Report No. 36285-BO Bolivia Institutional and Governance Review Towards an Inclusive Decentralization (In Two Volumes) Volume I: Main Report May 30, 2006 Public Sector Group Poverty Reduction and Economic Management Sector Unit Latin America and the Caribbean Region Document of the World Bank Acknowledgements This report was prepared by a team led by EdgardoMosqueira (Task TeamLeader) and Jonas Frank (Co- Task Team Leader). Core team members were: Sara Calvo (macroeconomic issues); Christian Gonzalez and Steven B. Webb (fiscal responsibility); Re'my Prud'homme, consultant (fiscal decentralization); James Robinson and Carlos Toranzo, consultants (governance and political economy); Patricia Alborta, consultant, and David Tuchschneider (institutional capacities of Prefectures); Jeffrey Rinne (human resource management); Raja Bentaouet Kattan, Georgia Pelaez, consultant, and Sergei Soares (education sector); Jose Pablo Gdmez (health sector); Jorge Shepherd, consultant, Maria Marcela Silva, and Elaine Tinsley (road sector); Christian Gonzalez and Renato Villela, consultant (subnational taxation). Contributions to different sections of the report were provided by Fernando Rojas. The team also included Ximena Coronado, Ximena Sermdo, Markus Steinich (all GTZ) and Adam Behrendt (DFID). Editorial support was provided by Christopher Humphrey and Herman Nissenbaum, consultants. General coordination and logistical support were provided by Judith Abele, Michael Geller, and Carmen Romero. Peer reviewers: HaraldFuhr (Potsdam University), and Yasuhiko Matsuda (LCSPS). The team worked under the overall guidance of Marcelo Giugale (Country Director), Connie Luff (Resident Representative), Vicente Fretes-Cibils (Lead Economist), and Ronald E. Myers (Sector Manager). We also acknowledge the useful suggestions by Guillermo Perry (Chief Economist). The Public Sector Group has provided inputs at all stages of the preparation of the report. We are particularly grateful to LinnHammergren, Alberto Leyton, Chris Parel, Kathrin Plangemann, and Mario Sangines. We would like to thank a wide range of Bolivian authorities who have participated and continuously supported the analysis and discussions duringthe elaboration of this report. Our acknowledgements go in particular to the following personnel. Ministry of the Presidency of the Republic: Sebastidn Michel, Fabian Yaksic. Ministry of Finance: Gonzalo Aviles Vargas, Juan Brunn, Jose' Camargo-Torrez, Waldo Gutierrez, Beatriz Muriel, German Molina-Diaz, Oscar Montaiio, Liliana Riveros. Ministry of Popular Participation: Gloria Ardaya, Martin Perez, Marcelo Renjel. Unit of Political and Economic Analysis UDAPE: Nicole Czerniewicz, Alvaro Lazo, Gabriel Loza. Civil Service Superintendency: Reynaldo Irigoyen, Karin Oporto. Ministry of Health: Alfred0 Ariiiez, Alvaro Muiioz, Cristian Pereira, Tatiana Ruiloba, Julio Von Vacano. Ministry of Education: Carmen Hada. National Road Agency (SNC): Jorge Avila-Mirabal, Patricia Ballivian, in addition to Mauricio Navarro. Municipal Association FAM: Jose Antonio Terh. Coordinating Agency for Constitutional Assembly (UCAC): Franz Barrios. Fondo Nacional de Desarrollo Rural (FNDR):Alfonso Garcia. Central Bank: Juan Antonio Morales. We are grateful for the support provided by personnel from the prefecture governments in the departments of La Paz, Santa Cruz, and Chuquisaca, who met with members of the team and whose inputswere critical for the completion of this report. Helpful comments were provided by Alejandro Simone, Ehtisham Ahmad, Simon Cueva, Esteban Vesperoni, Paul0 Medas, and Antonio Furtado (IMF), Antonio Guillermo Zoccali, Jaime Alberto Sujoy (IADB). We are grateful for the support provided by DFIDand GTZ for this report. Table of Contents EXECUTIVE SUMMARY ...................................................................................................................................................... i Introduction ............................................................................................................................................................................. 1 Background and the The First Phase o f Decentralization (1994- The Second Phase o f Decentralization and The IGR Instrument 1.The Context and Constraintsof RegionalDecentralization ............................................................................................ 9 The Driving Forces o f Decentralization ................................................................................................................................ 9 The Quality o f Public Services: The Supply o f Public Services Leaves Much to Be Desired The Fiscal Situation: There i s No Room for Increased Public Expendi Conclusion ........................................................................................... 2. Diagnostic: Decentralization Strategy, Institutional Arrangements, and FiscalDecentralization ............................. 17 Problem One: Inadequate Decentralization Strategy ........................................................................................................... Problem Two: Weak Institutional Arrangements ............................................... 17 Problem Three: Inequitable Fiscal Decentralization ...................................................................... 22 Conclusion ............................................................... 3. Building Blocks of a Decentralization Strategy .............................................................................................................. 31 Developing a Vision o f Regional Governments' Future Roles 3 1 Balancing Symmetric and Asymmetric Decentralization ........................................................................................................... ............................................... 34 Building IncentiveFrameworks.................................................................................................................. Entry Points for and Sequencing o f Implementation ManagingTransitions: SacrificingProductive and Allocative Efficiency for Greater Fiscal Responsibility 4. Policy Options for a Short-Term Decentralization Scenario ......................................................................................... 39 Decentralization Strengthening In Fiscal Decentrali ..................................................... 42 Expected Outcome o 5. Policy Options for a Medium-Term Decentralization Scenario .................................................................................... 47 Decentralization Strategy for the Medium-Tm: Maintain and Improve the Efficiency o f Public Services by Separating Roles and Responsibilities .................... Institutional Arrangements: Clarify Responsibilities and Provide More Autonomy for Resource Manage Fiscal Decentralization: Sector Transfers for Funding o f Sub-national Expenditures.. ............................... Expected Outcome o f the Medium-Term Decentralization Scenario. 6. Policy Options for a Long-Term Decentralization Scenario .......................................................................................... 51 Governments and Municipalities ............................................................ 5 1 ...................................................................... 5 1 y ....................................... 54 ........................................ ....................................................... 60 60 Bibliography .......................................................................................................................................................................... 65 Table 1. The Decentralization Strategy............................................................................................................. TABLES Table 2. GDP per capita (2005) and Growth Ratesby Region (inpercent), 1988-2003 Table 3. UnsatisfiedNeedsper Department (in percent), 2001 ......................................................................... Table 4. Quality of Public Services in Education, Health, and Roads Table 8. Impact of Hydrocarbon Law (HL) on Interregional Transfers and Disparities (inUS$per capita), ...................................................... ...................... 41 Table 13:Transfer Formulaein SelectedCountry Cases .................................. ..................... ...................... ...................... 56 FIGURES Figure 1. Survey Results: Demand for Political Decentralizat ...................... 10 Figure 2. Survey Results in SelectedBolivian Cities: "Will A Figure 3. Domestic and External Debt and Public Balance ..... Figure 4. Budget Decision-MakingAutonomy of Departmental Figure 5. Revenue and Expenditures per Level of Government ( Figure 6. Transfers from Hydrocarbon Law (in USDper capita) 2005-2006 (projected) ......................................................................................................... Figure 7. Evolving Roles of Regional Government inCountry Cases BOXES Box 1. Lessons Learned from the First Phaseof Decentralization(1994-2004). ........................................................ 20 Box 7.Achieving Fiscal Neutrality of Decentralization: Alternative Av Box 8. The Challenge of Fiscal Responsibility EXECUTIVESUMMARY The Challenges of the Second Phase of Decentralization ES.l Bolivia is now beginning the second phase of its long march to decentralization. The first phase o f decentralization started in 1994 and was municipal in nature. Municipalities now play an increasingly important role in public service delivery. The second phase i s focusing on strengthening the regional level of government. The driving force of this new phase is the regional autonomy ("autonomias regionales '7 and the proposed withholdingby departments of a struggle over dividing the natural resource pie, which has taken the cloak of a demand for greater share of "locally produced revenue." The driving forces behind Bolivia's new decentralization effort are not its needs for service delivery improvements, reduction of regional inequalities, or greater fiscal responsibility, despite their importance. ES.2 Two decisions for decentralization have already been taken for this second phase: (i) the popular election o fprefectus; and (ii) new Hydrocarbon Law (HL) will provide roughly the two-thirds of additional tax resources (an estimated US$240million in 2006) to the regions and municipalities without transferring new responsibilities. These new revenues to the regions and municipalities could increase significantly as an immediate effect of the nationalization of hydrocarbons recently made by the Government. ES.3 This second phase of decentralization presents three problems: 0) Inadequate strategy: Starting decentralization with the political election of regional authorities, without previously putting adequate institutional and fiscal arrangements in place; and transferring new resources without any requirement to take on new responsibilities, does not constitute an optimal sequence of implementation. This can lead to weak fiscal responsibility, unhealthy competition among regions and municipalities and consequently overlaps inresponsibilities for service delivery. (ii) Weak institutional arrangements: The institutional framework is inefficient and provides limited accountability. Departmental governments do not have minimum institutional capacities, sufficient autonomy to manage resources, and responsibilities among government levels are disjointed and fragmented. (iii) Inequitablefiscaldecentralization:Theintergovernmental fiscal framework does not distribute resources equitably or encourage proper fiscal responsibility by sub-national governments. Fiscal transfers from the HL allocate, for instance, US$16per capita to L a Paz (which has one of the largest communities of the poor) while Pando (among the least poor) receives US$407. ES.4 The country has only limited options to address these problems due to three constraints. The three constraints are the following: (i) decentralization i s being driven largely by strong inter-regional differences over the control of natural resources; (ii)there are considerable inequalities in public services, which are deficient across all sectors and regions; (iii) isnoroomforincreasedpublicspending.Thisenvironmentsetlimitsontheoptions there for decentralization as some of its elements are mutually inconsistent: reducing public expenditures is not easy to reconcile with improving public services; meeting the political demand for more autonomy might not necessarily contribute to better governance; strengthening municipalities might conflict with creating regions; increased autonomy might make it more difficult to reduce interregional disparities in terms o f service quality and availability of resources. Managingthese trade-offs representsa daunting challenge. ES.5 A failure to resolve the above-mentioned problems before moving forward could seriously jeopardize the country's governance, public fmances, and undermine social inclusion. Ifdecentralization is not well implemented there could be a further decline in service quality, and increasedpublic spending pressures could undermine the recent fiscal improvement. While today there i s macroeconomic stability, increased public expenditures could create a fiscal gap and as a result undermine the economy's ability to remain on a stable footing. Furthermore, inter-regional disputes could worsen, weakening the state's functioning and undermine governance. Finally, without addressing the problems outlined above, departmental governments cannot easily strengthen social inclusion. ES.6 As one of the most unequal countries in Latin America, Bolivia cannot afford inequitable spending on behalf of subnational governments. InBolivia the richest ten percent obtain 42 percent o f national income while the poorest only receive 0.3 percent. Poverty and exclusion are particularly harsh on the indigenous people, who comprise more than half o f the population. In2002,40 percent more indigenous Bolivians lived inpoverty than non-indigenous Bolivians, up from 30 percent in 1997. Infant mortality in rural areas i s almost twice that in urban areas. A decentralization process that shifts more and more resources to departmental and municipal governments in an inequitable manner does not contribute in any way to social inclusion. Public spending will continue to be inequitable. Also, a macroeconomic crisis as an outcome of fiscal indiscipline will hurt the poor the most. The poor have a great interest that decentralization does not hamper macro-economic management. ES.7 This Institutional and Governance Review (IGR) aims to assist Bolivia's decision- makers to address the afore-mentioned problems and constraints. More specifically, the report provides guidance on: (i) crafting a coherent decentralization strategy; (ii) structuring the required institutional arrangements; and (iii)underpinning the intergovernmental fiscal framework. This report provides analytical information; recounts pertinent experiences from other countries that faced similar challenges; and offers a set o f alternative policies-they are contained inVolume I(Main Report) and Volume I1(Detailed Technical Analysis). Crafting a Coherent Decentralization Strategy ES.8 A coherent decentralization strategy can help deepen social inclusion, protect national public finances, alleviate regional inequalities, while at the same time improve public service delivery and governmental responsiveness. It would have, however, to be generally agreed onby the different actors involved. Such a strategy should ideally: establish a vision o f the future roles andresponsibilities o f departmental governments; 0 coordinate transfer of expenditure responsibilities in light o f uneven institutional capacities at national and subnational levels; lay out a sequence for gradual implementation so that fiscal pressuresare kept incheck; buildincentiveframeworks for improvingservicesthat leadto more accountability, and; 0 give priority to fiscal responsibility over sub-national autonomy in use of resources to protect national fiscal andmacroeconomic interests. .. 11 ES.9 Once the decentralization strategy is set and agreed upon its implementation could occur inthree stages. The following arepolicyoptions for the short-, medium-, andlong-term. Policy Options for the Short-Term: EnsuringBasic Institutional Capacities and Correcting Inequalities ES.10 Decentralization should preferably be directed initially at ensuring basic institutional and implementation capacities at national and subnational levels; and at correcting fiscal inequalities inthe transfer system. 0 Institutional: (i)consensus building mechanisms would help facilitate political dialogue, reduce potential conflicts among level o f governments, and improve decision making on decentralization; (ii)up-grading public management tools would help foster trust among different levels o f government by increasing transparency as well as accountability. The following are key management tools for departmental governments that would ensure basic institutional capacities are inplace: o financial management systemto allow access to budget information by authorities at different levels o f government and citizens; o human resource management tools to make the public wage bill transparent as well as contribute to more competitive staff recruitment and hiring; o public procurement; and, o mechanisms for monitoring and evaluatingthe quality o fpublic services. 0 Fiscal: (i) improve resource distribution through establishment o f per capita formulas for transfers o f hydrocarbon revenues to departmental governments; (ii) basic fiscal install responsibility rules, e.g., curbs on new departmental debt. Policy Options for the Medium-Term: Separating Responsibilities among Levels of Government and Improvingthe Efficiency of Public Services ES.11 The strategy for the medium-term is focused on separating responsibilities among the national, departmental, and municipal levels to promote more accountability and efficiency. The departments also need to demonstrate that their services meet suitable standards and that they can improve them, before assuming more responsibilities. 0 Institutional: (i)clarify and separate roles and responsibilities assigned to the different levels o f government for greater efficiency and accountability in each o f the sectors; (ii) combine the decisions on services with adequate responsibilities for human resource administration (recruiting, hiring, firing, monitoring and salaries), as well as planning and budget management. 0 Fiscal: (i)gradually phase in a new funding formula for current services to provide incentives for their improvement as well as to reduce inequalities inresource distribution; (ii) once basic fiscal responsibility rules are working well, it would be possible to relax present fund ear-markings for more decision-making autonomy inresource use. Policy Options for the Long-Term: Deepening Responsibilities of Sub-national Governments ES.12 Once the short and medium-term measures have beenimplemented and greater capacities are created, departmental governments would be better able to adopt additional responsibilities and use more resources. 111 ... Institutional: (i) authorize sequential transfer o f additional responsibilities for secondary roads, health, and education; (ii) strengthen intergovernmental coordination. 0 Fiscal: (i) the distribution o f resources related to revenue-sharing and transfers from refine hydrocarbons revenue; (ii)consider permitting departments to raise taxes, e.g., domestic share o f ICE (Special Consumption Tax); and a gasoline and diesel sales tax. Expected Outcome ES.13 As an outcome of the policy options implemented over the short-, medium-, and long-term, there could be basic equity in resource distribution, and more accountable departmental governments. Decentralization can serve as a useful tool for better social inclusion. Achieving this requires a clear strategy o f sequenced measures that do not upset the country's finances or the capacity o f different levels o f government to deliver public services. This IGRis aimed at helpingthe Government develop and implement that strategy. iv INTRODUCTION I.BackgroundandtheContextofDecentralizationinBolivia 1. The Government o f Bolivia i s in the process o f further decentralizing power and resources. This is taking place against the backdrop o f the 1994 decentralization to the municipalities, the recent first popular election of governors, the proposed referendum on regional autonomy in July 2006, and the projected Constitutional Assembly in August 2006. These developments are also occuring in the new environment created by the election of President Evo Morales, who took office inJanuary 2006 as the country's first indigenous leader. 2. These momentous events are accompanied by a climate of national political crisis. Therefore, there i s a widespread feeling that the country needs a new social contract for future political stability and economic development. A central element o f this task clearly concerns how Bolivia can attain greater equality and inclusion. This is o f national urgency since the country i s among the most unequal countries in Latin America. It has significant disparities in the distribution of its assets (education, land, and housing) and earning differentials by gender, ethnicity, sector and types o f employment. According to the 1999 National Household Survey, the richest decile of its population obtained 42.3 percent o f national income while the poorest only received 0.3 percent. Infant mortality in rural areas i s almost twice that in urban areas. Poverty and exclusion have beenparticularly harsh on the indigenous, who comprise more than half o f the population-and their conditions have deteriorated. In 2002, the indigenous had 40 percent more persons livinginpoverty than the non-indigenous, up from 30 percent in 1997. 3. The issues o f equality and inclusion are intimately related to a set o f controversial issues which need to be resolved in the new social contract. The key topics would be the conflicts over land distribution, indigenous rights, distribution and uses o f hydrocarbon revenues, and the extent o f regional autonomy. Bolivians are deeply divided over these matters. And it i s quite likely that their differences will intensifyas opposing groups maneuver to increasetheir power or steer national policies towards their individual interests. 4. A central element is the struggleover dividingthe natural resourcepie. This hastaken the cloak o f demands for greater regional autonomy and for the departments' withholding o f shares o f "locally produced revenue." These demands are highly questionable from a methodological viewpoint. However, they clearly reflect the political impetus for more decentralization and how dividedBolivia is over the underlyingissues. 5. Decentralization can play a useful role inthis situation. It generally involves moving the control o f public services from the national to regional government levels, and shifting responsibilities for raising taxes and expending their proceeds. Through these changes, decentralization innately influences the distribution of power in a society. Therefore, reaching a new basic consensus on how political authority and responsibilities, revenues and expenditures, are to be distributedwould be an important contributionfor the new social contract. 6. Moreover, decentralization itself can have positive effects on socio-economic development. Bolivia i s a highly diverse country, with quite different ethnic groups, and extremely diverse climates, resource endowments, land uses and urbanization patterns. Given this diversity, it would seem reasonable to allow the departments greater autonomy to make decisions according to their needs and preferences. But policy makers also need to attend to the worrying signs of increasing regional inequality. Much o fthis reflects the growth in Santa Cruz's 1 economy and those other departments which benefit from hydrocarbon production. The resulting disparities have increased internal rivalries, fueling regionalist demands. For example, the lowland departments feel that their economic contributions to the country are not reflected in their political powers, while highland departments feel that they deserve a larger share of national wealth to help alleviate their poverty. 7. At the same time, it is important to recognize that decentralization can contribute to inclusion only indirectly. It i s not a panacea and by far not the only means to achieve inclusion. The increase inpolitical participationmade possible by decentralization creates opportunities for the poor for greater presence indecision-making bodies. Similarly, a more efficient and above all more equalitarian supply of public services will benefit the poor, provided that transfer systems to sub-national governments are also equalitarian, or clearly discriminate in favor o f poorer areas. The poor are the one who stand to benefit most from growth or to suffer most of the lack of growth. They therefore have a great interest at sound macro-economic management, and therefore in the making sure that decentralization characteristics do not hamper macro-economic management. But in any case decentralization cannot be seen as a substitute for an explicit inclusion policy that i s ideally basedon a variety o f other mechanisms. 8. Addressingthe above mentioned challenges is a daunting task that will require significant efforts. Whatever direction the country takes, any move forward will not occur in a vacuum. Particular relevant inthis regard is the "institutional landscape" that resulted from the municipal decentralization after 1994. Taking into account the outcomes and lessons learned from this first phase o f decentralization could be critical in designing a strategy for the second phase that the country i s now initiating. 