Bolivia: Ex-Post Assessment of Longer-Term Program Engagement

This paper presents key findings of the Ex Post Assessment of Longer-Term Program Engagement for Bolivia. Bolivia is a country that is perceived as having one of the best structural reform records in Latin America but experienced sluggish per capita growth, and made virtually no progress in reducing income-based poverty measures. The paper presents a summary account of Bolivia’s performance under IMF-supported programs. It emphasizes that to address Bolivia’s main economic problems—insufficient growth, and fiscal and financial fragility—a new medium-term program should be focused on fundamental institutional and structural reforms.


This paper presents key findings of the Ex Post Assessment of Longer-Term Program Engagement for Bolivia. Bolivia is a country that is perceived as having one of the best structural reform records in Latin America but experienced sluggish per capita growth, and made virtually no progress in reducing income-based poverty measures. The paper presents a summary account of Bolivia’s performance under IMF-supported programs. It emphasizes that to address Bolivia’s main economic problems—insufficient growth, and fiscal and financial fragility—a new medium-term program should be focused on fundamental institutional and structural reforms.

I. Introduction

1. The main questions addressed in this Ex-Post Assessment are as follows: how could a country that undertook one of the boldest and most celebrated economic reform programs of the 1990s and was among the first to benefit from debt relief under the HIPC Initiatives find itself, only a few years later, in a situation of large deficits, sharply increasing debt, declining per capita income, financial instability, and social unrest? What role did the IMF, which supported Bolivia almost continuously, play in this process? And what lessons can be drawn from the experience for the Fund’s engagement with Bolivia in the future?

2. In addition to reviewing IMF documents related to Bolivia since the early 1990s, this study draws on related studies conducted by academics, the World Bank (1998, 2000, 2001, 2004), the Inter-American Development Bank (IDB, 2004), the World Bank and IDB (2004), the IMF-World Bank Financial Sector Assessment team, and IMF staff, who wrote a Country Strategy Paper in 1998 focused on the question of why growth in Bolivia had not been higher. The purpose of the present study is broader, reflecting the broad mandate of an ex-post assessment of longer term program engagement, and also the significant turn for the worse that the Bolivian economy took after 1999. The study also draws on extensive interviews with IMF, World Bank and IDB staff, and the present and former authorities in Bolivia.

3. The sections that follow begin by stating critical facts that are part of the Bolivian “puzzle”: namely, that a country perceived as having one of the best structural reform records in Latin America experienced sluggish per capita growth, and made virtually no progress in reducing income-based poverty measures. This is followed by a summary account of Bolivia’s performance under IMF supported programs. The paper then turns to the main questions that are raised by Bolivia’s economic performance since the late 1980s. It is not difficult to see why growth slowed between 1999 and 2002—a period in which Bolivia was hit by a string of severe shocks—but why has long term growth not been higher? Why does the financial situation continue to be vulnerable, in spite of significant reforms? Why did the fiscal situation unravel after 2000? Should the HIPC Initiative be considered a failure in Bolivia, and if so, in what sense did it fail, and why? Was the 1996 privatization of the hydrocarbons sector—the subject of much controversy in 2003 and 2004—a mistake? The point of this analysis is not to minimize Bolivia’s successes in the 1990s (particularly relative to the preceding decade, which included a hyperinflation and a decline in output) but to understand why Bolivia did not do better. The answers raise obvious questions about the role of the Fund, which are discussed in the following section. Section IV takes stock, and proposes a strategy for future Fund engagement with Bolivia.

II. Facts

A. Reforms, Growth, and Poverty Reduction: Basic Trends

4. Between 1985 and 1999, Bolivia undertook some of the most extensive structural reforms of any Latin American country. According to a structural reform index compiled by Lora (2001), Bolivia’s structural reforms set out from lower base than in most other countries, and occurred in two main phases: from 1985 until about 1991, and from 1995 to 1997. After the first, Bolivia caught up with the Latin American average, while the second placed it among Latin America’s leading reformers (see figure).


Bolivia and Other Latin American Countries: Structural Reform Index, 1985–1999

Citation: IMF Staff Country Reports 2005, 139; 10.5089/9781451970500.002.A001

5. Bolivia’s growth performance over the same period mirrors that of other Latin American countries. As shown in Table 1, Latin America’s growth record since the debt crisis can be split in two periods. From the early 1990s until 1998, growth in the region averaged almost 4 percent, as the region benefited from high commodity exports and capital inflows (with a brief interruption at the time of the Tequila crisis). After the shocks of 1998–1999—a sudden stop in capital flows after the Russian crisis, and a sharp terms of trade deterioration—growth slowed to an average of 2 percent, and per capita growth turned slightly negative. The table shows that Bolivia’s growth performance was fairly typical in both periods. The question is why Bolivia did not do better in spite of its much lower initial income position—investment as a share of GDP remained far below average, contrary to what neoclassical growth theory would predict—and its extensive structural reforms.

Table 1.

Bolivia and Other Latin American Countries: Growth and Investment, 1991–2004

article image
Source: International Monetary Fund, WEO; Bolivian authorities; World Bank, World Development Indicators.

PPP weighted.

Bolivia, Ecuador, Peru, Colombia, Venezuela.

Trends in Poverty, Per Capita Income, and Inequality

Poverty as measured by income-based poverty lines slowly declined in the 1990s, but increased again after growth rates declined (1999–2002). As shown in the table, the high growth phase in the 1990s brought about only a modest decline in income poverty. The fraction of the population in extreme poverty, defined as having insufficient income to consume a food basket of minimum caloric intake, decreased by only 3 points in capital cities (for which comparable survey data is available) from 1993 to 1999. The fraction in poverty (which includes all those unable to meet basic non-food expenditures) declined 6 points from 52 percent in 1993 to 50.7 percent in 1997 and 46 percent in 1999. Since the economic slowdown started in 1999, poverty on a national basis has actually increased by just under 2 percentage points.

Table 1.

Headcount poverty indicators

(in percent of individuals)

article image
Source: INE household surveysNote: income based measures. The poverty indices expresses the number of individuals below the poverty line, while the poverty gap adds the distance of all poor individuals to the poverty line.

However, there has been substantial improvement based on non-monetary indicators of poverty. In the past decade, Bolivia witnessed considerable progress in living conditions, access to basic social services and outcome indicators. Poverty as measured by unsatisfied basic needs fell significantly from 71 percent of the population in 1992 to 59 percent in 2001, although mainly in urban areas, and with large disparities between municipalities.

Table 2.

Poverty according to Unsatisfied Basic Needs 1/

(percent of individuals)

article image
Source: INE based on 1992 and 2002 Census data.

The unsatisfied basic needs index is computed as an average of four sub-indices for housing, sanitation, education and health.

Sluggish growth in per capita income, increasing income inequality, and distance of the poor from the poverty line, are the main factors explaining the small income poverty decline. Per capita income levels have yet to fully recover from the debilitating effect of the economic and political crisis of the early 1980s and hyperinflation, with real GDP per capita still only just above 90 percent of its level in 1980 (see Chart 1). There has been some recovery since the height of the crisis in 1985, however on a PPP-adjusted basis, it only translates to around 6 percent. Also, over the 1990s, income inequality increased sharply (as in other countries in Latin America); with the Gini coefficient increasing from 53.8 in 1990 to 61.4 percent in 2002. The fall in per capita income since 1999 was largest among lower income deciles.

Chart 1.
Chart 1.

Annual Real per Capita GDP

Citation: IMF Staff Country Reports 2005, 139; 10.5089/9781451970500.002.A001

6. At the same time, Bolivia’s record at reducing poverty has been disappointing (Box 1). According to World Bank data, PPP-adjusted per capita income grew by only about 6 percent over the entire period. Inequality increased. The distributional effects of the 1999–2002 recession were particularly pronounced: income levels declined in the lowest rural population quintile, as well as in the urban population except for the top decile, where incomes rose (World Bank, 2004). Poverty defined in income terms stayed roughly the same between 1993 and 2002, as a modest decline from 1993 to 1999 was offset by an increase between 1999 and 2002. On the upside, social services and many social indicators improved in the 1990s, leading to a significant decline in poverty as measured by “unmet basic needs.” Regardless of how poverty is measured, Bolivia continues to have by far the highest poverty rate in Latin America, and one of the most unequal income distributions (together with Brazil and Chile) in the most unequal region of the world.

B. Performance under IMF-Supported Programs, 1994–2004

7. With the exception of a brief gap between June 2002 and April 2003, Bolivia has been in a continuous sequence of IMF arrangements since 1986 (see appendix, Table 6). The first three of these arrangements—a Stand-By Arrangement that was quickly augmented by an SAF, which was in turn replaced by an ESAF that included four annual arrangements—broadly met their objectives, though the ESAF was off track for about a year in 1993–94 because of fiscal slippages related to the 1993 elections, and some structural reforms were delayed. This section focuses on performance under the three most recent arrangements: the 1994–98 ESAF, the 1998–2002 PRGF and the 2003–2004 Stand-By. For background on the main economic and political developments at the time, the reader is referred to Appendix I.

8. The aim of the 1994–98 ESAF supported program was to implement an extensive set of “second generation” reforms that would put Bolivia on a higher growth path and improve public services, while making fiscal adjustments that would ensure that Bolivia could pay for these reforms in the short and medium run (in the long run, the reforms were assumed to pay for themselves, through higher growth and lower burdens on the budget). The package included: (1) a pension reform, from traditional pay-as–you-go system to a fully funded system based on privately managed savings accounts; (2) a privatization program designed to draw foreign investments into 5 major state enterprises (“capitalization”); (3) fiscal decentralization; (4) civil service and judicial reforms; (5) financial sector reforms, including central bank autonomy and new securities and insurance laws; (6) sectoral and infrastructure reforms; and (7) reforms in education and health. The package emphasized the complementarity of reforms; in particular, pension reform and capitalization were to be “closely linked,” with each “adult receiving one share of each of the six companies to be capitalized in an individualized pension account.”

9. While performance under the 1994–98 program is often regarded as impressive (particularly compared to that of other ESAF countries), a closer look reveals a more mixed picture. The program got off to a rocky start after slippages in early 1995, though this was corrected later in the year. Most structural reforms were delayed; for example, the capitalization of the state oil company, envisaged for July 1995, was not completed until May of 1997. As a result, waivers had to be sought for all structural performance criteria under the 1995 and 1996 programs (see appendix, Table 7). This said, core reforms in the areas of capitalization, pension reform, decentralization, and financial sector reform were eventually passed, the fiscal position improved beyond program targets in 1995 and 1996, and after an initial spike, inflation did come down in line with program projections in 1997 and 1998 (Figure 1). However, progress in other areas—civil service reform, the judiciary, health and education—was patchy. The much advertised link between capitalization and pension reform turned out to be less tight than initially suggested: rather than funding the new individual pension accounts, capitalization shares were used to create a new entitlement, the “Bonosol”, which paid a minimum pension to all retirees, including those that retired under the old system. The ultimate objectives of the program—to raise Bolivia’s growth rate to about 6 percent and to make adequate room for the fiscal costs of reforms—were not achieved, for reasons that will be discussed below.

