Report No. 27629-BA Brazil Country Assistance Evaluation January 14, 2004 Country Evaluation and Regional Relations Operations Evaluation Department Document of the World Bank Abbreviations and Acronyms AAA Analytical and advisory services AIDS Acquired Immune Deficiency Syndrome BNDES Banco Nacional de Desenvolvimento Economico e Social CAE Country Assistance Evaluation CAS Country Assistance Strategy CMU Country Management Unit ECLAC Economic Commission for Latin America ESW Economic and sector work FDI Foreign direct investment GDP Gross domestic product GEF Global Environmental Facility HIV Human immunodeficiency virus IBGE Instituto Brasileiro de Geografia e Estatistica IBRD International Bank for Reconstruction and Development IDB Inter-American Development Bank IFC International Finance Corporation IFIS International financial institutions IMF International Monetary Fund INSS Instituto Nacional de Seguridade Social IPEA Instituto de Pesquisa Economica Aplicada LAC Latin America LCR Latin America and the Caribbean MDGs Millennium Development Goals MIGA Multilateral Investment Guarantee Agency NGO Non Governmental Organization OECD Organization for Economic Cooperation and Development OED Operations Evaluation Department OEG Operations Evaluation Group OEU Operations Evaluation Unit PSD Private Sector Development QAG Quality Assurance Group SAL Structural Adjustment Loan SECAL Sector Adjustment Loan SMES Small and medium-size enterprises swc State Water Companies TA Technical assistance ~~ Director-General, Operations Evaluation : Mr. Gregory K. Ingram Director, OED Mr. Ajay Chhibber Senior Manager, OEDCR Mr. R. Kyle Peters Task Manager Mr. Roberto Rocha Peer Reviewer Mr. Ren6 Vandendries Peer Reviewer Mr. John Johnson External Peer Reviewer Ms. Eliana Cardoso, Georgetown University Contents Summary .......................................................................................................................................... i . 1 Purpose and Background ............................................................................................................ Long Period o f Growth Followed by the “Lost Decade” .............................................................. 1 Stabilization and Structural Reforms in the 1990s........................................................................ 2 2 . World Bank Assistance to Brazil in the 1990-2002 Period ..................................................... 4 World Bank Objectives and Strategiesfor the Decade ................................................................. 4 The First Phase (I990-1994): Contraction o f Lending and Shift to the Social Sectors and the Environment.......................................................................................... 6 The Second Phase (1995-98): Recovery o f Lending. Consolidation o f the Strategy. Initial Support to Structural Reforms ......................................................................... 7 The Third Phase (1999-2002): Further Expansion o f Lending and Shift to Adjustment Lending in Support to Stabilization and Structural Reforms ................................. 9 %4 3 . Assessment of Bank Products and Services ............................................................................ 11 Assessment o f Analytical and Advisory Services ......................................................................... 11 Assessment o f Lending Services .................................................................................................. 12 Factors that Have Contributed to the Improved Project Pe~ormance ....................................... 14 Assessment o f the Bank’s Role in Aid Coordination ................................................................... 15 4 . Assessing the Development Impact of the Bank’s Assistance ............................................... 16 4.1 Assessing the Relevance and Consistency of the Bank’s Strategy ........................... 16 4.2 Assessing the Contribution of the Assistance Program to Poverty Alleviation ......18 The Program ‘S Contribution to the Decline in Income Poverty Ratios ........................ 18 The Program’s Contribution to Improved Education and Health Outcomes ................21 The Program’s Contribution to Expanding Access o f the Poor to Basic Services ........24 4.3 Assessing the Contribution of the Bank’s Program to Sustainable Growth...........25 The Program’s Contribution to Brazil’s Growth Performance in the 1990s................. 25 The Program’s Contribution to Improvements in the Environment............................... 28 4.4 Assessing the Overall Impact ofthe Assistance Program......................................... 30 4.5 Assessing the Sustainability of the Positive Outcomes.............................................. 32 4.6 Assessing the Bank’s Performance ............................................................................. 34 4.7 The Counterfactual and the Bank’s Relevance ......................................................... 36 5. Lessons and Recommendations................................................................................................ 37 This evaluation was prepared by Roberto Rocha (Task Manager. OEDCR). with assistance from John Heath (OEDST). Andre Averbug (OEDCR). Dariush Akhavan. Kaizo Iwakami Beltrgo (PEA and IBGE). Regis Bonelli (PEA). Claudio Moura Castro (Pitagoras University). Peter Knight. Rosane MendonGa (IPEA and UFF). Ronald0 Seroa da Motta (PEA). Armando Castelar Pinheiro (BNDES). Eustaquio Reis (IPEA). Albert0 Mello e Sousa (Umu). and William Tyler (UERJ). The eva1uatio.n also benefited from project audits prepared by Helen Abadzi and Byniam Reja (OEDST). Contents (cont.) Text Boxes B o x 1: Examples o f Outstanding ESW ...................................................................................... 12 Text Tables Table 1: Selected Economic Indicators. Annual Averages. 1950-2002 ......................................... 2 Table 2: Size o f World Bank Lending. 1980-2001 ........................................................................ 4 Table 3: Sectoral Composition o f Bank Loans. 1980-2002 ........................................................... 6 Table 4: Sectoral Composition o f Bank Loans. 1990-2002 ........................................................... 8 Table 5: Q A G Evaluation o f ESW - Share o f Satisfactory ESW (in %). 1996-2002 ..................11 Table 6: OED Evaluation o f Closed Projects - Share o f Projects with Satisfactory Outcomes (in %). 1980-2002 ................................................................................... 13 Table 7: Q A G Evaluation o f Ongoing Projects - Share o f Satisfactory Projects (in %). 1996-2002 ...................................................................................... 13 Table 8: Average Lending Costs. per Project and per Commitment. 1990-2002 ......................... 15 Table 9: Poverty and Social Indicators. 1990-2001 ..................................................................... 18 Table 10: GDP Growth Rates in Brazil and Selected Comparators. 1990-2002 ............................ 25 Table 11: Environmental Indicators. 1980s and 1990s .................................................................. 28 Table 12: Brazil’s Progress in Meeting the Millennium Development Goals................................ 32 Text Figures Figure 1: World Bank Lending. 1980-200 1..................................................................................... 7 Figure 2: Q A G Evaluation o f Projects at Risk. 1990-2002 ........................................................... 13 Figure 3: Poverty Ratios and Income Growth by Deciles ............................................................. 20 Figure 4: Selected Education and Health Indicators ...................................................................... 22 Figure 5: GDP Growth. 10 Year Moving Averages (“A p.a.) and Real Fixed Investment (% o f GDP). 1965-2001 ........................................................................ 26 Annexes Annex 1 Statistical Tables ............................................................................................................ 41 Annex 2 Summaries o f Sector Evaluations .................................................................................. 65 Annex 3 An Evaluation o f MIGA Activities in Brazil, 1994-2002 .............................................. 83 Annex 4 Government’s Comments on the C A E ........................................................................... 87 Annex 5 Management Action Record........................................................................................... 89 Annex 6 Chairman’s Summary : Committee on Development Effectiveness (CODE) .............. 91 References ................................................................................................................................. 95 i Summary 1. Brazil entered the 1990s suffering the consequences of a “lost decade” of high inflation and slow growth. Between 1980 and 1990, per capita income had declined in real terms, and the share of the population in extreme poverty had risen from 16.5 to 19 percent-and from 36 to 42 percent in the Northeast. Income distribution had worsened. Key social indicators had improved little, particularly in the Northeast. These adverse conditions persisted in the early 1990s due to the Government’s failure to stabilize the economy and articulate a development agenda. Conditions improved only with the success of the 1994 Real Plan. Stabilization contributed to a decline in poverty and enabled the Government to articulate a development agenda with clear goals for education and other sectors during the remainder o f the 1990s. Key aspects of the Bank’s program The alleviation o f poverty was the central objective of the Bank’s assistance to Brazil in the 1990s and early 2000s. Growth was an intermediate objective and means to achieve the final objective o f poverty alleviation. The Bank’s strategy was developed in a period when the institution became less important as a source offinance to Brazil. Bank disbursements accounted for 16.5 percent of total financing to Brazil in the 1980s but only 6.5 percent in the 1990s. The strategy became more selective,focusing on the poor Northeast and on activities expected to address directly the roots o f poverty. Priority was given to education and health (their share in total commitments increased from 5 percent in the 1980s to around 20 percent in 1990-2002) and to the provision of basic services and assets to the poor (water, sanitation, rural electricity, land). The growth element o f the strategy supported the Govemment’s decentralization and privatization o f infrastructure through technical assistance in regulation and selective lending. These actions were expected to promote growth b y stimulating private investment in infrastructure and generating efficiency gains. Assistance to the environment expanded significantly in the 1990s and focused on green issues. At the end o f the 1990s the Bank shifted to adjustment lending to assist the Govemment in its stabilization and reform eflorts. The share o f adjustment lending in total lending increased from 6 percent in the 1980s to 26 percent in 1990-2002 and addressed fiscal and financial reform and social protection. Key program results 2. The Bank’s assistance strategy in 1990-2002 was relevant overall. The Bank’s strategy was underpinned b y high quality analytical work on poverty and growth, done with substantial participation by top Brazilian researchers. The shift to the social sectors and the focus on the Northeast were justified-given the very weak social indicators in the Northeast region-as was the shift to adjustment lending at the end o f the decade. A scenario o f crisis could have ignited inflation with negative consequences for the poor. When the Bank made adjustment and technical assistance loans, i t maintained assistance to the social sectors. .. 11 3. The main component o f the program, comprising roughly half o f total lending, addressed the roots o f poverty through human resource development and access to basic services. This component produced satisfactory results. Education gains were impressive. The net enrollment rate in primary education increased from 84 to 96 percent nationally, and from 72 to 93 percent in the Northeast. The youth illiteracy rate declined from 9.8 to 4.2 percent nationally, and from 22.7 to 9.6 percent in the Northeast. The Bank’s program contributed to these outcomes by providing infrastructure and education materials, strengthening state secretariats of education (so they could manage higher student enrollments), and implementing information and evaluation systems for primary education and minimum operational standards in all schools. The decline in infant mortality was also pronounced, particularly in the Northeast-from 48 to 29 per 1,000 live births nationally, and from 73 to 44 per 1,000 live births in the Northeast. Other key health indicators such as life expectancy and communicable disease incidence/mortality improved significantly. The Bank’s program contributed to these outcomes by combating communicable diseases and providing health infrastructure. Finally, the program helped expand basic infrastructure to the poor through i t s projects in rural poverty alleviation, water, sanitation, and urban transport. 4. A second component o f the program-designed to stimulate investment, economic eficiency, and growth-produced mixed results. The average growth rate between 1990 and 2002 (2.5 percent per year) was higher than in the 1980s (1.5 percent per year) and surpassed the Latin American average in recent years but remained disappointing. The Bank’s program strengthened the capacity o f Brazilian states to manage their own infrastructure, increased private sector participation in infrastructure, and built regulatory capacity in energy and water. The program generated efficiency gains, but did not succeed in removing critical bottlenecks that continue to constrain public and private investment and to affect economic performance. Regulatory deficiencies remain. Key reforms supported through adjustment lending (social security, financial sector) started late in the decade and have not been completed. Pension expenditures and deficits remain large (12 and 5 percent o f GDP, respectively), burdening public finances and depressing savings and investment. Interest spreads remain high and access to credit curtailed due to weak creditor rights and heavy taxes on financial intermediation. These reforms w i l l need to be concluded to improve growth performance. 5. A third component o f the program, involving support to the environment, produced satisfactory results. The Bank played a critical role in building environmental institutional capacity and in helping the Brazilian government address many important challenges. Brazil’s environment indicators s t i l l lag those of developed countries but improved during the decade. Average annual rates o f deforestation are s t i l l high but declined to 16,800 km2in the 199Os, down from 20,300 km2in the 1980s; the size of protected areas increased from 2.4 to around 6.5 percent o f the national territory; and pollution in large cities such as STio Paul0 declined. The average emissions o f carbon dioxide rose from 1.4 to 1.6 metric tons per capita between the two decades (as in most other regions), but this also happened in other regions and average emissions in Brazil remained substantially below the corresponding averages for Latin America (2.4), North America (19.2), and the OECD (12.1). 6. Adjustment loans had variable impacts, The fiscal reform loans achieved their main objectives-the reform o f the public sector and the imposition o f hard budget constraints at all levels of government. The social protection loan also achieved i t s objective o f preserving critical social programs during fiscal adjustment. The social security, financial sector, and iii energy loans supported important reform steps, including: changes in the pension benefit formula for private sector workers, restructuring and privatizing numerous insolvent state banks, and reforming energy tariffs. However, these reforms started late in the decade and have not gone far enough. Limited progress in reforming the pension system (especially for public sector workers) compromised the country’s performance by constraining the availability o f resources for investment and better targeted social programs. 7. The Bank’s assistance program to Brazil produced a suflcient mass o f positive outcomes to be rated as satisfactory. In the 1990s extreme poverty rates declined from 19 to 14.5 percent in the whole country and from 42 to 30 percent in the Northeast. The program contributed only modestly to the decline in income poverty rates-the gains resulted primarily from eliminating high inflation (which affected disproportionately the poor)-and the program contributed only indirectly by helping the Government sustain stabilization at the end o f the decade. However, the program contributed importantly to the impressive improvement in social indicators and to the expanded access of the rural and urban poor to basic infrastructure. All are key pre-conditions for further reductions in poverty and inequality. The education gains are likely to translate into reduced poverty and inequality because incomes in Brazil are strongly related to education levels. Expanded access of the poor to infrastructure both alleviates poverty directly and creates the conditions for subsequent growth in household incomes. 8. Goingforward, these gains will need to be enhancedfor Brazil to achieve further reductions in poverty and inequality. In the early 2000s the country faced volatile capital markets, a difficult external environment, and lingering fiscal vulnerabilities (a large share o f i t s public debt i s indexed to the exchange rate and short-term interest rates). Currently (May, 2003), market perceptions about Brazil are improving due to renewed efforts to complete key reforms and improve the fiscal situation, but vulnerability to market volatility may remain for some time. Policies must maintain fiscal discipline to ensure access to external finance, keep inflation low, and prevent an erosion of the gains o f the 1990s. The Government and the Bank must also continue addressing the roots o f poverty. Further reductions in poverty and inequality w i l l require improved growth performance and better targeting o f social expenditures. In education, increasing years o f schooling and improving education quality are s t i l l needed to obtain substantive results. Access of the poor to basic services also needs to be expanded. Completing tax, social security, and financial sector reforms i s critical. The Bank can continue to play a relevant role in Brazil b y helping the Government to increase gains in education and health and to remove the constraints to better growth performance. Ajay Chhibber Acting Director-General Operations Evaluation 1 1. Purpose and Background 1.1 This report evaluates the World Bank assistance program to Brazil during the 1990-2002 period, covering the administrations o f Presidents Fernando Collor de Mello/Itamar Franco (1990-94) and Fernando Henrique Cardoso (1995-02). The Country Assistance Evaluation (CAE) examines three main and inter-related questions. First, whether the objectives o f Bank assistance were relevant, given Brazil’s development needs and challenges in this period. Second, whether the Bank’s assistance program was effectively designed and consistent with i t s objectives. Third, whether the Bank’s program achieved i t s objectives and had a substantive impact on the country’s development during this period. Examining these three main questions also allows the C A E to assess whether the Bank can s t i l l play a relevant role in a large middle-income country such as Brazil, and to draw lessons and recommendations for future Bank assistance to the country. 1.2 The Bank’s program o f assistance to Brazil can only be evaluated with an understanding o f Brazil’s development challenges in the 1990s. The country entered the 1990s after a decade o f poor economic performance, characterized b y high inflation and low growth. Brazilian society had succeeded in restoring democracy, but had also become frustrated with the many failed stabilization attempts and the stagnation o f economic activity, after previous decades of high growth. The lack o f growth was particularly problematic, given Brazil’s high social inequalities and levels o f poverty, as i t tended to produce social tensions and unrest.’ Long Period of Growth Followed by the “Lost Decade” 1.3 During the 1950-1973 period Brazil had one o f the best growth performances among developing countries, as indicated by growth rates o f 7.5 percent per year (Table 1). Brazil was able to maintain high growth rates in the remainder o f the 1970s (after the 1974 o i l shock) at the cost o f large current account deficits and a rapid build up o f external debt, but debt ratios were considered acceptable and did not raise major concerns. Inflation was already high, but i t s effects seemed to have been largely neutralized b y extensive indexation arrangements. Income inequality was one o f the highest in the world, but all income classes benefited from the high growth rates, including the poorest segments o f the population. All in all, growth seemed to be sustainable, and the country seemed poised to become a success story. 1.4 The second o i l shock and the developing country debt crisis in the 1980s brought an abrupt end to this growth performance, and exposed the fragility o f Brazil’s economic strategy. First, the country’s inability to adjust internally to the lack o f foreign finance resulted ultimately in the acceleration o f inflation and a decline in investment ratios (Table 1). Infrastructure, then almost exclusively a responsibility o f the State, was especially affected. Second, i t became apparent that the country had failed to build i t s human capital to the levels required to maintain high growth rates. Finally, the economy was plagued by severe inefficiencies in resource allocation caused by trade and exchange restrictions, distorting taxes, and the pervasive presence o f the State in virtually all sectors o f the A historical review of Brazil’s development performance i s provided in Gordon (2001) and Pinheiro et a1 (2001). 2 economy. The poor segments o f the population were particularly penalized b y the stagnation o f economic activity and high inflation. 1.5 The 1980s were subsequently labeled as the “lost decade” for Brazil and most o f Latin America. In the case o f Brazil, the lost decade extended until the mid-l990s, due to the failure to stabilize the economy and implement a coherent program o f structural reforms. The administration o f President Collor initiated a number o f reforms in 1990, including the reduction o f trade restrictions and the privatization of some State enterprises. However, the failure to stabilize the economy and the emergence of corruption scandals reduced the reform impetus and caused the impeachment o f President Collor b y Congress in 1992. Vice- President Itamar Franco assumed the presidency for the remainder o f Mr. Collor’s term. Inflation remained unabated during most o f his administration, reaching 5,000 percent per year in June 1994. Table 1: Selected Economic Indicators, Annual Averages, 1950-2002 CPI Injlatwn GDP growth Real Fixed Current Current FDI External Debt Period (Average) (W Investment, Account Account (% of GDP) (% o f GDP) (%) 1980 Prices (US$ millwn) (% of GDP) (% of GDP) 1950-1973 29.3 7.5 19.1 -354 -1.3 0.4 17.6 1974-1980 41.6 6.8 24.0 -7,745 -4.5 0.8 21.0 1981-1994 507.2 1.9 16.8 -2,716 -1.1 0.6 21.3 1995-1999 16.6 2.3 17.0 -26,218 -3.7 2.7 42.5 2000 6.6 4.5 16.0 -24,225 -4.0 5.5 36.5 2001 5.7 1.5 16.5 -23,214 -4.6 4.5 41.7 2002 8.3 1.5 n.a. -7,696 -1.8 3.5 45.0 Notes: All variables are period averages except for extemal debt; 2002 figures are estimates. Stabilization and Structural Reforms in the 1990s 1.6 Success at stabilization was only achieved with the implementation o f the Real plan in June 1994. The success o f the Real plan in reducing inflation was initially threatened b y growing fiscal deficits in the first Cardoso administration (1995-1998), as they contributed to a real exchange rate appreciation, an expansion o f current account deficits, and a substantial increase in the public and external debts. However, during the second Cardoso administration (1999-2002) there was a substantial fiscal adjustment at all levels o f Government, as indicated by the shift in the primary balance (excluding interests), from deficits to surpluses of more than 3.5 percent o f GDP. The adoption o f tight fiscal policies for four consecutive years was instrumental for maintaining control of inflation, despite a large currency devaluation (following the move from a managed peg to a floating rate regime in 1999), and for reducing the current account deficit (Table 1 and Annex Table 9).* The professional literature analyzing fiscal and exchange rate policies in this period criticize the first Cardoso administration for not having contained the fiscal imbalances and for maintaining the managed peg regime for too long. At the same time, the literature also acknowledges that fiscal adjustment was complicated by several factors, including the rigidities introduced by the 1988 Constitution, the emergence of large hidden liabilities (e.g., past losses of public banks), and the imbalances at the level of states and municipalities. Tyler (1999) and Cardoso (1996 and 2000) stress the inconsistencies of the fiscal and exchange regimes in 1995-1998, and Giambiagi (2002) provides a comprehensive analysis of fiscal policy in the 1990s and early 2000s. 3 1.7 The Cardoso Government also implemented an ambitious program of reforms in other areas, including the abolition of State monopolies, the privatization o f State enterprises, the restructuring and privatization o f State banks, and land reform. Finally, there was also a more focused effort to improve education and health standards, as shown b y the impressive improvement o f most social indicators. For the sake o f illustration, the net enrollment rate in primary education increased from 84 to 96 percent during the 199Os, and infant mortality declined from 48 to 29 per 1,000 live births in the same period (Annex Table 10). 1.8 B y the year 2000 Brazil seemed to be harvesting the first fruits o f i t s stabilization and reform efforts. As shown in Table 1 , inflation had been brought under control despite a large devaluation o f the Real, and GDP growth had increased to 4.5 percent. The current account deficit had declined from a peak of US$33 billion to US$23 billion, and was fully covered b y a large volume o f foreign direct investment (FDI). Real fixed investment was s t i l l low by historical standards, but the large FDI flows enhanced the prospects o f large increases in productivity. The increase in external debt seemed to have been halted. 1.9 These positive trends were disrupted in 2001 and 2002, however, due to several adverse events. These included an international slowdown, a domestic energy crisis, spillovers from the Argentine crisis, and market fears that the 2002 elections would bring a new Government not committed to fiscal discipline and economic reforms. Market expectations about Brazil’s economic performance worsened considerably, resulting in a strong devaluation o f the currency and a sharp rise in international borrowing spreads, from 500 basis points in 2000 to 2,700 basis points in late 2002. The strong pressures on the exchange rate and interest rates in turn had adverse consequences for fiscal policy, as a significant part of public debt i s indexed to the U S dollar or at floating rates-during 2002 public debt increased from 50 to 63 percent o f GDP due to these effects. The current account deficit declined further but the growth recovery lost momentum with the energy rationing, the tightening o f macro policies, the restricted access to credit, and the fall o f confidence. 1.10 In January 2003 Mr. Luiz Inacio Lula da Silva became president, and announced his commitment to fiscal discipline and the targets that had been agreed with the IMF in September 2002, under a US$30 billion standby supporting Brazil through December 2003. This announcement had a positive impact on financial markets, but w i l l need to be followed by the implementation o f coherent policies in order to consolidate the hard-won stabilization gains and restore confidence further. Mr. Lula da Silva also faces a challenging development agenda, despite the positive achievements in the 1990s. Brazil’s per capita GDP i s around US$7,500 on a PPP basis and i t s social indicators have improved significantly, but its distribution o f income i s s t i l l very unequal and 34 percent o f the population i s s t i l l below the poverty line. T o restore growth and reduce poverty further, Mr. Lula da Silva w i l l need to release fiscal resources for investment and well-focused social programs, and remove the remaining structural bottlenecks that also hinder growth. Success in these areas w i l l depend, inter alia, on the implementation o f well-designed social security, tax, financial, and judicial reformse3 Brazil’s unfinisheddevelopment agenda i s examined in Scheinkman et al. (2002). 4 2. World Bank Assistance to Brazil in the 1990-2002 Period World Bank Objectives and Strategiesfor the Decade 2.1 The alleviation of poverty was the central objective of Bank assistance to Brazil throughout the 1990s and early 2000s. This central objective was stated in the 1993 assistance strategy, and reiterated in the 1995, 1997 and 2000 assistance strategies. Growth was usually defined as an intermediate objective, or as a means to achieve the final objective o f poverty alleviation. The strategy that was developed in this period called for elements expected to have a direct impact on poverty alleviation and elements designed to foster growth, but with an emphasis on the first set of elements. 2.2 The Bank’s strategy was developed in a period when the institution was becoming less important as a source o f finance to Brazil. As shown in Table 2, average Bank disbursements to Brazil increased slightly in nominal terms but declined sharply in relative terms between the two decades. Bank disbursements accounted for 16.5 percent of total debt financing flows to Brazil in the 1980s but only 6.5 percent in the 1990s. B y the end o f the decade Bank disbursements had declined to less than 4 percent o f total financing flows (4.5 percent including IFC-see Annex Table 1). The institution had also become smaller relative to other development institutions, such as the Inter-American Development Bank (IDB) and the National Development Bank (BNDES). 2.3 I t became apparent that the strategy had to reflect the diminishing importance o f the Bank as a financing source. In particular, the Country Assistance Strategies (CASs) indicated that the Bank had to become more selective and concentrate its resources in the sectors and regions where its poverty alleviation impact would be strongest. The strategy that was developed called for a switch o f assistance to the social sectors and other sectors expected to have a more direct impact on poverty alleviation. The strategy also called for a concentration o f efforts in the poorest regions of the country, primarily the Northeast. In assisting the poor regions, the Bank perceived the need to deal directly with Brazilian states and municipalities, given their responsibility for the provision o f public services mandated by the 1988 Constitution. Table 2: Size of World Bank Lending, 1980-2001 1980-1989 1990-2001 1990-1994 1995-1998 1999-2001 Annual IBRD Disbursements(US$ million) 894 1,100 663 1,254 1,621 % of total debt flows to Brazil 16.5 6.4 10.2 3.6 3.9 % of IDB disbursements 279.1 107.2 71.8 64.8 280.7 % of BNDES disbursements 24.2 15.8 20.2 10.0 16.4 IBRD Debt Outstanding, Disbursed(US$ million) 8,311 7,963 6,310 6,298 7,963 % of total Brazil debt stock outstanding 7.3 3.4 4.1 2.6 3.4 % of total IBRD stock of disbursed loans 9.8 6.1 5.7 5.4 6.1 Sources: IBRD, IDB, BNDES. 