11.The FirstPhase of Decentralization (1994-2004) 9. The first phase o f decentralization, triggered by the 1994 Law o f Popular Participation (LPP), created municipal governments. Although their functioning is far from perfect, municipalities now play increasingly significant economic and political roles. 10. Subsequently, in 1995 departmental governments (prefecturas) were established in the country's nine regions (departamentos). When they were created, they were not yet fully autonomous sub-national governments but were instead hybrid institutions. They were designed as deconcentrated units of the central government headed by presidentially-appointed governors (prefectos). But there are also Departmental Councils (Consejos Departamentales) with representatives appointed by the municipalities, which prepare the programs and budgets for departmental expenditures. The departments are still very weak, though, and have limited service delivery capacity. percentBefore public spending, and expenditures o f departments amounted to 8.6 percent. 11. o f totalthe approval of the HL, expenditures of municipalities represented about 18.3 1. Both municipalities and departmentsaccounted for some 27 percent of total public expenditures. However, sub-national governments raised less than 6 percent o f total taxes and royalties. They (especially the departments)have beenheavily dependent therefore on transfers. 1. This excludes payroll for health and education which is a nationalgovernment responsibility. 2 12. There are several outcomes as well as lessons from this first phase o f decentralization (Box 1). Considering them couldbekeyto moving forward inthe secondphase. Box 1.Lessons Learnedfrom the First Phaseof Decentralization (1994-2004) 1. "Voice" and inclusion: municipal decentralization and popular participationhave provided "voice" to citizens and have facilitated social inclusion. The number o f indigenous mayors and local elected representatives has increased steadily. Indigenous and campesino communities have acquired a greater say in public investment. Social control, through the civic control committee or other means, has served to improve the capacity o f local communities to direct resources to their priority needs. However, it is also true that social control has had a focus on inputs and process, not necessarily on results; and that the incentive to invest has often diverted attention away from service delivery. A new challenge arises now for governance at the national level: in the absence o f institutionalized mechanisms for conflict resolution-among levels o f government, and among different social groups-diverging interests are being put on the political agenda in an unmediated fashion. This creates dead-lock in the governance system. The lesson learned is that departmental decentralization, by only furthering inclusion without creating institutional mechanisms for conflict resolution, may perpetuate the governance challenge the country currently faces. 2. Equity: resource distribution to municipalities is equitable but resource distribution by Sector Ministries remains unresolved. Decentralization to the municipal level has changed both the distribution and focus o f national investment by being more poverty focused and prioritizing social programmes. This also implies a stronger distribution from municipal capitals to rural communities. However, still unresolved is the distribution o f resources by sector ministries in health and education, among others. Teacher distribution, for instance, is inefficient and inequitable: throughout the country, some 9,000 teachers are not needed where they work but they are needed inother places. Resources need to be better distributed within and across departments. The lesson learned is that regional decentralization needs to address inter- as well as intra-departmental resource distribution, but with a focus on improved services. 3. Fiscal Responsibility: low transparency, high informality, and absence of graduated sanctions have weakened fiscal responsibility.Arrangements for sub-national debt management are cumbersome and there is not sufficient data on municipal debt. Only top-down monitoring and administrative controls have proved insufficient to achieve more transparency. The central government as main provider has provided incentives to lend, but not to collect debt service. The status o f departmental debt is ambiguous. On the other hand, the transfer o f investment expenditures to municipalities allowed for "quick wins" to achieve fiscal neutrality, but this came at the cost of fragmenting decisions over resource allocation. The lesson learned is that departmental decentralization needs to create commitments for fiscal responsibility, while the central government alone cannot set good incentives for better debt management. 4. Assignment of responsibilities: politicaldrivers are the main reason for uncoordinated approaches for transfer of expenditure responsibilities. Following the political impetus o f decentralization, the transfer o f responsibilities was not carried out jointly among departmental and municipal levels. Decisions over investment and current expenditures, as well as service delivery, were disjointed. This has led to ambiguous assignment o f responsibilities and low efficiency. For this reason, it has been extremely difficult to achieve service improvements in the sector areas (health, education, roads) across all levels o f the state and regions o f the country. In the meantime, pilot experiences with regards to create associations o f municipal governments (mancomunidades) to achieve higher economies o f scale have not been sustainable. The lesson learned is that departmental decentralization should be designed such that decisions on investment and current expenditures are taken at the same level o f government; coordinated and sector-approaches are needed for decentralization to lead to better services. Source: World Bank. 3 111. The SecondPhase of Decentralization and its Challenges: Towards an Inclusive Decentralization 13. Bolivia i s now initiating its second phase of the country's long decentralization effort. This new phase has a much stronger emphasis on the regional level. It also differs from the earlier effort by being driven by inter-regional differences over the control and distribution of revenues from natural resources, particularly from hydrocarbon exploitation and land use. This issue is particularly heated inthe Santa Cruz and Tarija regions, which want to keep as much as possible of those resources for themselves. This has created a major national debate over how to share revenues from a non-renewable source. 14. Two important decisions have already beentaken on the secondphase: : (i) The election of department governors (prefectos). This represented a milestone because theprefectos derive their mandate from a new constituency--the people-- instead of the head o f state. Departmental governments will de facto be autonomous irrespective of their constitutional status. A legal framework for the new regional authorities must still await the decisions of the upcoming Constitutional Assembly. But negotiations are already underway among the Central Government, theprefectos, and the departmental councils, about defining the transitional stage of the newprefectos. (ii) Roughly two-thirds of the additional tax resources to be collected from the HL,(estimated to total US$240 millioninyear 20062)will begivento the regions and municipalities without their assuming new responsibilities. For prefectures, this represents on average an increase of more than 50 percent in the level of transfers. 15. Unfortunately, these decisions constrain the future path of decentralization. They limit some of the country's options because decisions that might otherwise have been used to set incentives for institutional strengtheningor better service delivery have already been taken. Also, Bolivia's earlier decentralization experience (and similar efforts elsewhere) showed that the desired benefits do not materialize ifthey are not implemented inabasic framework that requires fiscal transparency and responsibility, overcomes horizontal imbalances in resource allocations, and matchesresources with roles. 16. Given these recent decisions, the challenge o f the second phase is how to achieve inclusive decentralization. This is evident both in resource distribution as well as the quality o f public spendingand service delivery. 17. As noted, Bolivia is among the most unequal countries inLatin America. And the central government's fiscal transfers to departments from hydrocarbon revenues are themselves highly inequitable. They provide US$15.8 per capita to La Paz (which has one o f the largest incidences of poverty) while Pando (among the least poor) receives US$407.2. As a consequence o f HL, the new spendingo fdepartmental governments will also behighlyunequal inper capita terms. 18. Inaddition, the quality of services inall sectors is uneven-regions. The quality o f services in health, education, among others i s re also unequal within each department, showing significant disparities between urban and rural areas. However, improving public services is not 2. These values were calculated before the adoption o f Decree no. 28701 from May 1, 2006. 4 at the center of the current debate. This is unfortunate, because in fact, as shown in this report, inadequate resources are not the main causes o f inadequate service delivery, particularly in the cases of education and health. 19. Therefore, a more inclusive decentralization needs to be based on more equitable resource distribution as a means to achieve better and more uniform service provision. These two elements are key and can contribute to reducing inequality and achieving greater inclusion as part of a new social contract. 20. This second phase of decentralization faces three sets of problems. (9 Inappropriate Decentralization Strategv: The strategy for the implementation of decentralization is sub-optimal due to the entry points that have been selected as well as the sequence of implementation. Decentralization reforms are debated, enacted, and implemented in a disjointed manner among regional and municipal levels. (ii) Weak institutional arrangements: Departmental governments do not have sufficient autonomy to manage resources adequately; and responsibilities among levels of government are disjointed and fragmented. This leads to weak accountability and inefficiency. (iii) Inadequate fiscal decentralization framework: The intergovernmental fiscal framework leads to inequity in resource distribution and weak fiscal responsibility. Underlying to this are the ear-marking of revenues to certain regions; fiscal pressures generated by current royalty distributions; resources transferred without expenditure responsibilities; the lack of own revenue on departmental level and high transfer dependence on the Central Government; inequalities intransfers among regions; as well as non-transparent management of sub-national debt. Box 2. Latin American Experiences with Decentralization: Risksand Opportunities After more than two decades o f decentralization, Latin American countries have accumulated a wealth o f experiences that are useful for Bolivia. Decentralization has been primarily politically motivated. Since 2002, all countries in the region have held municipal elections. The strengthened legitimacy o f subnational governments appears to be the main impact o f decentralization so far inthe region. Beyond the political effects, decentralization has brought with it a series o f risks. One o f particular concern i s related to fiscal transfers. Often arrangements for intergovernmental transfers through revenue-sharing measures preceded transfers o f expenditure responsibilities. This has increased fiscal burdens when transfers were not linked to any changes in responsibilities. Their consequences have been added pressures to spend, except in such cases as Peru, which strictly applied the principle o f fiscal neutrality in its decentralization. Granting political autonomy to sub-national governments without installing a proper fiscal rules framework has also had adverse effects. The lack o f own revenue at local levels has been a stumbling block for decentralization, particularly for regional governments. Many regions or states in Argentina, Mexico, Peru and Venezuela, although nominally autonomous, still largely depend on transfers for their revenues. This has prolonged accountability to the center (instead o f local citizens and taxpayers), perpetuated transfer dependence, and limited borrowing capacity. And when taxing powers were infact granted to sub-national governments-such as inBrazil with its VAT-it resulted in sub-optimal tax collection. States competed in "fiscal wars" over tax rates in "races to the bottom." The Brazilian experience also demonstrated the complexity of sub-national tax administration with cross-jurisdictional sales. The transfer o f expenditures to sub-national governments has also been difficult. In most countries, 5 decentralization has not resolved ambiguities in assignments. These are evident in the inappropriate uses o f "exclusive" and "concurrent" hnctions, e.g., Ecuador, Nicaragua, Peru and Venezuela. Unfortunately, sectoral approaches were rarely followed. As shown in Chile and Spain, sectoral approaches have contributed to greater intergovernmental coordination and better services. On balance today, most countries in the region have not improved overall service delivery through decentralization. Only a few have materialized in sub-national governments that pursued consistent, long-term delivery policies, as inBogotl (Colombia). Fiscal responsibility is o f increasing concern. Argentina, Brazil and more recently Colombia and Peru, have all suffered painhl experiences from the profligacy o f sub-national governments. These have jeopardized the provision o f reliable, adequate public services-for which the central government may have ultimate political responsibility-the safety of financial systems, international creditworthiness, and overall macroeconomic stability. For Bolivia to ignore these experiences could be risky. This is particularly true because o f the decentralization strategy which is being followed, the weak institutional arrangements, the inadequate fiscal framework, and the country's fragile conditions. Source: World Bank. 21. Without tackling these three problems there i s a risk that the second phase of decentralization would not lead to the desired outcomes. Following international experience with decentralization (Box 2), this could imply a further decline in service quality, additional spending pressures that underminethe recent fiscal adjustment; and a weakening of governance that might weaken the social contract being negotiated. If decentralization is not well implemented, the country couldmiss an opportunity to enhance and achieve social inclusion. IV.The IGRInstrument 22. The Institutional and Governance Review (IGR) aims to trace the institutional roots of Bolivia's erratic government performance and to propose practical alternatives. IGRs assess weaknesses in the public sector manifested by poor regulation, service delivery, or high corruption; and analyze their possible causes, including political economy drivers. Hence, their distinctive feature is to present policy alternatives that seek "best fit" instead of "best practice," given the constraints ineach country context. 23. The Underpinnings of this IGR. The 2000 IGR, entitled "From Patronage to a Professional State: Bolivia Institutional and Governance Review," provided a series o f recommendations that underpin this new Review. The 2000 IGR assessed Bolivia's public sector organization and operations particularly in the structural areas o f financial and human resource management, as well the transparency and accountability in their performance. It found a gap between the "ideal" behavioral patterns prescribed in official norms and the patterns in actual practices. This breach affected budgeting, public expenditure, and humanresource management, among others. It also undermined transparency and accountability because, the 2000 IGR concluded, the bureaucracy had been co-opted by the political parties and used as the "currency" for political pacts. 24. As options to overcome this situation, the 2000 IGR suggested: (i) buildinga consensus for de-politization of the public sector; (ii)introducing greater transparency in civil service management; and (iii)improving central oversight o f the public administration for greater operational decentralization and decision-making. Its main proposals are currently being 6 implemented under the First Institutional Reform Project (PRI-I), Others o f these proposals figure inthe current decentralization agenda. 25. Objective of the new Bolivia IGR on decentralization. This IGR aims to support decision-makers in Bolivia inthe implementation o f decentralization particularly with regards to departmental governments. Particularly, it provides suggested guidance on how to (i) craft a coherent decentralization strategy; (ii) how to improve the institutional arrangements; and (iii) strengthen the intergovernmental fiscal framework. The policy options (Table 1) are aimed at how to best implement decentralization within a strategic framework. Instead of providing a closed set of recommendations, the report endeavors, to the extent possible, to present different policy alternatives to allow decision-makers to see trade-offs between different approaches. These trade-offs appear to be higher at the regional level as compared to the municipal level. Likewise, the report attempts to present policy alternatives that can be realistically implemented giventhe incentives prevailinginthe Bolivianpublic sector. 26. The process is as importantas the product.The IGR endeavors to support the process of discussion among stakeholders o f decentralization in Bolivia. To this end, the report was elaborated in close policy discussions with government officials, including a series of workshops on different thematic issues related to decentralization held in La Paz and Washington DC. The report benefited from useful comments and suggestions from representatives of the Vice- Ministry of Decentralization in the Ministry of the Presidency, the Ministry of Finance, the Ministryof Popular Participation (until2005), the MinistryofHealthand Sports, the Ministryof Education, the National Road Agency (SNC), as well as the Decentralization Coordinator for the Constitutional Assembly (UCAC). The dialogue also included field visits to the prefectures of Chuquisaca, La Paz, and Santa Cmz. The team held discussions with civil society, among others, the ComitC Civic0 de Santa Cruz, and the chamber of industry and commerce (CAINCO) inthe same department. A Policy Note on decentralization, summarizing the main recommendations from the IGR was discussed with several government officials in March 2006, in the scope o f policy dialogue on behalf o f the Bank. These activities reflect Bank's conviction o f the need for consensus to facilitate an open dialogue, both politically and technically, and to seek common and useful approaches. 27. Structure. The IGR consists of a Main Report (Volume I)and a Volume I1consisting of Detailed Technical Analysis. Following this Introduction, Chapter 1 o f the Main Report provides an overview o f the context inwhich decentralization inBolivia i s taking place, interms of actors and interests, the quality o f public services, and the fiscal constraints. Chapter 2 provides a detailed analysis o f the weaknesses in terms o f the decentralization strategy, the institutional arrangements, and fiscal decentralization. Chapter 3 provides the elements for a new decentralization strategy. Chapter 4 presents policy options for the short-term aimed at achieving minimum regional equity in resource distribution. Chapter 5 presents policy options for the medium-term aimed at dividing responsibilities among levels o f government and improving efficiency in services. Chapter 6 presents the long-term decentralization strategy aimed at deepening responsibilities of departmental governments, both on the expenditure as well as revenue side. It also establishesthe vision of future departmental governments. 7 I.. r i m a . 4, CHAPTER 1. THECONTEXTAND CONSTRAINTSOF REGIONAL DECENTRALIZATION 28. The second phase o f decentralization that Bolivia now initiates is taking place within a context that is characterized by three overarching facts: (i)a political economy o f interest o f actors in which conflicts over natural resources are the driving motivation behind decentralization; (ii) deficient public services across all sectors and regions; (iii) constrained a fiscal environment, that despite the recent improvements, remains fragile and vulnerable to increased spendingpressures impliedindecentralization, particularly inthe take-off period. Each o f these stylized facts provides guidelines for the design o f a decentralization strategy. Unfortunately, some o f these constraints are contradictory. How to manage these trade-offs will figure prominently inthe national debate on decentralization. I.TheDrivingForcesofDecentralization 29. This section aims to set out some of the main political facts underlying the next decentralization phase. Its objective i s to present analytical findings that should be useful for the designofthe decentralization strategy, which is presentedlater inthis report. 30. The driving forces behind Bolivia's new decentralization effort are not its needs for service delivery improvements, reduction of regional inequalities or greater fiscal responsibility, despite their importance. They are instead competing pressures for control and distribution of revenues originating from the country's natural resources, particularly revenues from hydrocarbon exploitation and land usage. The issue is particularly strong in the resource-rich regions, which want to keep control over and decision-making about these resources for themselves as much as possible through the decentralization of responsibilities. Conversely, the less endowed regions are likely to want these decisions to be centralized, from which they would expect more equally distributedrevenues. 31. Regional decentralization i s happening in the context of deepening regional inequality. The deep inter-personal inequalities are well known in Bolivia3.More important for the issue of decentralization are inter-regional inequalities. Most departments in the lowlands have experienced economic revivals and growth while the highlands are stagnant. A particularly important phenomenon has been the rise in the Santa Cruz economy, and of those departments that benefit from hydrocarbons production. This has been fueling regionalism. The lowland departments feel that their economic power is not matched with their political power. For these regions, the demand for political decentralization and popular elections i s one step into achieving this new equilibriumofeconomic andpoliticalpower. 32. A review of the regions' GDP per capita and their growth rates (in constant terms over 1988-2003) demonstrates large differences (Table 2). Santa Cruz's per capita GDP was twice Potosi's level. Regional growth rates are even more unequal. They are positively correlated (although loosely) to GDP per capita, which means that the interregional per capita GDP tends to increase, not to decrease. Bolivia's regional inequality i s thus substantial. It implies that the regions' interests have been diverging. 3. In 2002, an estimated 65 percent of the population had insufficient incomes for basic food and non-food expenditures while an estimated 4I percent had too low incomes to obtain food baskets o f minimum caloric intake. 9 Popular c ecftons for PtefeCtos thc very notionof national unity (Figure 2). Figure2, Sunw Resultsin Scl n8fion?"( 2005) Approual b l Cochabarnba NOanswer Sorrr.r.4.EncueataOpin1rjny mechanisms in place to discuss and resolve conflicts among the different actors involved, particularly between levels o f government. This i s all the more important as the election of prefectos has a deep impact on the governance system in Bolivia: in absence of a strong Legislative Power, the regional level may constitute the only effective check on the Executive. If no preemptive measures are taken, conflicts on responsibilities, roles, and resources, among different levels o f government are likely to increase as the decentralization process unfolds. Recent efforts to establish coordinating mechanisms among levels o f government have as o f yet not proven effective, thereby creating the risk of dead-lock inthe governance system. 38. The different interests of actors involved, along with the lack of institutional mechanisms to resolve conflicts, leads to the following decentralization strategy, that the country i s implicitly following. Decentralization proceeds separately between different levels of government. As a result of diverging interests, the decentralization process proceeds in a disjointed fashion among municipal and regional levels. This leads to a series of negative consequences, since responsibilities cannot be clearly assigned to each level o f government. Decentralization does not proceed sector by sector (health, education, roads, for instance), which from a technical point of view, would offer better opportunities for assigning responsibilities. The sequence of decentralization-political decentralization before building basic fiscal and institutional frameworks-is sub-optimal. The election o f the new regional government leadershas enhanceddemocratic participationinBolivia. But this will also bring nine new bargainers to the table seeking greater shares of the national resource pool. This will make it more difficult to agree on the size o f national public goods, like macroeconomic or fiscal stability. It will also limit the redistribution o f resources among regions for better service delivery. Likewise, political decentralization may lead to additional spending pressures inthe absence o f rules and incentives for fiscal responsibility. 