10. The main objectives of the 1998–2002 ESAF/PRGF were (1) to make a new attempt to lift Bolivia’s growth rate into the 6 percent range, through improvements in governance, infrastructure, education and health, as well as labor reform; (2) to clean up after the 1994–1998 program, by mitigating the fiscal consequences of decentralization and taking measures to offset the higher than expected cost of pension reform, and (3) to strengthen the financial sector, by improving the bank resolution process and through the establishment of a bank-financed deposit insurance system. Offsetting the costs of the pension reform meant targeting an improvement in the non-pension balance from about -0.8 percent of GDP in 1996–97 to surpluses after 1999. This was to be achieved through improvements in tax administration, backed by a new tax enforcement code, customs reform, and higher hydrocarbon revenues following the completion of a pipeline to Brazil. Drawing on recommendations of an FAD tax policy mission in early 1999, this package was later extended by a comprehensive tax reform plan which included introduction of a personal income tax. Submission of the tax code and tax reform to congress became structural performance criteria in the 2000 program.

11. Performance under the new program was initially comparable to that under the preceding program. All quantitative performance criteria for 1998 and 1999 were met, but most structural benchmarks were missed. The sole structural performance criterion for 1999 (publication of a bid for privatization of oil refineries) had to be waived. However, there were also some reform successes, notably the passing of a new customs law in July of 1999. Moreover, while not part of program conditionality, the superintendency of banks (SBEF) issued important norms on loan classification and provisioning which strengthened supervision. On the basis of this good overall performance, as well as its track record in implementing social policy actions required under the original HIPC Initiative (completion point in September 1998), Bolivia reached decision point under the enhanced initiative in February of 2000, and a second annual arrangement was approved.

Figure 1.
Figure 1.

Bolivia: Macroeconomic Outcomes versus Program Targets, 1991-2005

Citation: IMF Staff Country Reports 2005, 139; 10.5089/9781451970500.002.A001

Source: Central Bank of Bolivia; and Fund staff estimates.

12. In the first half of 2000, performance under the program began to slip as a result of an “economic stimulus” package introduced by the Banzer administration in response to the deteriorating recession and social unrest (see appendix) and a freeze in fuel prices. Moreover, a draft tax code that was submitted to Congress in April had to be withdrawn because it lacked congressional support, and the authorities argued that for the same reason, the planned tax reform was infeasible. As a result, in August and September of 2000, the program was renegotiated, resulting in a moderate easing of the deficit targets for 2000 and 2001 and the postponement of the tax reform until after the 2002 national election. At the same time, congressional approval of a modified version of the tax code, a law restructuring the internal revenue service (SIN), and a financial sector law including a partial deposit insurance scheme were made prior actions for completing the reviews under the 2000 program, and the presentation of the next annual arrangement to the Executive Board.

13. In the event, the modified fiscal targets and remaining quantitative performance criteria for 2000 were met and the law restructuring the SIN was passed in December, but the tax code passed only in the Senate (April of 2001) and the financial sector law was delayed. Following assurances by the Bolivian authorities that congressional approval of the tax code would be put on a special legislative track (including invoking an extraordinary session of congress in mid 2001, if necessary), Management nevertheless decided to go ahead with the request for a new annual arrangement. In June of 2001, the Board approved a third annual arrangement focused on a streamlined set of structural reforms in the fiscal and financial sectors (passage of the draft tax code and the financial sector law—though without the recommended deposit insurance scheme—became performance criteria for end-September and end-October, respectively). Other reforms envisaged in the original 1998 ESAF request—including legal and judicial reforms, health and education reforms, and labor market reforms, none of which had progressed much since 1998—were dropped from the list of structural benchmarks. At the same meeting, the Board decided that despite the weakening of reforms in 2000–2001, Bolivia had met completion point conditions under the enhanced HIPC Initiative, becoming only the second country to do so. The main reason not to delay the completion point seems to have been a desire to reward Bolivia for its overall reform record since the mid-1990s. In addition, the Fund did not want to be seen as undermining the newly created PRSP process (a PRSP had been completed a few months earlier in Bolivia).

14. Soon after approval of enhanced HIPC debt relief and the third annual arrangement, the fiscal situation deteriorated sharply. End-year targets on the fiscal deficit and domestic financing were missed by wide margins. The proposed tax code failed to pass after meeting strong opposition from the business sector, though the financial sector law was approved, in weaker form, in December. Attempts to bring the program back on track failed, and the first review was not completed. In late 2001, an informally monitored financial program was agreed to help the authorities resist pressures for expenditure increases during the run-up to the elections, and provide a basis for continued disbursements of World Bank and IDB loans; this was ultimately unsuccessful in containing the widening fiscal imbalance and limiting the accumulation of new non-concessional debt. Negotiations on a new PRGF-supported program were initiated in the fall of 2002 after the election of a new government headed by Gonzalo Sánchez de Lozada, but could not be concluded due to social unrest and congressional opposition to tax measures envisaged as prior actions.

15. Faced with an impending financial and political crisis, Bolivia and the Fund finally agreed on a new Stand-By Arrangement as a short term alternative to a new PRGF. The main objective of the program was to stabilize the economy and implement a set of structural reforms that were regarded as critical for returning Bolivia to debt sustainability and strengthening the financial sector. Unlike the proposed PRGF, the SBA envisaged no structural prior actions and could hence be approved quickly, in April of 2003.

16. Performance under the new SBA, which was extended in 2004, has been mixed. Most structural performance criteria related to the financial sector were met (with some delays), corporate sector restructuring lingered, and the revised tax code was finally approved by Congress in August of 2003. Most critically, the authorities were successful in avoiding a financial crisis, even in the face of political unrest leading to the resignation of the President in October of 2003. Bank capitalization was strengthened through actions taken with respect to weak banks in November 2003 and intensified on-site inspections by the SBEF. However, the financial situation remained very fragile, the original deficit targets for 2003 were exceeded by a large margin, and the deficit targets for 2004 needed to be revised upwards by almost two percentage points of GDP relative to the medium-term framework underlying the 2003 program. Several important revenue measures that were originally envisaged under the 2003 program could not be implemented; as a result, the 2004 fiscal program is based on substitute revenue measures that cannot be sustained in the medium term (a tax amnesty, and a financial transactions tax). Following a referendum on hydrocarbons in July of 2004, there remains uncertainty about policies in the hydrocarbons sector and the future of gas exports.

17. Over the last ten years, the Fund has collaborated extensively with the World Bank and IDB on fiscal and financial sector reforms as well as other structural reforms. With respect to the financial sector, the World Bank has generally taken the lead until recently. In particular, the Bank provided significant technical assistance to help create a modern banking law in the early 1990s and has since been actively involved in improving banking supervision and regulations. The Bank also took the lead in advising the authorities on capitalization (privatization). The Fund focused on central bank technical assistance until the late 1990s. Since then, and with the advent of FSAPs, the Fund has increasingly collaborated with the World Bank on financial sector issues. On the fiscal side, all three IFIs have been actively involved. The Fund generally took the lead on tax administration and customs reform, though specific financing for these reforms was provided by the World Bank and IDB. The World Bank took the lead on pension reform and the decentralization program, but beginning in 1997, the Fund has also offered technical assistance in some of these areas.

III. What Went Wrong?

A. Why Has Trend Growth Not Been Higher?

18. Between 1999 and 2002, the Bolivian economy suffered a protracted sequence of external and domestic shocks, including intensified coca eradication, curtailing of external credit lines, devaluation in Brazil, and the Argentina crisis (see appendix for details). This contributed to a credit contraction, a deposit run in 2002, and the political and financial near- crises of 2003. Hence, there is nothing puzzling about Bolivia’s slow growth after 1999. Indeed, one might ask why Bolivia did not do worse, in view of sharper output contractions elsewhere in the region—a question that is answered in part by Bolivia’s lack of exposure to private capital flows, and its continued access to official sources of financing.

19. The real puzzle is why Bolivia has not grown faster over the long term, following its bold reforms in the mid-to–late 1980s and again after 1994. This has been the subject of a substantial body of research in the last five years, as it became clear that structural reforms were not having the desired growth effects (see Kaufmann, Mastruzzi and Zavaleta, 2003, for a survey). The Fund’s 1998 Country Strategy Paper was mainly devoted to this question. Its main finding was that growth in Bolivia was constrained by gaps in the implementation of structural reforms; low public investment and inadequate provision of government services at the municipal level following fiscal decentralization; poor infrastructure; and low confidence related to political instability and the weak rule of law, discouraging private investment.

20. New datasets available since 1998 have confirmed this picture, while adding some important nuances and giving a better sense of how constraints on entrepreneurial activity, institutional weaknesses, and low growth are related in Bolivia. An extensive study by the World Bank (2001) examined the Bolivian manufacturing sector, and found some striking regularities, including very high inventory levels, low rates of capacity utilization, low domestic and international trade (firms tend to operate in local markets; a symptom of high transportation costs and unreliable supply chains), low investment in fixed assets, and a very high rate of informality. According to the study, “90 percent of manufacturing establishments, accounting for 72 percent of the manufacturing labor force, are informal. That is, they operate outside the law and so are deprived of the institutional arrangements that encourage innovations, improve the efficiency of labor markets, and enable economic units to realize economies of scale.” The causes of informality, according to the study, are costs to entering and remaining in the formal sector—red tape, corruption, burdensome labor regulations and tax administration—which deter firms that are small, unconnected, or both. The conclusion is that private sector growth in Bolivia would require a better transportation infrastructure, more predictable customs, lower indirect labor costs, and institutional reforms that improve government services and reduce the barriers to formality, including through better law enforcement and judicial reforms.

21. Several independent studies confirm the pernicious effect of corruption, a weak judiciary and law enforcement and political instability in Bolivia. Transparency International ranks Bolivia as one of the countries with the highest levels of perceived corruption in the world (106 out of 133, with 1 being the least corrupt). According to the World Business Environment Survey (WBES) conducted by the World Bank in cooperation with the EBRD, Harvard University and the IDB in 1999–2000, by far the most important constraints to businesses in Bolivia—in absolute terms and relative to other countries—were corruption, the judiciary, and poor government services. These are followed by poor infrastructure, political instability, burdensome tax administration and regulations, and lack of financing (Figure 2). Within the category of government services, roads, the judiciary, the police and customs rate poorly, while utilities—which were privatized in the mid 1990s—do fairly well. It is also interesting to note that on macroeconomic issues such as inflation, the exchange rate and the quality of the central bank Bolivia comes out well relative to other countries. Finally, according to a widely used set of governance indicators for 199 countries, in 2002 Bolivia ranked far below average with respect to the control of corruption, the rule of law, government effectiveness, and political stability.

