A case study from Reducing Poverty, Sustaining Growth--What Works, What Doesn't, and Why A Global Exchange for Scaling Up Success Scaling Up Poverty Reduction: A Global Learning Process and Conference Shanghai, May 25­27, 2004 Successes and Failures in Poverty Eradication: Chile Alejandro Foxley foxletsen@senado.cl (562) 236-6002 The findings, interpretations, and conclusions expressed here are those of the author(s) and do not necessarily reflect the views of the Board of Executive Directors of the World Bank or the governments they represent. The World Bank cannot guarantee the accuracy of the data included in this work. Copyright © 2004. The International Bank for Reconstruction and Development / THE WORLD BANK All rights reserved. The material in this work is copyrighted. No part of this work may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying, recording, or inclusion in any information storage and retrieval system, without the prior written permission of the World Bank. The World Bank encourages dissemination of its work and will normally grant permission promptly. SUCCESSES AND FAILURES IN POVERTY ERADICATION: CHILE Executive Summary Poverty eradication was at the core of development strategies in Chile during the Aylwin, Frei, and Lagos administrations of the 1990s. The social policies of all three governments were formed by a particularly difficult beginning. The Chilean economy went through a deep financial crisis in 1982, similar in severity to Argentina's crisis of 2001. In 1982 and 1983, Chile's GDP fell by 16 percent. The collapse of the financial sector cost Chilean taxpayers between 30 and 40 percent of GDP. Unemployment shot up to 30 percent. Around 50 percent of the population fell below the poverty line. Extreme poverty affected 30 percent of the population. Starting in 1985, the focus of economic policies shifted toward financial solvency and economic growth. Exports grew rapidly and unemployment went down. On the poverty front, however, results were less successful. People living below the poverty line still represented 45 percent of the population in 1987. Additionally, a key decision by the Pinochet government to reduce taxes and government expenditures in 1988 had a further negative impact in social policies. The decrease in social expenditures was equivalent to 3 percent of GDP, resulting in severe deterioration in the coverage and quality of public health services, lower wages for teachers, and lower pensions for the elderly. Growth with equity It was against this background that the newly elected government of Patricio Aylwin came into office in 1990. The new government immediately put the fight against poverty at the top of the agenda. Chile's new "growth with equity" development strategy was based on an explicit pro- growth strategy balanced by aggressive social policies; the strategy continued through the decade. Between 1990 and 2000, poverty was reduced from 40 percent of the population to 20 percent. The pro-growth strategy included: · Opening the economy to world trade · Conservative fiscal policy pursuing simultaneously a budget surplus and reduction of public debt · Expanding of the domestic capital market · Reform of labor and tax policies, including a tax increase that allowed the government to expand social expenditures by more than 200 percent through the 1990s The strategy enabled the Chilean economy to growth at 6 percent per year during the decade. The combination of high growth and active redistributive policies reduced by nearly one-half the extreme income disparity between the top 20 percent and the lowest 20 percent of the population. An empirical study comparing social policies in the 1980s and 1990s shows that about 60 percent of Chile's poverty eradication in the 1990s can be attributed to economic growth and 40 percent to social policies. The "easy" first phase represents a combination of high growth, 1 CASE STUDIES IN SCALING UP POVERTY REDUCTION increased wages and minimum wage, expansion in income-support schemes for low-income families, and improved minimum pensions with an immediate impact on poverty reduction. The shift from income-support schemes towards social investment aimed at upgrading education, skills, and access to health services produced a much more gradual effect on the incidence of poverty. The initial expansion in funding for public health and education did have an impact, as schools and new hospitals were built. Teachers, doctors, and health care workers received better pay, and coverage slowly expanded. But after a few years, pouring more money into inefficient public health services or unaccountable school systems did not result in the expected rate of poverty reduction. The slowdown in the second half of the 1990s had to do with decreasing returns to large expenditures in public health and education, after basic coverage problems were solved. Rigid, highly centralized, bureaucratic institutional arrangements in public hospitals, combined with active resistance from doctors and public health workers, resulted in great inefficiency. Though government expenditures in public health increased by 250 percent, output of healthcare services grew only by 22 percent. A similar outcome was observed in education. Following a tripling of public resources, learning scores initially surged, but did not continue to improve. A significant part of this disappointing result has