2.4 Education and health were elevated as priority sectors (especially the former), due to their perceived capacity to contribute to poverty alleviation and growth at the same time. Assistance to education would focus on early childhood development and basic education, whereas assistance to health would focus on the fight against transmittable diseases and improving the access o f the poor to health services. Assistance to rural development would 5 continue, but these projects would be restructured to sharpen their focus on rural poverty alleviation. 2.5 Assistance to infrastructure would also continue, but with some changes. The Bank would support the Government’s decentralization and privatization programs, assisting the Brazilian states and municipalities and encouraging private sector participation in infrastructure, in order to help Brazil meet i t s large infrastructure needs. Lending would focus on sectors more difficult to privatize, or needing an “honest broker,” or with a more direct impact on poverty alleviation (e.g. roads, urban rail, water and sanitation). Finally, assistance to the environment was rationalized on the grounds that it would contribute to sustain growth as well as to reduce poverty (as the poor were disproportionately affected by environmental degradation). 2.6 The decision to switch to the social sectors and the environment implied the need to reduce assistance to other sectors. The Bank scaled down traditional agriculture projects (irrigation, agriculture extension) due to unsatisfactory outcomes in the 1980s. Energy projects were also scaled down due to unsatisfactory outcomes in the 1980s and early 199Os, ~ CASs do not mention, however, caused primarily b y the Government’s tariff p ~ l i c i e s .The that the Government also had reservations on Bank assistance to energy due to the Bank’s safeguards, particularly the Bank’s resettlement policies, after a traumatic and costly experience with resettlement in the previous d e ~ a d e . ~ 2.7 The sectoral reallocation o f lending in the 1990s was generally consistent with the proposed strategy. As shown in Table 3, the share o f education, health, and the environment in the total portfolio increased substantially -from around 6 percent in the 1980s to about 25 percent in the 1990-2002 period. The share o f rural development/poverty remained roughly at the same levels. The share o f transportation increased due to the increase in lending to urban transportation, a sub-sector with a stronger direct impact on the poor. The share o f water and sanitation declined, although not b y a wide margin. The shares of traditional agriculture projects and energy were substantially reduced. If the analysis i s restricted to investment lending, the gains in the social sectors and the environment are even larger -their combined shares increase from around 6 to 34 percent o f total investment loans (Annex Tables 2 A and 2B provide more detailed information). 2.8 At the same time, Table 3 also reveals the emergence o f Bank assistance in areas such as public sector management and social protection. I t i s also evident that assistance in these two areas was accompanied by the emergence o f adjustment lending, which increased from 6.8 percent o f the portfolio in the 1980-1989 to 26.4 percent in 1990-2002. The shift to adjustment lending in the 1990s reflects critical developments in international capital markets and in Brazil, and an evolution o f the Bank’s strategy in the second half o f the decade. In order to appreciate the evolution o f the Bank’s assistance, i t i s useful to define three different phases: the 1990-1994 period o f contraction cum sectoral shifts; the 1995-1998 period o f recovery of lending, consolidation o f the strategy and initial support to stabilization and The completion report of a power distribution loan closed in 1992 includes an overview of the 1980s and points out the devastating effect of tariff policies on sector finances. I n the same year the Bank cancelled a power transmission loan, due to the Government’s unwillingness to adjust tariffs. In 1992, OED reviewed the environmental impact of Bank projects in Brazil, and noted that the final cost of rural resettlement supported by the Itaparica project had exceeded US$63,000 per family. 6 structural reforms; and the 1999-2002 period of shift to adjustment lending in support to stabilization and structural reforms. Table 3: Sectoral Compositionof Bank Loans, 1980-2002 (in percent of the value of commitments) Sectors 1980-1989 1990-2002 All Loans All Loans Investment Loans Agriculture 32.5 13.4 18.1 Rural Developmenfloverty 11.4 10.1 13.6 Other 21.1 3.3 4.5 Education 2.0 11.7 15.9 Environment 0.9 6.2 8.5 Finance and PSD 12.7 8.9 4.9 Health 2.8 6.9 9.4 Infrastructure 43.4 28.7 34.9 Transportation 13.0 16.5 22.4 Urban 3.8 7.4 10.0 Other 9.2 9.2 12.4 Energy 21.0 5.8 3.8 Water and Sanitation 9.5 6.4 8.6 Public Sector Management 0.2 11.0 3.8 Social Protection 0.0 9.9 0.1 Urban Development 5.4 3.2 4.3 TOTAL 100.0 100.0 100.0 Share of Adjustment Loans 6.8 26.4 The First Phase (1990-1994): Contraction of Lending and Shift to the Social Sectors and the Environment 2.9 The early 1990s were a turbulent period in Brazil, as noted in Section 1 . During this period Bank commitments and disbursements declined significantly and net transfers became large and negative, consolidating a trend initiated in the late 1980s (Figure 1). The sharp decline in the volume o f lending reflected the Government’s inability to provide counterpart funds, as well as the Bank’s hesitation to initiate new lending in view of the portfolio deterioration (the share of projects at risk had increased to 85 percent), in a scenario of extreme macroeconomic disequilibria and political uncertainties. 2.10 The 1993 country strategy, written in the middle o f this turbulent period, proposed a focus on poverty reduction and a sectoral reallocation o f lending to education and other sectors more consistent with this objective. The 1993 document had the merit o f outlining an assistance strategy that would be followed during most o f the decade, but was also an incomplete document. I t lacked a satisfactory discussion o f the Government’s program, did not assess the Government’s capacity to absorb the assistance, and failed to address many important details. For example, it perceived the need to encourage private sector participation in infrastructure but did not elaborate a clear assistance strategy in this area. 2.1 1 During this phase, the combined share of education, health, and the environment rose sharply to 43 percent o f new commitments (Table 4). The education portfolio comprised five projects dealing with basic education, whereas the new health portfolio included an innovative project to control AIDS and a project to strengthen health services in low income 7 Figure 1: World Bank Lending, 1980-2001 (in US$ millions) 2,000 1,500 1,000 500 0 -500 - 1,000 I 1980 1983 1986 1989 1992 1995 1998 2001 --t Commitments -8- Gross Disbursements +Net Disbursements areas o f the Northeast. The Bank's green environmental agenda was initiated in this period with three green projects (the Bank had two projects dealing with industrial pollution in the 1980s). The share o f infrastructure remained the same as in the 1980s, but with a strong reallocation o f resources to water and transportation and with a focus on decentralization. 2.12 Although the sectoral shifts were consistent with the Bank's strategy, most o f the new projects started facing the same difficulties experienced b y the previous ones: lack o f counterpart funds, poor institutional capacity, and slow implementation. Several projects that faced implementation difficulties had to be restructured b y the following administration. Some o f the main lessons that were extracted from this period were the need to ensure Government ownership o f the assistance program, to make a candid assessment o f the Government's implementation capacity, and to limit the set o f objectives to those that could be realistically reached. The Second Phase (1995-1998): Recovery of Lending, Consolidation of the Strategy, Initial Support to Structural Reforms 2.13 The conditions for Bank assistance improved dramatically after 1994. The Government's success in stabilizing the economy and elaborating a development agenda led to a sharp improvement in the quality o f the portfolio and opened the way for a resumption of lending. The quality o f the policy dialogue improved, especially after the decentralization o f the Country Management Unit to Brasilia in 1997.6 The Government became progressively more engaged in the elaboration o f the assistance strategy. The assistance strategies in 1995 and 1997 reaffirmed the focus on poverty alleviation with growth as a second objective, but elaborated the assistance strategy with more detail and coherence, benefiting from a poverty assessment and a private sector development (PSD) strategy. The World Bank country strategy in 1997 was the first one for Brazil jointly elaborated b y IBRD and IFC. The decentralization i s examined further in Section 3. 8 Table 4: Sectoral Composition of Bank Loans, 1990-2002 (in percent of the value of commitments) Sectors 1990-1994 1995-1998 1999-2002 Agriculture 6.1 26.3 10.3 Rural Poverty 0.7 24.7 7.4 Other 5.4 1.7 2.8 Education 22.1 5.2 8.2 Environment 11.2 7.1 1.9 Finance and PSD 6.3 4.5 13.8 Health 8.9 7.1 5.3 Infrastructure 40.2 34.0 16.4 Transportation 18.2 27.3 8.0 Urban 5.3 10.2 7.0 Other 12.9 17.1 1.o Energy 5.4 3.1 7.9 Water and Sanitation 16.6 3.6 0.5 Public Sector Management 0.0 10.0 20.1 Social Protection 0.0 0.1 24.0 Urban Development 5.2 5.7 0.0 TOTAL 100.0 100.0 100.0 Share of Adjustment Loans 0.0 0.0 63.9 2.14 There was a significant expansion o f Bank lending in this period, led by rural poverty (Table 4 and Figure 1). The Bank launched 13 new projects focused on rural poverty alleviation, most o f which were in the Northeastern states. This cluster o f projects was considered as the second generation o f rural development projects, more focused on poverty alleviation and more community-driven (the first generation had been initiated in the mid- 1980s). The basic objective o f these projects was to improve the standards o f living o f the rural poor through the provision o f basic services such as electricity, water and sanitation, on a community-driven basis. This cluster also included an innovative pilot project supporting market-based land reform. 2.15 There were few new education and health projects in this period, in good part due to the need to manage the existing projects and the fact that some o f them required restructuring. The new education project was the first o f a program o f three projects supporting improvements in school management and quality in poor regions. The new health project supported improvements in the delivery of health care in poor regions. These projects were highly consistent with the strategy, but the absence o f Bank assistance in the area o f early childhood development and nutrition was disappointing, as the CASs had been proposing a project in this area since the early 1990s. 2.16 Support to infrastructure continued, but under a strategy that was more elaborated and consistent with the broader PSD strategy developed b y IBRD and IFC. The PSD strategy had identified infrastructure deficiencies as one o f the major obstacles to private sector development and growth (other obstacles included deficiencies in financial intermediation and in the legal and regulatory frameworks). The infrastructure strategy that was developed maintained the support to decentralization, but also included support to the Government’s infrastructure privatization program, in order to help Brazil meet i t s large infrastructure needs. The strategy also assigned well-defined roles to B R D , IFC and MIGA. IBRD would provide technical assistance in the regulatory area and catalytic lending to specific sectors, 9 IFC would make pioneer investments in infrastructure, and MIGA would help attract foreign investment. 2.17 IBRD assisted the federal government in the development o f regulatory frameworks and agencies in the energy and water sectors. In the case o f energy, an existing loan was restructured to provide funds for technical assistance. In the case of water, assistance was provided through an existing water modernization loan in the water portfolio and a new water management loan in the environment portfolio. IBRD’s lending to infrastructure in this period was focused on the transportation sector, including two large projects accounting for ’ half o f the lending and supporting the restructuring and privatization o f federal railways and highways. These actions were all consistent with the strategy, but IBRD provided little assistance to the development o f regulation in the transportation sector, a surprising finding since most o f the lending was directed to this sector. 2.18 IFC started investing in infrastructure (primarily ports and roads), although not to the extent that would be expected-the share of infrastructure investments increased to only 25 percent o f new commitments and only 13 percent o f commitments in the whole period (Annex Table 3). MIGA, which had started operations in Brazil in 1994, provided guarantees on foreign investments with an accumulated value o f US$9 billion, mostly in infrastructure (Annex Table 3). 2.19 Another important development in this period was the Bank’s initial support to stabilization. The assistance strategies in 1995 and 1997 had pointed out the risks to stabilization posed by growing fiscal imbalances at the federal and state levels, and had also indicated the Bank’s willingness to support meaningful reforms through adjustment lending. The Government was not yet ready to start implementing a major reform, but requested Bank assistance in i t s efforts to sanitize the finances o f Brazilian states. In response to these requests, the Bank initiated a policy dialogue with several states, culminating in the elaboration of four loans. These loans accounted for 10 percent o f new commitments (Table 4) and supported the privatization o f banks and several infrastructure enterprises owned b y the respective states. These were de facto adjustment loans that had to be constructed as investment loans because of the lack of an appropriate lending instrument at that time.’ The Third Phase (1999-2002): Further Expansion o f Lending and Ship to Adjustment Lending in Support to Stabilization and Structural Reforms 2.20 The 1997 Asian crisis and the 1998 Russian crisis exposed the inconsistencies o f Brazil’s policy mix, ultimately forcing the Government to adopt a floating exchange rate regime and tighten fiscal policies considerably. These efforts were supported b y an international financial package led b y the IMF and including the Bank and the IDB. The Bank’s contribution would include several adjustment loans supporting fiscal and financial sector reforms, as described in the revised 1997 strategy approved b y the Board in December 1998. The package proposed in the revised strategy had to be submitted to the Board because the original 1997 assistance strategy did not foresee a substantial amount o f adjustment lending in the period. The Board had not yet approved sub-national adjustment loans. 10 2.21 The revised 1997 strategy and the assistance strategy in 2000 stress the need to avoid the economic stagnation and rising poverty that would result from a crisis and also indicate that the Government’s willingness to pursue critical reforms justified a shift to adjustment lending. The documents also point out that investment lending had to be temporarily curtailed anyway-the fiscal effort involved limitations on counterpart funds and borrowing restrictions on the states that would slow project disbursements. The documents do not mention, however, that the decision to shift to adjustment lending was also a response to a Government complaint that the Bank had been inconsistent, b y advocating fiscal adjustment at the macroeconomic level while pressing for project lending at the sectoral and state levels. 2.22 During the 1999-2002 period there was a further expansion o f lending, accompanied by a strong shift to adjustment lending, as shown in Figure 1and Tables 2 and 4. The Bank’s share in total financing flows to Brazil increased for the first time since the 1980s, although i t remained small. The package of adjustment loans included two loans supporting fiscal/public administration reforms, two loans supporting social security reforms, one loan preserving critical social expenditures, two loans supporting financial sector reforms and one loan supporting energy sector reform. The package was slightly different from the one originally envisaged in the revised strategy due to new policy developments. In particular, programmatic lending started in 2001 (this instrument did not exist previously), assistance to financial sector reforms was expanded, with a second loan approved in 2002, and one energy loan was approved in the same year, supporting changes in tariff policy and other sector reforms. These adjustment loans were accompanied b y technical assistance loans supporting reform implementation. 2.23 The Bank scaled down assistance to infrastructure in order to make room for adjustment lending and maintain the assistance to the social sectors. In education, the Bank initiated four new basic education projects in poor regions, two o f which focused on the states o f Bahia and Cearh. These projects show an evolution in design, including greater concern with teacher training. In health, the Bank launched three projects, including a second AIDS project building on the first successful project, a project strengthening the national disease control system (an evolution from two previous projects that dealt successfully with transmittable diseases), and a promising family health project that supports prevention and a pro-active contact with the communities. The Bank also started the third generation of rural poverty alleviation projects, encouraged b y the success o f the second generation, and a larger land reform project, building on the successful pilot project. In the environment area, the Bank initiated four green projects and finalized the preparation of an ambitious project designed to protect 10 percent o f the Amazon region financed b y Global Environmental Facility (GEF) and G7 grants. 11 3. Assessment of Bank Products and Services Assessment of Analytical and Advisory Services 3.1 Analytical and advisory services became an important component of the Bank’s assistance to Brazil in the 1990s. The greater importance o f non-lending services was due to the Bank’s awareness o f i t s declining importance as a source o f finance, and i t s potential capacity to contribute to reforms and the quality o f sectoral policies in Brazil. During the 1990-2002 period the Bank elaborated 75 formal reports (Economic and Sector Work- ESW) and numerous policy notes, covering all areas o f assistance. Half o f the ESW produced was prepared after 1997, indicating the increasing importance o f non-lending services in the Bank’s assistance strategy in the second half o f the decade. 3.2 Bank ESW on Brazil has received positive reviews by the Quality Assurance Group (QAG). As shown in Table 5, all ESW on Brazil reviewed b y Q A G has been rated as satisfactory, compared to 82 and 81 percent for the Latin American region and the Bank, respectively. The Operations Evaluation Department’s (OED) review o f Bank ESW on Brazil concluded that QAG’s positive assessment i s largely justified. The issues covered b y the ESW were relevant and the analyses were generally o f high quality. The Government has valued the Bank ESW, as well as the more informal policy notes. Some ESW have had a significant and positive impact in the formulation o f sectoral policies and the quality of Bank projects (Box 1). Table 5: QAG Evaluation of ESW Share of Satisfactory ESW (in %), 1996-2002 Brazil Latin America Bank Number % sat. Number % sat. Number % sat. 6 100 38 82 234 81 3.3 At the same time, OED has also found that the ESW program has fallen short o f i t s potential, and that there i s scope for improving ESW selection, clearance and, especially, dissemination. Government officials noted that on occasion ESW was initiated b y the Bank after informal requests from officials in sector Ministries, rather than as part o f a formally agreed work program. In some of these cases there was limited interest o f the relevant Ministry to participate in the final review phases. In these cases the clearance process took longer than usual, diminishing the value o f the ESW for policy purposes. At the time o f the OED review, the Bank and the Government had agreed to ensure clear sponsorship for every ESW product b y the relevant Ministry, to meet regularly to monitor the ESW program, and to establish a maximum period of 30 days for review and final clearance. 3.4 These measures w i l l enhance the relevance o f the Bank’s ESW if implemented effectively, but the main problem limiting the impact o f ESW - insufficient dissemination - s t i l l needs to be addressed. In some specific cases, the Bank made an effort at dissemination b y organizing conferences with the participation o f federal and state governments, politicians, academics, and NGOs, but in other cases active dissemination o f the ESW was poor. The Bank’s website in Brazil constitutes another channel for dissemination, providing a variety of useful information about the Bank, including i t s operations in Brazil, and an access to publications. However, very few publications are translated into Portuguese, creating a major obstacle for wider dissemination within the 12 country. The executive summaries containing the most important findings and recommendations have not been translated either, even though the cost o f translation would have amounted to a small fraction of the ESW budget. 3.5 Government officials have indicated that the Bank i s the multilateral institution that the Government mostly resorts to for technical advice. The Government has also indicated that the Bank can play an important “pedagogical” role in the country, informing Brazilian society about long-term structural issues, their potential solutions, and the trade-offs involved. To play this role effectively, however, the Bank must make a greater effort to lisseminate i t s work among the several potential audiences and to the population at large. Box 1: Examples of Outstanding ESW A report on environmental policies for the Amazon in 1992 provided an accurate diagnosis of the causes o f deforestation and recommended a coherent sustainable development strategy for the region. The report filled a strategic gap in the Bank and gave substance to the Government’s agenda. State Debt: Crisis and Reform (1995) was a timely report that examined the serious imbalances in the finances of Brazilian states that jeopardized the success o f the Real plan. The report was the basis for key policy discussions with the Government, the provision o f technical assistance to the National Treasury, and the starting point for defacto adjustment loans to state governments. A report to improve education in 1997 portrayed accurately the difficulties facing primary education in the Northeast and had a substantial impact on the design o f Government and Bank strategies to deal with this problem. The recommendations were incorporated in the design o f the Government’s FUNDESCOLA program, which has been supported by three Bank projects. Critical Issues in Social Security (2001) presented a comprehensive analysis o f the Brazilian pension system and detailed recommendations for reform. The report provided the analytical basis for the policy dialogue and the extensive Bank assistance, and i t s recommendationsremain valid. Attacking Brazil’s Urban Poverty (2001) and Rural Poverty Reduction in Brazil: Towards an Zntegrated Strategy (2001) are two outstanding reports on urban and rural poverty providing solid material for the design of coherent and integrated poverty reduction strategies. Assessment of Lending Sewices 3.6 During the 1990-2002 period OED reviewed a total o f 94 closed Bank projects in Brazil, representing US$13.2 billion in commitments. The share o f closed projects with outcomes rated as satisfactory b y OED amounted to 73 percent, showing a significant improvement over the previous decade and relative to the benchmarks, as shown in Table 6. During the 1980s the outcomes o f Brazil projects had on average worse ratings than projects in Latin America and Bankwide, whereas in the 1990-2002 period the ratings were comparable or better than these benchmarks. Computing the shares o f satisfactory projects b y the value o f commitments yields similar results (Annex Tables 5A-5D). 3.7 The improvement in project outcomes for the decade as a whole was due to the sharp turnaround in project performance after 1994. As shown in Table 6, in the first half o f the 1990s the share o f projects with satisfactory outcomes declined further to 33 percent, but that was more than offset b y the high performance in the 1995-2002 period. Furthermore, both investment and adjustment loans performed significantly better than the benchmarks during this period-90 percent o f the investment loans reviewed had a satisfactory rating, compared 13 with only 70 percent for the whole Bank, and all the four adjustment loans reviewed had a satisfactory rating, compared with 82 percent for the whole Bank (Annex Table 5D). Table 6: OED Evaluation of Closed Projects Brazil 59 73 33 91 Latin America 63 73 65 78 Bank 69 69 65 72 3.8 OED ratings on project sustainability and institutional development impact show a similar evolution-a pronounced decline in ratings in the first half o f the 1990s followed b y a sharp improvement after 1994. The sustainability o f 81 percent o f the projects closed after 1994 was rated as likely, compared with only 64 and 55 percent for the Latin American region and the Bank, respectively. The institutional development impact o f 67 percent o f the projects closed after 1994 was judged to be substantial, compared with only 45 and 39 percent for Latin America and the whole Bank, respectively (Annex Tables 5A-5D). 3.9 Q A G evaluations o f ongoing projects are consistent with OED evaluations. As shown in Table 7 , since 1996 (when Q A G started evaluating ongoing projects) Brazil projects have obtained better ratings than the benchmarks, both regarding quality at entry and quality at supervision. Finally, the annual computation o f projects at r i s k from supervision reports also indicates a pattern consistent with all the previous results. As shown in Figure 2, in the first half o f the 1990s 85 percent o f active Brazil projects were considered at risk, compared with 50 and 35 percent for Latin America and the Bank, respectively. Since the mid-1990s the share o f Brazil projects at risk has been equal or lower than the benchmarks. Table 7: QAG Evaluation of Ongoing Projects Share of Satisfactory Projects (in %), 1996-2002 Brazil LQtin America Bank Number %Sat. Number %Sat. Number %Sat. Quality at Entry 18 100 81 98 429 87 Quality at Supinision 21 86 147 78 800 77 Figure 2: Projects at Risk, 1990-2002 (% of Commitments) inn I 40 20 " 8 1990 1992 1994 1996 1998 2000 2002 I -Brazil +Latin America +Bank I 14 Factors that Have Contributed to the Improved Project Pe$omance 3.10 Several factors have contributed to the improved performance o f Bank projects in Brazil during the second half o f the 1990s. On the Government’s side two factors can be identified. First, success at stabilization lengthened the time horizons of policy-makers and allowed for a much better planning o f public expenditures. The problems caused b y deficient fiscal planning and lack of counterpart funds diminished considerably, benefiting all projects. Second, the Government that took office in 1995 was able to elaborate and implement a clear development agenda. Government ownership of the assistance program strengthened, several problematic projects were restructured, and project implementation units were generally strengthened as well. This also tended to benefit all projects. NGO involvement in the implementation o f projects in areas such as health and environment may also have contributed to improved project performance in these areas. 3.1 1 O n the Bank’s side there were also two contributing factors. The first was the Bank’s response to under-performing projects, involving changes in project design or curtailed lending to the sector. For example, some rural development projects o f the first generation were rated as unsatisfactory due to heavy overhead costs and failure to respond to the demands o f beneficiaries. The second generation was considered more effective at rural poverty alleviation (essentially because o f the community-driven approach) and was generally rated as satisfactory. Irrigation projects closing in the early 1990s were rated as unsatisfactory due to their failure to boost rural growth and reduce poverty, and were curtailed in the second half o f the decade. Most energy projects closing in the early 1990s were rated as unsatisfactory due primarily to low tariffs and the poor financial performance o f energy companies, leading the Bank to curtail lending to the sector. 3.12 The second contributing factor on the Bank’s side was the increasing Bank presence in the country. In 1996 the Bank increased the number o f staff working in the environmental area. In 1997 the Country Management Unit (CMU) was transferred to Brasilia and the number o f staff in the field increased considerably. In mid-2002 the Bank had a total o f 85 technical and support staff working in three offices, of which 10 staff were from headquarters. The Brasilia office had 70 employees, including the country director, several sector specialists, and an implementation team dealing with procurement, disbursements, and audits. The Recife office had 10 employees and focused in the management o f agriculture projects in the Northeast. The Fortaleza office had 5 employees and carried the policy dialogue in the Northeast. I t was being upgraded to also manage education projects in the region and handle distance learning programs. 3.13 The decentralization o f Bank operations has improved Bank knowledge o f the Brazilian economy and the quality o f the policy dialogue, resulting generally in better project design. Project implementation has also improved with the greater proximity o f staff to project areas. For example, the presence o f agriculture specialists in the Recife office has allowed the Bank to closely supervise a large number o f agriculture projects and sub-projects in the Northeast. Procurement and disbursement problems, relatively common in the first half o f the 1990s, have been reduced with the presence o f the implementation team in the Brasilia office. Government officials indicated to OED that the quality o f the Bank’s work had improved significantly after decentralization. Officials stressed that the Bank had become more sensitive to political constraints and more responsive to the country’s needs. 