22. Why is corruption so widespread and why do institutions remain weak in Bolivia, in spite of anticorruption initiatives and institutional reform projects? A recent study by Kaufman et al. sheds light on this question, in part from a historical perspective, and in part based on a survey of 1250 public officials in 100 public institutions in Bolivia. First, they corroborate the results of the WBES in finding large differences in the quality of public services across institutions: the Central Bank, the Ombudsman, the National Comptroller and the Constitutional Tribunal do well, while Police, Customs and other courts do poorly. Second, their findings suggest that “voice” (accountability to outsiders) and transparency are more relevant in explaining the quality of public services than conventional public sector management variables such as wages. Kaufman et al’s interpretation of these findings goes back to the often cited “clientelism” of the Bolivian political system, which has been based on coalitions between three or four parties of similar ideology and electoral strength. Forming coalitions involve bargaining over public jobs, which are the basis for political patronage: “obtaining electoral and political support by distributing public jobs and other types of rents.” As a result, government agencies are politicized, benefiting party interests and the interests of a well connected and relatively well-off minority. Institutions are successful in providing public services only to the extent that they manage to extricate themselves from the patronage system, and become accountable to society more broadly.

23. In sum, as has been recognized for some time, the root causes of low trend growth in Bolivia are poor infrastructure, poor institutions and government services, and fear of instability. The main insight of the more recent studies is that the political and social roots of these problems are deeper than recognized in the previous literature. At the industry level, these problems manifest themselves in the dichotomy between the formal sector, which benefits mainly powerful or well-connected firms, and a vast informal sector. At the state level, they lead to politicized agencies which—with important exceptions—function as providers of services and rents to their “clients,” at the expense of society as a whole.

Figure 2.
Figure 2.

Bolivia: Constraints to Growth

Citation: IMF Staff Country Reports 2005, 139; 10.5089/9781451970500.002.A001

Source: World Bank, WBES 200, and GRICS 2002.

B. Why Has The Financial Sector Remained Vulnerable?

24. Bolivia undertook important financial sector reforms since the late 1980s. In 1987, an independent superintendency of banks (SBEF) was created. A new banking law establishing minimum provisions for a modern banking system was implemented in 1993, 2 and a new Central Bank Law in 1995. New regulations on private pension funds and bank capital-asset ratios as well as a securities market law were passed in 1997, and in 1999 the SBEF also introduced stricter loan classification, and provisioning policies. Finally, a law establishing a comprehensive bank resolution framework was approved in December of 2001. Yet, Bolivia found itself at the brink of a systemic financial crisis in 2003. Why?

25. There are four main reasons for continued financial sector vulnerability in Bolivia: the timing of financial sector reforms; political interference which prevented or undercut critical reforms; financial dollarization, and deposit and loan concentration:

  • While financial sector reforms were continuously undertaken, the credit boom of the 1990s preceded the most important reforms. In the late 1980s and early 1990s, the banking system was operating under confusing and conflicting laws and regulations, and the coordination between the Banking Superintendency (SBEF) and the Central Bank (BCB) was less than satisfactory. Lending practices were distorted by strong links between the banking and corporate sector, which resulted in a high degree of related lending. In addition, given the lack of proper governance, management, and control systems in banks, lending practices were guided by an excessive reliance on collateral, rather than capacity to repay, in a period when property prices were experiencing a bubble. These practices only began to change in the late 1990s, as a result of regulatory changes in 1997 and 1999, and the entry of foreign banks (Morales, 2004);

  • The reform process has been undercut by political interference, in part due to lobbying by powerful local bankers. For instance, the adoption of a partial deposit insurance scheme found strong political opposition in Parliament, leaving the burden of an implicit full deposit guarantee to the central bank and hampering the bank resolution framework. In 1995 and 2001, the authorities created special funds (FONDESIF and PROFOP) to recapitalize weak local banks and help restructure their loans. These fed expectations of corporate bail-outs, thus discouraging corporations from restructuring their operations. Finally, in 2001–2002, the government undertook several measures that undermined the autonomy of the SBEF, through decrees introducing forbearance in loan classification and provision; and the establishment of new regulatory and “coordinating” bodies dominated by government representatives; 3

  • Banks are vulnerable to indirect exchange rate risk: while they do not suffer a currency mismatch due to dollarization of both deposits and loans, their private sector borrowers do. Indeed, almost two-third of bank credit is estimated to finance the non-tradable sector, i.e. borrowers with no regular source of income in foreign currency. This was one reason for the sharp rise in nonperforming loans in the 1999–2002 period, when the Boliviano depreciated by about 30 percent against the U.S. dollar;

  • Deposits and loans remain very concentrated, in part as a result of the high concentrated income distribution in Bolivia. Almost 60 percent of total deposits are held in one percent of the total number of deposit accounts. Hence, the system is vulnerable to withdrawals by a few large depositors.

26. Notwithstanding these shortcomings, the reforms have significantly improved the institutional setting, the capacity of financial authorities to handle crisis situations as well as of banks to manage their liquidity. The efforts made by the SBEF and the central bank to strengthen capitalization and liquidity of the Bolivian financial system, are the main reason why Bolivia managed to survive four deposit runs in less than three years. However, the system remains fragile because of the financial dollarization problem, as well as continuing political and social tensions, which are an impediment to further structural reforms as well as potential triggers of deposit runs. Furthermore, a weak legal framework for corporate restructuring remains a key impediment to reducing the vulnerability of creditor banks.

C. Why Did The Fiscal Situation Unravel?

27. Between 1996 and 2002, the defict of the combined public sector (after grants) rose from just under 2 percent to almost 9 percent of GDP. As Table 2 shows, the increase happened in two phases: first, an increase by almost 3 percent of GDP by 1998 (which was partly rolled back in 1999 and 2000); followed by an increase of over 5 percentage points between 2000 and 2002. The sources of these increases are now analyzed based on a decomposition of changes in the deficit.

Table 2.

Decomposition of Changes in the Fiscal Deficit, 1996–2004

(in percent of GDP)

article image
Source: Bolivian authorities, staff reports, and staff estimates

Includes pension contributions, which are eliminated in 1998, as the pension reform was implemented.

Includes privatization receipts.

28. Table 2 shows that the initial deterioration in the public accounts, after 1995, was mostly due to a surge in the pension-related deficit, by 4 percentage points between 1995 and 2000. The increase in the pension costs was only partially compensated by privatization revenues and lower capital and interest spending (most privatization receipts did not benefit the budget, as the shares of partially privatized large enterprises were used to subsidize the “Bonosol,” see paragraph 9). Excluding pension contributions, tax revenues grew only marginally after 1996—not enough to compensate for the loss of the pension-related revenues. As a result, the financing needs of the public sector grew. By 1999, as temporary sources of revenue (including privatization receipts) began to dry up, Bolivia’s fiscal position was vulnerable to the negative shocks that it endured in the period ahead.

29. The surge in the pension-related deficit reflected the far higher than expected costs associated with the 1997 pension reform and its subsequent implementation (see figure). While initially expected to reflect the loss of pension contributions to the budget, around 1¼ of GDP, the actual increase in annual costs reached 4½ of GDP. Three sets of factors contributed to this increase. First, the problems of the old pay-as-you-go system—including unsustainably high benefits, low retirement age, and poor administration—were not addressed before adopting the new system. This created a strong incentive for early retirement under the generous old regime, including through fraud. By 2001, the number of retirees under the old system exceeded initial (1997) projections by 50 percent. This could have been avoided, at least partly, by a more gradual transition involving a reduction in benefits under the old system, and a strengthening of administrative capacity. Second, in the process of shutting down the old system, the government took a number of decisions which significantly raised pensions costs. Rather than being reduced, benefits under the old system were increased three times: at the time of the reform, and twice in 2001. Furthermore, the government took over the liabilities of mostly insolvent “complementary pension funds” (the private sector pillar of the old system) at a cost of about 1½ percent of GDP on an annual basis, and allowed certain groups to benefit from early retirement under the old system (not envisaged in the original law). (In addition, the wages of public employees ended up being raised to compensate for the higher premiums under the new system.) Finally, certain macroeconomic and demographic assumptions proved too optimistic (see Cueva and Revilla, 2003, for details).


Pension deficit projections after pension reform

Citation: IMF Staff Country Reports 2005, 139; 10.5089/9781451970500.002.A001

1/ Assumes a fixed cost of pension reform of 0.5 percent of GDP. Here is added to the actual pension deficit in 1995, at which time the pension reform was supposed to take place.2/ based on study by UDAPE (1997) and Selected Issues (2003). Assumes reform takes place in 1997.

30. In contrast, the steep deterioration of the fiscal position after 2000 reflected mainly a surge in non-pension spending. As growth continued to falter, the authorities attempted to stimulate the economy via a countercyclical fiscal policy. In 2001 and 2002, non-pension spending grew by over 4 percentage points of GDP (Table 2, last column), or 27 percent in nominal terms—against nominal GDP growth of less than 8 percent—with capital spending and wages accounting for half of the increase. Spending related to the enhanced HIPC Initiative accounted for less than one-sixth of the total non-pension spending increase. The rise in interest payments, starting in 2000, occurred in spite of HIPC debt relief as a result of accelerating domestic debt accumulation. The decentralization process also created fiscal tensions, as the accumulation of municipal debt led to a partial bail out by the central government and a debt restructuring, which recognized arrears to the private sector (Box 2).

31. Policy measures in 2000–2001 also had an effect on the revenue side. Domestic fuel prices were frozen in 2000, leading to a gradual decline in excise taxes. Revenue collection by customs fell steadily, due to a reduction in tariffs, new tax exemptions, and the inability to contain smuggling and tax evasion, reflecting the undermining of the customs reform initiated in 1999. Tax revenues from higher gas and oil exports compensated for the loss in the non-hydrocarbons tax revenues, but not for the decline in nontax revenues, particularly as the privatization process came to an end. The postponement of tax reforms further undermined attempts to tighten the non-pension fiscal position.

32. The actions leading to the fiscal unraveling in 2001–2002—increases in expenditures, and the fuel price freeze—reflected in part a decision by the government to run a “countercyclical fiscal policy” through employment programs and high capital spending, and in part pressures by interest groups. The latter came from two quarters: business groups and banks, which called for a demand stimulus and stood to benefit from a freeze on diesel fuel; and public sector workers including teachers and the police. Business groups were also the main source of opposition against the introduction of a tax code, which would have ended the practice of withholding tax payments while tax disputes were adjudicated in court.

The Unintended Consequences of Fiscal Decentralization

The decentralization process began in 1985, but gained momentum during 1994–97, with the passage of the popular participation and decentralization laws. The main objectives were to improve the democratic process and the quality of public services at the local level, with an increased focus on pro-poor pending in recent years. However, the decentralization process has also created fiscal problems, in part due to an imbalance between revenue allocation and spending responsibilities, and the absence of effective mechanisms to control and monitor the quality of local spending.

In 2000, a financial restructuring plan (PRF) was implemented to reduce the high debt levels of some large municipalities. As decentralization progressed, the absence of a link between spending decisions and revenues (largely transfers), created incentives for debt accumulation by local governments. Under the PRF, short term debt (mainly arrears to the private sector), was converted to longer-term debt in 2000–01, and participating local governments were subjected to greater monitoring and spending control. However, since the PRF incorporated a bail-out scheme it also generated incentives for future accumulation of debt, particularly as municipalities graduated from the PRF.

