Report No. 20115-BO Bolivia From Patronage to a Professional State Bolivia Institutional and Governance Review (In Two Volumes) Volume l: Main Report August 25, 2000 Poverty Reduction and Economic Management Latin America and the Caribbean Region Document of the World Bank CURRENCY EQUIVALENTS (Exchange Rate Effective November 9, 2000) Currency Unit = Bolivian Peso Bs. 6.33 = US$1 US$0.15 = 1 Bolivian Peso FISCAL YEAR January 1 - December 31 Vice President: David de Ferranti Country Director: Isabel Guerrero Sector Director: Ernesto May Pillar Leader: Pablo Alonso Task Manager: Yasuhiko Matsuda Table of Contents PREFACE ............................................................. I EXECUTIVE SUMMARY .............................................................I Context and Objectives of the Bolivia Institutional & Governance Review ......... ............ i Informality in the Public Administration .............................................................. ii Options and Strategies for Institutional Reform in Bolivia ........................................... vii Conclusion: Institutional Reforms Are Urgent ............................................................. xi 1. INTRODUCTION ............................................................. I 1 .1. Transformation of the Bolivian Political Economy ........................... ..................... 1 1 .2. Institutional Reforms Since 1985 ............................ 4 1 .3. Public Sector Performance .......................... 7 1.4. The Current Reform Agenda .......................... 14 1.5. Summary .......................... 15 2. INSTITUTIONAL CHARACTERISTICS OF THE BOLIVIAN PUBLIC ADMINISTRA TION: IMMEDIA TE CA USES OF POOR PUBLIC SECTOR PERFORMANCE ..................................... 17 2.1. Absence of a Functioning Bureaucracy ............................................................. 17 2.2. The Legal Framework and the Organizational Structure of the Bolivian State ..... 18 2.3. Informality in Public Personnel Management .24 2.4. Micro Institutional Causes of Informality in Personnel Administration ................ 28 2.5. Informality in Public Expenditure Management ............................. ..................... 32 2.6. Micro Institutional Causes of Informality in Public Expenditure Management .... 40 2.7. Summary ............................................................. 41 3. THE POLITICAL ENVIRONMENT OF THE BOLI VIAN STA TE: DEEPER CAUSES OF POOR PUBLIC SECTOR PERFORMANCE .............................................................. 43 3.1. Patrimonial Dynamics of Party Politics ............................................................. 43 3.2. Coalition Politics and Inter-party Negotiations .................................................... 46 3.3. The Dynamics and Limits of Technocratic Governance ....................................... 49 3.4. The Impacts of Politics on Public Administration ................................................ 51 3.5. Summary ......... 59 4. BOLIVIAN FRONTIERS OF INSTITUTIONAL REFORMS ........................................... 61 4.1. Frontiers of Reform ............................................................. 61 4.2. Strengthening Internal Rules and Restraints: Regulatory Superintendencies ........ 62 4.3. The Insulation Strategy Without Consensus: Judicial Reform .............................. 69 4.4. Bringing the State Closer to People: Popular Participation and Local Accountability .............................................................. 71 4.5. Summary ............................................................. 80 5. REALISM WITH TEETH: RECOMMENDA TIONS ON THE INSTITUTIONAL REFORM IMPLEMENTA TIONSTRA TEGY ............................................................. 82 5.1. The Urgency of Reform .............................................................. 82 5.2. Tackling Informality by Taking Politics into Account ......................................... 84 5.3. Tackling Informality by Strengthening Institutions of Restraint ........... ............... 92 5.4. Other Institutional Reform Agenda .............................................................. 98 5.5. Conclusion .............................................................. 100 TECHNICAL ANNEX: BOLIVIA PUBLIC OFFICIALS SUR VEY ........................................... 102 Tables Table 1.1 World Development Report 1997, Private Sector Survey .............................8 Table 2.1 Legal and Funding Distinctions between Central Administration and Decentralized Agencies .................................................. 19 Table 2.2 Benchmark assertions on civil service arrangements .................................. 28 Table 2.3 Total and average budget deviations, 1990-92, 1994-96 ............................. 35 Table 3.1 Performance in public agencies .................................................. 58 Figures Figure 1. Informality and performance in four types of public agencies ..................... ix Figure 2.1 Central Government within the Non-financial Public Sector .19 Figure 2.2 Outline Architecture of Central Government .20 Figure 2.3 Perceived impact of the SAFCO Law, by types of agencies (Percentage of respondents answering "Yes" to the questions "Does SAFCO ... ?") .......... 23 Figure 2.4 Deviation of accrued spending from approved budget as a percentage of total agency budgets, 1995 .34 Figure 4.1 Superintendencies vs. Rest of the Public Sector .64 Figure 5.1 Satisfaction with Municipal Services, System Support and Tolerance. 83 Boxes Box 1.1 SNC: the decline of a service-delivery institution .1 Box 3.1 UDAPE .49 Box 3.2 Political Determinants of Social Investment Allocations, 1996-98 .53 Box 4.1 Bolivia's Autonomous Regulatory Systems .63 Box 4.2 Selected Municipal Survey Results .75 Box 5.1 Shared Patronage Regime and Eventual Political Decay in Venezuela . 82 Box 5.2 Hybrid Appointments in OECD Civil Services .86 Box 5.3 Semi-autonomous Tax Agencies in Latin America .89 Box 5.4 Peru's INDECOPI .90 Box 5.5 Getting the Basics Right: New Public Management in Developing Countries.94 Annexes (in Vol. 2) PREFACE ..I EXECUTIVE SUMMARY .. Context and Objectives of the Bolivia Institutional & Governance Review .. Informality in the Public Administration.. ii Options and Strategies for Institutional Reform in Bolivia . .vii Conclusion: Institutional Reforms Are Urgent ..xi 1. INTRODUCTION .1 1.1. Transformation of the Bolivian Political Economy .1 1.2. Institutional Reforms Since 1985 .4 1.3. Public Sector Performance .7 1.4. The Current Reform Agenda. 14 1.5. Summary .............................................................. 15 2. INSTITUTIONAL CHARACTERISTICS OF THE BOLIVIANPUBLICADMINISTRATION: IMMEDIATE CAUSES OFPOOR PUBLIC SECTOR PERFORMANCE ..................................... 17 2.1. Absence of a Functioning Bureaucracy .............................................................. 17 2.2. The Legal Framework and the Organizational Structure of the Bolivian State ..... 18 2.3. Informality in Public Personnel Management ...................................................... 24 2.4. Micro Institutional Causes of Informality in Personnel Administration ................ 28 2.5. Informality in Public Expenditure Management .................................................. 32 2.6. Micro Institutional Causes of Informality in Public Expenditure Management .... 40 2.7. Summary .............................................................. 41 3. THE POLITICAL ENViRONMENT OF THE BOLIVIAN STA TE: DEEPER CA USES OF POOR PUBLIC SECTOR PERFORMANCE .............................................................. 43 3.1. Patrimonial Dynamics of Party Politics .............................................................. 43 3.2. Coalition Politics and Inter-party Negotiations .................................................... 46 3.3. The Dynamics and Limits of Technocratic Governance ....................................... 49 3.4. The Impacts of Politics on Public Administration ................................................ 51 3.5. Summary .............................................................. 59 4. BOLIVIAN FRONTIERS OF INSTITUTIONAL REFORMS ........................................... 61 4.1. Frontiers of Reform .............................................................. 61 4.2. Strengthening Internal Rules and Restraints: Regulatory Superintendencies ........ 62 4.3. The Insulation Strategy Without Consensus: Judicial Reform .............................. 69 4.4. Bringing the State Closer to People: Popular Participation and Local Accountability .............................................................. 71 4.5. Summary .............................................................. 80 5. REALISM WITH TEETH: RECOMMENDA TIONS ON THE INSTITUTIONAL REFORM IMPLEMENTA TION STRA TEGY .............................................................. 82 5.1. The Urgency of Reform .............................................................. 82 5.2. Tackling Informality by Taking Politics into Account ......................................... 84 5.3. Tackling Informality by Strengtheninig Institutions of Restraint .......... ................ 92 5.4. Other Institutional Reform Agenda .............................................................. 98 5.5. Conclusion .............................................................. 100 TECHNICAL ANNEX: BOLIVIA PUBLIC OFFICIALS SURVEY ........................................... 102 Acknowledgement This report was prepared on the basis of missions in November 1998, and March and April 1999. The Bolivia IGR team was led by Yasuhiko Matsuda (task manager/political analysis/public expenditure management), and composed of Geoffrey Shepherd (regulatory systems), Nick Manning (civil service and personnel management), Linn Hammergren (justice system and corruption issues), Gary Bland (local governance), and Suzanne Dove (civil service and personnel management). The report also benefited from background materials based on a survey of public officials (conducted by a local firm, Encuestas & Estudios, and analyzed by Pablo Alonso, and Ranjana Mukherjee and Omer Gokcekus), a survey of sub-national governments (conducted by a team of Bolivian consultants led by Gonzalo Rojas-Ortuste), an assessment of public expenditure management (by Sofia Valencia), an analysis of social investment fund allocations (by Jonathan Hiskey), and an assessment of the role of the Contraloria General de la Repuiblica (by Suzanne Dove). The preface was prepared by Juan Carlos Zuleta. The peer reviewers were Malcolm Holmes and Philip Keefer. At various stages of drafting *the report, the team also received helpful comments from Guillermo Perry, Ernesto May, Pablo Alonso, Ronald Myers, Jose Antonio Gonzalez, Robert Ayres, John Heilbrunn, Cynthia Lopez, Paul Sisk, Allen Schick, Carlos Mollinedo, Gonzalo Rojas-Ortuste and Rene Rivera. The government counterpart was Alberto Leyton (Vice Presidency), with additional support from Paola Muller. In addition to her substantive contributions on civil service issues, Suzanne Dove provided valuable research and editorial assistance, including drafting Chapter 1 and editing and compiling the Annexes. The team received logistical support from the Bolivia Country Office, especially Fernanda Pacheco and Monica Claros. Meseret Kebede also provided assistance to the team throughout the preparation of the report. Acronyms and Abbreviations ADL - Administrative Decentralization Law ADN - Acci6n Democratica Nacionalista AGO - Accountant General Office BCB - Banco Central de Bolivia (Bolivian Central Bank) C'AS - Country Assistance Strategy C'GR - Controller General of the Republic C'ONFIP - Financial Prudential Norms Committee EDI - Economic Development Institute GOB - Government of Bolivia ICR - Implementation Completion Report IDB - Inter-American Development Bank IFMS - Integrated Financial Management System IGR - Institutional and Governance Review ILACO - Financial Decentralization and Accountability Project IMF - International Monetary Fund INE - National Statistics Institute IRP - Institutional Reform Project LOPE - Law of the Organization of the Executive Power MNR - Movimiento Nacionalista Revolucionario MOF - Ministry of Finance NIP - National Integrity Plan OECD - Organization for Economic Cooperation & Development ORA - Organizational Restructuring Agreement OTB - Organizaci6n Territorial de Base POA - Annual Operating Plan POAI - "Individual" Annual Operating Plan PPL - Popular Participation Law PSC - Civil Service Program SAFCO - Financial Administration and Control System SAMAPA - Servicio Aut6nomo Municipal de Agua Potable (Autonomous Municipal Water Agency) SBEF - Superintendency of Banks and Financial Entities SDC - Servicios Departamentales de Caminos SIRESE - System of Sectoral Regulation SIGMA - Sistema Integrado de Gesti6n y Modernizaci6n Administrativa SME - Small and Medium-size Enterprises SNAP - Servicio Nacional de Administraci6n de Personal SNC - Servicio Nacional de Caminos SPVE - Superintendency of Pensions, Securities and Insurance SRJ - Superintendency of Appeal SSC - Superintendency of Civil Service UDAPE - Unidad de Analisis de Politicas Econ6micas UESC - Unidad Ejecutora del Servicio Civil I PREFACE' The Bolivia Institutional and Governance Review (IGR) constitutes a major effort aimed at understanding the country's public sector and identifying challenges and possible directions for an ambitious and indispensable second-generation reform agenda. By exploring the concept and causes of informality, the IGR underscores the most significant obstacles facing the incipient Bolivian bureaucracy. Informality exists in several areas of public administration, including implementation of the SAFCO Law, public expenditure and personnel management. In most cases, its underlying causes have to do with the dynamics of politics. A weak private sector practically incapable of generating employment for the middle class, politicians' interest in obtaining electoral support from and control of the government bureaucracy, and a fragmented party system which forces political organizations to negotiate coalition agreements, are cited as sources of patronage and clientelism. In addition, deficiencies in the formal rules and procedures, weakness of supporting institutions, Ministry of Finance's (MOF) traditional concern about fiscal discipline at the expense of strategic allocations and efficiency in resource use, and delays in the implementation of the Statute of the Public Officials all present obstacles to the development of a well-functioning government. In the context of ongoing government reform efforts, the IGR provides four central recommendations, namely: (i) de-politicization of public sector personnel management; (ii) strengthening of central oversight capacity in financial and personnel management; (iii) learning from the recent successful experience of the autonomous regulatory agencies (Superintendencies) in an attempt to replicate it in other public sector entities; and (iv) use of "citizen voice" as a way to exercise social control over public administration. To cope with problems of corruption and the low quality of public services, the current government, with assistance from the Bank and donors (Denmark, The Netherlands and Sweden), has designed the Institutional Reform Project (IRP). The IRP, together with a new judicial reform project (in preparation), is a key element of the government's National Integrity Plan (NIP). As a broad-based public sector reform project, the IRP focuses on developing a series of horizontal reform measures (performance-oriented management and evaluation of public sector organizations; disclosure of government officials' assets and income of government officials; denunciation and prosecution of corruption; streamlining of procedures; procurement system reform; citizen education campaign; and civil service reform) to be vertically implemented in selected pilot agencies which will subject to major organizational restructuring. The IRP is now in its first phase of implementation. 1 Since the field work for the IGR was conducted, there have been several developments in Bolivia. This preface is intended to provide a brief update of these developments. II In this context, de-politicization of the public administration through a successful civil service reform is a sine qua non condition for moving from patronage to a professional state in Bolivia. This will only be possible with a credible Superintendency of Civil Service (SCS). Therefore the current delay in the nomination of the Civil Service Superintendent by the Senate is particularly worrisome. Likewise, completion of the initial stage of the SAFCO reform requires effective compliance mechanisms based on strengthening independent central oversight bodies such as the Controller General of the Republic, the SCS, and possibly, an autonomous regulatory entity for overseeing the government-wide financial management on behalf of the MOF, combined with an adequate information flow for monitoring agency behavior, and improved internal control capacities in public institutions. These tasks are being accomplished both through the Financial Decentralization and Accountability Project (commonly known as the ILACO II Project) which in the last year and a half has made significant progress and is now implementing an integrated financial management system (Sistema Integrado de Gesti6n y Modernizacion Administrativa, or SIGMA) in a number of central government entities, and through the IRP. Although the implementation of an integrated financial management system progressed only slowly and unevenly over the last decade, as described in the IGR, the Government has renewed its commitment to its implemenation, as evidenced, for example, in the recent enactment of decree 25875 that mandates SIGMA implementation. The vertical institutional reform focuses on organizations selected on the basis of demand for reform, political viability, support for other reforms, impact on poverty, economic impact, visibility, and their chance of being winners. Positive results from the reform of five pilot institutions (Ministry of Finance, National Customs Service, Internal Revenue Service, National Roads Service, and Ministry of Education) are crucial to the success of the Project. Moreover, IRP's financial sustainability is critically dependent on the performance of the first three of them, whose effectiveness is the basis for strengthening the state's fiscal management capacity. Successful reform of the latter two should provide a clear signal of the government's commitment to improving public service delivery in important areas of economic and social development. In all cases, positive results from implementation of vertical reforms would demonstrate that it is possible to develop non-politicized public sector organizations and that the government and the political system as a whole are serious about tackling the problem of a politicized public sector. Among the proposed vertical reform measures, reform of Customs is of particular importance. Customs was known to be one of the main illicit sources of political party financing, and it had gone through a series of corruption scandals that resulted in removal of its heads before the launching of the IRP. Hence enactment of the General Law of Customs (Law 1990) in July 28,1999 and the change of leadership to someone with international credibility, together with the fact that she has stayed in her post for a reasonable length of period so far, constitute two major breakthroughs in the history of institutional development in Bolivia.2 A recent progress report on Customs shows that 2 A couple of her immediate successors were replaced amidst allegations of corruption, each lasting only brief time in his position as the (pre-reform) customs diector. III during the first half of the year revenue collections increased by 10.1% with respect to a similar period last year, placing Customs at the forefront of the reform. Given the current economic crisis which has no doubt shrunk the revenue base due to reduced imports, the bulk of the increased collection might reasonably be attributable to improved efficiency in collection. Central to this success was the complete renewal of both management and front-line staff under the premise of demanding them a high-level professional expertise and effectiveness in exchange for an administrative career at Customs.3 As a semi- autonomous public entity, Customs appears to be following a similar path to institutional development as the Superintendencies. Although the progress made so far in Customs is by no means consolidated, its experience demonstrates that with an appropriate level of political consensus, it is possible to embark on a difficult vertical reform of even the most politicized public sector organization. By contrast, implementation of IRP in the other pilot agencies is generally lagging behind, indicating the political difficulty of implementing vertical reforms even in a limited number of priority areas. In this connection, prompt involvement of the MOF in IRP's implementation process remains, since the launching of the Project, a top priority, and the pending nomination by the Senate of a list of three candidates for the President and the Board of Directors of the National Roads Service (SNC) appears as a serious impediment to its institutionalization. Commitments by the Bank and donors to the ten year Adaptable Program Credit (APC) remain strong. However, this will not be sufficient to consolidate and continue the reform process, unless a consensus-building approach is initiated both among the government, the political opposition and civil society, and within the government itself for endorsement of the Project after change in administrations. Preliminary results of a recent evaluation of the NIP by an independent consulting firm ratify the initial perception that the role of the National Integrity Commission (NIC) in overseeing and enforcing consistency and coordination has been rather weak due to internal political problems. This indicates a need for reforming the NIC so as to involve the private sector and civil society, together with the government, in the fight against corruption. This is fully consistent with what the IGR terms the "citizen voice". Overall, notwithstanding the present obstacles and the challenges ahead in an increasingly difficult political environment aggravated by the social and economic crisis 3But merit-based hiring did not proceed without difficulties. Through the work of a specialized consulting firm a limited number of senior staff have been selected on a merit basis and hired by this entity. To overcome this problem, Customs has recently prepared, with the assistance of IRP's Technical Unit and the Bank, some innovative terms of reference of a new consultancy for selection of its personnel, according to which the firm will be paid both a fixed amount and a (variable) success fee. IV and the approaching elections, there are reasons to believe that institutional reform is indeed possible and that all efforts must be geared in that direction. EXECUTIVE SUMMARY Context and Objectives of the Bolivia Institutional & Governance Review 1. Transformation of the Bolivian political economy. Since the early 1980s, the Bolivian political economy has gone through significant transformation. The economic crisis that culmninated in hyperinflation and the political chaos that saw quick successions of nine governments between 1982 and 1985 and the early departure of the President in 1985 were replaced with consistent macroeconomic stability and evident consolidation of the democratic political regime. During the same period, Bolivia has implemented a number of bold reform measures to develop a market-oriented economic system. 2. In contrast to the notable progress with economic policy and certain structural reforms and in spite of the government efforts to modernize the central administration, the quality of public services remains low in Bolivia. This is because the core of Bolivia's public sector remains largely unreformed, and suffers from a variety of symptoms of institutional weaknesses. In 1990 the SAFCO Law, designed to turn the public sector into a modern, results-oriented institution, was promulgated, and efforts to implement it has received significant donor support, especially from the World Bank. The Bank and other donors have also assisted the Government of Bolivia (GOB) with its Civil Service Program that aimed at building up cores of competent public servants in key areas of public administration. However, these reform efforts have suffered from lack of government commitment and changing program focus, and to date have largely failed to produce noticeable results. 3. Current reform agenda. In the meantime, public discontent with the persistent corruption and poor public service delivery has raised the salience of the reform of public administration. Scandals after scandals are reported on the media highlighting both the seriousness of and public interest in corruption problems in the public sector. To address these problems, the Government has identified state modernization, fight against corruption, and judicial reform as three core components of its National Integrity Plan designed to tackle the remaining challenges. 4. A key instrument with which the Government is trying to combat corruption and strengthen public sector institutions is the Institutional Reform Project (IRP). The IRP is a comprehensive public sector reform project that aims at: * Development of a merit-based civil service; * Introduction of results-oriented management (e.g., performance budgeting and evaluation); * Strengthening the state's institutional capacities to control corruption (e.g., procurement reform, bureaucratic simplification); and * Reengineering and comprehensive reform of selected government agencies ii 5. The IRP's long-term emphasis on strengthening core institutions is the most appropriate strategy for combating corruption and other symptoms of public sector dysfunction. But the Project's success will depend critically on the Government's ability to confront difficult issues that arise from the reform's implications for the country's political process. The reform supported by the Project is expected to upset entrenched political interests that derive private gains from the politicization of the state and corruption, and also to encounter resistance from public sector employees, whose cultural and behavioral changes are crucial ingredients for the success of the Project. 6. Both the Government and the Bank have a high stake in the success of this Project as it constitutes a central element of the Institutionality Pillar, and thus of the Comprehensive Development Framework in Bolivia. The main objective of this Institutional & Governance Review (IGR) is to offer an assessment of the nature of the institutional weaknesses in Bolivia's public sector, analyze their probable causes, and offer some recommendations for devising a strategy for the IRP implementation. It is also hoped that by offering a cogent analysis of the political and institutional characteristics of the Bolivian state, the Review will also contribute to developing informed approaches to policy and institutional reforms in other areas. Informality in the Public Administration 7. Informality as a challenge for institutional development While there are several ways to characterize the institutional weaknesses in Bolivia's public sector, a concept ve find most useful to summarize various symptoms of the institutional dysfunction is that of"informality." The problem of "informality" exists when there is a significant gap between "ideal" or "desirable" behavioral patterns prescribed in a set of formal institutions (i.e., laws, rules, and organizational norms) on the one hand, and actual behavioral patterns that obtain without following the letter and the spirit of the existing formal rules on the other. Although the presence of informal dimensions itself is not necessarily a problem from the point of view of effective functioning of institutions, it becomes a source of institutional weaknesses when contradictions between the formal and the informal rules are resolved in favor of the latter, instead of having both sets of rules reinforce each other to produce predictable and desirable behavioral patterns. 8. Prevalence of informal rules over formal institutions causes trouble for reformers because it is primarily through strengthening formal institutions that interventions for public sector reform are expected to have an impact on bureaucratic performance. When the formal-informal gap is large, efforts to improve formal institutions may not only fail to produce the intended behavioral changes, but may even be subverted by the informal logic of behavior. Therefore, any institutional reform that aims at cultural and behavioral changes would have to address problems that arise from informality in order to bring about sustainable improvements in institutional performance. 9. In some sense, overcoming problems of bureaucratic informality is synonymous with institutional development, understood as a process of strengthening the abilities of formal rules and organizations for facilitating predictable human transactions within a given 111 domain. Informality is by no means a phenomenon unique to Bolivia. However, it is our view that the problems of informality have reached an excessive level in the Bolivian public sector. Unless they are addressed, efforts at institutional reform are likely to produce only limited results, if at all. 10. SAFCO and informality in the Bolivian state. One of the principal formal-legal instruments for public sector modernization in Bolivia is the SAFCO Law. Although the Law offers a reasonable legal framework for regulating public administration, its implementation has been incomplete and uneven. There is little evidence that SAFCO has had its intended impact on the functioning of the Bolivian state bureaucracy both in terms of promoting operational efficiency and results orientation and of ensuring probity and accountability. One simple reason for the incomplete implementation of SAFCO is the delay in the issuance of key implementing regulations (Normas Basicas). In fact, the Normas Basicas for the eight SAFCO systems became complete only in 1997, seven years after the promulgation of the Law. The Ministry of Finance, which is in charge of issuing seven of the eight Normas Basicas and overseeing their implementation by line agencies, has shown limited commitment to furthering the state modernization agenda including SAFCO implementation, focusing instead on the traditional role of macro fiscal management. It is also apparent that SAFCO implementation has been frustrated by resistance from public officials who do not necessarily welcome a rigorous system of accountability. Thus, for example, agencies' Annual Operational Plans (POAs) which are meant to guide annual budgetary allocations are not taken seriously by either the Ministry of Finance or the line agencies, and have become "rituals" to fulfill every year during the budget preparation cycle. Under the facade of formally applying some of the SAFCO management instruments (such as POAs), the informal reality of agency programming and resource allocations without regard to official POAs has persisted to date. Similarly, there is little evidence that "Individual" POAs (POAIs) are used effectively as an instrument of performance evaluation. 11. Donors, especially the World Bank, have their own share of responsibility for the limited progress Bolivia has made in improving its public financial management. While SAFCO implementation has been supported by three successive World Bank projects,4 it is now evident that these projects (except the last one) have failed to achieve greater efficiency and accountability in the ways in which public officials manage the state's financial resources. The projects have tended to emphasize technical aspects of information systems, while paying inadequate attention to altering the underlying incentives for better financial management. Until recently, the Bank continued to evaluate the SAFCO/ILACO projects favorably, and failed to recognize obstacles due to deep institutional and political difficulties. Nevertheless, in the last year and a half the ILACO II Project appears to have made significant progress and is now implementing the Integrated Financial Management System (SIGMA) software in a number of pilot agencies in the central administration. 4Public Financial Management Operation 1 (1987), Public Financial Management Operation 11 (1991), and Financial Decentralization and Accountability Project (1997). iv 12. A result of the incomplete implementation of the SAFCO Law is the lack of internalization of its norms and values by public officials. After nine years since the promulgation, there are substantial differences in the public officials' perception of SAFCO's impact on public administration. The survey of 692 public officials in 15 government agencies found some ambiguity in public officials' attitudes toward SAFCO. A little over 50 percent of the respondents thought that SAFCO did not encourage public officials to focus on results and that instead the Law made them afraid of possible prosecution ("juicio"). A similar percentage (51.7 percent) believed that SAFCO did not improve their agency's efficiency, although 56.9 percent thought it had an effect on reducing corruption. 13. Informality in public expenditure and personnel management. Symptoms of bureaucratic informality are also found in specific areas of public administration including public expenditure management and public personnel management. In the former, one of the clearest indications of the informality is the lack of credibility of annual budgets. Although annual budgets are supposed to be one of the most important formal-legal instruments that guides public administration, in Bolivia, as in many developing countries with weak fiscal institutions, budgets are rarely executed as they were approved by the Congress. Significant deviations of executed budgets from approved budgets point to both inadequate central control over details of resource allocations and use, combined with presence of unrestrained opportunistic spending behaviors by line agencies. Examination of budget data reveals that in the Fiscal Years 1990-92/1994-96, between 15 and 50 percent of the approved agency budgets were re- allocated across agencies.5 In five of these six fiscal years, the average agency-level deviations far exceeded the total deviation, implying that agency-level reallocations were not justified in terms of the need for macro fiscal adjustments. Mid-year reallocations appear to be more or less accepted behaviors within the government as long as the macro fiscal target is met, which is evidenced, among others, in the legal practice of budget reformulation by Congress in the middle of a fiscal year. 14. Informality is also amply observed in public personnel management. A well-known example is the widespread use of "consultants" for line functions. This phenomenon arises because agencies' ability to attract competent personnel is limited by the low salary level of regular government employees. The agencies try to circumvent this reality (based on the formal wage policy) by hiring "consultants" whose salary levels are not necessarily restrained by the government's wage policy. Another more obvious example is the persistent problem of corruption in the public sector, which is a clear case where formal institutional mechanisms for control and accountability are subservient to the informal reality in which opportunistic behaviors prevail without proper detection and sanctions. Through focus groups, in-depth interviews, and the survey of public officials, we have gathered evidence of other forms of informality in public personnel management, including: "Average" deviations were calculated on the basis of the absolute values of each agency-level deviation in order to avoid positive and negative deviations canceling out each other. The average was calculated with the data on approved and executed budgets at the agency level. Analysis of more disaggregated data (e.g., programs) is likely to reveal higher degrees of budget deviations. v * outside political pressure (28.5 percent of the respondents in the public officials survey), with the corresponding fear that refusal to go along with the pressure may result in the loss of either the job or salary (41.2 percent of the respondents); * deliberate destruction/concealment of agency-related information at the time of government change; and use of non-wage budget items (e.g., training) for salary supplements. 15. Survey results reported in Figure 1 indicate significant variations across agency types (autonomous agencies such as the Central Bank and the Superintendencies; ministries; servicios/institutios nacionales; and prefecturas) in degrees of informality (measured in terms of perceived degree in which internal regulations are credible; perceived seriousness of corruption in the agency; and perceived degree of political interference in personnel decisions such as hiring) and in performance (perceived degrees of agency mission accomplished). In general, autonomous agencies which are clearly better insulated from political interference demonstrate superior performance in terms of the perceived level of corruption, regulation enforced, and mission accomplished than the other three types. Figure 1. Informality and performance in four types of public agencies . _ EI~~~Mmission accomplished:, -.1 9 _ _ Credible regulationsX . . . ra~Corruption is not a problem , 8.3 -::--' 8 7. + _Nnpliiie hiin 6 5~~~~~~~~~~~~, 4~~~~~~~~~~~~~~~~~4 3 2 Autonomous Ministries Executive Prefecturas 16. Micro institutional causes of informality: weak control mechanisms. At one level, informality is a result of weak institutional mechanisms for management control and regulatory enforcement. The SAFCO Law was intended to introduce both results- vi oriented management practices (with the concept of operational decentralization) and to strengthen central oversight (with the concept of normative centralization). However, because of the weaknesses or indifference of the central oversight bodies such as the Controller General of the Republic (CGR) and the Ministry of Finance, and in the case of personnel management the virtual absence of an oversight entity, effective normative centralization has yet to be established. Furthermore, unavailability of information on agency performance and transactions, including on degrees of rule compliance, has made it virtually impossible for any central oversight body to exercise effective control functions. Thus, for example, the Treasury does not have timely information on budget execution during a fiscal year to adjust disbursement, if necessary, in order to ensure appropriate use of budgeted funds by line agencies. The CGR finds it difficult to audit line agencies' accounts because in many instances "there is no adequate financial statement to audit." For its part, the rector entity for government-wide personnel management, Servicio Nacional de Administraci6n de Personal (SNAP) has no access to agencies' salary and employment data, which would make it virtually impossible for SNAP to formulate a rational human resources policy for the whole government. Consequently, Bolivia's public sector has achieved a de facto "operational decentralization" without corresponding development of institutional capacities for "'normative centralization." This "operational decentralization" has contributed little to improving efficiency and results orientation of line agencies, and has probably permitted more opportunistic behaviors at the agency level. 17. Deeper causes of informality: political environment. Identifying specific institutional weaknesses that contribute directly to persistence of informality is only the first step of the diagnosis. The question is what motivates these informal behaviors in the first place. In the specific Bolivian context, one of the key reasons is found in the particular ways in which public bureaucracy is captured by political parties. Both anecdotal evidence gathered during the IGR field research and academic literatures on Bolivia's political economy point to the pervasive practice of patronage/clientelism as a defining characteristic of the relationship between the broader political environment and public administration in Bolivia. In general, sources of patronage politics are: * Weakness of the private sector to generate gainful employment, especially for the middle class, which encourages segments of the population to look toward the public sector for employment opportunities; * Politicians' interest in obtaining electoral/political support by distributing public jobs and other types of rents; * Politicians' interest in being able to control the government bureaucracy through placing "persons of confidence" in critical bureaucratic posts, either to ensure effective implementation of policies or to facilitate distribution of favors and rents. 18. In addition to these general reasons, the Bolivian politicians face another strong incentive to continue relying on widespread use of patronage appointments. Because of the fragmented party system (which in turn reflects deep social cleavages in the country) vii and the particular electoral institutions, the Bolivian political system has an overwhelming tendency to produce coalition governments. Given the lack of party identity on the basis of clearly differentiated policy programs, parties negotiate coalition arrangements with explicit agreements for sharing state patronage. This practice, called cuoteo politico (political quota), has been the basis for formation and sustenance of governing coalitions, and thus of democratic stability. Relatively stable coalition arrangements have also facilitated legislative passage of certain key reform measures, including the macroeconomic stabilization in the mid 1 980s. In that sense, patronage politics exercised through the practice of cuoteo politico has contributed to improved governance at the macro level. The problem is that this macro level governance has been achieved at the cost of deteriorating public services at the micro level because of the excessive politicization of the government bureaucracy and persistently weak institutional capacity as a result of it. Options and Strategies for Institutional Reform in Bolivia 19. Building consensus on the need of de-politicization. Advancing with the institutional reform agenda in today's Bolivia is a significant challenge both for the Government and for the donors that support the reform. Even though the Government's National Integrity Plan represents a sensible, comprehensive approach to combat corruption and improve public sector performance over the medium term, the good design alone does not guarantee eventual success of the reform. Expected political obstacles to the reform are significant not simply because public sector reforms almost always generate strong political and bureaucratic resistance as they inevitably affect distribution of political power but also because the status quo in Bolivia is based on a strong political logic of regime maintenance. It is unrealistic to expect that politicians whose political fortune depends on reliance on the cuoteo politico to easily give up their current source of power. Nevertheless, it is also probable that unless public sector performance improves noticeably, public discontent with the government (whoever is in power) and eventually with the current political system may lead to serious deterioration of the democratic governability in Bolivia as in some of the neighboring countries that are currently going through a period of substantial political uncertainty. Therefore, the institutional reform agenda must be pushed forward with a broad and sustained political consensus in order to transform the Bolivian state gradually from a source of patronage to a professional managerial entity capable of responding to citizen needs for high-quality public services. 20. A first step to building such a consensus is public recognition that one of the main causes of the institutional weaknesses of the Bolivian public administration is the excessive politicization. There are already a number of positive signs. Since the early 1 990s, the Bolivian political elite have been advancing with gradual reforms of democratic governance including the Constitutional and the electoral reforms in the mid- 1990s, and more recently the passage of the Political Party Law in 1999 that promotes greater internal democratization of the parties. High-ranking officials of the current Government describe the challenge ahead as one of"despartidizaci6n," and there is substantial media attention to the "ills" caused by cuoteo politico. This attention to the viii problems of politicization and efforts at "maturing" democracy should continue as one of the most important goals for the whole country. The government-wide Civil Service Reform and agency-specific "vertical" reform (e.g., ongoing reform of the Customs Agency) are two approaches to de-politicization, which requires such a broad societal support. 21. Transparency in the de-politicization process. In order to facilitate the inevitably difficult and gradual process of de-politicization/despartidizaci6n, it is critically important that the process be managed with transparency and fairness. If a major political actor, such as the main opposition party, perceived the process of implementing the Civil Service Reform to be manipulated in favor of other actors (most likely the governing coalition), the prospect of the reform would be dim. For this reason, the reform process must be transparent and overseen by an independent public entity with a high level of professional competence. The new Civil Service Law (Ley de Estatuto del Funcionario Publico) assigns this responsibility to the new Superintendency of Civil Service (SSC). By giving SSC an explicit mandate to monitor and regulate a reasonable number of political appointees who also meet minimum technical qualifications (hybrid appointees), the government may be able to increase transparency in the de-politicization process. In this regard, the Government might: / establish reasonable parameters for political appointments in government bureaucracy agreeable to major political parties; / publish information about the number of political appointees in each agency (possibly including their party affiliations), and make this information available to Congress and the public; and $ and commit to gradually decreasing the scope and magnitude of political appointments within the public sector. The fate of the reform, including the management of transparency-enhancing practices such as the ones suggested here, will depend critically on political neutrality and technical competence of the SSC. Thus, the Government 's initial efforts should be devoted to the creation of a credible and competent SSC. 22. Autonomization " and citizen "voice. " Bolivia should also take advantage of lessons that can be drawn from its own successful institutional reforms in the recent past. One of the clearest success stories in this regard is the establishment of effective autonomous regulatory agencies (Superintendencies). Unlike in many other countries, Bolivia's political elite have demonstrated both willingness and capacity to utilize autonomous agencies effectively. There is clear evidence from our public officials survey that autonomous agencies in Bolivia have achieved a much higher level of organizational development vis-a-vis the rest of the public sector. Thus, so long as a consensus can be forged to grant effective autonomy to particular agencies and sustain it, this option is not only attractive but also viable in the Bolivian institutional context. However, caution should be exercised in extending this "autonomization" strategy because autonomy can create other costs to efficient public administration in terms of ambiguous accountability arrangements, duplication of rules and functions, and difficulty of coordination within the ix public sector. In general, "autonomization" should be considered (with caution) in areas where political independence is sine quo non for the agency's successful performance. In contrast, given the weak accountability and coordination capacities in the public sector, "autonomization" may not be recommendable for most service delivery functions, which require both coordination and an effective accountability arrangement with the parent ministries. 