39. Implications. The political economy o f decentralization has a number o f obvious implications: decentralization must be gauged on its capacity to meet the multi-faced demand that exists and also on its capacity not to antagonize those who are cautious about it. Scenarios o f decentralization at the regional level cannot ignore municipalities. They must include and discuss possible relationships and conflicts between regions and municipalities, and be assessed in part on their ability to strengthen municipalities. At least a minimum consensus is necessary for decentralization to be successful. 11.The Quality of Public Services: The Supply of Public Services Leaves Muchto Be Desired 40. Public services are deficient across all sectors and regions. The many current unmet basic needs in education, health, water supply and housing are unequally distributed among regions. Improvements have been made but much remains to be done, particularly in certain regions. Table 3 shows the substantial horizontal imbalances that exist. Health services in La Paz, for example, are ten times more insufficient than in Santa Cruz and six times more than inTarija. 12 Table 3. UnsatisfiedNeeds Der DeDartment (in aercent). 2001 Department Inadequate Insufficient Inadequate Inadequate Insufficient Insufficient Housing Housing Space Water and Supply of Education Health Material Sewage Energy Services Services Services National average 39.1 70.8 58.0 43.7 52.5 37.9 Chuquisaca 53.7 72.1 62.2 62.5 70.7 40.4 La Paz 41.9 66.0 53.2 39.0 49.1 64.9 Cochabamba 37.3 68.2 55.1 42.2 52.6 28.3 Oruro 39.2 67.2 65.9 41.8 47.2 58.8 Potosi 60.3 67.1 71.5 65.0 72.4 59.6 Tarija 30.4 71.5 45.6 43.1 60.5 14.7 Santa Cruz 23.0 77.0 55.8 33.9 43.6 6.4 Beni 63.2 85.0 82.4 64.2 54.6 31.7 Pando 40.4 80.5 83.6 64.8 61.3 39.3 Source: INE, Unidad de Analisis de Politicas Sociales y Econ6micas. www.ine.aov.bo. 41. These data are consistent with a detailed analysis in three sectors, namely health, education, and roads, which are the focus of the IGR. These three sectors are critical for social and economic development, yet there are troubling deficiencies in the quality of public service, particularly amongthe nine different departments (Table 4). Table 4. Quality of Public Services inEducation, Health, and Roads Education 0 Low and inequitable access to upper primary and secondary education: in contrast with high Sector enrollments for primary education (97 percent), as recent as 2001, net enrollment in upper secondary was only 51percent. There are intra-departmental inequalities: within departments and between districts, per student expenditures are unequal. 0 Teacher distribution among districts is quite poor. Some 9,000 teachers are not needed inthe schools inwhich they work (but they are needed inothers). Health Sector 0 In indicators such as Infant Mortality Rate (IMR) and Maternal Mortality Rate (MMR), Bolivia occupies the last place in the Latin America region. There are large differences in health indicators within the country, particularly among different departments. According to the post Census 2002 survey maternal mortality was 320 in the Highlands (including La Paz, Potosi and Oruro), 147 in the Valleys (Cochabamba, Chuquisaca and Tarija) and 206 in the Plains (Santa Cruz, Beni and Pando). 0 In 2001, the percentage of professionally assisted deliveries was 23 percent in the poorer quintile, 53 percent in the following quintile and rose to 89 percent in the richest quintile in Bolivia. At the same time, both the Public Expenditure Review (2004) and the Poverty Assessment (2004) show differences inprofessional birthing coverage from 60 percent for the non-indigenous population to 37 percent for the indigenous population (year 2001). Roads Sector 0 The quality of roads is poor. Pavedroads account for only 7 percent of the total road network. Most o f these roads belong to the primary network. In international perspective these shares are very low since the average level is 27 percent inthe Latin America region. 0 There are regional inequalities interms o f the quality of road services. While Tarija has 16.2 percent o f its roads paved, this share is roughly 2 percent in departments like Potosi or Pando. Even in Santa Cruz, dependent on agro-export industry, has only a tenth o f its road network paved. Source: World Bank. 42. Implication. The implication of these facts is that decentralization should absolutely contribute to increase efficiency in public service delivery. Improvements have been made in 13 some sectors, but muchremains to be done, particularly incertain regions or zones. This requires a decentralization strategy that i s implemented with a view on service improvements, with regional and municipal politicians being accountable for the results they produce in terms of service improvement. 111.The FiscalSituation: There is No Roomfor IncreasedPublic Expenditures 43. The new decentralization effort has raised expectations about higher public spending. While today there i s macroeconomic stability, increased public expenditures could create a fiscal gap and as a result undermine the economy's ability to remain on a stable footing to generate employment and to provide quality public services. Indeed, in other-country experiences with decentralization, the launching o f a decentralization process involving increased public expenditure weakened the overall fiscal sustainability position, disproportionately affecting the poor as the needto adjust aroseS6 44. Bolivia's overall projected7 budget deficit dropped to 2.3 percent o f GDP in 2005 from 5.5 percent in 2004 (Figure 3). This achievement together faster GDP growth-at close to 4 percent in2005-have improved the public sector's debt sustainability situation, helping to keep the nominal exchange rate stable, which i s key to sustain solvency in a highly dollar indebted public and private sector. But Bolivia's public debt remains high at around 80 percent of GDP and the improvement inthe public finances was largely the result o f increased fiscal revenue due to higher international price o f hydrocarbons and the new Hydrocarbon Law. While this Law promises more revenues (about US$366.5 million in 2006), much o f it has already been earmarked. Furthermore, these resources are vulnerable to international price changes. Figure3. Domestic and External Debt and Public Balance 90% 0% 80% -1% 70% -2% 60% -3% 0 50% a -4% a n c3 -5% s 40% 0 -6% $ 30% -7% 20% -8% 10% -9% 0% -10% 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 External Debt(LHS) 0DomesticDebt(LHS) -o-Public Balance (RHS) Source: World Bank. 6. Annex Table 1shows the fiscal balances for the national, departmental, and municipal levels separately. 7. Final calculations are expected by end-May 2006. 14 45. Implication. The implication of this fact is that decentralization should by all means be builton fiscal responsibilityand shouldprotect the advancesinmacroeconomic managementthat already exist. Moving forward, the principle o f fiscal neutrality o f decentralization-that is, reforms should not result inadditional expenditures-should be applied. IV. Conclusion 46. Each o f these stylized facts gives us constraints or guidelines for the design o f a decentralization strategy. Unfortunately, some of these constraints are contradictory. Reducing public expenditures i s not easy to reconcile with improvingpublic services. Meetingthe political demand for more autonomy might not necessarily contribute to better governance. Strengthening municipalities might conflict with creating regions. Increased autonomy might make it more difficult to reduce interregional disparities in terms o f service quality and availability o f resources. How to manage these trade-offs will figure prominently in the national debate on decentralization. 47. These constraints and trade-offs raise the demandfor a specific decentralization strategy: There is need for a strategy that is built on a minimum consensus within an environment o f competing incentives and interests. There is need for institutionalreforms that contribute to improving services. There i s need for fiscal reforms that are compatible with fiscal constraints. At a minimum, they should not worsen the fiscal stance, but improve it through efficiency gains. 48. The following diagnostic aims at providing the elements to achieve this, in terms of strategy, institutional arrangements, and fiscal decentralization. 15 16 CHAPTER2. DECENTRALIZATION STRATEGY, INSTITUTIONAL ARRANGEMENTS, AND FISCAL DECENTRALIZATION 49. This chapter describes and analyzes the three main problems facing decentralization in Bolivia: (i) inadequate decentralization strategy, (ii) institutional arrangements, and (iii) the weak inequitable fiscal decentralization. As will be pointed out, these three problems need to be tackled so that the second phase of decentralization will lead to the desired outcomes: equitative distribution of resources and better services inbenefit of more social inclusion. The analysis in this section lays the base for the policy options that will bepresented insubsequent chapters. I.ProblemOne:InadequateDecentralizationStrategy 50. Bolivia lacks a coherent decentralization strategy for strengthening sub-national governments because ofthe following factors. 51. Political reforms launched the process without a previous definition of an adequate institutional and fiscal framework. While understandable in the country's social context, the primacy of political forces in absence of an adequate rules and incentive framework was unfortunate. Colombia, Peru and Venezuela, among others, were affected by the same dilemma. In so doing, they granted their regional governments more political autonomy before changing their fiscal relationships or expenditure responsibilities. This led to greater spending pressures, weakened fiscal obligations, heightened political cost of own-taxation (and thereby preferences for central government transfers), and overlapping service delivery assignments. Conversely, Chile was one of the few Latin American countries which has introduced institutional arrangements at the sub-national levels in benefit of better services and maintenance of fiscal discipline. These cases clearly demonstrate the benefits of creating and implementing institutional and fiscal frameworks previous to introducing deeper political decentralization. 52. Bolivia's decentralization process has not proceeded in an orderly, integrated fashion among different government levels. Its reforms were enacted separately, as exemplified in its 1994 reforms o f the municipalities, which were de-linked from the departmental governments. As pointed out in Box 1 above, the lessons learned from the first phase indicate that the country's future decentralization will require a more effective delineation of responsibilities among national, departmental and municipal levels. This would help clear the way for achieving the needed greater efficiency in service delivery and accountability. 53. The nature of regional governments creates constraints for the effective implementation of decentralization. This is apparent in political, fiscal, and administrative decentralization. 0 Fiscal decentralization. Regional governments are more transfer dependent due to their lack of own revenue. This requires especially demanding budget controls and fiscal regimes. Administrative (expenditure) decentralization. Regional governments are also more prone to having overlapping responsibilities vis B vis other levels due to their frequent responsibilities for "intermediation" and coordination. Political decentralization. Regional governments are not as close to their constituents as municipalities, and their policy-making is more "corporate" than "popular" in character. It 17 will benefit those groups that can organize themselves better at a regional level, which is unlikelythe case for local communities. 54. Giventhe constraints and the nature ofthe regionallevel, some elements of Bolivia's past decentralization strategy are not readily applicable to the second phase with a focus on departments. 11. ProblemTwo: Weak Institutional Arrangements 11.1 Lack of InstitutionalMechanisms to Discuss and Solve Conflicts among Levels of Government 55. As mentioned in Chapter 11, diverging interests are an integral element of any decentralization process. In Bolivia however, decentralization i s being implemented without adequate mechanisms in place to discuss and resolve conflicts among the different actors involved, particularly between levels of government. New coordination mechanisms are contained in the rules framework that regulates the operation o f recently electedprefectos until the Constitutional Assembly defines their final statusa8However, this rules framework is still incomplete and ambiguous, and it does not serve yet operational purposes. More importantly, recent disputes among levels of government related to resources allocation, the conditions and limits o f roles and responsibilities, among others, demonstrate that the preliminary conflict solution mechanisms are not effective and adequate. If no further action i s taken, there is increasing risk o f dead-lock inthe governance system. 11.2 Poor Management Tools at Central and Regional Levels 56. Financial and budget management at national and sub-national levels are deficient, as detailed inthe 2004 Public Expenditure Review and the 2005 CFAA (World Bank).There are no adequate systems for financial management, human resources (payroll), or procurement. The implementation of the SAFCO Law is weak, and the processes and systems for its adequate operation are not in place. The deficiencies o f these instruments will not allow regional government to provide reliable and transparent financial information to the Central Government. As demonstrated in other countries (Box 3), this is likely to impede the consolidation of decentralization because o f the Central Government's limited confidence about the appropriate uses of funds at regional levels. 57. On the other hand, the processes and systems to monitor and evaluate the delivery of services are practically non-existent. Much basic information in the education, health, and road sectors i s presently not available, e.g., details on payroll per primary and secondary health care, andcoverage indicators that measure service quality. The implication of these facts is that future decentralization will require improvements in the present levels o f capacities for good resource management and service delivery. These institutional weaknesses were already apparent in previous decentralization attempts. In 1999,there was a need to recentralize responsibilityfor the primary roads network from the prefectures to the national government, as a result of declining service quality. 8. Supreme Decree No. 28666 of April 5,2006. 18 Box 3. Transparency Underpinning Decentralization Lack o f transparency and limited exchanges o f information among different levels of government have been major factors impeding decentralization in many countries. Federal Mexico still has no unified budget classification among the federal level and the states. In Ecuador, disputes among provinces over accurate fiscal and budgetary information have obstructed the creation o f consensuses about "provincial autonomy". In Venezuela, poor reporting on transfers has ruled out coherent sub-national policy making and planning. In all countries, the main lesson learned was that good access to and transparent information underpins trust within and among governments. However, a few countries have made great strides by implementing integrated financial management systems at municipal and regional levels. The lion's share o f municipalities in Guatemala and Peru now provide regular, timely budget information through these systems that are integrated with national financial procedures. Argentina's provincial governments are helping their municipalities install such systems. Bolivia should follow other countries and deepen its effort to improve SIGMA (or install a better proven) system in all municipalities and departments for greater transparency. Source: World Bank. 11.3 LimitedRegionalAutonomyfor the Managementof Resources 58, Departmental governments have inadequate autonomy to manage their resources. They therefore cannot improve productive efficiency (the most efficient combination o f inputs to produce outputs)-one o f the key advantages o f decentralization. Large segments o f the funds they receive are ear-marked. This prevents allocative efficiency (the ability to produce services the population most needs). Over the past few years, the new transfers added have prescribed the uses o f their resources, which created a complex web o f rules. And, to worsen matters, these are neither well-monitored nor properly sanctioned. Therefore (Figure 4): Figure4. BudgetDecision-MakingAutonomy of DepartmentalGovernments(inpercent),2004 Group 2: Functional ear- Group 1: Fundswith no ear- marking (current and marking investment 8% expenditures) 20% Group4 Expe Group 3 Sectorai earmarking With no decisi (health, educalion, making roads, other) autonomy 4% 68% Source: World BankiGTZ, based on executed budgets. 0 Departmental governments can currently only spend 8 percent o f all their funds (from transfers and own revenues)' inline with their own priorities (Group 1 inFigure 4). 9. The sums considered here include resources provided for public sector payrolls. However, the decentralization ratios presented inTable 7 below excluded payroll expenditures. 19 0 More than two thirds o f the prefectures' funds (68 percent) are delegated to them without decision-making autonomy, e.g., the payment o f health and education personnel (Group 4 inFigure4). The regions' autonomies will be reduced in the future since the new HL resources are 100percent ear-marked. 59. These restrictions enable departmental governments to exercise only marginal control over the inputsthey must manage for producingtheir targeted outputs (public services). This also limits their allocative efficiency; that is, departmental governments have limitations inproviding the services the populationneedsmost. 60. Comparable constraints also impede effective human resource management, which mayors and regional governors need to exercise in order to obtain their desired outputs and services. However, a local government, which is formally responsible for delivering health or education services but does not control their personnel, cannot be held entirely accountable for what it fails to produce. 61. Inthis connection, too, ineffectual arrangements for civil service management can have significant fiscal consequences iftheir governing rules are unclear. This i s particularly important for Bolivia. The share of payroll in the education sector, for instance, amounts to 50.2 percent of total spending. It i s less significant in health, whose payroll cost was only 10.1 percent of total expenditures. (Departmental governments, though, can hire additional health personnel with up to 10 percent o f the funds they receive for investments). 62. Similarly, it is warrants noting that the ways inwhich humanresources are now managed impedes identifying their costs. One needs to differentiate between personnel costs in, for instance, primary versus secondary education, or primary versus secondary health care. Doing this would help sharpen responsibility assignments and accountabilities: each level of Igovernment needs to bear the cost associated with service delivery. Box 4. HumanResourceManagementinColombia: "Transfer-Push'' I Colombia has a history o f cost-based, supply-driven, and rigid transfers for financing public sector staffs. Intergovernmental grants have increased in proportion with salary and benefit increases or expanded payrolls. Its Ley 60, adopted in 1993, increased transfers to local governments under a revenue-sharing system which provided a fixed portion o f central government income. This was considered necessary because personnel costs for education and health had risen beyond the bounds o f Situado Fiscal transfers. In 1996 both the Situado Fiscal and the Participaciones became insufficient and the center created additional transfers for these needs, initially on a temporary basis. Later they were institutionalized and made permanent, e.g., through the Fondo Educativo Compensatorio. Also, the country's sub-national government have resorted to varied types o f debt when they were not able to pay their personnel. As a response to the "transfer push", the constitutional reform in2001 unified the transfers and set a ceiling on their growths. I t remains to be seenwhether this reform will successfully eliminate the underlyingdeficiencies o f the humanresource funding system. Source; World Bank. 63. Excessively "rigid" expenditures such as salaries can also have serious fiscal consequences. These could arise with the increasing share and volatility of hydrocarbon earnings inBolivia's total revenues.An even"worst-case" situation could develop ifthe country followed Colombia or others in implementing the kind of dysfunctional arrangements under which 20 subnational public sector salaries are established at the central level, irrespective of the payment capacity o f regional or municipal governments (Box 4). 11.4 Responsibilitiesare DisjointedandFragmented 64. In the Bolivian public administration, different levels of government intervene in the production of services in a disjointed fashion (Table 5). In education, for instance, teachers are paid by the central government; they are managed by the departments; their time i s controlled by the municipalities; and the districts are responsible for their hirings and firings. With such fragmented decision-making, accountability is diluted, and no single government or service provider is incontrol. Table 5. Responsibilities per Level of Government in the Health, Education, and Road Sectors Primary and Secondary Primary and Secondary Roads Health Education Central Government Financing Financing Sewicio Nacional de Planning Planning Caminos:National road network Prefectures Deconcentrated SEDES: Deconcentrated SEDUCA: SEPCAM: autonomous payment to healthpersonnel payment to teachers (payroll) institutionresponsible for ( P a Y W Management o f teaching departmental road Management o fhealth personnel. network personnel. Control of norms. Control o f norms. Delivery o f certificates. Infrastructure for secondary and tertiary care (ifnot transferred to municipalities). Municipalities Infrastructure (Hospitals) Infrastructure (Schools) Rural roads Source: World Bank. 65. Additional factors further complicate the situation. Departmental sector directors fill two-often conflicting-roles, They report to departmental governors but also dependon central government sector ministries. Moreover, departmental budgets are not clearly delineated. Department governments participate in the elaboration and partly also in the execution o f a number of ministerial budgets. Each sector has its own logic and pattern of involvement o f departments. A substantial part o f departmental budgets is, in truth, in the national budget. The road sector budget, for instance, i s agreed by each departmental SEPCAM individually with the Ministry of Finance directly, without being coordinated or rationalized with sectoral road policies, or departmental budget policy. The nine SEPCAMs, as autonomous institutions, illustrate the diffused, often confusing roles involved in the execution of departmental budgets. There i s a need therefore for a much clearer distinction between national and departmental functions (such as accounting and budgeting) to obtain more adequate accountability across levels o f government. This should be considered a prerequisite to the enhancement o f the sub- national governments' capacities. 21 111. ProblemThree: Inequitable FiscalDecentralization 111.1 Earmarkingof Revenuesto Certain Regions 66. Transfers from "regalias" and HL provide ear-marked revenues to hydrocarbon "producing" regions, This eliminates the possibility o f their being used for other beneficiaries and purposes, some of which might net higher social and economic returns. 67. Ear-marking has put additional fiscal pressures on the treasury. A provision o f Bolivia's regional pact on hydrocarbon earnings provides that, each time a "producing" department receives a higher share o f these revenues, "non-producing" departments have to be compensated with additional treasury funds. But this has further drained scarce treasury funds from non- hydrocarbon sources. For a related reason, the Departmental compensationFundhas not worked properly, as demonstrated firther below. 111.2 Transfer of Revenue Without Expenditures Assignments 68. As noted inthe earlier, roughly two-thirds of the additional tax resources to be collected from the HL," will be provided to the regions and municipalities. Their distribution, though, will change somewhat over time as follows (Table 6): 0 Central Government will keep 36 percent of the tax revenues o f the Direct Hydrocarbon Tax (Impuesto Direct0 a1 Hidrocavbuvos, IDH) in the first year. This percentage will decline to 32.5 percent in2007. 0 Regional governments will receive 33 percent of IDH's tax revenues, a share which will remainconstant. 