While the PRF and the HIPC-related transfers have helped improve the fiscal position of local governments, the structural imbalances created by the decentralization process remain. Following the 2000–01 National Dialogue, resources associated with the Enhanced HIPC Initiative are transferred automatically to municipalities, based on local poverty levels. These represent a significant new source of revenues, particularly for the poorer regions, and have helped to improve local public finances. In some cases, particularly small municipalities, revenues more than doubled from 2001 to 2002. However, the design of the decentralization process also created pressures on the central government, as it lost revenues to the regions while maintaining many of the spending responsibilities. In addition, as HIPC resources decline, there will likely be further pressures in the system. In particular, the accumulation of deposits by local governments has given the central government more room for borrowing from the Central Bank. A fall in these deposits could lead to either a monetary expansion or a credit crunch on the side of the central government.

The decentralization has also contributed to lower efficiency and transparency in fiscal policy, mainly due to a complex transfer system with excessive earmarking, limited ability to generate own resources at the local level, and lack of accountability of subnational governments. The dependency on transfers from central government and absence of clear spending responsibilities have weakened local government budget constraints: local governments can always exert pressure on the treasury for more resources. In addition, excessive earmarking has led to a divorce between spending decisions and actual needs. Monitoring of local governments’ budget execution remains poor, due to municipalities’ increased autonomy and weak control mechanisms, and a lack of effective sanctions in the case of untimely reporting. This prevents effective monitoring of municipal debt and spending, including pro-poor spending.

D. Did the HIPC Initiative Fail in Bolivia?

33. As noted in Section II, public sector nominal external debt at around US$ 5 billion in Bolivia is higher today than prior to enhanced HIPC debt relief in 2001 (see Table 3 as well as Table 5, Appendix). Indeed, it is higher than at any time in Bolivia’s history. In NPV terms, external debt increased to $2.9 billion at end-2003, almost 55 percent above projections at the enhanced HIPC completion point, while the external debt-to-export ratio increased from 117 percent at enhanced HIPC completion point to 176 percent at end-2003. Taking into account domestic debt, which has almost doubled since end-2000, the public debt burden has climbed to the precariously high level of 70 percent of GDP—about 14 points higher than prior to enhanced HIPC debt relief (see Table 5, Appendix). 4 In NPV terms, total public debt has climbed from 37 percent of GDP at enhanced HIPC completion point to 51.5 percent at end-2003.

Table 3.

Public Debt, 1997–2003

(in millions of US$)

article image

Source: Background work for Heavily Indebted Poor Countries (HIPC) Initiative-Status of Implementation, August 2004 ( The NPV debt values reported here differ from those underlying Box 4 due to differences in coverage (they refer to the consolidated public sector including the financial public sector, whereas the calculations in Box 4 refer to the non-financial public sector), and assumptions about discount rates and the delivery of HIPC assistance.

Source: Completion Point Document for the Enhanced HIPC, May 2001.

34. While GDP growth has fallen considerably short of what was expected at enhanced HIPC completion point (Table 3), a debt dynamics decomposition shows that the rapid accumulation of total public debt to GDP since the HIPC Initiative is largely due to the sharp widening of the primary deficit that began in 2001, which more than offset about 20 points of GDP in debt relief received in 1999 and 2001 (Table 4). This contrasts with the experience of most emerging market countries with debt difficulties, whose increasing real debt burdens in the 1990s were accounted for primarily by interest costs and real exchange rate depreciations (Cohen and Portes, 2004). In the case of Bolivia, interest costs have not so far been a significant factor driving the accumulation of debt—though they have been rising in recent years, as a result of the increasing share of domestic debt—and the effect of large primary deficits eclipsed the effect from real depreciations during 2001–02.

Table 4.

Debt Dynamics Decomposition, 1996–2004

(in percent of GDP unless otherwise stated)

article image

Reflects debt relief and other debt stock adjustments, privatization receipts, and measurement error.

35. How was Bolivia able to obtain the extensive additional financing that was required to cover its large deficits in 2001 and 2002, particularly in view of the fact that the program with the Fund was off track? Part of the answer is that almost 40 percent in the increase in the face value of total public debt between 2001 and 2003 came from domestic sources (Table 3). Of the remaining 60 percent, about half came from non-concessional lending by the Andean Development Corporation. In addition, Bolivia continued to receive concessional financing from the IDB and the World Bank, aided by “comfort letters” from the IMF based on its informally monitored program with Bolivia.

36. In sum, debt relief under the enhanced HIPC Initiative failed to put Bolivia on a sustainable lower debt path—not because the amount of debt relief granted was too low, but because shortly after receiving debt relief the country ran sharply higher deficits, financed mainly through new non-concessional debt, which more than offset the debt relief granted. As a result, under all but the most optimistic assumptions about hydrocarbon revenues (Box 3), Bolivia will need substantial fiscal adjustment to reverse the unsustainable trend in its public finances. Since a stated objective of the HIPC Initiative is to make debt relief available to countries with “restrained debt management policies—ensuring that new borrowing is not excessive, not too expensive, and can be repaid” (IMF, 2000a), it is hard to avoid the conclusion that the initiative was not well applied in this case. Starting in 1997, it was clear that Bolivia would need considerable fiscal adjustment to offset the medium term fiscal costs of structural reforms, particularly the pension reform. The 1998–2001 program hence envisaged major structural fiscal reforms, including better enforcement through a new tax code, and a reform of the tax system. By agreeing to HIPC completion point without insisting on these reforms as prior actions (paragraph 13), the IMF and the World Bank removed a critical incentive for their adoption. Without these reforms, spending pressures and lower revenue led to the large fiscal overruns in 2001 and 2002. Delaying the HIPC completion point until the reforms were completed may not have prevented the 2001 program from going off track, but it would have created a much stronger basis for preventing a new debt problem after Bolivia had received its second round of debt relief.

37. It is also possible that the HIPC Initiative and the associated PRSP process may have undermined themselves by raising spending expectations to unrealistic levels. A 2000 survey of 400 persons representative of the populations of La Paz and El Alto suggested that a large majority, particularly among the poor and the middle class, viewed the “national dialogue” on which the PRSP process was built as a “just a political maneuver” to bring in a billion dollars of debt relief (Andersen and Nina, 2001). At the same time, the HIPC initiative raised expectations on part of the municipalities, who submitted demands for funds in 2000 which were on average more than twice as high as in 1997, and were impossible to meet with HIPC assistance and the municipalities’ normal sources of income. This gap between expectations and outcomes might have contributed to the spending pressures and the social unrest of 2003.

38. The HIPC Initiative has had more success from the perspective of increasing social spending, which contributed to the large improvement in social indicators observed between 1992 and 2001. Under the enhanced initiative, the contribution of debt relief to social spending went up from 0.1 percent of GDP in 2001 to 0.8 percent in 2003. However, there are questions about the quality of spending in the health and education sectors, which account for over 70 percent of “pro-poor” spending, and are largely allocated to wages. The high wage bill crowds out spending on priority nonwage goods, such as textbooks, and while a large share of the education budget is directed to universities, which mostly benefit high-income households, secondary education is underfunded (World Bank and IDB, 2004).

Hydrocarbon Revenues and Debt Sustainability

Does the expected growth of the hydrocarbon sector, which generates almost one third of Bolivian exports, obviate the need for fiscal adjustment? To answer this question, consider first a baseline projection for gas exports, which includes continued gas exports to Brazil under the current contract and exports to Argentina, which started in 2004. Assuming that the non-hydrocarbon primary fiscal deficit is unchanged from its current level of 4.4 percent of GDP, as well as a medium term GDP growth rate of 4 percent, the NPV debt to GDP ratio would continue to rise in the short term and eventually stabilize at about 60 percent of GDP (see Figure, Scenario I). This is unsustainably high for a country with comparatively weak institutions (see IMF and IDA, 2004); in the sense that a relatively minor shock, such as a recession, could trigger a debt crisis. Hence, although hydrocarbon revenues are expected to lighten the burden of fiscal adjustment for Bolivia, they do not remove it.

To obtain a sense of the minimum medium term fiscal adjustment required under the baseline hydrocarbons assumption, one can ask what adjustment it would take to reduce the NPV debt to GDP ratio to around 45 percent by 2015. This corresponds to just under 30 percent of the NPV of external debt to GDP, the upper bound of what is considered sustainable for a county with weak institutions. As shown in the table (Scenario II), the non-hydrocarbons primary fiscal deficit would need to be adjusted by about 1.7 percentage points over 10 years, while also directing all extra hydrocarbon revenues to debt reduction (which may be difficult, given revenue-sharing arrangements with the regions and political pressures to earmark further revenues for specific spending).

Finally, how fast would hydrocarbon exports need to rise in the medium term to achieve a comparable debt reduction without fiscal adjustment? Scenario III is based on an optimistic scenario which assumes the same path until 2008 as the baseline (higher exports in the near term are unrealistic) after which gas production increases to around 50 percent above the baseline by 2015. Such an increase is very uncertain. Moreover, even under this optimistic scenario, the debt would continue to rise and remain at very high levels until about 2010. Hence, to reduce Bolivia’s high fiscal and macroeconomic vulnerabilities within the next five years, fiscal adjustment seems necessary, even under optimistic hydrocarbon projections.

Table 1.

Bolivia: Deficit Paths under Alternative Hydrocarbon Assumptions

(in percent of GDP)

article image

Bolivia: DSA Scenarios (NPV of Debt to GDP)

Citation: IMF Staff Country Reports 2005, 139; 10.5089/9781451970500.002.A001

E. Was the 1996 Hydrocarbons Strategy a Mistake?

39. In 1996, Bolivia introduced a new hydrocarbons law and partially privatized the state oil company, YPFB. One motivation for the reform was that YPFB did not have the financial capacity to undertake the large investments needed to even maintain the production levels over the medium-term. Hence, the objective was to create a legal framework that would attract private investment to the sector, be revenue neutral in the short run, and eventually increase revenues, as hydrocarbon production and export would increase. A new tax system introduced “downstream” levies (excises on fuel products) to compensate for the expected fall in “upstream” taxes (mainly royalties) as a result of privatization, moved from product taxation to profit taxation—including a surtax on exceptional profits, and differentiated tax treatment between “old” (under exploitation in 1996) and “new” fields, which would pay lower royalties to promote investment. The level of royalties for new fields was brought to an internationally competitive level.

40. The overall experience with the reform has been mixed. Measured against its first objective—attracting investment—the 1996 hydrocarbons strategy was clearly successful, leading to a surge in investment of about $2.5 billion dollars between 1997 and 2002, and the discovery of large reserves of gas. In part as a result of the new investments, gas production more than doubled, in volume terms, between 1998 and 2002. However, the revenue performance of the hydrocarbons sector has been disappointing. Royalties dropped by a large amount in 1997, and other hydrocarbon taxes have been flat (see text figure). In addition, the gradual recovery of royalties since 1997 has been offset by a decline in excise taxes as a share of GDP, after domestic fuel prices were frozen in 2000. Hydrocarbon revenues as a share of GDP stayed roughly at pre-reform levels.