15 3.14 The improved performance o f the Brazil portfolio did not result in high average costs relative to other areas of the Bank. As shown in Table 8, average project costs in Brazil were higher than average costs in Latin America, but not by a wide margin, and were significantly lower than average costs Bankwide. If average lending costs are defined as the ratio of costs to the value o f commitments, average costs in Brazil drop significantly below the two benchmarks. The high share o f local staff employed has probably helped the Bank maintain average costs under control. All in all, these indicators suggest that the decentralization enabled the Bank to manage effectively a large and complex assistance program, involving an average o f 55 active projects every year, spread over many states and a large territory. Table 8: Average Lending Costs, per Project and per Commitment, 1990-2002 Brazil Latin America Bank Lending Costs per Project (in US$ OOO) 655 623 763 O / W ~preparation costs 372 360 427 ESW Costs per ESW (in US$ 000) 146 143 125 Operational Costs per Project (in US$ OOO) 715 667 819 Lending Costs per Commitment (in %) 0.45 0.58 0.85 Assessment of the Bank’s Role in Aid Coordination 3.15 The Bank group and the IDB were the largest sources o f external development finance to Brazil in the 1990-2001 period, with roughly US$16.5 billion in total commitments, including IFC but excluding MIGA’s guarantees (Annex Tables 3 and 4). The Bank and the IDB performed largely complementary roles, with just a moderate overlapping between the two institutions. The IDB provided very limited assistance to agriculture, i t s education, health, and environment portfolios were smaller than the Bank’s and generally focused on different areas (e.g., secondary education). I t s infrastructure portfolio was much larger than the Bank’s and the institution was also more active in urban development, filling a gap left b y the Bank in the period. Japan and Germany were the third and fourth largest sources o f external development finance, with US$3.2 and US$0.2 billion o f lending, respectively, mostly to infrastructure. 3.16 The Bank held consultations with the IMF, the IDB and the bilateral agencies (all o f which have offices in Brazil), but coordination o f external assistance has essentially been performed b y the Government. The Ministry of Finance i s responsible for the overall policy dialogue with the IMF and the Bank, as well as the Bank’s adjustment loans. The Ministry o f Planning and COFIEX (an inter-ministerial committee) are responsible for screening and approving investment loans from all multilateral and bilateral agencies. The division o f labor between the Bank and the IDB emerged essentially from the Government’s dialogue with the two institutions. There does not seem to be a need for a more active Bank role in overall aid coordination, although the Bank and the IDB would both benefit from more frequent consultations and sharing o f experiences. 16 4. Assessing the Development Impact of the Bank’s Assistance 4.1 This section addresses the three major questions raised in Section 1 , namely, whether the objectives of the Bank’s assistance program were relevant, whether the program was well designed and consistent, and whether i t achieved i t s two central objectives o f poverty reduction and growth. The previous section showed the sharp improvement in the ratings of individual projects during the 1990s. The favorable ratings indicate that, in general, project objectives were relevant and achieved, but they provide little information on the overall impact o f the Bank’s assistance. T o assess the latter, i t i s necessary to go beyond individual project ratings, examine the country’s progress at poverty reduction and growth and assess the extent to which the Bank contributed to these outcomes. 4.1 Assessing the Relevance and Consistency of the Bank’s Strategy 4.2 Brazil entered the 1990s suffering the consequences o f more than a decade of high inflation and slow growth. The share of the population below the extreme poverty line had risen from 16.5 to 19 percent and income distribution had worsened further, due to a pronounced fall in the real incomes o f the poor (Figure 3 and Table 9). Poverty indicators for the Northeast were appalling. Under these conditions, the Bank’s focus on poverty alleviation was relevant and appropriate. 4.3 As mentioned in Section 2, the Bank’s strategy to reduce poverty comprised elements expected to address the roots o f poverty and elements designed to foster growth, with an emphasis on the first set o f elements. The first set included essentially assistance to human resource development (education, health) and expansion o f basic services to the poor, while the second set comprised financial and technical assistance to the build-up o f infrastructure and to the removal o f structural inefficiencies. 4.4 The decision to provide more intensive assistance to human resource development, especially education, was overall appropriate. Poverty and inequality are highly correlated with differences in educational outcomes,* and Brazil had made a very slow progress in improving key social indicators in the 1980s (Table 9). The country lagged Latin American and Middle-Income countries in many aspects, despite having a higher per capita income than these groups o f countries. The decision to pursue a diversified strategy o f poverty reduction (combining human resource development with other elements) was also justified by the finding that growth reduces poverty in Brazil, but with an elasticity lower than one.’ 4.5 The presentation of the assistance strategy in the CASs may be criticized for failing to explain and justify more clearly the greater emphasis on human resource development. For example, the dichotomy between poverty reduction and growth objectives pursued in most CASs was to some extent artificial. Education and health are arguably essential ingredients to long-run growth as well, although there i s an ongoing debate in the literature on the channels and the extent to which human capital affects growth.” Also, access o f the poor to * The empirical literature i s extensive. See, e.g., Almeida and Barros (1992); Barros, Henriques and Mendonca (2000), Ferreira and Litchfield (2001); and Menezes-Filho (2001). The literature is also extensive. See, e.g., Neri (1999) and Pinheiro et a1 (2001). lo Krueger and Lindahl(2001) provide a comprehensive and critical review of the literature. See also Pritchet (2001) and Temple (1999). 17 basic services not only alleviates poverty in a broad sense but may also foster subsequent growth.” Another criticism that may be raised i s the failure o f the CASs to clarify the time horizon within which these interventions would produce their effects. Education may be an important ingredient for both poverty reduction and growth, but i t may take several decades for i t s effects to materialize fully. 4.6 These criticisms on the presentation of the assistance strategy in the CASs do not imply criticisms on the substance o f the strategy, however. On the contrary, the assistance program was overall well designed, underpinned b y sound analytical work, consistent with the main program objectives, and with the Government’s own development agenda. The analytical support to the CASs improved substantially during the 199Os, with the elaboration of high quality poverty assessments, PSD strategies, and other ESW, prepared with the substantial participation of top Brazilian researchers. 4.7 The consistency o f the assistance program with the Government’s own development agenda also improved during the 199Os, with increasing Government participation in the CASs, macroeconomic stability, and the formulation o f clear sectoral programs (as defined in the 1996-1999 and 2000-2003 multi-year plans).12 For example, the Government defined primary education as a clear priority and established the goal o f reaching universal access to primary education by 2007. The Bank contributed to this effort through a significant assistance to primary education. In health, the Government’s agenda stressed the need to expand the access o f the population to health care and reduce infant mortality through reforms o f the public health system, family health programs, and control o f transmittable diseases. The Bank’s assistance to the health sector covered these areas. In infrastructure, the Government’s program included, inter alia, the recovery, decentralization, and partial privatization o f federal highways, the recovery of state highways, the privatization o f federal railways, and the decentralization o f urban rail transportation. The Bank also contributed with transportation projects in all these areas. 4.8 During the 1990s the Bank had to operate in an increasingly decentralized mode, dealing directly with Brazilian states. The focus on the Northeastern states was appropriate, given the region’s high levels o f poverty (Table 9), and the strategy to provide assistance to these states was also reasonable and cost-effective. Rural poverty alleviation projects were prepared and implemented in every Northeastern state, an appropriate decision given the high levels o f rural poverty in the region. Assistance to education and health was provided in part through federal projects covering all Northeastern states and other poor states. More specific assistance was provided to Bahia and Cearii, in the areas o f education, health, water and sanitation, and urban development. The selection o f these two specific states seems to have followed pragmatic rules. These are large states, accounting for 43 percent o f the Northeast population. The two states had active and reformist Governments and also had the capacity to borrow from the Bank, as a result o f previous efforts to adjust their public finances. 4.9 The shift to adjustment lending at the end o f the decade was also justified and consistent with the Bank’s central objectives o f poverty reduction and growth. The 1995 CAS had already indicated the need to provide assistance to fiscal, social security and I’ Pinheiro eta1 (2001). See Brasil em AqBo (1996) and Avanqa Brasil (2000), Ministkrio do Planejamento,Brasilia. 18 financial reforms to consolidate stabilization and create the conditions for growth. Poverty assessments stressed the need to maintain stability and improve the targeting of social programs, and PSD strategies identified distortions in financial intermediation as a major obstacle to higher growth. When the Government requested support to i t s fiscal, social security, and financial sector reforms the Bank acted promptly through adjustment and technical assistance loans, while also preserving the assistance to the social sectors. Table 9: Poverty and Soci, Indicators, 1 90-2001 1980 1990 2000 Poverty Rate (% of households) 40.0 40.7 33.6 Northeast 66.8 68.6 57.4 Northeast Rural 84.9 72.6 Extreme Poverty Rate (% of households) 16.6 19.1 14.6 Northeast 36.2 41.8 30.0 Gini Coefficient 58.4 61.6 59.6 Theil Coefficient 68.2 77.3 71.3 Ratio of Average Income 10% richest/40% poorest 21.5 26.7 23.6 Adult Illiteracy Rate (8of population ages 15 and above) 22.8 18.7 12.4 Northeast 41.6 36.4 24.3 Youth Illiteracy Rate (% of population ages 15-24) 12.0 9.8 4.2 Northeast 27.0 22.7 9.6 Net School Enrollment Rate, Primary Education (% of pop.) 80.1 83.8 95.4 Northeast 69.1 72.0 92.8 AIDS incidence (per 10,OOO pop.) 0.6 " 1.o Malaria Incidence (per 1,OOO pop.) 14.0 32.7 30.3 Life Expectancy at Birth (years) 62.6 65.6 68.6 Infant Mortality Rate (per 1O , OO live births) 79.2 48.0 29.6 Northeast 112.8 72.9 44.2 Access to Electricity (% of households) 74.9 87.5 96.0 Northeast 50.8 70.5 89.4 Northeast Rural 12.6 35.4 66.5 Access to ImprovedWater Source (% of households) 60.7 74.2 87.3 Northeast 30.8 43.3 67.0 Northeast Rural 4.9 10.6 22.8 Access to Sanitation (% of households) 51.3 63.2 71.3 Northeast 30.9 43.2 54.2 Households with Refrigerators (% of households) 56.6 70.6 85.1 Northeast 28.7 41.5 64.9 Northeast Rural 5.9 - ._ 14.5 33.9 Sources: IBGE (Household Surveys-PNAD): IPEA, INEP, Mi Notes: 11AIDS incidence increased to 1.4 per 10,OOO people in the mid-1990s (Annex Table 10). 4.2 Assessing the Contribution of the Assistance Program to Poverty Alleviation The Program's Contribution to the Decline in Income Poverty Ratios 4.10 During the decade Brazil achieved progress at poverty reduction, as indicated by the decline o f 7 and 5 percentage points in the share o f the population below the poverty and extreme poverty lines, respectively (Figure 3 and Table 9). Brazil seems to have achieved greater progress at reducing poverty than most Latin American countries, according to a comparative analysis made b y the United Nation's Economic Commission for Latin America (ECLAC).I3 Progress at poverty reduction was greater in the Northeast, as indicated b y the 10 percentage point reduction in poverty rates. In the rural Northeast there was an even greater reduction o f 12 percentage points (Table 9). l3UN-ECLAC (2002). 19 4.11 Most o f these gains are related to the success of the 1994 stabilization plan, which stopped the erosion o f the real incomes of the lowest two tenths of the income distribution (Figure 3). The gains at the end o f the decade were minor. Stabilization had a strong impact on poverty reduction because high inflation affected disproportionately the poor. l4 The successful reduction in inflation also helps explain why Brazil achieved greater progress at reducing poverty than the average Latin American country, despite a similar growth performance-between 1990 and 2002 i t s per capita GDP grew at an average rate o f 1.1 percent per year, similar to the Latin American average o f 1.2 percent per year. 4.12 Assessing the contribution of the Bank’s assistance program to the decline in national poverty rates in the 1990s involves assessing i t s contribution to stabilization, as the two events are clearly related. The Bank did not play any meaningful role in the 1994 Real Plan, but contributed to the consolidation o f stabilization in the second half o f the decade. The state reform loans contributed to the effort at fiscal reform and privatization within these states in the 1995-1998 period. After 1998 the Bank stepped up i t s support to stabilization, through a package o f five adjustment loans amounting to US$2.8 billion and several complementary technical assistance loans. The package o f adjustment loans (all of which were one tranche operations) included two fiscal reform loans, one social protection loan and two social security reform loans. 4.13 The two fiscal reforms loans supported the fiscal stability program that shifted the primary balance from deficits into surpluses o f more than 3.5 percent o f GDP, and the passage of an impressive body o f legislation reforming public administration and the finances of Brazilian states. This included the Law o f Fiscal Responsibility and the Law o f Fiscal Crimes. This body o f legislation has addressed a problem of fiscal imbalances and excessive debt in the states that seemed intractable in the mid-1990s. In particular, the Law o f Fiscal Responsibility has institutionalized fiscal discipline, transparency and accountability, changing radically the way in which government officials at all levels think and act. 4.14 The social protection loan was designed to minimize the impact o f the 1999 fiscal adjustment on the poor, by minimizing cuts to social assistance, health, and education, and preserving a number o f programs targeted at the poor. The loan achieved i t s objectives and was rated as highly satisfactory b y OED. Finally, the two social security loans supported a constitutional amendment that introduced the principle o f actuarial balance for both public and private sector workers, introduced a minimum retirement age for public sector workers and a less generous benefit formula for private sector workers. These were important reform steps, but the Brazilian social security system remains imbalanced, a problem that w i l l be further discussed below. 4.15 The program’s contribution to the consolidation o f stabilization must be acknowledged, both financially and in the provision o f technical assistance, but should not be l4Poverty had also been sharply reduced in the mid-l980s, but such reduction was short-lived because the stabilization attempts in that period (the Cruzado Plan) did not succeed (Figure 3). Inflation may affect the poor through more than one channel. See, e.g., Neri (1995), Ferreira and Litchfield (2001), and Cardoso, Barros and Urani (1995). 20 Figure 3: Poverty Ratios and Income Growth by Deciles Evolutionof Poverty and Esrtreme Poverty in Brazil, 1981-U)ol 60 - - - - - - - - - - ....................... -- > I -Poverty Rate -.--- EhemPov&y Rate I Growth of the Real Income of Each Tenth of the Income DistributionBetween 1981 and 1993 I -2" 1 2 3 4 5 6 7 8 9 10 Tenths of the Income Distribution Growth of the Real Income of Each Tenth of the Income Distribution Between 1993 and 2001 30 25 8 .8 v 20 v1 2 15 5 g 10 8 5 0 1 2 3 4 5 6 7 8 9 10 Tenths of the Income Distribution 21 exaggerated either. The IMF led the international financial package in 1998-1999, through a US$18 billion stand-by agreement, and the IDB also participated in the rescue package with two loans amounting to US$3.4 billion. Therefore, the Bank can claim only a moderate and shared contribution to the reduction in national income poverty rates in the 1990s. 4.16 The Bank’s contribution to the decline in poverty rates in the Northeast and the rural Northeast was probably larger, as the institution was focusing i t s resources on the region. However, i t i s difficult to link Bank interventions such as the rural poverty alleviation projects to increases in household incomes and reductions in poverty-it would require a more detailed analysis o f household incomes and poverty in municipalities covered b y the projects (an impact evaluation study i s planned for the third generation of rural projects but was not done for the second). Besides, there were also other factors at work, including the introduction o f rural pensions early in the decade. 4.17 The Bank’s main contribution to poverty alleviation in Brazil i s more clearly reflected in other related indicators also shown in Table 9. The Bank contributed to poverty alleviation in Brazil b y helping the Government address the fundamental determinants of poverty, including education and health outcomes among the poor and the access o f poor urban and rural populations to basic services. The Bank’s contribution to these two sets of outcomes i s assessed below. The Program’s Contribution to Improved Education and Health Outcomes 4.18 During the 1990-2001 period there were impressive gains in the education area, as shown by the sharp decline in the youth illiteracy rate and the sharp increase in the net enrollment rate in primary education (Table 9). The gains in the Northeast were particularly impressive-the net enrollment rate in primary education had increased b y only 3 percentage points in the 1980s but jumped from 72 to 93 percent in the 1990s, an impressive increase of 21 percentage points. The progress in the education area produced a double convergence o f key education indicators-the Northeast indicators converged with the national averages and the latter converged with the averages o f Latin American and Middle Income countries (Figure 4). This increase in net enrollment in the Northeast also implied that roughly one million and a half children in the lowest two tenths of the income distribution gained access to schooling in this period. 4.19 The Bank made an important contribution to these results. During the 1990-2002 period the institution launched ten projects focused on basic education, seven o f which benefited the Northeast and other poor regions. Bank projects have contributed to the positive outcomes by, inter alia, enhancing access o f poor children to education through the provision o f infrastructure and education materials (benefiting directly one million poor children in the Northeast and other poor regions), building the capacity o f state secretariats of education (enabling them to manage the increased student population), helping federal and state governments set up information and evaluation systems for primary education, implementing minimum operational standards in all schools, empowering local communities, and introducing programs for disadvantaged children. 22 25 - -.-... -.-- **-...__*_ .*..... - -. $20 2 g 15- ...... -. .... -.. 2 - e. &lo- -----------.- - e - . , 8 ru 0 5 - 8 0, ..-....Brazil (Northeast) -Brazil --- -Mid Inc. -LAC N e t Enrollment in Primary Education, 1990-2000 100 I I - - - - - - .Brazil (Northeast) -Brazil LAC -. _. . ---- - - - - - _. .* 63 - .__.----- * -* 23 4.20 The two basic education projects launched in the Northeast in the early 1990s received only a moderately satisfactory rating b y OED because they did not pay sufficient attention to teachers’ training and classroom effectiveness. Quality o f teaching in the Brazilian primary education system i s s t i l l unsatisfactory, especially in the Northeast, as revealed b y the national and state evaluation systems (which the Bank helped introduce) and international student comparisons. However, the design of Bank projects improved during the decade. Bank ESW provided important inputs for the design o f the national school improvement program (Fundescola) initiated in 1997 and supported b y three Bank projects. The full implementation o f the program should lead to better learning outcomes. The recent basic education projects in Bahia and Cearii show further improvements in design, including better targeting of the poorest municipalities, improvements in school supervision, and a greater concern with teacher training programs. 4.21 In the health area the gains have also been impressive, as indicated by the sharp decline in infant mortality rates (especially in the Northeast), the increase in life expectancy, the decline in the transmission of HIVIAIDS and the decline in deaths from communicable diseases, such as, Malaria, Chagas, and Schistosomiasis (Table 9). As in the case o f education, there was also a double convergence of key health indicators (Figure 4). The Bank contributed to these positive outcomes as well, although in the case o f health the extent o f the Bank’s contribution varied between the two main classes o f projects. 4.22 The first class o f projects focused on the fight against communicable diseases and included five projects that produced satisfactory outcomes. The projects targeted effectively poor populations by the nature of the diseases addressed. For many diseases, mortality rates dropped and incidence rates stabilized or dropped as well. A good example i s Malaria: a sharp increase in incidence in the 1980s was stopped in the 1990s and was accompanied b y a dramatic reduction in mortality. The second class comprised three projects focused on improvements in health infrastructure and policy. T w o o f these projects were exclusively focused in the Northeast, and the third was a national project benefiting primarily the North, Northeast, and Center-West regions. These projects created or recovered infrastructure, provided equipment, and trained staff in hospital management. This class o f projects also contributed to better health outcomes overall but was ultimately rated as only moderately satisfactory, due the partial success in expanding the access o f the poor to health care. 4.23 Improvements in infrastructure and equipment are a necessary condition for expanding access of the poor to health care, but sometimes are not sufficient. In many cases, access remained constrained b y factors not addressed by the projects, such as lack o f health professionals and lack of adequate financing. The project was designed on the assumption that these constraints would be removed b y other programs, but this did not happen to the extent that was expected, a problem that may be interpreted as a flaw in project design. In sum, the Bank’s contribution to better health outcomes in Brazil was significant, particularly due to the first class o f projects. The second class o f health projects also contributed to improved health outcomes among the poor, but this contribution fell short o f i t s potential, due to mismatches between the build-up o f infrastructure and the development o f human resources in the health sector.15 A previous OED evaluation of Bank assistance to the health sector had achieved the same conclusions, but without including the AIDS-I1and REFORSUS Projects, which had not progressed enough to be properly evaluated. 24 4.24 One disappointing aspect o f the Bank’s assistance to education and health was the absence o f a project addressing early childhood development and nutritional issues. These issues were examined in Bank ESW but the program did not include even a pilot project, despite the well-known positive impact of early childhood development on educational outcomes, and the fact that several CASs proposed assistance in this area. A recent education project in Cearii addresses this issue for the first time, but not with the focus and intensity that it deserves. The Program’s Contribution to Expanding Access of the Poor to Basic Services 4.25 During the 1990s and early 2000s there was significant progress in expanding the access o f the population to basic services, such as electricity, water and sanitation (Table 9 and Annex Table 10). I t i s difficult to pinpoint with accuracy the extent to which poor populations benefited from this expansion, but the relatively large gains in the Northeast and the rural Northeast suggest that large numbers of poor households were reached. The Bank participated in this effort through i t s rural poverty alleviation projects and through i t s water and sanitation projects, and can claim an important contribution to these outcomes. 4.26 The second generation of rural poverty alleviation projects was initiated in 1995 and comprised thirteen projects, eight of which located in Northeastern states. Roughly 60 percent o f disbursements in the Northeast projects financed the provision o f electricity and water to poor rural households, in approximately equal amounts. The number of households benefiting from access to water and electricity in these projects amounted to 541 and 294 thousand, respectively, corresponding to 2.1 and 1.2 million people. These numbers account for a large share o f additional households in the rural Northeast covered by these services in the 1995-2001 period, according to household survey data (PNAD), suggesting that the Bank made a fundamental contribution to these positive outcomes. 4.27 The Bank’s contribution to the provision o f water and sanitation to the urban poor was also significant. The water portfolio included three classes o f projects. The first class comprised two projects designed to extend water and sanitation to the urban poor with low cost technologies. The first project provided both water and sanitation to roughly one million poor people in the 1992-1997 period, accounting for about 15 percent o f the increase in total coverage, and probably for a much larger share of the increase in the coverage o f the poor. The second project i s expected to extend coverage to a similar number o f people. 4.28 The second class comprised three projects benefiting four states, and instituted a new approach to the recovery o f some o f the most polluted urban watercourses in the country, based on integrated water management and including environmental and land-use aspects. These projects benefited a large urban population and probably benefited a large number o f the urban poor, judging b y their location and designt6 Finally, the third class comprised three projects designed to improve the sector’s regulatory framework and encourage private sector participation. The need for modernization and investment in the water and sanitation sector i s obvious given the low coverage rates, particularly in sanitation. The Bank provided substantial assistance to the Government in the regulatory area but, unfortunately, progress at passing critical legislation has been slower than expected, due to the lack o f political l6Unfortunately, these projects do not always provide accurate estimates of the number of poor people that have benefited from greater access to these basic services. This deficiency can be interpreted as a flaw in design, as well as inadequate monitoring and evaluation. 25 passing critical legislation has been slower than expected, due to the lack o f political consensus on key issues, such as the concession of water companies. I t i s possible that the Bank did not explore i t s potential mediation role in this sector, a role that was actually proposed in the CASs. 4.29 Another criticism that can be raised i s the insufficient integration between different classes o f projects dealing with water. In the late 1990s the Bank launched two water management projects in the environment portfolio, designed to improve the capacity of the federal government and Northeastern states in the area of water management, and also improve water supply in the rural Northeast, Therefore, the Bank had water projects in three different sectoral programs (water, rural poverty, and environment), some o f them operating in the same regions and with similar objectives. I t i s not clear whether the policies that guide the Bank’s assistance in this area are being harmonized across the three programs and whether the potential synergies among these projects are being explored. 4.3 Assessing the Contributionof the Bank’s Program to Sustainable Growth The Program’s Contribution to Brazil’s Growth Performance in the 1990s 4.30 Between 1990 and 2002 Brazil grew at an average rate o f 2.5 percent per year and 1.1 percent per year in per capita terms, similar to the average growth rate of the rest of Latin America-2.9 percent per year and 1.2 percent in per capita terms. Brazil’s growth performance was much worse than the rest of Latin America in the early 1990s but improved in the mid-1990s and was better than the region at the end o f the decade (Table 10). Brazil seems to have been less affected b y the international slowdown than the rest of Latin America, suggesting that the stabilization measures in 1999 have paid off, at least to some degree. However, i t s growth performance in the 1990s was generally disappointing, by comparison with other large countries and with i t s own historical record. Brazil’s growth performance in the 1990s looks particularly disappointing considering that several structural reforms were implemented during the decade. 