0 Municipal governments and decentralized institutions (such as universities) will get 24.8 percent and 6.2 percent respectively o f the IDH revenues in the first year. These shares will later increase to 27.6 percent and 6.9 percent. Table 6. Oil and Gas Sector Royalty Revenuesby Levelof Government, Two HydrocarbonLaws, 2006 Royalties and IDH Former Law New Law New Law (US$ millions) (First Year, 2006) (Fourth year and afterwards) Central Government 132.0 225.7 208.9 Regional Government 203.1 326.5 326.5 Municipal Government 119.5 132.9 Decentralized Institutions 29.9 33.3 Total 335.1 701.6 701.6 Source: World Bank estimates. 10. Techcally, the HL eliminates a 31 percent tax on old fields (representing about a fifth o f hydrocarbon production), called participaciones, and introduces a 32 percent tax, called IDH, which applies to all hydrocarbon production. 22 69. In addition, after the HL's implementation, sub-national governments will carry out 35 percent o f public expenditures, up from 27 percent in 2004. These changes mean that Bolivia will become more decentralized infiscal terms. 70. Significantly at the same time, the HLprovides for the transfer of these resourceswithout additional expenditure assignments. In contrast with most Latin American countries, however, the resources made available originate from future revenue streams. From this perspective, the central government does not experience a revenueloss through decentralization (and could even still improve its fiscal position). Still, the additional revenue should be used by sub-national governments for additional expenditures and paralel reduction o f central expenditures, but this i s not the case. This violates the principle o f decentralization by which resources should be transferred with new expenditures and vice versa. 111.3 Volatile and UncertainResource Flow 71. A significant share of resources for sub-national governments is based on either volatile or uncertain resources. An increasing share of revenue i s from hydrocarbons which i s a volatile source; they are volatile in the future because o f the uncertainties o f future hydrocarbon output, as well as its prices. The Departmental Compensation Fund (analyzed further below) is not working properly. The HIPC funds for municipalities will be phased out in coming years.12 Moreover, resources from the general revenuesharing pool are sometimes not transferred in cash but inquasi-money instruments (Fiscal Credit Certificates; Certificados de Cridito Fiscal, CCF). This volatile anduncertain resource flow impliestwo sets ofproblems. 72. The first is the limited absorptive capacities of sub-national governments to cope with sudden increase in resources. Transfers to both municipalities and departmental governments will increase by 55 percent on average from 2005 to 2006, and much more in the smaller departments (e.g., Pando). These incremental funds will far exceed these governments' effective spending faculties. This limitation has already been quite evident. Departmental governments actually accumulated funds in 2000-2004.'3 They did not invest the added resources.Neither did the municipalities fully invest their HIPC proceeds, some of which simply went into bank accounts. Unfortunately therefore, these experiences indicate that just providing sub-national governments more resources will not necessarily help to improve their services. 73, Second, it i s unclear which additional expenditure responsibilities the sub-national governments might assume inthe context of a volatile resource flow. Ifthe added revenues went to meet rigid administrative expenses, this could complicate budget management, and even perhaps obstruct any adjustments needed for fiscal stabilization. This could happen, for example, 11. While Decree 28421 from October 2005 ear-marks the resources to certain sectors as well as foresees the transfer o f additional responsibilities inthe areas o f rural development, employment creation, and public health (brigadas mhviles, desease prevention, others) to municipalities and prefectures, it is not clear as o f yet whether the central government will stop providing such services, andor in which jurisdictions such activities were carried out in practice. These two factors largely determine if the measures foreseen in Decree 28421 will contribute to fiscal neutrality o f decentralization. Also, the Decree does not clarify responsibilities among departmental governments and municipalities. 12. HIPC funds will be phased out inthe coming years. 13. Aggregate fiscal balances for departments are in2000 (-10 million USD); in2001 (-5 million USD); in2002 (- 8 million USD); in2003 (20 million USD); in2004 (46 million USD).In2004, the surplus o f 45 million USD represents a quarter o f their current revenue (Data: Ministry o f Finance, 2005). 23 if the resources were used to cover long-term obligations, such as the wages for newly hired employees contracted for over many years. Volatile revenues could therefore limit the options for new expenditures insectors with highlevels o f current spending. 74. The net result of all these factors i s the fostering o f additional problems of fiscal management and transparency. A staggering resource flow can also undermine the trust among different government levels, which i s neededfor successful decentralization. 111.4 Vertical Fiscal Imbalances: HighLevels of Transfer Dependence 75. As all countries, Bolivia's levels of government have vertical fiscal imbalances: the own revenue of departmentsand municipalities i s smaller than their expenditures needs (Figure 5). In 2004, municipalities raised 1.8 percent GDP in revenue but their expenditures were 5.5 percent GDP. The departmental governments raised 0.9 percent GDP14but their expenditures were 2.5 percent GDP. Intergovernmental transfers-from revenue-sharing and royalties, among others- exist to fillthis fiscal gap among revenues and expenditures. Figure 5. Revenue and Expenditures per Level of Government (in percent GDP) and Vertical Fiscal Imbalances, 2004 35.01 ETransfersto Social Security System 0MunicipalGovernments 0DepartmentalGovernments Revenues Expenditures Source: World Bank. 76. Municipalities are relatively dependent on transfers. This dependence has been close to 68 percent, and is increasing to over 74 percent in 2006 as a result o f the HL (Table 7 below). There are huge variations among urban and rural municipalities, the latter being particularly dependent on transfers. Meanwhile, though, the municipalities have access to own tax revenue (automobile and property taxes)-which they underutilize. 14. This includes donations. The own revenue o fprefectures (fees) i s merely 0.1 percent of GDP. 24 77. Departmental governments are even more transfer dependent. Their revenues have relied 96.6 percent on transfers, which will rise to 99 percent in2006 with the new hydrocarbon funds. In the future, moreover, they could actually obtain larger transfers for decentralized expenditures. This would produce even greater vertical fiscal imbalances, which could be very problematic. Transfers create lines o f accountability towards the central government. Also, with the increased volatility of revenue streams because o f the higher share o f hydrocarbons, departmental governments will become vulnerable to greater financial risk. As mentioned, this in turncouldaggravatethepossible problemofinflexibleuses ofthe new HLresources. Table 7. DecentralizationRatios Before and After the HydrocarbonsLaw, 2004-2006 Before (percent, 2004) After (percent, 2006) Regions (prefecturus) Expenditure decentralization ratioa 8.6 13.2 Tax decentralizationratiob 0.5 0.4 Transfers dependency ratio 96.6 99.3 Municipalities Expenditure decentralization ratio 18.3 21.5 Tax decentralization ratio 5.4 4.5 Transfers dependency ratio 67.6 74.2 Sub-national governments Expenditure decentralization ratio 26.9 34.7 Tax decentralization ratio 5.8 4.9 Transfers dependency ratio 66.6 73.1 Notes: a =expenditures ofprefecturus/total public sector expenditures; = taxes ofprefecturusl total taxes of public sector; = transferdtotal resources. Analysis indepartmental governments excludes payroll. Source: World Bank. 111.5 Highand Increasing Regional Inequality inTransfers 78. One of the main obstacles to installingan effective decentralization scheme and obtaining greater interregional equality i s the current transfer system, particularly as it affects the departments. The existing inequities in resource distribution are a major impediment to achieve social inclusion. 79. The criteria used to allocate between the various regions the total amount of money to be transferred do not serve any o f the possible objectives o f a transfer system interms o f correcting for horizontal imbalances. Origin o f hydrocarbon revenue plays a key role; that is, "producing" departments obtain more. The "non-producing" departments obtain an equal share per region, irrespective o f the population or wealth o f each region. This implies significant regional inequalityindistribution o fresources. 80. Transfers are highly unequal across departments inper capita terms, as shown inTable 8 and Figure6. The La Paz department receives only US$15.8 intransfers per capita, while Tarija receives US$246 per capita, and Pando US$407.2 per capita. The new hydrocarbons law worsens this situation. The changes introduced by the law have an impact upon horizontal disparities among departmental governments. Transfers to departmental governments, which were already very unequal (on a per capita basis), are made even more unequal by the HL. Transfers were counter-redistributive before the new law, and now they are even more counter-redistributive. The same is true for expenditures per capita. Far from reducing interregional disparities in the 25 country, transfers to departmental governments actually contribute to increase these disparities. Such differences not only fuel regional resentment; they do not contribute to putting Bolivia's departmental governments on an equal footing with regards to the level o f public expenditures. Table 8. Impact of Hydrocarbon Law (HL)on Interregional Transfers and Disparities(inUS$ per capita), 2005-200616 (projected) BeforeHL(2005) After HL(2006, projected) Chuquisaca 20.7 48.5 La Paz 9.5 15.8 Cochabamba 20.1 30.3 Oruro 22.2 60.0 Potosi 15.0 35.8 Tarija 199.2 246.0 Santa Cruz 18.4 25.7 Beni 31.0 71.8 Pando 125.5 407.2 Note; For year 2005, numbers are based on the annual budget. For 2006, numbers are estimates only. Source: World Bank. Figure 6. Transfers from Hydrocarbon Law (in USD per capita) and Poverty Criteria (GDP per capita), 2005-2006 (projected) 450 400 350 'iJ 300 3k 250 Before HL (2005) 200 150 100 50 0 GDP per capita: From poor to rich Note; For year 2005, numbers are based on the annual budget. For 2006, numbers are estimates only. Source: World Bank. 81. The transfer system for municipalities that existed previous to the ap roval o f the hydrocarbons law does not create horizontal inequities inmunicipal expenditures. Transfers on 15. Measured among municipalities grouped in each department. The law increases transfers, but does so in a roughly proportional fashlon. Municipal transfers per capita, which were largely equal, remain equal. The new HLdistributes resources also onaper capita basis. 16. For year 2005, numbers are based on the annual Budget (IEDH+ Compensation Fund+ Royalties). For 2006, numbers are estimates only (IDH+ CompensationFund+ Royalties). 26 a per capita basis are fairly equitable and do not vary much from one region to the other. This situation i s likely to remain so inthe future given that hydrocarbons resources are distributed to municipalities on a per capita basis. 82. The increase in horizontal disparities-in a country where these disparities are already very large-can bringa series of negative consequences.It could favor interregional migrations. Some migrationmay be desirable. But too much too fast will involvevery highsocial and human costs. A valuable social and economic capital in out-migration regions will remain under- utilized. Conversely, in in-migration regions, the pressure on social and economic infrastructure will be much increased. The likely results are under-utilized schools in one place and over- utilized schools in another. Politically, a widening of regional differences will put a heavy burden on the very notion of national unity. Furthermore, the high inequality of transfers does not support the goal o f achieving social inclusion. This i s unfortunate since departmental governments could play a useful role to achieve higherinclusion. -\ 83. Inthe meantime, the mechanisms foreseento reduce inequality among departments have beenineffective.The Departmental Compensation Fundhasnot worked properly. There have not been sufficient funds to compensate "non-producing" departments. In the Compensation Fund there i s an increasing gap between the resources that are neededto compensate, and the resources that are effectively transferred. 111.6 Weak FiscalResponsibility and Poor Managementof Sub-National Debt 84. Ensuring sub-national fiscal stability is another major concern. In 1999, several municipalities could not pay their debt service after the 1994 decentralization process increased their expenditure responsibilities. At the time also, sub-national debt was under-regulated and poorly monitored, and the situation was aggravated by adverse macroeconomic developments. As a result, particularly the most populous municipalities (most of them "capitales de departamento`9 over borrowed. 85. Faced with this growing problem, the central government intervened in insolvent municipalities in 2000. It set up the "Programa de Readecuacidn Financiera (PRF)", a fund partially capitalized by revenue-sharing transfers complementing international financial support. The resources mobilized were used to pay suppliers and helped to reschedule the outstanding debt. This marked, for the first time since decentralization was launched, the Central Government's establishment of a tighter framework for municipal finances. This action brought down the aggregate municipal debt without actual write-offs, but debt reductionhas not followed in all the municipalities and prefectures. Some sub-national governments have failed to respect their annually set debt targets, including Santa Cmz and Cochabamba, two o f the larger municipalities. 86. Today, the total sub-national debt i s not fully or precisely known. Also, the size o f the departmental and municipal debt stock (roughly 5 percent of GDP) is known to be understated because o f unregistered borrowings and suppliers debt. However, data from the system o f debt 17. For this analysis, municipalities were grouped within a department. To c o n f i i the above mentioned hypothesis one would need to also know the fiscal capacity o f municipalities, but which is unknown. Since the (unknown) municipal tax base per capita is probably a function o f GDP per capita, this suggests that tax rates vary greatly from one region to another. For instance, the ratio o f municipal taxes to GDP-a proxy o f the municipal tax effort-is twice as large inthe L a Paz department than inthe Santa Cruz department. 27 capacity indicators now in place suggests that the new HL resources should give both departmental governments and municipalities room to acquire more debt. 87. It bears noting, too, that sub-national borrowing has poor oversight and very limited transparency. Most municipal debt today i s floating debt, particularly fiom suppliers, which is admittedly difficult to monitor. 88. The rules for sub-national debt are narrow and fail to encompass fiscal responsibility rules. It would be important to not focus exclusively on debt service but rather to adopt a wider fiscal responsibility perspective. For example, once the departmental governments are elected, their new leaders might conceivably not wish to recognize the debt stock contracted previously. Although that stock appears to be limited (roughly to 0.5 percent o f GDP), its size could increase rapidly ifthese governments do not follow responsible fiscal policies. An additional set o f fiscal issues i s related to new spending pressures that could arise, particularly during the take-off periods o f further decentralization. Most countries undergoing this process had to cope with these kinds o f pressures that surfaced when their central governments could not readily reduce their sizes in proportion to the sub-national governments' assumptions of new tasks. Theoretically also, there could potentially be new regional taxes to restrain excessive sub- national spending--but their magnitude is likely to remain limitedand therefore alone cannot set a new incentive framework for responsible fiscal management. 89. Bolivia's future thus seems particularly vulnerable to suffer some o f the typical ills that afflict decentralizing states. A failure to follow prudent fiscal policies may create negative externalities on a number o f fronts, some o f which already materialized in the recent municipal debt crisis. Such behavior, if unchecked, could enlarge the country's already existing "reputational risk" and perhaps even ultimately threaten the stability o f its financial system, creditworthiness and overall macroeconomic order. IV. Conclusion 90. The second phase o f decentralization in Bolivia faces significant challenges to make it a success. There are three basic problems that need to be addressed. (i) There is an inappropriate strategy for the implementation of decentralization. This i s apparent, among others, in the entry points for decentralization that have been selected andimplemented as well as inthe sequenceofprovidingrevenue without expenditures. (ii) The institutional framework is weak and inefficient. Departmental governments do not have sufficient autonomy to manage resources; and responsibilities among levels of government are disjointed and fragmented. This undermines accountability towards the citizens. (iii) The intergovernmental fiscal framework leads to inequity in resource distribution and weak fiscal responsibility. Underlyingto this i s the ear-marking o f revenue; fiscal pressures generated by current royalty distributions; resources transferred without expenditure responsibilities; the lack of own revenue on departmental level and high transfer dependence from the Central Government; inequalities in transfers among regions; as well as non-transparent management of sub-national debt. 28 91. There i s the risk that the second phase of decentralization will not lead to the desired outcomes if these three problems are not addressed properly. This would have serious consequences: a further decline in service quality; additional fiscal pressures that undermine the recent process o f fiscal adjustment; and weakening o f governance implied in regional conflict. Taken together, these conditions limit the opportunities to achieve social inclusion as one of the current administration's mainpolicy goal. 92. Taking this into account, the country needs a strategy that establishes a vision of the future role and responsibilities o f departmental governments; balances symmetric with asymmetric elements inlight o f uneven institutionalcapacities at national and sub-national levels as well as demands for new expenditure responsibilities; lays out a sequence for gradual implementation so that fiscal pressures are kept in check and capacities are not overstreched; builds incentive frameworks for improving services that lead to more accountability; and gradually provides more autonomy in resource management in line with fiscal responsibility rules. 93. This new strategy would need to be substantially different compared to the strategy that was underlying municipal decentralization in the first phase (1994-2004). This implies the following: 0 Instead o f merely transferring investment expenditures to sub-national levels (which has kept fiscal neutrality of transfers), future decentralization should createjoint decisions on current and investment expenditures, inbenefit o f higherproductive efficiency as well as delineation of responsibilities. 0 Instead of being mainly transfer-driven, future decentralization should create opportunities for own-revenue generation. 0 Instead of fostering participationprincipally o f local communities, future decentralization should create incentives and opportunities for participation o f the private sector and business. 29 30 CHAPTER 3. THEBUILDINGBLOCKS A DECENTRALIZATION OF STRATEGY 94. This chapter lays out the elements andbuildingblocks of a suitable strategy inlight ofthe analysis o f decentralization carried out above. The strategy takes into account five elements: (i) the vision of regional governments' future roles, (ii)asymmetry and symmetry in decentralization, (iii) points and sequencing of implementation, (iv) incentive frameworks, entry and (v) allocative efficiency and fiscal responsibility. 95. For decentralization to be "strategic," it i s useful to consider how best to build an incentive framework conducive to improving services, strengthening accountability and enhancing fiscal responsibility. It may, therefore, be helpful to review the principles for successful decentralization which are now widely recognized and accepted as good guidelines. The three main ones are: 0 Finance should follow functions: responsibilities should be transferred jointly with resources, and vice versa. 0 Define responsibilities clearly: each level o f government should know what tasks it i s responsible for. 0 There should be fiscal neutrality and fiscal responsibility: national, regional and municipal levels should function within hard budget constraints. Decentralization should not impose additional costs. I.DevelopingaVisionofRegionalGovernments'FutureRoles 96. European and Latin American experiences have been marked by the development of different models for regional decentralization and by varied organizational experiments. As a result, the roles and functions o f regional government in unitary countries have become increasingly diversified (Table 9). Table 9. Principal Roles of Regional Governments Roles Countries Planning Venezuela (regional planning authorities), Colombia (regional planning authorities), France (regional government; with some executive capacity) Political representation o f the central government Spain (Delegado del Gobierno), Ecuador (Gobernaciones), Peru (Prefecto) Agent o f the central government for investment Chile, Ecuador (theoretical role o f the coordination Gobernaciones) Agent o f the central government for direct service Bolivia, Peru (Consejos Transitorios de provision Administracidn Regional until2002) Elected regional government for direct service Colombia, Ecuador (Prefecturas), Peru (since 2003), provision Spain (with regulatory and legislative capacity), France (departments) Elected regional government ina federal system Venezuela (semi-federal) Steering regional competitiveness Chle, Argentina, Peru Source; World Bank. 31 97. These roles can be-and probably should be-envisioned in early stages of the reform. But it i s also true, that nobody can foresee the final model o f regional government as outcome of reform processesrequire several decades. Usually, the role o f regional government evolves over time and frequent adjustments aremade (Figure 7): Figure 7. Evolving Roles of Regional Government in Country Cases Years II II II II b 0 10 20 30 40 I France I I I b 1956. Creation o f 1972: Intergovem- 1986: Election o f Regional Planning mentalrepresentation regions and Authorities departments I Chile I I b 1976: Deconcen- 1992: Municipal 2002: Elected re-gional trated regional representation government? government I SDain I b 1979: Election of 2000: Discussions for new regional federalization o f Spain government I Cdomhia I I b 1991: Election o f 2002: Elimination of departmental Regional Planning government Authorities 1990-2002: Four Models: 1. Delegated - 2. Elected- 3. Delegated -4. Elected Regional Government I Source: World Bank. 98. Bolivia's circumstances warrant its consideration of three functions for its regional governments: Planning: which does not necessarily implymanagement ofresources; 32 0 Investment coordination: where regions have resources and autonomy in deciding where andon what theseresourcesare goingto bespent; 0 Service provisioninhealth, education, roads, or the like. 99. It is useful, therefore, to review the different conditions for success for each of these functions that need to be inplace in order to be effective (Table 10). For the present, though, it would appear that Bolivia's prefectures are likely to mainly be service providers, although roles o fplanningand investment coordination may evolve later on in a complementary fashion. Table 10. Regional Governments' Conditions of Success PotentialRoles Conditionof Success Regional governments as planning 0 Planning is complemented with financial incentives. institutions 0 Planning and budgeting are integrated. 0 Regional Planning contributes to national policies. 0 Regional Planning is linked to national planning. Regional governments responsible 0 Regional governments can offer sufficient incentives for other levels o f for investment coordination government (or service providers) to coordinate. 0 Clear debt service arrangements are inplace. 0 Transparent rules exist on investment decisions. Regionalgovernments as service 0 They have the freedom to decide on needed inputs (budget flexibility). delivery providers 0 They have spending autonomy 0 Clear institutional arrangements exist between national and regional levels, e.g., roles o f sector directorates inregional governments are clear; a professional civil service regime exists; there are well defined service responsibilities. 0 Regional delivery o f services is linkedto national monitoring and evaluation systems. Source: World Bank. 100. Assumptions for optimal implementation. The following are assumptions for achieving the full benefits regarding the future role o f regional governments: 0 There is open dialogue among participating actors and they are willing to join in an at least minimal agreement on the future roles o f regional governments. 