Total hydrocarbon revenues to the budget

Citation: IMF Staff Country Reports 2005, 139; 10.5089/9781451970500.002.A001

41. Fortunately, the disappointing performance of revenues so far (other than that of excise taxes) seems to be mostly attributable to factors that are either temporary or can be addressed through improvements in regulations and tax administration. Part of the initial drop in royalties in 1997 was due to a temporary fall in exports; however, a generous (i.e., narrow) definition of ‘old’ (higher royalties) relative to ‘new’ fields also played a role. 5 The latter was the result of a deliberate strategy to make the YPFB capitalization offer attractive to investors in order to ensure its success. Moreover, as new fields have been developed, the share of hydrocarbons produced in “old” fields has declined, reducing the overall government take. Second, large initial investments were amortized, temporarily reducing profits and hence tax receipts. Third, the design of the surtax, intended to partially compensate for the fall in royalties, was relatively generous in terms of admissible deductions allowed. Finally, the level of compliance may have declined over time, as the authorities have not adjusted quickly to the new tax system, while oil companies have learned to exploit loopholes. These problems can be addressed by strengthening regulations in areas such as transfer pricing and amortization as well as the institutional capacity of the tax agency.

IV. The Role of the IMF

A. Was Bolivia Encouraged to Undertake Reforms it Could Ill Afford?

42. As argued in Section III.C., two of the central structural reforms of the mid-1990—the pension reform, and decentralization—had unintended consequences which contributed significantly to the deterioration of the fiscal position that began in 1997. For the reasons explained in paragraph 29, the fiscal costs associated with the pension reform were much higher than expected, creating a significant adjustment burden over the medium term; while decentralization made it more difficult to control spending. The question is whether these potential problems were appropriately foreseen by the Fund, and whether steps were taken to minimize them.

43. At the outset, it is important to note that both sets of reforms were highly “owned” by the authorities, and that the Fund was not Bolivia’s main source of technical assistance on how these reforms were carried out. Moreover, the process by which the authorities decided to phase out the old system—which included benefit increases, the bail-out of the “complementary pensions funds,” and privileged treatment of specific groups—was unambiguously the responsibility of the Bolivian administration. Thus, it is not true that the Fund pushed Bolivia into an “unnecessary and costly” pension reform, as claimed by some (Comisión de Revisión del Gasto Público, 2004). This said, the Fund agreed with the Bolivian government and the other IFIs that a pension reform was necessary, in light of the financial unsustainability of the old system, its low coverage, and the hope that a fully funded system would stimulate private savings and growth. Hence, implementing a new pension reform law and establishing a regulatory system for privately managed pension funds was made a structural benchmark under the 1994–98 ESAF supported program.

44. While the Fund does not bear the responsibility for the pension reform, it should nevertheless have been more vigilant in recognizing its fiscal risks early on and pushing for remedial action. Strikingly, briefing papers and back to office reports written around the time of the 1994 program negotiations were very concerned about the fiscal costs of the overall reform package, but these concerns focused on other issues, such as severance payments to public sector workers, and the wage consequences of the education reform. Between 1994 and 1996, staff reports rarely mention the pension reform, and when they do, they tend to emphasize its potential benefits. In late 1992, a technical assistance mission had suggested possible risks, particularly regarding absorption of complementary pension funds, but this does not seem to have had an impact. Similarly, the equity-imbalance between the old and new system, and the timetable and modalities for shutting down the old system, were not discussed in Fund documents before the reform or even in the early stages of its implementation. The first full discussion of the potential costs of reform, and measures to offset it, occurred in the context of the 1997 Article IV Consultations. By that time, critical aspects of the reform were a fait accompli.

45. With respect to fiscal decentralization, the role of the Fund was somewhat different. First, the fiscal decentralization process was well underway prior to approval of the ESAF request in December of 1994 (one of its main pillars, the popular participation law, was approved in April of 1994). Unlike the pension reform, staff did not advocate quick implementation of the authorities’ decentralization plans, and no structural benchmarks were set in this area. Second, the Fund played an important role in attempting to contain the unintended consequences of fiscal decentralization once they became apparent (Box 2). Starting in 1997, the Fund has provided extensive technical assistance in this area—with some success—particularly towards improving the revenue sharing mechanism and the control of spending and debt levels. However, the situation remains precarious, and the Fund may well need to increase its involvement, including through structural conditionality, to push for further reforms.

B. Were Financial Sector Weaknesses Ignored?

46. The Fund recognized weaknesses in Bolivia’s financial system early on. In a 1987 report on the reform of Bolivia’s financial sector, an IMF-TA mission clearly pointed out the risks of high dollarization as well as the need for modernizing the central bank, revising and updating financial legislation and regulations, strengthening bank supervision, and introducing a deposit insurance system to ensure financial stability in the long run. Both the 1994–98 and the 1998–2001 programs contained extensive financial sector conditionality. This raises the question of why financial sector reforms, by and large, came too late to prevent significant problems in the banking system beginning in the mid 1990s (see Section III.B.), and why dollarization continues unabated. What role did Fund advice and conditionality play in that context?

47. A review of financial sector conditionality and technical assistance provided by the Fund shows that while the Fund was active in the financial sector area, most emphasis prior to 1997 was on the reform of the central bank as well as on the resolution of problem banks identified in 1995. Reforming the prudential and regulatory framework did not become the subject of IMF technical assistance until mid 1996, when a mission on banking supervision was dispatched for the first time, and only became part of IMF conditionality in the context of the third annual arrangement under the ESAF approved in September 1997. There are three plausible reasons for this. First, under the division of labor between the Bretton Woods sisters, financial sector reforms traditionally fell under the remit of the World Bank. In particular, the Bank provided extensive support to Bolivia during a first round of financial sector reforms in the late 1980s and early 1990s leading to the General Banking Law of 1993, and in the context of its 1991–95 Structural Adjustment Program. Second, regulations raising minimum capital-asset ratios and applying Basel standards for asset evaluations were tied to the approval of a new central bank law, which the Fund focused its conditionality on, and which was delayed until October of 1995. Finally, prior to the experience of the Asian crisis, the role of financial sector regulation in crisis prevention may not have been as keenly on the staff’s mind as it is today. This could explain why the 1996 second annual arrangement under the ESAF contained no conditionality in this area even though the central bank law had been passed at that point. In the event, the new Basel-consistent regulations were only issued in August of 1997, and the new standards did not come into effect until 1998–99.

48. After 1997, the Fund became more involved in banking issues, but may not have been sufficiently forceful in pressing ahead with some crucial reforms, such as deposit insurance (still inadequate) and consolidated supervision of financial conglomerates (still ongoing); highlighting moral hazard problems in government-supported recapitalization programs; and insisting on the reversal of decrees introducing regulatory forbearance. Beginning in 2000, concerns about economic, political and social fragility may have led the staff to compromise on financial sector reforms, even when this implied a loss in reform momentum. Backed by the recommendation of a Financial Sector Stability Assessment conducted in 2002–03, the staff has recently taken a firmer stance in encouraging bank restructuring and SBEF independence, while defusing political interference in the financial sector. In addition, in the last two years Bolivia has received extensive technical assistance in the areas of banking system liquidity management, regulation and supervision of bank credit risks, early warning systems, and increased use of domestic currency in the financial system.

49. Concerning dollarization, the Fund’s strategy hinged primarily upon the implementation of consistent macroeconomic policies but also envisaged an array of “structural measures”—including higher reserve requirements for foreign currency deposits, a small tax on dollar transactions, and changes in prudential regulations to internalize the credit risk embedded in the currency mismatch of borrowers’ balance sheet. 6 Changes in the exchange rate regime were considered, but the Fund’s advice changed over time in this respect, from advocating a rigid peg, to enhancing the credibility of a stable boliviano, to a more flexible exchange rate regime anchored on inflation targeting. 7 Moreover, the Fund never insisted on regime change, accepting the authorities’ view that, in the circumstances, the prevailing system (an informal crawl designed to keep inflation low while maintaining a competitive real exchange rate) had “served Bolivia well.” In the event, dollarization persisted, for a variety or reasons including the exchange rate regime, monetary and institutional credibility, and implicit guarantees in the financial system (Box 4).

The Causes of Dollarization Persistence

Dollarization has a long history in Bolivia, dating back to the abandonment of gold convertibility in the early 1930s (Morales, 2003). Yet it is puzzling why financial dollarization has persisted in Bolivia, in spite of almost two decades of moderate or low inflation and regulations favoring Boliviano deposits. Recent research, inspired in part by Bolivia’s dollarization “puzzle” as well the example of other Latin American and emerging market countries, suggests four factors that may have contributed to this hysteresis.

First, lack of macroeconomic credibility. The problem here is that in spite of the stability of the currency since the late 1980s, depositors and firms in Bolivia may still worry that high inflation and/or a sharp devaluation might happen in the future. This need not be driven by “long memories” of hyperinflation, but could reflect the view that the fiscal situation remains precarious, as well as a broad lack of faith in society’s current “institutions of conflict management”—a lack of confidence that the political system could handle a major disruption without resorting to desperate measures (Rajan, 2004). This lack of credibility encourages simple forms of contracts that are the most conflict-immune, including, in the financial system, demand deposits denominated in dollars. It also encourages borrowing in dollars by firms, as a “peso problem” may make borrowing in local currency prohibitively expensive unless a large inflation or devaluation actually materializes (Jeanne, 2003).

Second, the exchange rate regime. If households are interested in minimizing the volatility of the value of their financial assets in terms of consumption goods, then it might make sense to hold them in dollars rather than Bolivianos—even with moderate inflation volatility—as long as real exchange rate volatility is very low (Ize and Levy Yeyati, 2003). By targeting the real exchange rate, exchange rate policy in Bolivia might thus have contributed to the persistence of deposit dollarization.

Third, implicit guarantees in the corporate and financial sectors (Burnside, Eichenbaum and Rebelo, 2001). As argued in section III.B., the implicit guarantees of the banking system shouldered by the central bank, the creation of special funds to recapitalize weak banks, and the absence of an effective bankruptcy framework may all contribute to the perception that corporations that become insolvent as a result of a devaluation will be bailed out. As a result, the lower cost of dollar borrowing in the corporate sector is not offset by higher risk.

Finally, thin financial markets in local currency increase the costs of intermediation in Bolivianos, which in turn encourages dollar intermediation, keeping Boliviano financial markets thin (Ize and Powell, 2004).

C. Was the Need for Institutional Reforms Underestimated?

50. A casual reader of the literature on Bolivia’s growth experience in the 1990s might be led to conclude that “macro-structural” reforms supported by the IMF failed to raise growth because they ignored the need to improve institutions. In fact, institutional reform has featured quite prominently in all Fund supported programs, particularly in the period between 1994 and 2000. However—notwithstanding some important successes—implementation of most of these reforms was poor, and as a whole they proved insufficient in changing the character of the Bolivian public sector.

