70617 Towards a Fiscal Pact: The Political Economy of Decentralization in Bolivia1 May 2010 1 This report was produced by Jonas Frank, Sr. Public Sector Specialist at the World Bank. The author benefitted from the very useful comments received from peer reviewers for this report: Kent Eaton (University of California, Santa Cruz), Kai Kaiser (Sr. Economist, World Bank), Yasuhiko Matsuda (Sr. Public Sector Specialist, World Bank), and Rémy Prud’homme (Prof. emer. Université XII, Paris). Helpful comments were also received by Nick Manning (Sector Manager, LCSPS) and Chris Parel (LCSPS). Jonas Frank also appreciates the discussions with James Robinson (Harvard University) in the scope of the World Bank’s Institutional and Governance Review for Bolivia, titled “Towards an Inclusive Decentralization� (2006). In Bolivia he have benefited much from the insights and continued guidance by Carlos Toranzo (ILDIS). The author would like to take this opportunity to thank formally to the World Bank Country Office in La Paz for their valuable support provided, in particular through Julio Velasco and Julio Loayza and the Bolivians for their generous hospitality. 1 Abstract The decentralization game in Bolivia has been altered quite significantly with the presence of new bargainers at the departmental level. Two, opposing groups have emerged and which follow intricate strategies to enforce their claims. The highland departments are strongly aligned to the MAS party and the charismatic leadership of the country’s first indigenous leader Evo Morales. The Media Luna departments in the lowlands demand autonomy and seek a greater share of the national pool of hydrocarbon revenue. The paper discusses how a Fiscal Pact can be forged to bring agreement around the most pressing issues. It considers several bargaining packages which could be crafted and analyzes the extent to which decentralization principles need to be sacrificed in order to achieve agreement among competing actors. It is likely that these agreements may eliminate some inefficiencies, but also create others. An important insight is that the Fiscal Pact should be renegotiated at certain pre-defined times, in order to provide flexibility and adjust to the constraints of actors to engage in inter-temporal commitments. Explicit exit options would enhance the possibilities for agreement. 2 Selected Acronyms AND Political party CONALDE Consejo Nacional Democrático HL Hydrocarbons Law ICE Impuesto de Consumos Especiales (Special Consumption Tax) IDH Impuesto Directo de Hidrocarburos (Hydrocarbon Direct Tax) IEDH Impuesto Especial de Hidrocarburos (Special Hydrocarbon Tax) IGR Institutional and Governance Review MAS Movimiento al Socialismo MNR Movimiento Nacional Revolucionario PODEMOS Poder Democrático Social PRF Programa de Readecuación Financiera SEDES Secretaria Departamental de Salud (Departmental Health Agency) SEDUCA Secretaria Departamental de Educación (Departmental Education Agency) SEPCAM Secretaria Departamental de Caminos (Departmental Road Agency) SIGMA Integrated Financial Management System VAT Value-Added Tax 3 Index ABSTRACT............................................................................................................................................................................. 2 I. NEGOTIATING REGIONAL AUTONOMY ........................................................................................................ 5 II. INTERGOVERNMENTAL ARRANGEMENTS IN BOLIVIA ............................................................................. 7 III. FRAMEWORK OF ANALYSIS............................................................................................................................... 9 IV. ACTORS, INTERESTS AND INCENTIVES ........................................................................................................ 10 V. THE INSTITUTIONAL SETTING ........................................................................................................................ 12 VI. HYDROCARBON TRANSFERS: MAIN BONE OF CONTENTION .............................................................. 13 VII. ROLES AND RESPONSIBILITIES: WHO IS IN CHARGE? ............................................................................. 16 VIII. REGIONAL TAXATION ...................................................................................................................................... 20 IX. SUBNATIONAL DEBT AND FISCAL RESPONSIBILITY ................................................................................ 22 X. PROSPECTS FOR ACHIEVING A FISCAL PACT ............................................................................................ 24 BIBLIOGRAPHY ................................................................................................................................................................. 26 Table Index TABLE 1. DECENTRALIZATION RATIOS, 2004-2008.......................................................................................... 8 TABLE 2. REVENUE OF DEPARTMENTS IN PER CAPITA TERMS (IN USD), 2008............................... 16 TABLE 3. RESPONSIBILITIES PER LEVEL OF GOVERNMENT IN THE HEALTH, EDUCATION, AND ROAD SECTORS .................................................................................................................................................. 17 TABLE 4. UNSATISFIED NEEDS PER DEPARTMENT (IN PERCENT), 2001............................................. 19 4 I. NEGOTIATING REGIONAL AUTONOMY From Argentina to Afghanistan, from Malawi to Mexico, many countries around the world are struggling to adjust their intergovernmental models. This has not always led to a single process of decentralizing power to subnational governments but also of establishing centralized authority. Several decades of experience, and countless rounds of continuous adjustments, have led to a catalog of key advice and principles of decentralization that today are widely accepted and recognized. Resources and functions should be transferred together. Governments should honor the budget constraint. Politicians should be held accountable for their actions. Basic institutional capacity needs to be in place. Experts and practitioners debate whether regions or municipalities should be the focus of decentralization, but all agree that responsibilities among levels of government should be clearly delineated. Any country that does all this—so the wide-held view—can reap the potential benefits of decentralization. Yet in many countries there is growing awareness about the existing mismatch between the decentralization principles and how it is implemented in practice. Additional fiscal pressures, poor accountability, and questionable improvements in service delivery, are often mentioned as side effects in countries which have decentralized. This is not primarily a question of measuring the outcomes of decentralization, although this is of increasing concern in a number of countries. Behind it is the increasing awareness about political economy factors and how interests and incentives of stakeholders determine both the path and the outcomes of decentralization. The main emerging lesson is that it is often the early winners of partial decentralization who determine if decentralization can proceed and be completed in a later stage. These important insights have also led to a rethinking about the models of decentralization in countries. There is increasing awareness that in designing decentralization “good fit�—a model which is feasible and workable in light of incentive constraints—should be given prevalence over “best practice.�2 However, there is only little knowledge and systematic discussion, either from a theoretical or positive perspective, of how such “good fit�-models could look like. Too often, efficiency concerns guide the advice for the design of intergovernmental arrangements but fail to recognize fundamental constraints within countries. Likewise, the growing literature on the political aspects of decentralization has not dealt with the critical issues as to how the bargaining process could be structured such that it enhances the prospects of agreement towards these models. These insights—with a few exceptions like Webb et.al. (2006)—have not always translated into concrete and operational advice how conflicts of decentralization can be addressed in the context of competing actors. 2 In recent years, the World Bank has made a concerted effort in understanding these constraints. From civil service to budget management, from tax reform to decentralization, there is evidence that political economy constraints and incentive problems undermine successful implementation of public sector reforms. In a number of Latin American countries, Institutional and Governance Reviews (IGR) have been conducted that try to shed light on these limiting factors. 