0 Actors are willing to accept some o f the opportunity costs o f different organizational models. 0 There i s also accord that the decentralization strategy chosen may be adjusted to meet new circumstances. \ 11. Balancing Symmetric and Asymmetric Decentralization 101. Countries regularly consider the degree o f uniformity that should exist among sub- national responsibilities. From a theoretical point o f view, the higher the level o f uniformity (symmetry), the greater are the levels of coordination and their benefits. However, local capacities and sub-national demands are usually disparate, requiringasymmetric decentralization (under which regions can implement decentralization processes at different depths, scopes and paces). 33 102. The proposals that have been put forward in Bolivia about regional government's future responsibilities, as well as the country's structural conditions, seem to imply that its decentralization is likely to proceed in an asymmetric fashion. Some sectors, like education and health, might define their individual institutional standards for departmental governments that have to be met to be eligible for decentralization (e.g., "certification" in Colombia or Peru). Another factor i s that some local opponents of regional decentralization might successfully interfere with its implementation. The Government might therefore decide to negotiate one-to- one deals with individual regions on their adoptions of new responsibilities (as in Spain). Fiscal shortagesmight impede funds disbursements (as inEcuador). 103. Under these circumstances, the recommended approach is to balance symmetric with asymmetric elements so as to limit additional costs of coordination implied in an asymmetric decentralization process. This would mean that some areas o f responsibilitywould be identical in all regional governments, while individual regions may adopt additional or different responsibilities inother areas. 104. Assumptions for optimal implementation. We suggest consideration of the following factors for determining the blendo f asymmetric and symmetric approaches: There are clarity and technical information available to gauge which approach would have positive or adverse effects. 0 There are incentives available for proceeding with one or the other approach that can be agreed on and implemented. Bargaining schemes can be adopted for lowering transaction costs and providing oversight o f institutional arrangements. 111. Entry Points for and Sequencingof Implementation 105. There are no pure models for a vision of a decentralized state. And any decentralization process is, by definition, gradual. Its constituent measures, therefore, have to be appropriately sequenced. To illustrate this, the scenarios described inthis report set out phased approaches for decentralization inthe health, education androads sectors. 106. They address how these sectors' responsibilities can be assigned at national, regional and municipal levels. This i s very pertinent because Bolivia's intergovernmental system is featured by concurrent (or shared) responsibilities that are blurred. This is partly a result of how decentralization was (somewhat unfortunately) implemented in the past. It was a political decision to first decentralize at the municipal level; and only later complete it at departmental levels. (The process was initially perceived as a means o f strengthening departmental or municipal levels-but not both simultaneously.) So for future decentralization, it would be essential to employ different entry points and implement them differently. For this purpose, we recommend that the process follow a sector-oriented logic. This would facilitate the assignment of clearer responsibilities at different governmental levels. It should later permit the realization of the potential benefits o f more efficient service delivery and greater accountability. This IGR consequently identifies feasible entry points, cites the important conditions that should be present for their usage, and recommends suitable sequencing measures. 34 107. Assumptions for optimal implementation. The following are important factors for selecting an optimal sequenceo f implementation: 0 Actors agree on a vision o f decentralization, and commit to its realization. (This would be enhancedby obtaining some "quick wins" early on so that the decentralization process maintains credibility and popular support). 0 There is an incentive framework for sector decentralization. 0 Decentralization takes place as simultaneously as possible ineach region (and symmetricallyinthe pertinent sectors). 0 Intermediate targets are spelledout. 0 The Government is prepared to withstand pressures for suboptimal forms of decentralization (intheir depths, paces and scopes). IV. BuildingIncentiveFrameworks 108. The political pressures for further decentralization in Bolivia are not necessarily incompatible with the demands for improving public services. A well thought-out strategy can serve both objectives. This can be achieved through establishing "win-win'' situations between national and sub-national governments which address mutual interests. For this purpose, the central government could craft bargaining packages that contain the following components o f a suitable incentive framework: 0 a fundingformula, 0 service quality indicators that can be regularly, easily monitored, and 0 entry points for the transfer of responsibilities. 109. The proposal here is to depart from other Latin American countries' approaches. These often assumed that it was most important to meet a set of ex-ante conditions which tended to measure the past performances of municipal and regional governments. Instead, we recommend that Bolivia's decentralization framework be "forward looking": instead of transferring resources reflecting historic budget allocations, it would be better to transfer funds tied to incentive-based formulae (even if this results in different amounts), It would also be advisable to commit sub- national governments to providing improved services within an accountability framework of checks and balances. Also, instead of adopting "certifications"'8 of sub-national governments- 18. Decree no. 28666 o f April 5, 2006, establishes a "certification" system for Boliviandepartmental governments, without yet defining the technical criteria to be used to assess subnational capacity. As mentioned above, "certifications" o f subnational governments have-in the institutional context o f many Latin American countries-often not been efficient. The reasons are, among others, the difficulty in finding objective and measurable indicators not only to assess general local government capacity, but also sector-specific capacity that is required to deliver decentralized services in health, education, or roads. Such indicators necessarily measure only past performance, and not future potential o f local governments. Inaddition, there is a high level o f credibility and commitment that is needed from central government in refraining from transferring responsibilities to non-eligible subnational governments, in light o f regional pressures and from the "bottom- up" to do so. For these reasons, a more suitable approach is to assess minimuminstitutional capacities based on core systems for more accountability; track performance for given level o f service delivery; improve the quality o f these services; and start with transfer o f those responsibilities that are suitable for a large number o f municipalities or regions. 35 as followed in Peru and Colombia, for instance-it i s suggested that emphasis be placed on providing incentives for improved services. 110. At the same time, it is clear that the types o f sophisticated decentralizations implemented in Canada, Spain and other European countries that are more results-oriented are not workable without very clearly delineated responsibilities. Unless such conditions are in place, intergovernmental agreements are unlikely to be very effective. Therefore, at the present status o f Bolivia's transition towards a more decentralized state, its possibilities o f effectively pursuing results-based arrangements do not appear to be very strong. 111. Assumptions for optimal implementation. The following are important factors for the application o f incentive frameworks: The central level has sufficient leverage to offer bargainingpackages to regional governments that are o f mutual interest (e.g., raising taxes at local levels). Indicators o f successful attainment o f service improvement goals can be monitored easily and without the need for a technocratic apparatus. Indicators andtargets are published. V. ManagingTransitions: Sacrificing Productive andAllocative Efficiency for Greater FiscalResponsibility 112. Decentralization offers the potential benefit o f boosting efficiency. Sub-national governments are, in theory at least, more knowledgeable than central governments about local problems and combining factor inputs at that level (productive efficiency). Also, they are more knowledgeable about services which the population needs most (allocative efficiency). In Bolivia's transitional status, however, it would be well advised to sacrifice both productive and allocative efficiency in the foreseeable future for more effective expenditure coordination and fiscal responsibility. This implies that, among others: There i s ear-marking for sub-national expenditures. 0 There are fiscal responsibility rules which set ceilings on the hiring o f human resources. Sub-national debt capacity is assessedby taking into account fungible revenue in light o f rigid expenditures, such as debt service and expenses for payroll. 113. Fiscal neutrality o f decentralization is another concern. In recent years, Bolivia has had high fiscal deficits, e.g., the aforementioned NFPS deficit of 6.2 percent of GDP in 2004. It is therefore particularly important to consider if near term decentralization could help serve to lower, or at least maintain, the present level o f deficit (fiscal neutrality in the strict sense), or to determine if these deficits might instead increase (least desirable outcome). To be sure, there are other, more effective strategies to lower public deficits than decentralization. But if the latter is not implemented along with a related goal o f fiscal neutrality, it might derail the Government's reformprogram and worsen its macro-economic conditions. 114. On these accounts, sub-national governments need to concentrate on having more transparent revenues and expenditures, and on establishing mechanisms, which could enable them to monitor the fiscal neutrality o f their decentralized decisions. The measuring o f fiscal neutrality i s methodologically challenging, and made difficult for instance by general revenue- 36 sharing formulae as applied in Bolivia which imply fluctuating transfers according to macro- economic cycles. However, the following indicators could be used to assess the extent to which fiscal neutralityi s kept: (i) Sector budgets. Costing exercises demonstrate that the funding for newly devolved responsibilities would be available from current resources. (ii) Overall balance of the national, departmental and municipal governments. It i s important to assess the extent to which a given decentralization measure would affect the overall balance of each level. Hence, vertical imbalances for each and among the levels of government would have to be identified. (iii) Overall balance of the NFPS. Although decentralization actions can affect the balances of each o f the three government levels, they may still be consistent with overall deficit targets. The scenarios to be followed should therefore be assessed with regard to the overall NFPS balance. 115. Assumptions for optimal implementation. The following are important factors to implement a fiscal responsibilityframework: 0 Prompt actions can be taken to control deficit increases resultingfrom the new HL. 0 Redistributiono f funds happensbefore decentralization takes place inthe sectors. 0 Elected governors accept a new framework which limits their decision-making autonomy in the combination of inputs to produce services Oproductive efficiency) as well as the specific services required by the population (alocative efficiency). Fiscal responsibility rules are accepted and applied. 0 Ex-ante rules and ex-post conditions for prudent borrowing are in place and constrain lenders and borrowers. 0 Sanctions for non-compliance with precautionary fiscal rules can be enforced. 37 38 CHAPTER 4. POLICYOPTIONSFORA SHORT-TERM DECENTRALIZATION SCENARIO 116. This chapter lays out a set of policy options for the short-term. They comprise a set of initial steps that could be given priority and that serve to prepare the terrain to deepen measures inthe medium-, and long-term. This chapter first lays out the particular strategy that should be followed. In a second step, it describes the policy options to improve the institutional arrangements; and ina third step, it describes the policy options related to fiscal decentralization. The fourth and final section describes the expected outcome of the short-term decentralization scenario. I.DecentralizationStrategyfortheShort-Term: EnsuringBasicInstitutionalCapacities and Correcting Inequalities 117. An effective decentralization strategy requires that its measures would only be undertaken in the fiamework o f a well supported platform o f sequenced political decisions and reforms. A basic political consensus i s necessary on the following guidingprinciples (Box 5). Box 5. The Pre-conditionfor Successful Decentralization: A Basic Political Consensus I 0 All participating actors should agree that decentralization measures to be taken and implemented should be responsible and neutral infiscal terms (e.g. avoiding additional expenditures for decentralization). 0 All participating actors should agree that fiscal arrangements for decentralization, including actions on transfers, would mainly aim at the reduction o f horizontal imbalances in the delivery o f services. This agreement implies that the current distribution o f transfers inequal shares to departments is not necessarily the solution to the problem o fpoor service quality. ' All participating actors should agree that current service delivery conditions have to be maintained in the short-term, and need to be improved inthe medium- and long-term. Source: World Bank. 118. Once consensus i s reached on these principles, decentralization could be implemented gradually, with different steps to be taken inthe short-, medium-,and long-term. 119. In the first phase (Short-Term) the emphasis should be on ensuring basic institutional capacities and correcting inequalities. Institutional: improve regional management tools. This should encompass financial management, human resources, procurement and monitoring and evaluation, and corresponding systems. The strategy should focus on ensuring adequate basic institutional capacities before giving regional governments or municipalities more revenues or expenditure responsibilities. 0 Fiscal: (i)improve the distribution o f resources with a view on basic equity (per capita formula for transfers); (ii) Basic fiscal responsibilityrules. 39 11. Strengthening Institutions for Decentralization 11.1 Management Tools for Decentralization 120, The SAFCO Law created financial and control systems for public sector management. Some of these systems need to be implemented at regional levels for transparent departmental government management. So too do mechanisms to monitor and evaluate public expenditures and service delivery. In their absence, departmental authorities will not control their own operations; the Central Government and citizens will not be able to assess the use o f public funds; andthere will remainopportunities for inefficiencies andcorruptioninpublic activities. 121. At the departmental level, the implementation of the SAFCO systems is limited. Their fuller operation will require the definition and installation o f specific regulatory regimes, implementation of operating processes, and training o f responsible officials and staff. In the short-term, these activities should be developed for four basic management systems with the emphases indicated below: 0 Planning, to create adequate mechanisms for creating and obtaining agreement on operational plans and budget o f the regions. This should help make public investment and service delivery o f services more effective and efficient through enabling different levels of Government to coordinate their interventions. 0 Financial Management, to upgrade control of and transparency in expenditures. The Central Government financial management system "SIGMA" might be adapted to departmental requirements. 0 Procurement, to make the acquisition of goods and services transparent. The CPAR (a report prepared by the World Bank), provides a set o f recommendations to improve procurement mechanisms inBolivia. 0 Human Resources, to replicate the good practices in effect at the central level for competitive recruitment and hiringo fpublic employees, and control o f their wage bill. 0 Monitoring and Evaluation, to establish a set o f indicators for service delivery monitoring and methodologies for its evaluation. These will allow the Central and Regional Governments to supervise the quality o f the services and generate information to improve the public programs performance and the allocation ofresources. 122. These systems need to be in place to allow the Central Government to better supervise uses of public funds and, principally, the efficiency o f public service deliveries. Specific monitoringand evaluation methods should be developed for the main services now controlled by the Prefecturas and for those that will be decentralized in the near future. Chile provides good examples of these systems. 11.2 Autonomy for Efficient Allocation of Resources 123. The sub-national governments, particularlythe departments,have to have more autonomy for managing their responsibilities. For this purpose, the Central Government should establish a national incentive framework aimed at increasing productive and allocative efficiency at local 40 levels. When it i s installed, local governments could then gradually increase the scopes and depths of their responsibilities. The mechanisms to build an incentive framework are, among others: humanresources management; earmarking o f expenditures; sector formulae for service delivery; monitoring and evaluation systems; sector bargaining strategies for the transfer of responsibilities; and participatory and control mechanisms. The incentive framework i s aimed at raisingthe interest of departmental government for an efficient allocation o f resources, providing transparency in the use o f public funds, and creating accountability among levels of government as well as citizens. 124. One particularly important area i s human resource management. Useful modifications should include: 0 Establishing wage and career policies for sub-national civil servants. These would encompass rules and methods governing their wages and pensions, career management, and decisions on their hiringand firing. 0 Having departmental governments trained and provided technical assistance to fulfill their new roles effectively. This should include enabling the sector ministries to exercise stronger oversight. They also need to reinforce their units related to service delivery for functioning inmore decentralized environments. 11.3 Creatingthe Conditions for Better Services: Preparatory Phase of Institutional Strengthening 125. In addition, there are certain sector-specific requirements that should be met at departmental levels prior to any new transfers of expenditures (Table 11). Table 11. Preparatory Measures for Education, Health, and RoadDecentralization Education Health Roads Train departments on human e Train departments on human 0 Develop a national road sector resource management. resource management. strategy which involves the three Prepare inventory (physical e Prepare inventory (physical levels o f government. assets) for SEDUCAs previous assets) for SEDES previous to e Define the primary, secondary, to incorporation into the incorporation into the and tertiary road network. departmental government departmental government e Full disclosure of accounts of structure. structure. SEPCAMs. Complete information on 0 Complete information on number e Train departments on human number o f teachers per o f health personnel per service resource management. education level level (primaryhecondary health) Establish pool of public (primaryhecondary) in each in each department, and pay employees who would have to be department, and pay levels levels (including benefits). trained and gradually retired. (including benefits). e Complete information on health Prepare inventory (physical Complete information on personnel (doctors, nurses) assets) for SEPCAMs previous to teacher distribution among and among and within departments incorporation into the within departments (urbadrural). departmental government (urbadrural). 0 Monitoring arrangements in structure. Monitoring arrangements in place: select, discuss, and agree Contracting out o f all place: select, discuss, and agree on basic set o f indicators to maintenance activities. Train on basic set o f indicators to measure service quality; define departments on outsourcing measure service quality; define baseline in each department; strategies and tools. - baseline in each department; define responsibilities in Monitoring arrangements in define responsibilities in monitoring; define sanctions for place: select, discuss, and agree 41 Education Health Roads monitoring; define sanctions for non-compliance. on basic set o f indicators to non-compliance. Improve and disclose the measure service quality; define incentives framework for service baseline in each department; delivery, especially the use of define responsibilities in SUM1 monitoring; define sanctions for non-compliance. Source: World Bank. 126. To these ends, changes are needed for the adoption and monitoring o f service quality indicators (Table 12). They are different for the education, health, and roads sectors. Table 12. Service Quality Indicators in Selected Sectors Education Health Roads Enrollment rates Health coverage indicators: (i)Quality ofprimary, secondary, tertiary professionally assisted deliveries, (ii)roads (International Roughness coverage o fpentavalente vaccination, Indicator, IN). and (iii) coverage o f fourth Prenatal Care. Source; World Bank. 111. FiscalDecentralization: A First Package of Measures to Improve Regional Inequality 111.1 Improve RegionalInequity: Redefming the Transfer Formula 127. In the short term, the new HL resources should be targeted on meeting specific expenditure requirements. These should be transferred on the basis o f per capita formulas. This would help produce greater departmental revenue equality and contribute to social inclusion. A per capita formula is not only a sufficient proxy for expenditure needs. It is also acceptednow in Bolivia as a transparent formula (it has been working for municipalities already). Several other countries in Latin America have often adopted much more complicated, and therefore also less transparent, transfer formulae (Table 13). However, ifit facilitates the buildingof consensus on a new formula inthe country, a poverty formula could also be given consideration. Compared to a per capita formula it is less transparent-and it raises the question how much more a poor region gets than the wealthier region-but it is still a valid approach to establish a basic levelof equity andcompensating for expenditure needs. Table 13. Transfer Formulae in Selected Country Cases Mexico: Fondo General 45.17 percent distributed to the status on an equal per capita basis. de Participaciones 45.17 percent is allocated on a historical basis, starting with states' own revenues just before the system started in 1980 and modified gradually by relative tax effort. 9.66 percent is allocated ina way that compensates the previous two allocations. Colombia: general 60 percent inproportion to the number o f habitants with unsatisfied needs and revenue-sharing relative level o f poverty. (participaciones 40 percent according to population, administrative efficiency, and improvement in municipales) quality o f life. Ecuador: 15 percent o f 70 percent to municipalities. Fromthis share: 10 percent inequal shares; 40 percent general national revenue population; 50 percent unsatisfied needs. 42 (15 percent Law) 30 percent to provincial councils. From this share: 50 percent unsatisfied needs; 10 percent territorial extension; 40 percent population. Nicaragua: Municipal A fixed share (2.5 percent) for the Managua municipality. Transfer Law (Leyno. For all other municipalities (97.5 percent): 40) 25 percent on efficiency inadministration o f the property tax. 25 percent compensating fiscal capacity across municipalities. 25 percent for population. 