23. Another positive experience from Bolivia's own institutional reform efforts is the use of "citizen voice" as a means for social control over public administration. The Popular Participation reform has empowered the communities' ability to hold local governments accountable for their actions, at least in some of the municipalities, and this experience is worth replicating at the national level. However, the effectiveness of social control mechanisms depends on a number of factors including the ability of social organizations to avoid/withstand capture by the municipal governments themselves or by other political actors such as the political parties. A mechanical replication of municipal institutional mechanisms (e.g., Comite de Vigilancia) at the national level is unlikely to work effectively given the differences in institutional contexts between the two levels. Nonetheless, the principle of mobilizing societal support should be applicable for the Institutional Reform at the national level. In this sense, the Government should devise a strategy of engaging the private sector and the media as supporters of the reform agenda, consulting with them more closely on reform implementation issues, and providing them with information on institutional performance of the public sector so that these societal actors can monitor progress in implementation and sustain constructive pressure. It would also be useful if the MOF and the CGR adopted a more open policy of information disclosure in order to allow citizens to monitor and scrutinize efficiency and probity in public expenditure management. 24. Strengthening "normative centralization. " After the de-politicization, another important theme that emerges from the findings of this Review is the importance of strengthening central oversight capacity as a prerequisite for moving toward greater operational decentralization and delegated decision-making for results-oriented public management. Currently, the Bolivian public sector does not possess institutional pre- conditions for moving rapidly toward a more advanced model of public management practiced only in a handful of OECD countries. Concerted efforts must be made at the initial stage of the IRP implementation to establish effective oversight mechanisms in the core areas of personnel and financial management and well-functioning institutional mechanisms for generating and sharing management information within the public sector. Once again the role of the SSC would be critical in this regard. A strong central oversight body that is autonomous from short-term partisan political considerations is essential for implementing the de-politicization strategy through both the civil service reform and the public financial management reform. The Government might also re- consider the current institutional arrangement for overseeing adequacy of procedural compliance in financial management. The current arrangement of divided responsibilities whereby the Ministry of Finance sets the regulatory framework/standards of the seven SAFCO management systems and the CGR oversees the government control system has so far failed to ensure installation of minimum institutional capacity for x procedural compliance and accountability at the agency level. One option would be to expand the roles of the SSC, once its effectiveness and authority are clearly established, to include monitoring and ensuring development of agencies' capacity for internal control. 25. A second option is to revitalize the CGR by harnessing the widely recognized technical expertise of its personnel to improve government control. Even before government-wide implementation of all internal regulations for SAFCO systems is completed, there is room for the CGR to become more assertive in enforcing implementation of the audit recommendations it does make. Another way to strengthen the CGR's role is to improve communication between the CGR and public managers. Many managers expressed frustration at the CGR's reluctance to provide information on how to comply with SAFCO; managers learn that they have not complied only after they receive the audit findings, creating an atmosphere of fear among managers who want to follow the rules. 26. Certainly, much of the lack of effective government control is explained by the low frequency with which the CGR's findings are followed by visible consequences. For control mechanisms to function effectively, it is critically important to have a strong, independent judiciary as well as a professional body for prosecuting corrupt acts (e.g., Ministerio Puiblico). Recent research on autonomous regulatory bodies has also found that an independent judiciary is an important contributing factor to effectiveness of regulatory governance.6 Although Bolivia has been grappling with judicial reforms for the past several years, a clear consensus has not emerged to facilitate constructive collaboration among different institutional actors (e.g. Consejo de Judicatura, Supreme Court, Ministry of Justice) to design and implement a credible judicial reform program. Along with the establishment of an independent and competent Superintendency of Civil Service and a stronger CGR, forging a lasting political consensus on a judicial reform should be one of the highest priorities for Bolivia's public sector institutional reform agenda. 27. Credible and binding budget. Finally, another important institutional prerequisite for the eventual introduction of results-oriented management is predictability in budget allocations to line agencies. Without predictable resource allocation, line agencies could not take advantage of their increased managerial autonomy and flexibility, and thus could not be held accountable for their actions. Currently, there is a high level of uncertainty in budget allocations to line agencies because of the unreliable revenue estimates and the lack of firm political commitment behind policies indicated in the annual budget. Ultimately, a budget will not become binding and credible unless it truly reflects the country's strategic and political priorities as translated into the government's expenditure program. The Government has the intention to implement a new budgeting by results as part of the IRP, but development of this component by the Ministry of the Presidency (MOP) appears to be running behind the others. In this regard, more emphasis on coordination between the MOP and the MOF is required in order to ensure successful 6 Brian Levy and Pablo T. Spiller (eds.) Regulations, Institutions, and Commitment: Comparative Studies of Telecommunications. 1996. Cambridge University Press. xi implementation of a performance-oriented management and evaluation system, a necessary condition for the establishment of a solid and credible results dissemination system. The concern is that given the MOF's historically weak commitment to public sector reform, this important component may also suffer from similar lack of leadership during the IRP implementation. Clear signals of commitment, including concrete actions toward further developing a medium-term budget reform strategy, would be an important gesture from the Ministry of Finance. Conclusion: Institutional Reforns Are Urgent 28. Institutional & Governance Reviews (IGRs) are the World Bank's new analytical instrument that purports to explore "good fit" between a given country's institutional context and options for public sector institutional reform. Bolivia's frustrating experiences with the recent attempts at reforming/modernizing the central public administration (e.g., SAFCO, PSC) indicate the importance of better understanding "good fit." Simply importing "good ideas" from elsewhere without appropriate local adaptation would not work. In that spirit, this Bolivia IGR has explored incentives that shape behaviors within Bolivia's public administration. It is our view that the nature and the magnitude of politicization cautions us against "leapfrogging" into an idealized state of public administration we observe in a few OECD countries. There are prior steps to be taken before Bolivia can reach the "cutting edge." The strategy for implementing the Institutional Reform in Bolivia must balance the need to be realistic (by taking underlying political factors into account) and the need to have some "teeth" in its content (by strengthening accountability and control mechanisms to curb corruption and other forms of opportunistic behaviors). 29. Institutional reforms are urgent priorities for Bolivia. There are signs of reform fatigue among Bolivians, whose quality of life has not improved dramatically despite 15 years of fairly consistent reform. Unless the state's ability to respond to citizens' needs for better public services and more effective poverty alleviation efforts, the credibility of the current regime can be threatened. The recent social protests in Cochabamba and elsewhere are telling indications of the fragile legitimacy of the state. To arrest the trend of declining trust in the state, Bolivia should adopt a concerted, medium-term strategy to reforming public sector institutions and improving governance not only through the [nstitutional Reform Project alone, but also through other related efforts including decentralization and judicial reform. 1. INTRODUCTION l.1. Transformation of the Bolivian Political Economy 1.1.1 For much of the 20th century, Bolivia was one of the most under-developed countries in Latin America, plagued by political instability and economic stagnation. Since the mid 1 980s, however, the Bolivian political economy has undergone a r emarkable transformation. The country has become one of the most aggressive reformers in the region, and today finds itself at the forefront of fighting corruption and improving governance. This Institutional and Governance Review is intended to provide analytical support to Bolivia's ambitious institutional reform agenda. Any attempt to understand Bolivia's governance and institutional reality needs to take into account the remarkable transformation of the country in the last decade and a half. While Bolivia continues to suffer from one of the highest incidences of poverty and other social and economic problems in the region, it is no longer the "basket case" it was often considered until the mid 1 980s. Before considering some of the present institutional and governance issues, a brief historical review is in order to put the subsequent assessments in a broader context.7 1.1.2. Bolivia before 1985. Even within the Latin American context of political and economic instability, Bolivia has gone through a particularly tumultuous history. By bne count, there were more than 190 changes of government between independence in 1825 and the early 1 990s, more than one change of government a year.8 Between the Revolution in 1952 and 1978, when the military dictatorship fell, Bolivia was ruled first by a one-party revolutionary regime of the MNR (1952-64) and later by a military regime (1964-78). It was during this period that Bolivia advanced its model of state-led capitalist development and saw its state apparatus grow in both size and scope. Thus between 1970 and 1978, the public sector employees grew from 66,000 to almost 170,000.9 However, this relative stability was replaced by extreme political turmoil of 1978-82, when seven military and two civilian governments replaced each other in quick succession.10 1 .1.3. In 1982, Bolivia began its transition to democracy with the election of Hernan Siles Zuazo as president. The transition proved to be a painful process laden with social and political conflict and economic crisis. Siles was elected by the Congress with broad support from the military, party leaders, organized labor and the private sector. By this time, however, the economic mismanagement during the past several years had taken a toll and pushed the country into severe economic disequilibrium. Instead of supporting 7 A more detailed discussion on the country's recent historical evolution is found in the Annex 1. 8 Eduardo A. Gamarra and James M. Malloy, "Bolivia: Revolution and Reaction." In Howard J. Wiarda and Harvey F. Kline (eds.) Latin American Politics and Development, 3rd Ed. 1990. p. 362. 9 Ministerio de Trabajo y Desarrollo Laboral, Anuario de Estadisticas de Trabajo (La Paz: Instituto Nacional de Estadisticas, 1980,1981), and Estadistica Laboral Periodo 1970-75 (La Paz: Instituto Nacional de Estadisticas, 1976). '' The political decay reached a crisis proportion during the government of General Luis Garcia Meza, which was not only extremely repressive but was also notoriously corrupt with known ties to drug- traffickers 2 the new, fragile democratic government in its fight to restore economic stability, the same political forces that negotiated Siles' election, mainly the Congress, organized labor, and even Siles' own vice president, obstructed the government's legislative agenda, and even threatened to unseat him by way of "constitutional coups." Devoid of congressional support, the Siles administration relied on rules by decree. While the government negotiated several IMF stand-by programs in the three-year period, all faced strong opposition especially from organized labor, which forced the government to soften the content of the stabilization measures. While the government engaged in repeated episodes of announcing new reform packages and watering them down upon societal pressure, the economy deteriorated further with inflation reaching more than 20,000 percent. Given the inability to find a way out of the economic crisis, Siles was forced out of power one year early, and new elections were called in 1985. 1.1.4. Throughout this period in question, the Bolivian political economy was characterized by the extensive use of public sector employment as a source of patronage to be distributed among the politically-connected. While patronage politics is by no means unique to Bolivia, it has been argued that the extreme dependence of the Bolivian economy, with its limited ability to generate gainful employment for the urban middle class, has heightened the importance of public sector jobs. It was also a period marked by repeated, often violent, confrontation among competing groups such as the military, organized labor, regional interest groups (e.g., Comite Civico de Santa Cruz), and the political parties. With the political stability threatened so much by the problem of governability for much of the post-independence period, the country did not really enjoy an institutional pre-condition for sustained economic growth and social development. 1. 1.5. Bolivia since 1985. Against this historical background, major transformation of the Bolivian political economy began in the mid- 1 980s. First, the newly elected government introduced bold macroeconomic stabilization measures that later became a model of a sort for the region, and successfully controlled hyperinflation. The measures did encounter strong resistance from some Congressional opposition and from organized labor, which had repeatedly frustrated the previous government's economic stabilization attempts. However, the government (led by the MNR president) managed to reach an inter-party agreement ("pact") and form a viable legislative coalition with the main opposition party (ADN). Street demonstrations by non-parliamentary opposition groups were controlled by declaring a state of siege. In important ways, these measures went beyond economic stabilization, and re-established authority of the state and governability in a country which had been lost almost completely during the chaotic years in the early 1980s. 1.1.6. Since 1982, Bolivia has experienced four successful changes of government by democratic electoral means. While coalition governments that are necessitated by the combination of the fragmented party system and the country's unique electoral system have not always been without difficulty in reaching and maintaining political consensus on the government's legislative agenda, the system has functioned remarkably well by regional standards. Instead of allowing minority governments to be undermined by inter- party disputes and throwing the country into crises of governability, as has happened in 3 some neighboring countries, the Bolivian political system has permitted the maintenance of minimum governing coalitions and the broad political consensus on the market- oriented reforms. Based on their capacity to reach political consensus with each other, the Bolivian political elite continued to deepen democratic reforms of the state. In the mid 1990s, Bolivia drafted a new Constitution that introduced a plurality voting system (sistema uninominal) to 50 percent of the lower house seats, with a view to making members of the Congress more responsive to voters. The new Constitution also prolonged the terms of the president, Congress, mayors and municipal council members to five years, giving them a longer timeframe in which to govern the country. The institutional model for democratic governance was further developed with the Popular Participation Law of 1994, which created 311 municipalities, many of them new, and granted them significant fiscal and administrative responsibilities. 1.1.7. On the economic front, too, Bolivia's accomplishments have been considerable. In a region that has recently seen significant progress in economic and structural reforms, Bolivia stands out as one of the most aggressive and sustained reformers. Overcoming an acute economic crisis in the 1980s, successive Bolivian governments have continued to introduce important reform agenda with significant donor support.11 Sustained government commitment to reform, combined with on-going support from donors, have produced some encouraging results. In 1985, inflation reached 23,500 percent, external public debt was 540 percent of exports, and real GDP had declined by a total of 10 percent in the past five years. Today, inflation has fallen to around 7 percent,12 and the non-financial public sector debt was 4.1 percent in 1998.13 Between 1986 and 1996, Bolivia's average annual real GDP growth was 3.1 percent, as compared to an average of --2 percent from 1980-1985. Per capita GDP has recuperated some of the losses from the crisis period, and private investment and export performance have also improved in the 1990s. 1.1.8. The transformation of the Bolivian political economy from a volatile, often near praetorian, situation for much of its post-independence period to a relatively stable multi- party democracy since 1985 is a remarkable accomplishment. When several of its neighbors have seen their democratic regimes falter because of traditional political elite's inability to cope with their nations' political and economic governance (e.g., Ecuador, Peru, Venezuela), Bolivia's steady, though still fragile, progress toward consolidating the democratic political regime is particularly notable. The establishment of macro political governability has been one of the most important achievements in Bolivia's recent history. Although some aspects of the political system still appear fragile (e.g., will the major parties survive inevitable change of leadership?), there have emerged a relatively " Besides the macroeconomic stabilization with the New Economic Policy (DS.2 1060) in 1985, other important reform efforts include the reforms of the Central Bank and the Superintendency of Banks, financial management reform (SAFCO-ILACO), the civil service reform, education reform, popular participation/administrative decentralization, capitalization/privatization and the accompanying pension reform. The most recent example is the reform of the customs administration. l Poverty Reduction and Economic Management (PREM)/Development Economics and Chief Economist (:DEC) indicators, 1998. '3 1999 Bolivia Public Expenditure Review, p. iii. 4 predictable set of "rules of the party political game" that provides for an important degree of political stability. 1.2. Institutional Reforms Since 1985 1.2.1. Besides the well-known macroeconomic stabilization and democratic transition in mid to late 1980s, Bolivia has also introduced some major institutional reform measures in the past decade or so. As part of the overall macroeconomic reform, the government re-created in 1987 the Banking Superintendency as an autonomous agency to supervise the banking sector, preventing the slide of the banking system into insolvency,14 and established the formal independence of the Central Bank (strengthened further by the 1995 Central Bank Law). Other reform measures, which will be discussed below, include: * Public administration modernization measures through the 1990 Law of Financial Management and Control (SAFCO), and the Civil Service Program; * The 1994 Popular Participation Law (PPL) and the complementary 1995 Administrative Decentralization Law (ADL); * The 1995 Capitalization Law; and * The Judicial reform. 1.2.2. SAFCO. Since the early 1990s, a key component of the GOB's work on state modernization and reduced public sector corruption has been the SAFCO law, passed in 1990. SAFCO established a set of management systems to replace the tangle of contradictory, confusing and obsolete rules that had failed to provide direction for public administration in Bolivia for a decade. The SAFCO philosophy moved away from a focus on compliance with detailed rules relating to bureaucratic procedures, instead concentrating on the results achieved with public resources. 1.2.3. However, the implementation of the SAFCO Law has been delayed significantly first by the government's slowness in issuing the necessary implementing regulations (Normas Basicas), which was only completed in 1997 (except the Basic Norms for Government Control). Before the issuance of these government-wide regulations, none of the agencies would be able to issue and implement their own SAFCO rules even though the philosophy of operational decentralization entrusted them the authority to introduce SAFCO at the agency level. SAFCO implementation has advanced very slowly even after the issuance of all the Basic Norms in 1997. One reason is the failure of the Ministry of Finance, the rector entity for seven of the eight SAFCO systems, to develop its own internal SAFCO regulations, which the other agencies would have used as examples. Without the management systems defined by SAFCO, it is difficult to move forward with the other components of state modernization. Some officials complain that even in agencies where the systems are formally in place, actual practices have not changed. The dearth of information on inputs, objectives and outputs/outcomes 14 The Superintendency of Banks had existed as a regulatory agency since 1928, but had been merged with the Central Bank since 1970. The 1987 reform once again separated the Superintendency from the Central Bank, and gave it much greater operational autonomy. See Chapter 4 for more details. 5 makes it hard to check compliance with procedures and even harder to hold managers/agencies accountable for performance. 1.2.4. Civil Service Program. The GOB has also tried to address the importance of creating a corps of technically competent, motivated public employees. The Civil Service Program (PSC) was born in 1992 and aimed to create a critical mass of around 2,500 employees. After five years of erratic recruiting, the total number of posts filled was in fact so small that its impact on improved public sector performance was negligible. By 1997, 729 posts were supposed to have been filled under the PSC, but in November 1997 only 248 had been. In 1996, 290 posts were designated to be incorporated into the PSC; 232 recruitment processes were announced; 94 people entered the civil service, and 59 people left. Retention was also a problem; many people hired under PSC left in frustration, as they felt "sandwiched" between highly politicized upper and lower levels of public personnel. 1.2.5. It was evident that the Program suffered from lukewarm political support, in contrast to other reform measures that received clear and strong government commitment to implement (e.g., Popular Participation). Turnover of PSC staff in participating agencies remained fairly steady from 1995-1997 (average of about 11 percent), but after a change of government in mid- 1997, turnover of PSC staff more than doubled, reaching 24 percent in 1998. Moreover, the agencies which had lower levels of PSC staff by 1997 had higher levels of turnover, indicating possible sources of strong resistance to a merit- based system. In 1998,15 only 30 posts were assigned to the PSC; the project gradually ground to a halt. After the change of government in 1997, both the new Government and the Bank agreed that the PSC, which did not aim at improving the broader institutional environment within which a relatively small number of qualified technicians worked, was judged to be too limited in scope and unsustainable over time. As a result, the civil service reform was re-conceptualized and re-designed as the Institutional Reform Project as a means to support implementation of the Government's National Integrity Plan. 1.2.6. Popular Participation/Administrative Decentralization. Another important institutional reform in Bolivia is decentralization and deconcentration of many public services, embodied in the Popular Participation Law, the Law of Municipalities, and the Administrative Decentralization Law. Before these reforms, Bolivia's highly centralized government structure was similar to other countries in the region. The PPL and ADL reforms constitute part of an on-going effort at shrinking the state and increasing popular participation in government decisions that directly affect citizens' lives.16 The PPL, supported by the recently passed Law of Municipalities, makes local governments responsible for considerable resources and public investment decisions, and establishes mechanisms whereby local communities can monitor local government spending and hold their elected officials accountable for spending decisions. The ADL deconcentrates 15 September 1998 data. Reforma del Servicio Civil en el Marco del Plan Nacional de Integridad. Ministerio de Hacienda, 1998. 16 However, PPL/ADL did not necessarily shrink the overall size of the state, and instead created a bloated bureaucratic structure at the departmental level, which hampers efficiency and responsiveness of public services at the municipal level. For more details, see 1999 Bolivia Public Expenditure Review, Chapter 3. 6 resources and operational responsibility to the prefectures, in an effort to increase public sector efficiency and tailor services to departmental needs. In theory, at least, it also casts the departmental executive as an intermediary between the center and the municipalities. 1.2.7. The results of these reforms have been mixed. On one hand, some municipalities show considerable impacts from local investments in development, heretofore concentrated in a few large urban areas. On the other hand, some critics argue that the clientelistic culture of the central government has simply spread to the local level, as evidenced by high mayor and councilor turnover. This criticism is also leveled at the prefectures, where it may be especially problematic since prefects are named by the President. Low human resource capacity at local and departmental levels, combined with high turnover, are likely to have a negative effect on performance. We have seen evidence of limited administrative capacity in many municipalities. For example, the Contaduria has experienced difficulties in obtaining properly prepared financial statements from municipal governments, often because the people responsible for financial management at the local level are not properly trained. A similar problem exists with preparing the Annual Operating Program (POA). Three of the four municipalities studied in Annex 8 had donor assistance in preparing their POAs. Chapter 4 and Annex 8 contain more information on the successes and shortcomings of decentralization. 1.2.8. Capitalization and Privatization of Public Enterprises. One area of structural and institutional reform many governments in the region began to tackle in the 1 980s is privatization of state-owned enterprises. Bolivia, however, moved cautiously on the privatization front, when it embarked on bold macroeconomic stabilization measures in the late 1 980s. The government policies for divestitures of public enterprises began with the 1992 Privatization Law and with privatization of government-owned factories and airports in that same year. The privatization process received additional impetus in 1995 with the Capitalization program that transferred ownership of several large public enterprises in strategic sectors including hydrocarbons, telecommunications, electricity, rail and airlines. Under this program, five of the country's six large public enterprises were transferred to private management, generating $1.7 billion in foreign investment committed to capitalize the divested enterprises. A parallel privatization program transferred over 50 smaller enterprises to the private sector and brought $149 million in revenues. 1.2.9. In order to ensure successful transfer of asset ownership to the private sector, the government established a modem system of autonomous regulatory agencies in key sectors including electricity, hydrocarbons, mining, telecommunications, portable water, and transports. Later in 1998, the parallel financial regulatory system was consolidated into the current structure where one entity supervises the banking sector and another the non-bank financial sectors (i.e., pensions, capital markets, insurance). A third instance was established to deal with disputes unresolved by the first tier of regulatory system. As discussed later in this Review, the establishment of these autonomous regulatory agencies has proven to be among the most significant institutional reform measures taken in recent years. 7 1.2.10. Although full competition in some sectors will only begin after transitional periods end, the restructurings lay the underpinnings for competition and thus for increased productivity and growth. Already, there is evidence of expanded coverage in the infrastructure sectors. In the electricity sector, high levels of investment have increased installed capacity by around 40 percent. In telecommunications, the penetration rate has risen from 3.6 percent in 1994 to 5 percent in late 1997. The cellular phone market provides a particularly dramatic example of growth: the number of cellular phones grew from around 5000 in 1994 to 55,000 in mid-1997, to 200,000 in early 1999. In the hydrocarbons sector, the Bolivia-Brazil gas pipeline is being funded largely by capitalization investments, and is expected to make an enormous contribution to export growth, estimated to generate $1 billion per year.17 1.2.11. Capitalization and accompanying regulatory reform comprise an important contribution to strengthening the private sector environment. While the advances made in privatization would be nearly impossible to reverse, the sustainability of the regulatory agencies is less certain. Regulation is just one component of the overall framework for private sector activity, and in a country with no history of independent regulation, the Superintendencies are attractive targets for political influence. As the 1999 Capitalization ICR points out, "ultimately the rules of the game must be enforced by the courts, which in Bolivia have not been reliable or efficient arbiters."'8 1.2.12. Judicial reform. Aware of the damage that its poorly functioning judiciary was doing to the country's growth prospects, the GOB began judicial reform efforts in the early 1 990s with changes in the constitution and secondary legislation. Several donors, including the World Bank, offered funding to support the necessary accompanying transformation and strengthening of sector institutions. Reform attempts have encountered stubborn resistance from entrenched interests inside and outside the judiciary. This has meant the slow creation and even slower entrance into effective operations of such key organizations and institutions as the Judicial Council, judicial career and training program, and Constitutional Court. Progress made in 1999 was significant, but the country must continue to press for full implementation of the proposed changes as well as others which are still in the planning stage. 1.3. Public Sector Performance 1.3.1. Despite these efforts and some positive results, Bolivia's public sector continues to suffer from multiple weaknesses that undermine reform efforts. According to the survey of private entrepreneurs conducted in 1997, Bolivia's public sector requires businesspeople and citizens to wade through considerably more "red tape" than other countries in the region even though the level of regulatory burden as such is not perceived as excessively high. The country also scores poorly on the reliability of the judiciary and on corruption. ' Bolivia Capitalization Program Adjustment Credit Implementation Completion Report, 1999, p. ii. Ibid. p. iv. 8 1.3.2. The main consequences of these public sector dysfunctions are the poor quality of public services and the government's limited ability to make a significant dent on the country's poverty. The Bank's 1998 Country Assistance Strategy (CAS) reports 67 percent of all Bolivian households living below the poverty level; in rural areas 81 percent of households live in poverty. Bolivia's average GDP growth between 1985 and 1995 - 3.1 percent - was about the same as other lower middle income countries in the region, but as a poorer country, Bolivia needs higher growth rates for a sustained period in order to achieve significant poverty reduction. 19 Table 1.1 World Development Report 1997, Private Sector Survey Bolivia High Similar Full Lowest LCR Income Income Sample Income countries Law & Order Crime 3.89 3.54 4.75 4.52 4.96 5.21 Judiciary Unreliability 4.74 3.01 4.82 4.19 4.45 4.34 Policy Uncertainty 3.82 2.86 3.91 3.56 3.69 3.87 Corruption 3.53 1.93 4.02 3.24 3.60 3.11 Regulatory Obstacles Level 3.09 3.03 3.28 3.21 3.25 3.50 Time spent with* 3.00 1.52 2.56 2.26 2.61 2.18 Notes: High Income: Austria, Canada, France, Germany, Italy, Switzerland, United Kingdom, and United States. Similar Income: Kyrgyz Rep., Annenia, Macedonia, Moldova, Morocco, Bulgaria, Kazakhstan, Ecuador LCR countries: Colombia, Costa Rica, Ecuador, Jamaica, Mexico, and Venezuela. *Senior management's time spent with officials about changes and interpretations of laws and regulations. 1.3.3. Revenue administration. Revenue generation is an area that the GOB has been trying to improve since the adoption of the macroeconomic stabilization. To date, efforts to strengthen performance by the Tax and Customs Agencies have fallen short of expectations. Even though domestic tax collections increased from about 1 percent of GDP to 7.8 percent of GDP between 1986 and 1992, and by 1995 they were up to 10 percent, corruption is said to be a serious problem in tax administration. More problematically, customs revenues have stayed fairly constant at around 1 percent during the same period. The topography of the country's borders, high unemployment, and the government's poor supervisory capacity have made Bolivia a prime target for contraband, which has grown dramatically during the 1 990s: in 1990, the Bolivian government lost about $109 million per year in revenue; by 1997, the estimated annual loss had risen to about $430 million. 20 19 Bolivia Countrv Assistance Strategy, 1997, pp.19-20. 20 Jose Luis Valencia A. and Justo Alcides Casas, Contrabando e informalidad en la economia boliviana, (Opiniones y Analisis, 1998, p. 68). 9 1.3.4. According to the results of the survey commissioned for this Review, both the Internal Revenue Agency and the Customs Agency (before the recent change in the organizational structure and management) are highly politicized and prone to corruption. The respondents from the Internal Revenue estimated that 75 percent of hiring is done for political reasons. Customs was ranked worst in terms of officials' perception of levels of corruption by agency. Customs was also ranked worst by officials who were asked about the existence of a career path within the agency; Customs appears to have one of the highest turnover rates in the Bolivian public sector. These findings support anecdotal evidence, gathered in interviews and focus groups, of Customs posts being bought by people who wanted access to illegal sources of revenue for personal enrichment. An EDI-financed 1998 survey of private sector firms also showed the business community's concerns about high levels of corruption in Customs, particularly in the release of goods for import.21 1.3.5. Human development. Investment in human capital is another area that provides evidence of the pernicious effects of politicization on public service delivery. In Bolivia, as in many countries, the Ministry of Education provides fertile ground for distribution of patronage. Respondents to the public officials survey reported that nearly 50 percent of staff are hired for political reasons in the Ministry of Education. Teaching staff are hired at the prefecture level, where the survey shows politicization to be even more rampant. When teaching posts are distributed based primarily on political affiliation rather than professional qualifications, the quality of education is directly threatened. Moreover, teaching is a notoriously difficult activity for the government to monitor on a daily basis, making careful enforcement of quality standards nearly impossible. 1.3.6. Bolivia initiated education reform efforts in the early 1990s and, with donor assistance, the reform initiative has survived three administrations. Despite evidence of considerable politicization in some parts of the Ministry of Education, a technical team in the Ministry has worked on reform design and implementation without political interference for much of the decade. Still, the harmful effects of politicized personnel and absence of effective oversight are reflected in the poor quality of Bolivian education. Despite increased budget allocations for education and some improvement in enrollment rates, the GOB estimated in 1992 that "at least 55 percent of the adult population are functionally illiterate, including many who finished primary and even secondary school."22 More recent test scores in language and mathematics place Bolivian students in the third and fourth grades below the regional average. Repetition rates are also 2t Ascensio Villegas, Arcadio Morales y Neil Andersson, Corruption in the Public Services: the view of the private sector (First National Integrity Survey), CIET International, March 1998. In July 1999, Congress passed the Ley de Aduanas. This law recreates the Customs Agency as an autonomous entity, formally protected from political interference. Many customs agents were dismissed, and merit-based hiring procedures are required under the new law. The government hopes to improve customs administration by creating an agency staffed by highly-qualified professionals who are insulated from pressures exerted by political parties and other actors to participate in corrupt practices. However, the challenge is daunting. Powerful groups stand to lose from this reform. Political parties use the Customs Agency as a key source of party financing as well as a source for distributing highly attractive patronage positions to their supporters. 22 Bolivia 1992 Public Sector Investment Program Review, 75 10 disturbingly high at 86.6 percent. A 1997 national survey23 on academic achievement (SIMECAL) found significant differences in academic performance of students in public schools versus those administered by NGOs and the Church (convenio schools). After controlling for students' socioeconomic background, the 1997 study found that 44 percent of third and sixth graders enrolled in public schools were "at risk" in their academic attainment, compared to 26 percent of convenio students. 1.3.7. Basic health services are also woefully inadequate. Infant, child and maternal mortality rates are among the highest in Latin America. A 1992 World Bank report reported poor quality and low utilization of maternal child health services: nationwide, only 50 percent of pregnant women received some prenatal care, with many regions reporting coverage rates as low as 20-30 percent.24 By 1998, the percentage of births nationwide with some prenatal care by trained medical personnel had risen to 65 percent, 25 though the percentage of births that occurred in medical facilities was only 52 percent. Vaccination rates for children have remained stagnant or even worsened since 1989: the percentage of children with complete vaccinations for measles dropped from 57.5 percent in 1989 to 50.8 percent in 1998; for polio, the percentage dropped from 47.5 in 1994 to 39.1 in 1998.26 These averages mask significant discrepancies in health service delivery outcomes among socioeconomic groups. Not surprisingly, the poorest quintile of the population suffers much inferior health conditions than the richest quintile. 1.3.8. The Ministry of Health was not included in the public officials survey, but a wealth of anecdotes from Bolivian officials offer evidence that political interference adversely affects the quality and coverage of health services. Both government employees and international aid donors also told us of the problems caused by political interference in the implementation of donor-funded projects meant to address deficiencies in basic health services. A CGR audit of an IDB-funded Basic Health Services project provides a concrete example. Auditors found that the project had had six Project Directors over a seven-year period, and at times the Project Director post was left vacant for an extended period and had to be performed by the Principal Consultant. The Project Director is named by the Minister of Health, and the CGR report found that some Ministers had replaced the Project Director frequently, in some cases more than once per year. This interfered with project implementation, leading to noncompliance with IDB requirements; lack of timely legal action to resolve breach of contract by various firms hired under the project; low morale among project employees; and loss of credibility of the project among the community it was designed to serve.27 1.3.9. Roads. When responsibility for regional road construction and maintenance was 28 deconcentrated to prefectures in 1996, the quality of service delivery in the roads sector declined noticeably, partly due to the loss of highly qualified engineers from the SNC 23 Study cited in Bolivia 1999 Public Expenditure Review, 73. 24 Bolivia 1992 Public Sector Investment Program Review, p. 84. 25 Bolivia 1999 Public Expenditure Review, p. 104. 26 Ibid. 27 Auditoria Especial sobre la cooperaci6n internacional recibida por la Replblica de Bolivia al 30 de noviembre de 1998, Contraloria General de la Republica. 28 This decision was reversed by the current government, and the SNC was recreated in late 1997. 11 (see Box 1.1). The percentage of roads that received some maintenance activity fell dramatically from 30 percent in 1995 to eight percent in 1996, resulting in a marked deterioration of the already fragile roads network. In 1997, some departments executed as little as 19 percent of their required maintenance budgets.29 While the SNC has been reconstituted as a centralized agency, politicization of its workforce remains a problem. In the public officials survey, the SNC scored poorly in accomplishment of mission, corruption, and enforcement of regulations. Over 75 percent of employees are perceived to have been hired based on political affiliation. 1.3.10. Water and sanitation. Bolivia's water and sanitation sectors account for the next largest share of public investment after roads. Safe water and sanitation services have important positive externalities for public health. During the 1990s, Bolivia has made significant progress in improving access to piped water supply. At the beginning of the decade, water and sewerage service levels in Bolivia were the second lowest in the western hemisphere and the lowest in Latin America. About 45 percent of the population had access to piped water supply, while about 30 percent of the urban population was connected to a sewerage system. By 1998, 76 percent of the urban population had access to piped water supply, which is about average for the region. Access to sewerage connections remained low at about 32 percent.30 In urban areas, local governments have been responsible for water and sanitation services since the 1960s. The municipal water company of La Paz, SAMAPA, was a good example of a politicized, overstaffed and inefficient public utility. Although SAMAPA's financial situation improved somewhat during the late 1980s when it raised tariffs and was finally able to cover its own operating costs, it failed to invest sufficient funds in La Paz's water and sewerage systern, leading to poor quality services. A 1990 World Bank project to improve urban water and sewerage supplies encountered "continued political intervention in ... management and turnover in key staff [which] led to delays in the implementation of the projects."31 In 1997, the governnent decided to privatize SAMAPA and awarded the concession to a private water company (Lyonnais des Eaux). Since then, investment has progressed rapidly and the quality of water supply has improved. Box 1.1 SNC: the decline of a service-delivery institution Since its creation as a semi-autonomous agency in 1961, the SNC was widely recognized as a sound organization. Inadequate funds limited the agency's capacity to carry out urgently needed road construction, but the agency mnanaged to maintain the existing road network. Eventually. 29 Bolivia 1999 Public Expenditure Review, p. 155. 30 Bolivia 1990 Major Cities Water and Sewerage Rehabilitation Project Implementation Completion Report, 1998. The GOB has undertaken institutional reforms in the sector. The state is now cast in a regulatory role, while different service provision arrangements are emerging. The 1994 SIRESE law created the Water Superintendency, whose mandate is to regulate public water supply and sanitation services. Under the Popular Participation Law and the Law on Municipalities, ownership and operational responsibility for water and sanitation services were transferred to local governments. Municipalities have chosen one of three institutional arrangements for service delivery: cooperatives, private concession, and autonomous municipal water companies. 