1991-1993 1994-1998 1999-2002 1994-2002 Brazil 1.9 3.2 2.0 2.1 Latin America w/o Brazil 5.4 3.6 0.3 2.1 Latin America w/o Brazil and Argentina 4.2 3.7 1.8 2.8 Mexico 3.3 2.9 2.8 2.9 China 12.3 9.9 7.3 8.7 India 3.6 6.6 5.1 5.9 OECD 1.2 2.7 2.3 2.5 4.3 1 The Brazilian growth literature indicates that Brazil’s disappointing growth performance has been fundamentally due to the lower growth o f capital accumulation in the last two decades, and that this factor may explain as much as 70 percent of the decline in output growth rates after 1980. The growth studies also show that the structural reforms implemented in the 1990s actually generated significant productivity gains-total factor productivity (TFP) increased more in the 1990s than in any other period and explains a large share o f GDP growth in this period. The literature concludes that the reforms in the 1990s failed to restore high growth rates, not because they failed to generate productivity gains, but 26 because they failed to raise the levels of investment to a significant degree (Figure 5). The 2001 energy crisis provides a very illustrative example of how the insufficient level o f investment in infrastructure has affected the country’s growth performance-the rationing of energy in 2001 reduced GDP growth in that year alone by a full percentage point.17 4.32 The large flows of FDI in the 1990s have not contributed to substantive increases in investment in infrastructure and other sectors (excepting telecommunications) because a large portion of m>I was driven b y mergers and acquisitions, including the privatization of state enterprises. This type of FDI generally produces efficiency gains but may not raise investment levels significantly, at least in the early stages. The problem o f low investment levels during this period was essentially due to three factors, namely: (i) low domestic savings (arising in good part from public dissavings); ( ii)lingering legal and regulatory deficiencies that increased risk and deterred new investments (including expansion investments by new private owners); and ( iiilingering deficiencies in financial ) intermediation, resulting in large intermediation spreads and lending rates. Figure 5: GDP Growth, 10 Year Moving Averages (% p.a.) and Real Fixed Investment (% of GDP), 1965-2001 10% 25%% 8% E 20%.2 6% 26 15%- tl 4% 8 0 10% @ 2% j v1 0% 5% -2% 1 1965 1970 1975 1980 1985 1990 1995 2000 -GDP Growth - - - GDP Per Capita Growth eInvestmentRatio 4.33 The Bank’s assistance program contained many good elements addressing the causes o f Brazil’s weak growth performance. I B R D ’ s adjustment loans in support to fiscal and social security reforms aimed at reducing fiscal imbalances (caused in good part b y excessive wage and pension payments) and increasing public and national savings. IBRD also provided technical and financial support to the development o f regulatory frameworks and agencies in the areas of energy and water. IBRD’s transportation portfolio included operations supporting private sector participation in railways, highways, and urban transportation. IFC invested in toll roads, ports, and telecommunications, and also provided financial support to S M E s through lines o f credit and venture capital. MIGA guaranteed foreign direct investments with an accumulated value o f US$9 billion, mostly in energy and ” Pinheiro et a1 (2001) provide the most comprehensive analysis o f Brazil’s growth performance in recent decades. Abreu and Verner (1999, Bacha and Bonelli (2001), Bonelli and Fonseca (1997), Ferreira and Malliagros (1999) and Giambiagi (2002) also emphasize the weak pace o f capital accumulation after 1980 and the larger productivity gains in the 1990s. The results o f these studies contradict for Brazil the conclusions of Easterly and Levine (2001) based on cross-country data, that factor accumulation does not have a significant impact on growth. I t i s possible that an earlier correction o f the overvaluation o f the Real would have enabled somewhat higher growth rates in the late 199Os, but the supply constraints would s t i l l have limited Brazil’s growth performance in this period. 27 telecommunications. Finally, the Bank’s program also included support to financial sector reforms through two adjustment loans and two technical assistance loans. 4.34 The assistance was generally of good quality, but i t did not have a strong impact on investment and growth in the 199Os, due to several factors. First, the fiscal and social security reforms supported by the program started late in the decade, and have produced so far only a modest increase in national savings (Annex Table 9). The fiscal reforms achieved their main objectives (para 4. lo), but the social security reforms did not go far enough-the reforms have reduced the actuarially projected expenditures and deficits, but they have not reduced the disproportionately large pension expenditures and deficits (12 and 5 percent of GDP, respectively) that burden public finances and depress public and national savings. The reforms were less ambitious than expected because o f strong political opposition. Among other setbacks, Congress did not pass measures to increase the retirement age o f private sector workers and to reduce the large benefits of public sector retirees, and the Supreme Court overruled a law imposing contributions on public retirees. 4.35 Second, progress in developing the regulatory framework for infrastructure has been insufficient and uneven across sectors and layers of government. Federal regulatory agencies in the energy and water sectors were created with Bank support, but the transformation of the two sectors remains uncertain. In the case of water, a major piece o f legislation (prepared with Bank support) has not been approved b y Congress yet. The regulatory framework for energy was improved, including several changes in the structure o f tariffs designed to ensure cost recovery and protect low-income consumers, but the transformation of the energy sector remains incomplete and uncertain-distribution was largely privatized but generation remains largely state-owned. Finally, there has been little progress in developing the regulatory framework for transportation and building regulatory capacity at the level o f states and municipalities (which have responsibilities in several areas). 4.36 Third, the financial sector reforms supported b y the Bank have not been completed either. The two programmatic loans have supported a number o f important reform steps, including the liquidation or privatization o f most banks owned by the states (a major achievement, as these banks had been repeatedly used as a source o f finance b y state governments and had become insolvent), the restructuring o f federal banks, improvements in bank and corporate legislation, and improvements in financial supervision. However, federal state banks s t i l l account for 40 percent o f banking assets, the share o f directed credit remains large, financial intermediation i s heavily taxed (by explicit and implicit taxes such as reserve requirements) and creditor rights remain weak. As a result, competitive access to credit has remained restricted and intermediation spreads have remained large. 4.37 In sum, the part o f the Bank’s assistance program designed to improve efficiency and growth produced mixed results. The program strengthened the institutional capacity of states and municipalities to manage their own infrastructure, promoted private sector participation in infrastructure, contributed to the creation o f some federal regulatory agencies, and introduced several improvements in the financial sector. These outcomes have probably generated efficiency gains with some positive growth effects. However, Bank assistance in the regulatory area proved largely insufficient and some key reforms supported b y the program started late in the decade and have not gone far enough. The regulatory framework needs to be substantially strengthened and key reforms need to be completed to generate the 28 increase in investment and the additional gains in efficiency that are required for a much better growth performance." The Program's Contribution to Improvements in the Environment 4.38 The Bank was the first international organization that provided financial and technical support to environment-related activities in Brazil. During the 1990s and early 2000s this assistance was greatly expanded, through nine projects and the management of grants from the Global Environment Fund (GEF) and the G7 countries (PPG7). The green agenda (forest management) accounted for 60 percent o f the overall financial assistance, including loans and grants. Some o f the projects in the green agenda were designed to correct the environmental degradation caused b y infrastructure projects financed b y the Bank in the 1980s (e.g. the Polonoroeste project). The brown agenda (industrial pollution) accounted for 10 percent of commitments and the blue agenda (water management) accounted for the remaining 30 percent. In addition, there were also projects in the water and sanitation portfolio dealing with the recovery o f urban watercourses, as described above. 4.39 Bank assistance to the environment has been generally satisfactory, especially in the green area. Although environmental indicators in Brazil s t i l l lag those o f developed countries in several areas, they generally improved during the decade. As shown in Table 11, the average rate o f deforestation in the Amazon was s t i l l high during the 1990s but declined Table 11: Environmental Indicators, 1980s and 1990s (unless otherwise indicated) 1980s 1990s Average Annual Deforested Areas in the Amazon (km') 20,374 16,806 Share of Protected Areas in the National Territory (%) * 2.4 6.5 Proportion o f Inadequate Days in Air Quality in the city of Silo Paulo (%) 9.7 3.5 Average Annual Emissions o f Carbon Dioxide (metric tons per capita) 1.4 1.6 Average Annual Emissions of Carbon Dioxide (metric tons per capita) in other regions Latin America 2.3 2.4 North America (Canada and the US) 18.8 19.2 High Income OECD 11.5 12.1 Notes and Sources: 1/ Average deforestation in 1977-1989 and 1990-2000, source: INPE 2/ 1989 and 2002; source: WRI, World Bank, 31 1991 and 2000, source Cetesb; 4/ Average emissions in 1980-89 and 1990-1998; sources: World Bank and Carbon Dioxide InformationAnalysis Center. The Brazilian literature and OED audits of transportation projects generally conclude that the privatization of infrastructure has improved the financial and operational management o f enterprises, increased efficiency in the utilization o f the infrastructure, improved quality, and reduced the number o f accidents. However, there has not been any significant expansion o f capacity, with the exception o f telecommunications, due in good part to lingering regulatory deficiencies. Annuati et a1 (2002) examine the financial and operational performance of privatized firms, Pinheiro (2000) provides an overview of the first results of privatization, and Castro (2001 and 2002) examine the regulatory challenges in the transportation sector. 29 to lower levels relative to the 1980s. The size of protected areas increased significantly and air pollution in large cities such as Silo Paulo declined. The emissions o f carbon dioxide (measured in metric tons per capita) increased during the 199Os, but this also happened in other regions, and average emissions in Brazil remained substantially below the averages of Latin America, North America, and the OECD (Table 1l).19 The Bank’s assistance program contributed to these results not only through specific project interventions but also through a critical build-up o f institutional capacity in the Brazilian environment sector from relatively low levels in the early 1990s. 4.40 The Bank’s strategy in the green area had some shortcomings but evolved significantly in recent years. An OED report published in 2000 noted that the Bank’s forest strategy in most o f the 1990s was excessively conservation-oriented and did not support production activities.” The report added that the forest strategy had not constrained Bank operations in Brazil, but the lack o f balance between conservation and development had precluded a more active role for the Bank in forest management and the design of sustainable development policies. The OED report argued that pro-active development policies may not maintain forests intact but may succeed in reducing deforestation, mitigating i t s impact and deviating i t s course to less sensitive areas (in biodiversity parameters). 4.41 The Bank seems to have embraced this new forest strategy. A recent Bank ESW proposes a strategy that includes the imposition o f taxes on non-sustainable logging and adequate credit schemes to reforestation in degraded areas.’l Along the same principles, the Government has recently launched the National Forest Program with Bank assistance and the Bank has also finalized the preparation o f a project on sustainable forests as part o f the National Forest Program with strong support from NGOs. Finally, in 2002 the Bank and the Government also launched an ambitious project o f strict conservation with support from GEF and G7 funds designed to protect an area equivalent to 10 percent of the Amazon region. This implies an increase in the total share o f protected areas in the national territory o f 2-3 percent (from 4.2 percent in 2000 to around 6.5-7 percent in 2002). 4.42 While Bank assistance under the green agenda has generally produced satisfactory results, it i s more difficult to evaluate the outcomes o f projects under the blue agenda (broadly defined to include water projects in both the water and the environment portfolios). As mentioned in the previous section, the projects addressing pollution o f urban watercourses have generally achieved satisfactory results, but the projects designed to build capacity in water management are too recent to be evaluated. What can be stated at this stage is that there i s s t i l l much progress to be made in developing and implementing a meaningful blue agenda. As mentioned before, access to sanitation services increased during the 1990s but s t i l l remains low, resulting frequently in the pollution o f urban river basins and problems of quality in water supply. The lack o f faster progress i s to a large extent due to factors outside the Bank’s control, such as delays in passing critical water legislation. However, there also seems to be scope for a more effective Bank assistance in this area, b y pursuing more l9This i s due mainly to the predominance of renewable energy sources, such as hydropower and ethanol, in Brazil’s energy matrix. OED (2000). *’Schneider et a1 (2001). 30 integrated strategies and through closer collaboration with the IDB, which i s also providing a substantial amount of assistance in this area. 4.4 Assessing the Overall Impact of the Assistance Program 4.43 Assessing the overall impact o f the Bank’s large assistance program in Brazil requires assessing i t s main blocks or components. A major component o f the program was designed to reduce poverty in the long-run through human resource development and access to basic services. This component produced satisfactory results. Education and health outcomes improved significantly, and the access o f the rural and urban poor to services was considerably expanded. The Bank can claim an important contribution to these outcomes, as i t was the only international development agency providing substantial assistance to federal and state governments in most o f these areas. 4.44 A second component o f the program was designed to stimulate private investment, economic efficiency, and growth. This component produced mixed results. The Bank’s strategy was reasonable and probably contributed to the efficiency gains observed in the decade, but did not succeed in stimulating savings and investment to a significant degree, and in removing some critical structural bottlenecks that continue affecting economic performance. The actions and reforms supported b y the program were all necessary, but they were not sufficient. In particular, the regulatory framework in infrastructure needs to be substantially strengthened and some key reforms need to be concluded in order to generate a sharp turnaround in growth performance. 4.45 A third component o f the program involved support to the environment. This component produced satisfactory results. The Bank played a critical role in the build-up of institutional capacity in the environment area, from relatively low levels in the early 199Os, helping the Brazilian government address many important challenges in the decade. Brazil’s environment indicators still lag those o f developed countries in many aspects but improved during the decade. The Bank’s strategy was excessively rigid in the early 1990s but evolved during the decade, recognizing more explicitly the need to balance conservation and growth. 4.46 The package o f adjustment loans merits a separate examination, given i t s size and the fact that i t was proposed for both poverty reduction and growth objectives. Assessing the package requires assessing different clusters o f loans, as they had an uneven impact on Brazil’s performance in the period o f evaluation. The state reform loans and the fiscal reform loans achieved their main objectives, which were essentially the reform o f the public sector and the imposition of hard budget constraints at all levels of government. These loans could not be expected to produce much greater improvements in the public finances, because this would require other specific and complementary reforms, such as tax and social security reforms. The social protection loan also achieved i t s well-defined objective, which was the preservation of critical social programs in a scenario o f fiscal adjustment. The social security, financial sector, and energy loans supported important reform steps, but these reforms started late in the decade and have not gone far enough, particularly the social security reform. 31 4.47 The Bank’s assistance program to Brazil in the 1990s produced a sufficient mass of positive outcomes to be rated as satisfactory. The program had only a moderate contribution to the decline in income poverty rates and the slight reduction in inequality observed in the 1990s. As discussed above, these gains were primarily due to macroeconomic stabilization. However, the program helped the Government achieve some o f the key pre-conditions for further and more substantive reductions in poverty and inequality. As also discussed above, the education (and health) gains should translate into reduced poverty and inequality because the income differences in Brazil are to a good extent related to differences in educational levels. The expanded access o f the poor to infrastructure not only alleviates poverty in a broad sense, but also creates the conditions for a subsequent growth in household incomes. 4.48 Admittedly, these gains w i l l need to be consolidated and deepened further for Brazil to achieve further reductions in poverty and inequality. For example, in the case o f education i t w i l l be necessary to raise years o f schooling further and improve education quality to obtain more substantive results. A large share o f the poor s t i l l lack access to basic services and assets, despite the gains in recent years, because access was initially very limited. Therefore, further efforts in this area are also needed. However, these are developmental problems, whose solution w i l l require more time and ersistence with the right policies, and that do not diminish the achievements o f the 1990s. 2 P 4.49 Further reductions in poverty and inequality w i l l also depend on an improved growth performance and better targeting o f social expenditures. In this regard, the need to complete some critical reforms such as social security reform cannot be emphasized enough. Brazil’s large pension payments are an obstacle to growth because they have contributed to low public and national savings and hindered a substantial investment recovery. They have also been a more direct obstacle to poverty reduction by diminishing the volume o f resources available to better targeted social programs. Finally, they have also affected the poor b y preventing a reduction in payroll tax rates and an increase in formal employment. Brazil’s social security problem constitutes a major obstacle to development and should remain a priority item in the Government’s and the Bank’s agenda in the current decade. 23 4.50 The effectiveness o f the Bank’s assistance can also be assessed b y examining Brazil’s progress in meeting the Millennium Development Goals (MDGs). As shown in Table 12, during the 1990s Brazil made substantial progress in all areas covered b y the MDGs, and i s likely to meet the targets at the national level b y 2015.24 The Bank contributed to many o f 22 As shown in Table 9 and Figure 3, poverty rates in 2000 were below the 1980 levels, but the reduction in inequality in the 1990s was too slight to offset the legacy o f the previous decades, especially the damage inflicted in the 1980s. Reducing inequality has proved more difficult than reducing poverty, not only in Brazil but in most other countries. Li, Squire and Zou (1998) show that there i s much more variability o f Gini coefficients across countries than within countries over time. 23 The Government made the comment that income distribution in Brazil i s very unequal and must improve, but also pointed out that World Bank comparisons o f inequality coefficients across countries are flawed and tend to worsen Brazil’s ranking relative to other countries (Annex 4). 24 In December 2002, the United Nations named President Cardoso the first winner o f the Mahbub ul Haq Award for Outstanding Contribution to Human Development, in recognition for the improvement in Brazil’s social indicators during his administration. The U N ’ s 2002 Human Development Report (UNDP (2002)) also indicates that Brazil i s on track to meet the MDGs, in all the areas where progress was assessed. Curiously, the areas that were not assessed are precisely those where progress was faster-the education and gender areas. 32 these outcomes, primarily through i t s assistance to education, health and the environment. Areas that require closer monitoring by the Government and that warrant Bank support include child nutrition and maternal mortality. Brazil’s large regional inequalities also imply the need to monitor the progress o f the poorer regions in meeting the MDGs. 4.5 Assessing the Sustainability of the Positive Outcomes 4.51 A critical question in country assistance evaluations i s whether the positive outcomes achieved over a decade are likely to be sustained over time. Assessment of sustainability requires examining, inter alia, whether the hard-conquered macroeconomic stability w i l l be consolidated by the new government, whether the infrastructure built and the programs 33 supported b y Bank projects w i l l be maintained, and whether the institutional capacity that has been built at the federal and sub-national governments w i l l be preserved. 4.52 Macroeconomic instability could disrupt many o f the positive outcomes o f the 1990s. For one, poverty rates could increase again and growth performance could weaken. Macroeconomic instability could also affect the sustainability o f outcomes at the sectoral level. Bank projects have financed a substantial volume o f infrastructure benefiting poor populations and have also contributed to a substantial buildup o f institutional capacity. Several programs supported by Bank projects have become regular Government programs andor been expanded. The Water and Sanitation for the Poor, Urban Basin Management, School Improvement, and Disease Control programs are just a few examples. These gains could be eroded in a scenario o f fiscal instability. Sectors such as education and health command priority and would probably be less affected, but sectors that are not considered a social priority such as the environment would be more vulnerable to funding cuts. 4.53 Sustainability o f the Bank’s results in the 1990s i s closely associated with the preservation o f macroeconomic stability, but not exclusively. I t i s also possible that the institutional capacity that has been built w i l l be partly eroded with the termination of Bank projects, even without budgetary cuts to the sector. This i s because in some sectors (e.g., the environment) Bank projects have played a critical role in the buildup o f technical and managerial expertise. Some qualified experts recruited with project resources may find i t unattractive to remain in the public sector after project closure, due in good part to uncompetitive salaries. However, this i s a problem that affects Bank projects worldwide, not only in Brazil, and in any case poses a lower risk for sustainability than the r i s k o f macroeconomic instability. 4.54 As mentioned in Section 1, in January 2003 the new Government expressed its commitment to macroeconomic stability and the targets agreed with the IMF, in the context of a standby o f US$30 billion approved in September 2002. In the first quarter o f 2003 the primary fiscal surplus had increased to around 6 percent o f GDP, well in excess o f the target of 3.75 percent o f GDP agreed with the I M F and the Government’s own target of 4.25 percent o f GDP. Also, in April 2003 the Government submitted ambitious pension and tax reform proposals to Congress. Market reaction to these announcements was positive, and b y M a y 2003 the pressures on the exchange rate and international spreads had subsided significantly-the rate o f the Brazilian Real vis-&vis the US Dollar had declined from 4 to 3 between October 2002 and May 2003, and international spreads had declined from 2,700 to 800 basis points in the same period. The exchange rate appreciation led to a decline in public debt to 52 percent o f GDP in M a y 2003, from a peak o f 63 percent o f GDP in late 2002. If these trends persist i t would be reasonable to rate sustainability as likely. 4.55 Unfortunately, even with the recent improvement in market sentiments Brazil’s macroeconomic vulnerabilities w i l l still persist for some time to come. Brazil’s main vulnerability lies in the large share o f the public debt indexed to the exchange rate and to short-term floating interest rates, which makes the debt service burden very sensitive to shifts in market perceptions and price movements. A negative shift in market perceptions due to external or internal events can have a substantial impact on debt service flows and impose severe financing constraints. At the time o f writing (May 2003)’ insufficient time had passed 34 for the new Government to establish a track record that would permit a conclusion that current trends w i l l be consolidated and that debt ratios w i l l be placed on a sustainable path. Therefore, macroeconomic stability and the sustainability of the outcomes achieved in the 1990s and early 2000s were considered to be non-evaluable.25.26 4.6 Assessing the Bank’s Performance 4.56 Assessing the Bank’s performance differs from assessing the program’s performance because there are situations where the Bank’s program produces limited outcomes despite the quality o f the assistance, due to political limitations or other exogenous factors. The Bank had overall a good performance during the period of evaluation. The strategy that evolved in the 1990s was underpinned b y good quality ESW, consistent with the Government’s agenda, the country’s development needs, and the Bank’s own comparative advantages. The institution was able to learn the lessons from the early 1990s and to generally improve i t s modus operandi during the decade. The decentralization of the CMU and the build-up of technical expertise in the field were important factors contributing to the improvement in outcomes. The decentralization also seems to have enabled the Bank to manage a complex program in a cost-effective way (Section 3). 4.57 The Bank cannot be faulted for the lack of completion o f some key reforms. When the 1998 crisis created the conditions for the implementation o f structural reforms the institution responded quickly with financial and technical assistance. I t was difficult to predict the extent and the speed with which different reforms would progress, and when some reforms stalled the Bank stopped adjustment lending while maintainingtechnical assistance. What can be criticized i s the lack o f a clear assessment o f progress and of the measures required to complete the reform in some of the adjustment loans. 4.58 Likewise, the Bank should not be faulted for the energy crisis in the late 1990s. The institution had reasonable reasons to stop lending to the sector in the early 1990s, because the policy o f low energy tariffs resulted in weak financial results and low rates o f return in energy projects. The lack o f investment in the sector dates back to the 1980s and has been to a good extent due to the low tariffs and the weak capacity of the sector to generate funds internally. When the Government started implementing i t s restructuring and privatization program in the mid-1990s the Bank responded quickly b y restructuring an existing project and releasing funds for the development o f the regulatory framework. At this point there was probably little rationale to resume large scale lending to a sector which was being privatized. I t i s possible that the Bank did not provide sufficient technical assistance to the sector in the second half of the 1990s, but i t would be unfair to fault the institution for the crisis. 25 Brazil’s macroeconomic fundamentals are s t i l l not sufficiently strong to rule out scenarios o f instability. They are in fact in an intermediate range, giving rise to situations o f multiple equilibria, where market psychology dominates the final outcomes. Williamson (2002) presents a numerical exercise of multiple equilibria applied to Brazil and Williamson (2003) provides an updated evaluation o f Brazil’s current situation. 26 The Government expressed its disagreement with the CAE conclusion that sustainability could not be evaluated, stressing the positive evolution o f economic indicators and the submission o f key reforms to Congress (Annex 4). 35 4.59 Some Bank policies created difficult situations that had to be carefully managed b y the C M U in Brazil. For example, during the elaboration o f the assistance strategy in 2000, the Bank had to handle a delicate situation, resulting from i t s intention to discuss the assistance strategy with NGOs. The Government indicated that i t supported NGO involvement in project implementation but not in the formulation o f assistance strategies. The Government also pointed out that strategy formulation was the responsibility of democratically elected governments, and that the Government’s program was already strenuously negotiated within the Government coalition and in Congress. Bank attempts to introduce other players in the formulation of the strategy could increase enormously the complexity o f the exercise, contradict Government’s priorities, and even decrease the legitimacy o f the strategy. 4.60 As i t happened, the institution had separate consultations with NGOs and other segments o f civil society to discuss their views on key developmental issues. In general, the consultations did not reveal major objections against the main thrust of the Bank’s assistance program. Some NGOs voiced objections to globalization, the Government’s privatization program, and the effectiveness o f regulatory schemes. The Bank considered these objections, but concluded that they did not justify any fundamental revision o f the strategy. The situation was ultimately handled b y the Bank in a satisfactory way, but i t also exposed the potential conflicts that exist among some o f the Bank’s objectives (e.g. Government ownership, stakeholder participation, promotion o f NGO activity). 4.61 Although the Bank’s performance in the decade was generally good, i t i s also important to identify areas where Bank failed to provide adequate assistance. For example, the Bank failed to provide assistance in critical areas such as early childhood development and nutrition. The Bank may not have provided enough assistance to the development of regulatory agencies either, especially at the level o f Brazilian states. Some projects did not provide reliable estimates o f project benefits for the poor, a clear weakness, considering the poverty reduction objectives o f the program. Also, there was not sufficient integration between related projects and important synergies may have been missed. More generally, monitoring and evaluation o f projects was less than satisfactory-the Bank should be able to ascertain more systematically the impact o f i t s assistance on the poor and on the efficiency of the Brazilian economy. 4.62 There are issues o f a global nature that are more difficult to evaluate in a country assistance evaluation. Brazil’s economic performance was to some extent affected b y trade barriers in i t s main export markets. Multilateral trade relations are conducted in other international fora and the Bank has just a modest influence over the results o f trade negotiations. However, i t can be argued that the Bank could have been more vocal during the 199Os, advocating greater access o f developing countries, including Brazil, to the markets of developed countries. This issue had already been raised in the mid-1970s by former President McNamara*’ but was not pursued b y the institution in the following decades, until very recently. *’ Address to the Board of Governors in the 1976 annual meetings in Manila, the Philippines. 