5 This paper, then, attempts to do both: based on a country case, it analyzes actors and their incentives as well as assesses the possibilities for them to arrive at agreement. The paper discusses what intergovernmental model could emerge—often by sacrificing efficiency and accountability—in order to enhance prospects for consensus. In sum, it endeavors to systematically discuss as to how conflicts over power and resources can be resolved. In Bolivia these issues continue to be of immediate need and relevance. A new social contract among its citizens is being forged, and decentralization figures prominently among the issues which need to be resolved as part of this process. Along with issues of land distribution, indigenous rights, distribution and use of hydrocarbons revenue, the extent of regional autonomy3 are key elements of this new social contract. The country adopted municipal decentralization in 1994 but it was not until 2005 when the regional level of government played a more significant role: elections to the departmental prefect provided it with legitimacy derived from a popular vote. In 2006 a popular referendum on regional autonomy was held, and a Constitutional Assembly initiated discussions on the model of regional “autonomy�; it was more than two years until the new decentralization framework was adopted in a new Constitution in late 2008 eventually.4 Despite the adoption of the new Constitution however, Bolivians remain deeply divided over the extent and nature of decentralization as well as regional autonomy. More recently, there have been proposals to establishing a Fiscal Pact among the national and subnational levels of government. This would provide the opportunity not only to achieve consensus around the most pressing and salient issues, but also further regulate the intergovernmental model under the umbrella rules defined in the Constitution. There is growing consensus that such a Fiscal Pact should encompass the distribution of hydrocarbon revenue and transfers—which is the main bone of contention. Other areas of conflict are roles and responsibilities for each level of government; local taxation; as well as subnational debt and fiscal responsibility. The demands and interests of actors diverge substantially, but all are related to these areas and an agreement would need to be forged. This paper attempts to shed light on the key political factors for crafting a Fiscal Pact in Bolivia in these areas. Before discussing them within more detail, the next section describes the existing intergovernmental arrangements. This is followed by setting up a framework of analysis with the main arguments of the existing political economy literature. The paper then portrays the objectives and incentives of the main actors involved as well as the institutional environment in which these conflicts are taking place. The following sections examine the different decentralization 3 Before proceeding, it is good to emphasize that there is a great deal of ambiguity in the way some words are used in Bolivia. The word autonomía, which one can translate as “autonomy,� is much used, as is the word “decentralization.� Decentralization tends to conjure up connotations of moving the control of government services from the national to the regional level and shifts in the locus of responsibility for raising taxes or undertaking expenditures. Though these things are on the table it is not clear that this is what the debate is really about or what people are thinking of when they demand or reject autonomía. Indeed, demands for autonomy seem to be for fundamental changes in the balance of political authority. 4 These developments are occurring in the environment created by the election of President Evo Morales, who took office in January 2006 as the country’s first indigenous leader. 6 issues which need to be resolved as part of the Fiscal Pact. Although actors do not necessarily bargain separately for each of the issues of transfers, taxes, and expenditure assignments, for reasons of clarity and simplification these are being reviewed independently, yet pointing to possibilities for bargaining packages. The last section discusses the prospects and strategies in order to arrive at a consensus. II. INTERGOVERNMENTAL ARRANGEMENTS IN BOLIVIA Bolivia is a unitary country divided into nine departments (departamentos).5 The transition phase defined by the new Constitution and in effect until mid-2010 foresees that departments are headed by a governor (prefecto) who is elected popularly and subsequently appointed by the President. In their hybrid nature, they both represent citizens of the department as well as the head of state. There are also departmental councils which are elected popularly (since 2008); their responsibilities include approving the departmental budget. However, departmental budgets are not clearly delineated.6 Most expenditures of departments are in secondary roads, electricity, irrigation, and other physical works which they often execute in the rural areas. Departments do not administer own taxes but charge a series of fees. Their revenue depends mostly on transfers.7 In addition, there are about 330 municipalities with elected mayors and municipal councils. Municipalities are mostly responsible for urban services. There was the expectation that they gradually adopt powers in health and education, but in most cases this has been confined to the management of school and hospital infrastructure. The two most important tax revenues for municipalities is the property tax as well an automobile tax. The Constitution of 2008 also opens up the possibility for a creation of two new autonomous entities: “indigenous autonomies�, which in essence are municipalities based on ethnic considerations and traditional rights; “Regions�, in turn, can be established on the demand of neighboring municipalities and they would be funded out of resources assigned to departments. The functioning of these two new entities has yet to be clarified. There is also a provincial level which operates between the departments and municipalities. Often neglected, it serves the purpose of political representation of the prefect; however provinces do not execute expenditures themselves. The fiscal indicators of subnational governments have changed after the approval of the hydrocarbon law in 2006 (Table 1). Today, subnational governments collectively execute a higher 5 The terms “department� and “departmental government� are used interchangeably, except as otherwise noted. The term “subnational governments� refers to both departments and municipalities. 6 Department governments participate in the elaboration and partly also in the execution of a number of ministerial budgets. Each sector has its own logic and pattern of involvement of departments; therefore, a substantial part of departmental budgets is, in truth, in the national budget. 7 Transfers for departments include the hydrocarbon sales tax IEDH, the Departmental Compensation Fund (compensating for differences in revenue of the IEDH) as well as royalties. In 2007 the IEDH and Compensation Fund together constituted 14.3 percent of transfer revenue and the royalties were 48.6 percent of transfer revenue. The IDH was 37.1 percent. 7 share of public expenditures (about a third of total public expenditures). However they also have become more transfer dependent. This is particularly the case for municipalities. In comparative terms, the role of subnational governments in Bolivia is not as prominent as in federations of Brazil or Argentina; but the country has now a much deeper degree of decentralization than many other countries in Latin America, particularly in Central America. Table 1. Decentralization Ratios, 2004-2008 2004 2005 2006 2007 2008 Departments (prefecturas) Expenditure decentralization ratioa 7.3% 10.0% 16.4% 15.7% 11.0% Tax decentralization ratiob 0.3% 0.2% 0.2% 0.1% 0.1% Transfers dependency ratioc 96.0% 96.3% 93.7% 92.9% 92.1% Excluding wages and salaries for health and education 90.0% 93.0% 89.4% 88.4% 85.4% Municipalities Expenditure decentralization ratio 17.1% 14.9% 19.6% 20.6% 20.3% Tax decentralization ratio 5.5% 3.9% 3.3% 3.4% 3.1% Transfers dependency ratio 65.9% 68.7% 77.3% 74.1% 75.5% Subnational governments Expenditure decentralization ratio 24.4% 24.9% 36.1% 36.3% 31.3% Tax decentralization ratio 5.8% 4.1% 3.5% 3.5% 3.3% Transfers dependency ratio 82.6% 85.5% 86.8% 84.7% 83.8% Excluding wages and salaries for health and education 74.0% 79.7% 82.7% 80.4% 79.0% Notes: a = expenditures of prefecturas/total public sector expenditures; b = taxes of prefecturas/ total taxes of public sector; c = transfers/total resources. Source: IMF (2010); Ministry of Economy and Finance. There is not a uniform view on the outcomes of decentralization in Bolivia. However, there are weaknesses of the intergovernmental model which come to bear, among others, in the lack of tax revenue at the departmental level which leads to high transfer dependence. Municipalities, in turn, are somewhat less dependent on transfers than departments but—as mentioned above—this ratio has also increased as a result of the hydrocarbon transfers. Such transfer dependence can only be judged in light of the specific expenditure responsibilities. If departments (or municipalities) would be responsible for education and health services, their revenue should rely more on transfers so there are incentives to keep delivering social services with positive externalities. However, extreme degrees of transfer dependence, particularly for earmarked grants, are also undesirable on grounds of allocative efficiency. As well, in theory local taxation could act as an additional incentive for fiscal responsibility when politicians “feel the pain of the marginal tax boliviano they spend�. Another problematic area of the intergovernmental model is the assignment of expenditures. The past decentralization process has partially led to a fragmentation in the production of services, 8 for instance in health and education. The informality of the public administration on all levels of government is also an important factor for resources management and public service delivery. III. FRAMEWORK OF ANALYSIS Given the interests of actors in Bolivia and the characteristics of the existing intergovernmental system, it will be useful to draw on specific