3 Ibid, p. vii. 12 inflow of s-funsdoiv therblenytoorxtend itsef w thad thnreein costrvction sacfth weakneadei in the Sset performance in maintect1994, onsclrly aout 21 percn o Bolivia's road network wasS inf goo condition.t0g tiX v: accuse th. Crofscttn reprobding toletristzatednies and dcrimintiong agane ob arvticuna pafthe common causes. Politicization of almost the entire public administration appears to be having secrious consequences for the institutional capacities of the state bureaucracy to provide quality public services. One of the common problems sdo the sectors is coffuption. In a recent public opinion poll, 45 percent of respondents (1304 cases) thought there was "a lot of corruption" among public employees, and another 26 percent believed "some corruption" took plaee in the government bureaucracy. 12 Furthermore, 87 percent of respondents in the public officials' survey conducted for this Review reported that corruption was a very serious problem in the public sector, and another 10 percent admitted it of budget serious problem. What is worse, only 38 percent of the respondents reported that corrupt behaviors are either always or frequently 32 Encuestas y Estudios, March 1999. 13 denounced, indicating that a large majority of acts of impunity go unpunished within the public sector.33 1.3.12. Another problem that affects the bureaucratic performance throughout the public sector is the extremely high incidence ofjob insecurity, as measured by excessively high rates of personnel turnover which adversely affects organizational performance. Our survey data shows that only 33 percent of public employees were working in the same agency as in 1995, and 58 percent were working in the public sector in 1995. Since 1997, the year of the last election, around one-third of public employees are estimated to have left the public sector and only 52 percent are still working in the same agency as they were in 1997. Only 9 percent of current public employees believe they will be working in the same agency in five years. Also, 53.6 percent of public employees reported as likely or very likely the possibility of losing their jobs, and 74 percent believe that the high staff turnover is a "serious" or "very serious" problem that affects organizational efficiency. 1.3.13. The GOB has long recognized the importance of improving the quality of public sector employees, and administrations have made successive attempts at creating a professional, merit-based career civil service system since the 1950s. But program after program has succumbed to the strongly politicized/clientelistic culture that pervades Bolivia's public sector. The most recent effort (the Programa de Servicio Civil or PSC) was slow to be implemented, and in recent years most PSC employees have left the public sector. Although PSC employees were supposed to be protected from arbitrary dismissal, we learned of a number of cases in which civil servants hired under the PSC project were subject to intense pressure from upper management to resign from their posts, which is tantamount to being forcibly removed. 1.3.14. One potential contribution left behind by the PSC was the transformation of the project execution unit (UESC) to a semi-autonomous personmel administration agency called Servicio Nacional de Administraci6n de Personal (SNAP). But this agency has also become politicized in the last 18 months, particularly since its mandate was broadened to oversee implementation of the Normas Basicas de Administraci6n de Personal in January 1999. Between April 1998 and late 1999 around 75 percent of the original 22 staff were replaced, although the reasons for these staff changes remain obscure.34 The replacement of those employees, and the hiring of additional staff, has been done in ways that violate the Normas Basicas. Quite a few temporary (interino) employees have been retained for almost one year, although the Normas Basicas stipulate a maximum of 90 days for temporary contracts.35 Most other employees have been hired by "direct invitation," without the open recruitment and selection process mandated by 33 There is a distinct pattern in the self-assessment of the seriousness of corruption in particular agencies. Not surprisingly respondents from technocratic agencies including CGR, BCB, the Superintendencies of Banks and Telecom, and UDAPE tend to consider that corruption is not a very serious problem within the agencies. Respondents from Aduanas, Impuestos Intemos and the Ministerio de Trabajo, in contrast, consider corruption a very serious problem in their agencies. 34 As we describe in Chapter 2, agency heads often pressure employees to resign "voluntarily." 35 After 90 days, the agency must hold an open selection process to hire a permanent employee, or close the temporary post. 14 the Normas Basicas. According to one count, only three posts have been filled by the public recruitment method described in the Normas. 1.3.15. While it is hard to show a direct impact of these informal practices on public sector performance, it is likely that SNAP's politicization and its failure over the past year to set an example by implementing the personnel system set out in the Normas Basicas, threatens the credibility of the GOB's current efforts at creating a merit-based personnel system. Despite the creation of a Civil Service Superintendency to oversee implementation of the new Estatuto del Funcionario Puiblico, SNAP will continue to exist as a policy-making entity. Currently, it is staffed almost entirely with people who were chosen based on political affiliation rather than technical competence, which can only bode badly for the quality of public personnel policy in the future. 1.4. The Current Reform Agenda 1.4.1. The current Government is strongly committed to succeeding with state reform where others have failed. The main vehicle through which the Government is addressing these challenges is the National Integrity Plan (NIP), whose objectives are: "to prevent and combat corruption through the strengthening of state entities' institutionality, in order to produce more efficient services and more effective results for the citizenry; to promote complete legal security by effectively protecting citizens' rights through an accessible and transparent judicial system."36 NIP proposes a very sensible medium-term approach to fight corruption by strengthening the range of public sector institutions in a comprehensive fashion, instead of focusing on short-term measures with limited prospects for sustainable improvement in the public sector's institutional performance. If fully implemented, the NIP will likely have significant impacts not only on reducing corruption but on building robust public sector institutions that contribute positively to poverty reduction and economic development. 1.4.2. To achieve this overall goal, the Plan focuses on three specific objectives: improvement of the judiciary and legal framework; modernization of public administration; and development of effective mechanisms to fight corruption. One of the primary instruments for implementing the Plan's state modernization and anti-corruption components is the new Institutional Reform Project (IRP).37 This project aims to turn the state into "a national resource capable of enhancing the general welfare rather than a source of corruption, private gain, or an impediment to development."38 The IRP supports horizontal reforms in personnel management (i.e., civil service reform), budgeting by results, national integrity and a national performance evaluation system for both personnel and organizations. These measures will be implemented in selected pilot agencies which will go through major organizational restructuring, while simultaneously introducing the above-mentioned horizontal reform measures. 36 Plan Nacional de Integridad, 1998, p. 20. 37 A second Judicial Reform Project is also being considered to support the third pillar of the National Integrity Plan, but this report will touch on institutional issues of the judiciary only as they relate to the performance of the core public sector. 38 Bolivia 1999 Institutional Reform Project, Project Appraisal Document, p. 2. 15 1.4.3. The guiding philosophy of the IRP is consistent with the normative centralization and operational decentralization embodied in SAFCO, whereby line agencies are granted considerable discretion in managing themselves and fulfilling their missions. The project will begin with a diagnosis by each pilot agency of its mission, structure and function, culminating in the design of an Organizational Restructuring Agreement (ORA) which will specify concrete reform actions to be undertaken. The ORAs will "serve as transition instruments, acclimating the agencies to more rigorous development, achievement, and evaluation of performance results, and serve as a yardstick to trigger disbursements from the IRP...."3 9 As an ambitious ten-year program, eventual success of the IRP will depend on successive governments' sustained commitment to the spirit of the reform. 1.5. Summary 1.5.1 Bolivia has come a long way from being the poorest country with one of the most troubled political economies in South America to one of the top reformers in the region. Despite the notable progress in a number of areas including macroeconomic reform, democratic transition/consolidation, and certain structural and institutional reforms, the country continues to suffer from a high incidence of poverty, exacerbated by the poor quality of public services and the state's inability to create an environment that is truly conducive to private sector development. Poor governance manifested in systemic corruption and widespread politicization of the public bureaucracy is one of the main causes of the frustrated private sector and social development in Bolivia. Anecdotes as well as existing evaluations point to persistent patterns of politicization affecting government performance in a variety of sectors. 1.5.2 Although Bolivia has attempted to reform and modernize the central public administration in order to curb corruption and improve efficiency and effectiveness of its service delivery, these efforts have so far failed to produce visible results. This continued poor performance suggests fundamental weaknesses in the public sector institutions, which in turn would result from distorted incentives within the public bureaucracy. As will be detailed in the next chapter, Bolivia's public sector suffers from what can be characterized as a problem of bureaucratic informality, where formal institutions (understood as rules and regulations that govern public sector behaviors, as well as organizational arrangements that support their implementation) do not function as they are intended to but are circumvented or subverted by informal norms of behaviors. Informality makes formal accountability of the government virtually meaningless, and deprives the bureaucracy of its organizational efficiency. 1.5.3 The Government's current reform agenda addresses broad sets of institutional issues, and its success will depend, at least partially, on addressing these informality problems. To support the Government's reform agenda expressed in the National Integrity Plan and operationalized in the Institutional Reform Project, this Institutional 3 Ibid, 8. 16 and Governance Review attempts to document the problem of bureaucratic informality, analyze its probable causes, and offer a set of recommendations on a realistic reform implementation strategy. 17 2. INSTITUTIONAL CHARACTERISTICS OF THE BOLIVIAN PUBLIC ADMINISTRATION: IMMEDIATE CAUSES OF POOR PUBLIC SECTOR PERFORMANCE 2.1. Absence of a Functioning Bureaucracy 2.1.1. The low quality of Bolivia's public sector institutional environment and the consequent weak institutional performance described in the previous chapter stems from the near absence of a well-functioning bureaucracy capable of performing public tasks with impartiality, predictability and accountability. Low administrative capacities due to a high rate of personnel turnover, frequent and pervasive political interference in public management, weakness and/or non-compliance of formal administrative procedures and control, and systemic corruption are some of the problematic characteristics of the Bolivian bureaucracy. Without fully functioning internal core systems and procedures and adequate accountability mechanisms, it is unlikely that satisfactory bureaucratic performance will obtain. In Bolivia, as in many other developing countries: "Real bureaucracy is in scarcity, not excess. It is the absence of bureaucratic structures that leads to the utilitarian nightmare of the state as a collection of self- interested incumbents using their offices for purposes of individual maximization. Ineffective states are characterized precisely by the lack of predictable, rule- bound, bureaucratic norms and relations within the state apparatus.'-0 2.1.2. This "lack of predictable, rule-bound, bureaucratic norms and relations within the state apparatus" keeps the Bolivian state from functioning as a coherent organization in pursuit of a common set of policy objectives with a reasonable level of internal coordination. When internal procedures and control to ensure accountability are ineffective, pursuit of efficiency and client orientation that has captured the advocates of public sector reform in OECD countries remains an elusive dream. An effective public bureaucracy first requires internal institutional cohesion supported by adequate policy and legal frameworks, organizational structures and sound management practices. Ilowever, the effectiveness of formal institutional mechanisms to guide and control public sector operations hinges fundamentally on informal behaviors within the public sector, which reflect social and political as well as organizational factors surrounding individual public officials. As will be detailed in the rest of the report, the lack of rule- boundedness stems not so much from the absence of rules as from the incompatibility between the existing formal rules and procedures that prescribe ideal, expected behavior of public officials on the one hand, and informal rules and norms that govern their actual behavior on the other. This Review's main finding is that the political dynamics in play, especially the logic of coalition politics and patronage, define these informal rules in the Bolivian bureaucracy, often producing undesirable behavioral patterns that are detrimental to public sector performance. 40 Peter Evans, EmbeddedAutonomy: State & Industrial Transformation (Princeton University Press, 1995, p. 71). 18 2.1.3. This chapter will document the patterns of bureaucratic informality in Bolivia and point out some of the immediate causes, which are attributable to weak formal mechanisms for rule enforcement. In the following chapter, the Review will explore deeper causes of informality that arise from the nature of the political environment and that structure incentives within the public sector. First, however, in order to shed light on the problem of bureaucratic informality, the next section attempts to understand the "ideal" formal structure and procedures of the Bolivian state as envisioned in the existing legal and organizational framework. The report will then present illustrations of informal behaviors that undermine performance of the Bolivian public administration. The following sections will document systematic weaknesses of institutional mechanisms for rule enforcement within the Bolivian state, which are important immediate causes of informality and weak accountability. 2.2. The Legal Framework and the Organizational Structure of the Bolivian State 2.2.1. The Constitutional structure. The Bolivian Constitution defines the country's political structure as a unitary state, consisting of two levels of autonomous government, national and municipal, and three levels of public administration, national (central), departmental, and municipal. Municipal governments are autonomous within their territories, while departmental administration is a deconcentrated unit of government, whose chief executive, Prefect, is appointed by the President of the Republic. However, the governing structure of the departments also includes representation of municipalities at the departmental councils, whose members are selected from among the municipal councils in the same departments. The current structure and role of the departmental councils somewhat blur the role of the department as a deconcentrated arm of the central government. 2.2.2. The central government is one part of the non-financial public sector. It constitutes the major part in terms of aggregate budget funding but the minor part in terms of employment. Central government has some 16,500 employees, of which 80 percent are in the central administration and 20 percent in the decentralized agencies. By contrast, territorial government has over 130,000 employees, of which 88 percent are in the prefectures, many of whom had been absorbed into the prefectures from the former regional development corporations. The remaining non-privatized public enterprises have some 9,000 employees and the financial public sector around 1,000. The Central Government is then divided between the central administration and decentralized agencies with different legal bases and funding arrangements (Table 2.1). The figures 2.1 and 2.2 summarize the architecture of the Bolivian State. 19 Table 2.1 Legal and Funding Distinctions between Central Administration and Decentralized Agencies Entities Legal nature Funding sources Central Ministries Created by LOPE, Budget dependent - nil Administration Servicios core part of the state or minor revenue Nacionales No legal identity earnings separate from it Assets are the general Any liabilities property of the state - generated accrue to cannot be sold by the the state agency Decentralized Funds Legally distinct Assets are the general Agencies Superintendencias Created by specific property of the state - National Institutes enabling legislation cannot be sold by the agency Largely budget dependent - some revenue earnings Figure 2.1 Central Government within the Non-financial Public Sector sector~~~~~~~~~~~~~ prefectures o territorial enerr e Iwgovernment enepI social secuirity institution'si non-financial publc sector .cnrlgvrmn naioa enepIse 20 Figure 2.2 Outline Architecture of Central Government miscellasidneoyusd m inistries centralL_ || 2 admvinistratioe o te Oranzo onon-execmtive" | agencies l superintendencias| decentralized a e eagencies 1lnational institutes| 2.2.3. LOPE. Within this overall vision of the state structure, the 1997 Law of the Organization of the Executive Power (LOPE) assigns to central ministries "normative" (i.e., policy-making) roles while delegating responsibilities for policy implementation to several types of decentralized agencies and to departments. Since the Law of Popular Participation ( Fo994), municipalities also carry out significant portions of public investment execution.41 Through the recent capitalization/privatization, the governent has also divested a number of public enterprises in areas where state involvement entailed efficiency losses. 4 1 Further work is necessary to clarify and streamline divisions of responsibilities and fianctions across levels of government. The issues related to inter-governmental fiscal and institutional relations are not treated in this report. For details, see "A Strategic View of Decentralization in Bolivia" in the Bolivia 1999 Public Expenditure Review. 21 2.2.4. SAFCO Law. The most important piece of legislation that governs Bolivia's public administration is the Law 1178 of Financial Management and Control (Sistema de Administraci6n Financiera y Control Gubernamental, or SAFCO). SAFCO was conceived as a "law of principles" ("ley principista "), which would lay out the general direction of public sector operations, but which would require a considerable amount of secondary regulations to implement them.42 Some of SAFCO's primary objectives were: * to establish management systems, covering all aspects of public expenditure (from programming of objectives and the activities required to achieve them, to the development of indicators to evaluate results), that would create some degree of order in the public sector; * to reinstate the emphasis on individual responsibility/accountability by public employees; * to eliminate pre-control of expenditures and create a results-oriented management culture that would allow the Controller General of the Republic (CGR) to audit public expenditures with a view towards assessing what had been accomplished through spending of public money, instead of taking a narrow view limited to whether or not agencies had complied with procedures. 2.2.5. The key philosophy of the SAFCO systems is one of "normative centralization and operational decentralization," whereby each organization is to be held accountable for implementing policies, while the role of the center is to ensure adequacy of the procedures used by agencies, and to evaluate results. Operationalization of this vision received additional impetus with popular participation/administrative decentralization, and the creation of autonomous private sector regulatory bodies, which clearly separated from the central ministries significant portions of service delivery and regulatory functions. Under the SAFCO philosophy, the role of central ministries is seen primarily as one of policy-making, standard-setting and coordination. 2.2.6. Each of the eight SAFCO systems (operational programming, administrative organization, budget, personnel administration, procurement, treasury and public credit, integrated accounting, government control) required a specific subsidiary regulation (Basic Norms) that detailed how the Law would be implemented. Ministry of Finance (MOF) was designated as a "rector institution" to oversee the implementation of the first seven systems that were concerned with different spheres of public administration, while the CGR was made responsible for the Government Control system. As rector institutions, both had the responsibility to develop and emit basic norms for the eight SAFCO systems, and to approve specific procedures to be developed by each governmental entity throughout the public sector. 2.2.7. Delayed and uneven implementation. Although SAFCO applies a very modern conception of decentralized, results-oriented public management to the entire public sector, in practice, little sustained effort has been made to enforce the systems, and the 42 The logic behind this approach was to combat Bolivia's legalistic tradition, which tended to produce highly detailed laws that quickly became obsolete and were later very difficult to alter. The idea was "to draft a law that would last." Interview with a former member of the SAFCO reform team. 22 delay in passing supporting regulations created considerable skepticism. Even before SAFCO, agencies tended to operate quite independently of central regulation and coordination. However, the delay in serious implementation of SAFCO's systems has left less developed the corresponding question of agencies' and managers' accountability for results. Entities' adoption of internal regulations for the different SAFCO systems is spotty. For example, only around 12 public entities (all are small institutes and a few municipalities, none are ministries) have approved internal regulations for personnel administration; even the Ministry of Finance only has a "project" to work on the internal regulations. Few public entities have adopted internal regulations for the integrated accounting and treasury/public credit systems. 2.2.8. The delays in issuing enabling regulations for SAFCO (nearly seven years after the promulgation of the Law itself) suggest that leaders in both the Executive and Legislative branches, as well as public managers, were satisfied with the legitimate autonomy that it provided, but resisted the concomitant systems for evaluating results. Although our interviews with public officials indicate broad acceptance of SAFCO as a guiding framework that is here to stay, particularly strong support for it does not seem to have emerged among rank-and-file officials, either. In fact, our survey of 692 public officials in 15 government agencies found some ambiguity in public officials' attitudes toward SAFCO. A little over 50 percent of the respondents thought that SAFCO did not encourage public officials to focus on results and that instead the Law made them afraid of possible prosecution ("juicio"). The same percentage (51.7 percent) believed that SAFCO did not improve their agency's efficiency, although 56.9 percent thought it had an effect on reducing corruption. 2.2.9. These sample-wide results of the perceived impacts of SAFCO indicate lack of internalization of what SAFCO is about within the public sector, which is partly a reflection of the incomplete implementation of the Law. A more disaggregated look reveals an interesting pattern of variations across types of agencies. As Figure 2.1 shows, respondents from four different types of public agencies, "autonomous" that include the Central Bank, the Superintendencies and the CGR, ministries, servicios/institutos nacionales, which execute sectoral policies with some degree of organizational autonomy from parent ministries, and prefecturas, perceive impacts of the SAFCO Law quite differently. In general, the respondents from the technocratic autonomous agencies point out positive impacts of the Law in terms of promoting results orientation and efficiency while controlling corruption. In contrast, the prefecture staff in the sample tend to consider SAFCO as not having significant impact on improving public sector management. 23 Figure 2.3 Perceived impact of the SAFCO Law, by types of agencies (Percentage of respondents answering "Yes" to the questions "Does SAFCO ...?") 113Autonomous *Ministry DServicio OPrefectura | 8 0. 0........ 70,0~~~~~~~~~~~~~~~~~~~0 70.0 7 4 0 s0 0 - _ _ ___ 50.0 40.0 30.0 10.0 0.0 encourage results focus? cause fear of promote efficiency? control corruption? prosecution? 2.2.10. Overall, the institutional arrangements set in train by SAFCO are intended to focus largely on attainment of objectives, although compliance with a basic set of management systems and procedures is also important. We already have ample evidence that the ultimate objectives-based goal of the SAFCO Law has yet to be achieved. Furthermore, the survey results cited above indicate some contradictory perceptions of the effects of SAFCO among public officials. This is a sign that there is still some way to go before the SAFCO culture is fully instilled in the public sector and has positive impacts on public officials' behaviors. A common explanation provided is lack of commitment by successive governments to implement the SAFCO systems. Indeed, even a casual look at the history of Bolivia's efforts at civil service reform clearly indicates that weak commitment has been one of the primary problems. But political commitment is far from a sufficient pre-condition for reform. The feasibility of SAFCO reform requires a robust set of formal institutions and effective enforcement mechanisms; without these, the government would find it difficult to implement the SAFCO-related reforms even with the best of political intentions. The persistence of informal practices is a clear sign that most of the formal rules are not strong, and that enforceability is weak. As the next sections document, there are systematic patterns of informality in two core areas of public administration - personnel and expenditure management - which institutional reform should address in order to improve performance of the Bolivian public sector. 24 2.3. Informality in Public Personnel Management 2.3.1. As detailed in the next chapter, a major source of bureaucratic informality in Bolivia is political interference. Public personnel management suffers particularly from practices of clientelism. Through in-depth interviews and focus groups, we obtained considerable anecdotal evidence on informal mechanisms used by political parties to influence bureaucrats' behavior in two ways. Some groups of public officials apparently receive positive incentives ("carrots") through informal pay practices, which may be financed through donor-funded project budget, or through discretionary use of non-wage budgets such as travel and training. More often, parties use negative incentives ("sticks") to ensure public officials' compliance with party objectives. 2.3.2. Pay practices. One area in which significant amounts of informality are often observed, not only in Bolivia but also in many other developing countries, relates to use of unofficial, discretionary funds to compensate public sector managers. Topping up salaries of high-level public sector managers is often necessary to recruit and retain competent personnel in any institutional setting. However, bureaucracies in many developing countries suffer from extremely low salary levels with limited budgetary flexibility to compensate managers at a competitive level. A result is a creative use of discretionary, or sometimes even extra-budgetary funds (e.g., donor funding) to top up public officials' salaries. Bolivia is no exception in this regard. 2.3.3. In Bolivia, the aggregate budgeted public sector wage bill is Bs5.36 billion.43 However, from persistent anecdotal evidence, there is reason to think that there is significant usage of other budget line items for salary top ups.44 It is a cautious assumption that some Bs363.78 million is added to the total wage bill for competitive entry staff, suggesting an aggregate total of around Bs5.7 billion. In addition, anecdotal evidence suggests that salaries in the superior and directivo classes are supplemented by substantial amounts of additional remuneration from other funding sources. Finally, although the proportion of externally-funded consultants is not known, it probably represents a very significant additional flow of funds for remuneration. 2.3.4. The variation in supplementary pay between Ministries is considerable. The existing proposed pay reforms (staffing cuts of 27 percent in central government and wage increases to bring public sector salaries up from 28 percent of earnings in the private sector, to about 64 percent) will not fully address these inter-ministerial gaps. It appears that many senior level officials (Director General and above) are earning as much as (or even more than) private sector counterparts, when this discretionary funding is factored in. Reportedly, there is a strong political weighting in the allocation of these discretionary allowances and agency heads may be using these allowances as incentives for subordinates to comply with their directives. In fact, as merit protection increases job security for public employees, agency heads may logically be te.mpted to utilize the 43 1999 National Budget of Bolivia. 44 Specifically, line items 121 (personal eventual), 252 (estudios e investigaciones), 221 (pasajes), 222 (vidticos), and 460 (estudios y proyectos de inversi6n). 25 "carrot" of discretionary allowances even more than before, as the "stick" of job loss becomes less available. 2.3.5. Discretionary rewards are 'horizontal' decompression, in that employees at the same grade undertaking the same job can receive significantly different rewards. OECD experiences are that such discretionary allowances are used with considerable care, even in robust institutional environments, as they are likely to reduce transparency and credibility in motivating staff, for at least two reasons. First, they are unlikely to be related in any meaningful sense to performance. Second, extreme horizontal decompression encourages staff to comply unquestioningly with the agency head who can alter the allocation of these allowances. This can be a major contribution to corruption. 2.3.6. At face value the current vertical compression ratios between relatively low pay range staff (tecnico) and the highest pay (Minister) are, taking the examples of M. de Agricultura y Ganaderia, and M. de Justicia y Derechos Humanos, 1:12 and 1:7 respectively. These ratios worsen if an assumption is made that large amounts of additional resources are distributed to top-level officials. According to budget figures, the vertical compression ratio for directivo staff in relation to tecnico is around 1:3. However, the consistent story from officials is that many directivos also receive salary supplements, so it is likely that large differences in pay also exist between directivos and lower-level staff. 2.3.7. There is no 'right' answer to the level of vertical decompression, but the typical OECD decompression is around 1:8. There is every reason to assume that decompression radically in excess of this leads to cynicism and a perception that promotion is inconceivable. Given the extreme decompression that is certainly present in some Ministries, it is far from clear how this will be achieved in all agencies, particularly if the use of additional budget resources for superior salary supplements continues. 2.3.8. Informal methods ofpolitical control of the bureaucracy. In any public sector, politicians in both the Executive and Legislative branches have legitimate reasons to want to control behaviors of public officials, one of whose main responsibilities is to implement policies that the politicians formulate. However, in countries including Bolivia, where formal institutional mechanisms for controlling and holding the bureaucracy accountable are very weak, there are strong incentives for politicians to use informal methods of political control. This incentive gives rise to widespread use of patronage appointments. These in turn can take two forms: one is to appoint persons of confianza in key bureaucratic posts in order to ensure a certain level of performance; and the other is to use public posts as means to reward political supporters regardless of the persons' qualifications for the job. Clearly the latter pattern of patronage appointments is more problematic from the point of view of public sector efficiency. However, even the former pattern whereby competent professionals are appointed to key bureaucratic posts could also pose a problem of formal accountability as long these persons owe their 26 allegiance to the political patrons who appointed them rather than to the set of impersonal rules, which govern bureaucratic norms including performance objectives.45 2.3.9. In general, political parties use two types of incentives to ensure bureaucrats' compliance with party objectives while in power. In the face of extremely low salary figures, the "carrot" of additional income is essential for survival. So, salary supplements from various sources of discretionary funds as discussed above, as well as assignment to posts which enjoy access to rents (licenses, inspections, procurement, etc.), provide powerful positive incentives to employees. Most positive incentives go to loyal party members or people who have "invested" in the party by making large campaign contributions. Other prominent party members occupy posts which do not have access to rents, but are valuable in other ways (such as posts with access to information: Statistics Unit, Personnel Unit, etc.). These posts constitute a valuable career move (within the party) for their occupants. 2.3.10. "Sticks" appear to be more common than "carrots" in Bolivia, and are used to oblige public employees to abide by the party's informal rules. Given the precarious economic situation of many Bolivians, the threat of job loss is taken very seriously, and negative incentives play upon this fear. SAFCO and the Basic Norms of personnel management provide reasonable protection from arbitrary dismissal. But parties have found informal tactics so effective that it is unusual to have to initiate a formal dismissal procedure against an employee who refuses to resign. 2.3. 11. Informal tactics used to pressure people to leave their posts include: * request for employee's resignation combined with the threat of a poor performance appraisal which will go on the employee's record if the resignation is not presented * "freezing out" employees who refuse to leave by diverting work and excluding them from the organization's functioning * prolonged use of "temporary" contracts to fill posts * physical threats * threat of blackmail (for example, "cartas defidelidad": blank letters that are signed by a person in a sensitive post and can be used against the person if necessary). 2.3.12. In contrast to the lack of respect for formal rules and contracts, the parties monitor compliance with informal contracts closely. Parties rely on the bureaucracy to provide policy support, as well as financial backing for party activities. It is widely asserted that parties require public employees to pay contributions to the party that controls the agency in which they are employed. In the public officials' survey, nearly 30 percent of the respondents from ministries, national institutes and prefectures admitted to having made some amount of contribution to a political party.46 The contributions may be taken as an 4 This point is developed further in the next chapter. 46 This figure is certainly lower than what was implied in focus groups and confidential individual interviews we conducted during our missions. But there are good reasons to expect respondents to under- 27 automatic deduction from the paycheck, or a party may organize a monthly collection. During electoral campaigns, employees are often obliged to purchase multiple tickets to campaign events. In addition to requiring financial contributions, many parties also demand that public employees offer their time (usually attending campaign events). C'ompliaince is ensured through the distribution at the workplace of individualized passes, which must be returned at the event venue and the employee' s name is checked off an attendance list. 2.3.13. Effective monitoring of these contributions is one of the parties' key concerns: they must ensure that they are getting a valuable "payback" in exchange for securing jobs for people. To reduce monitoring costs, parties organize internally into smaller committees which are responsible for various areas (finance and fundraising; collecting intelligence on other parties' activities; publishing and distributing position documents; campaign organization; etc.). In this way, they are able to control their members and know who is contributing to the party, how much, etc. At the same time, it is common knowledge in Bolivia which party controls which ministry/vice ministry; in this way, each party can estimate how much of the "pie" other parties are getting. 2.3.14. Politics also interferes in the daily operations of public agencies, as parties sometimes exert pressure on officials as they carry out their duties. In this respect, 88 percent of the respondents from "autonomous" agencies never or only rarely received political pressure in their duties, while 40 percent of the respondents from the Ministries, 2.24 National Institutes, and Prefectures reported either very frequent or somewhat frequent political pressure. Fifty-two percent of the respondents from "non-autonomous" agencies replied that if one did not comply with political pressure, he/she would likely lose the job. 2.3.15. We would expect staff in the merit-based Civil Service Program (PSC) established in 1992 to enjoy greater protection from the kinds of dysfunctional informal practices described above. However, a focus group session with seven PSC employees (time in PSC ranged from 1 - 4.5 years) elicited some illuminating information, though anecdotal, about sub-optimal treatments they often receive. Some of the most common frustrations expressed by PSC focus group participants include: * Political pressure to leave their posts, especially when there is a change in government. Both Unidad Ejecutora del Servicio Civil (UESC) and now Servicio Nacional de Administraci6n de Personal (SNAP) are much too weak to provide any support for PSC employees. Moreover, SNAP itself has become highly politicized over the past two years. Thus, when PSC staff are subject to unfair treatments due to political reasons, there is no effective official recourse for redressing the situation. report this sort of behavior because of the sensitivity of the topic. We are therefore confident that at least 30 percent of public employees actually make these salary contributions to parties, and that the real figure is likely to be higher. Significantly, none of the respondents from "technocratic agencies" in our sample (CGR, BCB, the two Superintendencies, UDAPE, and INE) reported party contribution. 28 * Resentment by non-PSC employees (normally because of salary differences) leads to tense working relationships. For example, many of the focus group participants said that their paychecks are the last to be processed and they are often paid much later than their non-PSC co-workers; others gave examples of particularly strict application of rules in cases affecting PSC employees, whereas exceptions are the routine for other employees. Since there are strong incentives for agencies not to fire PSC employees, it seems highly plausible that politically- influenced managers try to make the work environment so uncomfortable for PSC employees that they will leave "voluntarily." * Salary scale is no longer competitive compared to consulting positions. Salaries for PSC employees do not rise, nor is promotion a realistic career goal. 2.4. Micro Institutional Causes of Informality in Personnel Administration 2.4.1. Evidence of informality suggests that formal policies, rules and procedures that are meant to serve as institutional restraints are not effective. The questions are whether the current institutional arrangements for personnel management are coherent, and whether there are formal mechanisms in place to ensure actual enforcement of the rules. To assess the coherence and the enforceability of these formal rules, we identify six areas of institutional arrangements that define the civil service. They represent a normative benchmark for civil service arrangements widely held by experts and implicit in much of the practitioner literature. Table 2.2 Benchmark assertions on civil service arrangements Civil service arrangements Rules and regulations should be should be coherent - andfitfor enforceable their purpose Civil service policy Arrangements should encourage a There should be sufficient and strategy conscious focus on managing and information in circulation to enable developing the civil service as oversight on adherence to civil circumstances change service policy Legislation and Access to civil service positions Staff should know the rules, consider regulations and responsibilities of civil them reasonable and follow them servants should be legally defined Civil Service Staff responsibilities, promotion The classifications systems should be structure and career prospects and career paths should maintained (they should matter management be credibly defined sufficiently) Pay and Employment Arrangements should enable Government should have the Framework government to act as a information on which to base responsible employer, restraining tradeoffs between remuneration and employment costs while ensuring staff numbers that remuneration arrangements Government should use this do not establish perverse information to provide reasonable incentives rewards Performance Arrangements should allow Individual and agency performance management performance to be defined should be monitored 29 (whether based on compliance or objectives) Accountability Standards should be in place and used to constrain the actions of the executive 2.4.2. Formal institutional arrangements for personnel management in Bolivia are reasonable overall. Especially with the passage of the new Statute of the Public Official (referred to hereafter as the Statute), the legal framework has much improved and provides a relatively clear set of guidelines for sound personnel management practices. However, it is in enforcement of the rules where Bolivia has traditionally demonstrated weaknesses. Mechanisms to ensure rule compliance are particularly weak in four areas: civil service policy and strategy, civil service structure and career management, pay and employment framework, and perfonnance management.47 2.4.3. Civil service policy and regulations. The government's civil service reform strategy focuses on the implementation of the Statute, which is to be done in a decentralized fashion by each public agency under the overall supervision by the new C ivil Service Superintendency (CSS). Enforceability of the new civil service rules will depend largely on the Superintendency's effectiveness as an oversight institution, beginning with its capacity to gather accurate and timely information on agencies' compliance with civil service policy, as well as on the commitment and capacity of the Ministry of Finance, as the "organo rector," to lead the agency-level implementation of the new rules. To date, a glaring lack of information impedes oversight of adherence to formal rules. The collection of public employment data is a serious problem in Bolivia; information is not collected on an annual basis and there is no civil service registry. In fact, no review of the position classification system has been conducted in the past five years (although the Basic Norms of personnel management require such a review), leading to an outdated and unenforced public employment structure and nonexistent administrative career path. What little information is available is not readily accessible, even within the government. Finally, the history of the past attempts at civil service 48 reform indicates at best lukewarm commitment by the Ministry of Finance. 2.4.4. Career structure and management. Consistent with the characteristics of an objectives-based civil service system, vertical progression, seniority and consistency in terms and conditions for public officials are strongly de-emphasized in Bolivia. There is no career path in Bolivia's public administration and advancement within the same agency is extremely rare. There is no incentive for vertical progression in careers. Mobility is extreme, which, to a significant degree, is the consequence of political pressures on an incoming government to dismiss existing staff. Public servants are hired for a specific position, and there are no formal, well-defined arrangements for promotion. Advancement takes place by public servants being selected for a higher position in another agency, or leaving an agency and returning at a higher level. Promotion is not seen as a fundamental instrument for developing a career path. The Basic Norms of 47 The detailed assessment results are reported in the Annex 3. 48 See Annex 4 for details. 