25 percent efficiency inexecution o f transfers. Source: World Bank. 128. A new transfer formula in Bolivia needs to be done promptly since these new resources will otherwise be absorbedinexpenditures that could be difficult to scale down later. This might also affect those departments which would receive fewer resourcesunder a replacement formula. Following the experience o f other countries (Box 6), prompt actions are also needed before new interest groups begin lobbying for the incremental resources. Box 6. ImprovingTransfer Systems: TimingReformsis Crucial From Argentina to Germany, from Peru to Spain, many countries have been, or are inthe process of, reforming their transfer systems. However, most were unable to completely overhaul their systems and many reforms failed. Germany has not been able to adjust its system for over 15 years even though its excessive compensation and redistribution have been long recognized. Neither has Argentina succeeded in revamping the inefficient and opaque revenue-sharing arrangement (the so-called "fiscal labyrinth") adopted in the early 1990s despite the very high transfer reliance o f most o f its provinces. Colombia also had long appreciated that its 1993 formula was contributing to more fiscal pressures. But it took more than 7 years before its transfers were adjusted for better fiscal controls. Only countries like Canada appear to have been able to win such new fiscal pacts as the 1999 arrangement with its provinces agreeing on fundamental changes to the country's equalization scheme and financing means. The lessons o f these experiences are evident: Try to introduce reforms quickly before there are new beneficiaries o f inefficient transfers who could block efficiency-enhancing reforms. Or try to reform transfer systems gradually and only at the margin, as in Colombia and others, for instance, Peru with the new formula for the Canon transfers. Bolivia would be well advised to introduce speedy changes so that it does not get dragged into the piecemeal reforms that have hamperedmost other countries o fthe region. Therefore, the introduction o f a per capita formula for the transfers related to the new hydrocarbon revenues would be an important near term policy measure. Source: World Bank. 111.2 FiscalNeutrality:Transfer of Revenuewith Expenditures 129. New resources from the HL have been transferred without transferring new responsibilities in a way that would clarify roles among levels o f government; and without providing opportunities for fiscal savings for the central government. Although the central government does not experience a revenue loss through decentralization (and could even still improve its fiscal position), the additional revenue should be used by sub-national governments for additional expenditures. 130. Inthe short term, to further improve Bolivia's fiscal situation, it would be desirable to consider decentralizing new expenditure functions without additional funding, both in departmental governments and municipalities, for example: 43 0 Departmental governments could adopt secondary roads expenditures in the amount o f US$70 million (in 2006) in return for the hydrocarbon resources revenues (in the amount o f US$120million in2006) without additional financing. 0 To compensate for the additional resources received, the at least nine capital municipalities could finance primaryhealth care. 131. This approach would need the verification that at least minimally acceptable institutional capacities are inplace before these expenditures are transferred. 132. Short of the possibility to transfer additional expenditure responsibilities, there are other options to improve the fiscal balance of the central government. The recent experience of other Latin American countries with regards to fiscal neutrality under decentralization can illustrate some useful options and measures (Box 7). Box 7. Achieving Fiscal Neutrality of Decentralization: Alternative Avenues and Strategies Peru was the first country in Latin America to decentralize in line with the principle o f fiscal neutrality. When it embarked on regional decentralization, the country established a one-to-one rule (pari passu). Under it, any new resource transfer to a sub-national government was to be accompanied by a corresponding cut at the central level. This has helped to protect fiscal sustainability and not distort vertical fiscal balances. The issuance o f the Law o f Fiscal Decentralization (2004), the regulations for the revamped Law o f Fiscal Responsibility, and the Law on Accreditation o f regional and local governments were designed to balance fiscal transfers and expenditure responsibilities in parallel with the reinforcement o f borrowing restrictions and strengthening managerial capacities inservice delivery at sub-national levels. Nicaragua is currently also attempting to apply this principle. Its 2004 increase in revenue-sharing to municipalities from 1 percent to 6 percent is projected to further rise to 10 percent in 2007. Since this would enlarge the fiscal deficit, several options for fiscal neutrality are now being discussed, including: (i) all central government agencies would cease activities that are assigned to municipalities but are nonetheless being administered centrally. (ii) Instead o f receiving transfers, municipalities would be provided vouchers which entitle them to receive investments by central governments intheir jurisdictions. (iii) New expenditure responsibilities are transferred to municipalities without additional funding. (iv) Municipalities would provide counterpart funds for foreign-financed public investment projects in their territories. They would be required to contribute larger shares o f project costs than inthe past. These options show clearly that fiscal neutrality for decentralization may not necessarily involve the transfer o f expenditure responsibilities-unfortunately often regarded as the only way o f achieving it. There are other ways that may provide quicker results, and more effective impacts, for achieving fiscal neutrality. Source: World Bank. 111.3 Creating a FiscalResponsibilityFramework 133. Bolivia has accumulated its own experience with fiscal responsibility at the sub-national level. Therefore, some lessons from the past could beusefully applied. These are that: The PRFpioneered some successfulincentive measures. 0 Sanctions need to be graduated to be credible. 44 0 The "20/200" rule system'' appears to be too lax an instrument for controlling sub- national debt. 0 Bolivia needs to rely mostly on administrative top-down regulation and oversight for controlling debt. 0 Mechanisms need to be put in place to clarify debt rules at municipal and departmental levels inorder to try to avert the enlargement of their debt problems. 134. Moving forward, it i s useful to assess the experience o f other countries facing similar challenges o f weak fiscal responsibility(Box 8). Box 8. The Challenge of Fiscal Responsibility Bolivia faces similar challenges to other countries whose sub-national governments gained fiscal autonomy without an adequate rules framework and fiscal discipline regime. To address this problem, some countries- including Brazil, Colombia, Peru, Argentina, India, New Zealand, and South Africa-have passed such legislation as fiscal responsibility laws (FRL). Their governments appear to have been motivated by related concerns: (i) an individual country's aversion to running an excessive deficit; (ii) a group o f governments in a single country want to enter into and enforce an agreement that each o f them would avoid such deficits. In either situation, the FRLwould function as a commitment device-in the first case, across time, to commit future governments; and in the second context, acrossjurisdictions, to wed governments ina unifiedpolicy. Several features o f these laws are o f particular relevance for Bolivia : 1. They require transparency through the publication o f information about the details o f programming and actual implementation o f finances throughout the public sector. 2. They restrain sub-national deficits by preventing them in advance andor by imposing penalties with early effects and supplementing normal curbs on fiscal imprudence, e.g., a. Impositions o f the national government on sub-national governments (SNGs). b. Impositions o f the SNGs themselves for managing their finances, including most importantly following the rules o f the national government 3. They restrain the national government from: a. Running unsustainable deficits b. Mitigating the consequences o f sub-national fiscal excesses. Source: World Bank. 135. Taking the international experiences with fiscal responsibility into account, Bolivia should consider adopting a set o f rules for sub-national governments within a national policy fiamework. These should be gradually implemented to give the new rules credibility-whose absence has beena major stumblingblock inseveral other Latin American countries. 136. For the short-tern, the following measures would be helpful, preferably to be implementedfor one year at atime: 0 Establish a different debt capacity indicator for municipalities that i s based on disposable revenue, not total current revenue as is currently the case. This implies taking into account rigid expenditures for humanresources and debt service. 19. The rule "20/200" establishes that a sub-national government cannot use more than 20 percent o f its current revenue for debt service, while the debt stock cannot exceed 200 percent o f current revenue. 45 While the Central Government continues to service departmental debt, the departments' access to new borrowing should be prohibited until they fully satisfy transparency requirements. Sub-national governments should: (i) the growth o f new hiringand salary increases; (ii) limit for those governments whose wage bills exceed a prudent percent of revenues, require a salary and hiring fieeze until attrition and/or economic growth bringthe wagehevenue ratio down to the acceptablerange; and iii)not enter into any new hiringor investment contracts in the last year of a department government's or municipality's term o f office. Require publication of information about the details of government programming and the implementation of funded activities throughout the public sector. 0 Link this publication with permissions to borrow (from abroad) and with financial sector regulations that only recognize loans as assets on bankbalance sheets iftheir key information i s available publicly. IV. ExpectedOutcome of the Short-Term Decentralization Scenario 137. The expected outcome of the short-term decentralization scenario is that there is a basic political consensus in place that would allow the national and departmental levels to focus decentralization on reducing inequities in resource distribution and in improving services. The intergovernmental system would be working in its basic elements: a minimum level o f transparency i s in place; and there i s basic equity in resource distribution from hydrocarbons revenue. These measures would the country allow to focus on clarifying roles and responsibilities and improving service delivery as part o f the pendingmeasures for the medium- term decentralization scenario. 138. To achieve the expected outcome, the following conditions should be givenconsideration and would need to be inplace: Actors can agree on abasic political consensus regarding the guidingprinciplesof decentralization. Actors are willing to introduce basic equality inthe distribution of resources. The central government is able to credibly commit to andwithstand a faster decentralization before basic institutionalcapacities are created. 46 CHAF'TER 5. POLICY OPTIONS FORA MEDIUM-TERM DECENTRALIZATION SCENARIO 139. This chapter lays out a set of policy options for the medium-term. Once the initial measures of the short-term scenario are taken, and the intergovernmental system is working inits basic elements-a minimumlevel o finstitutional capacities and equity inresource distribution is inplace-the country could move on and focus on the division of responsibilities and efficiency inservice delivery. This chapter first lays out the particular strategy that shouldbe followed. Ina second step, it describes the policy options to improve the institutional arrangements; and in a third step, it describes the policy options related to fiscal decentralization. The fourth and final section describes the expected outcome of the medium-term decentralization scenario. I.DecentralizationStrategyfortheMedium-Term:MaintainandImprovetheEfficiency of Public Services by Separating Roles and Responsibilities 140. The decentralization strategy for the medium-term is aimed at (i) maintaining, and (ii) increasing the efficiency of public expenditures. Departmental governments and municipalities need to prove that they are capable in maintaining a set of standards for quality of services (that have been defined as part of the preparatory phase); and then also improve them, previous to adopting more responsibilities. 141. For the second phase (medium-term) the essential measures to boost efficiency in departmental governments are the following: 0 Institutional: (i)separate clearly responsibilities among levels of government; (ii) basic once fiscal responsibility rules are working well, relax ear-marking for higher productive efficiency. I Fiscal: (i)Gradual introduction of funding formula for current services that provide incentives for producing better services; (ii) introduce mechanisms to reduce inter regional inequalities inthe services being delivered by regional governments. 11. InstitutionalArrangements: Clarify Responsibilitiesand ProvideMore Autonomy for Resource Management 11.1 Separation ofRoles and Responsibilities 142. Responsibilities among levels of government are disjointed. In light of the experience of other countries (Box 9), over the medium-term, it i s essential to clearly separate the roles of national, departmental, and municipal levels. This entails the following areas: Clarify the responsibility of departmental agencies like SEPCAM (road sector): either transform it into a deconcentrated agency o f a national road agency; or decentralize it in the third phase ofdecentralization towards departmental governments. Clarify roles of ministerial sector directors (health, education, others): nomination by central ministries or by departmental governments (ifthey are responsible for service delivery). 47 0 Clarify the management o f payroll on behalf o f departmental governments and match it with sectoral responsibilities (health personnel should only be managed if departmental governments are effectively responsible for providing health services). 0 Clarify the budget process: separate national and departmental budgets, so that over the medium-term the prefecturas have a budget that corresponds. Eliminate direct budgetary assignments from the Ministry o f Finance to departmental agencies (for instance for SEPCAM). Box 9. Separation of Responsibilities at the Regional Level: The Latin American Experience Intheir coordination of public functions, regional governments often appear to be more prone than municipalities to having overlaps and ambiguity inassigning responsibilities. An often overlooked reason for this is the influence o f problems inarrangements for human resource management, budgeting, and appointments o f key personnel (like sector directors at the regional level), rather than the contents o f laws and regulations. Regional governments throughout Latin America have had considerable experience with this. The Mexican states and Colombia's departmental governments are responsible for delivering primary education andor secondary education. However, only a small number o f their teachers are in fact hired locally. At the same time, mostly for political reasons, the central governments decide on these teachers' hirings and salary levels, with centrally and locally managed staff operating inparallel. Likewise, Peru's regional governments are responsible for managing payrolls for the national ministries o f health and education, even though these responsibilities are not yet decentralized and still under central government control. In Peru the process o f appointments o f regional sector directors (as in Bolivia) entails shared responsibility between regional governors and the national ministry. In Ecuador, there are both delegated and elected political authorities on the regional level. I t is clear that these and similar arangements are dysfunctional and undermine accountability. Mayors and governors cannot be held accountable for what they produce in the form o f services. Countries have responded with different strategies to these dilemnas. As exemplified by Peru and Ecuador, low performing regional management systems have been superseded. Although overlapping responsibilities and dual authority have been found to be dysfunctional, there have been new councils created inEcuador and Peru comprised o f both delegated and elected authorities (Peru: Consejo de Coordinacidn Regional, 2002; Ecuador: Comite Permanente de Desarrollo Provincial, 1997). Their usefulness is disputed, though, especially as they usually lack political and fiscal incentives to effect coordination. In the absence o f the political will in Bolivia to--or the possibility of- reorganizing regional structures, it seems essential to separate responsibilities at regional levels before they are deepened, rather than to try to invent new forms o fpolitical representation to address such problems. Source: World Bank. 11.2 Adjust ExpenditureEarmarking 143. It is important to adjust the ear-marking o f expenditures. Its rules and regulations are too complex, and its present level seems too highto permit more productive and allocative efficiency to materialize. The recommended corrective measures are the following: 0 Eliminate functional ear-marking (currenthnvestment expenditures) and replace it with the fiscal responsibility rules mentioned above that could help limit excessive human resource expenditures, and keep Bolivia within proper overall fiscal deficit parameters. 0 Establish sectoral ear-marking according to national and regional priorities (a specific percentage for health, education, roads). Once fiscal responsibility rules work well, gradually allow a higher share o f departmental resources to be allocated freely. Start by giving 10 percent more room in a given year, up to a 30 percent, so that a maximum o f 70 percent o f departmental resources is ear-marked. 48 0 Departments with quality o f services above a national standard are allowed to use resources inthose sectorsthat exhibit lower quality ofservices. 111. FiscalDecentralization: Sector Transfers for Fundingof Sub-national Expenditures 144. One of Bolivia's main challenges and opportunities for decentralization i s to bring about a change in incentives for better service delivery and correcting inequities. Current levels of funding, and the historic budget allocations are not conducive to accomplishing this goal. Therefore, it i s recommended to apply funding mechanisms as explained in Table 14. The proposedmechanism i s recommended due to its relative simplicity for implementation. There are other options for funding o f sub-national expenditures, which, for instance, would require redistributionamong departments. This inturnneeds conditions o f success that are not present in Bolivia at this point intime, and may evenremainunlikelyinthe long-tern. Type of Funding Explanation Condition of Success Transfer same amounts A formula is used to set an Fundingformula is simple and clear. as currently spent by incentive fiamework in each sector central government but that is output oriented, not input Funding formula is phased-in gradually so that with sector specific oriented: players accept the new incentive framework; and transfer formula a stable resource flow is guaranteed. 0 Education: per student formula Sufficient information is available so that current Health: per capita formula hnding levels can be identified. Sufficient information is available to make calculation o f new formula (number o f students, monitored year by year; population census, and so forth). Source; World Bank. IV. ExpectedOutcome of the Medium-Term Decentralization Scenario 145. The expected outcome of the medium-term decentralization scenario is that levels of government have taken steps to separate roles and responsibilities. Departmental governments would have clear expenditure assignments as a condition o f more accountability towards the citizens. There would also be incentives for more efficient service delivery, at the same that departmental governments obtain more autonomy in the use o f resources. Departmental governments would demonstrate that they are able to maintain, but also to increase the quality and efficiency o f public expenditures. This would allow the country to assign regions additional expenditures, and possibly also tax revenue. 146. To achieve the expected outcome, the following conditions should be given consideration and would need to be inplace. Departmental governments are willing and able to maintain standards o f service delivery and evenincreasethem. 0 Indicators for service quality are published. 49 0 The central government has the capacity to monitor departmental governments and to provide technical assistance. 0 Fiscal responsibility rules are working well so that ear-marking of expenditures can be relaxed inbenefit of higher productive efficiency. 50 CHAPTER6. POLICY OPTIONSFOR A LONGTERM DECENTRALIZATION SCENARIO 147. This chapter lays out a set of policy options for the long-term. Once the measures of the medium-term scenario are taken and effective-responsibilities among levels o f government are better defined and the quality o fpublic services hasbeenmaintained and improved-the country could move on and consider deepening the level of responsibilities o f departmental governments, both on the expenditures as well as revenue side. This chapter first lays out the particular strategy that should be followed. In a second step, it describes the policy options to improve the institutional arrangements; and in a third step, it describes the policy options related to fiscal decentralization. The fourth and final section describes the expected outcome o f the long-term decentralization scenario. I.DecentralizationStrategyfortheLong-Term:DeepeningResponsibilitiesof DepartmentalGovernments and Municipalities 148. Once there has been a successful application o f measures proposed in the short- and medium-term, it should be possible to deepen decentralization through giving regional governments and municipalities additional roles and resources.The key measures are: Institutional: (i)provide additional responsibilities, for instance inthe health, education, and road sectors; (ii) strengthen intergovernmental coordination. 0 Fiscal: (i) distribution of resources; (ii) refine consider providing departments with new own revenue inthe form o f taxes. 11. InstitutionalArrangements: Assignment of Additional Responsibilities and Strengthening Intergovernmental Coordination 11.1AssigningAdditional Responsibilities 149. Once the required institutional capacities are ensured in all departmental governments, the Government could then consider broadening their responsibilities. Inlight of the international experience (Box lo), this should be done through the recommended sectoral approach. This would facilitate the rationalization and coordination of responsibilities at the national, regional and municipal levels in an integrated fashion. This process should be guided by well designed and agreedconcepts ofthese future responsibilities (Table 15). 51 Box 10. Transfer of Responsibilities: The Benefits of a Sector Approach In most Latin American countries, the transfer of responsibilities has proceeded in a piecemeal fashion. Usually, municipal and regional levels have not done this inan integratedmanner. Unfortunately, these approaches have had high costs. While perhaps understandable politically, they have increased needs for coordination and made the assignment o f responsibilities more difficult. Colombia suffered this fate after its departmental governments became autonomous in 1991 following its municipalities' decentralization in 1986. Venezuela followed similar strategies during the 1990s. In Ecuador, the demands for "provincial autonomy" in 1999- 2000 and requests for the adoption o f functions were very disjointed, which led to substantial overlaps inresponsibilities. These and other cases showed that inappropriate decentralization can weaken accountability and cause unhealthy competition between different government levels. Technically, it i s best to follow a sector approach ingoing forward with transfers o f responsibilities. Chile did so over more than a decade with its explicit service delivery models involving municipalities. In Ecuador, in 2000 a sector approach was undertaken with framework agreements (convenio marco) set up in four sectors-roads, environment, agriculture and tourism. It was adopted after an assessment of the comparative advantages o f the provincial councils vis- a-vis municipalities (although an actual transfer o f responsibilities did not happen later). Spain has also followed an explicit sector approach: in health and education, among others. I t had intergovernmental committees set up that represented central government ministries and regional governments for the negotiations o f transfers. Bolivia is advised to follow such sector approaches. This will enable it to reap the benefits o f greater coordination, clearer assignment o f responsibilities and fewer overlaps between regions and municipalities. Source: World Bank. Education Health Roads Departmental Governments would have full Departmental Governments would be Departmental Governments responsibility to manage primary and responsible for primary and secondary incorporate SEPCAMs and secondary education. health care. obtain full responsibility for This entails: This entails: secondary roads within a national road sector strategy. 1. deciding when and where to open (i) Networkorganization:decide and close schools; how much health facilities a 2. human resource responsibilities Departamento would have, (according to options presented where they would be placed, o f above); what level o f complexity, and 3. deciding who will be the director o f a under which set o f referral rules school; they should be managed. 