insights of literature on the political economy of decentralization. Party politics is a central element to understanding the maneuvers of actors. O’Neill (1998) has demonstrated convincingly the importance of inter-temporal considerations of political parties; these are critical factors that can determine when and if decentralization can get off the ground. These were clearly a factor in the first phase of decentralization which led to municipal decentralization in 1994 promoted by the MNR, and these are still relevant for the MAS as it is confronted with demands for departmental autonomy. Haggard et.al. (1999), in turn, have discussed incentives shaped through party politics for Presidents and legislators in different political systems in Latin America. Brancati (2009) discusses the relevance of regional party interests in the decentralization process. Although Hellman (1998) developed his analysis of partial reform in a different context, they are applicable to decentralization. He demonstrated that, ironically, it is often the early winners of partial reform which can constitute obstacles for reforms to be completed in a later stage. This is relevant for Bolivia insofar as the roles and responsibilities are largely disjointed—there are winners of these early yet inefficient reforms which keep the system in a situation of partial decentralization. As demonstrated by Devarajan et.al. (2007), these situations can weaken accountability, and in turn lead to self-reinforcing governance traps. In a case study on Ecuador (Frank 2007), it was demonstrated how partial decentralization plays out in several rounds of bargaining. Actors form different equilibria and need to be offered new incentives to give up their benefits. This explains why inefficient decentralization arrangements can be eliminated but also why new inefficiencies are often added. These lessons are relevant for Bolivia as they underscore that sub-optimal intergovernmental arrangements are most likely to persist in the presence of actors who have a stake in inefficiency. As fiscal challenges came to bear during the take-off period of decentralization in Latin America, a group of literature focused on the issue of fiscal responsibility. Rodden et.al. (2003) and Webb (2004) have demonstrated that achieving hard budget constraints requires a balance of ex-ante controls and ex-post consequences which critically depend on the ability to enforce sanctions. Treisman (1999), in turn, has demonstrated that non-cooperative behavior can lead to the worst outcomes in terms of welfare and be entirely rational for the actors involved. Equally important is literature on rentier states and natural resource rich countries. Among others, Ahmad et.al. (2008), Bahl et.al. (2004), and Morgandi (2008), have discussed the intricacies of sharing these revenues among multiple layers of government and their effects on equity and efficiency. For the purposes of this paper, these lessons are relevant as the volatility of the resource limits the ability of actors to commit to inter-temporal agreements—these would be necessary to credibly engage in a Fiscal Pact. 9 IV. ACTORS, INTERESTS AND INCENTIVES8 The election of the nine departmental government leaders brought new bargainers to the table seeking greater shares of the national resource pool. As mentioned, the Prefects remain delegated by the President of the Republic during the transition period defined by the new Constitution—however de facto it is the popular vote which already provides them with a great deal of legitimacy. Among the nine prefects and departments are, however, significant differences as to the objectives and driving motivations to participate in the decentralization game, and which define the political context in which bargaining of a new Fiscal Pact is taking place. Regional autonomy has been pushed by the so-called Media Luna which comprises the four lowland departments of Santa Cruz, Tarija, Beni, and Pando. Some of them, like Tarija, are quite rich in hydrocarbon resources and hence have the desire to retain more of these revenues in the department (though this factor is quite marginal in Santa Cruz compared to Tarija). As measured by GDP per capita, the Santa Cruz department is the wealthiest jurisdiction and leading power among the lowland departments. Indeed, regional identities are so strong in Santa Cruz that the fact of being a Cruceño overrides most other group allegiances, whether it is to political parties, labor unions or business groups. One key demand of this department is to withhold 66.6 percent of “locally produced� tax and customs revenue with 33.3 percent transferred to the central government. These proposals are based on the questionable use of information comparing amounts of central or total government taxes collected with a department’s expenditures. Some of the assertions derived from these comparisons are erroneous. Among other factors, saying that a region should retain a certain share of total taxes paid in confuses local and national taxes. It is inappropriate to equate central government taxes collected in a territory with the central government taxes produced by activities of households and enterprises in that territory. Irrespective of these politically driven yet methodologically questionable proposals, the main reasons behind them seem to be: the view that La Paz does not care enough about the needs of Santa Cruz and central politics is too distant, inefficient, and corrupt; the viability of the economy in Santa Cruz and the stability of property rights are threatened by the central government; and the impression that decentralization will allow Santa Cruz to benefit from having to redistribute less income to other, poorer departments. The other departments in the lowlands, Beni and Pando, are less well represented in national politics and seem likely to be very marginal players in the decentralization process. They will likely accept a Santa Cruz proposal for autonomía if they can improve their political and financial position, but they do not have sufficient levers to stop Santa Cruz from getting what it wants either. The partial exception is Tarija. Because of its great hydrocarbon wealth, Tarija is the other department that might clearly benefit from a redistribution of resources which leads to departments keeping a higher share of hydrocarbon wealth. 8Some of the insights in this section are based on earlier political analysis conducted in the 2006 Institutional and Governance Review (IGR) on Bolivia, titled “Towards an Inclusive Decentralization�, World Bank. 10 The four highland departments of La Paz, Oruro, Potosí, and Chuquisaca are opposing regional autonomy as proposed by the Media Luna departments. These jurisdictions are an electoral stronghold of the President’s party MAS.9 This party has demonstrated that it can win presidential elections and hence endeavor to maximize the power of a central state. This was confirmed in the presidential and congressional elections end-2009, which even allowed the party to seize a greater electoral share. In turn, MAS does not believe it can gain electorally by having departmental elections10 and in consequence it does not believe that it will control decentralized resources. The possibility of creation of regions as defined in the 2008 Constitution can be regarded as a maneuver of MAS to undermine and hollow out the power of departments in a “divide-and-rule� strategy. As mentioned, regions would be funded from revenue assigned to departments. Likewise, giving departments autonomy may stop MAS getting things that it wants, for example hydrocarbon revenues from Tarija and land reform in Santa Cruz. At the moment MAS has far more power to get what it wants than the other parties. Aside from electoral support at the national level, it has developed the ability to paralyze La Paz and other major cities with collective action. Moreover, MAS anticipates that the 2008 Constitution will provide a way to get many things that it aspires to. The highland departments do not seem to be a united front that would allow them to resist any attempt from the Media Luna to co-opt them or else offer credible side payments. Oruro or Potosí, appear to have much less well defined positions, though it is possible that they will lose if a large redistribution of tax revenues is the outcome of decentralization. The position of La Paz seems difficult to characterize because of the conflict between El Alto and the city. Actors in La Paz may not form such a united front as in Santa Cruz mostly because of the role of the central government bureaucracy. This group could be one force within the department to oppose decentralization or autonomía if it affects rights and privileges of public employees. The political position of Cochabamba could be decisive in the process of agreeing on a new Fiscal Pact. Geographically located between the four departments of the Media Luna, and the four highland departments, its position has shifted from initially being opposed, to later subscribing to ideas of regional autonomy. However, its influence does not appear strong enough to adopt the role of a broker or mediator among conflicting departments—but it could still influence decisions as swing voter. From the above it is clear that the nine new bargainers will not always act and bargain in ways that are independent from Morales as leader of MAS; and that they therefore can be considered separate “veto players� or “bargainers� in all aspects of decentralization. They have acted in a united fashion when negotiating hydrocarbons revenue in 2005 as this was promising benefits to all nine departments. There is also an increasingly united position with regards to adopting autonomy statutes. However, over most other issues the divide between the “Media Luna� in the lowlands vis- à-vis the highlands will still determine most bargains that need to be struck. 