to do with the poor quality of teaching in the classroom. The National Teachers Unions succeeded in preventing performance evaluations of teachers or school administrators for 12 years. The lessons are that money does not guarantee effectiveness and that more attention should be paid to the political economy of institutional changes required to ensure cost-effective provision of basic health and education services. Institutional innovation As a complement to the poverty-eradication effort, an experimental program allowed community- based organizations to apply for public funds to develop projects to improve neighborhood infrastructure or start micro-enterprises. The program focused on the indigent population. To carry it forward a new institution was created: the Solidarity and Social Investment Fund (FOSIS). FOSIS has been successful in reaching the target group, but the coverage has been limited. As programs for the extremely poor were scaled up, intersectoral coordination emerged as a serious problem. But a more serious problem is that the percentage of the very poor reached by the programs has been small. In fact, eradicating extreme poverty has proved to be much more difficult than improving the lot of people near the poverty line. The public, meanwhile, sees a very unequal society unable to deal with hard-core poverty, engendering feelings of frustration and futility. Two approaches were developed to solve the scale-up problem. The first was the National Program of Poverty Eradication. In 1994, 80 municipal districts were selected on the basis of their high poverty ratios. Municipalities were asked to coordinate an effort to reach the poorest families. The program had some limited success, although sectoral coordination among 2 SUCCESSES AND FAILURES IN POVERTY ERADICATION: CHILE ministries proved difficult. Municipalities complained about lack of resources, and those municipal districts left out of the program complained about discrimination. A new approach was started in 2002. The Chile Solidario program targets the extremely poor in the country. It implements a decentralized approach where families are closely monitored by local public officials to make sure that they get effective access to income-support subsidies and public services, plus help in finding jobs or training for the head of household. It is too early to reach any conclusions about the success of the program, although preliminary evaluations do identify strong and weak aspects of the new approach. Internal and external catalysts It is remarkable that Chile's poverty ratio continued to decline after 1998, in spite of the fact that the economy was in recession. The countercyclical social expenditure policy was possible because of the budget surplus achieved throughout the decade. Public savings in good years could be used to protect the poor in bad years. This is where conservative fiscal policies and progressive social policies meet. The transition from a traditional approach of extending coverage in public services toward an integrated, family and community-centered poverty eradication effort was heavily influenced by holistic approaches developed by the World Bank. [identify] These were the result of theoretical contributions and hands-on experience by World Bank staff. In the case of Chile these contributions have proved to be a key factor in a learning process, still underway, to make antipoverty strategies more effective, particularly after the first easy phase has been completed. 3 CASE STUDIES IN SCALING UP POVERTY REDUCTION The Social Impact of a Pro-Growth Strategy Poverty eradication has been at the core of development strategies in Chile during the 1990s. The Aylwin, Frei and Lagos administrations have persisted in the fight against poverty as a high priority policy goal. We will explore in this paper the nature of the policies undertaken to reduce poverty, evaluate the results and draw lessons from what worked or did not work in the antipoverty policy approach pursued for more than a decade now. The social policies of the 1990s were heavily influenced by a particularly adverse starting point. The Chilean economy had gone through a deep financial crisis in 1982. When looking at the relevant figures, what comes to mind is the Argentina crisis of 2001, as a similar type of economic shock. GDP fell by 16 percent in 1982 and 1983 in Chile. The collapse of the financial sector cost Chilean taxpayers between 30 and 40 percent of GDP1. Unemployment shot up to 30 percent. Around 50 percent of the population went below the poverty line, a figure not unlike Argentina's in the early 2000s. Extreme poverty affected 30 percent of the population 2. Starting in 1985, a new Minister of Finance shifted the focus of economic policies towards recuperating financial solvency and economic growth. The effort was successful. Exports grew rapidly and unemployment went down. However, on the poverty front, results were less successful. People living below a poverty line still represented 45 percent of the population in 1987. Additionally, a key decision, that of reducing taxes and government expenditures in 1988, had a further negative impact in social policies. As shown in Table 1, a tax reduction of 4 percent of GDP undertaken by the Pinochet government in the late 1980s, forced a decrease in social expenditures equivalent to 3 percent of GDP. This meant a severe deterioration in the coverage and quality of public health services, lower wages for teachers and lower pensions for the elderly. Table 1. Taxes and Social Expenditures in Chile, as a percent of GDP 1987 1990 Net tax revenues 17.9 14.0 Government Social Expenditures 15.2 12.4 Total Govt. Expenditures 25.3 20.2 Source: Ministerio de Hacienda Chile (2002), Informe de Finanzas Públicas. It is against this background that the newly elected democratic government of Patricio Aylwin came into office in 1990. The new government immediately put the fight against poverty at the top of the agenda, as part of a "growth with equity", development strategy. 