51. The basic strategy for institutional reform in the authorities’ 1994–98 program was exceptionally coherent—in many ways, it reads as if it had been designed in response to the literature on institutional weaknesses in Bolivia that was written several years later. Not only did the program call for the reform of critical institutions—including the judiciary, the civil service, and the central bank—but it emphasized the need for widening political participation and redistributing resources to the disenfranchised. The objective of the political participation program, according to the 1994–1997 Policy Framework Paper, was to “incorporate indigenous, rural and urban communities into the political and economic life of the country, and to improve the quality of life for the Bolivian people through a more equitable allocation of resources.” The paper also called for a major reform to strengthen basic education, and improve health services. The 1998–2001 program emphasized intensified education and health reform, and public sector reforms including judicial reform, tax administration reform, customs reform, and civil service reforms.

52. What happened to these reforms? Political participation and decentralization was a mixed success. It led to substantially higher spending at the local level and contributed to improved social indicators, but strained the capacity of municipal institutions and caused municipal debt problems. Most importantly, it has had limited success in opening political decision-making to the disenfranchised, as “the clientelistic culture of the central government simply spread to the local level” (Kaufmann et al., 2003). Education reform led to an increase in the number of schools and educational coverage, but serious problems with the quality of education remain. Central bank reform was a success. Civil service reform, however, was a failure, with virtually no progress. Judiciary reform led to new institutions such as the Magistracy Council, the Constitutional Tribunal and the Ombudsman’s Office, but according to the World Bank (2004), “the judicial power has not been able to escape political pressure, and neither the magistracy council nor the constitutional tribunal has been able to function properly.” The customs reform was initially a success, with substantial progress in 1999 and 2000, but since then there have been attempts to re-politicize the agency, and by now, the gains from the initial reform have largely been reversed.

53. Does the Fund share blame for this generally disappointing outcome? At the outset, it is important to note that notwithstanding the inclusion of wide-ranging institutional reform plans in the 1994–98 and 1998–2001 Fund-supported programs, the Fund’s enthusiasm for various institutional reform plans varied considerably, and hence its willingness to fight for them. As stated earlier, the staff did not initially take great interest in the political participation and decentralization programs. Similarly, there were no structural performance criteria or structural benchmarks in the judicial and civil service areas, and they received scant attention in staff reports. In this sense, the need for fundamental institutional reforms was indeed underestimated, at least until the 1998–2001 program, which gave more prominence to education reforms, made customs reform one of its main priorities, and included structural benchmarks for judicial reforms.

54. The critical question is whether the Fund could have achieved more by pushing harder for a broad range of governance-related institutional reforms. Against this view, one could argue that the Fund’s structural conditionality needed to be parsimonious in order to be effective, and that the reforms which it chose to back were the most macro-critical. However, the recent literature on growth constraints in Bolivia implies that improving the quality of government services in Bolivia is in fact growth-critical. Moreover, the Fund was reasonably successful in areas in which it concentrated its efforts: privatization, central bank reform, tax administration reforms and (initially at least) customs reform. Clearly, whether or not the Fund threw its weight behind a particular reform plan had a significant impact on its chance of success, particularly in the 1994–99 period. Thus, more emphasis on civil service reform and judiciary reform in that period would have been desirable, even if other IFIs would have had to take the lead on specifics.

55. Regardless of which view is correct, two factors made it more difficult over time to push for a broad institutional reform agenda. First, as the costs of pensions continued to rise, economic activity began to weaken, and social pressures increased (see paragraph 32) specific fiscal-structural reforms gained priority relative to broader institutional reforms such as civil service reform. Second, the resistance of traditional elites and interest groups to new reforms became more intense over time, simply because the elements of the 1994–98 reform program that were broadly supported by these groups—such as privatization and the pension reform—were the first to be implemented. At the same time, the government became politically weaker. As a result, beginning in 2000, the government was unable to implement even reforms which the Fund viewed as critical, such as systemic tax reform and the tax code. Since 2001, Fund conditionality has concentrated on critical fiscal-structural and financial sector reforms, as the focus of conditionality shifted from growth to fiscal sustainability and crisis prevention.

D. Did Recommended Fiscal Adjustment Fall Short?

56. Given that the deterioration of Bolivia’s fiscal position that began in 1997 occurred mostly “on the Fund’s watch,” it is natural to ask whether the Fund should have been more forceful in pushing for fiscal adjustment, or more conservative in the way in which fiscal targets were designed. Since the rise in the 1997 deficit was largely due to the unexpected costs of the pension reform, it makes sense to ask the question separately for three periods: first, from 1998 to 2000; second, for the third annual arrangement under the PRGF approved in 2001, and finally, for the current Stand-By Arrangement.

57. As it became clear that the pension costs under the old regime would be higher than expected, the 1998–99 programs incorporated some measures to gradually contain the overall deficit, such as a large increase in excise taxes, which was partly implemented. However, the program did not make an attempt to reduce or contain the pension deficit more directly, by resisting increases in benefits, or calling for reforms to control fraud. In spite of temporary measures (including a tax amnesty), the primary deficit (before grants) grew to 4 percent of GDP in 1998, a deterioration of about 1½ percentage points in one year. The unexpected drop in economic growth in 1999 further undermined any efforts to implement sustainable measures to compensate for the increase in the pension deficit. It is difficult to assess if it would have been possible to have a significant larger fiscal effort starting in 1998, particularly as it was not expected that growth would collapse in the years ahead.

58. A striking characteristic of the 2001 fiscal program is that it envisaged large spending increases as a result of an optimistic nominal GDP growth projection (4 percent real growth plus 4.3 percent for the GDP deflator). This projection was based on the assumption that following an incipient recovery of the economy in 2000, growth would pick up in 2001. But by the time the program went to the Board in early June, there were mounting signs—including a public pronouncement by the central bank president—that the projected growth rate was too high. Nevertheless, the program was not revised, in part to maintain consistency with PRSP completed a few months earlier, as the Fund did not want to be seen as undermining the newly created PRSP process and the ongoing enhanced HIPC Initiative. Given the nominal growth outturn for the year (1.5 percent in real terms and 0.7 percent for the deflator) programmed spending increases would have made it very difficult to meet the deficit target, particularly since compensatory revenue measures were constrained by the lack of structural reforms prior to approval of the program (see paragraph 21). In the event, the authorities exceeded even the high programmed spending increase by a significant margin.

59. Like the 2001 arrangement, the 2003 SBA contained revenue assumptions that turned out to be too optimistic. But unlike 2001, this was initially compensated by cuts in spending. Only after the political and social crisis in October did the fiscal position rapidly deteriorate with the deficit target breached by a large margin. This could not have been avoided by programming more conservative spending targets at the outset. As far as the 2004 program is concerned, the problem is not so much the degree of programmed adjustment, but the way in which it is being achieved: after stronger measures (fuel price increases and a wealth tax) could not be implemented, these were substituted for low quality measures (a tax amnesty and a financial transactions tax) which serve at best a temporary purpose.

V. Lessons

A. Taking Stock

60. In the last ten years, Bolivia’s economy has experienced great reform successes, major failures, and bad luck. On balance, the country is better off than it was in 1994: average per capita income has grown slightly, and social indicators have significantly improved. Measured against the hopes and expectations accompanying Bolivia’s pathbreaking reforms, however, the overall outcome has been disappointing.

61. Bolivia’s disappointing performance had four main causes. First and most importantly, the structural reforms of the last ten years—including privatization, pension reform, fiscal decentralization and popular participation, and specific institutional reforms—have not fundamentally changed the nature of Bolivia’s governance problem. Patronage and clientelism continue to undermine the purposes of the state. The operating environment for business remains unfavorable, particularly for those that lack political connections, corruption is widespread, and government services are generally poor. Second, the Bolivian economy suffered a string of major adverse shocks, including the 1999 Brazil devaluation, stepped-up coca eradication, and the Argentina crisis of 2001/2002. These shocks harmed Bolivia both directly and by generating social pressures that made reforms more difficult and led to persistent increases in government spending, exacerbating Bolivia’s medium term fiscal problems. Third, critical financial and fiscal structural reforms were either never implemented or significantly delayed. In the absence of these reforms, the shocks of 1999–2002 had far more serious consequences than they otherwise would have had. Finally, financial dollarization—in essence, an economy-wide currency mismatch—was not reduced, and remains a constant source of financial vulnerability.

62. The fact that some critical reforms were either not implemented or were ultimately unsuccessful generally did not reflect a lack of ownership on the side of the authorities but political resistance to reforms, as well as the deep political and social roots of Bolivia’s governance problems. Political resistance came from two sides. First, skepticism against “neoliberal” reforms, such as privatization, on the side of the underprivileged majority, which grew increasingly impatient with the slow pace of improvement in living standards at the bottom. The main effect of this was to encourage populist measures to diffuse resistance to reforms, such as the Bonosol, and pension increases, but not to significantly slow or prevent reforms per se. 8 Second, resistance from interest groups, both within the traditional elites and the middle class. As argued in paragraph 54, this resistance grew over time as reforms with redistributive implications that would have hurt these groups moved up the policy agenda. Examples are specific financial sector reforms (which ran counter to the interests of domestic bankers), adoption of a new tax code (which ran counter to the interest of businesses paying little or no taxes), systemic tax reform (which by correcting Bolivia’s regressive tax system would have hurt the wealthy), civil service reform (which by professionalizing the state would have hurt the parties and certain groups of public employees) and education reforms which would have reduced the power of teachers’ unions.

63. Whether or not the authorities could realistically have done better in managing or resisting these pressures is impossible to say from the outside. However, it is clear that on several occasions, the price that the authorities were willing to pay in order to accommodate them was too high from the perspective of the long run interests of Bolivian society, by creating fiscal or quasifiscal burdens that sooner or later would lead to new and bigger problems, and not addressing financial vulnerabilities. Examples for this include the Bonosol, pension benefit increases under the old system, implicit guarantees and forbearance in the financial sector, and the spending increases of 2001–2002. In contrast, it is more difficult to blame the authorities for “sins of omission” such as the failure to pass systemic tax reforms or other critical reforms in times when congressional support was lacking. Here, the blame must go to the political interest groups that opposed these reforms.

64. The Fund can no doubt take credit for contributing to Bolivia’s reform successes, its stabilization and growth experience during the 1990s, and the prevention of a financial meltdown in 2003. However, it also shares some responsibility for the disappointments, along two dimensions. First, the Fund could have done better on some specific surveillance and program design issues. Second, the Fund exhibited a bias toward accommodating the political constraints faced by the authorities, and thereby contributed to the loss of reform momentum and the failure to adopt critical structural reforms.

65. The Fund’s specific successes include (1) helping to push privatization, which had the desired effects on investment and improved the quality of utilities (see appendix, Box 5); (2) insisting on central bank reform; (3) helping contain the adverse fiscal consequences of decentralization; and (4) advocating financial sector reforms which contributed to the sector’s resilience during 2002–2004, notwithstanding the fragility caused by dollarization. Areas where the Fund could have done better include (1) exercising firmer surveillance over the pension reform and more aggressively resisting policy decisions taken after the reform which resulted in rising pension costs; (2) placing more emphasis on governance and institutional reforms plans that were part of the 1994–98 program; (3) lowering the spending targets under the 2001 program as it became clear that growth would be much lower than originally expected; and (4) pushing critical financial sector reforms at an earlier stage.