9 The MAS and other indigenous social movements are committed to socialistic policy agendas and their rhetoric is based on the notion that they represent the exploited members of Bolivian society. 10 This logic was amply confirmed by the December 2005 elections where MAS won the race for prefecto in only three departments, Oruro, Potosí and Chuquisaca. 11 This divide is not as deep and visible at the municipal level. Although the electoral success of MAS is uneven across the different municipalities, it generally sees in municipalities a possible political ally. Further decentralization to this level is a means of undermining the department’s claims for autonomy, very much in line with the political tactics discussed by Dickovick (2006). Municipalities perceive departmental decentralization as a threat to their power. They existed before the Law of Popular Participation (1994). While they were not autonomous and hence it is difficult to attribute agency to municipalities from this perspective, they have secured a “first-mover advantage� over regional governments in the initial phase of decentralization starting in 1994. They were strengthened through the Law of Popular Participation, and a now perceive an increased role of departmental governments as a threat to their resources and political power. The one element which acts to “lock in� the interests of municipalities is the automatic revenue-sharing transfer (co- participación). Any change in national or departmental taxation would directly affect their revenue flows, while the formula-based and automatic nature of the transfer make it much harder—though not impossible—to be dismantled. V. THE INSTITUTIONAL SETTING These conflicts are played out within a unitary state where there is, by definition, no formal decision-making power of subnational units at the national level.11 Although not always on a continuous basis, several arenas of bargaining and decision-making have emerged and are used in parallel. The Constitutional Assembly played a leading role to formalize agreements on regional and indigenous autonomies but could not cope with the deadlock which was developing to forge a new social contract. As Van Cott (2001) has noted, actors which propose changes and are represented in constitutional assemblies can exercise credible threat to not to sign the new constitution. This can lead to a stalling of these efforts and a continuation of crises of legitimacy and governability. It was precisely in Congress and not in the Constitutional Assembly that important agreements were forged for the approval of the new Carta Magna. This in turn led many pro- autonomy forces in Santa Cruz to feel betrayed by the PODEMOS senators who agreed to these concessions in Congress. Irrespective, departmental actors continue lobbying through legislators at the national level—both through the Senate as well as House of Representatives—as a way to obtain material benefits for their jurisdictions. Another arena which played a role is CONALDE (Consejo Nacional Democrático) since it served—temporarily at least—to integrate opposition prefects in the difficult negotiations between the Media Luna and the national government. The role of these arenas needs to be assessed in light of the new institutional feature of Bolivian politics: much of it has become la Política en las calles (Sábato, 1998). The rise to power of MAS has to a large extent moved political decision-making into the street. 11In contrast to other decentralizing unitary countries in Latin America, Bolivia has not established Decentralization Commissions or the like, which can be considered substitutes of national decisionmaking mechanisms and which are a natural feature of federal states. 12 A significant factor to explain this trend is the decay of traditional parties (MNR, MIR, ADN). Many groups feel that their interests and ideas have not been reflected in their policies. The political gridlock which has overtaken Bolivia over the past few years and a rising tide of disillusionment with these parties who have governed the country since the return to democracy in the early 1980s, appears to have severely weakened their political support. Currently, none of them clearly sees in the shift of decision-making to local levels a way to increase their electoral support. These institutional features and incentives clearly add to the complexity of how agreement for a Fiscal Pact can be achieved. Based on the considerations above, the following are the main political conditions for a Fiscal Pact.  Each of the actors most likely will bargain for a different Fiscal Pact. This is likely to increase the asymmetric elements in Bolivia’s intergovernmental model.  Bargaining packages could increase the prospects for agreement, but it is not clear that they would be efficiency-enhancing or include all elements which—based on international experience—would usefully strengthen the intergovernmental model, particularly with regards to tax authority, expenditure responsibilities, and fiscal responsibility.  There is no strong and powerful broker or mediator, which means that repetitive games among actors are likely to be necessary in order to arrive at consensus.  Actors need to manage a high level of uncertainty. This likely leads to short-term agreements with exit options. As well, none of the actors has the ability to make credible inter-temporal commitment which is crucial for “contracting� or formalization of agreements. The following sections will elaborate in more detail about the prospects to arrive at consensus on the key features of the intergovernmental model. VI. HYDROCARBON TRANSFERS: MAIN BONE OF CONTENTION 1. The Challenge The agreements over the Hydrocarbon Law adopted in 2005 do represent the new social contract which is being forged in Bolivia.12 As a result of the claims for nationalization of resources, the Law introduces a new production tax for natural gas which has resulted in one of the most significant sources of revenue for the government. While the recipients of resources are different social groups—among them indigenous people—the Law was also pushed by natural resource rich regions. The Media Luna departments would be able to bear the political cost of claiming these resources which, in turn, would allow the highland departments to free-ride on these demands. The Law provides additional resources to the regions and municipalities as earmarked funds without The initial decision on the distribution of the hydrocarbon law was based on Executive Decree no. 28421 from 12 October 2005. 13 transferring new responsibilities:13 in 2008, subnational governments obtained nearly half of these resources although changes were introduced later. These transfers to subnational levels are widely regarded as a symbolic victory of regional interests. In contrast to other situations in Bolivia’s decentralization process, this bargain was about dividing a new resource pie and initially did not imply a reduction in resources from any actor. All gained in the approval of the Law, including the central government. However, complex bargains and renegotiations developed in subsequent years. Morales used his discretion to reduce the share of both the department’s and municipalities’ hydrocarbons revenue by 30 percent in order to finance the national government’s social program Renta Dignidad. Strapped with financing needs, this maneuver curtailed the power of subnational governments. Other modifications followed. These have reduced the power and influence of the departments in particular. The shares being distributed to municipalities and departments have changed over the course of implementation: since October 2007, municipalities receive proportionately more resources than departments,14 even taking into account the deductions for Renta Dignidad. Quite naturally, departments feel shortchanged and these adjustments have further fueled distrust. While such political maneuvers are rational from the perspective of actors and will always be an ingredient in decentralization, it is the nature of the resource itself which shapes the incentive of the different players involved. Hydrocarbon revenue is volatile in the future because of the uncertainties of future hydrocarbon output; this is also related to the changing international macroeconomic environment and swings in commodity prices. As a result of volatility and discretionary adjustments, in 2008 for instance, resources transferred to departments from this Law were US$78 million (down from the initially projected US$263 million for this year) and for municipalities $267 million (up from the initially projected US$160 million). 15 It is clear that faced with these changes actors will unilaterally abandon previous agreements and bargain with a short- term horizon. With the growing importance of natural resources in the intergovernmental system, it is more unlikely that actors will be able to make credible inter-temporal commitments. 2. Policy Options and Strategy It is clear that the Fiscal Pact will need to include agreements on funding of subnational governments. While in the past bargaining was based on an ever growing resource pie, it is likely that in the future retrenchment budgeting due to declining hydrocarbon revenue will be a politically sensitive element. This is all the more difficult as additional resources have been spent over a few years already at the local level and interests are created which likely favor the status quo. 13 The initial decision was based on Executive Decree no. 28421 from October 2005. 