30 Personnel Management require that all vacant posts be published, either internally (for internal competition) or publicly (the top three civil service posts must be filled by open competition). Some officials report, however, that it is not uncommon for agencies to post an announcement for a vacancy that has already been filled, in order to obtain a number of resumes that can be kept as "proof' that the recruitment process was followed. However, the intention of the arrangements is to encourage external entry, not to reward seniority or internal promotion based on merit. 2.4.5. The PSC was initiated partly to establish a modicum of merit-based career structure and stability in government employment. Although all civil servants in the PSC program are on common terms and conditions, they represent a very small percentage of total public employees (145 employees, less than .01 percent of total public employees). Other public employees are subject to a number of different formal and informal employment conditions. The combination of diverse regulations for each agency, and the employee performance evaluation system (POAI) is intended to establish an arrangement between the agency head and the employee which is more akin to an individual contract than to a common set of terms and conditions. Even then, evidence is weak that POAI's are used systematically and effectively as means for career management. 2.4.6. Pay and employment framework. In the central administration, positions are broadly classified into three categories: appointees; competitive entry officials; and consultants. Appointees are divided into two groups: high-level appointees (Ministers, Vice-ministers and political advisers); and support appointees to work in the ministers' office (specialists to cover identified skill gaps and confidential support appointees such as personal secretaries). Competitive entry public employees are divided into five groups: director level; mando medio level; profesional level; tecnico/administrativo level; and auxiliar staff. At the competitive entry level, there is no position classification system, only general descriptions in the Basic Norms for personnel management. Finally, the PSC offered another hiring mechanism, whereby donors channeled funds through the Civil Service Executing Unit (later Servicio Nacional de Administraci6n de Personal, SNAP) to pay PSC salaries (donor funding was to be gradually phased out and replaced by government resources). 2.4.7. Another important category of people working in the public sector are so-called "consultants." In principle, these people perform temporary work on specific projects. In fact, however, they often fill line positions and may spend several years in the agency. People hired as consultants are usually well-qualified and/or well-connected, and their salaries are considerably higher than those of their colleagues. This is because consultants are paid with donor funding for development projects, or using certain line items within the agency's budget.49 Estimates of the total number of consultants in Bolivia's central administration range from 2,500 (Servicio Nacional de Organizaci6n del Poder Ejecutivo) to 4,200 (Ministry of Sustainable Development). Needless to say, this is an important example of informal personnel management practice, which is now intended to be corrected by the new Institutional Reform Project. 49 Such as the line item 252 for "studies." 31 2.4.8. Pay policy is significantly distorted. First, there is no government-wide salary scale; even within the same ministry, salaries can vary dramatically. Most public sector salaries are notoriously low (with the exception of some autonomous agencies), and discretionary salary supplements are pervasive. There is significant horizontal decompression (employees doing the same or a very similar job but earning different salaries) in actual pay practice. There are some indications that to date in Bolivia's public administration salaries have been tied to the individual rather than to the post. Furthermore, as discussed in the earlier section, there appear to be rather problematic informal pay practices in the Bolivian bureaucracy. 2.4.9. Performance management and accountability. The SAFCO systems emphasize explicit corporate objectives and plans for public agencies (through the Programa Operativo Anual, or POA and Programa Operativo Anual Individual, or POAI), with annual progress reports prepared. The POA and POAI are intended to hold agencies and individuals accountable for performance, moving beyond the implicit work programs for agencies set out in the establishing legislation, and to define concrete and measurable objectives. In practice, however, most government officials report that the POA is a mere formality and that most employees (even senior officials) have no idea what indicators are outlined in the document. According to comments gathered in focus groups, individual performance appraisals are conducted sporadically (except for PSC employees) and there are no clearly defined, consistent standards of evaluation. 2.4.10. Focus group participants also expressed unanimous frustration that rewards and sanctions are based on factors unrelated to employees' performance of job-related responsibilities. Rather, these depend primarily on political affiliation and obeying top managers' decisions regardless of their quality or appropriateness. This view was supported by 64 percent of survey respondents from Ministries, Servicios Nacionales, and Prefecturas, who revealed that individual performance is evaluated by mechanisms other than the official annual performance evaluation. Moreover, while formally there are supposed to be links between performance appraisal and other personnel management procedures (rewards, promotion, termination, training, etc.), in practice these links do not exist. 2.4.11. Overall, the institutional arrangements set in train by SAFCO are intended to focus largely on attainment of objectives, although compliance with a basic set of management systems and procedures is also important. We already have ample evidence that the ultimate objectives-based goal of the SAFCO Law has yet to be achieved. The persistence of informal practices is a clear sign that most of the formal rules are not strong, and that enforceability is weak, which in turn point to the existence of informal "rules" that supplant or supersede these formal rules. We will turn to analyzing these informal "rules" later, when we discuss the interplay between the political environment and public administration in Bolivia. 32 2.5. Informality in Public Expenditure Management 2.5.1. Ample evidence of informality and weak accountability is also found in the area of public expenditure management. In a well-functioning system, the annual budget serves as an expression of the government's political commitment to a set of expenditure programs, for execution of which the government is held accountable. For a budget to serve this purpose by guiding the government's strategic resource allocations and facilitating efficient use of allocated resources to implement public programs, the decisions contained in the budget document must be credible and binding. In other words, the expenditure programs should be executed as closely as practicable to the policies and strategic plans embodied in the budget document. 2.5.2. In Bolivia, as in many countries with dysfunctional budgetary systems, there are always nontrivial discrepancies between the approved budget and actual executions. While quite effective in meeting its aggregate fiscal targets, the MOF has extremely limited knowledge of, and thus inadequate control over, how exactly funds are being spent during a fiscal year. It is common practice for line agencies requesting cash release from the Treasury to submit what are essentially false claims on their spending (C-3 1 forms). Public entities habitually fail to submit their financial statements to the Accountant General of the State by the deadline mandated in the Constitution, and at the end of a fiscal year, the Controller General (CGR) often has little to audit. Due to such informal practices, the formal mechanism to capture expenditure records - the Integrated Financial Information System (IFMS) - is rendered virtually meaningless, forcing the staff of the Viceministry of Budget to collect information on each agency's expenditures to date through direct visits to the agencies themselves.50 High transactions costs that arise from this practice (as well as from similar ones by other central entities in other related areas) are itself a reason for public sector inefficiency. 2.5.3. One striking characteristic of Bolivia's public expenditure management is the significant reallocations of budgeted funds during the fiscal year, which are even legally sanctioned by Congress both during and at the end of a fiscal year. This allows the government to use budgets as a flexible means to accommodate shifting preferences of coalition partners and interest groups as they implement the annual budget. Likely consequences of this practice are a lack of credibility of the budget as the government's strategic management tool, and erosion of accountability for use of budgeted funds, which makes it virtually impossible to ensure that the funds are used efficiently for appropriate purposes. 2.5.4. Institutionalperformance in budgeting. Good public expenditure management requires simultaneous achievements of three inter-related performance objectives: maintenance of aggregate fiscal discipline (Level 1); facilitation of resource allocations according to the government's strategic policy priorities (Level 2); and efficiency in the use of budgeted funds (Level 3). While the achievement of Level 1 performance may be a pre-requisite for any effective public expenditure management, many countries achieve 50 This analysis does not take into account recent progress made in implementation of the ILACO project, which may qualify some of the statements here. 33 this by weakening performance at the other two levels. The challenge is to develop an institutional framework that effectively balances the three objectives. 2.5.5. Since the early 1980s, a number of diagnostic studies have been conducted of Bolivia's public expenditure management. The past and present diagnoses point out both a significant improvement in aggregate fiscal discipline (level 1) over the years and continued weaknesses facilitating strategic allocations of budgetary resources (level 2), and in ensuring efficient use of allocated resources (level 3).51 These studies indicate that the system fails to establish an effective link between policies (to be determined in a consultative process within and between the Executive and Legislative branches) and the budget, and between operational programming (i.e., definition of program objectives and performance standards by line agencies with MOF approval) and the budget. The budget process continues to be heavily oriented toward minute control of line items without actually achieving effective expenditure control except at the aggregate level. Uncertainty about actual allocations of budgeted funds undermines line agencies' ability to plan and execute their programs, thus reducing their operational efficiency. Moreover, the absence of timely information on actual expenditure patterns during the fiscal year, or even soon after the end of the fiscal year, seriously weakens accountability of the whole government for its use of budgetary resources. 2.5.6. The unreliability of the approved budget as a guide for understanding actual r esource allocations is amply demonstrated in a simple analysis of the budget data comparing the approved and the executed amounts and allocations. Figure 2.1 show degrees to which spending by budget entities deviated from the amounts specified in the approved budgets for selected agencies in 1995.52 These figures show significant re- allocations of funds across agencies during a given fiscal year. In 1995, the Ministry of Sustainable Development "underspent" by 48 percent, while the Vice Presidency "overspent" by 48.7 percent. Mid-year corrections may be necessary in some cases, for example, due to less-than-projected revenues. However, in this particular year, the overall deviation between the approved and the executed budgets was mere -1.3 % (Total (Central adm)), indicating the absence of a need for large scale fiscal adjustment during the year. The average agency-level deviation in 1995 was 14.8%. 51 These studies include Bolivia Public Sector Expenditure Review with a Special Emphasis on the Social Sectors, 1989; Bolivia Public Sector Investment Program Review 1992; "Evaluaci6n del Sistema IFMS," a consultant report, 1998; and Bolivia Public Expenditure Review 1999. 52 The figures show the deviation between the approved budgets and the accrued spending for each agency. 34 Figure 2.4 Deviation of accrued spending from approved budget as a percentage of total agency budgets, 1995 --------------i S__~. ... *Y .:^ ._...v ....^m.i...__._ ._ __._. .. _ _i... . -60 00% -40.00% -20.00% 0 00% 20.00% 40.00% 60.00% 2.5.7. On average, in the Fiscal Years 1990-92/1994-96, between 15 and 50 percent of the approved agency budgets were re-allocated across agencieS.53 In five of these six fiscal years, the average agency-level deviations far exceeded the total deviation (Table 2.2). Although some of these deviations may result from necessary mid-year fiscal adjustments, they are more likely to indicate lack of predictability in resource allocations from the center to line agencies, as well as re-allocations across line items by spending agencies uncontrolled by the Ministry of Finance. 54Both unpredictability of resource availability to spending agencies, and weak control of the final destinations of allocated funds are indications of the unsatisfactory "level-3" performance (i.e., inefficient use of allocated budgetary resources to achieve policy/programn objectives). Furthermore, these deviations are also a sign of the weakness of the budget as a credible, binding expression of the government's policy intentions ("level-2" performance). 53 "Average" deviations were calculated on the basis of the absolute values of each agency-level deviation in order to avoid positive and negative deviations canceling out each other. 540It is important to note, however, that the degree of budgetary unpredictability is likely to be inflated in these figures if indeed the MOF uses the "presupuesto vigente" with reduced allocations for agencies soon alfter the congressional approval (as discussed below) and if the adjusted allocations in the "presupuesto vigente " are known to the spending agencies in advance. 35 Table 2.3 Total and average budget deviations, 1990-92, 1994-96 Entity 1990 1991 1992 1994 1995 1996 Total Central Administration -24.5% -10.50% 1% 3.9% -1.3% 11.3% Average (absolute value) 22.1% 25.8% 49.4% 14.8% 14.8% 27.6% Source: Contaduria General del Estado - Division de Estadisticas Fiscales." 2.5.8. The budget data analysis as well as past and present diagnoses point to serious weaknesses in Bolivia's public expenditure management. To achieve reasonable efficiency in use of funds at the operational level, line managers must be given a degree of flexibility matched by reliable and credible mechanisms to hold them accountable. However, an appropriate balance between flexibility and restraint is not achieved under the existing arrangements for public expenditure management in Bolivia. The current practices of budgeting in Bolivia do not conform to the philosophy of SAFCO that aimed to redirect public management toward a results orientation, which would require clear statements of policy priorities, predictable allocations of resources for those stated purposes, and clear accountability for results delivered through the use of those resources. 2.5.9. Institutional dysfunctions in the budget process. Lack of predictability and accountability is a source of informality. Unpredictable flow of budgetary resources necessitates some coping strategies by spending agencies in order to continue operating in the uncertain environment. For its part, weak accountability system permits or even encourages opportunistic behaviors and produce questionable financial management practices. Both patterns of informality are very clearly observed in the budget process. What formal rules and procedures do exist are not effective guides to budgeting and financial management in Bolivia. As a result, neither existing formal institutions nor their reforms can be relied on fully to improve either operational efficiency or accountability in public financial management. 2.5.10. Budgetformulation and approval: In Bolivia, the budget formulation and approval processes suffer from inadequate information base for decision-making and from insufficient consultations between the MOF and line agencies. The MOF on the one hand tends to set conservative ceilings in order to control escalation of the aggregate budget total, while line agencies hesitate to disclose their own revenues at this stage for fear of receiving even lower ceilings. Although line agencies' annual operating programs (POAs) are supposed to be incorporated into the government's budgetary decisions, POAs are perceived as a mere formality by both the line agencies and the MOF, and do not serve as a real basis for program budgeting. The MOF's decisions on budgetary allocations to line agencies are reportedly made irrespective of the content of the agencies' POAs.56 At the Congressional approval stage, although proposals for new spending are supposed to be considered only when financing sources are properly identified, in fact Congress often proposes to cover new expenditures by increasing the 5 The years 1993 and 1997 are not included in the table because the governmental changes resulted in ministerial re-organization which made across-the-year comparison between approved and executed budgets at the agency level meaningless. 56 Even if the MOF officials tried to take POAs seriously, their format is such that they are not realistic instruments for decision-making. 36 (assumed) efficiency of revenue collection, without technical justification. For the 1998 budget, there were many reallocations and new spending proposals during the approval stage; to fund them, the revenue collection efficiency rate was raised almost 30 percent.57 2.5.11. In short, budget formulation and approval follows the formality of prescribed steps to give the appearance of compliance. However, as some officials stated: "Everyone knows that the budget formulation is a 'salute to the flag,' (i.e., paying lipservice) but we all have to comply with the formalities to obtain resources." Under this system, agencies do not have an incentive to improve their current methodologies of budget formulation and to improve the quality of information they submit to the Vice Ministry of Budget. 2.5.12. Budget Execution: Informality is variously observed at the budget execution stage as well. Many payments with or without budgetary provisions are made with both Treasury funds and with agencies' own funds. Although this behavior is sometimes detected by the Internal Audit Unit and/or the CGR, sanctions against spending funds without proper authorization generally go unapplied. One of the fundamental weaknesses that gives rise to this weak control is the inadequacy of information systems, and the absence of strong incentives for keeping accurate financial information. Although the approved budget is supposed to be registered in the IFMS, thus tying all spending authorizations to the approved budget, the budget passed by Congress is not the budget that actually takes effect. Instead, the day after Congressional approval, a "consolidation" process begins. Based on a revenue projection for the fiscal year, the Treasury Viceministry reportedly cuts the budgets of most agencies between 20 and 30 percent in all categories except salaries. The remaining budget, known as the "budget in force" (presupuesto vigente), is then entered into the IFMS. 2.5.13. Disbursement of public funds is determined by the Treasury, which sets monthly quotas for each agency. The Treasury informs agencies at the end of each month what their quotas will be for the next month. Monthly quotas are supposed to be based on the approved budget, but in fact are based on both the "budget in force" and on the most current estimate of public resources for a particular month. Frequently, agencies' monthly quotas are lower than resources allocated in the approved budget, and may even be less than allocations assigned in the "budget in force." Budget cuts are decided centrally, without dialogue with the affected agencies. 2.5.14. Agencies can request changes in their monthly quotas by submitting a Form C31 with amounts identified by budget categories. Only those requests that are within quotas and have budget allocations are registered in the IFMS and become part of the allocation requests of all agencies for a particular month. Requests for additional budgetary funds are supposed to be made according to the annual budget law and relevant norms. In 57 According to anecdotal evidence, the commission's debate is dominated by the majority chairman, while minority members' observations tend to be suppressed and there is little partisan consensus over conflictive issues. There is not a set of imposed priorities to direct the debate. 37 58 practice, these requests are usually made outside the normal authorization procedure. Agencies also use their own funds to fill the gap between what they need and what the T reasury actually authorizes them to spend. The Ministry of Finance has stipulated three cases in which it will not authorize changes in budget appropriations: changes that imply an increase in total expenditures in the budget law approved by the Congress; changes that imply an increase in salaries; and reallocations of funds from investment to operational accounts. Theoretically, changes in any of these three cases can only be made with prior authorization from Congress. In fact, many agencies reallocate between or within budget categories without proper authorization and outside the limits of annual budget appropriations. In many cases, these unauthorized reallocations are overdrafts. 2.5.15. Regularization of extra-budgetary expenditures: Some legal and "extralegal" procedures have been established for regularization of expenditures, but these are usually used to formalize much of the expenditures made under the practices described above. Three of these mechanisms are: * Reversion, which can be done at any time during the year. Under this procedure, the agency changes the information provided in the C3 1 after having used the funds for purposes different from what was originally intended. * Restructuring, or reformulation of the budget by the Congress, which is usually done in August. This process is used to formalize: * transfers made by agencies without authorization; * transfers authorized formally or informally by the MOF outside the budget law; * additional budget appropriations authorized by the MOF, which required prior authorization by Congress; * agencies' own-source revenue that was not declared during budget formulation and was later registered in the Treasury and used to fund budget categories; and * expenditures funded with international donations and credits which were not registered in the approved budget. * Reassessment, which is done at the close of the fiscal year. During this process, funds are transferred from budget categories with positive balances to those that have negative balances and additional spending carried out by agencies with their own resources are finally registered. In this way, public information on the executed budget does not show overdrafts. 2.5.16. These practices of regularization give rise to three formal budgets apart from the real budget that is made and executed by the Ministry of Finance (budget-in-force), which are: the approved budget, the restructured budget, and the final budget at the end of the fiscal year. The real budget is made and remade throughout the year through the Treasury's cash management at the levels of aggregate total and allocations to agencies. However, even the Treasury or the Accountant General do not really monitor and control exact uses of the funds once they are disbursed to line agencies. While it is quite likely that some of these unauthorized uses of budgetary resources represent instances of opportunism and corruption, it is also clear that agencies have had to develop coping 58 Most of the time this is done by submitting debit notes (notas de debito) to the Treasury which commit more resources than those initially budgeted and later have to be regularized. 38 strategies for handling the problem of unreliability of the budget and unpredictability of resource availability. The practice of unauthorized reallocations of funds is encouraged by line agencies' knowledge that deviations from the approved budgets are likely to be "regularized" either during reformulation or reassessment. 2.5.17. End-of-year reporting: The MOF, through the Accountant General's Office (AGO), is supposed to follow up and evaluate the physical and financial execution of the budget. To this end, agencies are required to submit information on their execution within the first ten days of each month. When agencies do not submit this information on time, different sanctions can be applied: * agencies' fiscal accounts could be frozen; * transfer of Treasury funds could be suspended; and * requests for additional budget could be denied. 2.5.18. In fact, there is no systematic follow-up of physical and financial executions, as only 10 percent of public entities submit information to the AGO on time. Nonetheless, sanctions are only used as a threat in case of non-compliance, and even then somewhat arbitrarily. Agencies' fiscal accounts are seldom frozen when agencies submit their information late. For example, in 1996 and 1997 Congress's observations to the Annual General Account Reports were not submitted to the Executive. Thus, until now, the Congress has not enacted the laws to approve the General Account Report for either year. 2.5.19. Audit: Under SAFCO, central missions of the CGR as the supreme audit institution cover three main areas: * to improve the state's capacity to execute government policies, by evaluating the effectiveness, relevance, reliability and timeliness of the information generated by public management and control systems; * to improve public sector transparency by developing norms for government control (auditing) and by evaluating public investments and operations; and * to hold public sector managers accountable for the use of public funds (both in terms of the process by which funds are spent as well as the results obtained) by overseeing public agencies' expenditures and providing training on SAFCO's administrative and control systems to all public employees 2.5.20. There is general consensus, both within Bolivia and the donor community, that the CGR possesses a high level of technical competence. However, capacity does not automatically translate into impact. Continued corruption in the public sector is the most telling evidence of the ineffectiveness of the Government Control system, which the CGR is charged with leading.59 Two factors in particular account for this deficient performance: the slow and uneven installation of the SAFCO management systems, and the CGR's dependence on other institutions for follow-up of its own findings. 59 CGR officials are reluctant to accept any responsibility for the on-going corruption that affects Bolivia's public sector. One senior CGR official argued that "it is not the CGR's responsibility to fight corruption." 39 2.5.21. The minimum requirement for SAFCO's implementation was the approval of the Basic Norms and subsequent passage of internal regulations within all public agencies so that the management systems could be (at least formally) implemented. Without these systems in place, the information available for the CGR to audit is limited and unreliable. However, issuance of secondary regulations and agency-level application of the SAFCO systems was delayed considerably after the passage of the SAFCO Law itself. It was not until 1997, under considerable pressure from the donor community, that all the Basic Norms were issued. However, even in the two years since then, agencies are behind in drafting their internal regulations. The MOF is charged with approving agency-specific regulations, but many managers we spoke to expressed frustration over the MOF's perceived slowness in setting up its own internal guidelines, which in turn may be used as models by other agencies. 2.5.22. Without the management systems defined by SAFCO, it is not possible to set up an effective control system. The dearth of reliable information on inputs, objectives and outputs/outcomes makes it hard to check compliance with procedures and even harder to hold managers/agencies accountable for performance. With annual inputs of approximately 500 person years by CGR staff, in addition to the agencies' internal auditors and external auditors from private firms, the CGR conducts financial audits of about 20 percent of public sector agencies each year. Private auditing firms cover another 10 percent. The remaining 70 percent of the public sector, including the Central Administration, receives no financial audit; that is, the financial statements rendered by the executive in the Informe Anual de la Administracion Central which comprise approximately 52 percent of total public expenditure and around 80 percent of the state's assets, are not audited.°° 2.5.23. Even if adequate financial accounts were available for audit and financial mismanagement were more effectively detected, the CGR alone could not sanction it. UJnder SAFCO, the CGR must turn over its findings to other bodies which are apparently unable or unwilling to apply consistent, meaningful sanctions against public employees who break the rules. Different types of responsibility require the CGR to present the case to different institutions, and the more serious the infraction the less power the CGR has over its final treatment. One member of the CGR's upper management admitted her dissatisfaction with the current situation: "the CGR can apply social pressure during the auditing process and make it unpleasant for the person being audited," but in most cases that is as far as the process goes. Recent statistics quoted in the Bolivian media put the total "notas de cargo" issued by the CGR at $11.9 million, of which only $1 million have been collected (just over 9 percent). 60 In 1998, 149 special audit reports were produced, 114 of which resulted in presentation of findings (dictamen de responsabilidad) by the CGR. This represents an increase from 86 findings issued in 1997 and 44 in 1996 (we do not know how many special audits were conducted in those years). Action against the officials involved was initiated by the CGR for 44 percent of the findings. Bolivia Country Financial Accountability Assessment, July 1999, p. 28. 40 2.6. Micro Institutional Causes of Informality in Public Expenditure Management 2.6.1. Deficiencies in the formal rules and procedures.61 The set of formal rules that govern public financial management procedures in Bolivia suffer from vagueness and even some internal contradictions. In some areas, the law leaves too much room for interpretation. In the absence of a stable informal consensus to supplement the formal rules, legal ambiguity results in unrestrained opportunistic behaviors by agencies. In addition, there are some specific rules that need to be defined but are still missing from the law. For example, the Treasury Basic Norms and the Annual Budget Law mandate that resources should be used only on expenditures for which funds were requested. At the same time, some exceptions are defined in a ministerial guideline. Many agencies use this procedure to regularize unauthorized expenditures, claiming that the guideline could be applied more generally. Indeed, the ministerial guideline allows for different interpretations of specific circumstances under which the regularization procedure could be used. Contradictory procedures can be found in different parts of the regulations. For instance, the SAFCO law determines that the Annual General Account Report must be submitted to Congress approximately three months after the closing of the budget. However, Ministerial Resolution 704 establishes that deadline as August 6 of the new fiscal year. Some agencies comply with the first deadline, but most claim that it is the second deadline, which is in force. 2.6.2. Weakness of supporting institutions. Incentives for opportunism provided by deficiencies in formal rules and procedures are reinforced by weaknesses in certain institutions that should support enforcement of procedural compliance. The state's ability to deter opportunism through ex post monitoring and sanctions, as envisioned for the SAFCO Control System, is very limited. The CGR is the primary supporting institution charged with guarding the state's financial probity. Although the CGR is known to be technically competent and relatively free of patronage, a number of obstacles keep it from playing its monitoring role effectively. The Judiciary, the ultimate means for accountability, has been highly politicized and suffers from serious capacity problems (see Annex 6 for more details).62 2.6.3. The weaknesses of the institutional mechanisms to enforce procedural compliance are exacerbated by unavailability of timely and reliable information about government financial transactions. Information on the final allocation of Treasury funds is mostly unavailable throughout the fiscal year. Treasury funds could be allocated to expenditures without budgetary allocations. Ideally, central authorities should be able to monitor this departure and apply proper sanctions. However, once transfers are made to line agencies, information on the actual final allocation becomes extremely difficult to obtain. 2.6.4. It would be a mistake to conclude from this assessment that the main problem is simply poorly functioning information systems. The real problem is a deeper institutional one that arises from well-entrenched practices of circumventing existing 61 The law is defined here as the set of laws and regulations having to do with the budget process. The annex contains a description of those laws and regulations. 62 See Annex 6. 41 procedures for information processing and reporting actual financial transactions. The most egregious example is the use (abuse) of Form C3 1. Ideally, IFMS should capture accurate information on spending patterns by line agencies so that the MOF can carry out effective, timely monitoring of budget execution. However, because of what amounts to false claims with the C3 1, the IFMS information does not allow the MOF to verify agencies' final use of funds. The unreliability of IFMS has forced the MOF to rely increasingly on parallel information systems to carry out and follow up budget execution and to verify the information available from IFMS. 2.6.5. MOF's institutionalpriority. As seen so far, public expenditure management practices in Bolivia are satisfactory in achieving only the first of the three inter-related performance objectives. Sound fiscal and monetary policies combined with tight cash management by the Treasury achieve reasonably responsible handling of aggregate fiscal discipline. Given the recent history of extreme macroeconomic instability, and in contrast to a number of other countries in the region which have so far failed to bring aggregate fiscal discipline under control, this is a very commendable achievement. Having demonstrated the ability to maintain macroeconomic discipline on a sustainable basis, however, the Government should seriously consider stepping up its efforts to further strengthen its fiscal and financial management capacities. 2.6.6. However, in clear contrast to the MOF's concern about fiscal discipline, the other two areas of budgetary performance - strategic allocations and technical efficiency in resource use - appear to receive little, if any, attention from central government. This lack of commitment to improving public financial management partly explains the delay in issuing most of the SAFCO management system norms, and the slow progress in implementing the series of SAFCO/ILACO projects. However, attributing these problems to "lack of commitment," while accurate in our view, trivializes the true nature of the problems. The real question is: why has this reform-minded government been so slow to effect public financial management reform despite its own public commitment and significant donor support over the past decade? As in the case of public personnel management and civil service reforms, persistent informality and lackluster implementation of public financial management reform has its root in the nature of the political economy and its impacts on public administration. 2.7. Summary 2.7.1. This chapter has documented systematic informality in two core areas of public administration, namely, public personnel and expenditure management. In both areas, reasonably sound norms and regulations exist. However, in some cases, the norms themselves need further clarification as ambiguity and even contradictions reduce their effectiveness in encouraging rule-bound behaviors. There is some evidence that the SAFCO norms are yet to be fully internalized by public officials, thus facilitating ad hoc (i.e., not predictable according to formal rules and regulations) behaviors within the public sector. Even when internal norms and regulations are relatively clear, often they are not effectively enforced. Glaring lack of critical information impedes effective management personnel and finance in the Bolivian public sector, and the formal 42 mechanisms to enforce rules and sanction non-compliance are often under-utilized, while sets of informal practices have been developed to circumvent formal procedures. In some cases, especially in public expenditure execution, both legal and extra-legal practices have been established to formalize questionable informal practices. Systematic presence of informality in public administration indicates breakdown of the rule of law within the state. There is every reason to expect such an institutional breakdown to have serious impacts on the effectiveness of government bureaucracy. 2.7.2. SAFCO is an attempt to establish an institutional order within the Bolivian public sector. But, it has encountered persistent difficulty in its implementation. In fact, in the absence of a strong, central oversight body, the decentralized management structure that SAFCO has legitimated may be working against any effort at uniform application of government-wide internal regulations and norms. Besides, the present patterns of informality, especially in the area of personnel management, reveal pervasive influence of political parties on public administration. While politicization of public administration is a common problem in developing countries, the degree to and the manner in which political parties exert influence on the Bolivian government bureaucracy are a serious cause for concern from the point of view of public sector performance. In order to better understand deeper causes of bureaucratic informality and the nature of the trade-off between macro political governability and institutional performance of the public sector, we will turn to an analysis of the broader political environment and its interplay with the public administration in the next chapter. 43 3. THE POLITICAL ENVIRONMENT OF THE BOLIVIAN STATE: DEEPER CAUSES OF POOR PUBLIC SECTOR PERFORMANCE 3.1. Patrimonial Dynamics of Party Politics 3.1.1. Systematic presence of bureaucratic informality points to the existence of "informal" rules that govern behaviors of public officials in Bolivia. This section identifies salient features of the political environment that surrounds Bolivia's public administration, and offers some explanations for the observed problem of informality. These features refer to the extreme importance of patronage in politics; the dynamics of coalition politics that necessitate constant inter-party negotiations; and the crucial role of technocratic governance. After identifying these features in general terms, the chapter will conclude by developing an explanation as to how these features of the Bolivian political system contribute to the observed bureaucratic behaviors, especially to informality and weak formal accountability. 3.1.2. Problems of Patronage Politics: The importance of patronage and clientelism has been a central feature of Bolivian politics. Both academic experts and our informants single out clientelismo, where political criteria dominate the organization and management of the state bureaucracy, as one of the most serious obstacles to effective functioning of the state. Academic observers have stressed the "patrimonial dynamics of party politics" in Bolivia: 3.1.3. In a region where political parties have often been primarily vehicles to capture and circulate state patronage among the dependent middle classes, Bolivia stands out as an extreme case of this essentially patrimonial political dynamic. Parties in Bolivia have been quite permeable ideologically, party loyalty among the political class has been low, and more often than not the behavior of parties has been driven more by access to patronage than by constituting programmatically focused governments.63 3.1.4. The phenomenon itself is by no means unique to Bolivia. Still, capture of the state apparatus by political forces has significant implications for performance of public administration, and therefore is an important concern for this report. The effects of patronage politics on public sector performance are all too clear. Use of political over technical criteria for recruiting, selecting, and promoting public sector employees prevents development of professionalism in the public service, and lowers efficiency and accountability of its operations. Prevalence of particularistic considerations in government decision-making undermines impartiality of government policies. Widespread use of patronage appointments in government bureaucracies can be explained by politicians' interests in garnering/maintaining support of key constituents (by offering patronage jobs) as well as by their interests in controlling the government bureaucracy in order to use it to advance their political/policy/personal goals. 63 Gamarra and Malloy (1996, p. 399). 44 3.1.5. Patronage as a Means of Political Control: An application of the principal-agent theory to the politician-bureaucracy relations suggests that the degree to, and the manner in, which politicians delegate operational discretion to bureaucrats depends on methods of political control available.64 In a developed institutional setting, there are a variety of methods available for politicians to control the bureaucracy, albeit imperfectly. In theory, politicians (for example, ministers or parliamentarians) could monitor their subordinates' behavior very closely to prevent opportunism, and reward or sanction them depending on their performance/behavior (perhaps with middle-level managers as supervisors). Politicians could also rely on ex post oversight mechanisms such as audit, special investigation and public complaints to detect and sanction undesirable bureaucratic behavior. Finally, by clearly specifying and enforcing administrative procedures, politicians can bind public officials' abilities to exercise substantive discretion. 3.1.6. The most traditional form of bureaucratic control mechanism is ex ante control of public officials' behavior and bureaucratic input (such as use of budgets). Bolivia's current budget system does not provide adequate, timely information on the actual use of financial resources, and thus is not an effective means to monitor and control bureaucratic behavior. Until the mid- 1 980s, Bolivia had a monitoring system that relied heavily on ex ante control by agents of the CGR, called interventores. This system created a perverse moral hazard problem, whereby the monitors became (or at least were perceived as) co- responsible parties in opportunistic behavior, and undermined the accountability mechanisms in the public sector. In essence, no one monitored the behavior of the monitor. Although interventores may have served as means for some politicians to exercise influence over bureaucratic behavior, the requirement for detailed ex ante control unarguably affected efficiency of bureaucratic operations significantly. The interventores and the ex ante control regime were abolished with the reform of the CGR in the mid 1980s. 3.1.7. As for the second method, expost oversight, Bolivia has yet to develop a reliable system of performance evaluation (as envisioned in the SAFCO Law and contemplated in the Institutional Reform Project) and of public sector audit, let alone a capacity to conduct timely and reliable special investigations of suspected opportunistic behavior in government. Information flow from the government to the public is limited, making it difficult for interested citizens to monitor public sector behavior so as to be able to "whistleblow" when necessary.65 Furthermore, Congress lacks both the capacity and the incentives for institutionalized mechanisms of monitoring and control over the government bureaucracy. 64 Matthew McCubbins, Roger Noll, and Barry Weingast, "Structure and process, politics and policy: administrative arrangements and the political control of agencies", Virginia Law Review 75 (1989, pp. 431- 482). 65 However, the Ombudsman (Defensor del Pueblo) was created to facilitate citizen oversight of public sector misbehavior in the latest Constitutional Reform. 45 3.1.8. The third mechanism, administrative procedures, is even weaker in Bolivia than ex post oversight.66 A critical difference between Bolivian and U.S. institutions in this respect is the reliability of formal legal means of controlling the bureaucracy. While SAFCO offers a good overarching legal framework to guide public sector behavior, its implementation has been incomplete, and the system of enforcement necessary to "give teeth" to SAFCO is weak. Besides, pervasive informality in Bolivia's public sector means that formal legal mechanisms, as a means to control bureaucratic behavior in order to make it responsive to political principals, are far from reliable. Finally, the effectiveness of administrative procedures, even once they are implemented, is likely to suffer from the judiciary's inability to enforce these procedures in a fair and predictable manner. 3.1.9. Given this reality of inadequate instruments, given politicians' interest in being able to control the bureaucratic apparatus for either legitimate public policy reasons or for purely political and personal gains, and given the strong societal demand for employment, one obvious course to take is continued reliance on patronage. By placing their "confidantes" in key bureaucratic posts, politicians can maximize the probability that the bureaucracy, at least those parts under their direct influence, stays responsive to their interests, while minimizing the cost of monitoring individual officials' behavior since those occupying patronage posts are (presumably) trustworthy. 3.1. 