1G. Held and L. F. Jimenez (2001) 2D. Raczynski (1998), J. P. Arellano (1995). 4 SUCCESSES AND FAILURES IN POVERTY ERADICATION: CHILE In thus defining the government programme, growth with equity, the new administration was making a strong statement: a successful poverty eradication would not be possible unless it was based on an unambiguous pro-growth strategy but one that would be accompanied by active social policies. As described elsewhere3, the pro-growth policies included a commitment to a permanent budget surplus, to retiring government debt as an insurance against future financial shocks, as that experienced in 1982; to a further unilateral reduction of external tariffs that would be followed by free trade agreements to be negotiated with Chile's main trading partners, including Latin American countries, the European Union, the United States and Canada. The pro-growth policies also considered a deepening of the capital market, by allowing for a gradual diversification of financial instruments, where privatized pension funds and other financial intermediaries could invest in. All these policies were pursued with vigour and continuity by the three Center-left governments of Aylwin, Frei and Lagos. The pro-growth agenda produced generally good results, as shown in Tables 2 and 3. In fact, the Chilean economy grew by more than 6 percent a year for the decade, in spite of a slowdown caused by the Asian crisis in 1998 and 1999. Productivity grew at a respectable 4.4 percent annually. Real wages increased by 3.3 a year. The rate of investment reached an all time high. The same is true of domestic savings. The budget surplus objective was achieved. It reached 1.3 percent of GDP as an average for the decade. Table 2. Chile Macroeconomic Results, 1990­2000 1990 ­ 2000 Annual growth rate 6.1 Domestic saving, percent GDP 22.0 Investment, percent GDP 25.0 Productivity growth, per year 4.4 Real wage increase, per year 3.3 Employment growth, per year 1.7 Government savings percent GDP 4.5 Budget surplus, percent GDP 1.3 Source: Banco Central de Chile (2001), Indicadores Económicos y sociales de Chile 1960 ­ 2000 3A. Foxley (2003) 5 CASE STUDIES IN SCALING UP POVERTY REDUCTION Table 3. Chile, Economic Indicators, 1990­2000 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 Annual growth 3,7 8,0 12,3 7,0 5,7 10,6 7,4 6,6 3,2 -1,1 4,2 rate Inflation 27,3 18,7 12,7 12,2 8,9 8,2 6,6 6,0 4,7 2,3 4,5 Unemployment 7,8 8,2 6,7 6,5 7,8 7,4 6,5 6,1 6,2 9,7 9,2 Budget surplus 0,8 1,6 2,3 2,0 1,7 2,6 2,3 2,0 0,4 -1,5 0,1 Public debt over 43,0 37,0 30,6 28,4 23,1 17,7 15,1 13,2 12,5 13,8 13,7 GDP Source: Banco Central de Chile, 2000, Indicadores Económicos y Sociales de Chile 1960 ­ 2000 The robust growth performance provided a solid foundation for the anti-poverty objective. Both real wages and employment increased significantly. New job opportunities for the unemployed and higher disposable incomes had an immediate impact on poverty reduction. But such a positive result had to be reinforced by aggressive social policies designed to further improve the lot of the poor. The financing of such an effort could only be achieved by redressing the drastic tax reduction undertaken in the later years of the Pinochet regime. A tax reform was proposed in 1990 increasing taxes across the board, including corporate and personal income taxes, the value added tax and reducing tax evasion in sectors like the farm sector. The tax reform passed the Congress by a large majority in less than two months4. As a consequence, tax collection increased by 3 percent of GDP in subsequent years. The government committed itself to spend a hundred percent of the additional revenues in the social sectors. The commitment was fulfilled, with a dramatic reduction in the number of families living below the poverty level. Indigence also decreased sharply. Both these results are shown in Table 4. Table 4. Poverty and Indigence as percentage of Chile's population, 1987­2000 1987 1990 1992 1994 1996 1998 2000 Poverty 45.1 38.6 32.6 7.5 23.2 21.7 20.6 Indigence 17.4 12.9 8.8 7.6 5.8 5.6 5.7 Poor non indigent 27.7 25.7 23.8 19.9 17.5 16.1 14.9 Source: Banco Central de Chile (2001), Indicadores Económicos y Sociales 1960 ­ 2000 and, Mideplan (2001), Pobreza e Indigencia e Impacto del Gasto Social en la Calidad de Vida. These results are the most successful when compared with other Latin American countries in a similar period of time as described in Table 5. The results also compare favourably with similar efforts undertaken by countries such as Indonesia, Malaysia, Pakistan and others5. 