14 With Executive Decree no. 29322 from October 2007 the shares allocated to each level of government have been amended. The initial share for departments was 56.9 percent and it has been scaled down to 24.4 percent. Municipalities initially received 34.5 percent and now receive 67 percent. 15 Comparing the initial allocation for 2008 based on Decree no 28421 (1.848 million Bolivianos) with the 2008 allocation (554 million Bolivianos) underlying the Decree no. 29322 as well as taking into account the reduction for Renta Dignidad. Data: FIAM (2008). 14 Given this context, there are several possible avenues for renegotiation. A new formula could be phased in gradually and at the margin. Each year a certain percentage could be distributed to a new formula, irrespective of the amounts which would be distributed each year (be they higher or lower). Ideally, a new formula would be equity-enhancing and compensate for expenditure needs. The formula would need to be simple and transparent so that all stakeholders understand how resources are distributed. A per-capita formula has been working already for municipalities and the same approach would certainly be valid for the transfers from hydrocarbons.16 In order to make such new formula attractive to departments, a bargaining package could be crafted by which the new formula is introduced in return for departments obtaining more autonomy in resource use. In 2006, departments were able to only spend about 8 percent of all funds according to their own preferences and discretion.17 Gradually reducing expenditure ear-marking, and thereby providing more autonomy over expenditures might be a complementary measure which could increase the prospects for a re-negotiation of transfers. On the other hand it is also clear that on political grounds it is warranted—and on economic grounds18 probably justified—to accept the claim of hydrocarbon-“producing� departments to obtain a larger share of the revenue than “non-producing� departments. Necessarily this share would need to remain small so that equity is not being undermined. It is not clear how to address the very high degree of inequity resulting from the distribution of the hydrocarbon revenue. These transfers are highly unequal as measured in per capita terms (Table 2). This can be explained by the fact that the distribution of resources is based on dividing the resource pie by equal shares, plus an additional compensation for hydrocarbon-“producing� departments. Consequently it is no surprise that in 2008 transfers from the hydrocarbon law allocate, for instance, about US$23 per capita to Santa Cruz and La Paz (with one of the largest communities of the poor) while Pando (among the least poor) receives US$466. Although there are statements from municipalities19 that the resulting inequities need to be addressed in order to improve public services and to put all territories onto an equal footing with regards to their ability to execute public expenditures, reducing them will not be straightforward given the political claims over resources. 16 A poverty formula would also be valid, however it raises the need to define how much more a poor region receives, compared to a non-poor or less poor region. 17 No more recent data are available. However, the share of conditioned resources might even be higher in 2009, given the increased proportion of hydrocarbon resource in total resources of departments and which are earmarked. However, in some departments there are unspent resources which can be spent freely but under the condition they are investment expenditures. 18 Higher transfers for hydrocarbon-“producing� departments would compensate for negative externalities (for instance, environmental damage) associated with exploitation of a non-renewable natural resource. 19 This is, for instance, recognized by the Municipal Association FAM although municipalities have been clearly benefitting from the recent redistribution. Refer to: FAM 2008. 15 Table 2. Revenue of Departments in per capita terms (in USD), 2008 Departments: Poverty IDH Royalties Revenue Total from poor to rich rate sharing Potosí 79.7% 45.0 64.9 56.6 166.5 Beni 76.0% 81.6 38.8 55.8 176.2 Pando 72.4% 466.0 113.4 78.3 657.6 Chuquisaca 70.1% 61.9 27.7 53.6 143.2 Oruro 67.8% 79.1 30.2 57.9 167.1 La Paz 66.2% 23.4 2.5 50.8 76.7 Cochabamba 55.0% 27.8 19.2 49.1 96.1 Tarija 50.8% 95.2 366.4 51.6 513.3 Santa Cruz 38.0% 23.1 17.6 46.2 86.8 Source: World Bank, with information from Ministry of Economy. Poverty rates: INE (2001). Note: Includes departments and municipalities. Revenue sharing includes the IEHD transfer. IDH = Hydrocarbon Direct Tax. Another, complementary reform is to redefine the distributable pool of revenue. By defining a national pool of revenue—including all tax and hydrocarbon revenue—the conflict over hydrocarbon revenue could be transformed since this would take into account all available resources. The department’s acceptance to be financed out of a national pool are likely to be higher if they realize that this would provide a smoother revenue flow, instead of them depending only on a single and highly volatile source. It is questionable however if actors at the departmental level have a sufficiently long time horizon to consider such an arrangement as beneficial to them; but it would increase possibilities that actors can develop and commit to inter-temporal agreements. Additional stabilizing rules—for instance, basing the transfers on moving averages—could complement such a bargaining framework. This would not avoid that actors bargain for future revenue streams, and still perceive it as a zero-sum game, but it would give both the central government as well as departments more time to adjust should a shortfall in revenue occur. VII. ROLES AND RESPONSIBILITIES: WHO IS IN CHARGE? 1. The Challenge The early statutes of Autonomy proposed by the Santa Cruz department foresaw a complete decentralization of all responsibilities of the public sector. Whether practical or operationally feasible or not, this demonstrates that a Fiscal Pact would also need to address the issue of subnational responsibilities and align it with the funding mechanisms. In contrast to conflicts over fiscal resources and hydrocarbon revenue, “decentralizing� the provision of education, health and infrastructure is not an issue which should create conflicts among departments or among municipalities; rather, it affects the vertical division of authority and levels of government. 16 It is too early to judge if such proposals as the one put forward by Santa Cruz would clarify the assignment of responsibilities. Expenditure responsibilities are largely disjointed and fragmented across levels of government and in the different sectors, for instance in education, health, and roads (Table 3). In education, for instance, teachers are paid by the central government; they are managed by the departments; their time is controlled by the municipalities; and the districts are responsible for their hirings and firings, in coordination with the central level. With such fragmented decision- making, accountability is diluted, and no single government or service provider is in control. Following Devarajan et.al. (2007), such low levels of accountability contribute to poor governance. Table 3. Responsibilities per Level of Government in the Health, Education, and Road Sectors Primary and Secondary Health Primary and Secondary Roads Education Central Financing Financing Servicio Nacional de Caminos: Government National road network. Planning Planning Prefectures Deconcentrated SEDES: Deconcentrated SEDUCA: SEPCAM: autonomous payment to health personnel payment to teachers institution responsible for (payroll). (payroll). departmental road network. Management of health Management of teaching personnel. personnel. Control of norms. Control of norms. Infrastructure for secondary Delivery of certificates. and tertiary care (if not transferred to municipalities). Municipalities Infrastructure (Hospitals) Infrastructure Rural roads (Schools) Source: World Bank. Additional factors further complicate the situation. Departmental sector directors fulfill two— often conflicting—roles. They report to departmental governors but also depend on central government sector ministries. Moreover, departmental budgets are not clearly delineated.20 Due to the fragmented institutional arrangements, some agencies—such as the Departmental Road Agency SEPCAM, for instance—can be termed “floating institutions� since they neither report to the central government, nor to prefectures, nor to road users.21 20 Department governments participate in the elaboration and partly also in the execution of a number of ministerial budgets. Each sector has its own logic and pattern of involvement of departments. A substantial part of departmental budgets is, in truth, in the national budget. 21 The road sector budget, for instance, is 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. 