36 4.7 The Counterfactualand the Bank’s Relevance 4.63 The relevance of the Bank has been challenged in more than one occasion, particularly regarding i t s assistance to middle-income countries such as Brazil. Critics usually point out that these countries have access to private financial flows, that the links between Bank financing and project performance are weak due to money fungibility, and that the recent decentralization of Bank operations has intensified the problem o f duplication and conflict with regional development banks2* 4.64 This evaluation concluded that these criticisms were not justified in the case of Brazil. The Bank did become less important as a source of finance during the 1990s, as Brazil gained greater access to private capital markets. However, the Brazilian authorities stated that Bank resources were still an important element o f the Government’s financing strategy, particularly in periods o f volatile capital markets, curtailed access, and high spreads. Bank adjustment loans in the late 1990s and early 2000s helped the Government build a reserve cushion and manage the timing o f their international bond issues more effectively, yielding important savings for the country. 4.65 The criticism that the link between Bank financing and projects i s weak disregards the transfer o f knowledge and technical assistance that takes place through Bank projects. The Bank played an important role in the transfer o f knowledge and institutional building in Brazil during the 1990s. Although this transfer occurred in part through policy dialogue and ESW, a large share took place during the design and execution of projects. The Bank’s contribution seems to have been particularly important in the Brazilian states, which have a much weaker institutional capacity than the federal government while being responsible for delivering basic services to the population. 4.66 The criticism that the expansion of regional offices would cause duplication and conflict with regional banks-the IDB in the Brazilian case-is not supported and overlooks the gains from a greater presence in the field. Bank and IDB programs were highly complementary, as noted in Section 3. Although the Bank and the IDB could both gain from closer cooperation, Brazil’s large size and development needs open a wide room for the two institutions to operate without conflicts. Also, greater Bank presence in the field produced important improvements in project design and supervision. 4.67 This evaluation concluded that the Bank can s t i l l play a relevant role in a middle income country such as Brazil. The Bank is most relevant as a source o f knowledge and technical assistance. I t i s less relevant as a source o f finance but not irrelevant, particularly for poorer regions and in a scenario of volatile capital markets. Besides, financing and transfer of knowledge frequently come together. Substantial learning and experience accumulate during the life of a project, both on the side o f the Government and the Bank, and this experience can be capitalized in future Bank projects and Government programs. The evolution in the design o f education projects i s a good illustrative example. *’ For example, the well-disseminatedMeltzer report (US Congress (2000)) raises these and other criticisms. 37 4.68 The assessment of a counterfactual scenario (how Brazil would have performed without the Bank) and the assessment of relevance are closely related. Brazil has a substantial amount of human resources but they are not evenly distributed across regions and are generally scarcer in the North and the Northeast. The country could eventually have achieved the same results without Bank assistance, but i t would have taken longer to achieve these results. Several Government programs were started, accelerated, or expanded due to Bank assistance. Also, the technical assistance loans that accompanied the adjustment loans provided knowledge at a pace that would not have occurred otherwise. For example, the implementation o f the inflation targeting system by the Central Bank in a very short period of time (the first half o f 1999) was only possible due to the intensive use o f a technical assistance loan. The blending of technical assistance and finance provided by the Bank was effective and probably would not have been provided by any other institution or bilateral donor in the Bank’s absence. 5. Lessons and Recommendations 5.1 The greatest lesson learned from a decade o f assistance was the need to ensure macroeconomic stability and Government ownership o f the Bank’s assistance program. The unsatisfactory project outcomes and the low effectiveness of the Bank’s assistance in the early 1990s reflected fundamentally the lack o f a stable macroeconomic framework and a clear development agenda. The Bank’s central objectives o f poverty alleviation and growth cannot be achieved under these circumstances. Other major lessons include the need for the Bank to build technical capacity in the field, maintain an open and frequent policy dialogue with federal and state authorities, employ the resources of the local academic community, and have the ability to closely supervise i t s projects. 5.2 The Bank can continue playing a relevant role in Brazil in the coming years, b y helping the Government consolidate the important achievements o f the 199Os, correcting some flaws in the assistance, and entering critical areas that remain largely uncovered. The following recommendations address the quality o f the Bank’s assistance. They include recommendations o f a general and strategic nature and sector-specific recommendations. 5.3 Strategic Direction o f Bank Assistance. I t is very difficult to determine the optimal m i x o f human resource development projects and other projects (e.g. infrastructure) in the Bank’s assistance program because the empirical literature on poverty and growth does not provide precise parametric information on the impact o f all relevant variables. However, there i s a strong case for continuing Bank involvement in the social sectors, particularly education. Primary education has been universalized but enrollment in secondary education i s s t i l l low, and quality o f education i s deficient on both levels, particularly in the Northeast. The Bank can play an important role in this area, assisting the federal and state governments to address the problem o f student learning, particularly in the poorer regions of the country. 5.4 Zmproved Framework for Programmatic Lending. There i s justification for further programmatic lending in coming years to help the country complete critical reforms, such as the social security and financial sector reforms. However, programmatic lending requires discipline in the setting o f the overall reform program, i t s final objectives and the 38 intermediary benchmarks. For example, the second social security loan supported improvements in the pay-as-you-go (PAYG) system for private workers but did not stress sufficiently that the imbalances of the pension system remained large and s t i l l constituted a major obstacle to Brazil’s development, and that addressing the pension problem would require a more ambitious reform. This loan was not technically a programmatic loan (this instrument had not been approved b y the Board yet), but had similar characteristics and should have followed similar rules. Agreement with the Government on final reform goals and meaningful benchmarks i s essential to maintain programmatic lending on track. 5.5 Building the Capacity of Sub-National Govemments. At the sub-national government level, public finance management and administration appear weak, especially in relation to the federal government. Possible Bank assistance to states and municipalities might reap high returns. For example, during the early 1990s the Bank helped Southern states such as Paranfi strengthen the administrative capacity o f municipal governments, enabling municipalities to manage their finances and meet their expanded obligations under the 1988 Constitution (through a state-wide program of technical assistance, training, and equipment). The Bank should explore the possibility of providing a similar type o f assistance to Northeastern states. 5.6 Further Support to Private Sector Development. There i s also scope for improving Bank support to private sector development. The Bank might consider assistance to judicial reform, as the judiciary has been identified as a bottleneck in private sector development ~~ strategies produced b y the Bank and d ~ m e s t i c a l l y .Protracted judicial procedures, unpredictable outcomes, weak enforceability o f contracts and rights are frequently identified as important constraints. The Bank i s already providing some initial assistance through ESW. Depending on the results and the interest o f the Government and the Judiciary, the assistance could evolve into an operation. More assistance to regulatory agencies i s also needed, especially at the level of states, in order to reduce the lingering uncertainties and encourage private sector investment. 5.7 Design of More Zntegrated Strategies. There i s scope for producing more integrated strategies and explore better the potential synergies across projects, b y changing the focus o f assistance from sectors to regions. The move towards more integrated projects may imply some trade-offs, such as the reduction in the geographical areas covered b y the projects. However, the gains in project efficiency and sustainability may offset the possible losses. The Bank seems already to be proceeding in this direction in the state o f to cant in^.^' 5.8 Regarding the design and objectives of new projects, there i s a need to review the outcomes of the 1990s and identify the changes that are needed. The following paragraphs provide the main recommendations for improved project design. More detailed recommendations are provided in Annex 2 and in the background papers prepared for each sector. *’ Pinheiro, A. (2001). 30 The Prosanear projects provide a good illustration. These projects were able to extend water and sanitation to a large poor population at a low cost, but maintenance of the infrastructure became an issue, due to the lack of parallel investments in paved streets and electricity. A more integrated urban development project could have had to sacrifice geographical coverage, but could also have gained in sustainability. 39 5.9 Education. The Bank must change the emphasis of basic education projects, from improving management to improving quality o f teaching. Operationally, this should be achieved b y moving from system and efficiency goals to classroom and learning goals. The priority for Bank assistance to primary education should be to get inside the classroom in order to ensure better learning outcomes. This change will require a major role for technical assistance and supervision, to ensure the quality leap that i s required to tackle student learning. Many poor states in the Northeast may need more focused assistance, and several o f these states may not be able to borrow from the Bank due to fiscal constraints. Small and focused technical assistance loans may constitute one alternative to deal with specific learning problems faced b y these states (e.g., teacher training and effectiveness). The Bank must also consider entering two areas that remain largely uncovered: early childhood developmentlpre-school and secondary education. 5.10 Health. The Bank has been only partially successful in targeting the poor and offering effective solutions for their problems vis-&vis health care. Focus on diseases that by nature affect the poor i s an effective manner of targeting, but i s not sufficient. I t i s necessary to design projects to deal with the structure of the public system and that address specifically the barriers o f the poor. These issues could be addressed through adjustment lending. The Bank must also consider a pilot project addressing child nutrition in the Northeast, preferably integrated with an early childhood development project. Another area to be considered i s that o f maternal mortality, due to i t s high rate and devastating impact. 5.1 1 Rural Development. The rural development projects have made an important contribution to alleviating poverty and providing the infrastructure on which agricultural growth can take place. However, there may now be a need for a more focused effort to raise farm productivity in areas of higher agricultural potential. This could be achieved through a combination o f different approaches: (i) increasing the share o f resources in rural development projects channeled to productive activities, including mechanization, commercialization, technical assistance, and training o f community councils; (ii) land reform can also contribute to higher agricultural productivity. The pilot land reform project has built on the design o f community-based projects and shows promising results; ( ii i) Despite past failures, irrigation may also contribute to an increase in productivity. The Bank could pursue a limited program, focusing on areas with good soils, access to markets and sound state finances. 5.12 Water and Sanitation. The Bank should explore i t s potential educational role in this sector. The key issue of which government level - state or municipal - has the jurisdiction to make concession o f water and sanitation services has not been overcome. Studies on asset valuation o f current water and sanitation infrastructure and compensatory schemes for the transfer o f concession power could contribute to a satisfactory resolution o f this problem. Regarding privatization, an issue that i s highly misunderstood, studies should cover themes such as universality o f services, economies o f scale, partial against total concession, cross- subsidies, and regulation o f natural monopolies. Bank strategies towards the water and sanitation sector must be more integrated with the Bank’s environmental and poverty strategies. As a starting point, a common strategy plan should be prepared with the collaboration o f the three areas. In this regard, there should be a greater effort to estimate the share o f the poor benefiting from water and sanitation projects. 40 5.13 Environment. Regarding analytical work, the Bank should undertake a sound analysis o f the private and social costs and benefits of deforestation, as well as studies integrating blue and brown issues and i t s activities in the water and sanitation sector. Regarding project work, the Bank should adopt more realistic criteria for the determination o f the regional coverage and scale o f projects, in relation to the institutional capacity of the implementing agencies. The Bank should also engage representatives o f the private sector (timber producers, industries and farmers) in project formulation and implementation, forging additional sources o f financing and risk sharing, and contributing to the political sustainability o f i t s projects. 5.14 Better Dissemination of ESW and Bank Activities in Brazil. Last but not least, the Bank’s relevance as a source o f knowledge and technical assistance may not be sufficiently explored. Greater dissemination o f its work may help the Bank enhance i t s credibility and provide more effective assistance to the Government in the design and implementation o f complex reforms. The Bank should consider allocating a larger share o f i t s budget in Brazil for translation, organization o f workshops, distance training, and other activities designed to disseminate i t s work. ANNEX 1 STATISTICAL TABLES 43 Annex Table 1: World Bank Lending:to Brazil. 1980-2001 Average Annual Flows I 80-89 90-01 90-94 95-98 99-01 US$ millions IBRD Commitments 1,221 1,143 1,001 1,044 1,512 IBRD Disbursements 894 1,100 663 1,254 1,621 IFC Disbursements 57 153 106 232 127 IFC Disbursements + IFC "B" loans 93 381 154 745 274 80-89 90-01 90-94 95-98 99-01 Shares in Total Debt Finance (YO) [BRD 16.5 6.4 10.2 3.6 3.9 [BRD + IFC 17.4 7.6 11.9 4.2 4.2 [BRD + IFC + IFC "B" Loans 17.7 8.3 12.4 5.6 4.5 Relative to Other Institutions (YO) [BRD/BNDES 24.2 15.8 20.2 10.0 16.4 71.8 64.8 280.7 1989 2001 1994 1998 2001 US$ millions [BRD Disbursed 8,316 7,963 6,3 11 6,298 7,963 [BRD Undisbursed 4,672 2,770 4,950 3,546 2,770 [BRD Total 12,988 10,733 11,261 9,844 10,733 Shares in Total Stocks (%) [BRD BraziVTotalBrazil Stock 7.3 3.4 4.1 2.6 3.4 [BRD BraziVIBRD Worldwide 9.9 6.1 5.7 5.4 6.1 44 45 46 1980-89 1990-2002 IBRD IFC IBRD IFC MIGA" US$ million # JSS million # JS$ million # JS$ million # IS$ million # Agriculture 3870.3 36 54.1 6 2,047.2 31 400.2 15 Education 238.5 5 1,798.7 12 3.3 1 Envirmnt 108.0 2 958.0 14 Finc & PSD 1,518.4 5 88.5 46 1,362.6 7 1069.0 56 770.5 Health 338.0 5 1,060.0 6 40.3 3 150.5 Infrastructure 5,179.7 28 0.4 1 4,395.8 28 235.9 11 12,297.2 Telecom 106.1 5 4,005.2 Transport 1,547.0 10 2,533.1 16 99.8 5 Urban 449.0 3 1,130.1 9 0 Other 1,098.0 7 1,403.0 7 99.8 5 Energy 2,505.1 11 0.4 1 888.4 5 30 3 8,292.0 W&S 1,127.6 7 974.3 6 PSM 29.0 1 1,691.5 6 Social Prot 0.0 0 1,525.2 5 -- -- Urban Dvlp 640.0 8 4911n 4 ~ 11,921.9 90 143.0 53 15,329.0 113 1,748.7 97 23 r TOTALS % IBI D I 11.3% 13.4% IBRD 27.4% - 22.9% IF - 15.5% Agriculture 32.5% 40.0% 37.8% Education 2.0% 5.6% 11.7% 10.6% 0.2% 1.O% Envirmnt 0.9% 2.2% 6.2% 12.4% Finc & PSD 12.7% 5.6% 61.9% 86.8% 8.9% 6.2% 61.1% 57.7% 5.8% 52.2% Health 2.8% 5.6% 6.9% 5.3% 2.3% 3.1% 1.1% 4.3% Infrastructure 43.4% 31.1% 0.3% 1.9% 28.7% 24.8% 13.5% 11.3% 93.0% 43.5% Telecom 6.1% 5.2% 30.3% 13.0% Transport 13 .O% 11.1% 16.5% 14.2% 5.7% 5.2% Urban 3.8% 3.3% 7.4% # 8.0% Other 9.2% 7.8% 9.2% 6.2% 5.7% 5.2% Energy 21.0% 12.2% 5.8% 4.4% 1.7% 3.1% 62.7% 30.4% W&S 9.5% 7.8% 6.4% 5.3% PSM 0.2% 1.1% 11.0% 5.3% Social Prot 0.0% 0.0% 9.9% 4.4% - -- - - - Urban Dvlp 5.4% 8.9% 3.2% 3.5% TOTALS 100% 100% 100% 100% 100% 100% 100% 100% 47 Annex Tal - - 1980 189 1990-: US$million % # % US$ million % # % Lgriculture 328.4 11.7% 8 15.4% 130.8 0.8% 2 2.4% :ducation 218.2 7.8% 4 7.7% 862.0 5.3% 5 6.1% hvironment 701.9 4.3% 6 7.3% 'inc & PSD 397.4 14.1% 14 26.9% 3,436.0 21.0% 8 9.8% Iealth 535.0 3.3% 2 2.4% nfrastructure 1,758.8 62.6% 24 46.2% 6,094.7 37.2% 41 50.0% Transport 715.5 25.5% 10 19.2% 3,090.5 18.9% 21 25.6% Urban 560.6 3.4% 4 4.9% Other 715.5 25.5% .lo 19.2% 2,529.9 15.5% 17 20.7% Energy 592.19 21.1% 9 17.3% 818.8 5.0% 9 11.O% W&S 406.61 14.5% 3 5.8% 2,185.4 13.3% 11 13.4% General 44.51 2 3.8% 0.0% 'SM 1,083.0 6.6% 8 9.8% ocial Prot 2,242.0 13.7% 2 2.4% -- 107.5 3.8% 3.8% 1,285.5 7.9% 8 9.8% - -- Jrban Dvlp 'OTALS 2,810.4 100% 52 100% 16,370.9 100% 82 100% 48 I Annex Table 5.A - OED Ratings of Closed Proiects 1980-89 Net Commitments - by net commitments World LCR -- - Brazil World LCR Brazi -- Brazil Agriculture 21,677 4,553 905 58 31 30 6 18 A . LI 20 49 Education 3,241 468 84 87 85 75 66 80 100 42 46 51 Electric Power & Other Energy 10,701 3,479 1,514 75 63 44 69 62 21 40 13 0 Environment 80 34 34 100 0 0 100 100 100 100 100 100 Finance 6,000 1,571 251 73 65 12 41 27 0 34 50 0 Health Nutrition & Population 475 76 31 57 43 0 65 0 0 31 0 0 Industry 4,405 896 464 70 57 90 33 0 0 .25 0 0 Mining 638 346 239 37 17 0 73 84 100 40 0 0 Multisector 8,903 3,024 351 66 78 0 55 74 0 37 44 0 Oil & Gas 2,688 295 0 88 56 0 65 53 0 38 39 0 Private Sector Development 782 23 1 0 71 37 0 45 0 0 5 0 0 Public Sector Management 297 110 0 63 81 0 30 40 0 16 27 0 Social Protection 10 10 0 100 100 0 100 100 0 100 100 0 Telecommuiications 1,640 249 0 94 100 0 98 100 0 18 43 0 Transportation 11,132 3,260 1,169 74 56 62 54 48 69 27 19 21 Urban Develooment 3,195 1,772 319 62 44 100 45 26 98 20 17 54 Water Supply & Sanitation 3,764 --- - - - - - -- 1,264 678 63 49 38 40 43 43 37 24 43 Total 79,629 --- - - - - - -- 21,639 6,038 69 55 47 48 42 39 30 26 21 -- - - - - - - - - - - Adjustment 11,256 3,810 653 58 61 0 42 57 0 31 48 0 Investment 68,374 17,828 5,385 70 53 52 49 37 47 29 20 23 by # o f projects # o f Projects % LIK Agriculture World 689 -- Brazil LCR 98 - 20 59 - 56 - - 45 World 34 I LCR 18 - Braz 36 22 I 22 50 Education 180 36 3 80 79 67 58 69 100 30 46 50 Electric Power & Other Energy 172 53 12 79 73 67 68 74 50 31 19 0 Environment 4 1 1 100 0 0 100 100 100 100 100 100 Finance 159 34 2 69 68 50 34 26 0 31 42 0 Health Nutrition & Population 28 7 2 54 43 0 42 0 0 25 0 0 Industry 85 14 7 63 62 86 29 0 0 26 0 0 Mining 25 8 1 63 50 0 38 25 100 36 0 0 Multisector 82 24 1 63 61 0 35 44 0 22 33 0 Oil & Gas 63 11 0 89 64 0 57 45 0 43 36 0 Private Sector Development 25 5 0 76 60 0 47 0 0 7 0 0 Public Sector Management 47 15 0 57 60 0 20 20 0 17 25 0 Social Protection 1 1 0 100 100 0 100 100 0 100 100 0 Telecommunications 46 9 0 87 100 0 86 100 0 33 67 0 Transportation 298 68 11 72 53 73 41 34 80 19 13 20 Urban Development 87 31 5 79 74 100 59 56 75 24 33 50 Water Supply & Sanitation 130 --- - - 34 7 72 59 43 35 33 -33 33 Total 2,121 --- - - 449 72 69 63 59 43 39 -51 34 -- - - - - - - Adjustment 98 28 2 60 52 0 33 35 0 0 Investment 2,023 42 1 70 69 64 61 43 40 55 26 I 24 35 49 I Annex Table 5.B: OE D Ratings of Closed Projects 1990-02 N e t Commitments % SAT % LIK %SUB by net commitments World LCR Brazil Agriculture 32,457 6,859 3,364 74 77 69 54 64 63 42 49 41 Economy Policy 20,213 1,977 0 83 88 0 68 14 0 42 46 0 Education 11,632 2,490 978 80 76 98 64 73 81 43 37 28 Electric Power & Other Energy 23,202 4,356 1,240 67 37 4 63 58 49 41 17 4 Environment 1,417 56 1 337 74 84 89 65 91 87 41 51 62 Finance 21,873 6,585 483 14 69 0 68 59 0 44 50 0 Health Nutrition & Population 5,602 1,915 647 79 87 91 66 70 91 41 56 49 Industry 9,606 2,656 0 65 80 0 62 85 0 33 21 0 Mining 2,097 592 0 86 99 0 75 81 0 58 81 0 Multisector 12,222 4,410 0 81 86 0 62 79 0 46 47 0 Oil & Gas 6,148 1,408 664 70 78 100 75 91 100 30 48 61 Private Sector Development 4,128 1,945 170 64 74 100 57 66 0 45 61 0 Public Sector Management 10,029 4,023 834 86 91 97 71 85 85 52 61 85 Social Protectioii 6,632 3,693 1,515 86 99 100 79 78 100 42 55 50 Teleconnnunications 2,497 49 0 90 45 0 91 100 0 65 0 0 Transportation 23,626 6,693 1,418 86 91 100 68 61 47 50 64 59 Urban Development 9,798 3.191 516 82 1 85 I 79 1 59 1 80 1 47 I 29 1 27 I 42 Water Supply & Sanitation 7,945 2,349 1,056 58 56 70 32 35 44 22 20 20 Total 211,125 55,754 13,222 77 78 75 63 69 66 42 46 42 -- Adjustment 71,771 21.278 2.502 80 85 100 70 78 100 44 53 50 I Investment 139,353 34,476 10,720 75 I 74 1 69 I 60 I 64 I 58 I 41 I 41 ~ I ~ 40 # o f Projects % SAT Yo LIK %SUB by # o f projects World LCR Brazil World L C R Brazil World LCR Brazil World LCR Brazil Agriculture 657 I 96 I 32 64 I 68 I 72 41 I 50 I 66 33 I 63 1 I I 1 I I IElonomy Policy 134 16 0 1 Education 239 42 9 :I : IElecmc Power& OtlierEnergy 233 41 6 IEnvironment L i c e Health Nutrition & Populatioti I Pi3 I 6 2 6:’ 67 83 0 83 65 52 52 87 63 55 83 0 83 47 38 29 47 43 39 50 0 50 I llndusav 59 78 0 53 78 0 32 22 0 Millillg 71 88 0 62 75 0 62 75 0 Multisector 71 88 0 46 68 0 27 42 0 Oil & Gas 74 92 100 58 83 , 100 49 50 67 Private Sector Development 55 18 1 46 53 0 43 56 0 26 29 0 Public Sector Management 179 44 4 69 86 75 58 72 50 40 53 50 Social Protectlon 79 25 3 86 96 100 53 52 I 100 I 47 56 67 Telecommunications 46 3 0 0 Transportation 301 55 10 90 Urban Development 148 32 6 83 Water Supply & Sanitation 128 29 6 60 48 67 38 43 50 30 25 33 Total 2815 537 95 69 73 73 51 60 63 36 41 49 Adjustment 397 87 5 76 86 100 60 71 100 38 52 60 Investment 2418 450 90 68 I 71 I 72 1 50 I 58 1 61 I 36 I 38 1 48 I Annex Table 5.C: OED Ratings of Closed Proiects, 199 1-94 by net commitments I Net Commitments I % SAT I % LIK %SUB World LCR Brazil World LCR Brazil World LCR Brazil World LCR Brazil Agriculmre 15279 I 3652 1637 70 77 58 48 65 46 36 42 3 Economy Policy 2768 1179 0 88 89 0 53 89 0 46 51 0 Education 2873 360 116 90 96 100 79 82 74 63 78 62 Electric Power & Other Energy 10062 2651 1240 64 15 4 71 66 49 34 3 4 Finance 5683 2013 483 56 59 0 50 53 0 38 40 0 Health Nutrition & Populatioii 666 74 57 76 0 0 57 0 0 31 0 0 Industry 6455 2085 0 72 74 0 68 81 0 32 12 0 Mining 572 111 0 55 100 0 68 6 0 52 6 0 Multisector 10012 3818 0 78 84 0 64 79 0 49 45 0 Oil & Gas 2050 575 0 80 45 0 92 97 0 42 43 0 Private Sector Development 1153 773 0 77 92 0 52 70 0 43 64 0 Public Sector Management 2165 956 24 68 71 0 53 61 0 62 70 0 Social Protection 71 47 0 94 100 0 66 100 0 94 100 0 Teleconlmuiications 671 27 0 75 0 0 78 100 0 37 0 0 Transportation 8478 2422 347 80 88 99 58 56 0 32 51 0 1~ Urban Develoument 3663 1307 132 77 71 18 53 82 18 31 34 0 Water Su 1 & Sanitatioii 2044 553 313 72 37 0 28 18 0 29 18 0 74665 22605 4349 73 69 34 59 69 34 39 38 4 Adjustment 25198 10798 482 73 82 100 62 82 100 45 49 0 I 49467 I 11807 I 3Xh7 I 72 5R 2h I 58 57 26 I 36 26 5 # o f Projects % SAT % LIK %SUB by # ofprojects World LCR Brazil World LCR Brazil World LCR Brazil World LCR Brazil Agriculture 323 I 41 I 10 60 61 30 35 39 20 29 27 11 Economy Policy 14 4 0 79 75 0 54 75 0 31 50 0 Education 83 14 3 80 77 100 63 54 67 38 46 33 Electric Power & Other Energy Ill 24 6 64 52 17 57 67 50 30 24 17 Finance 64 18 2 56 65 0 40 59 0 28 35 0 Health Nutrition & Population 29 2 1 69 0 0 45 0 0 17 0 0 Industry 56 7 0 63 71 0 55 71 0 34 14 0 Mining 13 2 0 58 100 0 54 50 0 46 50 0 Multisector 67 19 0 67 84 0 50 63 0 31 33 0 Oil & Gas 36 4 0 75 75 0 50 75 0 61 50 0 Private Sector Development 13 4 0 46 75 0 23 50 0 8 25 0 Public Sector Management 52 13 1 49 77 0 33 62 0 27 62 0 Social Protection 4 2 1 0 75 100 0 50 100 0 , 75 100 0 Telecommunications 17 2 0 76 50 0 76 50 0 41 0 0 Transportation 129 22 72 73 67 51 45 0 29 32 0 Urban Develoument 55 11 75 45 50 38 45 50 24 18 0 Water Supply & Sanitation 47 9 2 70 33 0 33 25 0 36 25 0 Total 1113 198 30 65 65 32 45 53 26 31 33 10 Adjustment 141 42 1 67 80 100 52 71 100 34 44 0 Investment I 972 I 156 I 29 I 65 61 30 I 44 48 24 I 31 29 10 51 7 Annex Table 5.D: OED Ratings of Closed Proiects, 1995-02 I b y net commitments Net Convnitments % SAT Yo LIK YOSLB World LCR Brazil World LCR Brazil World LCR Brazil World LCR Brazil Agriculture 17,178 I 3,207 I 1,727 78 77 79 58 63 79 48 56 83 IEionomy Policy I 17,444 I 799 I 0 82 87 0 70 52 0 41 39 0 Education 8,759 2,130 862 77 73 98 59 72 82 36 30 24 Electric Power & Other Energy 13,140 1,705 0 70 70 0 56 44 0 46 39 0 Environment 1,417 56 1 337 74 84 89 65 91 87 41 51 62 Finance 16,190 4,572 0 80 74 0 75 62 0 47 54 0 Health Nutrition & Population 4,936 1,841 590 80 90 100 68 73 100 42 59 53 Industry 3,151 571 0 52 100 0 50 100 0 34 56 0 Mining 1,525 48 1 0 98 99 0 78 99 0 60 99 0 Multisector 2,210 592 0 94 100 0 53 81 0 29 58 0 Oil & Gas 4,098 833 664 64 100 100 67 86 100 25 51 61 Private Sector Development 2,975 1,171 170 59 60 0 59 63 0 45 58 0 Public Sector ,Management 7,864 3,067 810 90 97 100 76 93 87 49 58 87 Social Protectioii 6,561 3,646 1,515 86 99 100 79 78 100 41 55 50 1 Telecommunications 1,826 22 0 95 100 0 96 100 0 75 0 0 Transportation 15,149 4,272 1,071 90 94 100 73 64 63 60 72 78 Urban Development 6,135 1,883 384 85 95 100 63 79 56 28 22 56 5,901 ~ 1,796 743 53 62 100 34 40 62 20 20 29 136,460 33,149 8,873 79 85 93 66 70 81 44 51 60 Adjustment 46,573 10,480 2,020 84 89 100 74 74 100 43 56 63 Investment 89,887 22,669 6,853 76 83 91 61 67 75 44 49 59 % SAT % LIK %SUB by # o f projects World LCR Brazil World LCR Brazil World LCR Brazil Agriculture 68 74 91 46 58 86 38 48 86 Economy Policy 120 12 0 81 90 0 64 60 0 32 44 0 Education 156 28 6 76 79 83 51 64 67 32 29 33 Electric Power & Other Energy 122 17 0 64 71 0 51 65 0 40 47 0 Environment 43 15 6 72 73 83 65 87 83 47 47 50 Finance 99 26 0 67 76 0 60 67 0 44 48 0 Health Nutrition & Populatioii 100 21 5 66 76 100 54 60 100 32 43 60 Industry 18 2 0 50 100 0 44 100 0 28 50 0 Mining 16 6 0 81 83 0 69 83 0 75 83 0 Multisector 33 6 0 79 100 0 38 83 0 20 67 0 O i l & Gas 42 8 3 74 100 100 66 88 100 38 50 67 Private Sector Development 42 14 1 46 46 0 50 58 0 32 31 0 Public Sector Management 127 31 3 78 90 100 69 77 67 46 50 67 Social Protection 75 23 3 87 96 100 53 48 100 45 52 67 Telecommunications 29 1 0 86 100 0 83 100 0 55 0 0 Transportation 172 33 7 86 81 100 67 61 71 56 58 71 Urban Develoomeiit 93 21 4 71 81 100 48 57 75 32 33 75 Water Supply & Sanitation 81 20 4 55 55 100 41 50 75 26 25 50 Total 1,702 339 64 72 78 92 55 64 81 39 45 67 Adjustment 256 45 4 82 91 100 65 70 100 40 60 75 Investment I 1,446 I 294 I 60 I 70 76 91 I 53 63 80 I 39 43 66 52 d rf - I FA FA V I V I y V I - - w d m 6 6 m z z g 4 C t-* & cc 7 VI FA hVI r- rf W h FA N O W d d b V rf 0 0 0 Q 64- N U FA 0- O b C F Am W 64 4 O N V 6 4 0 W \c FA 7 g 2 - FA c Q g 2 - 64 cc 5 g 2 N h h 64 7 O N C 64w N 64 'c w w e ~ ~e ~ m i m ~ ~ m ~ a m- r - o o o o a m : m ~ m o r o m~ : m c e m w e e m e m m " " m m . 54 ffi d - m a W W N 1 m2 x 0 m N r- W 0 2 0 2 2 2 N - o m m N W m r- m m 00 r - m - a m d N a N N N m 55 SectodReport T i t l e Date Report # Country Assistance Strategies Brazil - 2002 Country assistance strategy progress report (V01.1) 512012002 24 182 Brazil - 2001 Country assistance strategy progress report (Vol. 1) 51112001 22116 ~ Country Report Bahia State Economic Memorandum 1012612001 21377 Ceara State Economic Memorandum 812 112000 19217 Agriculture Rural poverty alleviation in Brazil: towards an integrated strategy 1212712001 21790 Decentralized Rural Development, and Enhanced Community Participation: a case study from Northeast Brazil IBrazil - The management o f agriculture, rural development and natural resources (Vol. 1) I 713111994 I 11783 Rural Development : F r o m Vision to Action? 913011998 23925 Education Brazil - Teachers development and incentives - a strategic framework 12/6/2001 20408 Brazil - Early Child Develop. A Focus on the Impact o f Preschools 912412001 22841 Brazil - Assessment o f the Bolsa Escola Programs (Vol.1) 3/15/2001 20208 Brazil - Higher education sector study (Vol. 1) 613012000 19392 Brazil - Early Childhood Development: a Focus o n the Impact o f Preschools 912412001 22841 Environment Brazil - Managing pollution problems : the brown environmental agenda (Vol. 1) 212711998 16635 Brazil - Managing environmental pollution in the state o f R i o de Janeiro (Vol. 1) 812211996 15488 Brazil - National Biodiversity Project 31111996 14523 Finance & PSD Brazil - The Custo Brasil since 1990-92 (Vo1.1) 1211011996 15663 Brazil - The development o f Brazilian capital markets (Vol. 1) 101711994 11581 Brazil - Private sector and social services in Brazil : who delivers, who pays, who regulates (Vol.1) 613011994 13205 Brazil - An assessment o f the private sector (Vol. 1) 612411994 11775 Health Brazil - Maternal and Child Health 212612002 238 11 The Brazil Health System: Impact Evaluation Report 613011998 18142 Brazil - Addressing nutritional problems in Brazil l o l l 811996 16010 Brazil - The organization, delivery and financing o f health care in Brazil : agenda for the 90s 613011994 12655 Brazil - The new challenge o f adult health 813111990 PUB7807 56 Infrastructure Brazil - Rural electrification with renewable energy systems in the Northeast : a preinvestment study 713 112000 ESM232 Brazil - Multimodal freight transport : selected regulatory issues 1011511997 16361 Brazil - Hydro and thermal power sector study 913011997 ESM197 Brazil - Natural gas pricing and regulatory study 1012511996 12772 Brazil - Energy efficiency and conservation: strategic pathways for energy efficiency in Brazil 11111995 ESM170 Brazil - Attacking Brazil's Poverty - poverty report focus o n urban poverty reduction policies 313112001 20475 Brazil - R i o de Janeiro - a city study 611911999 19747 Public Sector Management Brazil - Public expenditures for poverty alleviation in Northeast Brazil 611112001 22425 Brazil - Financing municipal investment : issues and options 412012001 203 13 Brazil - State debt : crisis and reform 1111411995 14842 Brazil - An agenda for stabilization 101711994 13168 Brazil - F r o m Stability to Growth through Public Employment Reform 211711998 16793 Brazil - Issues in Fiscal Federalism 61412002 22523 Social Protection Brazil - Attacking Brazil's Poverty - Poverty Report Focus Urban Poverty 313 112001 20475 Brazil - Critical issues in social security 51112001 22513 Brazil - Gender Review: Issues and Recommendations 112312002 23442 Brazil - Poverty assessment 612711995 14323 Brazil - Social insurance and private pensions 112511995 12336 Brazil - Public spending o n social programs: issues and options 512711988 7068 Urban Development Brazil - R i o de Janeiro: A Citv Studv 611911999 19747 57 Annex 88: List of Projects Loan# I Name Commitment Approval Date Closing Date AGRICULTUI 31300 ]Agriculture Research 111 24-Oct-891 30-Jun-96 31600 BR LND MGMT Il-S. CATAR 30-Jan-90 18-Oct-91 31700 NE IRRIG I 30-Jan-90 30-Jun-9S grant ITAPARICA SUPPLEM 20-Feb-90 grant ITAPARICA 20-Feb-90 31-Dec-97 grant SCIENCE CENTERS-EMERGENCY ASSISTANCE 1-J~l-94 grant EMERGENCY ASSISTANCE 28-Oct-94 grant EXTRACTIVE RESERVES 16-NOV-94 grant INDIGENOUS LANDS 19-Jun-95 39170 RURAL POV. (BAHIA) 29-Jun-95 29-Jun-96 39180 RURAL P0V.