4. in-service teacher training; (ii) Humanresources:decisionto 5. acquiring and distributing textbooks hire and fire according to and other pedagogical materials; national rules (refer to section on 6. expanding, reducing, and human resources above). maintaining school infrastructure (in (iii) Infrastructure provision. coordination with municipalities); 7. deciding upon method o f service Although not ideal from a perspective delivery and pedagogical approach. o f assigning responsibilities, municbalities would continue to Although not ideal from a perspective o f providd for health infrastructure. assigning responsibilities, municipalities would continue to provide for school infrastructure. 52 150. To successfully proceed with the recommend sector approach, it would be important that there be an incentive framework governing the departments and sector ministries. These entities should only be eligible to receive new responsibilities and resources if they follow the framework's policies. This should entail the following: There needs to be a formal framework agreement (convenio rnarco) among all departments and the sector ministry (one agreementper sector). To keep transaction costs low, municipalities could sign a convenio rnarco collectively with the ministries, but grouped within a department. 151. Given the limited institutional capacity at national and sub-national levels, and the volatile resource flow from transfers, the following sequence should be followed: 0 first, transfer secondaryroads; 0 second, transfer responsibilities for health; 0 third, transfer responsibilities for education. 152. There could be different treatment for municipal and departmental governments in dealing with primary vs. secondary education, as well as primary vs. secondary health care. Although due to the small scale in many o f the rural areas of the country a splitting of these responsibilities i s not practical and recommended, it should be given consideration what role capital municipalities could play in assuming these responsibilities intheir jurisdiction. 11.2 Intergovernmental Coordination 153. In the long-term, it is essential to strengthen intergovernmental coordination. The framework agreements mentioned above are only one way for assuring clear responsibilities, as well as improvinghuman resourcemanagement. Both should exist as long as the transfer process continues. In the future, institutions should be created to help ensure good coordination o f expenditures and policies in each of the sectors. Following Spain's example (Box 1l), national- regional sectoral coordinating bodies could be useful (depending on the expenditures that are transferred). Box 11.Asymmetric Decentralization and the Emergence of Sectoral Coordination: the Case of Spain Spain is a good case o f balancing well the dynamics o f an asymmetric and a sectoral decentralization strategy. Individual regions were given the possibility to adopt a broader range o f functions in an asymmetric arrangement; and for each o f the sectors involved, such as health and education, intergovernmental committees were set up that represented central government ministries and regional governments in the negotiations o f transfers (convenios). Once the transfer process was completed, participants became aware o f the usefulness o f this institution. In fact, the sectoral committees still exist and fulfill important functions for intergovernmental coordination. However, there has been criticism that these committees cannot meet on initiative o f regional governments. Nonetheless, this arrangement is a good example o f how decentralization triggered intergovernmental coordination; and how the center maintained its ability to articulate sectoral policies while keeping costs in check. It may constitute a reference for all countries in which the impetus for reform comes from bottom-up, or individual regions, but have to be kept inline with good national policies. Source: World Bank. 53 111. FiscalDecentralization: Improve Distribution and Increase FiscalAutonomy 111.1 AddressingHorizontalImbalances: Improvingthe Distribution of Resources Among Departments 154. Once a new per capita-basedtransfer system i s operative and has contributed to achieving a minimum of inter-departmental equity, it could be considered to refine the methodology and transfer formula. Such refinedmethodologies however require better information and are often less transparent than the simple per capita formula. Such methodologies include, inter alia: weighted indices of relative need; 0 per capita (per client) expenditure norms; bottomup or physical expenditure norms (representative expenditure approach); sector formula (ifthe approach i s to finance different sectors, and not to opt for a general purpose grant). 155. An additional consideration that could betaken into account is that, once there is taxation at the departmental level (discussed below), the formula could change to one which would compensate for relative lack of fiscal capacity (as in Canada, for instance). This opens the possibility of using less complex methodologies that are implied when estimating expenditure needs. The idea of an equal share per region in order to meet fixed costs is also acceptable, providedit remains a small part o f the total. 111.2 Elimination of Earmarking of Revenue to Regions and Poolingof Different Resources 156. The total amount of central government transfers to regions and municipalities should be determinedinrelationto total central government resources, not inrelation to royalties. There is nothingabout royalties that justifies their ear-marking as a source o f intergovernmental transfers. Its use should be eliminated. 157. Pooling of revenue is also a first step to reduce volatility in flow o f revenue to regions in particular. In addition to this measure, the country might deliberate on other, explicit mechanisms to stabilize intergovernmental transfers. Indoing so, a decision needs to be taken on which level of government would assume the risk of volatility: the national government alone; subnational governments; or whether the risk is shared among both levels. Whatever mechanism i s selected, a stabilizing rule-particularly with regards to withholding revenue increases during boom periods-could assist subnational governments in avoiding the boom-bust trap that has afflicted so many countries in LatinAmerica. 158. Designing a good rule requires determining whether the economic shock is a temporary or permanent shock, and whether policy makers will be able to determine the difference. If the income shock i s permanent, then by necessity subnational governments will have to adjust downwards their expenditures. To stabilize the permanent shock (or if permanence i s not known), it would make sense to follow an intermediate path. One alternative i s to implement a moving-average type of transfer rule, not fully protecting the subnationals from the required adjustment, but giving them more time to adjust because o f their usually more limited (costly) access to credit compared to the central government. Another alternative would be a set o f 54 individual subnational stabilization funds, so-called "rainy day" funds. Countries in Latin America have experimented with different mechanisms, often as a combination o f moving averages, floors, and/or fixed sums (Table 16). Table 16. StabilizingMechanismsfor IntergovernmentalTransfers in Selected CountryCases Country StabilizingMechanisms Cases Argentina Between2003-2005, provinces would start to receive a moving average o f the three most recent years shared revenue amounts. Incase this moving average were to coincide with recessionary or low growth years, a minimumamount i s set: $1.4 billion per monthin 2003, $1.44 billion in2004, and $1.480 billion in2005. These minimumamounts represent approximately 2.6 to 2.8 percent increase per year innominal terms. Colombia The country has tried to limitthe growth o f total transfers. Duringa transition of 2002- 2008, transfers would grow annually at the rate o f annual inflationplus 2 percent for 2002-05 and 2.5 percent for 2006-08, thereby creatingboth a floor and a ceiling. Escape clause for the ceiling: ifGDP growth exceeds 4 percent, the ceiling would no longer apply andtransfers would increase inproportion to the national revenue growth. Ecuador FODESEC (Fondo de Desarrollo Seccional): transfers can never be lower than inthe previous year. Venezuela Oil stabilizing fund Source; World Bank. 159. It i s essential to bear in mind that some o f these mechanisms have seriously jeopardized the fiscal position of national governments in certain periods o f time, or have had only limited effectiveness. Ecuador, for instance, has beenafflicted by the infamous "ratchet effect" by which transfers can never be lower than in the previous year. In Argentina, floors on transfers have contributed to the fiscal stress under which the federal government was inthe crisis following the year 2000. InVenezuela, the federal government had only weak discipline in effectively saving funds ofthe oil stabilization fund andwere spent. 160. These country experiences therefore underscore that stabilizing rules for intergovernmental transfers have to be complementary to national and subnational fiscal responsibility rules, budget planning, and debt management. Also, if the national government itself i s at the brink of insolvency, or if there are long-term structural imbalances at the subnational level that are being financed by the current system o f transfers, then stabilizing transfers may only complicate or even worsen these initial problems. Therefore, before any decision i s taken on specific stabilizing mechanisms in Bolivia, the following conditions o f success would need to be inplace before establishing a stabilizingrule to national-to-subnational transfers (Box 12). Box 12. Three Pre-conditions for the Stabilization of Intergovernmental Transfers (i) Subnational governments are credit constrained-rationed insome way out of the market, or subnational governments confront substantially higher cost o f borrowing. (ii) The national government possessesstable access to credit and quality debt management. (iii) There are no severe structural fiscal imbalances, either within levels o f government or across levels o f government. In other words, neither individual level o f government is facing unsustainable cyclically- adjusted fiscal deficits. In addition, the subnational governments are not spending excessively, with financing by automatic national transfers. Source; Gonzalez, Rosenblatt, and Webb (2002). 55 111.3 ReducingTransfer Dependenceby StrengtheningOwn RevenueBases 161. Strengthening own-revenue bases and diversifying income are important strategies for future decentralization, as the experiences of other Latin American countries have shown. The lack o f own revenue bases was a major stumblingblock for many regional governments intrying to deepen autonomy. Peru, Colombia, Ecuador and Venezuela have demonstrated that, if regions do not obtain reasonable access to their own sources of revenue, their fiscal dependence on the center will continue to dictate their day-to-day management. Additional advantages o f greater dependence on own-revenue include the importance of local politicians "feeling the pain o f the marginal tax boliviano they spend." Then too, local taxation can contribute to more efficient spending, fostering hardbudgetconstraints, and stimulating greater accountability to citizens. 162. It is notable thus that, compared to other countries inthe region, regional governments in Bolivia are particularly transfer dependent (Table 17). In the long-term, therefore, providing departmental governments access to own revenue sources would be desirable to reduce their transfer dependence (which i s otherwise likely to increase even more). Table 17. Transfers andRevenueof RegionalGovernment inSelectedCountries(percent) Transfers as Share of Total Revenue OwnRevenue as Share of Total Revenue Bolivia (2005) 96.60 3.4 Venezuela (2000) 98.39 1.61 Colombia (2000) 59.9 40.1 Ecuador (1998) 92.8 8.8 Peru (2000) 93.8 6.2 Spain (2001) 76.69 23.31 Source: World Bank. 163. The selection of appropriate tax sources for regions is extremely challenging, partly because o f competing claims. The national government keeps many taxes for equity reasonsjust as the muicipalities keep others for administrative or efficiency considerations. The central government i s considered more to administer revenues from the personal income tax, corporate income tax, and those related to natural resources. Conversely, municipalities are believed to fare better in administering the property tax, the residence-based individual income tax, and payroll taxes levied at the place o f employment (when considering the principle of subsidiarity). An additional particular problem for Bolivia i s that the regions are relative late-comers to decentralization while the municipalities gained a "first-mover advantage." They were thereby been granted new tax bases, some of which would be suitable for regional administration. 164. In the face of similar constraints elsewhere, regional governments in several countries have either had very limited own revenue, or have been quite creative, e.g., in some cases, imposing surcharges on national or municipal taxes (Table 18). 56 Table 18. Co ntry Casesof RegionalGovernmentTax Administration Type of Tax Country Cases Property 0 Spain, Ecuador: municipalities administer these; regional governments are allowed to levy a taxes tax on changes in land ownership. 0 France: departments share these taxes with municipalities (which obtain roughly a third o f the total), and regional governments are allowedto piggy-back on top o f them. Mexico, Argentina: state/provincial taxes with revenue-sharing for municipalities. Personal 0 Spain (tax i s administered by the central government, but regional governments can regulate Income Tax the regional tax schedule, create their own deductions for individuals andhouseholds, non- corporate investments and the application o f income-provided these do not affect other taxes levied on any category o f income). 0 Mexico 0 Argentina (Impuesto a Ingreso Bruto) Payroll tax Mexico 0 Australia South Africa Excise Tax Colombia: departments collect `sin taxes' ("impuestos del vicio") as charges on liquor, beer and tobacco. Spain: regional governments charge taxes on bingo and surcharges on gaming (which are administeredby municipalities). 0 Basque country (Spain) specific-status regional government charges excises on manufacturing. Retail Sales 0 USAstates Tax 0 Specific-status government o f Navarre (Spain); also charges sales taxes on luxury goods. VAT 0 Brazil 0 Canada (Dual VAT; 5 different VATS). Canary Islands government (Spain) levies an indirect general tax, with a similar structure to that o f VAT, albeit with fewer rates and without taxing the retail trade stage. Business France: departments share these with municipalities (which obtain roughly a third o f the total Taxes value). Spain: single-province regional governments can set surcharges on the municipal tax on businessactivities. Basque country (Spain) establishes flexible rates and concessions on the corporate income tax. Automobile Colombia: departments administer a vehicle tax. registration 0 France: regional governments apply a tax on driving licenses; departmental governments own and annual and administer the yearly automobile tax. license taxes Road tolls 0 Interregional highways have beenprivatized or concessioned inChile, Colombia, Peru and Ecuador, with several regional governments participating intoll proceeds. Fuel taxes 0 Canada: administered by the federal government and provinces ina tax-sharing arrangement. Source: World ank. 165. In light of this international experience, and given the particular tax structure and conditions present in Bolivia (no personal income tax inplace; no functioning tax administration at regional levels; low transparency and poor exchange of fiscal information among levels o f government), there are two possible revenue sources that could provide a starting point for regionaltaxation there: (i) 1:thespecialconsumptiontax(domesticshare); Option (ii) 2:asalestaxongasolineanddieselfuel. Option 57 166. With sometax administration considerations explained below (Table 19), these two taxes would inprinciple accurately mirror local fiscal capacity, while also not creating adverse effects on equity or efficiency. Giventhe experience o f other countries (e.g., Colombia), they could also constitute suitable regional taxes. They cannot be easily exported from the taxing jurisdictions (contrary to foreign trade taxes, for example). They would not readily or very much influence locational choices (contrary to the corporate income tax, for example). They are not too difficult to administer locally (contrary to the VAT); and they are not used for inter-regional redistributive purposes (contrary to the personal income tax. Table 19. Options for Regional Taxation inBolivia Tax Base Tax Rate Tax Administration Appropriation of Tax Proceeds Option 1: Regional Different Alternatives: Regional Domestic share o f percentages a) National with a Special Consumption depending on commission for central Tax (ICE) goods. government. b) Regional, revenue collected at the retail level. c) Regional, collected at the producer level. Option2: Regional Between zero and 5 Regional. Regional Sales tax on gasoline percent. Alternative: national and diesel fuel Region can choose with a commission for the rate. central government. Source: World Bank. 167. Other possible options o f regional revenue include the automobile tax, although that i s a municipal tax. Furthermore, on political economy grounds, it seems unlikely that this tax be transferred to regions. The same holds for the property tax. 168. I t i s essential to bear in mind that the introduction o f regional taxation based on the two options suggested above is not fiscally neutral to either level of government. This i s of course true for regional governments which would obtain more revenues. This i s different for the national government which would experience a revenue loss by transferring the domestic share o f the ICE to regions. Municipalities would also be affected due to the decline inrevenue-sharing as a result o f the changes occurring at the national level. These fiscal impacts need to be carefully assessedprior to granting regions more access to additional revenue (Table 20). 169. Tax decentralization could be combined with expenditure decentralization. This could be aimed at phased transfers o f responsibilities in conjunction with increases in own tax revenues and fiscally-neutral (i.e., corresponding expenditure cuts at the national level) transfers. If Bolivia proceeds asymmetrically-that is, not all departments proceed at equal paces and depth-one could first transfer responsibilities inthe more fiscally able departments, then extend the process to the others as they improve their fiscal capacities. In any case, this would require the guarantee of sufficient transfers to the less fiscally capable departments to enable them to cover minimal service requirements. 170. Another option i s to explicitly condition the transfer of tax authority with expenditure responsibilities. For instance, if a departmental government wants to administer the secondary 58 roads network, the central government would condition the adoption o f this responsibilityfor the collection o f the sales tax on gasoline and diesel fuel (Tax Option 2). This would set an additional incentive to collect taxes. On the other hand, ear-marking o f specific taxes to specific expenditures should be avoided.20 Table 20. Regional Taxation and Revenue Net Gainfor Departmental Governments (in Million US$), 2006-2010 2006 2007 2008 2009 2010 Central Government Option 1: Revenue loss to allow dep. to collect dom. ICE -62.8 -67.3 -71.3 -75.6 -80.2 Total Net Gain -62.8 -67.3 -71.3 -75.6 -80.2 Departmental Government Option 1: Tax collected through domestic ICE 41.9 75.1 79.6 84.5 89.6 Option 2: Gasoline and Diesel Sales Tax 31.4 33.6 35.6 37.8 40.1 Total Net Gain 73.3 108.7 115.2 122.3 129.7 Municipalities Total Net Gain (Revenue loss in co-participation) -12.6 -13.5 -14.3 -15.1 -16.0 Note; differences in tax revenue from ICE result from the fact that currently the domestic ICE is administered as a VAT, Incase this tax is decentralized to departments, its administration should only be done as a final sales tax. Source: World Bank. 171. Whatever the final choice is, there are several conditions of success that need to be met for desirable regional taxation: 0 The central government is willing to accept revenue shortfalls inrevenue due to losses of domestic shares of the ICE. 0 Departmental governments are willing to deal with the political cost o f new taxation, particularly a sales tax on gasoline (which yields the highest amount of revenue). They would have to have efficient tax administration departments (although in the first instance, the central government could administer regional taxes). 0 Municipal governments can bear the temporary shortfalls in co-participation transfers inferredinrevenue changes at the central level. 111.4 Consolidating FiscalResponsibility 172. The country should gradually buildingup commitment to fiscal responsibility among all levels o f government. Over the long-tern, the following measures are important: 0 Restraint o f sub-national deficits by preventingthem in advance and/or by imposing extra penalties that take effect quickly, in addition to the normal measures taken to deal with the consequenceso f fiscal imprudence. 20. Ear-marking is questionable: it reduces the freedom o f regional governments in resource allocation, ignores regional specificities (road maintenance may be more costly in some regions, or river transportation may be an option only incertain regions), and implies complicated controls. 59 0 Once annual fiscal targets are working well, they can be extended to multi-year bases. These targets could then be mandatedinrevisedfiscal responsibilitylaws. IV.ExpectedOutcome of the Long-TermDecentralization Scenario 173. The expected outcome of the long-term decentralization scenario is that departmental governments have deepened their responsibilities both on the revenue as well as expenditure side. An increasing share o f expenditures i s financed from own revenue in the form o f taxes. This would gradually create more lines of accountability towards the tax-payers andcitizens. 174. To achieve the expected outcome, the following conditions should be given consideration and would need to be in place. They are different for central, departmental, and municipal governments. Central Government: 0 Has the capacity to redistribute resources within sectors and among departments previous to transfer of responsibilities (so that decentralization i s fiscally neutral). 0 Can negotiate transfers o f health and education personnel to departmental governments. 0 Has the fiscal space to provide additional tax bases or tax revenue for departmental governments. Departmental governments: 0 Can cope with new expenditure responsibilities and have both the institutional as well as absorption capacities to deliver the new sector services. 0 Are willing to collect taxes. Effectivetax administration canbe established. Are willing to hireand fire sector personnel. Use their own resources to finance sector infrastructure. 0 Will have additional funding available inthe future, e.g., for secondary education for which improvements are a so-called "unfunded mandate." 0 Fill redistributive roles intheir jurisdictions (education, above all). 0 Clarify their institutional frameworks particularly with regards to the hybrid roles of sector directors, and budgeting. 0 Have effective budget constraints inplace, especially limits on personnel expenditures. Municipalities: 0 Establish effective coordinating mechanisms for delivery of infrastructure works inthe main sectors. 0 Accept expandedroles of departmental governments inservicing key sectors. V. Vision of FutureDepartmentalGovernments 175. The implementation of the policy options in the short-, medium-, and long-term should contribute to a strengthening o f departmental governments. Decentralization represents an opportunity to achieve social inclusion. Successfully strengthening regional governments for these purposescouldbe very salutary. 60 176. Inline with the above, the following are the expected outcomes as a result of the policy options laid out inthis Report. There are mechanisms inplace to process conflicts among levels of government inbenefitof better governance. There would be a minimumlevel o f transparency and flow o f information that would allow all actors to constantly revise and adjust the decentralization strategy in light ofnew circumstances. There is a minimum level of horizontal equity in resource distribution from hydrocarbons revenue, which would lead to more equitable spending. Responsibilities o f levels o f government would better separated. The different sectors would have coordinating mechanisms inplace that make overlaps inresponsibilities transparent. There would be more autonomy in the use of resources, which in turn would allow departments to use inputs more efficiently. Citizens would have the ability to monitor the quality of public services at the departmental level. Departmental governments would bemore accountable to the citizens and fax-payers. 61 62 Annex Table A.l. Balances per Level of Government %sa share of GDl) 2004 2005 2006 2007 200x 2009 2010 :entral Government went Revenue 24.7 28.8 32.3 31.8 30.1 29.2 28.4 Tax Revenues 19.4 24.6 27.1 26.7 25.9 25.3 24.8 Hydrocarbons 5 0 9 2 12.5 11 9 11.0 10.3 9 6 Royalties 2 3 2 2 4 0 3 8 3 5 3 3 3 1 IDH I 1 3 2 2 3 2 1 2 0 1 s 1 7 Other 1 6 3 8 6 3 6.0 5 5 5 2 4 8 Other 14 4 I 5 4 14 6 14 8 14 9 I 5 0 15 2 Direct Taxes 3 6 4 4 3.8 3.6 3.6 3 6 3 8 Indirect Taxes 10 8 I 1 0 10 8 11.2 1 1 . 