17 Aware of this complex situation and inefficient assignments of responsibilities, in 2006 the government issued a regulation to the Hydrocarbon Law. The aim was to earmark the use of resources for sector-specific investments, and thereby also re-define responsibilities of each level of government. To date, however, enforcement appears to be weak. The new Constitution further complicates the assignment of responsibilities. Recognizing the need to clarify responsibilities, the Magna Carta defines four different types of responsibilities: concurrent, shared, delegated, and exclusive functions, which are not well separated. In addition, the possibility to establish “regions� as well as “indigenous autonomies� as new jurisdictions and levels of government further makes any assignment of responsibilities more complex. 3. Policy Options and Strategy The assignment of responsibilities is without doubt an important element of the Fiscal Pact. But addressing unclear responsibilities is not a legal problem. More fundamentally, to a great extent this is an issue of altering the control over inputs for the production of services. Latin American countries, almost throughout, achieved very little progress with regards to a clear delineation of responsibilities. Given that control over production factors can often not be altered, , there is a trend towards “coordination� rather than “clear divisions� of responsibilities.22 Arguably, strengthening any sector strategy is running counter to the interests and incentives of prefects and mayors, which favor separate approaches for regional and municipal levels, respectively. This was clearly visible in the claims for autonomy. The prospects for achieving a sector strategy in the Bolivian context do not seem all too dim however: to a large extent the prefectures today do still function in a sectoral manner. In budgetary matters, for instance, investment plans and resources are agreed upon in the sectors of education and roads, and not in an integrated fashion at the departmental level. Some management approaches could create the incentives for a sector approach. For instance, national and subnational co-financing—which has been working in Bolivia in the past—and which is strictly tied to sector expenditures in roads, education, and health, and so forth, are a powerful tool to achieve this. About two-thirds of investments are executed by subnational governments, so better organizing these expenditures would be one important entry point. Another option, which could be implemented in a complementary fashion, is to start introducing results-based management approaches based on considerations of social inclusion. This figures prominently in the public agenda particularly of MAS. Many public services leave much to be desired and there are inequities in access and coverage of services (Table 4). A results-based approach could be crafted in the following manner: in return for letting departments adopt their statutes of autonomy, the central government would make departments publicly responsible for certain services. This could possibly trigger a process of more accountability—vertically towards the central government and horizontally towards the users. As part of this approach, service quality indicators could be defined early on and then tracked over time. In willing or self-selected departments, simple and clear indicators could be defined such that they reflect commitment to at least maintaining the quality of services and improving it at a later stage. The likely outcome of this process is asymmetric 22One common though not the only way is to establish coordination councils at the intermediate (or other) level of government with representation of several levels of government. 18 decentralization by which different subnational governments would formally obtain different levels of responsibilities. Table 4. Unsatisfied Needs per Department (in percent), 2001 Department Inadequate Insufficient Inadequate Inadequate Insufficient Insufficient Housing Housing Water and Supply of Education Health Material Space 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 Potosí 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 Análisis de Políticas Sociales y Económicas. www.ine.gov.bo This is a controversial approach for Bolivia, since many pre-requirements for effective application are not yet met. These are, among others, a fair degree of autonomy and control over resource use; transparent human resource management; and clear lines of accountability. Yet even if these technical requirements are not met, results-based management approaches have some clear political advantages and could be crafted as described below. Such strategy would not rule out that Bolivia could move away from mere “coordination� and arrive at a “clear division� of responsibilities in the long term. The political pressure generated by publicly making responsible the departments for a service would allow shifting the cost of producing services gradually to the individual levels of government. In education, for instance, the cost of teacher salaries could gradually assumed by departments who start paying them based on national rules. However, this requires departments who recognize and derive a benefit in controlling teachers as one of the most powerful segments of public employees. On the other hand, such arrangements would also require a credible budget constraint so that the increasing cost of providing a service is not shifted back to the center.23 One possibility is to combine this, early on, with the ability to establish taxes, discussed further below. 23 This has happened in Colombia during the 1980s and 1990s. 19 VIII. REGIONAL TAXATION 1. The Challenge The Santa Cruz autonomy statute proposed either the withholding or administration of national taxes. Though methodologically questionable since these proposals do not adequately take into account the fiscal capacity of the department,24 and although Santa Cruz overestimates its capacity to obtain more resources, these proposals touch upon one of the critical areas for any decentralized system: own revenue as a base for financing expenditure responsibilities and service delivery. As mentioned above, departments rely nearly completely on transfers. As Eaton (2007) has noted correctly, more control of revenue by richer regions would constrain the redistribution capacity of the central government and this can lead to opposition from poorer regions. Santa Cruz is aware that it has to redistribute to poorer regions. The proposal of autonomy foresees that of the 66.6 percent of locally withheld revenue a 10 percent share (of the total tax and customs revenue) would be destined for redistribution purposes among departments. Jurisdictions like Pando or Beni have few prospects for establishing buoyant departmental taxes, and it is entirely possible that they would lose in a scenario of large tax revenue decentralization. This explains why demands for departmental taxation are not shared by all departments within the Media Luna. The same holds for highland departments like Oruro or Potosí which appear to have much less well defined positions regarding taxation. Interest in departmental taxation cannot be taken for granted. It is not clear that newly elected prefects are willing to bear the political cost of taxation—the existing proposals already demonstrate that this willingness will not be uniformly high. In the current arrangement, it is the central government which assumes this political cost. There are also several technical complexities in finding suitable regional taxes for Bolivian departments. This is a constraint inherent to regional levels of government, but it is particularly strong in the case of Bolivia. Although the fiscal situation of the central government was exceptionally positive in recent years,25 the macro environment is likely to remain vulnerable to external shocks due to the dependence from hydrocarbons. This might make it even harder to find suitable sources of revenue which can be ceded to departments—and which the center will be willing to cede. Also, some potentially buoyant taxes, like the personal income tax, do not exist in Bolivia. Municipalities will also be affected as any increase or decrease in the revenue yield from national taxes would directly result in corresponding changes in the amount of revenue-sharing (co- participación) they receive. As mentioned earlier, the revenue-sharing arrangement has locked in 24 As noted earlier, claiming that a region should retain a certain share of total taxes paid confuses local and national taxes; it is methodologically incorrect to equate central government taxes collected in a territory with the central government taxes produced by activities of households and enterprises in that territory. 25 Bolivia has enjoyed the benefits of exceptional external conditions that have persisted until the end of 2008. Fiscal surpluses, mainly resulting from high oil prices and changes in the hydrocarbons taxation regime, reached 4.6 percent of GDP in 2006, 1.7 percent in 2007, and have reached 3.2 percent in 2008. Since 2000, GDP growth was above 3.5 percent; it has reached 4.6 percent in 2007 and accelerated to 6.15 percent in 2008. 20 powers they received as part of the early decentralization reforms. Part of this dilemma is that departments are relative late-comers to decentralization. Municipalities, in turn, were granted new tax bases which would have been suitable for departmental administration, for instance the automobile tax and the property tax. Likewise, there are no functioning tax administrations in departments and there is low transparency as well as poor exchange of fiscal information among levels of government. It is not clear how this would influence individual departments in their willingness to bargain for more taxation. However, poor information exchange can lead to distrust among the different actors and lower prospects to arrive at consensus. 