- CEARA 29-Jun-95 30-Jun-01 39190 RURAL P0V.-SERGIPE 29-Jun-95 30-Jun-01 40600 BR (PR)R.POVERTY 27-Jun-96 30-Sep-02 40600 BR(PR)R. POVERTY 27-Jun-96 30-Sep-02 grant FOREST RESOURCES MNGMT 12-Dec-96 41200 R.POVERTY(RGN) 12-Dec-96 30-Jun-02 grant R.POVERTY(PE) 12-Dec-96 31-Dec-01 41210 (PIAUI)R.POVERTY 12-Dec-96 31-Dec-01 41470 LAND RFM PILOT 22-Apr-97 30-Jun-02 41480 RGS LAND MGTIPOVERTY 22-Apr-97 30-Jun-Of 41690 AG TECH DEV. 22-Ma y-97 22-May-97 41900 CEARA WATER PILOT 12-Jun-97 12-Jun-97 42380 BR LAND MGT 3 (SA0 PAULO) 28-Oct-97 28-Oct-97 42510 PARAIBA R.POVERTY 20-NOV-97 20-NOV-97 42520 MARANHAO R.POVERTY 20-NOV-97 30-Jun-0: 44800 ANIMAL&PLANT DIS. CO 27-May-99 31-Dec-0: 45310 CEARA WTR MGT (PROGERIRH) 6-Jan-00 30-Jun-OE ‘0370 7037T001 LAND-BASED POVERTY ALLEVIATION I 30-NOV-00 30-Sep-04 46260 RURAL POVERTY REDUCTION PROJECT - CE 26-Jun-01 30-Jun-OE 46250 RURAL POVERTY REDUCTION PROJECT - PE 26-Jun-01 30-Jun-OE 46240 RURAL POVERTY REDUCTION PROJECT - PI 26-Jun-01 30-Jun-OE 46230 RURAL POVERTY REDUCTION PROJECT - BA 26-Jun-01 30-Jun-OE 30-Dec-9E 36040 BR NE BASIC EDUC II 212 13-May-93 31-Dec-9E 36630 BR NE BASIC EDUC 111 206.6 23-NOV-93 31-Dec-9: 37330 BR- MINAS GERAIS BASIC EDU. 150 17-May-94 28-Jun-0; 37660 BR- PARANA BASIC EDUC 96 28-Jun-94 31-Dec-01 42660 BR- SC. & TECH 3 155 18-Dec-97 31-Dec-0; 43110 BR- FUNDESCOLA I 62.5 2-Apr-98 30-Jun-01 44870 BR- FUNDESCOLA 2 202.03 8-Jun-99 31-Dec-04 45920 BR- BA BASIC EDU PROJECT (PHASE I) 69.6 20-Dec-00 30-Jun-0: 45910 BR- CEARA BASIC EDUCATION 90 20-Dec-00 30-Jun-OE 58 ENVIRONME r 34800 INATL IND POLLUTN 50 28-May-92 30-Jun-97 34920 MAT0 GROSS0 NAT RES 205 18-Jun-92 31-Dec-01 grant NATURAL RESOURCES POLICY PROJECT 22-NOV-94 39240 BR ENV/CONS(CVRD) 50 11-JuI-95 30-Jun-OC grant N'TL BIODIVERSITY 16-Apr-96 grant BR BIODIVERSITY FUND 16-Apr-96 42320 BAHIA WTR RESOURCES 51 11-Sep-97 31-Dec-Oi 43100 FED.WTR MGT 198 2-Apr-98 31-Dec-0: grant MONlTORlNGiANALYSlS 14-Aug-98 43890 BR EMER. FIRE PREVENTION 15 10-Sep-98 31-Dee-01 45240 INATL ENV 2 151 9-Dec-991 30-Jun-031 32680 PRIVATE SCTR FIN 300 29-NOV-90 14-Mar-91 42450 CEN.BANK TAL 20 18-NOV-97 2-Mar-98 45540 NE MICROFINANCE DEVELOPMENT 50 30-May-00 6-Sep-00 70530 Prgrm Financial Sect Adj I 404 24-May-01 23-Jun-01 46370 TA FINANCIAL SECTOR 14.46 4-Sep-01 HEALTH 26990 NE BASIC HLTH SRV I 59.5 22-May-86 31-Dec-95 29310 NE ENDEMIC DIS. CTL 109 31-Mar-88 30-Jun-96 30720 AMAZON BASIN MALARIA 99 25-May-89 30-Jun-96 31350 NE BASIC HLTH SRV II 267 30-NOV-89 30-Dec-97 36590 BR: AIDS CONTROL 160 9-NOV-93 30-Jun-98 40470 BR- HEALTH SECTOR REFORM - REFORSUS 300 20-Jun-96 31-Dee-01 43920 BR- AIDS & STD CONTROL II 165 15-Sep-98 9-Feb-99 43940 BR- DISEASE SURVEILLANCE - VlGlSUS 100 17-Sep-98 17-Feb-99 31690 HWY MGMT AND REHAB 310 13-Feb-90 31-Dec-96 32270 ELEC TRNS&CONSRVTN 385 14-Jun-90 30-Jun-95 33760 HYDROCARBN TRNSP/PRO 260 26-Jun-91 31-Dec-01 34570 METRO TRANSPSPAULO 126 31-Mar-92 31-Mar-98 35480 STATE HWY MGMT 88 22-Dec-92 31-Dec-99 36330 METRO TRANSP. RIO 128.5 29-Jun-93 14-Oct-93 37150 STATE HIGHWAY MANAGEMENT II 220 15-Mar-94 25-Mar-94 39160 BEL0 H M.TSP 98.6 29-Jun-95 30-NOV-95 39150 RECIFE M.TSP 102 29-Jun-95 3-Apr-96 40460 RAILWAYS RESTRUCTURG 350 20-Jun-96 17-Sep-96 41650 RGS HWY MGT 70 15-May-97 31-Dec-03 41880 FED HWY DECENTR 300 12-Jun-97 31-Dec-02 42650 GASSCTRDEVPROJECT 130 18-Dec-97 31-Dee-00 42910 RJ M.TRANSIT PRJ. 186 5-Mar-98 30-Jun-02 43120 (BF-R)SP.TSP 45 7-Apr-98 31-Dee02 44940 SALVADOR URBAN TRANS 150 17-Jun-99 31-Dec-03 45140 ENERGY EFFICIENCY (ELETROBRAS) 43.4 5-Oct-99 31-Dec-O3 grant BR ENERGY EFFICIENCY (GEF) 5-Oct-99 grant GUARANTEED NOTE TRANSPORTADORA BRASlLEl 14-Dec-00 46360 GOIAS STATE HIGHWAY MANAGEMENT 65 23-Aua-01 31-Dec-04 59 27210 /PUB SECTOR MGMT 29 19-Jun-86 31-Dec-93 41390 BR RGS ST.REF 125 4-Mar-97 31-Dec-98 41890 BR MT STATE PRIV. 45 12-Jun-97 30-Sep-01 421 10 BR RJ ST.PRIV. 250 17-JuI-97 31-Dec-98 43180 MG STATE PRIV. 170 28-Apr-98 31-Dec-99 45430 BR ADMIN & FISC REF 505.06 30-Mar-00 31-Dec-00 70460 BR PRGM.FISCAL REFORM SAL 757.6 18-Jan-01 31-Mar-02 46040 BR FISCAL & FIN. MGMT. TAL 8.88 24-May-01 31-Dec-05 43690 IBR PENSION REFORM LIL 51 30-Jun-981 31-Dec-01 44310 BR SOC SECURITY REF I 757.57 7-Jan-99 30-Jun-99 44300 BR SOCIAL PROTECTION 252.52 7-Jan-99 30-Jun-99 45370 INSS REF LIL 5.05 7-Feb-00 31-Dec-02 45420 SECOND SOCIAL SECURITY SPEC. SEC. ADJ LN 505.06 30-Mar-00 31-Dec-00 25320 RURAL WS & S PILOT 16.3 7-May-85 30-Sep-90 26230 STA CATARINA TOWNS 24.5 19-Sep-85 31-Dec-93 26450 NE URBN RECON (FL'D) 100 19-Dec-85 31-Dec-88 26810 SALVADOR METRO DEVT 55 22-Apr-86 30-Jun-97 29750 RIO EMERG FL'D RECON 175 24-Jun-88 30-Sep-95 29830 W&S/PROSANEAR 80 29-Jun-88 30-Dec-96 31000 MUNIC DEV 100 22-Jun-89 31-Dec-95 31020 WATER SCTR S A 0 PAUL0 280 28-Jun-89 30-Jun-97 31290 MUN DEV (RIO GRANDE) 100 24-Oct-89 31-Dec-95 34420 WATER SECTOR MODERNIZATION PROJECT 250 10-Mar-92 30-Jun-00 35050 BR WTR Q/PLN(SP/PR/FED) 245 2-JuI-92 31-Dec-02 35540 WTR Q/PLN(MINAS GERA 145 5-Jan-93 30-Sep-99 36390 BR MINAS MNCDEVELOPMT 150 20-Jul-93 31-Dec-01 37670 ESPSANTO WATER 154 28-Jun-94 31-Dec-02 37890 CEARA URBAN DEVELOPMENT &WATER RESOURC 140 6-Sep-94 31-Mar-02 41400 BAHIA MUN.DV 100 4-Mar-97 31-Mar-02 42920 WATER S.MOD.2 150 5-Mar-98 30-Jun-04 45.170 PROSANEAR 2 30.3 6-Jan-00 31-Dec-04 60 61 - 0 0 N 0 0 0 N 5. h E; m m 2 . r. 5 9 9 9 m F NL? m o u s - h mwr- m ' 0 6 IDN W 0' vi e U 01 e 01 0 I N h I # e h 8 I 62 Annex Table 11: Brazil at a glance 9112102 Latin Upper- POVERTY and SOCIAL America middle- Brazil &Carib. income Developmentdiamond’ 2001 Population, mid-year (millions) 172.6 524 504 Life expectancy GNI per capita (Atlas mefhod, US$) 3,060 3,560 4,460 GNI (At/as method, US$ billions) 528.7 1,862 2,248 T I Average annual growth, 1995-01 Population (%) 1.3 1.5 1.3 ” >%, + Labor force (%) 1.8 2.2 1.8 GNI Gross per -, i ’ :c- Primary , ’ ~ Most recent estimate (latest year available, 1995-01) capita t enrollment ti I <-“- Poverty (% of population below national poverty line) 22 i. Urban population (% of total popuiafion) 82 76 77 Life expectancy at birth (years) 68 70 71 i Infant mortality (per 1,000 live births) 30 29 24 Child malnutrition (% of children under 5) 6 9 9 Access to improved water source Access to an improvedwater source (% ofpopuiafion) 87 85 87 Illiteracy (% of population age 15+) Gross primary enrollment (% of schoot-age population) 14 154 11 130 10 127 ..--- Brazil Male 156 131 128 Upper-mlddle-incomegroup Female 152 128 126 KEY ECONOMIC RATIOS and LONG-TERM TRENDS 1981 1991 2000 2001 Economic ratios’ GDP (US$ billions) 263.6 407.7 593.8 502.0 Gross domestic investmenffGDP 23.1 19.8 21.7 21.0 Trade Exports of goods and serviceslGDP 9.4 8.7 10.8 13.4 Gross domestic savingsiGDP 22.7 20.5 20.1 19.9 T Gross national savingdGDP 18.6 18.6 17.2 16.4 Current account balancelGDP Interest paymentslGDP -4.5 3.1 30.9 -0.3 0.5 29.7 -4.1 2.2 39.8 -4.6 2.1 45.0 Domestic savings - , cTa- . s Investment “;I! Total debffGDP Total debt servicelexports 66.4 23.3 92.0 60.5 37.7 f_l Present value of debffGDP Present value of debffexports 327.4 Indebtedness 1981-91 1991-01 2000 2001 2001-05 (average annual growfh) GDP 2.9 2.9 4.4 1.5 2.8 Y_-x Brazil GDP per capita 0.9 1.5 3.1 0.2 1.6 Upper-middle-income group Exports of goods and services 6.8 6.4 11.4 12.1 3.7 I STRUCTURE of the ECONOMY 1981 1991 2000 2001 Growth of investmentand GDP (Oh) I (% of GDP) 20 7 Agriculture 10.7 7.8 7.4 9.3 I Industry 43.7 36.2 28.3 33.9 10 Manufacturing 32.9 25.3 17.4 21.0 Services 45.5 56.1 64.3 56.8 0 I I Private consumption 68.0 61.6 60 6 60.1 , -mi General government consumption 9.3 17.9 19.3 19.9 -it GDI +GDP Imports of goods and services 9.8 7.9 12.4 14.4 (average annual growih) 1981-91 199161 2ooo 2o01 1 Growth of exports and imports (%) I 40 Agriculture 2.5 3.4 3.0 5.1 ~ Industry Manufacturing 2.2 1.4 2.5 1.9 4.9 6.1 -0.6 0.6 1 20 Services 3.5 2.9 3.7 1.5 l o Private consumption 1.2 5.0 3.7 2.6 General government consumption 7.5 -1.0 0.0 4.8 1 I -20 Gross domestic investment 3.9 3.5 11.2 -2.0 I I ---Exports +Imports Imports of goods and services 2.9 10.0 12.4 0.7 Note: 2001 data are preliminary estimates ‘ I The diamonds show four key indicators in the country (in bold) compared with its income-group average. If data are missing, the diamond will be incomplete, 63 Brazil PRICES and GOVERNMENT FINANCE 1981 1991 2000 2001 inflation ( O h ) Domestic prices (% change) looT Consumer prices 101.7 432.8 6.0 7.7 implicit GDP deflator 107.2 415.3 8.0 7.4 Government finance /a I (% of GDP, includes current grants) Current revenue 21.9 23.0 96 97 96 99 00 Current budget balance 1.8 1.8 * -GDP deflator "0"CPI Overall surplus/deficit 3.2 3.8 O' TRADE 1981 1991 2000 2001 Export and import levels (US$ mill.) (US$ millions) Total exports (fob) 31,620 55,086 58,223 Coffee 2,870 3,048 2,932 6osooo T Soybeans 2,031 2,188 2,726 Manufactures 20,483 41,027 41,144 Total imports (cif) 21,041 55,783 55,581 Food 1,275 1,507 1,169 Fuel and energy 3,371 6,362 6,276 Capital goods 5,966 13,591 14,803 95 9s 97 98 99 00 01 Export price index (1995=100) 81 81 97 94 Import price index /7995=700) 70 68 118 114 W Exports CI Imports Terms of trade (1995=100) 116 120 82 82 BALANCE of PAYMENTS (US$ millions) Exports of goods and services 1981 25,522 1991 34,333 2000 64,584 2001 67,545 Current account balance to GDP (%) I imports of goods and services 27,200 26,142 72,443 72,652 Resource balance -1,678 8,191 -7,859 -5,107 Net income -10,272 -11,154 -17,886 -19,743 Net current transfers 10 1,556 1,521 1,638 Current account balance -1 1,940 -1,407 -24,224 -23,212 Financing items (net) 12,286 -3,272 31,904 19,468 Changes in net reserves -346 4,679 -7,680 3,744 Memo: Reserves including gold (US$ millions) 6,693 8,763 33.01 1 35,886 Conversion rate (DDEC, local/US$) 3.38E-11 1.48E-4 1.8 2.4 EXTERNAL DEBT and RESOURCE FLOWS 1981 1991 2000 2001 (US$ millions) Composition of 2001 debt (US$ mill.) Total debt outstanding and disbursed 81,454 121,020 236,157 226,067 iBRD 2,286 8,165 7,377 7,963 7963 ... 8337 IDA 0 0 0 0 29292 Total debt service 17,874 8,340 62,891 42,977 IBRD 317 1,917 1,351 1,362 IDA 0 0 0 0 Composition of net resource flows Official grants 8 50 54 Official creditors 1,230 -1,062 -1,780 786 Private creditors 7,512 1,701 7,936 3,615 Foreign direct investment 2,520 1,103 32,779 22,457 Portfolio equity 0 803 3,076 2,481 152416 World Bank program Commitments 1,039 1,348 1,290 1,609 A - IBRD E - Bilateral Disbursements 387 840 1,692 1,639 B - IDA D -Other multilateral F Private C - IMF ~ Principal repayments 136 1,248 887 828 G -Short-term Net flows 251 -408 805 810 Interest payments 181 669 464 533 Net transfers 70 -1,077 34 1 277 Development Economics 9/12/02 a. Data refer to central government ANNEX 2 SUMMARIES OF SECTOR EVALUATIONS 67 1. Public Sector Management The Bank’s Assistance Strategy 1. The Real Plan succeeded in establishing greater macroeconomic stability and with i t a conducive environment for a more active Bank role, The Bank’s assistance included, inter alia, ESW, technical assistance loans and adjustment loans dealing with fiscal reform, social security reform, privatization o f the state banks, and economic regulation. 2. In general, the E S W was relevant to the major policy problems facing the country, and the assessment o f the problems and issues addressed was o f satisfactory quality. The proposed solutions and policy suggestions were generally well founded, although not always realistic nor practical, given the political realities and constraints faced by the Government. Moreover, inadequate attention was given to report dissemination. 3. In great part derived from the E S W and related policy dialogue, the Bank undertook a number o f lending operations as part o f i t s assistance in the area o f public sector management during the period 1995-2001. This lending consisted of: (a) four state reform loans; (b) two fiscal reform loans; (c) two social security reform loans; and (d) a series o f technical assistance loans supporting the implementation o f these reforms. The four state reform loans were prepared and approved in the 1995-1998 period, when it became apparent that fiscal imbalances in the states were part o f the overall fiscal imbalances that jeopardized the Real Plan. The remaining loans were approved after 1998, in the aftermath o f the Russia crisis and in support o f the Government’s effort at fiscal adjustment. 4. State Reform and Privatization Loans (SRLs). F r o m the Bank’s ESW, independent w o r k within the Government, and the policy dialogue between the Bank and the Government, it was clear by 1995 that there was an emerging fiscal problem in the states o f major proportions - a problem which threatened macroeconomic stabilization. T h i s was the point o f departure for the development o f the SFUS, with the overall goal being to support fiscal adjustment in the states. As a partial means to this end, the divestiture o f state-owned enterprises or the concession o f their services became the main stay objective for the loans. Also, the establishment o f a state regulatory agency for the newly concessioned or privatized public services was supported in all the SRLs. The S R L objectives were consistent with the CAS objectives o f macroeconomic stability and public sector modernization, as well as private sector participation in infrastructure. The effects o f the privatizations supported by the loans, with a few possible exceptions, were positive. There i s also evidence o f fiscal adjustment in these states. 5. Fiscal Reform Loans (FRLs). Brazil’s increasingly weak macroeconomic fundamentals in 1996-98 and a looming currency crisis prompted the Brazilian Government to undertake the necessary and overdue fiscal adjustment. Assistance from the Fund, the Bank, and the IDB was requested in support for a fiscal adjustment program. A Fund Stand-by Arrangement, involving substantial fiscal adjustment, was negotiated and approved by the Fund’s Board in November 1998. The Bank’s participation in the special international financing package took the form o f two successive fiscal reform loans (FRLs), along with two adjustment loans for social security reform and one adjustment loan for social protection (discussed below). 6. The fiscal reform loans entailed two separate, but sequentially linked, adjustment operations. Common features o f both loans included: (a) support t o overall improvements in Brazil’s fiscal performance at all Government levels (including adherence to a number o f quantitative performance criteria specified in a Fiscal Stability Program and supported by the 68 Fund SBA); and (b) specific components supporting, inter alia, public administration reforms and improvements in debt management. B o t h operations were designed as one-tranche loans, with up-front conditionality. 7. The outcomes and results o f the FRLs have been positive. The primary balance o f the General Government shifted from deficits to surpluses around 3.5 percent o f GDP, exceeding the FRL targets. The proportion o f government spending for personnel has been progressively reduced. Greater fiscal discipline has been imposed o n the states, as supported by the FARL, through the implementation o f legislation affording the National Treasury a central role in refinancing the debt o f the states. Refinancing contracts between the state governments and the National Treasury n o w involve fiscal performance criteria. The PFR-SAL supported inter alia the implementation o f the Fiscal Responsibility Law, which i s currently being successfully pursued as a new modus operandi for govemment at all levels. 8. Social Security Reform Loans. As a part o f the Bank’s efforts to support fiscal adjustment, two adjustment loans were made t o support reforms in the country’s problematic social security and pension systems. The social security loan involved support for a Constitutional amendment which, upon enactment in December 1998, introduced the principle o f actuarial and fiscal balance for both private and public sector workers, eliminated the benefit formula for private sector workers from the Constitutional text, introduced a minimum retirement age for public sector employees, eliminated most special regimes, and tightened eligibility criteria. The second loan was centered on the introduction o f a new benefit formula for private workers in an attempt to reduce actuarial imbalances and to tighten contributions and benefits. It also included the restructuring o f the Social Security Organization (INSS), a L a w o f Crimes Against Social Security, tightening o f disability criteria, and incentives for expansion o f coverage among the self-employed. 9. B o t h the Government and the Bank have acknowledged that these initial reforms would not by themselves reduce significantly the severe actuarial imbalances o f the pension system. However, the reforms undertaken to date have introduced several important changes and, most importantly, opened the door for subsequent and more specific reforms through enactment o f ordinary legislation. f the Bank’s Assistance Overall Evaluation o 10. During the macroeconomic stabilization crisis o f the early 1990s, the Bank did not appear disposed, or possibly able, to assist the Brazilian Govemment find a solution t o the crisis. There i s n o evidence that the Bank had any role in the formulation or implementation o f the Real Plan. Since the Real Plan the Bank has played a much more positive and supportive role. A central and consistent Bank assistance objective-as articulated in the recent CASs-has been to support the underlying fiscal adjustment to make stabilization sustainable. The objective was highly relevant and Bank loans supporting public sector reforms during the 1995-2001 period were very consistent with the CASs. 11. The overall assessment o f the Bank’s assistance to Brazil in the area o f public sector reforms, i s rated as satisfactory, encompassing the S R L cluster, the fiscal reform cluster, and the social security cluster. Within the areas o f focus that the Bank set for itself, Bank assistance was both relevant and effective. The Bank contributed to the establishment o f a fiamework and control mechanism for improving fiscal performance at several layers, including the federal government, states, municipalities and social security. This assistance has come in several ways, 69 including lending for adjustment and technical assistance, ESW, informal technical assistance in loan preparation and execution, and informal personnel exchange. 12. Brazilian policy-makers had a clear understanding o f the macroeconomic issues and problems, including those related to public finance questions. The fiscal adjustment was delayed until 1998, but not for lack o f understanding o n the part o f the policy-makers but a lack o f political consensus. The unfolding currency crisis, beginning in 1997, created the political will to finally proceed with fiscal adjustment, and the Government effectively used that opportunity to implement a number o f much needed reforms. If some reforms, such as social security reform, did not go as far as originally expected, i t was not because o f lack o f Government commitment and ownership, but because o f lack o f political consensus and support. 13, Apart from the Brazilian Government, the IMF was unquestionably the institution that played the most central and critical role in supporting the fiscal adjustment beginning in 1998. The Fund exercised general leadership in defining the parameters o f the fiscal program with the Brazilian authorities and in orchestrating financial support for that program. However, both the Bank and the IDB became important partners in providing necessary financial, as w e l l as technical, support. The Bank supported fiscal adjustment and reform during the 1994-98 period through state reform loans. Once the Government finally decided to proceed with a concerted and significant fiscal adjustment program beginning in late 1998, the Bank responded quickly with adjustment lending in support o f deeper fiscal reforms and social security reforms as well. Recommendations for Future Assistance 14. Continued fiscal adjustment and public sector modernization will continue to be important challenges for the new Government that took office in January 2003. Bank responsiveness and potential support for these challenges should constitute high priorities for Bank assistance, and accordingly it i s recommended that Bank assistance continue in the area o f public sector reform and public sector management. Three broad areas o f assistance are suggested, covering fiscal management, public governance, and social security reform. 15. First, under the heading ofJiscal management, several specific interventions could include: (i)continued support for fiscal adjustment and reform, possibly in the form o f additional programmatic adjustment lending, conditional o n the generation o f substantial primary fiscal surpluses and a reduction o f public sector debt, the deepening o f the administrative reforms, and the implementation o f the long-delayed tax reform; and (ii) assistance to state governments and selected municipalities in public finance management, in the form o f investment or technical assistance loans. 16. Second, in relation t o public governance, specific Bank interventions could include: (i) support for improvements in the regulatory framework, assisting the recently established federal and state regulatory agencies t o overcome their start-up problems; and ( ii) assistance to “new” areas o f public governance such as judicial system reform, the strengthening o f property rights and the reform o f labor legislation. 17. Third, for proceeding with high priority social security reforms, the Bank could: (i) provide assistance to the Government in the elaboration o f a well-articulated educational/public relations campaign and opinion surveys designed to reduce resistance to further and deeper reforms; and (ii) support actual efforts o n the part o f the Government to reduce imbalances in the PAYG systems for private and public workers through further adjustment lending. 70 2. Agriculture The Agriculture Sector in the 1990s 18. Agriculture in Brazil accounted for 11 percent o f GDP and 22 percent o f employment in 2000. Livestock production has outpaced crop production and now accounts for 40 percent o f output value. The agriculture sector i s particularly important to trade, with about 35 percent o f merchandise exports coming from agriculture and processed food products. The country accounts for about 4 percent o f global agricultural trade. I t i s the world’s largest exporter o f coffee, sugar and orange juice, the second biggest exporter o f soybeans and the third largest exporter o f beef and chicken. 19. Brazil has one o f the most concentrated patterns o f land distribution in the world. The last agricultural census (1 996) revealed that 11 percent o f rural property owners controlled 68 percent o f arable land. There has been progress at land reform, as indicated by the redistribution o f about 18 m i l l i o n hectares (equivalent to 6 percent o f the land in farms) to 542,000 families (almost 2 m i l l i o n people). However, land reform has not yet changed significantly the skewed distribution o f land in Brazil. Since the early 1990s Brazil has made progress in reducing rural poverty, but the incidence o f poverty in rural areas remain extremely high, especially the Northeast-72 percent in 1999, down from 85 percent in 1990. The Bank’s Assistance to Agriculture 20. The share o f total Bank commitments devoted to agriculture (broadly defined) slumped from 32 percent in 1980-89 (US$3.8 billion) to 14 percent in 1990-02 (US$ 1.9 billion), with important variations across types o f interventions. There was an increasing focus on rural development/poverty alleviation projects in Northeast states. Most o f the poor are located in the rural Northeast and the focus o f the Bank’s strategy towards the country and the sector has adequately reflected this. The share o f this class o f projects in total commitments increased from 47 percent in the 1980s to 74 percent in the 1990s. There was also a change in design in this class o f projects in the 1990s, with greater focus on community-based poverty alleviation. Projects supporting land reform were also launched in the 1990s, as part o f the poverty alleviation strategy. The Bank scaled down more traditional projects such as irrigation projects, due to unsatisfactory outcomes in the 1980s and early 1990s. f the Bank’s Assistance Overall Assessment o 2 1. Since the middle o f the last decade agriculture projects have been turned around. Between 1980 and 1989 agriculture projects in Brazil had on average worse ratings than the rest o f the Brazilian portfolio, and the Brazilian portfolio had on average worse ratings than projects in the region and the rest o f the Bank. In the 1990s, the performance o f the Brazilian portfolio in general and agriculture projects in particular was turned around, primarily because o f the sharp improvement after 1995, when Brazil outperformed the region and the Bank by a wide margin. O f the projects with an unsatisfactory outcome in 1990-2001, two were irrigation projects, two were credit projects, two were for Amazonian development, one was for rural water supply and sanitation and one for land administration. O f the projects whose outcome was rated satisfactory, roughly 60 percent correspond to the community-driven rural development program which covered all nine states in the Northeast plus Minas Gerais. 71 22. The overall assistance to agriculture for the decade as a whole i s rated satisfactory, primarily due to the positive outcomes o f rural development and land reform projects. The rural development projects were successfully restructured in the early 1990s, by building a demand- driven approach in a cost-effective manner. These projects have delivered electricity, water, and other essential services to large numbers o f poor households in the region. Roughly 60 percent o f disbursements were channeled towards the provision o f electricity and water alone, benefiting 835,000 poor households and an estimated rural population o f more than 3 million people. Second, although i t i s too early to pass judgment on the land reform projects, the early evidence shows that the cost o f delivering parcels to small farmers i s lower than in previous land reform models and the existence o f supporting infrastructure and financing suggests that there i s a viable base for raising productivity. 23. The less satisfactory project clusters were reduced in number and size o f commitments. Research and extension efforts are rated moderately satisfactory. The research projects provided top-quality training to a critical mass o f professionals, but were less successful in transferring appropriate technologies to poor Northeastem farmers. On the other hand, within the last three years, a concerted effort has been launched to conduct research more relevant to that region. Finally, the irrigation projects are rated unsatisfactory owing to persistent high costs and conspicuous design flaws. However, in important pockets, crop diversification exceeded appraisal expectations, farmers used their own resources to upgrade to drip irrigation, and there was significant growth o f farm incomes. 24. The Bank i s the only multilateral organization s t i l l devoting a significant attention to the agriculture sector. The Inter-American Development Bank has scaled back i t s involvement in agriculture to a much greater extent than the W o r l d Bank, with only one project in 1995-02. The Bank maintained a direct policy dialogue with most Northeastern states and contributed to better outcomes in the rural area, particularly through the rural development and land reform projects. It i s probable that, at least in some states, these results would not have occurred without Bank involvement. Bank performance was stronger in the latter part o f the period, owing in n o small measure to increased Bank presence in the field. The Recife-based supervision team has been particularly effective, responding rapidly to requests and queries from the various state project management units. Less positive, however, has been the neglect o f impact evaluation. 25. The outcomes o f agriculture projects are likely to be sustained over time. I nparticular, rural development projects show good prospects for sustainability based o n the maintenance record for infrastructure investments and the likelihood that skills created in the municipalities will endure. Land reform projects require groups o f beneficiaries to take the initiative, keeping costs l o w to avoid excessive debt. I t s outcomes should also be sustained for a significant number o f new farms. It i s to be expected that some o f the newly-created farms will fail, but given the gross under use o f farmland before the reforms, even if only h a l f or so o f the new farms prove sustainable, this will be an improvement on the past. Recommendationsfor Future Assistance 26. Although the overall rating o f the Bank’s program i s satisfactory there i s s t i l l r o o m for improvement. There are four areas in which more attention i s needed. First, i t i s vital to avoid complacency in relation t o the community-driven rural development projects. The most important challenge i s to press forward with the comprehensive impact evaluation that i s n o w being sponsored. There i s good reason to believe that funds are adequately targeted, but supervision and implementation completion reports need to document this routinely. 72 27. Second, raising the productivity o f Northeastern farms remains the biggest challenge for the government and the Bank. The rural development projects have made an important contribution to providing the infrastructure on which agricultural growth can take place. However, more i s needed in generating and disseminating technology packages. There may n o w be a need for a more focused effort to raise farm productivity in areas o f higher agricultural potential. 28. A third area to address i s the possible shortfall in support to schooling in rural areas. In the 1990s, Brazil made great strides in raising school enrolments, but i t i s less clear what success there has been in reducing the gap that separates rural f r o m urban areas. Monitoring the gap between urban and rural schools i s an essential aspect o f improving performance. If rural schooling i s explicitly monitored and the results are widely reported there will be more pressure o n politicians to provide the funding needed to boost performance o f rural schools. Closer coordination i s needed between sector units within the Bank. Finally, the case for Bank assistance t o rural finance needs to be examined. Probably this should take the form o f institutional development rather than a credit line. 3. Water and Sanitation M a i n Developments in the Water and Sanitation Sector in the 1990s 29. The 1988 Constitution stated that public services should be provided by a public authority, either directly or through concessions, and also empowered municipalities to make these concessions. However, the 1988 Constitution and the 1995 Concession L a w have left important gaps, creating obstacles for private sector participation and for a faster development o f the sector. A Draft L a w establishing the National Water Policy should have resolved a l l these ambiguities, by setting tariff policy and concession criteria and establishing the autonomy o f the regulatory agency. However, the Draft L a w has not been passed by Congress yet, because it i s being challenged o n some key privatization issues, and because it i s facing strong opposition from municipalities (which disagree with the transfer o f the conceding power to the states). 