4A. Foxley (2003) 6 SUCCESSES AND FAILURES IN POVERTY ERADICATION: CHILE Table 5. Poverty Indicators in Latin America in the 1990s (percent of population) 1990 1999 Brazil 48 38 Chile (2000) 39 20 Colombia (1991) 56 55 Costa Rica 26 20 Guatemala (1998) 63 60 Mexico (1989, 1998 48 47 Venezuela 40 49 Latin-America 48 44 Source: A. Arenas and P. Benavides (2003), Protección social en Chile, published by International Labor Office, Santiago, Chile Antipoverty Strategies Having inherited deep social imbalances and massive poverty, the Aylwin government took strong initial action to provide a minimum income to poorer families. The minimum wage was increased by 17 percent in real terms between 1989 and 1991 . Family allowances and various 6 subsidies for disadvantaged families, for disabilities and for the unemployed were also increased, as well as pensions. The next step was to expand social expenditures in public health, housing for the poor and education. The public health system had been seriously neglected by the military regime in the hope of forcing a complete privatization. Hospitals had to be provided with basic equipment, including bed sheets, heating and standard drugs required by patients. Investment in public hospitals and primary care units increased from US$ 10 million a year to US$ 100 million annually during the Aylwin administration. Sharp increases in primary health care subsidies administered by municipalities were implemented, as well as a significant improvement in salaries paid to hospital personnel. However, no changes were made to highly centralized institutional arrangements and deficient management practices in public hospitals, something that would be regretted at a later stage. The initial effort to increase the coverage and quality of public health was continued under Presidents Frei and Lagos. In a ten year period, public expenditures in health had increased by 250 percent, as shown in Table 6. 5K. Cowan and J. De Gregorio (1996) 6A. Mizala and P. Romaguera (2002) 7 CASE STUDIES IN SCALING UP POVERTY REDUCTION Table 6. Social Expenditures, 1990­2000 Annual growth rate Accumulated growth rate Total social expenditures 8.0 214 Health 9.4 247 Housing 5.1 164 Social Security 6.1 181 Education 10.6 274 Subsidies 5.9 177 Other 13.3 346 Source: Ministerio de Hacienda (2003), Estadísticas de las Finanzas Públicas, fothcoming. Housing subsidies were also expanded by 160 percent in the decade, with an emphasis in the lower income quintile, in rural communities and in urban renewal programs. But where the main effort was concentrated, was in improving the quality and equity in education. Public resources increased almost three fold in the decade. The notion was to improve on the coverage of education by including pre-school children, and to upgrade the quality of primary and secondary education . 7 The first stage in educational reform involved concentrating resources on the 10 percent of primary schools with poorest learning outcomes. Funds were used to repair schools, improve teaching methods, provide access to books and, later, to computers, as well as reinforcing the school lunches program. After a few years, the two thousand schools in the program, had an improvement of 12 percent in the reading and math skills scores, vis a vis a 9 percent in the rest of municipal schools, a modest but positive result. This was followed by a systemic intervention at the primary and secondary school levels. The effort included all schools public and private, oriented towards low income families. Upgrading teaching materials, libraries and networking schools through Internet in the Enlaces program was part of the effort. The rest consisted in providing funds for school or teacher generated innovative projects. The idea was to induce changes in teaching methods from below. A third phase in education reform extended the school day to 8 pedagogical hours, 1200 hours a year, in order to allow for a more sustained effort at improving the quality of education in poor schools. The whole process was accompanied by huge wage increases for teachers, more than doubling their salaries in real terms in ten years. However, teachers and particularly the National Teachers Union were reluctant partners in the effort. The Teachers Union forced the government to accept a Teachers Statute in 1991, that essentially gave tenure to teachers and school administrators, irrespective of their performance, and established a highly centralized national collective bargaining process for teachers salaries. Teachers could not be fired even if school enrolment fell. And they would be paid identical salaries irrespective of local conditions. 7C. Cox and N. J. Lemaitre (1999). 8 SUCCESSES AND FAILURES IN POVERTY ERADICATION: CHILE Income distribution The real dimension of the social expenditures effort can be appraised in Table 6. Social expenditures more than doubled in the decade. As many incidence studies have demonstrated, the impact of these programs in reducing income disparities was very significant. The World Bank concludes that by 1996, a monetary income differential of 1 to 20 between the poorer 20 percent of the population and the 20 percent richer had been reduced to 1 to 11 because of the redistributive effects of social expenditures8. In fact, income distribution when considering only monetary income, improved between 1990 and 1994 but had deteriorated by 1998, as shown in Table 7. However, when income distribution is adjusted for cash and in-kind transfers, the situation drastically improves, as indicated in Table 8. Anti-poverty programs had a clear