3 I 1 4 I 1 4 Financial Transactions Tax 0 5 0 8 0 7 0 7 0 7 0 7 " 7 of which: addirional fax revenues 0 7 0 7 0 7 0 7 Non Tax Revenues 2.8 2 0 2 3 2 2 2.2 2 0 I.8 Grants 5.5 2.2 2.9 2 9 2 0 1 9 1 8 ' w e n t Expenditures 22 0 23 7 23 6 24 0 23 3 23.0 22.6 Health 0 3 0 8 0 3 0 3 0 3 0 3 0 3 Education 0 3 0 3 0 3 0 3 0.3 0 3 0 3 Roads 0.5 0.5 0.5 0.5 0.5 0.5 0 . 5 Interest Payments 2.9 3 1 2.6 2 7 2.7 2.7 2.7 Transfers 9.3 11 0 10.5 10.3 10.0 9 8 9 6 Coparticipation 2.4 2.6 2 5 2 . 5 2.5 2 6 2 6 Royalties 1.6 1.6 2 8 2 7 2.5 2 3 2 2 IDH 0.6 2 0 1.I 1.4 1.3 1.3 1.2 Education Payroll 3.9 3.9 Health Payroll 0.8 0.8 Tronsfer for social services 3 7 3 7 3.7 3 7 3 7 Other Current Expenditures 8.8 8.5 9.2 9.9 9.5 9 4 9.2 Capital Revenues 0.2 0 0 0.0 0 0 0.0 0 0 0.0 Capital Expenditures 5.1 4.3 6.3 8.5 7 5 7.2 6.9 Health 0.0 0 1 0 1 0 1 0 1 0.1 0 1 Education 0.0 0 0 0.0 0 0 0.0 0 0 0.0 Roads 2.3 2.5 2 6 2.6 2.6 2.6 2.6 Compensation Fund 0.1 0.2 0 2 0.2 0.2 0 2 0.2 OthR 2.6 I 5 3-4 5 6 4.6 4 3 4.0 herall Balance o f CG -1.2 0.8 1.4 -0.7 -0.7 -1.0 -1.1 lepartmental Government ' w m t Revenue 7 2 8.0 7 8 7.7 7.4 7.2 7 0 Tax Revenue 0 1 0.1 0 1 0 1 0 1 0 1 0 1 Transfers from CG 7.1 7.9 7 7 7 6 7 3 7.1 6 9 Royalties 1.6 I.6 2 8 2 7 2 5 2 3 2.2 IDH 0 3 I 1 0 7 0 7 0.7 0 6 0 6 Educationpayroll 3.9 3 9 Health payroll 0 8 0 8 Transfer for social services 3 7 3 7 3 7 3 7 3.7 OthR TranSfR.S 0 5 0 5 0 5 0 5 0 5 0 5 0.5 'urrcnt Expenditurss 5 7 6 3 6.1 6 0 5 8 5 8 I 8 Health 0 8 0 8 0 8 " 8 0 9 1 0 I O Education 4 " 4 0 4.0 4 1 4 1 4 2 4 2 Roads 0 1 0 1 0 1 0 1 0 1 0 1 0 1 Interest Payments 0 1 0 1 0 0 0 0 0 " 0 0 0 0 Other Current Expenditures " 7 1 3 I 2 I O 0.8 0 6 0 5 Capital Revenues 0 4 0 5 0.5 0 5 0 5 0 5 0 5 Compensation Fund 0 1 0 2 0 2 0 2 0 2 0 2 0 2 0th- Transfers 0 3 0 3 0 3 0 3 0 3 0 3 0 . 3 Capital Expenditures 1 7 2.0 2 3 2 3 2.1 1 9 1 8 Health 0 1 0 1 0 1 0 1 0 1 0 1 Education 0 0 0.0 0 0 0 0 " 0 " 0 Roads 0 5 0 6 0.6 0 6 0 6 0.6 Other I.3 I.6 1 6 1 4 1 2 1 , WeraII Balance of DG 0.3 0.1 0.0 0.0 0.0 0.0 0.0 lunicipalGovernment went Revenue 4 3 5.2 4 8 4.6 4.6 4.6 4.6 Tax Revenue 1.1 1 1 1 1 1.0 1 0 1 0 1.D Non Tax Revenue 0.8 0 7 0 6 0.5 0 5 0.5 0.5 Transfers from CG 2.4 3.4 1.1 3.1 3 1 3 1 3 1 Coparticipation 2.4 2.6 2.5 2 5 2 5 2 6 2 6 IDH 0.8 0 6 0 6 0 5 0 5 0 5 w e n t Expenditures 1 4 1 7 2.5 2 s 2.4 2.4 2.4 Health 0.2 0.2 0 2 0 2 0.2 0 2 0.2 Education 0.2 0.2 0 2 0.2 0 2 0.2 0.2 Roads 0 1 0 1 0.1 0.1 0.1 0.1 0.1 Interest Paymen- 0 . 1 0 , 0.1 0.1 0.1 0.1 0 1 Other Current Expenditures I 1 1 . 1 2 0 2 0 I.9 1 9 1 9 of which: new responsabiliries 0 9 0 9 0.8 O B 0 8 Capital Revenues I 5 1.4 1.6 1.5 1.5 1 5 L S Capital Expenditures 4 1 4 6 3.8 3 6 3.7 3 7 1 7 Health 0 2 0.2 0.2 0.2 0 2 0 2 0 2 Education 0 6 0 6 0 6 0.6 0 6 0 6 0.6 Roads 0 5 0.5 0 5 0.5 0.5 0 5 0.5 0th- 2 7 3.2 2 4 2.2 2 3 2.3 2 3 ,vera11Balance of MG 0.3 0.3 0.0 0.0 0.0 0.0 0.0 Iemo Items: Jb116 entorpn,seoperatins balance 0.1 0.1 1.1 0.7 0.7 0.6 0.5 >cia1Seovriry -4.3 -4.1 -4.1 -4.0 -3.7 -3.5 -3.6 vera11 Balanoe o f M P S -5.9 -2.6 -0.5 -4.0 -3.8 -3.9 -4.2 entral Bank operating balonEe 0.3 0.3 0.4 0.5 0.5 0.5 0.5 vernll Ealan~e f CPS o -5.6 -2.3 -0.1 -3.6 -3.3 -3.5 -3.8 63 64 BIBLIOGRAPHY Agosta, Roberto S. (2005): Informe Final--EvaluacGndel funcionamiento del Servicio Nacional de Caminos como resultado de la reforma institucional aplicada desde 2002. Republica de Bolivia, Programa de Reforma Institucional, Abril2005, La Paz, Bolivia. Akizuki, Kengo (2002): Controlled Decentralization: Local Governments and the Ministry o f Transportation and Home Affairs in Japan. World Bank Institute, May 2001, Washington, DC. Alborta, Patricia, Montalvo, Marcel0 and Zapata, Marco (2005): Bolivia: Regimen de transferencias intergubernamentales para las autonomias regionales. USAID, La Paz, Bolivia. Alonso-Biarge Jose M., and Alexandra Ortiz (2000): Decentralization o f Road Network Managements: Lessons from Bolivia and Some hero-American Countries. Private Sector Development and Infrastructure, Transport Division, Discussion Paper TWU-37- World Bank, Washington, DC. Arrow, KennethJ. (1951): Social Choice and IndividualValues. Wiley, New York. Ayo, Diego (2002): Municipalismo-Participacih Popular, Apuntes de un proceso. Muela del Diablo, La Paz, Bolivia. Ayo, Diego (ed.) (2004): Municipalizacion: Diagnbstico de una dkcada, 30 investigaciones sobre Participacih Popular y Descentralizacibn. Friedrich Ebert Siftung, USAID, Plural, La Paz, Bolivia. Ayo, Diego (ed.) (2004): Voces Criticas de la Descentralizacibn, Una decada de Participacion Popular, 14 entrevistas. Friedrich Ebert StiftungPlural, La Paz, Bolivia. Banco Central de Bolivia (2002): Boletin Estadistico No. 316. Diciembre 2002, La Paz, Bolivia. Banco Mundial (2002): Evoluci6n de 10s PRF a Diciembre del 2001 y Plan de Acci6n 2002. Consultoria por parte de Cesar Vargas, Washington D.C. Bardhan, Pranab and Dilip Mookherjee (2000): Capture and Governance at Local and National Levels. American Economic Review 90:135-139. Barrios Suvelza, Franz X. (2004): Dictamen especial sobre la descentralizacih en Bolivia: Vision sistemica, relaciones fiscales y reingenieria prefectural. Ministerio sin Cartera Responsable de Participacibn Popular, Viceministerio de Fortalecimiento para la Gesti6n Departmental, La Paz, Bolivia. Barrios Suvelza, Franz X. (2002): El Estado Triterritorial. Friedrich Ebert StiftunglPlural, La Paz, Bolivia. Barrios Suvelza, Franz X. (2004): La descentralizacih en la Constitucih Boliviana, Diagnostic0 Critico. Ministerio de la Presidencia de la RepublicdAgencia Espaiiola de Cooperaci6n Internacional, La Paz, Bolivia. 65 Barrios Suvelza, Franz X. (Coord.) (2002): Estudio de Evaluation de Capacidades Institucionales y de Gestion Prefectural. Ministerio de la PresidencidEmbajada de 10s Paises Bajos, La Paz, Bolivia. Barro, Robert J. (1991): Economic Growth in a Cross Section o f Countries. Quarterly Journal o f Economics 106:407-444. Behrman, J., P. Senguptaand S. Parker. (2005): Progressing through PROGRESA: An Impact Assessment o f a School Subsidy Experiment in Mexico, in: Economic Development and Cultural Change (forthcoming). Blanes, Jose (1999): La Descentralizacion en Bolivia: Avances y Retos Actuales. Parlamento Andino, Organizacion de Estados Americanos, Facultad Latinoamericana de Ciencias Sociales Sede Ecuador. Cabana, Marcelo; L. Liataud, and F. Faiz (1999): Area-wide Performance-based Rehabilitation and Maintenance Contracts for Low-Volume Roads. Paper presented at the Seventh International Conference on Low-Volume Roads, May 23-27, 1999. Chrdenas, Marina (2004): Estudio de cuentas nacionales de financiamiento y gasto en salud. Ministerio de Salud y Deportes, DFID, La Paz, Bolivia. Chrdenas, Marina (2004): Estudio de cuentas nacionales de financiamiento y gasto en salud. Ministerio de Salud y Deportes, DFID, La Paz, Bolivia. Contreras, Manuel and Talavera, Maria Luisa (2004): Examen Parcial: La Reforma Educativa Boliviana 1992-2002. Fundacion PIEB, La Paz., Bolivia. Coronado, Ximena; Fernandez, Fernando; Lora, Oscar; Santa Cruz, Ludmila (2003): Situation Fiscal-Financiera y de Endeudamiento de 10s Municipios en Bolivia. Redde Analisis Fiscal (RAF), Ministerio de Hacienda, LaPaz, Bolivia. Czerniewicz, Nicole K. (2004): Analisis exploratorio de la descentralizacion de 10s sectores agropecuario, salud, educacion, saneamiento basico, medio ambiente, caminos y poder judicial. Ministerio de la Presidencia, Unidad de Coordinacion para la Asamblea Constituyente, La Paz, Bolivia. Czerniewicz, Nicole K. (2005): Desconcentracion Territorial de 10s Sectores: agropecuario, salud, educacion, saneamientobasico, medio ambiente, caminos y poder judicial. Plural, La Paz, Bolivia. Danish Ministry of Transport (1998): Road Safety in Latin America and the Caribbean Analysis of the Road Safety Situation inNine Countries. RoadTransportation Directorate. Faguet, Jean-Paul (2000): Does Decentralization increase Government Responsiveness to Local Needs? Decentralization and Public Investment in Bolivia. London School o f Economics, Center for Economic Performance, Working PaperNo.999. Faguet, Jean-Paul (2004): Does decentralization increase government responsiveness to local needs? Evidence from Bolivia, in:Journal of Public Economics 88:867-893. Fiske, E.B. (1996): Decentralization o f Education: Politics and Consensus. World Bank, Washington D.C. 66 Fondo Nacional de Salud de Chile (2005): Cuenta Pcblica de Gesticin2004. www.fonasa.gov.cl, Santiago, Chile. Frank, Jonas (2002): Competencias: Que Descentralizar?.GTZ-CONAM,Quito, Ecuador. Frank, Jonas (2004): Decentralization in Ecuador: Actors, Institutions, and Incentives. Doctoral Dissertation, Potsdam University. Frank, Jonas (2006): Decentralization to the Regional Government in Latin America. World Bank,Washington DC, forthcoming. French Ministry for Infrastructure, Transport, Housing, Tourism and the Sea (2003): Roads management and land transport in France. Direction des Affaires Economiques et Internationales, France. Galindo, Mario: Analisis comparado de propuestasde Descentralizaci6n. ILDIS,La Paz, Bolivia Gamarra, Eduardo A. (1997): HybridPresidentialism and Democratization: The Case of Bolivia, in: Scott Mainwearing and Matthew S. Shugart (eds.): Presidentialism and Democracy in LatinAmerica, Cambridge UniversityPress, New York. Gamarra, Eduardo A. and James M.Malloy (1995): The Patrimonial Dynamics o f Party Politics in Bolivia, in: Scott Mainwaring and Timothy R. Scully (eds.): Building Democratic Institutions: Party Systems inLatin America. Stanford UniversityPress, Stanford. Garcia Salaues, Alfonso (2004): Deuda municipal, in Municipalizacibn: diagnbstico de una dkcada. Tom0 I,ILDIS/USAID, La Paz, Bolivia, p. 63-100. Gibson, EdwardL.ed. (2004): Federalism and Democracy in Latin America. The Johns Hopkins Press. Ba1timore. Gonzalez, Christian Y.; Rosenblatt, David; Webb, Steven (2002): Stabilizing Intergovernmental Transfers in Latin America: A Complement to NationaVSubnational Fiscal Rules? World Bank Policy Research WorkingPaper 2869, Washington D.C. Gordo, Luis; Hernhndez de Cos, Pablo (2000): The Financing Arrangements for the Regional (Autonomous) Governments for the Period 1997-2001. Banco de Espaiia, Servicio de Estudios, Documento de Trabajo no. 0003, Madrid. Greenwood, P.W., K.E. Model, C.P. Rydell and J.R. Chiesa. (1998): Diverting Children fiom a Life o f Crime: Measuring Costs and Benefits. RAND, Santa Monica. GTZ (2005): Aspectos institucionales sobre el control del endeudamiento subnacional en Bolivia. Analisis para el Institutional and Governance Review (IGR 2005-2006), Banco Mundial, La Paz, Bolivia. GTZ-CONAM (2002): Linea de Referencia (Segunda Edicibn), Los procesos de descentralizacih en Argentina, Bolivia, Brasil, Chile, Colombia, Ecuador, Peni, Venezuela. Proyecto Descentralizacih, GTZ-CONAM, Quito, Ecuador. Gupta, Monica and Rani, Manju (2004): India's Public Health System:How well does it function at the national level?. World Bank policy researchworking paper 3347, Washington D.C. Hans Seidel Stiftung, Fundemos (2003): Descentralizacih y Autonomia Regional, Opiniones y Analisis. La Paz, Bolivia. 67 Harral, C., and A. Faiz (1988): Road Deterioration in Developing Countries. World Bank, Washington, DC. Heggie, Ian (1999): Review o f Selected Roads Funds. World Bank Institute, March 1999, Mimeo. Hiskey, Jonathan T.; Seligson, Mitchell A. (2003): Pitfalls of Power to the People: Decentralization, Local Government Performance, and System Support in Bolivia, in: Studies in ComparativeInternational Development 37(4):64-88. IADB (1994): Bolivia: Road Administration Decentralization Plan (ATNhF-4443-BO). Technical Assistance by the Italian Trust Fund o f the Istituto per la Ricostruzione Industriali, Washington, DC. IADB (1998): Bolivia: Plan of Operations, GIs-Based Decision Support System for Transportation Planning and Infrastructure Management. Technical Cooperation Project TC 9801053-BO, Finances and Basic Infrastructure Division I,Washington, DC. IADB (2002): Bolivia: Land Regularization and Legal Cadastre Program (BO-0221). Executive Summary, Washington, DC. IADB (2003): Politicas y estrategias de administracion vial prefectural, Sistema integrado de administracion de mantenimiento (SIAM). Volumen I,Resumen ejecutivo, Banco Interamericano de Desarrollo, Programa de Fortalecimiento a Entidades Subnacionales Viales y de Transporte-CAEM/Booz, Allen & Hamilton. IADB(2004): Northern Corridor Highway Program, LaPaz-Caranavi Segment. LoanProposal Document BO-0200, Washington, DC. ILDIS (2003): La descentralizacion que se viene, Propuestas para la (re)constitucion del nivel estatal intermedio. FriedrichEbert Stiftung/Plural, La Paz, Bolivia. IMF (2004): Bolivia: ImprovingBudget andDecentralization Processes. LaPaz, Bolivia. IMF (2005): Bolivia: FifthReview Under the Stand-By Arrangement. IMF Country Report No. 05/146, April 2005, Washington, DC. INE (2003): Bolivia: Caracteristicas Educativas y Cambios Entre 1992 y 2001. October. La Paz, Bolivia. Instituto Nacional de Estadistica de Bolivia (2005): Estadisticas de Poblacih. La Paz, Bolivia (en www.ine. gov.bo) Japan's Ministry o f Land, Infrastructure and Transport (2004): Road Administration Management in Regional Level in Japan. Performance Management Office, Road Bureau, MinistryofLand, Infrastructure andTransport, March2004, Tokyo, Japan. Juan Carlos Aguilar Perales (Coord.) (2002): Estudio de Evaluacibn de Capacidad Institucionales y de Gesti6n de 10s Gobiernos Municipales. Ministerio de Desarrollo Municipal, Viceministerio de Participacion Popular y Fortalecimiento, La Paz, Bolivia. Klein, Herbert S. (2003): A Concise History o f Bolivia. Cambridge University Press, New York. 68 Kohl, Benjamin (2003): Democratizing Decentralization in Bolivia, The Law of Popular Participation, in:Journal of PlanningEducation and Research 23: 153-164. Lazar, Sian (2004): Personalist Politics, Clientelism and Citizenship: Local Elections in El Alto, Bolivia, in:Bulletin of Latin American Research 23(2):228-243. Leipzinger, Daniel (2001): Decentralization Models and Issues in Latin America. World Bank, mimeo, Washington, DC. Lora, Enrique. (2005): Estudio de Analisis sobre la capacidad de absorcion de las actuales competencies prefecturales. Ministerio de Participacion Popular y Viceministerio de descentralizacibn Departamental, Diciembre 2005, La Paz, Bolivia. Mantel, A., Albino, M. y Jimenez, J. (2004): Bolivia: Evolucibn de la Descentralizacibn Fiscal y el Control de la Deuda de 10s Gobiernos Municipales. IMF, La Paz, Bolivia. McMeekin, R. 2003: Network o f Schools. Education Policy Analysis Archives 11(16). Ministerio de Educacion. (2004): La Educacion en Bolivia: Indicadores, cifras y resultados. La Paz, Bolivia. Ministerio de Hacienda (2004): Informe de la Cornision de Revisibndel Gasto Publico. La Paz, Bolivia. Ministerio de Hacienda (2005): PresupuestoGeneral de la Nacion (Proyecto), Titulo I:Resumen General. Direccibnde Presupuestoy Contaduria, 2005, La Paz, Bolivia. Ministerio de Hacienda (2006): Estrategia de Descentralizacih Fiscal: Marco Base de Acciones. Enero 2006, La Paz, Bolivia. Ministerio de Hacienda (2005): Marco Estrategico 2005-2007 para el Fortalecimiento de las Finanzas Publicas. La Paz, Bolivia. Ministerio de Servicios y Obras Publicas (2005): Taller de Planificacion Institucional. Viceministerio de Transportes, Febrero de 2005, La Paz, Bolivia. Montero, Alfred P. and David J. Samuels eds. (2005): Decentralization and Democracy in Latin America. Universityo fNotre Dame Press, South Bend, Indiana. O'Neill, Kathleen M. (2005): Decentralization inBolivia: Electoral Incentives and Outcomes in: Alfred P. Montero and David J. Samuels eds. (2005): Decentralization and Democracy in LatinAmerica. Universityo fNotre Dame Press, South Bend, Indiana. Pan American Health Organization (2002): Public Health in the Americas. Scientific and Technical PublicationN.589, Washington D.C. Patrinos, H.A. 2002: Contracting Out inEducation. World Bank (Processed). Prefectura de Santa Cruz (2003): Reglamento especifico de las direcciones provinciales de caminos. Santa Cruz de la Sierra, Bolivia. Prefectura de Santa Cruz (2004): Reglamento de Desconcentracibn Prefectural. Santa Cruz de la Sierra, Bolivia. Prud'homme, Remy, Herve Huntzinger and Sonia Guelton (2001): Descentralizacih en Bolivia. IADB,Agosto de 2001,Washington, DC. 69 Rodriguez, Eduardo (2005): Diagnostico del Sector de Transporte de Bolivia. Ministerio sin Cartera Responsable de Participacion Popular, Mimeo, Mayo de 2005, La Paz, Bolivia. Romer, Thomas and Howard Rosenthal (1978): Political Resource Allocation, Controlled Agendas and the status quo. Public Choice 33:27-43. Ruiz, J. (2005): CIR: Central de Informacion y Riesgo. Ministerio de Hacienda, La Paz, Bolivia. Ruiz, J., Teran, J. (2003): Relaciones Fiscales Intergubernamentales, Lecciones Aprendidas a la Luz de la Evidencia. Ministerio de Hacienda, La Paz, Bolivia. Sabato, Hilda (1998): La Politica en las Calles: Entre el voto y la Movilizacion. Editorial Sudamericana, 1862-1880, Buenos Aires, Argentina. Salcedo, L. (2004): Escenarios de Departamentalizacion de la Deuda Publica en Bolivia. Ministerio de Hacienda, La Paz, Bolivia. Salcedo, L. (2005): SAIDS: Analisis de Deuda Subnacional. Ministerio de Hacienda, La Paz, Bolivia. Saltman, Richard and Figueras, Josep (1998): Analyzing the evidence on European health care reforms. HealthAffairs, MarcWApril 1998. Saric, Drina (2005): Transferencias fiscales y modelacion de escenarios para autonomias. Plural, La Paz, Bolivia. Schultz, T.P. (2004): School subsidies for the poor: evaluating the Mexican Progresa poverty program. Journal o f Development Economics 74(1):199-250. SEPCAM Santa Cruz (2005): Primer foro sobre conservacion y gestion sostenible de caminos vecinales. Servicio Prefectural de Caminos, Santa Cruz, Bolivia. Shefter, Martin (1994): Political Parties and the State. Princeton UniversityPress, Princeton. Shepherd, Geoffrey; Waagstein, Thorbjsrn (coord.) (2005): Bolivia Institutional Reform Project Independent Evaluation.World Bank and Donors, La Paz, Bolivia. Shepsle, Kenneth A. (1979): Institutional Arrangements and Equilibria in Multidimensional Voting Models. American Journal of Political Science23:27-59. Shepsle, Kenneth A. and Barry R. Weingast (1984): Uncovered Sets and Sophisticated Voting Outcomes with Implications for Agenda Institutions. American Journal of Political Science 28:49-74. Shleifer, Andrei and Daniel Treisman (2000): Without a Map: Political Tactics and Economic Reform inRussia. MIT Press, Cambridge. Singer, Matthew M.; Morrison, Kevin M. (2004): The 2002 presidential and parliamentary elections inBolivia, in:Electoral Studies 23:143-182. SIS, Seguro Integral de Salud de Peni (2005): Estadisticas Enero-Diciembre 2004, Oficina de Informacion y Estadistica del Seguro Integral de Salud-Ministerio de Salud, Peni, www.sis.minsa.szov.pe. 70 Skoufias, E. and S. Parker. (2001): Conditional CashTransfers and their Impact on Child Work and School Enrollment: Evidence from the PROGRESA Program in Mexico. Economia 2(1):45-96. Slater, David (2003): On the Spatial Dynamics of Democratic Politics: Analysing the Bolivian Case, in:Development and Change 34(4):607-632. SNC (2003): Memoria de Gestibn 2002-2003. Servicio Nacional de Caminos Institucionalizado, La Paz, Bolivia. SNC (2004a): Memoria de Gestibn 2003-2004. Servicio Nacional de Caminos Institucionalizado, La Paz, Bolivia. SNC (2004b): Estadistica Vial 1999-2003. Gerencia de Planificacibn y Desarrollo Tecnologico, Servicio Nacionalde Caminos Institucionalizado, La Paz, Bolivia. Spahn, Paul Bernd (2003): Strategies for Decentralizing Government While Maintaining Macroeconomic Stability, processed. Steinich, Markus(2006): Comentarios a1debate sobre el futuro del nivel intermedio en Bolivia a la Luz de la experiencia alemana. GTZ, La Paz. Stevens, Andrew (2004): A variety o ftaxes funds Japan's prefectures and municipalities, in: City Majors Government, Local government inJapan. Swope, J.Y and M. Latorre (1998"): Estudio Eficiencia Interna de las Escuela Formales de Fe y Alegria. CIDE, Santiago, Chile. Swope, J.Y and M. Latorre (1998b): Comunidades Educativas, Donde Termina a1 Asfalto, EscuelasFeY Alegria en America Latina. CIDE-PREAL-USAID-IABD, Santiago, Chile. Swope, J.Y and M. Latorre (1998~):Fe y Alegria: una propuesta alternativa de educacibn primaria en America Latina. Teran A. y Ruiz, J. (2003): Relaciones Fiscales Intergubernamentales: Lecciones Aprendidas a la Luz de la Evidencia. Mimeo, La Paz, Bolivia. Thevoz, Laurent (2001): Decentralization in Bolivia: A Model under Construction. Community Studies inPlanning and LandUse Management (CEAT), La Paz. Treisman, Daniel (2000): Decentralization and the Quality o f Government. Unpublished document, Department of Political Science, UCLA. Urenda, Juan Carlos (2003): Autonomias Departamentales. Academia Boliviana de Estudios Constitucionales (ABEC), Santa Cruz de la Sierra, Bolivia. Urquidi, Jorge (2005): Descentralizacibn sectorial, Gestibn y Calidad en debate. PADEM, La PaZ. Urquiola, M.,Jimenez, W., Talavera, M. L.,and Hernany, W. (2000): Los Maestros en Bolivia: Impacto, incentivos y desempeiio. La Paz. Van Cott, Donna Lee (2003): From Exclusionto Inclusibn: Bolivia's 2002 Elections, in:Journal ofLatin American Studies 35:751-775. 71 Vargas, John D. (2004): La Reforma Agraria desde las regiones. CIDES, Plural editores, La Paz, Bolivia. Villaroel, Sergio (2005): Balanzas fiscales departamentales y costeo en la administracibn publica. Plural, LaPaz, Bolivia Villaroel, Sergio (2005): Servicio Civil y Escalafbn hibrido para puestos jerirquicos. Plural, La Paz, Bolivia. World Bank (1992): Staff Appraisal Report, SRMP approved in 1992. World Bank (1996): Colombia: Reformingthe Decentralization Law: Incentives for an Effective Delivery of Services. Mimeo. World Bank (2000): Bolivia Institutional and Governance Review: From Patronage to a Professional State. Washington, D.C. World Bank (2002): Proposed Credit for the Road Rehabilitation and Maintenance Project. Report No. 23625-BO, March 15,2002, Washington, DC. World Bank (2003). Bolivia Education Sector Briefing.Bolivia Country Office, Mimeo. World Bank (2003). Bolivia Public Expenditure Review. Mimeo. World Bank (2003): Implementation Completion Report on a Loan for the Highway Management Project of Venezuela. Report No, 25746, June l7,2003,Washington, DC. World Bank (2003): Human Development Report 2003. Washington, DC. World BaWIADB (2004): Bolivia Public Expenditure Review (PER). Report No. 28519-BO, World Bank and InterAmerican Development Bank, Washington D.C. World Bank (2004): Bolivia: Public Expenditure Management for Fiscal Sustainability and Equitable and Efficient Public Services Report No. 28519-BO. La Paz, Bolivia. World Bank (2004): Implementation CompletionReport for PSAC I. Washington, D.C. World Bank (2004): Implementation CompletionReport for PSAC 11. Washington, D.C. World Bank (2004): Impact Evaluationo f InterculturalBilingual Education. Mimeo. World Bank (2004): Health sector reform in Bolivia, A decentralization case study. Washington D.C. World Bank (2005): Beyond Survival: Protecting Households from the impoverishing effects o f health shocks. Washington D.C., forthcoming. World Bank (2005): Bolivia Policy Note: Manejo Fiscal de la Renta Petrolera: Situacibn Actual y Opciones de Politica. Washington D.C. World Bank (2005): Mexico Determinants o f LearningPolicy Note. Report No. 31842-MX. World Bank (2005): Program Document for a Social Sector Programmatic Credit I1 Bolivia (Report 32066-BO). Washington D.C. 72 Laws and regulations Circular SB/249/99, Superintendencia de Buncos y Entidades Financieras Decreto Supremo 25907, September 22,2001 Decreto Supremo No. 25737, Plan de Readecuacidn Financiera,April 14,2000. Decreto Supremo No. 26868, December 17,2002. Ley 1834 del Mercado de Valores Ley deMunicipalidades Resolucidn Ministerial No. 122,March18, 2003. Ley de Administracidn Presupuestaria No. 2042, 1999. Resolucidn Suprema No.218041 (Normas Basicas del Sistema de CriditoPziblico) 1997 Decreto Supremo No. 26944, February 28,2003. Decreto Supremo No. 28213,2005. 73 IBRD 33374 70°W 65°W BOLIVIA SELECTED CITIES AND TOWNS B R A Z I L DEPARTMENT CAPITALS NATIONAL CAPITAL To RIVERS Porto Velho 10°S Abunã MAIN ROADS RAILROADS Miles PA N D O DEPARTMENT BOUNDARIES Riberalta Cobija Madre de Dios INTERNATIONAL BOUNDARIES Asunción Asunción 60°W This map was produced by the Map Design Unit of The World Bank. The boundaries, colors, denominations and any other information shown on this map do not imply, on the part of The World Bank Puerto Heath Yata Guaporé Group, any judgment on the legal status of any territory, or any Lago Lago endorsement or acceptance of such boundaries. Madidi Huaitunas Rogaguado P E R U Beni Magdalena Lago B E N I Rogagua Santa Ana Lago de Yacuma San Luis Reyes Apere amoré San Apolo M Trinidad Paraguá San Borja S an 15°S L A PA Z Puerto Miguel Blanco Martín 15°S Acosta Lago Nevada Caranavi Ascención Ascención Titicaca Ichoa To Illampu Puno (6,362 m) Guaqui Quiquibey Concepción Concepción LA PAZ AZ Nevada Las Petas Viacha Illimani pay) (6,462 m) Chapare Gran d e San Ignacio C O C H A B A M B A Ichilo Yapacaani (Gu Cochabamba Montero S A N TA C R U Z Desaguadero San José de Chiquitos José Oruro Santa Cruz To Nevada Sajama Arica (6,542 m) Cordiller O R U R O Banados del Lago C o rdiller Aiquile Roboré Roboré Izozog Puerto Poopo Sucre Santa Suárez Suárez Salar de o Ana Coipasa Potosí Potosí To 20°S Iquique O Central Tarabuco C h a c Salar de C Pilcomayo G r a n To Campo Grande Carniri 20°S na Uyuni H Uyuni U Q U I S PA R A G U AY ce Pila y a A C A P O T O S Í To O C H I L E ccidental Villa Montes Mariscal Estigarribia Pacific To Calama G ra n d eeLd Yucuiba Tarija ípez TA R I J A Viljazón iljazón To Tartagal To BOLIVIA Abra Pampa 0 50 100 150 Kilometers To San Ramón de la Neueva Orán 0 50 100 Miles A R G E N T I N A 70°W 65°W 60°W SEPTEMBER 2004