10. Unlike formal methods of control such as legislative and judicial oversight and administrative procedures, which are meant to provide universalistic (i.e., non- personalized) form of accountability, patronage appointments tend to make bureaucrats respond primarily to their patrons and perform functions that may or may not be among their official duties. The absence of formal methods of control is likely to mean subordination of universalistic/public interests to particularistic/private interests when political control is exerted over the government bureaucracy. 3.1.11. This implication does not change even if we assume that patronage appointees are always technically qualified for their positions, for their incentive is to be responsive to their political patrons. Of course, a well-known problem arising from clientelism is that technical competence is often given a secondary consideration in appointment decisions. Given a choice between technical competence and political loyalty, politicians are more likely to value loyalty. According to this logic, public employment's primary function is to serve as a mechanism of political control; its role as an instrument to carry out policy programs is secondary.67 66 A new Administrative Procedures Law is being discussed in Congress at the timne of this writing. 67 This point does not negate possibilities that patronage appointees are equally or more competent than non-patronage appointees in some cases. Most of the so-called technocrats in Bolivia and other Latin American countries are essentially patronage/political appointees, who have been chosen for their positions on the basis of their technical competence and political allegiance to the patron (i.e., presidents, ministers). 46 3.2. Coalition Politics and Inter-party Negotiations 3.2.1 Another striking feature in Bolivia's recent political history is the fact that dramatic instability until the mid- 1 980s has been replaced with remarkable stability since then. Political actors have accepted the democratic "rules of the game," centered around political representation through officially registered political parties, which abide by constitutional means to reach and exercise power. Furthermore, Bolivian political parties have displayed abilities to arrive at negotiated settlements for power sharing and implementation of certain public policy programs. The parties' adroit use of coalitions and their respect for institutional independence of the National Electoral Court as an arbiter of electoral and party competition are clear signs of their willingness to engage each other within a more or less clearly established set of rules. It is to the credit of Bolivian political leadership that the country now enjoys a stable, democratic system with no visible threat to its sustainability, either from inside or outside the system. 3.2.2 However, as the episodes of governrment-opposition conflict (e.g., between the Acuerdo Patriotico and the MNR-led opposition during the Paz Zamora administration, disagreement over appointments of the members of the Judicial Council) indicate, mistrust between parties (or at least between certain personalities across party lines) still appears deep and strong. Maintenance and management of political coalitions, which are necessary for the government to carry out its legislative agenda, is therefore a difficult task requiring a high degree of negotiation and leadership skills.68 3.2.3 Two sets of institutional factors necessitate and encourage inter-party negotiations. First, negotiation is inevitable for parties to overcome collective action problems and govern the country effectively simply because of the fragmented nature of the party system. Given the current situation of multi-party regime, no single party can realistically expect to govern without forming a coalition. Thus, fragmentation forces negotiations. Second, negotiation is encouraged by the peculiar executive-legislative relations in Bolivia, which has been characterized as a "hybrid" between pure presidential and parliamentary systems.69 This institutional arrangement facilitates inter-party negotiations as in multi-party parliamentary systems. 3.2.4 Fragmented party system: The Bolivian party system has historically been quite weak, and it is only since the democratic transition that the system has come to display a 68 Again, academic observers note that the success of the Pactopor la Democracia between the MNR led by Paz Estenssoro and the ADN led by BAnzer depended to a large degree on the two party leaders' capacities to control their respective party factions and rank-and-file members. Gamarra reports that "throughout the three-year duration of the pact, members of ADN and the MNR complained about the alliance. MNR militants were upset by the loss of sources of patronage and the perceived loss of the party's traditional populist electorate. ADN members, in turn, did not trust the MNR's promises of support for Banzer's presidential bid in 1989. Weekly meetings of representatives from each party resolved many of these issues. But in the main, the role and presence of both Banzer and Paz Estenssoro prevented party discipline from breaking down earlier" (1997, pp. 375-376, ffi. 15). 69 Gamarra (1997). 47 degree of stability in its structure.70 The system is quite fragmented, with five parties controlling similar shares of seats in the Chamber of Deputies - currently the shares of the Lower House seats among the five major parties are: ADN (27%), MIR (19%), MNR (19%), UCS (15%), and CONDEPA (14%).71 This party fragmentation is no doubt a reflection of the socioeconomic, ethnic and geographic diversity of the Bolivian society, but is also reinforced by certain features of the electoral system. 3.2.5 The closed-list proportional representation system, which had been used to elect all of the deputies until the 1997 elections and half of the deputies since then, tends to strengthen the influence of party leaders, who control selection of candidates and their position on party lists. When voters are not well-organized or well-informed, which may be the case of a large number of Bolivian voters, this system creates an incentive for politicians to be more responsive to party (or faction) bosses, thus weakening the crucial link of electoral accountability. On the other hand, party discipline appears to be only nmoderately strong among the major parties. There are known cases of politicians switching party affiliations. There are also reports of some rank-and-file members with multiple affiliations. Parties are factionalized internally, and apparently depend more on the personalities of particular leaders than on well-developed internal organizations. Relatively weak party discipline may place party bosses in potentially tenuous positions vis-a-vis some of the more powerful party members, thus necessitating further intra-party negotiations to maintain organizational cohesion. 3.2.6 "Hybrid" executive-legislative relations: The process of government formation in Bolivia is peculiar in that when no presidential candidate has won a majority of votes, Congress elects the president from the top two (three until the electoral reform in 1995) contenders. 72 Similar to a parliamentary system (hence the characterization as "hybrid"), the legislature, instead of the electorate, has the power to decide on the most crucial aspect of government formation. 73 This system of indirect presidential election, in the context of a fragmented party system, has encouraged coalition formation and constant inter-party (sometimes intra-party, inter-factional) negotiations before and after actual voting. 3.2.7 In essence, the effect of this system is that "any one of the three largest parties needs at least one of the other two to win: It needs either for the other not to present a 7o "Between 1979 and 1993 the number of parties legally registered with the Electoral Court fell from seventy-one to sixteen," according to Carol Graham, Privatizing Markets for Public Goods: Raising the Slakes in Economic Reform (The Brookings Institution, 1998). ' The electoral system favors "large" parties in the Senate, and as a result, the ADN currently controls 11 out of 27 seats, followed by the MIR (7), the MNR (4), Condepa (3) and UCS (2). 72 Since the return of democracy in 1982, no presidential candidate has won enough votes to be elected without a congressional run-off. 73 Unlike in some presidential systems (e.g., US), Congress does not have the authority to confirm/disconfirmn the President's ministerial appointments - the only civilian officials whose appointments need to be approved by Congress (Senate) are ambassadors. The Constitution is somewhat vague about Congress's power to censure ministers, and the procedures to follow. However, it appears that Congress's ability to "censure" a minister does not go beyond a recommendation to the President, as the Constitution clearly states that a "censure" could lead to the minister's resignation, but this could be accepted or rejected by the President. 48 candidate or for the other to vote for it after the election"74 (i.e., in the run-off in Congress). The need to have an ally for the purpose of winning the presidency induces the major parties to seek compromise with potential coalition partners. Smaller parties, which might be marginalized in countries with dominant parties, can also expect a reasonable prospect of sharing power. 3.2.8 Through these inter-party agreements, in the Bolivian political system, proportional representation of parties is reproduced not only in Congress but also in the Executive branch, which makes a divided government (i.e., the Executive under one party's control and the Legislature under another) unlikely to result. As long as the governing coalition is able to maintain its internal cohesion and each party its internal organizational discipline, the system also facilitates effective governance in the legislative arena, which is one of the reasons for Bolivia's good reform records in recent years. 3.2.9 According to the predominant view among students of Bolivian politics, a glue that keeps governing coalitions together is the careful sharing of patronage opportunities. Academic analyses of the political economy of reform in Bolivia since 1985 emphasize the critical role of various pacts, which all rested on patronage sharing arrangements and some sort of basic agreements to support certain policy programs that the presidents wished to implement (e.g., D.S. 21060, Capitalization, Popular Participation). Thus in Bolivia, the ruling party's and the President's need to maintain a governing coalition is an additional factor (besides their desire to control the bureaucracy) that encourages widespread use of patronage as a bargaining chip in inter-party negotiations. 3.2.10 Since the more veto subgroups there are, the harder it is to change the status quo,75 this situation of bargaining among multiple players should be an obstacle to fundamental reforms in the absence of a consensus among the veto players. However, Bolivia has been able to implement a number of important reform measures, which in other countries might have faced strong political and ideological oppositions. Some of the reasons for the impressive reform record may lie in the pragmatic orientations displayed by the main political parties, which are not driven primarily by strong commitment to particular ideologies and policy programs based on them. Furthermore, by now all the major parties have at some time been part of a governing coalition that has furthered economic and institutional reforms, making it difficult for them to criticize the fundamental directions of these reform efforts. Paradoxically, to the extent the legislative coalitions are built among several political parties on the basis of shared patronage, recent reforms in Bolivia owe their successes to the calculated use of patronage, even though the latter fundamentally contradicts the logic of most of these reform measures. The question remains, however, as to whether deeper and broader institutional reforms as contemplated 74 Matthew Soberg Shugart and John M. Carey, Presidents and Assemblies: Constitutional Design and Electoral Dynamics (Cambridge University Press, 1992). The "winning" parties have often needed more than one party to form a coalition as even two of the three major parties combined have more often than not failed to win a simple majority. 75 Gary Cox and Richard McKelvey, "Ham Sandwich Theorems for General Measures," Social Choice and Welfare 1: (1987, pp. 75-83), cited in D. Roderick Kiewiet and Matthew D. McCubbins, The Logic of Delegation: Congressional Parties and the Appropriations Process (Chicago University Press, 1991, p. 34) 49 in the National Integrity Plan, which would further "squeeze" the sources of patronage and rent pool in the public domain, will be compatible with the patronage logic. 3.3. The Dynamics and Limits of Technocratic Governance 3.3.1. Almost invariably, the recent major reforms in Bolivia have been led by teams of technocrats who are relatively insulated from political interference with short-term partisan, electoral concerns.76 It is evident that the traumatic experience of hyperinflation in the early 1980s (which was in many ways a culmination of the weak governance capacity of successive governments) and the indisputable success of the stabilization measures have contributed to the emergence of a clear political consensus to preserve operational autonomy of technocrats in key economic policy-making arenas. The Bolivian political elite have effectively delegated macroeconomic management to these technocrats by respecting formal organizational autonomy of government agencies charged with macroeconomic management, such as the Central Bank and the Superintendency of Banks. With the independent Central Bank and the relatively insulated policy-making apparatus, Bolivia's record of macroeconomic management has been noteworthy. Box 3.1 UDAPE A particularly well-known example of an effective technocratic agency is the economic analysis unit UDAPE (Unidad de Analisis de Politicas Econ6rnicas). UDAPE was created in 1983 with USAID support, with the original mandate of advising the then Ministry of Planning and Coordination on macroeconomic and some sector-specific economic issues. Its staff have traditionally received relatively high salaries, and benefited from extensive training opportunities. A number of UDAPE staff had their academic training at the Catholic University of Bolivia, and many of them have had graduate training abroad. As a small unit of competitively-recruited technocrats, UDAPE seemed to enjoy a high level of esprif de corps and professionalism. Any policy proposals with fiscal implications must go through UDAPE for review before they reach the Economic Cabinet. As a technical advisory body, UDAPE and its policy recommendations appear to be highly respected by politicians. Only rarely would the Economic Cabinet override UDAPE's recommendations. 3.3.2. Besides UDAPE and the Central Bank, the Ministry of Finance (MOF) is staffed with competent technocrats, many of whom have had a relatively long tenure in the Ministry despite generally high staff turnover in the public sector as a whole. One of the reasons for the relative stability in the Ministry of Finance is the technocratic nature of rnany of its tasks, such as budgeting, accounting, and financial programming, which cannot easily be fulfilled by political appointees without seriously affecting the smooth 76 The academic study that has emphasized this aspect most aptly is Catherine Conaghan, James M. Malloy, and Luis A. Abugattas, "Business and the 'Boys': The Politics of Neoliberalism in the Central Andes." Latin American Research Review, 25 (1990). The term "boys" here refers to technocrats in government, originally borrowed from the famous "Chicago boys" who led economic reforms under the military regime in Chile. 50 operations of the entire government. Since the MOF does not enjoy legal autonomous status as the Central Bank and the Superintendencies do, there is greater room for political interference in the MOF's affairs. Nonetheless, it seems fair to describe key parts of the MOF as informally-insulated technocracy. 3.3.3. Clearly technocrats have been allowed to play much larger and more important roles in Bolivia than in some neighboring countries which are still struggling with the "first-generation" reform agenda. Like Bolivia, countries which have made significant strides in economic reforms have all relied on technocrats to design and implement reform measures (e.g., Argentina, Brazil, Chile, Mexico). However, in spite of the important roles technocrats have played in furthering the economic and institutional reform agenda, the Bolivian situation does not match countries where technocrats appear to have established much stronger institutional presence within government, such as in Mexico, where even presidents and party leaders have emerged from the technocratic corps and have established commanding positions over control of the government.77 While Bolivia has seen its share of important politicians with technocratic background, they have had to co-exist with traditional politicians in a difficult coalition environment, in which political negotiations are essential elements of governance. 3.3.4. The realm within which technocrats can act autonomously from political influence are always defined by politicians, even with regard to institutionalization of autonomous spaces such as the granting of legal independence to the Central Bank. If autonomy is not fully institutionalized, the effectiveness of technocratic governance is even more fragile. Given the coalition dynamics of Bolivian politics, protection of institutional autonomy of technocratic segments within the state must of necessity be maintained via formal or informal inter-party agreements. In the area of macroeconomic management, the severe economic crisis of the early 1 980s and the dramatic success in stabilization in the late 1980s apparently produced such an inter-party consensus to leave macroeconomic management largely out of the partisan influence. This is understandable because macroeconomic instability is a public bad which hurts all parties. The political parties, who possess significant capacity to disrupt sound macroeconomic management through political interference in relevant government agencies, have collectively chosen to constrain their own behavior in order to save themselves from their own (potentially) destructive behavior.78 3.3.5. A similar consensus does not seem to have emerged on mnany other issue areas, including anti-corruption and public sector institutional reforms. This is understandable because patronage opportunities are private goods, which parties have little incentive to give up. It remains to be seen whether the current attempt to reform the customs administration and to establish an independent Civil Service Superintendency, along similar lines of granting autonomy, will be supported by the same degree of inter-party consensus. 77 Miguel Angel Centeno, Democracy Within Reason (Pennsylvania State Univ. Press, 1994). 78 Another important factor that explains the sustained macroeconomic stability is the weakening of organized labor, which posed serious challenges to governments' efforts to control the economy in the early 1980s. 51 3.4. The Impacts of Politics on Public Administration 3.4.1. Patronage andpersonnel management in the Bolivian bureaucracy. Once the three salient features (pervasive patronage; the importance of coalition building among parties; and the presence of a political consensus surrounding technocratic macroeconomic management) of the Bolivian political system are taken into account, deeper causes for pervasive political influence in the public administration become explicable. As in other countries, patronage politics in Bolivia allows politicians to satisfy some of their supporters by providing them with scarce job opportunities in the public sector, and to control the bureaucracy by placing individuals who enjoy the politicians' personal confidence. In addition to these relatively common reasons for the persistence of patronage politics, Bolivian politicians face an additional and strong incentive for resorting to patronage because of the dynamics of government formation. Any government in Bolivia has a pressing need to build and maintain a coalition. This gives minority parties a clear bargaining chip in relation to the President's party as it attempts to form a government. The President's party must entice minority parties into a coalition government, and must establish a reputation as a trustworthy party to team up with. This political necessity has reinforced the Bolivian society's clientelist culture, and institutionalized the practice of "cuoteo politico", a quota system whereby ministries, viceministries and even directorates are divided among the parties in the coalition. 3.4.2. The distinctive need to build coalitions arises in Bolivia because of the extremely fragmented party system from which no single candidate is realistically able to gamer a majority vote to win the presidency, and because of the peculiar electoral procedures that defacto charge Congress with electing the President. Once elected, the President cannot pass any legislation without a majority control of the Congress, and has no power to rule by decree other than within the substantive boundaries of existing laws.79 The task of building and maintaining a coalition is particularly difficult in Bolivia because of the number of parties. The largest share of lower house seats that a single party has won since the transition to democracy in 1982 is 41 percent (MNR in 1993). Although the collapse of a coalition does not lead to the fall of the government as in a Parliamentary system, collapses carry a cost. Potential coalition allies of the next government will be wary of parties that could not maintain their authority. 3.4.3. If the imperative of coalition building provides a particularly strong motive for offering coalition partners inducements to join the government, then the over-ambitious institutional arrangements for central government employment under SAFCO provide the opportunity. The decentralized operational framework of SAFCO has facilitated the systematic award of discretionary payments and patronage positions to parties in exchange for agreement to enter the coalition. By establishing control mechanisms which cannot realistically be monitored, there is no obstacle to patronage as the trading currency for coalition-building. If this interpretation is valid, then it becomes apparent why SAFCO's decentralized operational framework has gained wide political acceptance 79 Major quasi-constitutional legislative actions (e.g., congressional ratification of the state of siege, which was used both during 1985-89 and 1989-93 to facilitate economic governance) requires a 2/3 majority. 52 while its enforcement mechanisms are not fully buttressed. The SAFCO managerial logic of operational decentralization suits the political logic of fragmented party capture, while the ideal of centralized normative framework does not. 3.4.4. The Political Economy of Public Expenditure Management. A similar political explanation might be applied to understanding continued informality in public expenditure management. One of the more obvious reasons why financial management reform has progressed so slowly in Bolivia is the apparent lack of political commitment to accelerate its implementation. The CGR has been aggressive, at least in the early years immediately following SAFCO's promulgation, in fulfilling its obligations for implementing the SAFCO framework. But its mandate was limited to instituting the government control system, which, as seen above, depended for its effectiveness on the functioning of the other seven management systems. But the MOF was very slow to issue the Basic Norms for the seven SAFCO management systems over which it has been assigned the role of "rector institution." 3.4.5. One plausible (and understandable) reason is that, given the inability of the informatics side of the integrated financial management (ILACO) projects to provide the Ministry with a reliable IT system, various parts of the Ministry have had to rely on sub- optimal but sufficiently functional existing systems to carry out their day-to-day operations. The result is continued postponement in integrating the financial management systems, and prolonged existence of parallel systems that are incompatible with each other. Another possible reason for the delays is resistance, not only from within the MOF itself but also from the rest of the public sector, to introducing a more transparent financial management system, which would make some of the questionable financial transactions more difficult to conduct. We have no concrete evidence to support this argument, but based on anecdotal information, it is reasonable to assume that at least some resistance would have arisen for this reason. 3.4.6. Both of these problems could only be addressed with strong political commitment and leadership. It is evident, however, that such leadership has not been forthcomning. There are two possible explanations for the lack of political commitment. It may be that the MOF has always concentrated its attention on macro-fiscal management, and simply has not considered improvement of strategic prioritization and efficiency in resource use (budgetary performance at the levels 2 and 3) important priorities. It is also plausible, however, that some underlying political factors inhibited full-scale introduction of public financial management reform. The three characteristics of Bolivian political institutions (pervasive patronage; the importance of coalition building among parties; and the presence of a political consensus surrounding aggregate fiscal discipline to be guarded by competent technocrats) identified earlier are relevant for understanding some of the dynamics at the heart of public expenditure management. 3.4.7. First, pervasive use of patronage criteria in public sector employment weakens the government's human resources, both in terms of staff's technical competence and continuity of personnel in key financial management positions, resulting in observable informality and lack of accountability. We heard of cases where political-patronage 53 considerations allowed financial management positions to be filled with staff with no training in accounting. The absence of a government-wide merit-based civil service to maintain professional ethos, combined with ineffective rule enforcement mechanisms, means that individual public officials with authority to spend are not sufficiently constrained from opportunistic behaviors. 3.4.8. Second, the most serious impact of Bolivia's peculiar political dynamics on public expenditure management might plausibly come from the government's need to negotiate and maintain an inter-party coalition throughout the presidential term. Just as posts in the government bureaucracy are divided among coalition partners through the practice of "political quotas," it is probable that allocations of some of the financial resources follow the same logic of bargaining and sharing among the parties in the governing coalition, at least to some degree. Furthermore, it is quite probable (and consistent with anecdotal evidence) that expenditure allocations across geographic areas are influenced significantly by political factors. Thus Prefectures apparently spend significant amounts of budgetary resources in municipalities with a high degree of discretion, over which the central government appears to have limited control. There is also some evidence that allocation decisions by the investment funds may also be influenced by partisan, political criteria (Box 3.2). Box 3.2 Political Determinants of Social Investment Allocations, 1996-98 Bolivia's Social Investment Fund (FIS) represents an important element in the Bolivian governments efforts to increase the fiscal autonomy of the country's newly created municipal governments. Based on an analysis of FIS spending in 296 of Bolivia's 311 municipalities, two features of the country's political landscape emerge as critical determinants of the allocation of program funds-the partisan leaning of a municipality's electorate and the stability of the municipal government during the years under analysis. Of the 296 municipalities with spending data, 72 (24%) did not receive any funds from the FIS between 1996 and 1998, 112 municipalities (38%) received one year of funding, 67 (22%) received two years of funding, and 45 (15%) received money from the program in each of the three years. The single most important predictor for the number of years a municipality received FIS funding was its level of electoral support in the 1995-96 municipal elections for the ADN and the Movimiento Bolivia Libre (MBL, the primary coalition partner of the MNR party during the previous administration). A consistent and highly significant pattern emerges from the data showing a strong negative correlation between electoral support for the ADN and the number of years a municipality received funding. The higher the support for the ADN in the 1996 municipal elections, the less likely it was to receive more than one year of funding from the Social Investment Fund. Conversely, the higher the level of electoral support for the MBL in the 1996 elections, the greater the likelihood a municipality received funding for all three years under study. Figure I below clearly demonstrates these patterns. 54 SIF funding and electoral support, 1996-1998 ;23- : .Q 'n 18 - -4 MBL. > 1~3 13 - = _ .AD cE 0 0~8 0 1 2 3 Years of funding A second finding of the study concerns the deterninants of the per capita levels of funding among those 224 municipalities that received at least one year of funding. With each municipality receiving a relatively equal amount, the data reveal a wide variance in funding levels, ranging from those municipalities that received just over I boliviano per capita to those that received more than 180 bolivianos per capita. Among the most important predictors of a municipality's level of funding, particularly for 1996 and 1997 (the last two years of the Sanchez de Lozada government), were the percent vote received by Sanchez de Lozada's party, the MNR, in the 1996 municipal elections, and whether or not the municipality retained its elected mayor during all three years of the analysis. After controlling for an array of municipal-level social and economic variables, each 10 point increase in electoral support for the 1996 MNR mayoral candidate translated into an additional 6 bolivianos per capita of FIS funds received by a municipality in 1996 and 1997. Likewise, if a mayor managed to stay in office through all three years under study, the municipality received on average t0 more bolivianos per capita than those municipalities where the elected mayor was removed from offce through a town council's "voto constructivo" (a Constitutional provision that allows a town council vote to remove the mayor from office). To summarize, the partisan leaning of a municipality and the stability of its municipal government both appear to have greatly affected the likelihood of a municipality receiving funding from the FIS and, if it did receive money, the amount of funds allocated to it by the program. These findings are strikingly similar to other studies of FIS-type programs that allow for a high level of spending discretion in the name of an enhanced targeting capability by program planners. A study of Peru's FONCODES program found spending patterns for this program greatly affected by electoral considerations. Several studies of Mexico's National Solidarity Program of the early 1990s also produced evidence of a link between program spending and electoral patterns. Based on this preliminary analysis of Bolivia's Social Investment Fund, it appears that Bolivia provides yet another example of the ubiquitous presence of politics in the allocation of public expenditures for programs purportedly designed to enhance the fiscal autonomy of sub-national governments. 55 3.4.9. Coalition dynamics are expected to complicate coordination during the budget process, as parties try to agree on allocation of budgetary resources among competing claims from different agencies, regions, and programs. Given that the approved budget lacks legitimacy (it does not accurately represent the government's true commitment to a set of public expenditure programs), political actors in the budget process naturally see the opportunity to reserve for themselves space to intervene throughout the fiscal year to adjust and re-adjust allocations. 3.4.10. Then, one of the reasons why budgets are not binding in Bolivia is because they do not truly reflect the political consensus on the country's strategic priorities. At this stage, a particular institutional dilemma, called the tragedy of commons, afflicts budgeting for strategic resource allocations. The dilemma is that as each claimant of budgetary resources demands a share, the sum of their demands would exceed the total resources available. Thus, all would be collectively worse off for not achieving strategic prioritization, but each claimant would have a strong incentive to protect his/her own share. Given the dynamics of bargaining among claimants, the best strategy available to each is to over-claim, thus making it more difficult to arrive at a collectively rational budget allocations. The solution requires coordination among key participants in the budget process. For decision-makers to arrive at a binding decision on budgetary resource allocations, information is required on trade-offs that result from particular decisions. Furthermore, there needs to be a single forum, such as a cabinet, where a binding decision can be made with all the relevant information taken into consideration. 3.4.11. Two institutional solutions to the tragedy of commons problem have been suggested. One is to delegate to an independent agent the authority to impose a binding decision, by monitoring activities of the others and, if necessary, sanctioning those who overspend (delegation). The other solution is for all participants with ability to veto budget decisions to negotiate and agree on a set of ex ante fiscal targets (contract). Delegation is more difficult to achieve in a coalition government. A relatively large number of veto players makes it inherently difficult for the veto players to reach an agreement about the extent of delegation. However, if a sufficient number of veto players collectively agree to delegate certain responsibility (e.g., to set and impose fiscal targets) to an executive agent (e.g., MOF), then the delegation approach would be effective. 3.4.12. Indeed this seems to be the case with aggregate fiscal management in Bolivia. The Government enters an agreement primarily with the IMF on the aggregate fiscal targets, and the MOF ensures its compliance through tight cash management control throughout the fiscal year. As mentioned earlier in this report, this is one area where a political consensus has allowed technocratic governance to ensure stable and socially optimal policy directions (i.e., deficit control). While technocratic governance has proven quite effective, its limitation is very clear in the case of budget management: the underlying political consensus only extends to maintenance of fiscal discipline to avoid the macroeconomic instability that seriously plagued the country in the 1980s. 56 3.4.13. A similar political consensus has not emerged to abide by the allocation decisions nominally arrived at in the approved budget. Without the necessary consensus, the delegation approach described above does not work, leaving the government with only the contract approach. Bargaining inherent in the contract approach requires information, and must be made in a legitimate forum to render agreements binding for all participants. However, good information is a scarce commodity within the Bolivian bureaucracy, and decisions on budget allocations are not binding, as evidenced by the common practice of reallocations throughout the fiscal year. Instead, the budget is subjected to a continuous process of negotiation and re-negotiation starting "the day after the budget is approved."80 Negotiations are conducted bilaterally between the MOF and the individual claimant (e.g., line ministry, legislator). The bilateral nature of these negotiations, combined with the fact that they occur not at one given point in time but throughout the fiscal year, makes it virtually impossible for the government to take into account its true strategic priorities in making and re-making spending decisions. 3.4.14. This practice of repetitive budgeting is a common feature in countries with limited budgetary resources and high uncertainty in their availability.81 In the Bolivian case, we see the additional ingredient of coalition politics that makes the negotiation process more politically important and thus more delicate and complicated. In a one-party government with a strong president/prime minister, it should be relatively easy for the center of government to discipline line agencies that make excessive claims on the budget. The ultimate disciplinary measure against a recalcitrant minister would be dismissal from the cabinet post. The head of a one-party government can easily find an alternative minister without jeopardizing the government's survival. In a coalition government, the head of government does not have the same degree of disciplinary authority over members of parties other than his own, and thus the center's ability to force coordination is much more limited. The result is a greater fluidity in budgetary resource allocations throughout the fiscal year, and a resulting lack of accountability in government program implementation. 3.4.15. Fragmentedparty capture and institutionalperformance. Constant and repeated negotiations among multiple actors are a fact of life in Bolivian politics. An elaborate, well-institutionalized system of patronage sharing acts as a glue to maintain coalitions. This situation is best characterized as one of party capture of the state. Since the party system is fragmented, the state that is captured by the parties has also become fragmented. The state's fragmented capture is facilitated by the practice of a political quota system, which produces some positive effects on governance in that it (a) allows for sustenance of the governing coalition through the presidential term, (b) facilitates passage of legislative measures by assuring congressional support by the coalition partners, and (c) as a result of the first two effects, contributes to stability of the democratic regime as exists presently. Unfortunately, this same practice has serious negative effects on the functioning of the state apparatus. First, it is probable that control of different units within the same agency (e.g., vice ministries within a single ministry) by different parties which, despite being coalition partners, do not fully trust each other, so Interview with a GOB budget officer. 81 Naomi Caiden and Aaron Wildavsky, Planning and Budgeting in Poor Countries, (1974). 57 creates serious internal coordination problems. Second, fragmentation of the bureaucracy and lack of internal coordination is likely to make uniform implementation of government-wide rules and procedures difficult. Finally, it is usually (but not always) the case that when personnel selection for the public bureaucracy is made on partisan criteria, considerations of technical competence often take second place. Except in those parts of the state which have been largely protected from partisan dynamics (mainly those that handle macroeconomic management), the captured parts of the state tend to respond more to party or individual political interests than to public goods considerations. 3.4.16. The fragmented nature of the institutional environment is clearly summed up in the results of the survey specially conducted for this Review, covering 738 public officials at different hierarchical levels in 15 state organizations of different types.82 Five measures are used to compare institutional performance of 15 selected agencies, representing different organizational types. The measures are: * Mission accomplished: Degrees to which agency heads perceive their agency's mission to have been accomplished. * Vertical solitude: Degrees to which there is a discrepancy in perceptions about a set of issues between senior managers and middle to lower-ranking staff of the same agencies. This is a proxy for vertical cohesion within the organization. Internal cohesion across hierarchical levels is essential for the agency's operational efficiency. * Corruption: Degrees to which corruption is perceived to be a serious problem in a given agency. - Credible regulations: Degrees to which there is an adequate set of internal regulations in a given agency. * Political hiring: The ratio of persons perceived to have been hired for political reasons in a given agency. 82 The survey was administered to 692 general officials and 46 agency heads. The two groups were asked different sets of related questions. 58 Table 3.1 Performance in public agencies83 Typeof aenc Orgnizaion Mission Vertical Cruto" Credible Political Type of agency Organization accomplished* solitude CoITuption** regulations * hiring*** Autonomous BCB 9.4 26% 1.8 7.0 11% SI-Telecom 5.0 22% 1.8 5.7 21% SI - Banking 10.0 27% 2.1 7.7 0.2% CGR 6.7 28% 0.2 6.8 5% Average 7.8 26% 1.5 6.8 6% Ministries Labor 6.7 36% 7.3 5.0 68% Education 8.9 40% 4.3 5.1 44% Agriculture 4.2 29% 3.6 5.5 58% Finance 8.9 28% 4.3 5.2 38% Average 7.2 33% 4.8 5.2 54% Executive agencies Roads (SNC) 5.8 46% 7.3 5.1 76% Customs 7.8 28% 8.8 5.8 56% Tax 6.7 30% 7.0 4.8 77% INRA 3.3 37% 4.4 6.0 15% TNE 6.7 26% 2.7 6.0 23% Average 6.1 33% 6.0 5.5 56% Prefecturas Santa Cruz 6.7 78% 5.6 4.1 80% Chuquisaca 10.0 67% 6.0 2.9 74% Average 8.3 73% 5.8 3.5 78% BOLIVIA 7.0 36% 4.6 5.6 43% * The higher the score, the better the performance. ** Lower score indicates lower perception of corruption. *** "Political hiring" is not a performance measure but rather one of the determinants of institutional performance. It is reported here to illustrate different degrees to which agencies are subjected to political interference. N.B.: Figures in bold show the agency with the best result for each column; figures in italic show the agency with the worst result. 3.4.17. What stands out from examining the survey data are significant variations in institutional performance across agency types. Not only do a small number of statutory autonomous agencies such as the Controller General's Office (CGR), the Central Bank 8 See Technical Annex for the logic and the method used to construct these indices. 59 (BCB) and the Superintendencies consistently score well in all five dimensions, but the measured differences between this group of agencies and the rest of the public sector is substantial. On the other hand, the survey results underscore institutional weaknesses of the Prefectures in most dimensions. Ministries and executive agencies such as Servicios Nacionales and Institutos Aut6nomos are placed in the middle, though closer to the Prefectures in their performance scores. Particularly noteworthy is the correlation among the perceived frequency of politically-motivated hiring within a given agency, the perceived level of corruption, and the reported degrees of credible regulations in that same agency. 3.4.18. The high degree of vertical solitude also stands out in the Prefectures. The survey results show that in most agencies, the heads and the general officials who work under them share a fairly consistent view about agency performance (average vertical solitude of 33 percent for Ministries and executive agencies, and 26 percent for autonomous agencies). In the Prefectures, however, this measure rises to almost 75 percent. This indicates a striking disconnect between senior managers vis-a-vis middle and lower-level officials. High levels of divergence in the views held by the heads and by the people who work below them may result in poor communication and low employee satisfaction. It is also worth noting that, in light of this high degree of vertical solitude, the Prefectures' scores on "mission accomplished" may not reflect a widely held opinion within the agencies.84 3.5. Summary 3.5.1. The political-institutional analysis presented in this chapter has explored deeper causes of bureaucratic informality in the Bolivian public sector. It has been suggested that important aspects of informal "rules of the game" within the Bolivian state stem from the particular nature of the political environment surrounding the bureaucracy. Although politicization of public administration is by no means unique to Bolivia, this Review' s contention is that the particular patterns of politicization as well as its magnitude are peculiar to Bolivia. Ironically there is evidence that this distinctive pattern of politicization has contributed to the stability in macro political governability, at least since 1985, but that it is affecting performance of public bureaucracy at the micro level. In the long run, this balance between macro political stability on the one hand and poor bureaucratic performance on the other is likely to shift. Unless public sector performance is improved, there is a risk that macro political stability itself may be jeopardized as the state loses its legitimacy for its inability to deliver quality services to the citizens. The current public interest in public sector corruption is a strong sign that people are demanding a more accountable and effective government. 3.5.2. The survey finding reported at the end of this chapter lends support to this Review's argument that political interference is one of the fundamental reasons for poor 84 "Mission accomplished" was obtained based solely on the responses of agency heads; given the high degree of vertical solitude in the Prefectures, it would not be surprising if general officials also held different views on how well their agencies fulfill their mission. 60 public sector performance, including corruption and rule non-compliance, but that the Bolivian political system has given autonomous space with little politicization to a small number of technocratic agencies. This finding suggests that (a) depoliticization of the public bureaucracy is an important reform goal for Bolivia and (b) selective depoliticization has already taken place in Bolivia with positive impacts on organizational development of the public agencies. Thus, depoliticization should be gradually expanded to cover the rest of the state over the medium term. In the next chapter, we will examine some of the Bolivian experiences with institutional reforms with a view to drawing country-specific lessons for politically feasible and effective reform options. 61 4. BOLIVIAN FRONTIERS OF INSTITUTIONAL REFORMS 4.1. Frontiers of Reform 4.1.1. Challenges of institutional reforms facing Bolivia are significant. Even with the best of political intentions, concerted efforts based on careful reflections of past successes and failures are required. In particular, any serious attempt at institutional reform must strike a balance between accommodating present political realities and at the same time gradually restraining perverse influence of politics on public sector operations. For example, an effort to build a politically neutral, professional civil service emphasizes the latter objective, but if it is not politically feasible, the reform will be derailed. Of course, a civil service reform proposal that is politically acceptable to all vested interests may be a toothless one. Similarly, a purely technical solution, such as introduction of a new informatics system, to what is fundamentally a political problem may not offend political interests and thus may be politically more feasible. But, unless the reform also tackles the distorted incentives that give rise to informality, it is unlikely to produce desired outcomes. 