66. At a more fundamental level, the Fund’s main failing in Bolivia was its willingness to accommodate the authorities’ political constraints even when important reforms were delayed or had stalled completely. The clearest manifestation of this problem was the decision to award Bolivia debt relief under the enhanced HIPC Initiative in 2001 even though critical structural performance criteria under the 2000 program had not been met. The irony of this particular situation was that Bolivia received a large transfer from the outside even while it was rejecting reforms that would have had redistributive consequences on the inside. Another example was the use of “comfort letters” based on the Fund’s informally monitored program in 2002, although this program was weak, and quickly went off track. By continuing to support Bolivia even when there was little reform progress—either directly, or through signals to the donors—the Fund became a source of moral hazard. A tougher stance would have increased the chances of reform implementation, at least over the medium run.

67. For the most part, the Fund’s accommodative bias in Bolivia was not a consequence of errors in judgment, but a reflection of deeper problems. The first is the well known “Samaritan’s Dilemma:” caring about a country’s long run interests, it makes sense to be tough, but by the same token, it is difficult to deny support when the alternative might be a crisis, an interruption of reforms, social tensions, and so on. Concerns of this kind were one of the reasons why Management and Staff decided not to postpone the request for a third annual arrangement under the PRGF in June of 2001. Second, in Bolivia, as in many other low income countries, the Fund’s dilemma was compounded by its gatekeeper role with respect to aid flows—in other words, pressure from donors and other IFIs. However, there are ways in which the Fund can make itself less sensitive to these tensions and pressures, as discussed below.

B. A Strategy for Future Fund Engagement

68. In view of Bolivia’s experience in the last ten years, any proposal for a new PRGF-supported arrangement needs to address two questions. 9 First, given that the main constraints to growth and poverty reduction in Bolivia are related to poor governance, corruption, and political and social instability, what contribution can the Fund make? Are Bolivia’s key problems not outside the Fund’s core expertise? Second, how would a program—any serious reform program—overcome the political resistance that derailed the last PRGF, and undermined negotiations for a new PRGF in 2002? The political context has, in the meantime, become more difficult, as the social situation is even more volatile, and populist pressures have increased, adding to the constraints imposed by interest groups. How can one hope to get anything done in this context, other than continuing to muddle through?

69. The first question—regarding the potential contribution of the Fund—has an easy answer. First, one of the key constraints on growth in Bolivia is its continuous state of near-crisis and vulnerability to shocks, which depresses investment. This has political and social roots, but also financial and fiscal causes, which are within the Fund’s core area of expertise. Second, fiscal-structural reforms and financial sector reforms tend to reduce rent seeking, improve accountability, and eliminate special treatment of privileged groups, all factors that are at the core of Bolivia’s growth problem. Thus, through their impact on transparency, fairness, and redistribution, these reforms have direct relevance for growth, in addition to their impact via macroeconomic stability. Third, the fact that some growth-critical reforms are outside the Fund’s core area of expertise should not prevent the Fund from taking these reforms very seriously, including through the use of conditionality, although in designing this conditionality the Fund must of course look to other institutions for help.

70. The second question—how to overcome political resistance—is harder. The basic answer is that a PRGF-supported program should in fact not go forward, unless there is social consensus in favor of a critical mass of structural reforms. However, building such a consensus may not be quite as hopeless as it seems as first. The fluid social and political situation in Bolivia—with loss of influence of the traditional parties, large parliamentary representation of the traditionally disenfranchised for the first time in history, a President and cabinet composed of political independents, and a constitutional assembly scheduled to take place soon—harbors opportunities as well as risks. The risk is that populist solutions may be embraced too easily. But there is an upside too, namely, that there may be a growing sense in Bolivia that the concerns of traditional outsiders need to be taken seriously, even if this conflicts with the narrow interests of established interest groups.

71. Thus, the critical question is how to gauge whether an adequate social consensus for a strong PRGF-supported program truly exists. The answer could be to begin by defining a critical set of reforms for removing critical growth-constraints and addressing the causes of fiscal problems and financial fragility. Next, a key reform must be identified that would test the country’s capacity to implement the program; this should be made a prior action under a new program. An obvious candidate would be an important reform that has failed in the past because of resistance from the same interest groups that could also undermine a new program—for example, the introduction of a personal income tax, critical expenditure measures, or financial sector reforms. If it is implemented, this would constitute a strong signal that the political and social consensus has indeed evolved to an extent that gives the rest of the program a chance of being implemented.

72. To define the core conditionality that is consistent with this basic strategy, the Fund will have to look to the National Dialogue, and the resulting PRSP. Since the premise of a new PRGF should be social consensus, the PRSP process—provided it is reflective of such a consensus—will be particularly important. However, the analysis of this paper suggest that the following reforms should be pillars of any future PRGF:

  • Systemic tax reform. “Tax reforms that simultaneously improve efficiency and equity—for instance, the removal of regressive exemptions or loopholes for the better off” have long been recognized as a key feature PRGF supported programs (IMF, 2000b). Bolivia is in dire need of such a reform for three reasons: Fiscal sustainability, equity—given its current regressive tax system—and ultimately growth, given the link between social equity, reforms, and growth in Latin America (Birdsall and de la Torre, 2001). The main elements of this reform should be: (i) the elimination of the current payroll tax against which VAT receipts can be credited and its replacement with a personal income tax, taking care that this does not increase the tax burden on the poor; (ii) the elimination of special tax regimes that result in unequal treatment of taxpayers with the same contributive capacity and facilitate tax evasion. A successful tax reform needs to incorporate a sustainable strengthening of tax and customs administration to combat widespread corruption and tax evasion.

  • Expenditure policy and civil service reform. As in the case of systemic tax reform, reforms in this area serve more than one purpose. First, current expenditures in Bolivia are high relative to its neighbors and other HIPC countries, and reducing them is important from a fiscal sustainability perspective. Second, a PRGF-supported program should better focus expenditures on growth and poverty reduction. Two aspects of this should be a reduction in the wage bill and pension costs, to allow for higher priority nonwage spending (for example, in the education sector). The inequality between the old and new pension regimes must be reduced by rolling back increases in benefits under the old regime. Finally, there can be no lasting reduction and reorientation of public expenditure in Bolivia without reviewing the size and composition of the civil service. At the same time, civil service reform is critical for higher growth, as poor government services, corruption, and patronage are a key growth constraint. Hence, it is imperative that the civil service be depoliticized and professionalized.

  • A transparent and incentives-friendly hydrocarbons regime. The attempt to reform the hydrocarbons regime must not throw the baby (an incentives-friendly regime) out with the bathwater (tax loopholes and poor administration). Bolivia needs a system that attracts investment, generates a rising revenue stream from current levels, and is transparent in the way in which revenues are being generated and used. Some degree of earmarking of revenues may be acceptable as long as these principles are met. However, if the present reform does not meet these principles, approval of a satisfactory hydrocarbons law should be a prior action under a PRGF-supported program, given its importance to Bolivia’s medium-term growth and development.

  • Other growth-critical institutional reforms. Based on the analyses of growth constraints surveyed in this paper, two areas stand out: labor reform and judicial reform. Labor reforms must address indirect labor costs, which are very high in Bolivia. Judicial reforms must address corruption, shorten court proceedings, and generally improve contract enforcement.

  • A strategy to encourage greater use of Bolivianos. The analysis of the causes of dollarization persistence in Bolivia (Box 4) leads to a four-pronged strategy for reducing financial dollarization: (1) fiscal adjustment to address the credibility problem; (2) the adoption of an inflation targeting regime (possibly in gradual steps), both to strengthen monetary credibility and to move away from real exchange rate targeting; (3) the development of local currency financial markets, beginning with the gradual replacement of dollar denominated government debt by medium term CPI-indexed debt; (4) regulatory changes to encourage greater Boliviano use, both to offset distortions due to implicit guarantees, and to overcome the thin market problem (Ize and Powell, 2004). The open question here is timing and sequencing: as argued by a recent TA mission, one can think of a “one go” approach or a gradualist approach. The former would have the advantage of constituting a clean break but would require strong upfront fiscal, monetary and foreign exchange market measures and entail greater risks.

  • Key financial sector reforms, including an efficient corporate and insolvency framework to strengthen bank soundness, measures to strengthen the capitalization of the banking system, and a partial deposit insurance scheme to replace the existing blanket guarantee on deposits. A potential prior action in this area could be the removal of the current prudential forbearance.

  • Last but not least, given the upcoming Constituent Assembly and growing pressures for regional autonomy, a coherent strategy to improve the budget and decentralization processes. This should include: (i) a better matching of revenue and spending responsibilities at various levels of government; and (ii) a significant improvement in the budget process, starting with the introduction of an organic budget law and eventually the enacting of a Fiscal Responsibility Act.

73. It is clear that these reforms, while critical for growth, fiscal sustainability, and crisis prevention, do not all fall within the Fund’s core areas of expertise. In those cases, the Fund will need to rely on the World Bank and the IDB for specifics, as prescribed by the Conditionality Guidelines (IMF, 2002), but it must monitor progress and must apply conditionality even to these areas. It is also clear that not all of these reforms can be undertaken at once. Systemic tax reform needs to be undertaken up front, both because it is critical from a fiscal stability, equity and efficiency perspective, and because it serves as a signal. In other areas—including expenditure policies and civil reform, financial de-dollarization, expenditure management and corporate restructuring—reforms should initiated immediately, but completing them will take time. Yet other reforms may have to wait until the second or third year of the program, given limited capacity, and given the objective of keeping the set of structural conditionality parsimonious in each given year.

74. Notwithstanding the ambitiousness of some of the reforms outlined above, it is important that a new PRGF be conservative in its growth projections. It is simply unrealistic to assume that the proposed set of reforms, even if implemented on schedule, will push Bolivia on to a higher medium term growth path in a matter of three years. The basic assumption should be that, in light of Bolivia’s growth performance since the late 1980s, trend growth will remain in the 3–4 percent range in the program period, though actual programmed growth could of course be higher depending on gas export projects coming on stream, terms of trade, and other developments at the business cycle frequency.

75. The strategy proposed sets a high bar for a new PRGF arrangement. What if the Bolivian government cannot jump over this bar in the time period that is considered reasonable for PRGF negotiations? The answer depends on the cause of the delay:

  • If the main obstacle is technical—for example, the time required to undertake critical prior actions—then the current Stand-By Arrangement could be extended for a brief period. Steps towards undertaking these prior actions could become part of the conditionality of the extended Stand-By;

  • If the authorities, while capable of implementing key upfront reforms that would be required under a PRGF arrangement, cannot commit to a three year program because of the political and constitutional uncertainty that accompanies the upcoming constitutional assembly, then a new one-year arrangement—or a one-year extension under the current Stand-By arrangement—should be considered. The conditionality under this Stand-By should be similar to that under the first year of a PRGF arrangement, except for reforms that can only be planned and implemented over a three year horizon.

  • If the reason is that the requisite social consensus for critical reforms is lacking—as evidenced, for example, by the fact that a critical prior action is blocked in congress—then an interruption of formal Fund support until such a consensus emerges may be inevitable. However, in a situation where Fund support is urgently needed to prevent an imminent financial collapse and provide a framework for continued donor support, a Stand-By that meets adequate standards, along the lines of the 2003 Arrangement, could be justified.