30. Coverage o f water through network services increased from 66 to 78 percent o f households in the 1990s, while coverage o f sewerage through network services increased from 3 1 to 45 percent o f households in the same period. Total water coverage compares relatively well with other Latin American countries, particularly for urban areas, but can get as l o w as 9 percent in poor rural areas and, although wells and springs supply safe water to some o f the non-served rural population, a significant percentage s t i l l lacks access to safe water. Sewerage coverage in Brazil i s one o f the worst in the continent, reaching less than 60 percent o f the urban population. Urban sewerage coverage reaches only 80 percent in the richest Southeast and bottoms 2 percent in the North. Coverage o f sewerage networks in rural areas i s extremely small although somewhat alleviated by septic tanks, which can provide safe disposal in low-density areas. The Bank’s Assistance to the Water and Sanitation Sector 3 1. The Bank’s strategy for assisting the water and sanitation sector in Brazil has comprised three main components: (i) service extension to the poor; (ii)institutional and regulatory reform; and (iii)water quality, pollution control and urban drainage. The Bank’s program o f assistance to the sector seems t o be based o n a good diagnostic and t o be consistent with the sector’s needs. The three components were highly relevant: the first because o f the l o w levels o f service t o the poor; the second because o f the difficult regulatory issues that have hindered the development o f 73 the sector; and the third for having introduced an integrated approach to the problem o f water quality and pollution control. 32. The Bank’s assistance to the sector since the late 1980s has comprised eight projects with total commitments o f US$1.3 billion. Bank investments were concentrated in the early 1990s (about 85 percent o f total commitments) when high inflation was the main bottleneck for domestic investments in the sector. With macroeconomic stabilization, domestic sources gradually returned to finance the sector and the Bank has moved i t s assistance towards modernization and regulation reforms in order to sustain the sector’s development. Overall Evaluation of the Bank’s Assistance 33. The overall outcome o f the Bank’s assistance has been satisfactory. The objectives o f individual projects and clusters o f projects were relevant, and the projects already closed or near completion have met most o f their objectives. The Bank has pioneered the introduction o f river basin management, and the development and application o f low-cost technologies in poor urban areas in the PROSANEAR projects. Extension o f water and sewerage services to the population increased significantly, especially among the poor. Despite the ongoing controversy surrounding regulatory issues, the Bank has provided critical assistance to the revision and elaboration o f the National Sanitation Policy. 34. ESW work was unsatisfactory, however, due mostly to i t s scarcity. Bank assistance could have been more effective if it had undertaken ESWs on specific areas such as privatization, concession, and pricing criteria, and had promoted discussion o f these issues among the several interest groups. I t i s possible that such an effort could have reduced the current resistance to the approval o f the Draft Law. 35. The institutional development impact o f the Bank’s assistance i s rated as substantial. Bank projects required a significant degree o f inter-governmental integration and stakeholder participation, with positive effects in institution building. Several projects were launched as pilot projects and have been replicated at a broader, national level. Sustainability o f the outcomes i s judged as likely, despite the problems that still affect the full development o f the sector. The introduction o f sound managerial practices was not fully implemented in some State Water Companies (SWCs), and some low-income areas may lack resources to prevent deterioration o f the physical infrastructure. However, tariff pricing has generally improved with macroeconomic stabilization to sustain most o f the investments, and several SWCs have improved their performance. The adoption o f integrated approaches and river basin management units i s also a step in the right direction and contributes to the sustainability o f investments in the sector. More importantly, there i s a growing awareness among policy-makers and politicians that this sector has been neglected for too long, and that further neglect would pose a risk to the health o f the population. Recommendations for the Future 36. Despite the satisfactory evaluation o f the Bank’s assistance during the past decade, there i s room for further adjustment and improvements: First, the Bank should explore more effectively its knowledge and educational role. The key issue o f conceding power, although a matter o f law interpretation and political decision, can largely benefit from studies on asset valuation and compensatory schemes. Regarding privatization, an issue that i s poorly addressed and highly misunderstood, studies should cover themes such as universality o f services, 74 economies o f scale, partial against total concession, cross-subsidies, and regulation o f natural monopolies. 37. Second, achievements o f PROSANEAR should be fully scrutinized and disseminated to promote their effective mainstreaming into government programs and more effective coverage o f poor populations. Third, removal or mitigation o f financing constraints i s crucial for the future o f the Bank’s assistance. One alternative i s partnership schemes with the private sector in concessions, thereby reducing the need for public expenditures. Another i s to match funds with domestic sources (BNDES and CEF, for example). Attempts to forge partnerships with IDB must be made so that both institutions can share experiences in overcoming financial constraints in project analysis and implementation problems. 38. Finally, Bank strategies towards the water and sanitations sector must be more integrated with the Bank’s environmental and poverty strategies. As a starting point, a common strategy plan should be prepared with the collaboration o f the three areas. The new institutional arrangements in the Brazilian water management system have to be taken into account in future Bank projects. Issues such as water charging, water committee representation, and subordination to water agencies, require careful analysis in the design o f future projects. 4. Environment The Bank’s Assistance Strategy 39. Bank assistance to Brazil in the environmental area started modestly in the 1980s, with only two projects amounting to US$l08 million, representing only 1 percent o f total commitments during the decade. By contrast, during the 1990s the environment was identified as one o f the priority areas for Bank assistance. The intensification o f Bank involvement in this area was enabled by the increased awareness o f environmental problems in Brazil, as well as the improvements in the Brazilian institutional framework (in particular the creation o f IBAMA and the Ministry o f the Environment in the late 1980s). This translated into nine projects and almost US$900 million in commitments, accounting for 7 percent o f new commitments in the 1990- 2000 period. I naddition, the Bank managed U S $ 350 millions in grants from the Global Environment Fund (GEF) and the PPG7. 40. Bank assistance was channeled through three separate agendas. I t started in the 1980s with industrial pollution control projects (the Brown Agenda) and during the 1990s it focused on ecosystems, particularly the Amazon (the Green Agenda). Since the late 1990s the assistance has been expanded to include river basin management (the Blue Agenda). The Brown Agenda included three operations amounting to 10 percent o f the total environment portfolio. The Green Agenda comprised seven operations amounting to 60 percent o f the portfolio. In addition, there were also grants supporting work on green issues. Finally, the new blue agenda includes two operations amounting to 30 percent o f the portfolio. 41. In 1980 the Bank approved i t s f i r s t environmental project in Latin America-an industrial pollution control project for the state o f S2io Paulo. This project was renewed in 1986 and a nation-wide version, the National Pollution Control Project, was launched in 1992. During the remainder o f the 1990s the Bank seemed to struggle to distill the experiences o f these projects and elaborate a Brown Strategy. A few reports were developed to stress the need for institutional strengthening, but they did not set up a clear agenda, nor elaborate a careful analysis o f the main causes and consequences o f pollution problems in Brazil. 75 42. In 1990 the Bank launched the Green Agenda with the first National Environmental Project to protect other national ecosystems, reedited later in 1998. H a l f o f Bank commitments in the green agenda was allocated to two Amazonian states with specific projects, P L A N A F L O R O in Rondonia and PRODEAGRO in Mato Grosso. These projects focus on conservation, monitoring and financing o f sustainable practices, and o n correcting the adverse effects o f some infrastructure projects supported by the Bank in the 1980s. The Green Agenda also included management o f GEF grants for biodiversity (PROBIO and FUNBIO) and PPG7 grants to Amazonian projects. Overall, the Green Agenda seems t o have been well timed and tuned with Brazil’s policies and environmental challenges. However, most green projects lacked sound private and social analysis o f costs and benefits o f deforestation as well as a good understanding o f issues related to international compensation for forest preservation. 43. The blue agenda was launched in the late 1990s with two large projects o n water management in the Northeast (Federal P R O A G U A and Bahia PROAGUA). Water provision in the Northeast has been a secular problem, not only because it has a large semi-arid region, but also due to political interference and lack o f institutional capacity. The Bank’s participation and support to the new blue agenda i s welcome, but this work seems to be poorly connected with the Bank’s work in the closely related area o f water and sanitation, despite the recognition o f the tight connections between sanitation and the environment in the CASs. Overall Assessment of the Bank’s Assistance 44. Bank projects generally had relevant objectives, consistent with the country’s o y n agenda and needs. Some o f these projects were designed to deal with the environmental degradation associated with infrastructure projects supported by the Bank in the 1980s. Although some individual projects have not been able to generate the expected results, outcomes have been generally positive. Some environmental indicators for Brazil are lagging behind those o f developed countries but are in many cases ahead o f those in developing countries and are showing a consistent trend o f improvement. For example, air pollution in the city o f SBo Paulo i s s t i l l severe but the number o f critical days dropped in the last ten years. The share o f protected areas has increased by 75 percent. The average rates o f deforestation in the Amazonian region are s t i l l high but declined relative t o the 1980s. It i s admittedly hard to measure with any degree o f accuracy the extent o f the Bank’s contribution t o these outcomes, but the Bank provided invaluable assistance to the Government through projects, ESW, and the policy dialogue, and contributed at least indirectly to these results. 45. Except for the three projects under the B r o w n Agenda, the institutional development impact o f Bank projects i s judged to be substantial. Environmental projects have promoted participation by stakeholders and ownership through integrated management among government institutions, NGOs, and social actors. Technical assistance components are present in all projects, in a mix o f institutional strengthening, social infrastructure investments, and the pursuing o f ecologically sustainable activities. 46. There i s a r i s k that these positive outcomes might not be sustained over time. Environmental concerns are s t i l l a luxury good for Brazil (compared to education, health, and basic infrastructure) and, consequently, the environment could become a l o w priority sector during a fiscal crisis. Although the risk o f this unfavorable outcome i s considered to be modest (the degree o f environmental awareness has increased and a fiscal crisis i s not expected), i t cannot be entirely ruled out. 76 Recommendationsfor Future Assistance 47. Despite the favorable assessment o f the Bank’s assistance to Brazil in the environmental area, there i s s t i l l room for improvements. Regarding analytical work, the Bank should undertake a sound analysis o f the private and social costs and benefits o f deforestation, as w e l l as studies integrating blue and brown issues and i t s activities in the water and sanitation sector. Regarding project work, the Bank should adopt more realistic criteria for the determination o f the regional coverage and scale o f projects, in relation to the institutional capacity o f relevant implementing agencies. Managerial plans and r i s k assessment analyses could be improved. The Bank should engage representatives o f the private sector (timber producers, industries and farmers) in project formulation and implementation, forging additional sources o f financing and r i s k sharing, and contributing to the political sustainability o f i t s projects. The Bank should also promote the inclusion o f new donors in the PPG7 and IDB partnerships on the B r o w n and Blue Agendas. 5. Health M a i n Developments in the Health Sector in the 1990s 48. During the 1990s there was a sustained improvement in major health indicators such as l i f e expectancy and the infant mortality rate, although Brazil s t i l l lags Latin American and Middle Income countries in some areas and there remains room for improvement. The new constitution o f 1988 set the goal o f providing universal, free health care through the health care system (SUS). I t also initiated a progressive decentralization o f administration to municipalities. Financing mechanisms began to be reoriented from production-based payments to capitation, and new legislation was created to guarantee the sustainability o f finance at all levels. Progress was made towards the regulation o f the private health sector through the creation o f an independent agency. The Family Health Program began to be implemented as the framework o f a health care system which emphasizes prevention. 49. Despite these important advances, many challenges lie ahead. The process o f decentralization o f health care i s n o t yet consolidated. Ensuring the financial sustainability o f the system remains a challenge. The expansion o f coverage by the Family Health Program will require special attention for the detection and prompt correction o f errors. Chronic human resource deficiencies will need to be addressed in the public system. The Bank’s Assistance to the Health Sector 50. Bank assistance to health increased significantly in the 1990-2002 period. Projects initiated totaled US$992 million, representing 8 percent o f the Bank’s portfolio. Including three projects started in the late 1980s, the Bank maintained eight active health projects during the 1990s. T w o new projects are set to initiate in 2002-2003: the Family Health Project, and the Bahia Health Project. Bank projects during the period o f evaluation can be categorized into two classes. Class 1 comprises projects dealing with the control o f transmittable diseases (PCDEN, PCMAM, AIDS-I, VIGISUS, AIDS-11), while class 2 comprises projects dealing with health infrastructure, health policy and finance, and aiming ultimately at improving access and quality (NE-I, NE-11, REFORSUS). ‘ 5 1. Class 1 projects began with a focus on control o f specific infectious diseases such as Malaria, Chagas’ Disease, Schistosomiasis and AIDS. They evolved into the strengthening o f the national epidemiologic surveillance system. There were many problems with disbursement and 77 implementation, but ultimately financial and operational goals were achieved. PCDEN and PCMAM made significant contributions to disease control in the North and Northeast, introducing n e w frameworks and important lessons. The AIDS projects were singularly important in mounting an adequate response to the advance o f the H I V / A I D S epidemic, establishing an institutional framework, training human resources and creating indispensable partnerships. VIGISUS has been somewhat impeded by barriers within government, but has s t i l l made important progress in strengthening the surveillance system. 52. Class 2 projects began with a traditional focus on constructing clinics in the Northeast, and evolved t o include infrastructure rehabilitation, provision o f equipment, and managerial and policy contributions to SUS. The Northeast projects had significant disbursement and implementation problems and lessons learnt from the first project were not incorporated into the second. Some difficulties were alleviated when the Bank became more flexible in adapting i t s rules to the reality o f the national situation and to changes in a dynamic health sector. REFORSUS also has not met i t s disbursement goals yet. Inadequate assessment o f implementation capacity and the excessive fragmentation o f resources into many subprojects are part o f the cause. Overall Assessment of the Bank’s Assistance 53. Class 1 Projects: Control o f Transmittable Diseases - The projects aimed at control o f specific infectious diseases had highly relevant objectives which were met with satisfactory success. PCDEN allowed the continuation o f disease control in the Northeast which was threatened by lack o f funds; transmission o f Chagas’ Disease was significantly reduced and important progress was made in controlling Schistosomiasis and treating Kala-Azar. P C M A M made an important contribution to the government’s efforts in Malaria control in the Amazon, which dramatically reduced mortality and incidence, and clearly demonstrated the effectiveness o f emphasizing rapid diagnosis and treatment as opposed to vector control alone. Despite significant external difficulties created within the government for VIGISUS, it has managed to organize, strengthen and enhance the national epidemiologic surveillance system. I t has also played an important and effective role in some aspects o f disease control, and in environmental and indigenous health. These projects adequately target the poor by the very nature o f the diseases they a i m to control. 54. The impact o f these projects on institutional development has been significant at all levels o f government. This included the acquisition o f new managerial capacities and skills in monitoring and evaluation. Also o f importance were innovative actions in infectious disease control such as those supported by P C M A M , which had great impact at the international level. The AIDS projects brought about a more sophisticated approach to prevention as well as new working relationships with institutions o f different categories, most notably NGOs. The sustainability o f the outcomes f r o m the transmittable disease projects i s rated as likely, as many o f the new skills and approaches have been thoroughly assimilated into the institutional culture, although part o f the built up capacity may be lost after project completion. 55. Assessment o f Class 2 Projects: Health Infrastructure, Policy, and Finance. These projects ultimately attained several o f their operational goals, but their impact o n the health o f the population has been only moderately satisfactory. Regional targeting o f resources was satisfactory, with most project resources effectively directed to poorer regions, but i t i s difficult to ascertain the extent to which the project benefited the poor, as there are n o reliable data showing the number o f health interventions by income group over time. A major flaw o f these projects i s that they have aimed to improve access and quality primarily through the strengthening o f health 78 infiastructure and equipment, but without addressing effectively other barriers, such as chronic human resource deficiencies and other financial constraints. These projects share some o f the credit for the tangible improvements in the Northeast, especially the sharp decline in infant mortality, but their impact has been moderate compared to their potential. Class 2 projects also introduced diverse new skills at different levels o f government, and their impact o n institutional development has been substantial. Recommendationsfor Future Assistance 56. General Strategic Approach. The Bank’s assistance has been composed o f a mixture o f policy lending, technical assistance, and investment lending. In Brazil, it would be appropriate to designate a larger role for policy lending and technical assistance. T o prevent delays in disbursement and execution, the Bank should undertake a careful evaluation o f operational and managerial capabilities before finalizing project design, and maintain a reasonable level o f flexibility in procedures. E S W should be more focused and investment loans also more focused and smaller. The Bank should consider addressing key issues such as decentralization, quality o f care, regulation o f the private health sector, and the creation o f mechanisms for equitable distribution o f resources. M o r e difficult issues could be addressed in pilot projects. Direct lending to states could also lead to more effective health system solutions and more receptivity for their implementation. 57. Poverty issues. The Bank has only been partially successful in targeting the poor and offering effective solutions for their problems vis-a-vis health care. Focus o n diseases that by nature affect the poor i s an effective manner o f targeting, but i s not sufficient. I t i s necessary to design projects to deal with the structure o f the public system as a whole and that address specifically the problems and barriers o f the poor. Malnutrition i s one o f the major barriers to better educational outcomes for children. One particular pilot project to consider i s an early childhood nutrition project in the Northeast ultimately aiming to show a concrete improvement in educational outcomes. The Bank also needs to understand and provide input to the evolving approach o f regionalization and the efforts t o bring about equity in health finance. 58. Specific Disease Groups. There are certain disease groups o f high impact which need urgent and focused attention. Non-communicable diseases cause a large share o f mortality in the Brazilian population; non-intentional injuries and certain types o f neoplastic disease would b e prime targets. Cardiovascular disease i s the largest single cause o f mortality and i s susceptible to appropriate and well-designed preventive interventions. Programs for the detection and control o f risk factors could have a potentially huge impact o n the health o f the population. Another area to be considered i s that o f maternal mortality, due to i t s extremely high rate and devastating impact. It would be appropriate to continue assistance in the area o f H I V / A I D S control, but this should gradually taper o f f over the next few years. 6. Education Main Developments in the Education Sector in the 1990s 59. In the early 1990s, political turmoil and failed stabilization program affected adversely all sectors, including education. Since 1995, economic stability and political continuity have provided a fertile ground for education reforms. Among them, one can l i s t the introduction o f evaluation systems for basic, secondary, and higher educations, (SAEB, ENEM and PROVAO), the improvement o f information systems, the establishment o f curriculum guidelines (PCN), the formula for redistributing resources among municipalities, assuring a minimum level o f unit 79 expenditures in all schools (FUNDEF), changes in procurement, quality control and distribution o f textbooks, and the promulgation o f the National Education L a w (LDB), which provides flexibility for reforms. M a n y states and municipalities have also made substantial efforts to improve management and to accommodate the enrollment expansion. 60. There was an impressive gain in educational achievements during the 1990s, particularly in enrollment rates. The net enrollment rate in basic education increased from 84 to 95 percent for the country as a whole and fi-om 72 to 93 percent in the poor Northeast. The net enrollment rates in secondary education doubled to 33 percent. Other indicators such as dropout and retention rates also show impressive improvements. Improving the quality o f education i s a battle that i s j u s t starting, however, as indicated by the poor results o f Brazilian students in national and intemational evaluation tests. The expected income gains among the poor will increasingly depend o n h o w fast leaming outcomes reach acceptable standards. Bank Assistance to the Education Sector 61. Bank assistance to the education sector became an important component o f the overall assistance strategy to Brazil during the 1990s, as indicated by the significant increase in the number and dollar value o f projects, as well as the number o f ESW. Whereas in the 1980s there were only 5 education projects, amounting to US$239 million in commitments, and representing only 2 percent o f the portfolio, after 1990 the Bank launched 12 education projects, amounting to more than US$1,800 million in commitments, and representing more than 11 percent o f the p~rtfolio.~ ’ Bank’s greater attention to the education sector i s also reflected in the body o f The E S W produced-eight studies addressing specific education issues and additional analytical material in poverty assessments and other related studies. 62. The Bank’s E S W covered a broad range o f educational issues, but Bank projects were focused in assisting basic education in poor regions o f the country. The assistance was provided through federal loans covering poor regions (North, Northeast, Center-West) and direct loans to specific Northeastem states (Bahia, Cearh)). In the early 1990s there were also basic education loans to the states o f Paranh, Minas Gerais, and S5o Paulo. f the Banks Assistance Overall Assessment o 63. The overall objectives o f Bank assistance in the education sector were highly relevant. The focus on primary education was correct and the attention t o the poor regions o f the country (especially the Northeast) was also correct, because this double emphasis assured that poor students would be the main beneficiaries. Although focusing o n poor regions tends to tax management capacity, the benefits o f promoting convergence between regions cannot be underestimated. The Bank’s assistance comprised three education projects in higher income states in the first h a l f o f the decade. The selection o f these states can be justified for at least two reasons. First, these projects focused o n primary education, benefiting primarily poor students. Second, the experience acquired in these states was relevant for the Bank-these states had a better institutional capacity t o execute Bank projects and the lessons leamed could be replicated in poorer states. 64. The assessment o f the actual outcomes o f education projects needs t o take into account the evolving conditions o n the country’s and the Bank’s side. As mentioned before, the 1990-94 period was characterized by economic and political turmoil, coupled with the lack o f a clear 3’ These numbers include three projects developed in the late 1990s but only approved in 2001 and 2002. strategy for the education sector. After 1994 the projects benefited from macroeconomic and political stability, a well-articulated strategy for the education sector, and a more efficient project unit in the Ministry o f Education. On the Bank’s side, conditions also improved during the 1990s, with the transfer o f the country management unit to Brasilia and the placement o f education specialists in the Brasilia office, leading to better project design and supervision. 65. The projects reviewed can be divided into two groups. The first group comprises two federal projects focused in the Northeast (the NEBE projects) and a project in Silo Paulo, all initiated in the early 1990s. These projects suffered from design and implementation problems. The degree o f project ownership and absorption capacity were low. All projects in this group required subsequent restructuring and more time to complete the disbursements and achieve their objectives. The NEBEs tried to solve too many education problems in the Northeast in a troubled period. Their inadequate design was reflected in confusing criteria for the choice o f schools, inappropriate financial arrangements for poor states and faulty performance indicators. The Siio Paulo project introduced several pedagogical innovations, such as automatic promotion, but with mixed results. The Siio Paulo project was also overly complex, with school actions and health interventions requiring an implementation capacity unavailable, even in a rich state. 66. The first group o f projects contributed to the increase in enrollment rates and to some other improvements, such as the introduction o f evaluation systems. However, they did not achieve the improvements in management and teaching quality that are essential for improving actual learning outcomes. Possibly it was unrealistic to expect that these achievements could take place at all in a short period o f time, especially in the Northeast. This cluster o f projects i s rated as only moderately satisfactory. 67. The projects pertaining to the second group comprise the sequence o f three FUNDESCOLA projects and the Paranb and Minas Gerais projects. These projects are generally characterized by a strong emphasis on management and o n attaining minimum standards in all schools. FUNDESCOLA i s a federal project covering poor regions o f the country-the North, Northeast and Center-West regions. All schools in urban and peri-urban areas meeting minimum size criteria can participate. The definition o f participating schools has avoided political influences and has forced the rationalization o f the use o f classrooms in the area. The FUNDESCOLA program has set minimum operational standards that all schools have to satisfy, strengthened local empowerment and planning, through the school development plan and greater community participation, promoted institutional development in State and municipal secretariats, and introduced special learning programs such as escola ativa and accelerated classes. The Paranb and Minas Gerais projects have common features and similar performances. In both cases there was strong government ownership as well as good implementation capacity. The two projects proved to be fertile grounds for consistent policies such as school-based management, assessment and information systems, and training for state and municipal teachers. 68. Projects in this second group seem to have improved access and promotion rates, and to be generating gains in management and efficiency that are pre-conditions for better learning outcomes. They represent a very important step in the effort towards better quality, relative to the projects in the first group. For this reason, they are rated as satisfactory. However, the evidence regarding learning i s s t i l l ambiguous, especially in the poor regions, as suggested by SAEB results. These disappointing results may be partly related to natural lags and the enrollment o f poor and disadvantaged children from uneducated parents. However, they may also be due to insufficient attention to teacher training and classroom practices. FUNDESCOLA 1 11, the last project in the FUNDESCOLA program, has started dealing with these issues, but i t i s too early to make an assessment o f this project, as it has been initiated very recently. 