4.1.2. Several options are available for reformers intent on curbing negative influence of partisan politics on public sector institutions. One is to insulate parts of the state from political interference by giving them formal autonomy. Bolivia has effectively used this "autonomization/insulation" strategy in some areas including monetary policy inanagement (Central Bank independence), private sector regulation (regulatory Superintendencies), and election administration (National Electoral Court). Unlike in some neighboring countries, the Bolivian political system is capable of generating the necessary political consensus to insulate certain parts of the state from partisan influence. In the case of the reform of regulatory agencies, insulation from political interference in day-to-day management has had clearly positive impact on their organizational development. The regulatory superintendencies are now among the few public sector agencies with robust internal rules and restraints on opportunistic behaviors, and clearly stand apart from the rest of the Bolivian state.85 4.1.3. However, an insulation strategy is not a magic bullet even in Bolivia. A similar approach has so far failed to produce positive results in the Judiciary, whose independence from party politics is yet to be consolidated. Similarly, its effectiveness in freeing the corrupt customs administration or developing an independent Civil Service Superintendency remains to be proven. Another option is to give citizens greater roles in holding the government accountable for its actions. Bolivia has introduced a bold attempt to "bring the state closer to the people" with local government reform under the Popular Participation Law (PPL). The PPL experiences have also been a mixed one, with its impact on improving local government accountability depending heavily on the nature 85 A recent study of effective public sector organizations in developing countries also found autonomy (in the sense of being free from political interference) in personnel management as a pre-requisite for developing an effective public agency. See Merilee S. Grindle, "Divergent Cultures? When Public Organizations Perform Well in Developing Countries." World Development, (1997, pp. 481-495). 62 and strength of local political parties and social organizations. The PPL also suffers from incomplete inter-government institutional reforms, including politicized prefectures and distorted incentives provided by the absence of a national co-financing policy. 4.2. Strengthening Internal Rules and Restraints: Regulatory Superintendencies 4.2. 1. To the extent we identify informality as the principle institutional weakness in the Bolivian public sector, one of the logical solutions would be to strengthen internal rules that restrain opportunistic and informal behaviors within the government bureaucracy. Procedural reforms such as SAFCO and the current civil service reform aim at building institutional mechanisms and capacities for internal restraints. However, our analysis also indicates that as long as public agencies are subjected to the kind and the level of political interference that is observed throughout the Bolivian bureaucracy, it is unlikely that a robust regime of internal rules and restraints would be developed. Therefore, a necessary prerequisite for this approach to institutional development would be insulation of public agencies from day-to-day political interference. This can be accomplished by giving public agencies autonomous status, as has been attempted with the regulatory superintendencies. 4.2.2. Bolivia has made one of the most sophisticated efforts seen outside the OECD countries to construct an efficient, honest, and sustainable mechanism to regulate private activity in financial markets and natural monopolies. Two considerations are critical in institutional design of autonomous regulatory agencies. First, regulatory agencies are prone to "client capture" due to information asymmetry between the regulator and the regulatee, which places the latter in an advantaged position. Second, there is a problem of political interference. Sometimes, political interference may be at the behest of regulated entities - an attempt to use politics to achieve regulatory capture. At other times, politicians and/or other agencies may interfere to achieve their own objectives, which may be to maintain control of patronage or to pursue certain policies (which involve influencing the price of regulated services). In Bolivia, the challenge to regulatory mechanisms is posed at least as much by political capture as it is by the direct, non-political manifestations of regulatory capture. The formal solution proposed for the problem of political capture was to create a substantial degree of political and administrative autonomy, in effect protecting the regulators from the rest of the state. As "home-grown" examples of apparently successful institutional reform to guard public entities against political interference, the experiences of the Superintendencies are of particular interest to this Review. We examine the institutional arrangements and performance of the Superintendencies with a view to drawing possible lessons for the broader public sector. Intemational Bank for Reconstruction and Development FOR OFFICIAL USE ONLY SecM2000-696 FROM: Vice President and Secretary December 8, 2000 From Patronage to a Professional State: Bolivia Institutional and Governance Review (Report No. 20115-BO) Corrigendum The attached pages 63 and 64 should replace those in Report No. 20115-BO entitled "From Patronage to a Professional State: Bolivia Institutional and Governance Review" which was printed with an incorrect Figure 4.1. Distribution: Executive Directors and Alternates President Bank Group Senior Management Vice Presidents, Bank, IFC and MIGA Directors, LAC, Bank This document has a restricted distribution and may be used by recipients only in tie performance of their official duties. Its contents may not otherwise be disclosed without World Bank authorization. 63 Box 4.1 B lAto uRegulatory -Systems In the.absence of a.strong external a aimchanismAhe Bolivian government_sought to - create a degree of self-regulation. Thishasentailedcatthe.very leastthecreation of a second- instance. appeals procedure agais Sulperintendet' .decisions (the Supreme Court being the third- and finalinstance of appeal). PFor some Superintendencies this second-instance a l procedure. is IocSt&itpin GeneraSuperinten deny hiclhalso,has the.takof eei ienr.i3cy and effl!tivenessof the Superintendencies under: it Fbuciaharketr_ ThSu tn 9asandhania tEnties-(SBEF hasexi.sted,in its presentfbrm, sice 1987. A Superorndeo surancehanbe estabIishein-m 1975. in 199k the financial egul sstemasJmntegraoethro6gh the creatinr ofra newE,-? Superintetndency-of Pensions, Securities,-ancLLnsuanc(SPVF) , wich replacedividual superintindencies in each oftthese areas.. -ach Superintendency hears6appealagansi-s decisio,and a Superintenden i-yofAOppaifASR.Tysits above both these S rintendeis sa second-instanceof appeaL(t SR -wascieatdi 99&toeple-aGeneralSpeitedency.- which,-i .addition-to hearing appeals supersd-the twoi ies..A Committee for Financial-Prudetia-Regu_atinC o esfancial-sector-regulatioainongthe,9 Superintendenciei andrelevant parts ofthe eiecutiv& Ney30polareemplo3ed int-. regulof financialentitie. . -V Natur1al dlegal monopohFes. 1994 saw the-t creatonof a System of Sectoral Regulatin (SIRESE)3to regulate the-newly privatized (and some alreadyprivatte)natural and legal monopolies. SIRESE comprises five-first-tier Superintendencies -in transport;-communications (telephones and broadcasting), hydrocarbons, electrici, and water -and a second-tier-General Superiintendency to supervise the first tier (much like.the financial. sector in 1996 to T998) he system is still in its infancy: not all the sectoraf legislation is in place (transprt an-d water are stilL tcome) and the Supeindcies iavenotfall compleiedtheir initial programs. Almost 300-. - r people ^rwo in six eititieo SIf ESE.- - Renewable reso'rces. _In19'95 a GeneraI Superintendenc for Reulaon of Natural Renewable.-. Resources (SRENARE) was created tsupse private concessions o renewSbk natuira - resources with a view-to protecting the enviuonment' promoting biodiversity,-and encourage. the economically rationaLand socially usefulexpleoitaon of newable resources. Following ther - same two-tier logic as, SIRESE, there are Superintendencies. for agriculture and forestry.Zhis - .. systerm is-still in its infancy---. 4.2.3. Internal Organizations of the Superintendencies. As these Superintendencies enjoy substantial political and managerial autonomy, they sit like "islands" within the Executive, in effect as an alternative approach to public-service provision. As "islands" the autonomous regulatory systems have already demonstrated many of the traits of well- functioning bureaucracy that are absent in much of the rest of the state entities such as a much higher degree of career stability and professionalism, and much lower degree of politicization. The survey results clearly demonstrate differences between the two Superintendencies in the sample (SBEF and Telecom) and the rest of the public sector in a number of critical dimensions (Figure 4.1). The staff of the Superintendencies tend to be better trained, better remunerated, and better protected from political interference and 64 job insecurity. The Superintendencies have also developed better procedures for personnel management including merit-based hiring and protection from undue firing. Figure 4.1 Superintendencies vs. Rest of the Public Sector 1 2 0 0 %---- ~----.----.. ....... OS. uperintendencles ERest of public sector j 100.0% 100 OS 80.0% 60 0% 4 0.0% 20.0% 0 0% say salary is have not think they say staff say do not make say jobs are say there is Ratio of non- adequate received are likely to turnover is a corruption is salary publicly hiring by test political salary late lose their serious a serious contributions announced of hiring jobs in the problem problem in to political competence shortterm theiragency parties 4.2.4. The superintendencies: internal environment. A key to the Superintendencies' superior institutional performance lies in their autonomy from political interference. The experiences of the Superintendencies show that even in the politicized environment of Bolivia's public sector, sustained autonomy can build a strong public agency, which is most clearly demonstrated by the Superintendency of Banks and Financial Entities (SBEF), the oldest of all. Over 12 years, SBEF has built a strong cadre of professional bank supervisors through careful recruitment, strong salary and non-salary incentives, the enforcement of behavioral rules, and the development of a strong corporate culture. This regime, along with the design of the supervisory process, has been effective in keeping the agency relatively free of corruption. 4.2.5. SBEF has a formal recruitment process. Candidates are chosen based on public announcements and a competitive entry system based on references and technical and psychological tests. Through these tests a recruiting committee creates a short list of candidates from which the Superintendent chooses. The initial strength of SBEF was based on a similar exercise in the late 1980s in which 30 candidates were chosen from a pool of 800. The advantage of this system is not only to have a systematic approach to finding the right skills and aptitudes, but also to provide an easy way to turn away politically motivated requests to make patronage appointments. SBEF's ability to offer market wages to its staff has clearly been an important factor, though this has recently been challenged by a recent Supreme Decree which limits the maximum salary payable to public sector employees to the salary of the President of the Republic. 63 Box 4.1 Bolivia's Autonomous Regulatory Systems In the absence of a strong external accountability mechanism, the Bolivian government sought to create a degree of self-regulation. This has entailed at the very least the creation of a second- instance appeals procedure against Superintendents' decisions (the Supreme Court being the third and final instance of appeal). For some Superintendencies this second-instance appeal procedure is located within a General Superintendency which also has the task of overseeing the efficiency and effectiveness of the Superintendencies under it. Bolivia now has three such regulatory systems. F;inancial markets. The Superintendency of Banks and Financial Entities (SBEF) has existed, in its present form, since 1987. A Superintendency of Insurance had been established in 1975. In 1998. the financial regulation system was integrated through the creation of a new Superintendency of Pensions, Securities, and Insurance (SPVF), which replaced individual superintendencies in each of these areas. Each Superintendency hears appeals against its decisions, and a Superintendency of Appeal (SRJ) sits above both these Superintendencies as a second-instance of appeal (the SRJ was created in 1998 to replace a General Superintendency which, in addition to hearing appeals. supervised the two first-tier entities). A Committee for Financial-Prudential Regulation (CONFIP) coordinates financial-sector regulation among the Superintendencies and relevant parts of the executive. Nearly 300 people are employed in the regulation of financial entities. Natural and legal monopolies. 1994 saw the the creation of a System of Sectoral Regulation (SIRESE) to regulate the newly privatized (and some already private) natural and legal monopolies. SIRESE comprises five first-tier Superintendencies - in transport, communications (telephones and broadcasting), hydrocarbons, electricity, and water - and a second-tier General Superintendency to supervise the first tier (much like the financial sector in 1996 to 1998). The system is still in its infancy: not all the sectoral legislation is in place (transport and water are still to come) and the Superintendencies have not all completed their initial programs. Almost 300 people work in the six entities of SIRESE. Renewable resources. In 1995 a General Superintendency for Regulation of Natural Renewable Resources (SIRENARE) was created to supervise private concessions of renewable natural resources with a view to protecting the environment, promoting biodiversity, and encourage the economically rational and socially useful exploitation of renewable resources. Following the same two-tier logic as SIRESE, there are Superintendencies for agriculture and forestry. This system is still in its infancy. 4.2.3. Internal Organizations of the Superintendencies. As these Superintendencies enjoy substantial political and managerial autonomy, they sit like "islands" within the Executive, in effect as an alternative approach to public-service provision. As "islands" the autonomous regulatory systems have already demonstrated many of the traits of well- functioning bureaucracy that are absent in much of the rest of the state entities such as a much higher degree of career stability and professionalism, and much lower degree of politicization. The survey results clearly demonstrate differences between the two Superintendencies in the sample (SBEF and Telecom) and the rest of the public sector in a number of critical dimensions (Figure 4. 1). The staff of the Superintendencies tend to be better trained, better remunerated, and better protected from political interference and 64 job insecurity. The Superintendencies have also developed better procedures for personnel management including merit-based hiring and protection from undue firing. Figure 4.1 Superintendencies vs. Rest of the Public Sector 10% 120 0% T 0 0 00 000 ] EiSuperintendencies l 0 <~~~~~~~~~~ 0f f00:;0000J 0:000000 _ Rest ofpublic sector t) f 0; 0 00 00 00.0% 7-= _ - - w1> 40.0% 20.0% 0.0 % say salary is have not think they say staff say do not makn say jobs are say there is Ratio of nor- ad.eqeate recenved are likely to orn-over is a corruption is salary publicly hiring by tesi political salary late lose their serious a serious contribulions announced of hiring jobs in the problem problem in to political competence short term their agency parties 4.2.4. The superintendencies: internal environment A key to the Superintendencies' superior institutional performance lies in their autonomy from political interference. The experiences of the Superintendencies show that even in the politicized environment of Bolivia's public sector, sustained autonomy can build a strong public agency, which is most clearly demonstrated by the Superintendency of Banks and Financial Entities (SBEF), the oldest of all. Over 12 years, SBEF has built a strong cadre of professional bank supervisors through careful recruitment, strong salary and non-salary incentives, the enforcement of behavioral rules, and the development of a strong corporate culture. This regime, along with the design of the supervisory process, has been effective in keeping the agency relatively free of corruption. 4.2.5. SBEF has a formal recruitment process. Candidates are chosen based on public announcements and a competitive entry system based on references and technical and psychological tests. Through these tests a recruiting committee creates a short list of candidates from which the Superintendent chooses. The initial strength of SBEF was based on a similar exercise in the late l 980s in which 30 candidates were chosen from a pool of 800. The advantage of this system is not only to have a systematic approach to finding the right skills and aptitudes, but also to provide an easy way to turn away politically motivated requests to make patronage appointments. SBEF's ability to offer market wages to its staff has clearly been an important factor, though this has recently been challenged by a recent Supreme Decree which limits the maximum salary payable to public sector employees to the salary of the President of the Republic. 65 4.2.6. As a result of the salary cap, the average real wage in 1999 is 83 percent of what it was in 1994, according to SBEF figures. Since 1995, 104 people, 60 of them SBEF- trained professionals, have left the agency. This represents two-thirds of today's total staff. SBEF managers attributes these worrying losses to the salary cap. This "budget battle" is an expression of the ambiguity of SBEF's autonomy given that the Superintendency is part of the executive and subject to the budgetary supervision of the MOF. The battle is not being fought on the issue of control of SBEF's policies - SBEF and the Ministry are allies on this. But the battle does reflect the political problems generated by large differences in the salary levels of the superintendencies and the ministries. 4.2.7. SBEF is dominated by formalized, centralized procedures designed to avoid conflicts of interest and promote a high standard of ethical behavior. These rules deal with conflicts of interest inside the institution and in dealings with the banks.86 Staff are aware that transgressions will be punished, but serious transgressions are thought to have been minimal: over the last 10 years, only two people have been dismissed for mnisbehavior (both for providing internal reports to banks). SBEF staff credit these rules with protecting them from the informality - the susceptibility to family and friendship ties and to corrupt acts - that they see as characteristic of the ministries. SBEF's policies to professionalize the agency have led to the development of a strong esprit de corps. Staff are typically proud of their agency - its unique technical skills, and its achievements. They appreciate the respect they receive from the banks they supervise. 4.2.8. In contrast to SBEF's relatively long existence, the Sectoral Regulatory System (SIRESE) is still in the process of establishing itself. Nonetheless, a nascent professionalism is evident in the System. The Superintendencies are small - the largest, telecommunications, has 80 employees. Personnel regimes differ from agency to agency. The Superintendencies use a mix of direct invitation or advertising to solicit candidates, then a committee-based selection process, typically on the basis of technical and psychological selection criteria. As with the SBEF, the committee- and criteria-based process helps the Superintendencies to deal with attempts to use political patronage to secure jobs. 4.2.9. Operating under private labor law, the Superintendencies tend to set salaries at market rates and to emphasize non-salary elements, such as early promotion, training, and travel, to broaden the package of incentives. The Superintendencies also tend to cultivate a strong corporate culture of apolitical professionalism and corporate loyalty and teamwork. An informal approach to cultivating these attitudes seems to play a larger role than more formal approaches such as ethics rules. This approach has so far been S6 These rules are now being developed into a formal code of conduct. 87 All the agencies have to deal with similar problems of asymmetric information and the capture of their employees by their operators. In this, they appear to face the common situation of having to select their employees from small, specialized labor markets where, typically, the operating companies are the largest employers. This makes competitive recruitment difficult. 66 effective in creating a reputation for professionalism and a bulwark against outside political or family influences in several of the Superintendencies. 4.2.10. SIRESE nonetheless faces the same challenge to its autonomy from the government as does SBEF: the government is using its budget-setting powers to control salary levels in the Superintendencies. Some people within the Superintendencies see the government as stepping beyond its powers, but the conflict reflects the broader conflict in the public sector over privileged islands of highly-paid, often donor-supported workers. The government's policy is undermining the competitiveness of salaries at the most senior levels and, in some cases (notably Hydrocarbons) at technical levels. 4.2.1 1. The External Environment. Key to understanding the source and nature of the Superintendencies' autonomy, and its likely sustainability, is the external environment that surrounds them. In Latin America, history of short-lived "islands of excellence" has repeated itself in a number of countries, and Bolivia is by no means immune from this same problem. The dominant tendency is that autonomous agencies flourish as long as they enjoy political (usually presidential) support, but they lose autonomy and effectiveness as political masters change. 4.2.12. Formally, SBEF enjoys substantial autonomy. As part of its managerial autonomy, SBEF is allowed to levy a certain percent of the banks' assets and contingencies to pay for its running expenses. The Superintendent's decisions are subject to a three-instance appeals process. The first level of appeal is carried out within the Superintendency. The second is carried out by a specialized body, an Appeals Superintendency (SRJ) which hears the second-instance appeals of both financial-sector Superintendencies (i.e. SBEF and SPVF). The Supreme Court provides the final level of appeal. In practice, these legal arrangements provide effective defenses against political interference by the three branches of government, but they by no means banish it. SBEF has also faced some opposition from some of the banks - those banks whose ability to make profits has been most threatened by the application of prudential rules. These banks have 'often been adept at political manipulation and effective use of different branches of government to press their cause.8' Thus far, SBEF has defended itself successfully from these pressures thanks in part to strong political support from the Central Bank, MOF, the President and some international donors. 4.2.13. The five Superintendencies in the SIRESE system and the General Superintendency that oversees them enjoy a legal autonomy very similar to that of SBEF. The President selects the Superintendent and General Superintendent from three names submitted by the Senate on the basis of a two-thirds majority vote. The Superintendents serve for five years - seven for the General Superintendent - and can only be removed, for misconduct, by the Supreme Court. The Superintendencies are funded by a statutory 88 The most notable case concerns SBEF's liquidation of a bank in 1997. The bank's principal owners used their positions in Congress - one was elected a deputy in 1998 - to conduct a campaign against the Superintendency. They also used other instruments, including suing a number of the senior officers of SBEF, who did not enjoy legal protection as SBEF staff because the law only protects the Superintendent from litigation against him for acts conducted in the course of carrying official duties. 67 maximum levy on operators' turnover varying from 0.8 to three percent. Similarly, the Superintendents' decisions are subject to two instances of appeals beyond the first level of appeal within the Superintendency. The second level is carried out by the General Superintendency of SIRESE. The Supreme Court provides the final level of appeal. The General Superintendency also has oversight of the efficiency and effectiveness of the five Superintendencies under it.89 These formal arrangements providing SIRESE with autonomy have provided a basis upon which the system can build its legitimacy, but there is still a long way to go in this. The Superintendencies have to face an executive and legislature that are often unhappy with their powers and privileges and regulated companies which are not all inclined to cooperate. 4.2.14. The pressure on the young Superintendencies can be seen from the instability in their leadership, in some contrast to SBEF. The first Superintendents of the five Superintendencies were named between the end of 1995 and the middle of 1997. There were several changes in leadership before the change of government in 1997, and all Superintendents, except the General Superintendent, changed with the new Government. Moreover, some of the appointments have been interim, thereby allowing the Superintendent less autonomy. Some of the appointments have been of political allies, though in general Superintendents have also been named for their technical expertise. 4.2.15. The high turnover of Superintendents around the time the new government came to power shows the fragility of the formal rules in the face of the weapons of politics. One of the Superintendents was found by the Supreme Court not to have the professional title required by the SIRESE law by virtue of the fact that his foreign degree had not been registered in Bolivia. Yet it is clear that the rules are changing: the government has not in general tried to extend its powers of patronage to the staff of the Superintendencies or to intervene on behalf of operators. Nonetheless, there have been some frictions between the Superintendencies and the Ministries or Vice-Ministries under whose "tuition" they fall. The issue is often one of power and the line at which the Ministries' policy making ends and the Superintendencies' policy implementation begins.90 4.2.16. Bolivia's new regulatory mechanisms, as exemplified by SBEF and SIRESE, represent a remarkable achievement. These mechanisms are young and at various stages in the incomplete process of political legitimization. Still their design promises an effective approach to avoiding regulatory capture, both by regulated entities and by the rest of government or politicians, and thus represents some of the rare instances of effective organizational development in Bolivia. Protected from frequent political interference, the Superintendencies have managed to develop rather effective systems of internal rules and restraints to limit informality that pervades the rest of the public sector. 8 The financial system has no such similar function. Some have argued that combining general supervision and a second-instance function creates a conflict of interest; others claim that abandoning this general supervision encourages laxness and that the Auditor General does not fulfill this function. 90 Generally, the MOF, with its technocratic interest in market-led growth, has proven an ally of the Superintendencies (as it has of the financial-sector Superintendencies). The Ministry has, however, faced some problems with the Hydrocarbons Superintendency because the government's fiscal-policy interests have been at odds with the Superintendency's regulatory interests. 68 4.2.17. Remaining Challenges to Legitimizing the Regulatory Systems. Much of the thrust of institutional reform has been to protect the regulatory systems from the rest of the state and to replace this with some form of self-regulation. But this poses the question: to which external actors are the systems accountable? The systems have some accountability to the executive, the CGR, and the judiciary. These accountability mechanisms are generally limited: they may be designed to control various forms of misconduct or departure from formal procedures, but they do not address the effectiveness and efficiency of the systems. The public has so far played little role in demanding standards of performance. The most important external agent addressing effectiveness and efficiency issues is, informally, the World Bank, which supports institutional development of the superintendencies through ongoing projects. Without more formal accountability procedures, we may very well expect a deterioration in the performance of the system as time progresses and the initial impetus of reform wears off. 4.2.18. Autonomy may thus have short-term advantages, but pose long-term problems. In the long term, as administrative reform proceeds in the executive and judicial branches, the Superintendencies and the rest of government may have to find a way of re- establishing closer accountability links. Administrative modernization that improves conditions of public sector employment in general may also tackle another political obstacle to legitimizing the regulatory systems: the resentment generated in large parts of the executive by "islands" of high pay. 4.2.19. Implications for broader public sector reform. Large-scale replication of the Superintendency model in the entire public sector is neither possible nor desirable, because of a fallacy of composition that limits the extent to which autonomous agency models can be extended: the larger the sector of autonomous agencies of the Superintendency type, the greater will be the accountability problem and the likely abuse of autonomy. Nevertheless, the examples described so far in this chapter clearly demonstrate that strong organizations can grow within the Bolivian state given autonomy from political interference and adequate resources. In this sense, the most important impact of the Superintendencies on more generalized administrative reform may lie in the demonstration effect: the Superintendencies are demonstrating that it is possible to have a professionalized public service as long as external (political) conditions and internal administrative incentives and rules encourage this. Based on these experiences, potential broader but selective application of the model should be considered as an important option for institutional reform in Bolivia. 4.2.20. It is important to bear in mind, however, that in terms of specific institutional design, successes in autonomous regulatory agencies may not be directly applicable to similar attempts to make service delivery agencies independent of political interference. There is a critical difference between institutional characteristics of regulatory agencies and service delivery agencies, which give rise to fundamentally different motives for "autonomization" and which may make it easier on balance to create and maintain autonomous regulatory agencies than autonomous service delivery agencies. Although de-politicization is the most immediate objective in both cases, at least in Bolivia and in many other developing countries, once that objective is achieved, the two types of 69 agencies will require different accountability arrangements. Autonomous regulatory agencies (including those which enforce internal public sector regulations such as the Controller) are deliberately set up outside the hierarchical structure of the government to protect them not only from interest group capture but also from capture by the government in power (and political parties behind it). Their accountability arrangements often depend more on the abilities of the Legislature and the Judiciary to intervene ex post than on a hierarchical arrangement of control within the Executive branch. In contrast, service delivery agencies will inevitably have to remain connected to the government's hierarchical structure for purposes of control and accountability even if they are given a degree of autonomy and flexibility to improve their operational efficiency. No matter how much operational autonomy a service delivery agency may acquire, it is unrealistic and undesirable to expect it will be completely free of ministerial control. Effective functioning of semi-autonomous service delivery agencies would require an ability to manage and enforce a quasi-contractual arrangement between the agency and the parent ministry (e.g., the Servicio Nacional de Caminos and the Ministry of Economic Development). This is a capacity that is scarce in Bolivia's public sector. 4.3. The Insulation Strategy Without Consensus: Judicial Reform 4,3.1. If the experiences of the regulatory superintendencies demonstrate positive potential of the insulation strategy for institutional development in Bolivia, the record of the judicial reform, which has also aimed at strengthening judicial independence from partisan-political influence, exemplifies difficulties of this same approach. A clear lesson is that when political consensus is weak or absent, the insulation strategy is unlikely to succeed. 4.3.2. Lack ofjudicial independence and conflictivejudicial reform. Through the first decades of the present century, Bolivia's justice sector developed much like that of the rest of Andean South America. Under the civil code tradition (mixture of Spanish and French influences) the judiciary was the third, but weakest branch of government, with a role confined to conflict resolution and rule enforcement. Despite a 1931 constitutional amendment reconfirming judicial independence, there were sixteen massive purges of the judiciary (generally with each new government) between 1936 and 1982. This broke with the elite tradition and made courts more openly subservient to the specific interests, partisan identifications, or ideological leanings of the government in power. With the opening to democracy in 1982, the multi-party system and shifting government coalitions increased competition for control of public sector employment and decreased job stability for office holders. Although judges were selected by constitutional means for the first time in decades, they were no exception, and the parties quickly developed means for dividing up the judicial positions. This is a clear example of negative impacts of the patronage practices on a public sector institution that have been intensified with the democratic transition. 9' Most of Bolivia's constitutions have recognized judicial independence, but in the early years judges, despite their "permanent," tenure were paid by the Executive and replaced with some frequency. The 1880 Constitution introduced term limits, ten years for Supreme Court Justices. (Gamarra, pp. 22-3.) 70 4.3.3. Against this historical backdrop, Bolivia initiated judicial reform in the mid 1 990s. Though a late-comer in the region, it has moved rapidly, and thus without substantial national discussion or a broad-based consensus either on the need for change or the form it should take. The reforms themselves are typical of those being implemented elsewhere: the creation of a Judicial Council, Constitutional Tribunal, Public Defense Office and Human Rights Ombudsman, and the rewriting of the substantive, organizational, and procedural codes shaping the sector's operations. The 1993 Constitutional amendments were followed by several years of delays as the necessary enabling legislation was drafted and the selection of the new bodies was argued in Congress. Only the new Ministry of Justice and the Public Defenders Office were created immediately. In the case of the council and tribunal, opposition from the Court, which would lose the powers they gained, posed a further obstacle. Additional delays derived from the Court's loss, through retirement or resignation, of all but five of its 12 Justices. This also impeded the selection of lower level judges to fill vacant positions (by March 1999 numbering 42 district court judges, 147 trial judges and an additional 100 judges whose period of office had expired) and delayed appointments in the Fiscalia. 4.3.4. Important progress was achieved in more recent years. In March 1999 with the selection of the missing members of the Supreme Court and Constitutional Tribunal, the impending election of a new Supreme Court President, and the passage of the Criminal Procedures Code, slated to enter effect within two years, the country appeared to be moving out of the worst of its institutional crisis. However, having new organizations staffed and laws approved is no guarantee that they will work according to plans, and it is evident that the transition will not be automatic. 4.3.5. The Judicial Council, fully staffed and operational since early 1998, is plagued by internal conflicts and external criticism. Its involvement in the pre-selection of candidates to the Supreme Court has been criticized as too proactive (by the head of the Congressional Constitutional Committee) and not proactive enough (by a majority of observers including the Minister of Justice). Similar debates surround its selection of candidates for the nine Superior Courts and the Supreme Court's subsequent choice of the new judges. The members of the Council are themselves divided as to what their role should be, and one of them has charged both the Council and the Court with succumbing to personal and political pressures. This kind of conflict is unprecedented in Latin America and thus indicative of the lack of consensus behind the Bolivian reforms. The further consequences are that the Council has been unable to advance such key activities as the creation of a training institute, the presentation of a career law, the rationalization of the judicial administrative structure, and the vetting of candidates for the vacant judgeships. Until the Constitutional Tribunal is fully operational, questions having to do with the Council's powers cannot be answered authoritatively, thus encouraging the continuing battles among its members and with outside actors. In summary, although the country has begun an ambitious restructuring of its justice sector, the reform is still too young to show consolidated results. The process suffers from many conflicts, which are indicators of weak political consensus among key players in the reform. 71 4.4. Bringing the State Closer to People: Popular Participation and Local Accountability 4.4.1. Another approach for restraining negative influence of politics on public administration is to empower the beneficiaries of public services as agents of accountability. In a number of places, giving people "voice" has proven effective in disciplining governments, making them more responsive to citizens' needs and improving efficiency of their service delivery.92 The local government reform brought about by the 1994 Popular Participation Law (PPL) explicitly aimed at strengthening citizens' ability to oversee actions of the municipal governmnents and hold them accountable. By bringing society and the state closer together, the reformers believed municipal government would become more effective in articulating people's needs and in delivering public services.93 This final section aims to determine the degree to which one of the prime objectives of the reform - improving local government effectiveness through community participation and oversight - is being achieved, and indirectly assess the potential of the "voice" option in Bolivia's institutional environment.94 4.4.2. The Popular Participation Law. Trhe Law aimed at bringing the state closer to people by establishing municipal governments in rural as well as urban areas, where none had existed, and by setting up a series of institutional mechanisms to allow citizen oversight. Such mechanisms include: * Regular election of town councils, separate from national elections, for five- year terms of office (the length of the first term under the new system is four- years, 1996-2000; the five-year term is established in the Constitution).95 Popular elections of mayors and members of municipal councils provide the most basic means of citizen control through electoral accountability. The council can remove a poorly-performing mayor from office by "constructive vote of censure" (voto constructivo) with three-fifths vote (with the exception of those elected directly by majority vote) after at least one year in office; * The institution of a process of public participation and social control over the municipal investment decisions through the establishment of 92 A well-known example is the so-called "report card" method used in Bangalore, India. For details, see Samuel Paul "Making Voice Work: The Report Card on Bangalore's Public Service." World Bank Policy Research Working Paper No. 1921, 1998. 93 Merilee S. Grindle, Audacious Reforms: Institutional Invention and Democracy in Latin America, draft manuscript, October 1998, forthcoming, 125. 94 The assessment is based on a field study of the impact of the PPL in four municipalities of various sizes-- Luribay, the city of Oruro, Padcaya, and La Guardia--located in four of the country's diverse departments-- La Paz, Oruro, Tarija, and Santa Cruz, respectively. Each municipal government presented a distinct set of political and socioeconomic circumstances. Short visits to four municipalities are not, of course, sufficient to draw firm conclusions about the state of Bolivian local government generally. We considerably broadened the investigation to an additional 38 municipalities, where we conducted surveys of local officials about their municipal government. Likewise, we extended the study to the departmental level with the completion of surveys in six of the seven prefectures in which the sample of municipalities is located. 95 The lead candidate of a party receiving an absolute majority of the vote is automatically elected mayor; otherwise, the town council selects one of the two top vote-getters as mayor. 72 "vigilance committees." Vigilance committees have the mandate to publicly issue a statement, which is to be published in the media and transmitted to the national government, on how co-participation funds are budgeted and invested. If the vigilance committee finds a problem with the investment of those resources, it may file a complaint to the national executive, which is to investigate and attempt to resolve the issue. If the executive fails to secure a remedy, it may petition the national Senate to consider the case and suspend further disbursement of coparticipation funds; * The establishment of a series of rights and duties of territorial base organizations (OTBs), including the ability to propose, request, and control the performance of public works and provision of services. The OTBs themselves are legally required to render account to the community of all actions taken on its behalf. Finally, the OTBs are responsible for electing leaders to represent them on the vigilance committee; and * The introduction of a participatory investment planning process through development of an annual operative plan (POA). 4.4.3. Administrative Decentralization. One year after the Popular Participation Law, the Administrative Decentralization Law was passed to re-define the role of the departments in the newly decentralized state structure. In the new arrangement, the deliberative and oversight functions of the department are carried out by the departmental council. Council members are elected from each province by a vote of the province's town councilors. They have the legal mandate to approve the departmental budget and a report on how those resources are expended; approve development projects and public investments and oversee their execution; and they are required to authorize contracts for investment project and public services. Generally, however, the councils are viewed as weak entities with little real power to take initiative. Among their difficulties is that the prefect presides over the council presidency. The council does not have a budget of its own, and members have no way to hire technical support. 4.4.4. Findings of the four municipal case studies. The increased "voice" developed by the PPL brought with it new opportunities for citizens and political parties. By devolving responsibility to local officials for designing and implementing projects and charging them with using a participatory planning process, the PPL creates a space where greatly improved municipal development may emerge. It also opens a window for the opportunistic patterns of political party behavior at the national level to be recreated in the municipalities. The oversight and participatory instruments created by PPL - the vigilance committees and the OTBs - sometimes become attractive targets for political parties and interest groups to co-opt, thereby weakening accountability and transparency. 4.4.5. The evidence from the four case studies conducted for this Review points to a conflict-ridden struggle to gain partisan control over municipal resources. Three 73 manifestations of this conflict are: elected officials' decisions about how to distribute coparticipation funds and what projects to carry out with the money; the oversight and participatory roles played by the vigilance committees and the OTBs; and the use of the voto constructivo. * In three of the four municipalities the bulk of coparticipation funds were spent on high-profile projects in heavily populated areas. Only in one case, Luribay, was a major investment of resources undertaken that clearly benefits a broad area of the municipality. * Vigilance committees' oversight functions were subverted by either capture (Padcaya) or circumvention (Oruro and La Guardia) by politicians and interest groups. In the latter cases, mayors established relations directly with the OTBs, bypassing the vigilance committee. In many cases, it appears that the OTBs are also instruments of political or corporate interest, and are not effective at demanding accountability. * The censura constructiva appears to be abused to overturn political rivals in a grab for power, often in connection with the national party structure. This is evidenced by the fact that several mayors (including the mayors of Padcaya and La Guardia) were removed when ADN came to power at the national level in 1997. 