2. Policy Options and Strategy Given the incentives of the actors in Bolivia—particularly due to departments favoring autonomy and the Santa Cruz proposal—the most likely outcome of debating subnational taxation is tax-sharing.26 Shared responsibilities exist on the expenditure side, and it is not unlikely that this will also be reflected on the revenue side. Shared tax bases are applied in France and Spain, but in the Bolivian context some of the incentive problems are likely to be even more prominent: both municipalities and departments would either overgraze the common resource pool and raise tax rates above an efficient level; or they would not charge any at all, in the hope the other level of government will bear the political cost of taxation. Both dilemmas could occur simultaneously in Bolivia. If actors can agree on individual tax bases for each level, there are two starting points for regional taxation.27  There is the option to cede the domestic share of the Special Consumption Tax (ICE) to departments. It is currently administered by the central government and would imply a revenue loss both to the central government and municipalities, whose consent would be needed for this option to be politically feasible.28  Another option is to create a new tax consisting of the sales tax on gasoline and diesel fuel to be administered by departments. Coupled with some important administrative provisions that would ensure that these taxes mirror accurately local fiscal capacity, these sources could be used as a starting point for departmental taxation.29 26 Applying this approach in Bolivia would imply that departments can use and exploit tax bases jointly with the national government or municipalities. For instance, the municipal property tax could be structured such that municipalities collect taxes on property ownership but departments collect taxes on change and sales of property. 27 These options have been discussed with more detail in the 2006 Institutional and Governance Review on decentralization. 28 Contrary to federal countries, the departments in Bolivia are unlikely to introduce revenue-sharing systems which establish transfers to municipalities. Such a system works in Argentina, for instance. 29 Given the experience of other countries (such as Colombia), these two taxes are also suitable departmental taxes since they cannot be easily exported out of the taxing jurisdiction (contrary or foreign trade taxes, for 21 Given that most departments with proposals for autonomy statutes—among them Santa Cruz, and more recently La Paz—have signaled their willingness to adopt new expenditures there is the possibility to craft a bargaining package: departments are entitled to new responsibilities if they are also willing to administer new taxes. Provided some of the operational challenges could be addressed,30 this could open a path towards combining revenue and expenditure decentralization in an asymmetric fashion. IX. SUBNATIONAL DEBT AND FISCAL RESPONSIBILITY 1. The Challenge Rent-seeking and demands for seizing greater shares of the national resource pool are related to the issue of fiscal responsibility. Subnational debt has not been a salient issue in the discussion of autonomy. However it would not be surprising if departments, once autonomous, would reject existing obligations based on considerations of “odious debt�. Departmental decentralization happens against the backdrop of municipal indebtedness problems, which have only been partially resolved.31 The department’s spending responsibilities have been increased, but debt rules are weak and incomplete.  The total subnational debt is not fully or precisely known. The size of the departmental (about 0.5 percent GDP in 2006) and municipal debt (4.5 percent GDP, in 2006) is known to be understated because of unregistered borrowings and suppliers debt. Most municipal debt today is floating debt, particularly from suppliers, which is admittedly difficult to monitor. In general, subnational borrowing has poor oversight and very limited transparency.  The “20/200-rule�32 is too lax by international standards and has other limitations.33 The rule implies that both departmental governments and municipalities obtain more room to acquire more debt with the new hydrocarbon resources. instance); they do not influence much locational choices (contrary to the corporate income tax, for instance); they are not too difficult to administer locally (contrary to the VAT, for instance); and they are not used for inter-regional redistributive purposes (contrary to the personal income tax, for example). 30 Among others, the operational challenges are: effectively addressing horizontal imbalances arising from tax decentralization through the transfer system; selecting taxes and expenditures which are proportionate in amounts so that departments find the bargaining package attractive. Estimating tax revenue is difficult since revenue is a function of the tax rates which are applied. 31 In 1999, several municipalities could not pay their debt service and were intervened by the central government; the Programa de Readecuación Financiera (PRF) was established as 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, however this did not happen in all municipalities. Yet this marked, for the first time since decentralization was launched, the Central Government’s establishment of a tighter framework for municipal finances. 22  The rules for subnational debt are narrow and fail to encompass fiscal responsibility rules. New spending pressures could arise, particularly during the take-off periods of further decentralization. As discussed above, there could potentially be new regional taxes to restrain excessive subnational spending—but their magnitudes are likely to remain limited and, therefore, alone are unlikely to set a new incentive framework for responsible fiscal management.  More recently, some departments—particularly those which have high hydrocarbon revenue income and limited absorption capacity—have accumulated unspent balances. This creates a fiscal coordination problem among levels of government, since all levels should follow similar fiscal adjustment policies which might be required in light of decreasing hydrocarbon revenue. Addressing this however is all the more difficult since departments can use unspent resources freely without being tied to earmarking rules.34 Addressing these limitations are all the more difficult given that there is no national rules framework for fiscal responsibility. 1. Policy Options and Strategy It is not obvious that fiscal responsibility rules can be included as part of the Fiscal Pact. Nevertheless, some aspects can be addressed gradually. For a start, fiscal responsibility rules should be introduced for one year at a time only and should also encompass unspent balances of departments. Given the short-term horizons of actors in the context of volatile resource flows, multi- year commitments—as applied in Colombia, Peru, or Chile, among others—are likely not credible and enforceable. . In light of the limited enforcement capacity of the central level, the politically important strategy is to include the private sector. Past attempts at introducing market-based controls have largely failed,35 and the small domestic financial market implies structural limitations. One measure consists of private banks being forced to disclose debt held by municipalities and departments. For instance, financial market regulations could be established ruling that loans are only accepted on balance sheets of banks if certain information about them is made available and public. This measure would not control the floating debt problem; but it would establish basic transparency on a 32 The “20/200-rule� foresees that debt service of municipalities or departmental governments must not exceed 20 percent of current revenues and debt stock must not exceed 200 percent. In the meantime, the indicators of the “20/200-rule� should be tightened, and de-linked from transfers and current revenue of subnational governments. 33 The rule does not include commitments that are being converted into debt, and it does not offer a rules spanning over a period of time more than a year. 34 The only rule applying to unspent balances is to use these resources for investment; the earmarking rules do not apply for these resources. 35 The two most populous municipalities—La Paz and Santa Cruz—were required to have credit assessments by private banks, which however, were never accomplished. The domestic financial market seems to be too small for a large-scale private sector participation. 