APPENDIX I Main Economic Developments, 1985–2004

1. Bolivia’s recent economic history began with a “big bang” (see appendix for a chronology). On August 29, 1985, the newly elected government of Victor Paz Estenssoro introduced the “New Economic Policy”, a stabilization and reform program that put an end to years of political and economic instability which had included a debt crisis, declining output, fiscal collapse, and finally hyperinflation. The central administration’s cash deficit was cut from over 20 percent to around 6.5 percent, hyperinflation disappeared, and by the Spring of 1986, inflation had stabilized at low levels. This success was all the more remarkable as the program came on the heels of six previous unsuccessful attempts to stabilize the economy within the preceding four years, was implemented without initial IFI support, met intense political resistance from the trade unions and coincided with a period of severe terms of trade shocks (Morales and Sachs, 1989). The program broke with Bolivia’s decades old tradition of “state capitalism”, eliminating all price and interest rate controls, unifying the exchange rate, liberalizing trade, allowing banks to operate in foreign currency, raising public sector prices to world levels and enacting a tax reform based on a VAT.

2. After stabilizing, Bolivia negotiated a debt restructuring with its bank creditors which reduced the face value of its $650 million commercial bank debt by $473 million—including by buying back $253 million at 11 cents on the dollar, using funds donated by foreign governments (IMF, 1995; Boughton, 2001). A second agreement covering most of the remaining commercial bank debt followed in 1992. In addition, between 1986 and 1995, Bolivia benefited from six Paris Club debt reschedulings on increasingly concessional terms, debt forgiveness by the U.S., a stock-of–debt operation with Argentina, and a concessional rescheduling by Brazil (Brooks et al, 1998). According to World Bank data, total (face value) of debt forgiveness or reduction received by Bolivia from 1986 to 1995 amounted to $1.74 billion—about 38 percent of its 1986 debt stock—though the corresponding NPV debt reduction was much higher (Daseking and Powell, 1999). As a result, Bolivia’s external debt-to-GDP ratio fell from 115 percent to 70 percent by 1995.

3. Notwithstanding occasional slippages—particularly ahead of the 1993 presidential elections—Bolivia continued on a path of slow fiscal consolidation, declining inflation, and economic recovery through the mid 1990s. The fiscal deficit (after grants) was brought down from 7 percent in the late 1980s to about 5 percent in the early 1990s and finally 2 percent in 1995. Inflation fell to around 10 percent in the early 1990s and about 5 percent in 1997. Growth remained stuck in the 2–3 percent range in the late 1980s—too low to raise living standards, given Bolivia’s high population growth—but rose to around 4 percent in the early 1990s. This was accompanied by a widening current account deficit, financed mostly by official capital flows, rapid growth in credit to the private sector, and a recovery of investment from very low levels in the mid 1980s to around 15 percent of GDP (Table 5). Structural reforms also continued, albeit at a slower pace. The banking sector underwent significant restructuring; state-owned banks were liquidated or sold, and a number of private banks were intervened and either closed or sold to foreign banks. In some areas, such as central bank reform and public administration, reforms were delayed, either in Congress or due to slow implementation.

Table 5.

Selected Economic Indicators, 1986–2004

article image

Public sector external debt reflect actual delivery of assistance under the original and enhanced HIPC initiatives (contracts for HIPC relief not yet signed are not included).

Sources: Central Bank of Bolivia; Ministry of Finance; and Fund staff estimates and projections.

4. The reform process received new impetus after the election of Gonzalo Sánchez de Lozada, who as planning minister during 1985–89 had been one of the architects of the “New Economic Policy,” to the Presidency in 1993. The new government put forward an ambitious “second generation” reform program centered on privatization, pension reform, fiscal decentralization, education reform and improvements in governance and accountability. The implementation of these reforms was judged a success; for example, the 1997 ESAF review applauded Bolivia’s record of structural reforms as among the best of all ESAF countries. Based on this strong record, in early 1998, Bolivia was one of the first countries to receive debt relief under the HIPC Initiative ($448 million in NPV terms, with Japan granting additional debt relief of $371 million in NPV terms in October), reducing its debt-to-GDP ratio to about 55 percent.

5. The reforms of the Sánchez de Lozada administration had important macroeconomic consequences. As a result of the “capitalization” program (Box 5) foreign direct investment increased tenfold, from $90 million in 1994 to about $900 million in 1997, and private investment as a share of GDP jumped from around 5 percent to 12 percent in 1997 and over 16 percent in 1998. GDP growth also rose, to 5 percent per annum in 1997 and 1998. However, the switch from a pay-as-you go system to a fully funded pension system created significant and growing fiscal costs, in the order of 2.5 percent of GDP in 1997 and about 4 percent in 1998.


“Capitalization” was a privatization scheme designed to secure a minimum level of foreign investment rather than maximizing privatization revenues. Under capitalization, private investors acquired a 50 percent stake and management control of public enterprises in return for a commitment to undertake capital expenditure at least equivalent to the enterprise’s original net worth. The remaining 50 percent of the shares were distributed equally to all Bolivian citizens above the age of 18 in 1997. They are managed by a fund which uses the dividends to partly finance an annual payment (the “Bonosol”) to all Bolivian retirees. This approach implied lower fiscal revenues from privatization than the norm. However, without this concession, the privatizations would probably have been unviable politically.

Despite a widespread perception in Bolivia that the capitalization process was mishandled, it has largely met its investment objectives, and employment losses have been relatively small. Only in the telecommunications sector did capitalization marginally fail to reach its investment target (Table 1). Services improved as the share of the population with access to electricity increased by nearly 20 percentage points from 1994–97 and telephone connections increased significantly. There were some employment losses—mainly in the hydrocarbons sector, reflecting previous overemployment and inefficiencies—but total job losses from the capitalization process are estimated at less than five thousand.

Table 1:

Investment in Capitalized Industries until 2001

(millions of US$)

article image
Source: Ministerio de Comercio Exterior e Inversión

6. Beginning in 1998, the economy was hit by four shocks: spillovers from the 1998 Russian crisis to Latin America, which led Brazil to devalue in February 1999; the acceleration of Bolivia’s coca eradication program by the new government of President Banzer (Box 6), changes in regulations and the take-over of the country’s largest bank by a Spanish bank, leading to a reduction in credit; and the Argentine crisis of late 2001, which further strained competitiveness and lowered the dollar value of remittances by Bolivian workers in Argentina. In addition, some large investment projects ended (in particular, the construction of a natural gas pipeline to Brazil). As a result of these factors, GDP growth fell to 0.4 percent in 1999; FDI flows declined by about 30 percent after peaking at about $1 billion, and non-FDI private capital flows turned negative. The deterioration of the external environment also exposed the fragility of the financial system, which had become highly dollarized since the late 1980s, as the central bank was forced to trade off the competitiveness benefits of a faster devaluation with the risk of exacerbating currency mismatches in corporate balance sheets. Non-performing loans began to climb, from about 5 percent of total loans in 1998 to 18 percent in 2003. At the same time, credit to the private sector contracted sharply, reflecting the early repayment of costly foreign credit lines by banks, a rise in risk perception, and the application of stricter loan policies.

7. With slow growth and social unrest in April and September of 2000, the government’s structural reforms program became progressively weaker. Bolivia nevertheless received a second round of debt relief under the enhanced HIPC imitative (June 2001, $854 million in NPV terms), reducing external debt from 53 percent to about 41 percent of GDP. Subsequently, the fiscal situation deteriorated sharply, with the deficit widening to 7 percent in 2001, 9 percent in 2002, and 8 percent in 2003, financed partly by rapidly growing domestic debt and partly by new multilateral lending. At end-2003, Bolivia’s external debt to GDP ratio was back to over 50 percent, and its total public debt to GDP ratio stood at 73 percent.

Coca Eradication

Coca leaves have been cultivated in Bolivia for centuries, with many traditional uses. Beginning in the mid 1970s, however, cultivation expanded rapidly as a result of the rise of the global cocaine trade, and began to have a significant impact on the economy. The recessions and public enterprise reforms of the 1980s led many former public sector workers—especially former miners—to become coca growers. The lucrative nature of coca production also encouraged corruption, crowded out traditional crops and pushed up labor costs in other parts of the economy. With virtually all transactions in US$, the growing coca trade also boosted dollarization.

In 1986, with U.S. support, efforts began to reduce coca cultivation and cut drug production, but the net impact was small until coca eradication was sharply boosted during the Banzer-Quiroga administration (1997–2002). According to U.S. State Department data, net cultivation of coca crops fell by two thirds between 1997 and 2000. However, during 2001–2003, replanting has exceeded destruction under the eradication program, and net cultivation has recovered to about 60 percent of its peak in the mid-1990s.

Despite efforts to encourage crop substitution, coca eradication has resulted in a large negative income shock to the economy. Coca trade as a percentage of GDP fell from a peak of 5.6 percent in 1988 to about 0.7 in 2003, which significant effects on economic activity, particularly in the informal sector.

Coca cultivation and eradication

Coca cultivation and eradication (ha)

article image
Source: US State Department, International Narcotics Control Strategy Report.

The reported leaf-to–HCl conversion ratio is estimated to be 370 kilograms of leaf to one kilograms of cocaine HCl in the Chapare. In the Yungas, the reported ratio is 315:1.

8. The combination of economic stagnation, social unrest, fiscal deterioration and financial sector fragility brought Bolivia to the brink of a major political and financial crisis. Election-related uncertainty led to a run on deposits in mid 2002 (Figure 3). Deposits recovered after former President Sánchez de Lozada was elected President, having won a slim margin over indigenous leader Evo Morales. But violent social unrest erupted in February and October 2003, the former triggered by police protests against budget and taxation proposals, and the latter by opposition against a project for natural gas exports through Chile. Both episodes led to severe runs on deposits. The October protests led to 70 deaths and the resignation of the President. Vice-president Carlos Mesa, an independent, assumed the presidency (until 2007) on a platform of social dialogue—including a referendum on hydrocarbon policies and gas exports—and a call for a constitutional assembly in 2005 to modify the existing constitution. The July 18, 2004 gas referendum showed large support for a revision of the 1996 hydrocarbon law in the direction of a stronger role for the state, while also indicating supporting for the export of natural gas, albeit by a much slimmer margin.

Figure 3.
Figure 3.

Bolivia: Selected Financial Indicators

Citation: IMF Staff Country Reports 2005, 139; 10.5089/9781451970500.002.A001

Source: Bolivian authorities; International Monetary Fund, INS, IFS, WEO; and IMF staff estimates.1/ Weights based on trade with ten countries, excluding trade related to natural gas, in 1996–97.
Table 6.

Chronology of Main Economic and Political Events

article image
article image
Sources: de Mesa et el. (2003), IMF and World Bank reports, Country reports of the Economist Intelligence Unit.
Table 7.

Bolivia: History of IMF Arrangements, 1986–2004

article image
Source: Staff Reports, Finance Department.
Table 8.

Bolivia: Performance Criteria, 1994-2004

article image