4.4.6. Findings of the Survey. To obtain a broader perspective of the municiVal system, we conducted semi-structured surveys in thirty-eight additional municipalities. 6 This sample included local governments, again of varying size, in seven departments: La Paz (six municipalities), Santa Cruz (six), Beni (six), Chuquisaca (three), Oruro (four), Cochabamba (six), and Tarija (seven). The surveys were aimed at gathering additional information that might affect our initial findings and at providing further insight into the progress of Popular Participation and the performance of local government (see table X, below, for selected survey results). Key findings include: * High turnover of mayors. The 38 municipalities have had an extraordinary total of 92 mayors in the first 3.5 years of the 1996-2000 term. That amounts to an astonishing average of 2.4 mayors per municipality. More than a quarter of our municipal sample had had a new mayor every year of the four-year term. * Instability in council membership. Some 269 councilors served in 37 of the 38 municipal governments surveyed. Had there been no turnover, only 198 96 For both the municipal and prefectural surveys, which were conducted in July and August 1999, some of the questions involved simply collecting factual information from the municipal governments. Otiers were open-ended questions, the answers to which required, after interviewing at least two individuals (usually many more), some subjective assessment on the part of the researcher. 74 originally-elected councilors would have served. An additional 71 councilors - or 36 percent more - served on the local councils in our sample. Strength of local organizations. Strong local organizational life tends to be associated with, but certainly does not guarantee, civil society's influence in the municipal investment decision-making process. And a considerable portion of all Bolivian localities have well-organized, local campesinos unions and other groups, such as the coca farmers. Some 13 of 38 of the municipalities we surveyed were characterized by such organizations, and, generally, they were considered quite involved with the municipal government to ensure their interests are protected under the terms of the PPL. Eight of our surveyed municipalities were characterized by few or no organizations; some 17 of 38 evidenced moderately organized communities. * Civil committees' limited roles. Local civic committees appear to have little impact at the municipal level. Only 3 of 38 municipalities had civic committees that were rated strong and influential in municipal affairs (and only one of these was a civic committee in the traditional sense; two others were a women's civic committee and a civic action group linked to the mayor). Twenty-one of the 38 municipalities had civic committees that were rated active, but with no influence on decision-making in municipal affairs. The remaining 14 municipalities - more than one-third - either had civic committees that were weak (in 8 municipalities), in some cases barely existing, or their committee was nonexistent (in 6 municipalities). Thus, except in rare cases local civic committees are not centers of municipal influence. Power resides in the mayor and in other organized groups in the community. Civic committees often have been politicized, they face competition from legally mandated vigilance committees, and important members, such as the private sector, may no longer participate in them. This contrasts sharply with departmental civic committees, which are considerably stronger and more influential. * Limitedpower of vigilance committees. From 1996 to date, 12 vigilance committees of the 38 in our sample - roughly one-third - attempted to freeze the transfer of coparticipation funds under the provisions of the PPL. Two of the 12 made two attempts, so a total of 14 attempts were made. Only three were successful and one case was pending.97 This 37 percent (14 of 38 cases) is a large percentage of petitions, tending to confirm what we have seen in our field cases. Elected officials and committee members have difficulty reaching a consensus on budget issues. Relations between the mayor and vigilance committee were considered good in 37 percent (14 of 38) of the municipalities. Yet, there was at least some tension or cooptation of the 97 The Ministry of Finance, a bank, and a deputy (unsuccessfully) were involved in freezing funds in two municipalities (3 occasions) in our sample. 75 committee in 61 percent of them (23 of 38). In 29 percent (1 1 of 38), mayor- committee relations were reportedly poor. Bypassing of vigilance committees by mayor. When a mayor has problems working with the vigilance committee, he or she will attempt to plan investments without it. Mayors will bypass the committee members, who are chosen representatives of the OTBs in each local canton or district, and instead work directly with OTBs. Twenty-three out of 38 municipal governments were reported to be working with both the vigilance committees and OTBs. But when a mayor was working more closely with one or the other (in 11 cases), it was almost invariably (with one exception) with the OTBs. Box 4.2 Selected Municipal Survey Results Mayoral Turnover. Question: Since the last elections in 1995, how many times has the mayoralty changed hands? In 13 of the 38 sample municipalities, the mayoralty had not changed hands at all; the same mayor had governed for the entire period. Eight of the 38 local governments were on their second mayor, while 7 of 38 had had three mayors. Seven of the 38 municipalities were on their fourth mayor since 1995. while 3 of 38 had had five mayors. Council Turnover. Question: Since the last elections in 1995, how many councilors have served on the town council, including the councilors serving at present? Some 269 councilors had served in the 37 municipalities in which this question was answered (the council composition in one municipality could not be determined). In only 8 of 37 had there been no change in the council's composition. Most of the councils had witnessed the replacement of from one to three of the original councilors. Mayor-Vigilance Committee President Relations. Question: How would you describe relations between the mayor and the president of the vigilance committee (Choose all that apply)? Fourteen of 37 municipalities reported that relations between the mayor and vigilance committee were good (one vigilance committee was too new to report). Twelve of 38 reported that relations were fair with some tension. Eleven of 38 reported that relations were poor. Suspension of Municipal Coparticipation Funds. Question: Has the vigilance committee attempted to have the municipal government's coparticipation funds suspended by petitioning the Senate? From 1996 to the date of the survey, 12 vigilance committees of a total of 38 in our sample attempted to suspend the central government's transfer of coparticipation funds as provided under the PPL. Two of the 12 committees made two attempts, so a total of 14 attempts were 76 made. Of those 14 attempts, 3 were successful and one was pending. Developing Annual Operative Plans. Question: In developing the annual operative plan (POA), how would you describe mayor-council relations vis-a-vis the vigilance committee and OTBs? Twenty-three of 38 municipal governments were reported to work with both the vigilance committee and the OTBs in preparing the POA. Ten of 38 reported working more closely with the OTBs than the vigilance committee. Only one was reported to work more closely with the vigilance committee than the OTBs. Four of the 38 were reported to work with neither or to have been unable to develop the POA. Mayor-Prefect Relations. Question: How would you describe relations between the mayor and the departmental prefect? Seventeen of the sampled 38 mayors were reported to have good relations with the prefect. Fourteen of 38 were reported to have fair relations with the prefect, while 7 of 38 reported that they were poor or non-existent. Distribution of Investment of Coparticipation Funds. Question: One of the sources of considerable tension at the municipal level (among the mayor, council, vigilance committee, and community organizations) is the local government's inequitable distribution of coparticipation funds. On a scale ranging from non-existent to high, what is the level of tension created your municipality by such distribution? This issue was reported to have created low levels of tension in 18 of 38 of our sample municipalities. It had created moderate tension in 16 of the 38 municipalities and high tension in 4 of them. Local Civic Committee Presence. Question: On a scale ranging from non-existent to strong, how is the presence of the local civic committee in municipal affairs besttdescribed? Only 3 of 38 sampled municipalities reported having civic committees that were a strong influence in municipal affairs (and only one of them was a traditional civic committee). Twenty-one of the 38 had civic committees that were active, but not decisive in municipal decision-making. The remaining 14 municipalities reported having civic committees that were either weak (8 municipalities) or non-existent (6). Strength of Local Community Organization. Question: On a scale ranging from non- existent to strong. how is the social organization of the municipality best described? A total of 13 of 38 sampled municipalities were reported to have strong social organizations, with considerable influence in local elections and municipal affairs. Seventeen localities were considered moderately organized, with some organizations and activity and little influence or involvement in municipal decision-making. Eight of the 38 municipalities reportedly had weak community organization, with few or no organizations and little activity. Nature of Municipal Investment. Question: How is the nature of municipal government investment in the community best described? Thirty of the 37 municipal governments sampled (we received no response from one 77 municipality) reported spending most of their resources on providinig new or improving public services, including: education in particular, health, sanitation, and sports. Twenty- nine of the 38 municipal governments reported focusing investment in infrastructure, especially roads, but also urban infrastructure. Three municipalities reported investments in agricultural production, including irrigation. One mentioned investment in "institutional development." 4.4.7. Prefectures as institutional bottlenecks? Relationships municipalities have with their respective departments can have important implications for municipal government performance. The prefect, who can be a direct line to ministries and resources in La Paz, if not the president, and a good relationship, signifies benefits in the form of project financing and technical assistance to the municipality. On the other hand, opposition mayors tend to be isolated from the departmental administration or to receive minimal support. Similarly, departmental councilors who have good relations with the prefect can work on behalf of local elected officials; opposition councilors have little recourse. When assessing the local system as a whole, then, it is clear that party division in the Bolivian system negatively affects municipal performance. 4.4.8. Our survey showed signs of the politicization of intergovernmental relations. Most of the six prefectures examined - all of which are ADN or ADN allies (in one case) -- reported having close contact only with municipalities led by the ADN or alliance parties. On the municipal side, 13 of 17 mayors who were reported to have "good" relations with the prefect represented the ADN or parties allied in the national coalition. Virtually all of the 7 mayors who reported "poor" relations with the prefect indicated that party divisions explained the nature of the relationship. 4.4.9. These findings are all the more understandable when one considers the degree to which, in the Bolivian system, the party or governing coalition controls the institution on the party's behalf. Our survey results found, for example, strong political party clientelism in the hiring of departmental personnel - 5 of 6 prefectures reported that the governing parties play a major - equal or secondary, and in one instance, greater - role in deciding who is to be hired and fired. Job stability and merit-based promotion do not exist: clientelism is theformal mechanism of the system. Hence, efficiency-related and management improvement cannot occur, and the attendant effects reverberate at the local level. 4.4.10. Citizen participation at the departmental level is weak. The departments are expected to be part of the new participatory process, consolidating the municipal planning process for the establishment of a regional development plan, registering OTBs, and serving as an intermediary for municipal governments and community residents. This does not appear to be occurring to a significant extent. In our surveys, 4 of 6 prefectures reported that they had not taken steps to promote participatory governments. Only 2 had been engaged in participatory workshops for the creation of development plans. In Santa Cruz, where nongovernmental organization is the strongest in the country, the influential civic committee was closely involved in designing and executing the planning strategy. 78 4.4.11. Explaining the Performance of the New Local System. To date, it appears that the results of municipalization under the Popular Participation framework are mixed. The same can be said for the specific objective of building local accountability: clear progress can be found alongside serious limitations in the system. The following identifies some of the factors responsible for the relative successes and failures of the PPL. 4.4.12. Party influence. The system created by the PPL reforms contains several weaknesses. Popular Participation is less likely to reach its potential, and sometimes breaks down completely, where the political parties, as opposed to rural unions and other organizations, have a strong presence. In comrnmunities politicized along political party lines, especially where nonparty organizational life is weak, loyalties to the parties tend to take precedence over the broader interests of the municipal institution and the community as a whole. In largely urban municipalities, and in the large departmental capitals like Oruro, vigilance committees must fight to exist and carry out their functions against the party militants who have taken control of many of the societal organizations.98 4.4.13. Mayor and councilor turnover, especially the former, and the party politics that generate it are obstacles to achieving greater public accountability and better municipal performance.99 In practice, the censure procedure for removing "incompetent" or "corrupt" mayors is little more than a political tool used by local officials to win power for themselves and their parties. The procedure breeds corruption (payoffs to councilors, etc.) and disillusionment with the democratic process. '00 4.4.14. Weak vigilance committees. Generally, the vigilance committee is highly suspect in serving as a mechanism for municipal accountability. In each of the four cases assessed here, the vigilance committees functioned poorly or not at all. In only one instance, La Guardia, was the "committee" functioning with a member other than the president. Committee members frequently do not participate because they receive no pay; they receive no reimbursement for travel and food costs incurred; they often lack the technical training to assess municipal budgets; and they often, as we have seen, are refused the information they need to carry out their responsibilities. Members who are on the best terms with the mayor, meanwhile, are likely to have been co-opted."0' Although OTBs have considerable authority under the Popular Participation Law, they, too, are 98 Ruben Ardaya Salinas, "El comite de vigilancia al auxilio de la democracia municipal," Descentralizaci6ny participaci6n, numero 3, Friedrich Ebert Stiftung and ILDIS, 1998, 28. 99 Oruro is the obvious example. In Oruro, seemingly every issue--from who held the mayor's seat and how coparticipation funds were spent to how the vigilance committee was operating - was determined along party lines. The multitude of interests in the city appeared largely captured by the parties. In such an environment, the mayor was accountable, in a sense, primarily to his political clients, providing support to them--favors, jobs, projects, etc. - in return for their support on election day. '00 For an excellent study of the voto constructivo in municipal governments elected after 1995, see Gonzalo Rojas Ortuste, "Censura constructiva, inestabilidad, y democracia municipal," Descentralizaciony participaci6n, riunero 1, Friedrich Ebert Stiftung and ILDIS, 1998. 101 Bolivian analysts of the local system well recognize the need for reform of the committees. See, for example, Ardaya Salinas, "El comite de vigilancia." 79 subject to politicization and manipulation. In the Bolivian context of informal rule compliance, they, like the vigilance committees, are often effectively ignored. 4.4.15. Ineffective participatory planning Finally, Popular Participation's planning process is not working well as an accountability mechanism, although municipalities are developing their annual and five-year plans. It is difficult to get people to participate, and the outcome of community planning meetings is usually much more a listing of projects than any type of planning strategy. Local officials manipulate the process and do not follow the plans, and so far there is little institutionalization of municipal planning capacity. The five-year POAs were developed or are being developed by outside consulting agencies in three of the four municipalities studied here. Nongovernmental organizations and international donors are providing strong, important support for the development of municipal planning and management capacity. 4.4.16. Community empowerment. There are, of course, some positive developments as well. In their localities, Bolivians recognize that the decision-making power in the hands of municipal officials is real and have pressed their demands through OTBs and vigilance committees, despite the problems. There is a degree of community representation that was weak or nonexistent in the past. "To the extent that one of the principal concerns of the [community] leaders is responding to and obeying the grassroots," one analyst writes, "this concern is communicated to and produces demands on municipal authorities." 102 We saw clear evidence of this in the sample of localities with moderate-to-strong organizational life. As a result, every mayor faces pressure to distribute resources equitably. Mayors list investment projects in their budgets according to their location in the municipality for good reason: communities view these resources as their property and they look to see how they are spent. A new, obviously imperfect type of constituent- representative relationship, one characterized by some measure of accountability, may be emerging.'03 4.4.17. Improved electoral accountability. Electoral accountability has obviously improved. Local officials are now elected throughout the country, providing opportunities for traditionally marginalized groups to take part in municipal government. We saw evidence of progress in the rise of power of a campesino mayor in Padcaya and a female mayor (before her removal) in La Guardia. Much remains to be accomplished in this regard, however. The Bolivian Constitution retains its barrier to the establishment of local parties and the electoral system is fashioned such that parties - not individuals - can 102 lbid, 46. i03 Ibid., 47. It should be noted, however, that the development of the collective civic consciousness that pushes people to want to have their say in municipal decision-making affecting their communities appears virtually nonexistent. Unless they take place within the Popular Participation framework (that is, unless they are events organized by the municipal government to discuss the distribution of resources organized by the municipal government), town meetings on local issues do not take place. The ideas of local issue referenda or attendance at council meetings, for example, draw no interest at all. It is striking that in the several council meetings attended while carrying out this investigation, not one citizen was seen attending the sessions, with the exception of the capital city of Oruro, which attracted three or four individuals (including the press). Local officials confirmed, moreover, that no one attends their public sessions. Popular Participation, so far, has done little to effect a local civic sense. 80 run for office. Both provisions help preserve the monopoly of traditional national parties over the political system at the expense of the emergence of new, dynamic local leaders. 4.4.18. Popular Participation, municipalization, and decentralization are failing in some respects and producing gains in others. There can be no doubt that Bolivian municipal governments are alive. They are developing plans, investing resources in major public works, improving public services or providing them for the first time in many areas, mobilizing resources from funds, departmental capitals, and La Paz, and often feeling pressure from or being forced to answer to community demands and protests. New electoral forces are emerging, and OTBs, vigilance committees, and council members are pressing for their political space. Still, in a number of municipalities, political parties exert disproportionate influence on local affairs, supplanting the civil society's oversight mechanisms. Clearly, for the reform that relies on citizen voice to be fully effective, institutional development of civil society organizations is a necessary precondition. 4.5. Summary 4.5.1. The review of the two strategies for institutional reform attempted in Bolivia shows mixed results but clearly with strong signs of positive potentials. Both the insulation strategy and the voice strategy have contributed to improving public sector performance in selected parts of the Bolivian state. The experiences clearly demonstrate that negative influence of excessive politicization can be contained by providing formal autonomy to some public agencies, or countered by empowering citizens' capacity for public sector oversight. The experiences also indicate, however, that both approaches are by no means trouble-free. The insulation strategy requires a strong a priori political consensus to respect the agencies' autonomy. The evidence is that politicians may be willing to give autonomy to some segments of the state (e.g., regulatory agencies) but not necessary others (e.g., the judiciary). When political conflict frustrates minimal respect for autonomy, internal organizational development that is necessary to build effective internal rules and restraints against opportunism and arbitrary behaviors is not possible. 4.5.2. Giving "voice" to citizens as overseers of government functions has proven an effective method of improving performance of local governments, but only in certain municipalities that possess particular social and institutional pre-conditions. Offering legal status as overseers alone is insufficient to empower vigilance committees and OTBs, if political parties are organizationally stronger than civil society organizations and are able to co-opt the latter. When the mechanism for electoral accountability is inefficient (because of the de facto indirect elections of mayors and the abuse of constructive vote of censure), "voice" without teeth often produces disappointing results. Citizens may raise their voice and complain but politicians are able to ignore them without any political cost (e.g., the fate of the petitions to freeze co-participation funds). 4.5.3. Despite the shortcomings, both approaches, as genuinely home-grown efforts at institutional reform with significant positive results, promise potential for wider applications within the Bolivian context. The challenge is to figure out when it is appropriate and feasible to resort to these reform options, and when it is necessary to seek 81 other approaches. With both strategies, it is also important to recognize the potential risks involved. The insulation strategy may work in a short run and help develop effective public sector agencies, but unless proper mechanisms for external accountability are established, these agencies are likely either to evolve into a "state within the state" or to succumb to renewed pressure to bring them under closer political control. The voice strategy might be captured by minority interests at the local level, and exacerbate inequality and inefficiency that the PPL reform was intended to address. 82 5. REALISM WITH TEETH: RECOMMENDATIONS ON THE INSTITUTIONAL REFORM IMPLEMENTATION STRATEGY 5.1. The Urgency of Reform 5.1.1. In the past decade and a half, Bolivia has transformed itself from a South American "basket case" characterized by political instability and economic crises to one of the more stable democracies and one of the most committed reformers in the region. Nonetheless, the achievement of macro political stability has not been without its cost. Since the Bolivian political system relies heavily on the use of shared patronage for maintaining governability, there are reasons to believe that public sector performance has suffered in exchange for the democratic consolidation and political stability as well as for facilitating implementation of the first generation reforms. More importantly, if experiences of neighboring countries that once relied on systematic use of shared patronage such as Colombia (the National Front) and Venezuela (the "Punto Fijo" system) are any guide, there can be significant long-term costs to the continued use of this "model" of governance. In both Colombia and Venezuela, shared patronage not only weakened performance of the government bureaucracy over the years, but the capture of the state by small elite segments and their followers also alienated the general population. Both factors have contributed to substantial erosion of state legitimacy, and in Venezuela, have culminated in the recent demise of the old political order (Box 5. 1). Box 5.1 Shared Patronage Regime and Eventual Political Decay in Venezuela For much of the second half of the 20th century, most of Latin America was characterized by political instability caused by military coups, revolutions and civil strife. One of the few exceptions, together with Costa Rica and to a lesser extend Colombia, was Venezuela, which established a stable democratic regime with constitutional alternations of powers between the two main parties, Accion Democrd!ica (AD) and COPEI. The consolidation of the democratic stability in Venezuela rested initially on a formal pact among the main parties signed in Punto Fijo (el Pacao de Punto FU), whereby these parties agreed on democratic rules of the game codified in a new constitution of 1961. As the main parties, especially AD and COPEl, dominated the political scene, they have also become virtual captors of much of the political space including the state itself and important segments of civil society (e.g., through its influence over the labor union and other organized segments of society). By the late 1 980s/early 1990s, popular discontent with this elitist political system had reached an explosive level, as manifested in the violent riot in 1989, and the two failed coup attempts in 1992. By the mid- 1990s, the legitimacy of the political system and the traditional political parties had eroded so much that anti-system candidates gained the presidency in the two successive elections in 1994 (Rafael Caldera who, despite being a founder of COPEI, left COPEI and ran on an anti- establishment platform) and in 1999 (Hugo Chavez, the leader of the first coup attempt in 1992). In particular, Chavezs victory was accompanied by the virtual obliteration of the traditional political parties, and thus the Punto Fio regime that was once seen as a model of democracy in the turmoil region. Today Venezuela finds itself in the midst of charged confrontation between the President 83 Chiavez (and his supporters) and the opponents to Chavismo. Although a new constitution has been put in effect, the ;institutionality"4 of the Venezuelan political system is still fragile, with continued serious impacts on the country's governability. 5.1.2. Bolivia may run similar risks unless its political system evolves into one that relies less on exclusionary sharing of patronage than on universal norms of transparency, accountability, the rule of law and responsiveness to citizens' needs. More importantly, Bolivia will need to develop a state apparatus that is capable of providing public services efficiently and effectively to the population as citizens' satisfaction with public services are positively correlated with their support for the political system (Figure 5.1). The current public interest in public sector corruption is a telling signal of citizens' longing for cleaner and more efficient government that serves their needs. The latest episode of social explosion over the water tariff issue in Cochabamba, and the rapid spread of protests in other seemingly unrelated areas, is another example of public discontent over inadequate public service affecting governability of the entire nation. Figure 5.1 Satisfaction with Municipal Services, System Support and Tolerance Satisfaction with Municipal Services, System Support and Tolerance 5 0 4 8 -- - - t li at 468 436 -00 ,, TSeac 4 4~~~~~~~~~~~4 42 -~~~~~~~~~4 4 0 - - Sup portlAlie nta tion 3 8 361 Tolerance Very poor Poor Fair Good Excellent Evaluation of Service Both sig. < .001 Source: Mitchell A. Seligson, La Cultura Politica de la Democracia Boliviana Encuestas & Estudios. 1999, p. 89.'°4 104 The result is from a survey of 3476 Bolivians in a nationally representative sample conducted in summer 1998. "System support" measures respondents' support for the political system irrespective of the incumbent government in power, and was constructed from the following questions: "To what extend do you think courts guarantee fair justice?"; "To what extent do you respect political institutions in Bolivia?"; "To what extent do you think basic rights of citizens are protected by the Bolivian political system?"; "To what extent do you feel proud to live under the Bolivian political system?"; and "To what extend do you think you should support the Bolivian political system?" "Satisfaction with municipal services" is measured with the responses to the following questions: "Would you say that the services that the 84 5.1.3. Recognizing the need for improving public sector performance, Bolivia has embarked on a very ambitious path toward strengthening its public sector institutions as key vehicles for fighting poverty. Some promising results have already obtained from the number of reform efforts such as Capitalization and Popular Participation. However, for the fruits of these reforms to translate into concrete benefits for ordinary Bolivians, additional concerted efforts must be made to improve the coverage and the quality of public services. In order to better serve citizens, the public sector will need to transform itself from the current patronage-laden entity to a professional bureaucracy capable of delivering public services with efficiency and accountability. 5.1.4. The Government's current reform agenda, within the framework of the National Integrity Plan, identifies the right set of challenges and proposes sensible medium-term approaches to institutional reforms in key areas of concern (i.e., state modernization, judicial reform, and anti-corruption). However, the proposed reforms are inherently difficult as they are likely to upset vested interests that have derived private gains from the politicized public administration. The reform measures are also likely to receive at best lukewarm support from most public officials, who will find themselves in a high degree of uncertainty as rules of the game are changed. Thus the stakes are high, and so is expected resistance. The challenges of managing reform implementation will be considerable. Strong leadership and a sustained political consensus on the need for and the direction of the reforms is necessary for eventual success. More than anything else, the reform requires a realistic implementation strategy. 5.2. Tackling Informality by Taking Politics into Account 5.2.1. "Politicization is the problem. " To generate and maintain a necessary political consensus on both the desirability of the reform and the direction it should take, a necessary starting point is a public recognition, by key government officials, civil society, and the politicalparties themselves, that much of the institutional weakness of the Bolivian public sector (e.g., corruption, limited service delivery capacity) is due to the strong political interference in public sector management and that a serious reform effort would entail de-politicization of the public bureaucracy. Of course, the need for de- politicizing the public sector is not unique to Bolivia. Political use of government bureaucracy is a common phenomenon in most developing countries and a number of OECD countries. In any country, public sector reforms are inherently political because they amount to changes in the distribution of political powers, which are exercised through the government bureaucracies. But both donors and governments in developing countries have often ignored this reality and designed reforms that rely heavily on technical fixes (e.g., information technology, "new" management techniques). In Bolivia, as elsewhere, experiences of implementing such technically-oriented public sector reforms in the absence of political commitment, are not positive. Efforts to reform public sector institutions cannot hide behind politically innocuous technical solutions municipality is giving its residents are excellent, good, regular, bad or very bad?" "How have they treated you or your neighbors when you have gone to get some paper work done at the municipality? Did they treat you very well, well, regular, badly, or very badly?" 85 without addressing the underlying political and institutional factors that affect public sector behaviors. 5.2.2. Managed de-politicization. Among the components of the Institutional Reform Project, the Civil Service Reform is the most critical one that attempts to de-politicize the public administration by introducing a system of merit-based personnel management. A successful civil service reform is absolutely essential for transforming the Bolivian state from a source of patronage to a professional bureaucracy and for introducing the ambitious set of measures to promote performance and results orientation in the public sector. Without de-politicization, none of the other measures is likely to bear fruit. 5.2.3. Significant reduction of patronage politics would require institutional development not only in the public sector but also in the private sector and civil society. The private sector must become capable of offering attractive employment opportunities to the large number of citizens who currently seek public sector jobs for lack of alternative opportunities. Civil society has to demand more professionalism from the state instead of seeking private goods through bribery and exploitation of other privileged access to the state that some enjoy. The need for maturing of the private sector and civil society is a generic problem afflicting many developing countries, which will take many years of persistent structural reforms to overcome. In this regard, the Government should continue deepening measures that promote private sector development and improve labor market flexibility. The findings from the proposed World Bank study on "Microeconomic Constraints and Opportunities for Higher Growth" are expected to offer some policy recommendations that may be useful in this regard. 5.2.4. Besides this structural dimension, the question that is uniquely important for Bolivia is how to address the pressures for patronage that arise from the particular problem of coalition building and maintenance. This is where a dose of realism is especially important as de-politicization of the bureaucracy cannot and will not be accomplished over night. Because the use of patronage is closely related to the government's need to build and maintain a viable governing coalition, de-politicization of the public sector must be done in a "managed" fashion, or the political resistance would become insurmountable. This is by no means an easy task. But there are reasons for some optimism as the Bolivian political system has demonstrated its capacity to arrive at open consensus among political parties that is necessary for "managed" de-politicization (e.g., parties' respect for the National Electoral Court). 5.2.5. One way in which the process of de-politicizing the bureaucracy may be made more manageable is to recognize explicitly somne limited, legitimate space for political appointees, instead of aspiring to establish a purely merit-based state in a short time period. Politicians/political parties have a legitimnate interest in being able to exert some influence over the bureaucracy (as discussed in Chapter 3 of this Review). Currently, the unreliability of formal administrative mechanisms for controlling bureaucratic behaviors raises the importance of direct, informal control mechanisms including patronage/political appointments. It is unrealistic to expect that politicians will easily relinquish the influence they are currently able to exert through their appointees in key 86 bureaucratic posts. In this context, the only way in which de-politicization may be accomplished is to provide politicians with some abilities to retain legitimate political control over the state bureaucracy through limited and transparent practices of hybrid appointments, whereby governing parties can fill certain senior posts in the government on the basis of political/partisan criteria as well as merit. Box 5.2 Hybrid Appointments in OECD Civil Services Most civil servants in QECD countries are hired based on the principle of individual merit for a given position. However, a closer look at some OECD bureaucracies shows the existence of a significant number of hybrid appointments in which merit criteria are accompanied by subjective political judgments. For hybrid appointments, merit is a necessary but insufficient condition for appointment. There are two problems to be solved in hybrid appointments: I) attracting the best candidates for the position, since the best candidates may be deterred in light of the threat of apparently arbitrary political dismissals. and 2) striking the right balance between two conflicting types of recruitment criteria. In OECI) countries, the most common solutionl for the first challenge is the creation of a "pool" system. Applicants must satisfy merit criteria to enter the pool, which is then available for subsequent political selection. "Pool management" requires that the pool be sufficiently attractive to recruit candidates who meet the merit criteria but will not always satisfy the political criteria. The pool system ensures that civil servants will not face redundancy. In France, for example, all discretionary appointees are guaranteed ajob in the career civil service, so that the prospect of dismissal from the hybrid post does not imply unemployment. In Germaniv, appointments to senior positions automatically lapse when there is a change in government. The incoming government can reappoint the civil servants. Civil servants who are not reappointed are retained as "ruhestand" (resting). They receive an allowance and may be reappointed at any time. In Canada, Deputy Ministers serve "at the pleasure" of the Crown, and the Prime Minister may terminate the appointment at any time. In the event of dismissal, convention dictates that a comparable position be found or that a voluntary retirement package be arranged. The classic solution to the second issue, striking a balance between merit and political recruitment criteria, is to make an initial selection based on merit. In Canada, the Cabinet Secretary's Office has a highly scientific approach for identifying the competency profiles necessary for specific posts and for assessing candidates. This Office proposes a list of candidates for Deputy Minister posts to the Prime Minister. From this list, the Prime Minister selects a candidate, acting on behalf of Cabinet and using purely political criteria. In the UK, a senior selection committee makes recommendations to the Prime Minister on the appointment of permanent secretaries. These recommendations take into account the views (but not the veto) of the relevant Minister. Application of the merit criteria in hybrid appointments is enforced either by standing independent commissions (UK, US, Japan), or by committees made up of senior civil servants (Belgium). 5.2.6. One instrument that might facilitate political consensus-building over this hybrid approach (or any other approach to de-politicization) is transparency about ways in which political parties exercise control over the government bureaucracy. To the extent all the parties can be more or less certain that no other party is taking unfair advantage of the new rules of engagement, the likelihood of successful de-politicization will increase. 87 On the contrary, if opaqueness prevails about the process of civil service reform, parties in opposition, for example, may suspect that the reform is a government's ploy to staff the bureaucracy with its sympathizers, whom the opposition party would be unable to fire if and when it regains power. The challenge is that the quality and availability of information in Bolivia is extremely low. Thus, one key to achieving the consensus described above is to improve the quality of information about personnel management, and especially about the political parties' control over the state. For example, the Government might: V establish reasonable parameters for political appointments in government bureaucracy agreeable to major political parties; V publish information about the number of political appointees in each agency (possibly including their party affiliations), and make this information available to Congress and the public; and /' and commit to gradually decreasing the scope and magnitude of political appointments within the public sector. 5.2.7. These tasks can be given to a credible central oversight body such as the Civil Service Superintendency. The idea is to allow the political parties themselves, as well as the general public, to monitor compliance with the reasonable standard on political appointments to be set in the implementing regulation of the Ley de Estatuto de Funcionario Pfiblico. Unless there is some assurance that no single party would be breaking the rule to take unfair advantage by continuing to engage in patronage appointments, there will be no incentive for the other parties to also comply with the new rule of de-politicization. It is equally important that the agency in charge of handling this information be seen as politically neutral and professionally competent in order to avoid the perception that information may be manipulated or mishandled. 5.2.8. Careful use of autonomous agencies. Within the broad understanding of "managed de-politicization," the Government might consider selective use of autonomous agencies for functions which lend themselves to autonomization. This is an attractive strategy, which has been more widely used in some other countries including Peru, but with its own costs and risks. Bolivia has a relatively coherent set of autonomous regulatory agencies, and benefits from the political system's demonstrated capacity to create and respect agency autonomy, at least in some areas. While autonomization is no magic bullet, Bolivia's own experiences with the superintendencies, the Central Bank, Controller General of the Republic, and the National Electoral Court, offer reasons for optimism. Based on other country experiences, candidates for further autonomization could include tax administration and other regulatory functions that are capable of generating their own revenues (to guarantee financial autonomy). Some service delivery functions may be potential candidates for "autonomization," but a consideration must be given to the benefits of autonomy versus potential costs and risks of weakened accountability and problems of coordination with parent ministries. 5.2.9. Cautions are needed, however, in expanding the pool of autonomous agencies. "Autonomization" is not a panacea. It does not work in all circumstances and, in some 88 cases, even if it "worked" (in the sense of establishing effective autonomy and corresponding organizational development), it could cause additional institutional dysfunctions that outweigh the benefits of autonomy. First, it is important that these agencies not become "islands" with no accountability to any other part of the state. Autonomy without accountability could be equally dysfunctional as excessive politicization. The challenge is to establish an effective accountability mechanism while avoiding the undermining of effective autonomy by, for example, de facto input control (typically budget) by a central agency. 5.2.10. Second, to the extent autonomous agencies utilize more favorable employment conditions than the rest of the public sector (e.g., better salaries), there is a real risk that their proliferation will generate resentment among public officials and perhaps even encourage exodus of qualified personnel from the "non-autonomous" parts of the state. If taken to the extreme, the proliferation may also result in a serious lack of institutional cohesion within the state, negatively affecting policy consistency and the state's institutional capacity for internal coordination. In the long-run, the "islands" cannot exist without corresponding modernization of the rest of the state. 5.2.1 1. Third, long-term fates of autonomous agencies in the region are generally poor. As political leadership changes, once-effective autonomous agencies often see their autonomy erode as they lose the initial political support. One possible countervailing factor is public support for "islands of excellence." Thus autonomous agencies' fate often depends on (a) their own effectiveness, and (b) public recognition of their effectiveness and support for their sustenance. An aggressive communications strategy may be needed to inform the public of the achievements and potentials of existing autonomous agencies (e.g., SIRESE, customs) so that public opinion can be mobilized in their support, in case their autonomy is threatened. 