23 significant share of subnational debt. It would provide a starting point and the leverage to expand reforms later on. The circumstances outlined above also warrant a rather conservative and cautious approach for managing debt through administrative controls. For instance, a moratorium on departmental debt should be established until minimum transparency requirements are being met; the status and ownership of departmental debt is clarified; and basic fiscal responsibility rules are adopted and effective. The “20/200-rule�36 should continue but it should be tightened and complemented with broader fiscal responsibility rules outlined above. X. PROSPECTS FOR ACHIEVING A FISCAL PACT The decentralization game in Bolivia has been altered quite significantly with the presence of new bargainers at the departmental level. Two, opposing groups have emerged: the highland departments strongly aligned to MAS, vis-á-vis the Media Luna departments in the lowlands. None of these groups, however, do always act like a united bargainer. Their interests diverge which leads to a complex political process, compounded by the fact that there are swing voters as in the case of Cochabamba. The central government under the leadership of Morales clearly sees in departmental decentralization a threat to its power. It has used “divide-and-rule� strategies as well as its discretion to hollow out the increasing power of departments. The government’s constituents are mainly on the municipal level. Hence, it needs to be anticipated that bargains may eliminate some inefficiencies, but also create others. Actors play nested games with regards to the division of the resource pool; local taxation; expenditure assignments; and subnational debt. The outcome is likely to be an intergovernmental system with efficiency constraints. Moreover, as examples in other countries clearly illustrate, it is often the early winners of partial decentralization which provide the greatest obstacle to complete decentralization in a later stage. Transition arrangements create rents and in order to convince the veto-players to relinquish them, it is very likely that new rents will need to be created. Bolivia will not be an exception. Predicting outcomes of this bargaining process are fraught with uncertainty. In theory, the nine prefects and the central government constitute 10 bargainers which could develop not less than 49 bilateral relations among themselves. More importantly however, much depends on how the symbol of “autonomy� will be used by the central government or the departments in its bargaining 36The “20/200-rule� foresees that debt service of municipalities or departmental governments must not exceed 20 percent of current revenues and debt stock must not exceed 200 percent. In the meantime, the indicators of the “20/200-rule� should be tightened, and de-linked from transfers and current revenue of subnational governments. 24 process. In recent years, the different referenda and constitutional changes have created and maintained expectations that “autonomy� can be implemented eventually. The notion of “autonomy�, however, is ambiguous which makes the process of agreeing on the basic intergovernmental arrangements all the more difficult. There is likely no “end game�: a muddling through scenario is probably the most likely outcome. Hence it is needs to be expected that conflicts will continue. If resource rents get squeezed, episodes of non-cooperative games (Treisman 1999) are likely. Therefore it would not be unwise to use the Fiscal Pact initially solely for purposes of discussing the basic rules for decentralization. In a first step a rudimentary rules framework could be established without defining precise fiscal or expenditure targets or other decentralization measures. Once these basic agreements around the principles are achieved, prospects would be increased that agreement around other conflictive issues could be reached gradually: defining the distributable pool for transfers; allocating responsibilities and/or creating coordination mechanisms; and setting fiscal responsibility targets for each year. Such agreements could be renegotiated at certain pre-defined times, in order to provide flexibility and adjust to the constraints of actors to engage in inter-temporal commitments. Such explicit exit options would enhance the possibilities for agreement. 25 Bibliography Ahmad, Ehtisham; Brosio, Giorgio (2008): Political Economy of Multi-Level Tax Assignments in Latin American Countries: Earmarked Revenue versus Tax Autonomy, IMF Working Paper NO. WP/08/71, Washington DC. Ayo, Diego (ed.) (2004): Municipalización: Diagnóstico de una década, 30 investigaciones sobre Participación Popular y Descentralización, La Paz, Friedrich Ebert Siftung, USAID, Plural. Ayo, Diego; Bonifaz, Gustavo (2008): ¿Hegemonía indígena o interculturalidad?, Friedrich Ebert Stiftung, La Paz. Bahl, R.; Tumennasan. B. (2004): How Should Revenues from Natural Resources be Shared in Indonesia?, in: Reforming Intergovernmental Fiscal Relations and the Rebuilding of Indonesia: The 'Big Bang' Program and its Economic Consequences. Edward Elgar, Cheltenham, UK & Northhampton, MA, USA, p. 199-233. Barragán, Rossana; Roca, José Luis (2005): Regiones y poder constituyente en Bolivia. Una historia de pactos y disputas. Informe sobre Desarrollo Humano, Cuaderno de Futuro 21, La Paz. Barrios, Franz (ed.) (2006): Nuevo Pacto Territorial para Bolivia. Propuesta de nivel prefectural para la Asamblea Constituyente, PADEP/GTZ, Unión Europea, La Paz. Barrios, Franz (2007): Evaluación de algunas propuestas constitucionales de autonomías indígenas en Bolivia, Mayo de 2007, La Paz. Barrios, Franz (2008): Procesos de desconcentración prefectural en Bolivia. Teoría y práctica de la desconcentración, del ordenamiento espacial y de la cuestión regional. La Paz. Barrios, Franz (2009): Hacia un Pacto territorial en Bolivia. Conflictos, Conceptos, Consensos en torno a las autonomías, PNUD, La Paz. Brancati, D. (2009): Peace by design. Managing intrastate conflict through decentralization. Oxford University Press, Oxford, New York. Devarajan, Shantayanan; Khemani Stuti; Shekhar Shah (2007): The Politics of Partial Decentralization, The World Bank, Washington DC. Dickovick, Tyler (2006): Municipalization as Central Government Strategy: Central-Regional-Local Politics in Peru, Brazil, and South Africa, Publius 37(1):1-25. Eaton, Kent (2007): Backlash in Bolivia: Regional Autonomy as a Reaction against Indigenous Mobilization, Politics and Society 35(1):71-102. Eaton, Kent ; Kaiser, Kai et. al. (2009): The Political-Economy of Decentralization Reform, Washington, DC. FAM (2008a): 14 años de municipalización, Boletín Informativo no. 4, Junio, La Paz. FAM (2008b): Cuanto tienen los municipios para invertir y gastar en el 2009? Poder Local, Diciembre de 2008, La Paz. 26 FAM (2008c): ¿En Qué Gastan Los Municipios?, Estadísticas Municipales 2006-2007, La Paz. FAM (2008d): Inversión Municipal, Boletín Informativo no. 5, Mayo, La Paz. FAM (2008e): Pacto Fiscal. Propuesta de la FAM Bolivia, Septiembre de 2008, La Paz. FAM (2008f): Recursos IDH: tema pendiente, Boletín Informativo no. 6, Junio, La Paz. FAM (2008g): Situación IDH, Octubre de 2008, La Paz. Frank, Jonas (2007): Decentralization in Ecuador: Actors, Institutions, and Incentives, Nomos, Baden- Baden. Haggard, Stephan, Cristopher da C.B Garman, and Eliza Willis (1999): “The Politics of Decentralization in Latin America,� Latin American Research Review 34:7-56. Hans Seidel Stiftung, Fundemos (2005): Debate sobre las autonomías regionales, en: Opiniones y Análisis, La Paz. Hellman, Joel S. 1998. “Winners Take All. The Politics of Partial Reform in Postcommunist Transitions,� World Politics 50:203-234. Herbst, Jeffrey (2001): The Politics of Revenue-Sharing in Resource-Dependent States, WIDER Discussion Papaer NO. 2001/43. Hurtado, Hervas (2008): Consideraciones sobre el Estatuto de Autonomía de Santa Cruz, Marzo de 2008, Santa Cruz. ILDIS (2003): La descentralización que se viene. Propuestas para la (re)constitución del nivel estatal intermedio, La Paz, Friedrich Ebert Stiftung/Plural. La Razón (2008): El pacto fiscal se ofrece como una alternativa para abrir el diálogo. Ventana Ciudadana no. 95, 26 de Agosto, La Paz. Madrid, Raúl L. (2005): “Indigenous Parties and Democracy in Latin America�, Latin American Politics and Society 47(4):161-179. Morgandi, M. (2008): Extractive Industries Revenues Distribution at the Sub-National Level: the experience in seven resource-rich countries, Revenue Watch Institute, New York. O'Neill, Kathleen (1998): Understanding the Great Power Give-Away: Decentralization in the Andes, Paper presented to the 1998 meeting of the Latin American Studies Association, Chicago. O'Neill, Kathleen (1999): “Decentralization as an Electoral Strategy,� processed, Ithaca: Cornell University. PIEB (2008): Autonomías, Bolivia frente al espejo. Boletín del Programa de Investigación Estratégica en Bolivia, No. 9, Abril de 2008, La Paz. Prud’homme, Rémy; Huntzinger, Hervé; Guelton, Sonia (2001): Descentralización en Bolivia, La Paz, Interamerican Development Bank. Rodden, Jonathan A. et.al. (2003): Fiscal Decentralization and the Challenge of Hard Budget Constraints, MIT Press, Cambridged. 27 Toranzo, Carlos (2009): Economía política de los hidrocarburos en Bolivia. Unpublished report, La Paz. Tordo, Silvina (2007): Fiscal systems for hydrocarbons : design issues. Working Paper No. 123, World Bank, Washington, D.C. Torres Villagomez, Andrés (2008): Los Municipios en el Debate Constitucional, FAM, La Paz. Treisman (1999): “Political Decentralization and Economic Reform: A Game-Theoretic Analysis�, American Journal of Political Science Vol. 43, p.488-517. Van Cott, Donna Lee (2001): “Explaining Ethnic Autonomy Regimes in Latin America�, Studies in Comparative International Development 35(4):30-58. Webb, Steven B.; Gonzalez, Christian (2006): Bargaining for a New Fiscal Pact in Mexico, World Bank, Washington DC. Webb, Steven B. (2004): Fiscal Responsibility Laws for Subnational Discipline: The Latin American Experience. World Bank, Washington DC. World Bank (2005): Comparative Study on the Distribution of Oil Rents in Bolivia, Colombia, Ecuador, and Peru, ESMAP, Report No. 304/05, Washington DC. World Bank (2006): Towards an Inclusive Decentralization, Bolivia Institutional and Governance Review, Washington, D.C. Yashar, Deborah (2005): Contesting Citizenship in Latin America: The Rise of Indigenous Movements and the Postliberal Challenge, Cambridge University Press, Cambridge. 28