governance measures-not only across the FBH and RS, but also urban and rural. Municipalities in Bosnia and Herzegovina are quite diverse in population size, area/density, and fiscal capacity. The average municipal population is 27,043, but sizes range from 62 in Istocni Drvar in the RS, to 224,647 in the municipality of Banja Luka. Estimated per capita revenues range from 62 KM per capita in Ravno in the Federation to 1,466 in Novo Sarajevo in the Federation. The micro municipalities of Petrovac (189 inhabitants) report per capita revenue of 4,825 KM, and Istocni Drvar (62 inhabitants) 6,291 KM. The smallest municipalities are typically the result of the fragmentation of larger municipalities across the entities that occurred with the Dayton accords. 3. Municipal selection proceeded in four components. In the first component, fifteen municipalities were selected randomly across three per capita municipal revenue strata and according to population shares. The 2005 income strata served as a proxy for fiscal capacity/expenditure levels. Nine municipalities were chosen in the FBH and six municipalities in the RS. 118 Second, the sampling dropped all municipalities under 2,500 inhabitants (12 out of 142). The third component selected two municipalities within the city of Sarajevo. Sarajevo city comprises four municipalities within FBH and three within RS. One municipality was selected randomly for FBH and RS respectively. Finally, the study also selected municipalities associated with three of the largest cities outside of Sarajevo, namely Banja Luka (RS), Mostar (FBH), and Tuzla (FBH). 4. The approach considered and rejected several alternatives for stratification based on assumptions that these would influence diversity in service delivery and local governance across municipalities. These included size (population size) and population density (as a proxy for urbanization). Indicators pertaining to the degree of urbanization of municipalities were not available. Given the limited overall sample, only one stratum criterion was selected under the entity category. Figure A1: Municipal Survey Sample 118. In RS, the study drew upon all per capita revenues for 2005 as reported to the Ministry of Finance. The income stratum within FBH excluded own-source/non-tax revenues, as these data were not available at the entity level Ministry of Finance. 83