5.2.12. In the context of implementing the Institutional Reform Project, this Review's recommendation would be to emphasize first providing truly effective autonomy to the Superintendency of Civil Service and the National Customs Agency. Autonomy of the Superintendency would be a key for ensuring effective implementation of the Civil Service Reform, while the sustained autonomy of Customs would send a strong signal of the Government's (and the political system's) seriousness about de-politicizing key government functions. Application of this strategy to other sectors/agencies should be reviewed on a case-by-case basis, taking into account the extent of political commitment/consensus, characteristics of the agency's functions, and the adequacy of accountability arrangements that can be designed into the new agency structure. In general, however, it may be prudent to limit autonomization to those entities in charge of some aspects of regulatory enforcement, which requires a high level of neutrality (and perception that they are politically neutral). Service delivery functions (e.g., roads maintenance), over which parent ministries would have legitimate interest in effective control and accountability, have the potential to suffer from conflicts of interests and objectives between the parent ministries and the semi-autonomous executive agency. Given the current ineffectiveness of accountability arrangements within the Bolivian state, and the likelihood that parent ministries would try to exert some control over the 89 agency behavior on the basis of legitimate policy concerns, an attempt to "autonomize" service delivery agencies may not be very promising. There is already some signs of this difficulty in the agency-ministry relationship in the customs administration. 5.2.13. From such a point of view the autonomization of the SNC, which has just received a congressional approval, may pose a considerable challenge for several reasons. First, road construction is a highly lucrative activity that poses a serious temptation for political interference. Although the current Minister of Economic Development appears committed to guaranteeing the SNC's technical independence, there is limited continuity in the ministerial post, so future officeholders may threaten the reform. Second, the agency's technical capacity is weak and may lead to an even greater deterioration of Bolivia's road infrastructure. Since undertaking its new supervisory role in 1997, the SNC has encountered conflicts with construction and maintenance companies holding government contracts, for various reasons including the SNC's inability to pay companies in a timely manner. This may indicate a liquidity problem in the SNC which in turn could signal the lack of financial autonomy. More than anything else, the most important reason to be concerned about the plan to autonomize the SNC comes from its own history of moving from an effective semi-autonomous executing agency in the 1960s to a highly politicized and corruption-prone entity today (Box 1.1). Box 5.3 Semi-autonomous Tax Agencies in Latin America In recent years. a number of Latin American countries (e.g., Peru, Venezuela, Mexico, Guatemala) have established semi-autonomous tax agencies in order to circumvent bureaucratic rigidities and political interference in tax administration. Although agency autonomy is expected to improve performance of tax administration, the semi-autonomous tax agencies in Latin America have proven less effective and sustainable than their founders intended. Some of the country experiences indicate that the main threat to autonomy often stems from rivalry with the ministry of finance whose political and bureaucratic incentives may not be compatible with the existence of a semi-autonomous tax agency. From a political perspective the main benefit of establishing an effective tax administration reform is greater revenues. Yet it is the president, the congress and other "spending' ministers, not the finance minister as a "guardian of the public purse," who stand to benefit from increased spending that the greater revenue will allow. These benefits are obtained at the cost of lost patronage opportunities for the minister of finance (because he/she can no longer appoint a large number of tax administration staff at will), and less control over policy and operational decisions of tax administration (because of the greater autonomy). Adapted from Robert Taliercio's "The Political Incentives for and against Administrative Reform: The Design, Establishment, and Sustainability of Semi-Autonomous Revenue Authorities in Latin America", draft, March 2000. 5.2.14. Promoting more citizen "voice. " In areas where it is not easy or desirable to grant formal autonomy to the agencies in charge (e.g., most ministries), the Government might also consider a more aggressive approach to encouraging public oversight of 90 government performance. Philosophically, this might entail replicating successful experiences of the Popular Participation reform at the central level, perhaps specifically targeting selected service delivery functions where government-citizen interface provides fertile ground for accountability by citizen "voice." Ultimately, citizen "voice" should translate into improved government performance at all levels through the mechanisms of electoral accountability whereby voters urge politicians to care about public sector performance and thus to support the institutional reform agenda. 105 Given the differences in institutional contexts between municipal and national levels of government, however, mechanical application of social control mechanisms in the Popular Participation framework would be unpractical, or even undesirable. Options for such a direct measure of social accountability are more limited at the national level, where direct interface between the government and citizens is less frequent. It would be critical, nonetheless, to involve non-state actors who are capable of mobilizing themselves at the national level in support of the reform, such as the private sector organizations. In this sense, the Government might consider short-term impacts on private sector activities in selecting pilot agencies for the first phase of the Institutional Reform Project. One interesting possibility, combined with the aforementioned recommendations on the "autonomization" strategy, is to replicate the very successful example of Peru's INDECOPI, an autonomous, market-regulation entity that specializes in anti-trust, and protection of consumer and intellectual property rights (Box 5.4). Box 5.4 Peru's INDECOPI Peru's Institute for the Defense of Competition and Intellectual Property (INDECOPI) was created as an autonomous regulatory agency in 1993 to serve as a catalyst for the sorely needed "culture of free market," after the populist policies of the Garcia years. Since its creation, INDECOPI has developed various information and partnership strategies to involve citizens and establish its credibility and reputation as an open, transparent government agency that promotes public interests (consumer protection) and serves as an example for the rest of the public sector. As a result, despite fears of potentially excessive intervention in private sector activities, even INDECOPI's fiercest critics acknowledge its achievements during Peru's transition to a market economy. Today, INNDECOPI is widely known as a complaint center for consumers who demand better products and services from the private sector. One of INDECOPI's products is Market Information Services. It provides businesses and consumers access to information such as patents, market-related legal transactions, technical norms, and product comparisons. In addition, it conducts public campaigns by disseminating information such as delays in bank lines and on-time arrival records of airlines; trains school teachers in market education; and actively disseminates market competition concepts in the media. As a result of these efforts, significant improvements have been made in private sector performance. For example, bank and airline delays are noticeably reduced. A novel feature of INDECOPI is the use of Decentralized Offices (ODIs) in the provinces to delegate some of its functions. such as trademark registry and dispute resolution, to reputable 105 If the first two options, "managed de-politicization" and "expanded use of autonomous agencies," take politics into account by reducing its scope, this third option airns to exploit positive aspects of politics as a way of restraining perverse political influence on public administration. 91 civil organizations. This partnership with the civil society, such as universities and chambers of commerce, is part of INDECOPI's "stakeholder-building' strategy and the propagation of the Institute's free-market gospel in Peru. Future plans include incorporating its civil society partners into INDECOPI's Board or its high level decision-making structure. 5.2.15. For the expanded "voice" strategy to be effective, it is essential to improve transparency of government action. Relevant information on government performance must be made available in a fornat that is easily understandable to the general public. This could be done by, for example: regularly conducting client surveys for specific government services and publishing their results in the local media; $ publishing the government budget and its execution (detailed below), and; / clarifying and publicizing administrative procedures that citizens are supposed to follow in receiving public services, including specifications of procedures that citizens can follow to report grievances to the Public Defender or other competent public authorities. 5.2.16. Certain state agencies such as the CGR are in a privileged position to contribute positively to the sort of transparency-enhancing strategy proposed here. Unfortunately of the CGR takes a very low profile and does relatively little to strengthening the state's information disclosure practices. However, if certain results of the CGR audits are regularly made public, that would provide society with a powerful instrument to monitor and scrutinize probity and efficiency in government financial management. 5.2.17. It would also be in the interest of the Government, which has a very sensible institutional reform agenda, to try to influence the nature of the public debate on public sector performance and reform. While the current media attention to corruption is a welcome movement that exerts healthy pressure on politicians and the Government to tackle the problem, the media also tend to pursue "scandals" and pay more attention to "whose head rolls."106 An effective anti-corruption strategy goes beyond isolated incidents of punishing specific individuals involved in corruption. It should be a comprehensive strategy that tackles institutional strengthening for preventing and detecting corruption and enforcing sanctions. The National Integrity Plan offers such a comprehensive strategy, but it does not appear to be fully disseminated in society. The media should be encouraged to pay more attention to the systemic nature of the problem and the merit and progress in the Government's medium-term institutional and structural reform agenda. Client surveys and other methods that direct public attention to performance of government agencies and institutional roots of poor performance might be one way of "upgrading" the nature of public debates on corruption. The Government should also develop a more aggressive communications strategy to inform the media, and through them, the broader society, of the reform's objectives and approaches. 106 This point was also made in "Increasing civil society participation in the PNI: Trip report to La Paz," April 15-20, 1999, Walter Guevara. 92 5.3. Tackling Informality by Strengthening Institutions of Restraint 5.3.1. Without tackling informality, or without developing effective institutional mechanisms to enforce a minimum set of formal rules, an ambitious agenda of performance-oriented public management is unlikely to produce desired results. De- politicization of the public sector is a necessary first step to this end, without which other measures of institutional strengthening are likely to prove very effective. Beyond de- politicization, tackling informality means strengthening institutions of restraint. In the Bolivian context, this would entail, among other measures, more aggressive efforts to implement the normative centralization envisioned in the SAFCO Law and its subsidiary regulations to establish a robust regime of internal public sector regulations. Also to the extent patronage practices are encouraged by the current weaknesses of institutional mechanisms for controlling bureaucratic behavior (see Chapter 3), strengthening internal oversight mechanisms might have some positive impact on reducing incentives for patronage appointments. 5.3.2. Sequencing SAFCO implementation. SAFCO is an ambitious framework that aims at transforming Bolivia's public sector into a modern, performance-oriented entity. The key guiding principle for this objective is the notion of operational decentralization. Through normative centralization, SAFCO was also intended to introduce a degree of institutional order into the public sector by establishing sets of internal norms, rules and regulations. But the ambitiousness of the Law has provided the Bolivian Government with a significant challenge. Results-based management systems, such as the one envisioned in the SAFCO Law, depend on the exchange of accurate, timely information. The information on outputs that it requires is costly, but failure to obtain it undermines any prospect of enforcing the contractual arrangements which underpin it. Unfortunately, Bolivia has made this move toward results orientation without the information and enforcement mechanisms in place. As a result, the expected transformation of the Bolivian public sector has yet to happen. Without minimally effective central oversight mechanisms, which rely on good information flow, operational decentralization in the context of fragmented party capture is likely to encourage more opportunism than performance orientation and efficiency. 5.3.3. Gathering and transmitting information on performance is costly in any circumstances, and the OECD tradition for over 100 years has been to reduce the cost by defining performance in terms of compliance with due process, with minimal room for discretion. The key elements have been: * career paths consisting of a largely predetermined vertical progression, with an emphasis on seniority; * placing all civil servants on broadly common terms and conditions; * tying jobs or job families to preset academic qualifications; * constraining agencies' spending by narrowly-defined line-item budgets; and * controlling position creation by a central body. 93 5.3.4. In contrast to compliance-based systems, performance-based arrangements create a formal linkage between performance in achieving objectives and incentives. The latter is usually achieved by a combination of formal 'contracts' within the public sector, in which agencies are expected to deliver specified outputs in exchange for defined resources, and performance-related benefit schemes. Typically, formal contracts can be between senior managers and Ministers - when they link the pay rates or continued employment tenure of the agency head with monitored performance - or they can be between funding bodies and the agencies and departments, when they establish a 'purchasing' arrangement of some sort with specified outputs to be provided in exchange for guaranteed budget, or allow the agency some retention of fees in exchange for monitored performance. 5.3.5. In Bolivia, there are three principal and inter-related contracts defined by SAFCO: the POA, the POAI (or other individual annual performance appraisal/programming document), and the budget. As discussed in the Chapter 2 of this Review, the POAs and the POAIs are not respected as management tools by line agencies, by individual employees, nor by the Ministry of Finance. These are perceived as mere formalities that agencies must produce in order to receive their budget allocation. In addition, one of the primary complaints that public managers expressed in focus groups and in individual interviews was the unreliability of the budget process. Thus the formal contracts that make up the backbone of SAFCO are relatively worthless. An inability to develop the performance information allows all parties to renege on their contracts - the funders have every reason to doubt that the required performance has been achieved, and the line agencies and public officials have equal reason to doubt that their resourcing will be as predicted. 5.3.6. Govermnent-wide institutional reform efforts must take into account the very partial implementation of the SAFCO framework, and the underlying reasons for it. It should be kept in mind that the first set of SAFCO objectives - establishment of administrative and control systems to ensure minimal procedural compliance before being able to move to a more ambitious model of results-orientation - is yet to be operationalized. Thus, a reform strategy should aimfirst at completing this initial stage of the original SAFCO reform, with most emphasis given appropriately to instituting effective compliance mechanisms. Effective compliance mechanisms would be built by a combination of developing/strengthening independent central oversight bodies (e.g., CGR, Civil Service Superintendency) accompanied by effective flow of necessary information for monitoring agency behavior, and fostering better internal control capacities in line agencies. Only then would it make sense to begin to fully implement operational decentralization and results-oriented management practices. 5.3.7. This is a proposal for sequencing reforms to achieve the ultimate objective of developing a results-oriented public sector in Bolivia. As this Review has emphasized, institutional pre-conditions for moving rapidly toward results orientation do not exist in today's Bolivia, which calls for a phased approach that first attempts to build those pre- conditions. It is also a proposal for a transitional approach that pursues a sensible middle ground between an advanced performance-oriented system as we find only in a handful 94 of OECD countries today on the one hand, and a traditional, compliance-based system on the other. The proposed approach aims at building agencies' capacity for effective internal control (i.e., institutional capacity and incentive to be their own guardians of agency probity and accountability), while gradually relaxing tight central control over administrative inputs. Box 5.SGetting the Basics Right: New Public Management in Developing Countries Since the late 1980s, a number of OECD countries have implemented public management reforms based on contract-like relationships between the government and ministers as purchasers of goods and services and departments and other entities as suppliers. The main objective of the reform is to improve operational efficiency of public agencies by giving their managers and staff greater managerial discretion in exchange for delivery of pre-specified performance goals through some forms of formal contractual arrangements between the government (usually represented by a minister) and the agency's chief executive, or between the agency head and staff. New Zealand is widely known as advancing the most coherent form of this so-called New Public Management (NPM) reform, and the idea is quickly spreading in developing countries (including some in Latin America). where similar approaches to public sector reform have been either attempted or, at least, contemplated. However, it is important to bear in mind some critical differences in institutional pre-conditions between New Zealand (and other advanced OECD countries) and most of the developing countries, which can affect the advisability of an ambitious NPM reform. Of particular importance is the enforceability of rules (including "contracts") within the public sector. By the time the NPM reform was launched, New Zealand had developed a rule-bound, relatively corruption-free public bureaucracy, in which budgets "effectively controlled spending and corresponded to actual tnsactions" and a civil service regime "govemed how public employees were hired and paid." Althouh NPM reforns are sometimes misconstrued as reducing the amount of bureaucratic rules, the culture of voluntary rule compliance and robust institutional mechanisms for rule enforcement (in case of rule non-compliance) are essential prerequisites for a public bureaucracy to move in the direction of performance/results-oriented management as enbodied in the NPM. In most developing countries, in contrast, bureaucratic informality prevails within the public sector, and contract enforcement is weak even in the private sector. Therefore, it is important for developing countries aspiring to move in the direction of NPM to "get the basics right" before "leapfrogging." One way of "getting the basics right" could entail: I . Parallel advances in the market sector together with those in the public sector so that a culture of rule compliance and contract enforcement can be developed. 2. Establishing reliable external controls as building blocks for a formal, rule-based, honest public sector. 3. Strengthen the capacity for controlling inputs - i.e., the public managers must be able to account for cash before they are asked to account for cost. they must abide by uniform rules before it is authorized to make their own rules; they must operate in integrated, centralized departments before being authorized to go it alone in autonomous agencies. 95 iJnless adequate capacities and incentives for effective internal control are developed first, mere delegation of managerial decisions can result in increased corruption and weakened accountability without necessarily improving public sector performance. From -Why Most Developing Countries Should Not Try New Zealand's Reforms" by Allen Schick, The World Bank Research Observer, vol. 13 no. I (February 1998). pp. 123-131. 5.3.8. In the context of implementing the Institutional Reform Project, this approach would translate into demanding that agencies vying for IRP support first demonstrate a capacity to comply with the minimum set of internal rules and regulations. As the agencies acquire demonstrable capacities to manage and enforce their own internal regulations, they can be given greater managerial autonomy and budgetary flexibility, while still being subject to ex-post audit and performance evaluations. At this initial stage, "performance" would be defined more in terms of actual compliance with these rules as opposed to output/outcome indicators used in more advanced performance management/evaluation systems. The newly-created Civil Service Superintendency should vet the adequacy of the agencies' capacity to manage their own personnel, and monitor their compliance with the rules and regulations. Agency heads would still hold ultimate responsibility for setting up SAFCO systems and other norms in their own agencies, but the closer supervision by the central oversight body would increase the agency heads' incentives for instituting the norms and for building institutional capacity for self-compliance. For its part, the oversight body would be responsible for ensuring that the quality of the systems in place and the capacity of the agency to regulate itself is adequate. 5.3.9. A similarly strong oversight mechanism is necessary to advance with a more robust system of regulating government financial management. Currently the task is divided between the Ministry of Finance, as the rector entity for the seven SAFCO management systems, and the Controller General, as the rector entity for the SAFCO control system. Given this current arrangement has contributed relatively little to accelerating the implementation of the SAFCO financial management norms across the public sector, consideration of an alternative arrangement may be in order. One option would be to expand, at an appropriate moment, the responsibilities of the Civil Service Superintendency to include oversight of public financial management. Alternatively, an autonomous regulatory entity that oversees the government-wide financial management system and practice on behalf of the MOF may be created. Finally, the government might continue with the current arrangement. It is not clear, however, whether the Ministry of Finance, whose priorities have always been on management of macroeconomic policy (including aggregate fiscal discipline) and donor relations, has either the commitment or the administrative capacity to fulfill this expected role. This decision could be made after assessing the initial performance of the Civil Service Superintendency in order to judge whether a parallel entity should be created for financial management regulation, whether the Superintendency itself should take on this new task, or whether the current arrangement should be preserved with appropriate adjustments. 96 5.3.10. Whichever entity is charged with overseeing the adequacy of line agencies' internal regulatory capacities, it is critically important that these oversight entities be provided with the necessary level of operational autonomy and political backing. It is highly probable that some agencies would violate some aspects of the government-wide internal regulations. An episode of this nature happened in the Ministry of Agriculture a few years ago, when the incoming minister tried to fire all the personnel hired under the Civil Service Program despite the employment guarantee these staff presumably enjoyed. At that time, the inappropriate decision by the minister was reversed only with the strong stance taken by the Civil Service Execution Unit. The new oversight bodies must possess similar ability to stand by their mandate and resist political pressure that will likely arise from some corners. 5.3.11. Credible and binding budget. Simultaneously, the Government should enhance its effort to improve public expenditure management, as an absolutely necessary prerequisite for subsequent introduction of results-oriented management in the public sector. As argued in Chapter 2, without predictability of budgetary resource flows, line agencies will not be able to take advantage of increased operational autonomy, which the IRP envisions, and be held accountable for their performance. A pressing objective is to ensure greater stability in budget allocations within a given fiscal year (i.e., significantly reduce the extent of in-year reallocations). If implemented according to the design, the ongoing efforts through the ILACO Project would provide a necessary information base for effective public financial management. But without tackling broader institutional issues in budgeting including the perverse incentives the current procedures create (e.g., unrealistic congressional budgets, mid-year reformulation), overall outcomes of budgeting are unlikely to improve much. 5.3.12. From the point of view of supporting eventual introduction of results-oriented public sector management, key is to make the budget a credible, binding document that truly reflects the Government's commitment to strategic resource allocations. Government budgets in Bolivia currently do not fully serve this purpose. In addition to continued implementation of the revamped ILACO project, the Government should consider adoption of a more strategic approach to public expenditure management aiming at making the budget more realistic and binding with regard to both inter- and intra-sectoral allocations. This would require reform of certain aspects of the budget procedures. Budgets should reflect sectors' policy and program priorities as accurately as possible, while abiding by the aggregate fiscal constraints. For this, ways must be found to facilitate systematic incorporation of POAs' into budget allocation decisions, including revision of their format and broader publication/dissemination of their content. This in turn would mean reformatting POAs from the current thick vohlume that cannot be digested during the budget preparation process to a more focused document that links the agency's program objectives to its budgetary needs. V In order to make the annual budget more binding, the Government might consider publicizing its inter-sectoral and sectoral policy priorities in advance of m 97 annual budgets and demonstrating the consistency between the actual budget and the announced policy priorities. It is important that these priorities are arrived at after consultations with line agencies and key interest groups in sectors. $ The revenue base of the budget should be realistic in order to avoid artificial inflation of approved budgets. To achieve this, Congress's ability to raise revenue assumptions without technical foundations must be restrained, and the technical capacity for revenue projection must be improved. v Elimination of the practice of mid-year reformulation should also be considered as it introduces a perverse incentive that legitimizes and encourages unjustified as well as justifiable re-allocations during the fiscal year. To the extent aggregate expenditures are stable (i.e., indicating limited external shock), the practice of mid-year reformulation loses its original rationale, which was presumably to allow the government to respond to external shocks by adjusting expenditures. / Besides efficiency and effectiveness of public expenditure programs, the Government should also strive to evaluate its own budgetary performance in terms of the degree to which actual budget execution approximates the allocations in the approved budget as an indicator of the credibility of the budget. This information should be made public as a way of strengthening citizen oversight of public expenditure management. V Cash management based on monthly quotas is an effective method for controlling aggregate expenditures, but has the significant downside in terms of limiting the predictability of budgetary resource flows to line agencies. The Government might consider relaxing the monthly quota system to a quarterly quota system with a simultaneous introduction of an obligations-based accounting practice at least to those agencies that meet the IRP criteria for operational decentralization. In this new approach, the MOF would concentrate on monitoring agencies' compliance with their quarterly quotas at the level of expenditure commitment as opposed to actual spending on detailed line items. This should ease the cost of monitoring spending, and is expected to improve the effectiveness of expenditure control at a more aggregated level. For their part, the line agencies would enjoy somewhat greater managerial flexibility and improved predictability in resource flows. 5.3.13. In a presidential system such as Bolivia, the Congress has an important role to play during budget approval and evaluation. 5.3.14. Finally, strong cornmitment by the Ministry of Finance to advance institutional reform of the public expenditure management systems and to work toward eventual introduction of budgeting for results is essential for the success of the Institutional Reform Project. Both the history of SAFCO implementation and the recent behaviors of the MOF indicate weak commitment to the institutional reform agenda. While the MOF has demonstrated commendable capability to ensure fiscal discipline and macroeconomic stability, its stronger support for public sector modernization has so far been an important missing ingredient for the eventual success of the Institutional Reform Project. 98 5.4. Other Institutional Refonn Agenda 5.4.1. Although this Report has focused its analysis and recommendations on devising a realistic implementation strategy for the Institutional Reform Project (and by extension the National Integrity Plan), there are other institutional reform agenda that are critical for improving Bolivia's governability over the medium term. In particular, improving governance in Bolivia depends crucially on parallel progress on two other major institutional reform initiatives, decentralization and judicial reform. 5.4.2. Decentralization. Effective decentralization can be an effective tool for improving efficiency of public services and strengthening governance by making local governments more responsive to citizens' needs and demands. In today's Bolivia, however, the state's capacity to provide efficient public services at the local level is still limited. More importantly, its inability to resolve localized conflicts at the local and regional level is evident from the recent episodes of social explosion. These events bring to center stage the importance of strengthening sub-national governments' legitimate autonomy and administrative effectiveness. As sub-national responsiveness to citizens' demands becomes an essential pre-requisite for governance at all levels, clear distribution of resources and responsibilities and strengthening each government's accountability have become urgent challenges for political stability. However, as our discussion of Popular Participation shows, local politics are still heavily influenced by the national political agcnda through political parties, politicized Prefecturas, and other policy and political instruments of the central government (e.g., the investment funds). Innovative institutional mechanisms for social control that were introduced with the Popular Participation Law are not functioning as intended partly because of the distorted incentives and accountability relations across the three levels of government. 5.4.3. As our survey results clearly show, the weakest link in Bolivia's decentralized public administration system is the intermediate level of Prefecturas, which are highly politicized, extremely weak in their administrative capacities, and very ambiguous in their accountability arrangements. While Prefects are appointed by the President, the Departmental Councils represent municipalities (though often based on partisan criteria). Combined with the strong influence of regional political interests, the current institutional arrangement at the intermediate level has produced an entity that is fully accountable neither to the national nor to the local governments. A result is a highly discretionary style of management at the intermediate level that rewards or penalizes municipalities (e.g., with resource transfers) depending on the latter's political affiliations. 5.4.4. it is clear from evidence that the political use of Prefecturas stems from the same party coalition dynamics that produces cuoteo politico within the central administration. At the departmental level, the need to use the bureaucratic apparatus to meet political needs gets further complicated by the importance of regionalism as an important political cleavage in the country. Although it may be legitimate, up to a point, to use the intermediate level as political "glue" for a nation that potentially suffers from centrifugal tendencies of strong regionalism, the cost of this arrangement to the state's effectiveness 99 and legitimacy is already visibly high. Continued politicization and institutional weakness of the intermediate level further frustrate the fragile achievements of Popular Participation by suppressing municipal autonomy, and negatively affect the system's legitimacy and Bolivia's governability. 5.4.5. There is a strong need to clarify accountability arrangements, rationalize the institutional framework of inter-governmental relations, and de-politicize the Prefecturas to enhance municipal autonomy and responsibility. Given the current state of Prefecturas, the most advisable approach to defining its roles is to follow the principle of "matching responsibilities and resources to capabilities." It is a well-known fact that Prefecturas today maintain more resources than they need given their responsibilities, and more responsibilities than they can manage given their capacities. While the impacts of the Institutional Reform Project can only be expected to materialize in the long term, reforming the decentralized public administration holds the promise of a more immediate improvement in public service delivery to citizens, greater impact on poverty reduction, and thus increased legitimacy of the state. In fact, perceptible improvements at the decentralized level may catalyze stronger citizen demand for institutional reforms at the national level. 5.4.6. Judicial reform. Another critical area that Bolivia has been grappling with is judicial reform. The experience so far has been a frustrating one, marred by partisan, organizational, and personal conflicts (Annex 6). An independent and efficient judiciary plays a central role in government-society relations and economic development, by resolving conflicts, protecting citizens' rights, and upholding the rule of law. Bolivia's judiciary has some of the poorest performance indicators in the region, including large backlogs, low clearance rates, little public confidence, a lack of uniformity in decisions, and frequent charges of corruption. The judiciary is the weakest branch of government, partly due to the civil code tradition but also because of consistent political and personal interference from the Executive and Legislature. 5.4.7. Since 1994, Bolivia has tried to modernize its judiciary through an array of initiatives. These range from revising its organizational structure (creation of the Judicial Council, the Ombudsperson's Office, the Ministry of Justice, the Constitutional Tribunal, etc.) and basic legal framework, to introducing judicial training, new appointment mechanisms, and modern management technologies. Some reform efforts focus on improving efficiency (e.g., case processing times), while others target the quality of performance by changing the judicial incentive structure (altering the relationships with other branches of government or within the judiciary itself). So far, results have been limited on both fronts. Specifically, there is growing concern about the existence of formalized external control over judiciary appointments and decisions which simultaneously decrease the organizations' potential for self-governance. It is unlikely that either efficiency or quality of output will improve significantly until the institutional issues are resolved. 5.4.8. Our examination of the judicial sector concluded that the composition and operations of the key judicial institutions (courts, Fiscalia, defense, police) are shaped by 100 a variety of political inputs into the appointment process and operational matters which clearly detract from institutional unity and individual performance. Over the last decade, internal conflict has increased; institutional leadership's ability to control subordinates has declined, and there appear to be increased opportunities for the operation of internal cliques and horizontal penetration (the well-linked law firms who work through ties with individual police,fiscales, judges and administrative staff). These weaknesses have contributed to a growing frustration with the judiciary among the public, and skepticism about the possibilities of change. And yet, criticisms are rarely accompanied by concrete suggestions for improvement. Media coverage of sectoral issues has probably contributed to greater demand for judicial reform, but Bolivia is still far from reaching a consensus about the direction that reform should take. 5.4.9. Under recent reforms, greater judicial independence (which it is believed will lead in the long run to better performance) is allowed but not guaranteed. Still, as the last decade of partial reforms suggests, imposing new rules or adding agencies without addressing the more basic understandings of those who will apply or staff them is an invitation to unexpected and often undesired results. Unilateral change that does not seek to build consensus through broad debate is a frequent source of inconsistencies, contradictions, and new loopholes. Lack of consultation in a sector that so directly affects government-society relations can result in an even greater chasm between politicians, public officials, and the citizens they are supposed to serve. 5.5. Conclusion 5.5. 1. Institutional and Governance Reviews (IGR) are the World Bank's new analytical instrument that purports to explore "good fit" between a given country's institutional context and options for public sector institutional reform. Bolivia's frustrating experiences with the recent attempts at reforming/modernizing the central public administration (e.g., SAFCO, PSC) indicate the importance of better understanding "good fit." Simply importing "good ideas" from elsewhere without appropriate local adaptation will not work. In that spirit, this Bolivia IGR has explored incentives that shape behaviors within Bolivia's public administration. It is our view that the nature and magnitude of politicization cautions us against "leapfrogging" into an idealized state of public administration we observe in a few OECD countries. There are prior steps to be taken before Bolivia can reach the "cutting edge." The strategy for implementing the Institutional Reform in Bolivia must balance the need to be realistic (by taking underlying political factors into account) and the need to have some "teeth" in its content (by strengthening accountability and control mechanisms to curb corruption and other forms of opportunistic behaviors). 5.5.2. Other institutional reform agenda including decentralization and judicial reform are equally important for the Bolivian state to enhance its efficiency and effectiveness, and thus to regain its legitimacy. Although no quick fix is available for any of these reform agenda, and concrete results in terms of improved governance performance may not obtain for some years, it is a matter of urgency that the government takes credible actions in the short run. In tackling any of these reform challenges, it is important to bear 101 in mind that legal imperfections are not primarily responsible for slow progress in administrative and judicial reforms and the fight against corruption. While the legal framework governing public administration and the judiciary has defects, a more serious problem is persistent lack of effective enforcement of existing laws and public sector rules. Tackling the problem of informality which gives rise to the systemic rule non- compliance and weak accountability is analogous to confronting the difficult task of addressing the political roots of public sector behavior. 5.5.3. This final chapter is our attempt to suggest several realistic measures to facilitate reform implementation. Given the inherently difficult nature of the subject, both analytically and operationally, we do not pretend to have offered definitive answers to the number of puzzles that motivated this study in the first place. Our modest aims are to systematize and make explicit, to the extent possible, our understanding of the nature of the institutional environment of the Bolivian public sector, and contribute to informed discussions on how to improve public sector performance, and thus governance over the medium term. 102 TECHNICAL ANNEX: BOLIVIA PUBLIC OFFICIALS SURVEY The Bolivia Public Officials Survey was commissioned by the Bolivia Institutional and Governance Review team for the purpose of obtaining information from public officials on characteristics of the institutional environment of Bolivia's public sector. The survey was carried out from May - August 1999 with a sample of 738 public officials at different hierarchical levels, from 15 public organizations. This Technical Annex describes how the indicators in Table 3.1 of the main IGR report were constructed. * Mission accomplished: Degrees to which agency heads perceive their agency's mission to have been accomplished. * Vertical solitude: Degrees to which there is a discrepancy in perceptions about a set of issues between senior managers and middle to lower-ranking staff of the same agencies. This is a proxy for vertical cohesion within the organization. Internal cohesion across hierarchical levels is essential for the agency's operational efficiency. * Corruption: Degrees to which corruption is perceived to be a serious problem in a given agency. * Credible regulations: Degrees to which there is an adequate set of internal regulations within a given agency. * Political hiring: The ratio of persons perceived to have been hired for political reasons in a given agency. Following is an explanation of how the indicators were constructed. Mission accomplished: Agency heads were asked: 28. "What percent of the objectives of [your agency's] mission do you think was achieved?" 1. 0% to 25% 2. 25% to 50% 3. 50% to 75% 4. 75% to 100% A response of 1. (0-25%) scored 0; a response of 2. (25-50%) scored 3.3; a score of 3. (50-75%) meant 6.6; and 4. (75-100%) scored 10. The column totals were averaged to obtain the agency score for mission accomplished. Vertical solitude was calculated by first scoring agency heads' perception of each dimension of the institutional environment (rule credibility, policy credibility, and resource reliability) and then the general officials' perception of the same. Then the difference (absolute value) of these scores in each dimension was averaged as the vertical solitude. 103 Corruption was probed by asking whether officials saw it as a problem, and whether it is reported. To calculate corruption, Q. 88 and 92 were used. 88. Would you say that, in your organization, corruption is a Very Serious Problem, a Serious Problem, a Somewhat Serious Problem, or Not a Serious Problem? 1. Very Serious 2. Serious 3. Somewhat Serious 4. Not Serious At All 92. In general, cases of corruption are always, frequently, occasionally or never reported to proper authorities? 1. Always 2. Frequently 3. Occasionally 4. Never In Q. 88, a response of Not Serious At All produced a score of 10, Somewhat Serious a score of 6.6, Serious a score of 3.3, and Very Serious a score of 0. In Q. 92, Always gave a score of 10, Frequently 6.6, Occasionally 3.3, and Never meant a score of 0. The column totals of Q. 88 and 92 were averaged to obtain the score for corruption. The existence of credible regulations was tested in four areas: recruitment, evaluation of performance, training, and record keeping. Six questions were used: 31. Are positions like yours generally advertised to the public? 1. yes 2. no 36. In your organization, would you say that the employees are hired more often for: 1. performance on an examination 2. personal connections 3. political connections 4. family relations 5. payment or gift to some public authorities 6. (good) cv 38. In your organization, which of the following methods have you used to learn your job? 1. A written procedural guide 2. Verbal instructions from a superior 3. Verbal instruction from a colleague 4. Self-taught 44. Does your organization have Individual Annual Operations Plans (POAI) for its employees? 1. yes 2. no 46. Are there any other performance evaluation methods in your organization? 104 1. yes 2. no 51. Does your organization have records from the last five years? 1. yes 2. no 3. n/a In Q. 31, 44, 46 and 51, "yes" produced a score of 10, and "no" a score of 0. In Q. 36, a response of either 1. (performance on an examination) or 6. (good cv) both meant a score of 10; if neither response was checked the score was 0. Respondents could receive a score of either 0 or 10 on this question. In Q. 38, a checkmark next to 1., 2. or 3. was scored as 10, while a checkmark next to 4. received a score of 0. The respondent could receive a score of either 0 or 10 on this question. The column totals of the questions were averaged to obtain the score for credible regulations. World Bank User \\prints3\proj-sup\PersTeam\kebede\Bolivia IGR.doc 2/10/00 5:33 PM