81 69. The Bahia and Ceara basic education projects are very recent. Therefore, the evaluation o f these projects also has to be restricted to their design. The two projects expand the area o f coverage o f FUNDESCOLA, with similar components. Actions such as early childhood development, pre-service teacher training and school supervision appear for the first time. Furthermore, the two state projects target the poorest municipalities for some activities, thus reaching directly the rural areas. The design o f the two projects i s satisfactory and reveals an enormous accumulation o f lessons and experience from previous Bank projects. The prospects for favorable outcomes are positive, but close supervision will be required for many o f these activities. 70. Bank projects have contributed significantly to the build-up o f institutional capacity in the education sector, through their support to the implementation o f school autonomy, greater participation o f the communities, the introduction o f information and evaluation systems, and a number o f other innovations that are being mainstreamed by federal and state administrations. There are reasons to believe that the positive outcomes will be sustained in the kture. The education reforms implemented in the 1990s, such as FUNDEF and Bolsa Escola, command broad political support and are likely t o be maintained in the future. The management practices and the innovations implemented by the projects are also likely to be sustained in the future. M a n y states are mainstreaming project initiatives into the regular activities o f the secretariat. The increased participation o f the stakeholders has provided a firm ground for sustaining the projects’ achievements. Recommendationsfor Future Bank Assistance 71. The Bank should elaborate focused E S W on early childhood development, the performance o f automatic promotion policies, the performance o f SAEB and state evaluation systems, the problem o f teacher training and classroom performance, and the problem o f over aged students in secondary education. 72. Regarding future project work, primary education s t i l l deserves support from the Bank, given the l o w levels o f proficiency. The emphasis should shift from management to quality o f teaching and leaming outcomes. The Bank should continue i t s main focus o n the Northeast and other poor regions. Improving access and quality in rural areas may deserve a specific treatment. For this purpose, the targeting o f poorest municipalities in Ceara and Bahia projects may suggest viable strategies. The Bank should consider the use o f smaller Technical Assistance Loans (TALs) and Leaming and Innovation Loans (LILs) in critical areas such as teacher training, and in states facing severe borrowing constraints due to fiscal problems. 73. A pilot project dealing with early childhood development and nutritional issues needs to be considered in the immediate future, especially for the Northeast. The Bank has touched o n secondary education in its recent projects. There i s an increasing space for projects in this sector, as many states face critical shortages o f inputs and inadequate management. The need to focus on quality o f teaching will require increasing supervision efforts from the side o f the Bank. F o r this reason, i t i s recommended that the Bank places a senior education specialist in the Fortaleza or Recife offices. 83 ANNEX 3 An Evaluation of MIGA Activities in Brazil in the 1994-2002 period 1. MIGA’s Strategy in Brazil MIGA’s mandate i s to promote the flows of private investments to developing countries and to complement the activities of IBRD and IFC. To fulfill i t s mandate, MIGA offers insurance to foreign investors against the risks of expropriation, war and civil disturbance, transfer restrictions, and breach of contract. Additionally, MIGA provides technical and advisory services related to foreign direct investment (FDI). The 1997 assistance strategy called on MIGA to focus on infrastructure and manufacturing, while the 2000 strategy indicated that MIGA should also consider guaranteeing investments in services, agribusiness, and S M E projects. The 2000 country strategy noted that MIGA should also support highly developmental projects which would help Brazil reduce i t s current account deficit, as well as insure Brazilian companies investing in other developing countries. Additionally, the country strategy suggested that MIGA should generate additional insurance capacity to meet the demand for risk insurance, in the context of limited capacities. 2. MIGA’s Products: Political Risk Guarantees for FDI As of June 30,2002, Brazil was MIGA’s largest host country, with outstanding gross and net exposures (before and after reinsurance) of US$909 million and US$307 million, respectively, as shown in Table 1. Demand for MIGA guarantees was high during most o f this period, stretching the Agency’s capacity to meet requests for coverage. 32 During 2000-2002, MIGA made extensive use of opportunities to reinsure and coinsure investments with public and private insurers in order to leverage i t s own capacity, especially for large infrastructure projects. This trend i s reflected in the growing difference between gross and net exposure shown in Table 1. Table 1: MIGA Guarantee Projectsand Exposure (in US$ millions, unless otherwise indicated) MIGA’s Portfolio * Brazil was MIGA’s largest client country in fiscal years 1996-1997 and 2000-2002. 32 In M 0 2 , MIGA’s limit on individual country exposure was US$420 million, defined as MIGA’s net exposure plus 10 percent of the amount o f reinsurance to account for the credit risk o f reinsurers. MIGA’s adjusted net exposure on Brazil was US$367.4 million as of June 30,2002, about US$50 million below the country limit. 84 MIGA’s guarantees are estimated to have facilitated US$9.3 billion o f FDI during the 1994-2002 period (Table 1), or roughly 6 percent o f the total accumulated flows o f FDI in the same period. Only 8 percent o f the FDI facilitated by MIGA was related to privatizations, compared with 20 percent for the total flows o f FDI in this period. This implies that the FDI facilitated by MIGA had a greater component o f greenfield investments and contributed more to capital accumulation in this period. Infrastructure projects accounted for a large share o f MIGA guarantees45 percent o f total exposures and 91 percent o f the total FDI insured. Recently, MIGA has also guaranteed a banking project supporting SMEs and mortgage financing and a project in the health sector. The regional distribution o f the portfolio was biased towards the more prosperous regions around Rio de Janeiro and Sao Paulo. However, MIGA insuredthree projects in the poorer northeastern states, in the automotive sector, municipal transportation, and power generation. Table 2: Sectoral Breakdown o f M I G A Guarantees 1 Manufacturing I Finance I Services I Energy 1 Telecom I Transport I Shares in 7 27 21 29 12 4 Exposure Shares in 2 5 2 58 31 2 FDI insured 3. MIGA’s Products: Non-Guarantee Activities MIGA has provided some limited investment marketing services in Brazil whose impacts have not yet been assessed. This has included a forum on new tools and techniques to promote FDI for the heads o f South American investment promotion agencies, and a cooperation agreement with Investe Brusil to participate as a partner in MIGA’s new FDI e-mail alert service that provides users with customized updates on new investment opportunities and market analysis. N o claims have been brought under policies issued by MIGA in Brazil. However, MIGA’s Legal Department has mediated an investment dispute unrelated to a MIGA guarantee, in support o f a Canadian claimant against the state o f Cearh. 4. Assessment o f M I G A Activities MIGA’s activities have generally followed the priorities set out in the CASs. The strong infrastructure component was consistent with World Bank Group’s strategy to attract private capital in order to meet Brazil’s large infrastructure needs, and also consistent with the Government’s liberalization and privatization programs. However, MIGA has been only partially successful in diversifying i t s portfolio into the other sectors identifiedby the 2000 assistance strategy. Facing capacity constraints imposed by country limits, MIGA has leveraged i t s country exposure well. By intensively pursuing cooperation with national and private insurers in reinsurance and coinsurance, MIGA has been able to facilitate large amounts o f additional FDI, while limiting its net exposure. Each dollar o f gross insurance in Brazil facilitated estimated investments o f US$9.5, a very high leverage ratio. However, MIGA does not seem to have adopted clear selection criteria (e.g., developmental impact) or processes according to which it would select projects from a universe o f potential projects under capacity constraints. Although MIGA’s Investment Marketing Services have not been evaluated, the Agency’s decision to devote little resources to investment services activities appears to have been appropriate, given Brazil’s record in attracting large amounts o f FDI. It i s consistent with MIGA’s strategy to focus these activities on countries that attract little FDI and need to develop institutions to target foreign investors. OEU has evaluated seven MIGA projects in Brazil guaranteed between FY94-96: four projects in the financial and three in the manufacturing sector. All seven projects had substantial effects in the development o f human capital through general and specific training courses for their staff. Over 70 percent o f the evaluated projects had high ratings for their downstream effects. For example, a financial project contributed to increasing the availability o f medium-term U S dollar capital equipment financing for the middle market in Brazil. All real sector projects had substantial upstream effects and achieved high development impacts in the dimensions o f technology transfer and know-how and social/infrastructure provision. The evaluated projects did not have a strong direct macroeconomic impact, as their scale was too small to generate sizeable tax revenues or improvements to the balance o f payments. The direct impact o n employment was either negative or negligible, although this result i s not surprising given that entry o f foreign capital usually contributes to sector restructuring. T w o real sector projects had positive impacts in the areas o f environmental performance and corporate citizenship, respectively. The modernization o f an automotive plant included the expansion o f the wastewater treatment system and other improvements to attain I S 0 14001 accreditation. Similarly, a soft drink bottler and distributor launched a successful aluminum can recycling program in partnership with local schools, the local recycler and the municipal government. OEU conducted a survey o f MIGA guarantee holders in Brazil, with a response rate o f 7 1 percent. For 63 percent o f the respondents, MIGA insurance was indispensable for the investment to go forward, another 25 percent considered MIGA coverage important, and 12 percent peripheral. F r o m those investors who responded that MIGA coverage was indispensable, it can be inferred that foreign investments o f US$5 billion would not have gone forward. 5. Lessons and Recommendations Given M I G A ’ s capacity constraints in Brazil, the Agency should continue to leverage its exposure well in order to maximize beneficial FDI flows into Brazil. MIGA should also explore alternative forms o f cooperation to help meet the demand for political risk insurance o f potential foreign investors in Brazil. MIGA should establish transparent criteria for i t s selection o f projects for coverage in Brazil and should give priority to projects with high developmental impact. MIGA also needs to diversify i t s existing portfolio in Brazil to take account o f priority areas identified by the CAS, as well as reduce portfolio risk. 87 ANNEX 4 Government’s Comments on the CAE MINISTERIO DA FAZENDA SECRETARIA DE WSUNTOS INTERN.ACION;US Esplanada dos hlinisltrios Bloc0 P sala 225 Brasilia. DF 70.048-900 Tclefonr (61) 412 2226 E.mail sain.d~~fazmda.goi..hr Letter SAIN no:, IO3 Brasilia, April 16. 2003 Mr. Roberto Rocha Lead Economist OED - W o r l d Bank Washington, D.C. Re: Brazil Country Assistance Evaluation (CAE) Dear Sir, The Country Assistance Evaluation provides a balanced and insightful analysis o f the W o r l d Bank’s assistance t o Brazil over the past decade. W e are generally in agreement with the main conclusions and recommendations o f the report and believe that the report provides important inputs for the next phase o f the Bank’s assistance. At the same time, we would like t o take advantage o f this opportunity t o present the Government’s opinion o n some o f the specific issues pointed out in the report. Although we agree that the Bank’s shift from infrastructure into the social sectors in the 1990s was generally appropriate and justified by the poor social indicators at the start o f the decade, w e also feel that the Bank’s exit from infrastructure was too abrupt, especially in the case o f energy. The document mentions that the Bank could have provided more assistance to the development o f regulatory frameworks (in energy and other infrastructure sectors), but we think that this problem was not sufficiently elaborated in the report. W e appreciate the document’s analysis o f Brazil’s achievements in the social areas during the 1990s, but also believe that the analysis could be improved in some aspects. For example, the shifi t o social sectors followed the Government increased expenditures in this sectors, with the establishment o f measures such as earmaked funds or floors o f expenditures for specific sectors such education and health. The overall consequences for budget administration o f the relationship o f such measures and o f Bank financed projects for the same sectors was not h l l y explored. Another example, the document stresses that income distribution remains highly unequal, despite the reduction in poverty and the improvement in social indicators during the decade. Although Brazil’s income distribution must be improved, most international comparisons o f inequality presented in World Bank documents contain t w o important flaws that worsen Brazil’s ranking relative to other countries. First, the international comparisons frequently involve comparisons of 88 Gini income and Gini consumption coefficients across countries. Countries with income-based coefficients (such as Brazil) tend to fare worse in these comparisons. Second, Brazil’s Gini income coefficient does not capture properly the impact o f Brazil’s income transfer programs o n the incomes o f poor families, and that also tends to overestimate the level o f inequality in Brazil. Considering that Brazil spends substantial amount in income transfer programs, we think that this methodological problem should be taken into consideration in the document and the conclusions revised accordingly. W e also think that the document should do more justice t o the improvements in social indicators by providing comparisons with other countries, particularly in Latin America, Moreover, the document should highlight Brazil’s progress in meeting the Millennium Development Goals. In particular, the report should reflect the fact that some o f these goals have already been met. The document criticizes the Bank for not having provided more assistance t o Brazil in the area o f early childhood development and recommends attention to this area in the future. We agree with this conclusion and the related recommendation. I n this regard, we feel that the Bank may play an important role in the associated area o f maternallreproductive health. W e disagree with the analysis o f sustainability which i s developed and presented in the document, or more precisely with the conclusion that sustainability cannot be evaluated. We believe that the Government has fully demonstrated its commitment to stability and its intention to honor international obligations. W e would also l i k e to point out the sharp improvement in Brazil’s external accounts which has taken place in recent months, and stress that this improvement does not reflect only the effect o f the currency devaluations that occurred after 1998. M o r e importantly, the recent improvements reflect also the effect o f structural reforms that started being implemented in previous years, and that have generated permanent gains in efficiency and competitiveness, leading to a steady increase in the production o f tradable goods (both exportable goods and goods competing with imports). Additionally, the Government i s working hard to present to Congress in the next few months important structural reforms in taxation and social security areas that will strengthen and consolidate even further the solid macroeconomic environment. Therefore, we think that this conclusion should be reviewed and elaborated further. L e t me finalize by congratulating you and your team for producing a good, comprehensive, balanced and insightful report, rich in recommendations that will prove useful for the joint elaboration o f the next country assistance strategy b y the Government and the Bank. Sincerely , -_ - LA-- . 7- -31. ,L 1 \ r C Otaviano Canuto dos Santos Filho Secretary o f International Affairs 89 ANNEX 5 90 91 ANNEX 6 Chairman's Summary: Committee on Development Effectiveness (Meeting o f August 27, 2003) Brazil Country Assistance 1. The Committee on Development Effectiveness (CODE) met on August 27,2003 to solicit members' views on the Country Assistance Evaluation (CAE) for Brazil (R2003-94) as input to the Country Assistance Strategy (CAS) for Brazil, scheduled for Board discussion in October 2003. OEG's Country Impact Review (CIR) for Brazil (CODE2003-53) was circulated as a background document. 2. Background. The C A E evaluated the World Bank assistance program to Brazil during the 1990- 2002 period. I t examined (i) whether the Bank's assistance during this period was relevant; (ii) whether the Bank's assistance program was effectively designed and consistent with i t s objectives; and (iii) whether the Bank's program achieved i t s two central objectives o f poverty reduction and growth. The report concluded that the Bank's assistance strategy in 1990-2002 was relevant overall. The strategy was underpinned by high quality analytical work on poverty and growth, conducted with substantial participation by top Brazilian researchers. The report also found that the shift to the social sectors and the focus on the Northeast were justified - given the very weak social indicators in the Northeast regron - as was the shift to adjustment lending at the end o f the decade. It noted that when the Bank made adjustment and techmcal assistance loans, it maintained assistance to the social sectors. Government's comments, and management's response to the C A E were included in the annexes. 3. The CIR reviewed IFC's operations in Brazil from FY90 to FY02. I t concluded that IFC made a strong contribution to Brazil's sustainable development through financing many viable and pioneering projects that have substantial benefits on the ground. Overall within the large IFC portfolio, the success rates for development and combined development-and-investmentoutcomes were somewhat better than in other countries. However, IFC's returns on assets in Brazil have been low and fallen well below the average corresponding returns for the rest o f IFC's portfolio. Key issues identified for further attention were corporate governance reform, strengthening the financial sector, and improving the profitability o f IFC activities. 4. Overall Conclusions and Next Steps. Members welcomed the timing o f the discussion o f the C A E prior to Board consideration o f the Brazil CAS. They broadly concurred with the CAE's conclusion that the Bank's assistance strategy in 1990-2002 was relevant overall, and that the shift to the social sectors and the focus on the Northeast were largely justified. At the same time, there were some differences o f views on the ratings, with some members suggesting that the assessment was too positive and uncritical, and others indicating that it could have been even more positive. The Committee's discussion focused on the issues o f ESW, lending services, growth, development impact, consultation process, the Bank's role in middle income countries, and the review o f IFC's operations. Some members felt that the CIR deserved fuller attention from members, and i t was agreed that the C O D E Subcommittee would schedule a separate discussion o f the CIR. 92 ESW. Members h i g h l i g h t e d the importance o f ESW, and agreed that i t was generally o f high quality. They also agreed with the finding o f the study that better dissemination o f ESW could help the Bank provide more effective assistance to Brazil. A key step would be to ensure that the Bank's work i s available in Portuguese. However, several members raised questions with regard to the abundance and relevance o f ESW and suggested a more critical appraisal o f the program based on the country's demand for it. Lending services. Members agreed that programmatic adjustment lending was warranted, to address critically needed reforms. They supported OED's recommendation that the Bank continue to further build on i t s approach on programmatic lending, particularly with respect to establishment o f objectives linked to real economic outcomes, quantified and time bound benchmarks, and a framework for measuring and assessing outcomes and benchmarks. Some members felt that the Bank's assistance to infrastructure should be maintained or scaled up, particularly in regard to regulatory and environmental aspects. A member agreed with the Government view that the Bank's exit from infrastructure may have been too abrupt. Members also supported the shift from a sectoral to a more regional approach, and noted that the method f o r selecting the regions to receive the Bank's attention will be crucial. Growth. Members noted the report's finding that the component o f the program designed to stimulate investment, economic efficiency, and growth, produced mixed results. Brazil's growth performance in the 1990s was generally disappointing considering that several structural reforms were implemented during the decade. Some members suggested that the Bank should focus its support more o n the growth agenda, and expressed the v i e w that B r a z i l w i l l n o t b e able to sustain the necessary expansion o f coverage and improvement o f quality o f social services without a higher level o f economic growth. A suggestion was made that the Bank pay attention to the qualitative determinants o f growth and increase its focus on secondary and higher education. Development impact. Members commented on the complexity o f attribution and the d i f f i c u l t y o f assessing development impact o f the Bank's assistance. They acknowledged the efforts made by the evaluation team in this regard, but suggested that further work and refinement was needed to assess more thoroughly the Bank's contribution. Members also encouraged the Bank to bring to bear i t s experiential knowledge and analytical resources to demonstrate the cost-benefit o f reforms. The importance o f addressing the issue o f income inequality to further the poverty reduction agenda was highlighted. Consultation process. Some members agreed with the Government's position that the NGOs' involvement should be limited to project implementation rather than formulation o f assistance strategies. Another member, while concurring with the view that Government bears the responsibility for policy and strategy formulation, suggested that broad public consultations i n f o r m and improve policy and strategy formulation and increase ownership, and urged the Bank and the Government to consult fully with the public in formulation o f the new CAS. Bank role in middle income countries. Members agreed that the Bank can s t i l l play a relevant role in large middle income countries such as Brazil. The Bank i s most relevant as a source o f knowledge and technical assistance which could also enhance i t s relevance 93 as a source o f finance. In this context, a member highlighted the need to address the issue o f IBRD competitiveness and IFC's l o w profitability. IFC support for second-tier companies. In response to a member's recommendation o f strong promotion o f I F C support for rigorously screened second-tier companies, OEG indicated that screening w o u l d have to b e dramatically tighter to raise the current 16% win-win rate to the 55% achieved for the 53 evaluated first-tier company projects. Demonstration effects o f second-tier company lending can be valuable, and the preferred instrument for supporting them i s wholesaling via credit lines with banks. However, for that altemative to be feasible, there needs t o be a systemic solution to what is a systemic problem o f public sector borrowing requirements dnving up interest rates and driving out banks' appetite for lending to corporates. IFC equity performance. A member suggested that the depreciation o f the Real since 1998 played a central role in IFC's poor equity performance. OEG responded that the impact o f the currency decline was indeed significantly negative, as it had also been in both Argentina and Mexico over the same period, but it had been compounded by other Brazil- specific factors described in the report. F o r example, the sectoral mix o f I F C investments was both unfavorable and more vulnerable to currency losses. Finn Jonck, Chairman 95 REFERENCES Abreu, M., and D. Verner. 1997. Long-Term Brazilian Economic Growth: 1930-94. Organization for Economic Cooperation and Development, Paris. Almeida dos Reis, J., and R. Barros. 1992. “Desigualdade Salarial e Distribuiqiio da Educaqiio: Um Estudo da Evoluqiio das Diferencas Regionais da Desigualdade no Brasil Metropolitano.” Trimestre Econdmico, (May): 195-229. Anuatti, F., M. Barossi, A. Carvalho, and R. Macedo. 2002. “Financial and Operational Performance of Newly Privatized Firms in Brazil: A Panel Data Analysis.” Unpublished Manuscript. FPE, Universidade de Siio Paulo. Bacha, E., and R. Bonelli. 2001. “Crescimento e Produtividade no Brasil: 0 que nos diz o Registro de Longo Prazo.” Unpublished manuscript, PEA, Rio de Janeiro. Barros, R., R. Henriques, and R. Mendonga. 2000. “Pel0 Fim das DCcadas Perdidas: EducaqBo e Desenvolvimento Sustentado no Brasil.” In R. Henriques, ed., Desigualdade e Pobreza no Brasil, IPEA: Rio de Janeiro. Bonelli, R. 2002. “Uma Nota sobre Retomos A Educaqiio no Brasil.” Unpublished Manuscript, IPEA. Rio de Janeiro. Bonelli, R., and R. Fonseca. 1998. “Ganhos de Produtividade e de EficiCncia: Novos Resultados para a Economia Brasileira.” Pesquisa e Planejamento EconGmico 28 (2). Cardoso, E. 1996. “Lessons from the Mexican Crisis for Reforming Economies.” Cairo: Egyptian Center for Economic Studies, Distinguished Lecture Series, March 1996. Cardoso, E. 2000. “Brazil’s Currency Crisis: The Shift from an Exchange Rate Anchor to a Flexible Regime.“ I n C. Wise and R. Roett, eds., Exchange Rate Politics in Latin America. Washington, D.C.: Brookings Institution Press, pp. 70-92. Cardoso, E., R. Paes de Barros, and A. Urani. 1995. “Inflation and Unemployment as Determinants of Inequality in Brazil: the 1980s.” Chapter 5 in Dombusch, R. and Edwards, S., eds., Reform, Recovery and Growth: Latin America and the Middle-East. Chicago: University o f Chicago Press for the NBER. Carpena, L., and J. Oliveira. 2002. “Estimativa do Estoque de Capital Humano para o Brasil: 1981 a 1999.” Unpublished manuscript. PEA. Rio de Janeiro. Castro, N. 2002. “Estrutura, Desempenho, e Perspectivasdo Transporte Ferroviirio de Carga.” Pesquisa e Planejamento Econdmico 32 (2). Castro, N. 2001. “Privatizaqiio e Regulaqiio dos Transportes no Brad.” I nCaixeta, J. and R. Martins, eds., Logistics do Transporte de Carga. Editora Atlas. Rio de Janeiro. Easterly and Levine. 2000. “It’s Not Factor Accumulation: Stylized Facts and Growth Models.” The World Bank Economic Review 15 (2). Washington, D.C.: World Bank. 96 Elbers, C., J. Lanjouw, P. Lanjouw, and P. Leite. 2001. “Poverty and Inequality in Brazil: New Estimates from the Combined PPV-PNAD Data.” UnpublishedManuscript. Vrije University Amsterdam, Yale University, New Haven, Catholic University, Rio de Janeiro. Ferreira, F., and J. Litchfield. 2001. “Education or Inflation?: The Micro and Macroeconomics of the Brazilian Income Distribution during 1981-1995.” Cuadernos de Economia 38 (114): 209- 238. Ferreira, P. C., and T. Malliagros. 1999. “Investimentos, Fontes de Financiamento e EvolugPo do Setor de Infra-Estrutura no Brasil: 1950-1996.” Ensaios Econdmicos N. 346, EPGE-FGV, Rio de Janeiro. Giambiagi, F. 2002. “Do Deficit de Metas i s Metas de Deficit: A Politica Fiscal do Govemo Fernando Henrique Cardoso - 1995/2002.” Working Paper No. 93. BNDES, Rio de Janeiro. Giambiagi, F. 2002). “Restrigks ao Crescimento da Economia Brasileira: Uma VisPo de Longo- Prazo”. Revista do BNDES 9 (17). Rio de Janeiro. Gordon, L. 2001. Brazil’s Second Chance, En Route to the First World. Washington, D.C: Brookings. Krueger, A., and M. Lindahl. 2001. “Education for Growth: Why and for Whom?.” Journal of Economic Literature, March. Li, H., L. Squire, and H. Zou. 1998. “Explaining International and Intertemporal Variations in Income Inequality.” The Economic Joumal, 108 (January): 26-43. Menezes-Filho, N. 2001. “A Evoluglo da Educaglo no Brasil e seu Impact0 no Mercado de Trabalho.” Instituto Futuro Brasil, Rio de Janeiro. Neri, M. 1999. “A Note on the Growth Sensitivity of Poverty across Brazilian States.” IPEA, Rio de Janeiro. Neri, M. 1995. “Sobre a Mensuraglo dos Salirios Reais em Alta InflagPo.” Pesquisa e Planejamento Economico 2 (3): 886-902. OED (Operations Evaluation Department). 2000. Brazil, Forests in the Balance: Challenges of Conservation with Development. Washington, D.C.: World Bank. , 1998. The Brazil Health System. Washington, D.C.: World Bank. Pinheiro, A. 2000. “Apbs a Privatizaglo.” Unpublished Notes. BNDES, Rio de Janeiro. Pinheiro, A. 2001. “0Judiciirio e a Economia na VisPo dos Magistrados.” IDESP, Rio de Janeiro. Pinheiro, A., IGill, L. Serven, and M. Thomas. 2001. “Brazilian Economic Growth, 1900-2000, . Lessons and Policy Implications.’’ Unpublished manuscript. BNDES, Rio de Janeiro. Pritchet, L. 2001. “Where Has all the Education Gone?’ The World Bunk Economic Review 15 (2). Washington, D.C.: World Bank. 97 Ravallion, M. 200 1. “Growth, Inequality, and Poverty: Looking Beyond Averages.” World Development, November. Scheinkman, J. et al. 2002. “A Agenda Perdida: Diagndsticos e Propostas para Retomada do Crescimento com Maior JustiGa Social.” Unpublished manuscript. Rio de Janeiro. Schneider et al. 2001. “Amazonia Sustentivel: Limites e Oportunidades para o Desenvolvimento Rural.” Serie Parcerias 1, IMAZON/The World Bank, Brasilia, 2000. Temple, J. 1999. “The New Growth Evidence.” Joumal of Economic Literature 37 (1): 112-156. Tyler, W. 1999. ‘The Pursuit of Monetary Policies under the Transition from High Inflation to Stabilization: The Experience of Brazil.” Paper prepared for the Conference “The Economic Issues Facing the New Indonesian Government” Jakarta, August 18-19, 1999. United Nations Commission for Latin America-ECLAC. 2002. Social Panorama of Latin America, 2001-2001. Santiago de Chile. UNDP (United Nations Development Program). 2002. Human Development Report. New York. US Congress. 2000. International Financial Institution Advisory Committee. The Meltzer Commission Report. Washington, D.C. Williamson, J. 2002. “Is Brazil Next?’ International Economics Policy Briefs, Institute for International Economics, Washington, D.C. (August). Williamson, J. 2003. “Lula’s Brazil.” Foreign Afsairs 82 (1).