redistributive impact. Table 7. Income Shares per Quintile, Chile, 1987, 1994, and 1998 Quintile 1987 1994 1998 1 3.7 4.0 3.1 2 7.1 7.5 6.7 3 10.9 11.5 10.8 4 17.9 18.5 18.3 5 60.2 58.5 61.1 Source: World Bank "Chile's High Growth Economy, Poverty and Income Distribution, 1987-1988", World Bank, 2002. Table 8. Income Distribution adjusted for Cash and In-Kind Transfers, Chile, 1998 Quintile Monetary Income Total, incl. Transfers 1 3.1 5.2 2 6.7 8.2 3 10.8 11.6 4 18.3 18.0 5 61.1 57.0 Q5/Q1 20.0 11.1 Source: World Bank (2002), op. cit. A key to understanding the rapid reduction in poverty ratios in Chile, lies in the balanced nature of the equation: an active pro-growth agenda, and aggressive social policies focused on poverty reduction. Studies by Meller9 and by Cowan and De Gregorio ask the following question: did active social policies make a difference in the case of Chile or is good growth performance the sole driving force behind poverty reduction? These authors compare the effectiveness of each 8World Bank (2002) 9P. Meller (1999), K. Cowan and J. De Gregorio (1996) 9 CASE STUDIES IN SCALING UP POVERTY REDUCTION additional point in growth rates, on the percentage of poverty reduction achieved with two different growth strategies: the trickle-down strategy of the 1980s, as compared to the growth with equity strategy of the 1990s. The conclusion is that active social policies do matter. In fact they explain 40 percent of poverty reduction. Economic growth explains the rest . Another way of looking at these results is that, with social expenditures maintained at the level of the late 1980s, and identical growth rates as experienced in the 1990s, poverty in 1996 would have been 28 percent of the population, instead of the 23 percent actually achieved. Social policies did make a difference. Deacceleration in Poverty Reduction: Why? After a first impulse characterized by a fast decrease in the number of poor between 1990 and 1996, the process slowed down considerably thereafter, as shown in Table 9. The "easy" first phase represents a combination of high growth, increased wages and minimum wage, expansion in income supporting schemes for low income families, and improved minimum pensions with an immediate impact on poverty reduction. . The shift from income supporting schemes towards social investment, in order to upgrade education, skills and access to good health services on the part of the poor, produces a much more gradual effect on the reduction of poverty. The initial expansion in funding for public health and education does have an impact. Schools and new hospitals are built. Teachers, doctors and health care workers get a better pay. Coverage slowly expands. But after a few years, pouring more money into inefficient public health services, or unaccountable school administrators, does not improve the poverty reduction impact of the first upsurge in increased social expenditures. Table 9. Poverty Reduction (thousands per year) 1990 1992 1994 1996 1998 2000 Poor -536 -634 -551 -492 -128 -79 Extremely poor -466 -490 -133 -223 -7 +29 Source: Banco Central (2001) and Mideplan (2001), op. cit. Government expenditures on the public health sector increased by 190 percent between 1990 and 1999 in Chile, and yet health services increased by only 22 percent . A good fraction 10 of this unsatisfactory result is explained by inefficient, bureaucratic management of public hospitals, rigidity in hiring and firing practices (tenured jobs), and lack of incentives. Similar results are observed in the education reform outcomes. Expenditures increased by almost 300 percent in a decade, and yet that effort did not translate in higher scores in math and language skills. After a 10 percent increase in learning scores in the 1990-1996 period, scores stagnate in the 1996 ­ 2000 period . 11 10J. Rodríguez and M. Tockman (2000) and H. Bayer (2001) 11C. Cox and M. Lemaitre (1999) and Ministerio de Educación, Chile (2002) 10 SUCCESSES AND FAILURES IN POVERTY ERADICATION: CHILE The results suggest that for social sector reforms to be effective in reducing poverty ratios, a lot of attention has to be paid to the political economy of social reforms. To make the story short, public health workers, doctors employed in public hospitals and teachers often represents conservative forces, resisting efficiency oriented changes in the way hospitals and schools are run. Doctor and health care labor unions have consistently opposed health sector reform in Chile for 12 years, and governments have not been willing to endure work stoppages in hospital and emergency care units, with the consequent political cost attached to it. It took three governments and 12 years for teachers to accept an external evaluation of teacher performance, and agree on very timid and slow procedures for a removal of the worst performers. And yet a study by Brunner and Elacqua12 concludes that two thirds of learning performance differentials in Chilean schools, is explained by the quality of teachers, their previous training and quality of teaching methods. On the other hand, the extraordinary rigidities vis a vis wage negotiations, and de-facto guarantees of lifelong employment in the teaching profession accepted by the government in the Teachers Statute imposed by the National Teachers Union in 1990, has proved lethal for a more rapid improvement in the quality of education in Chile. A lesson from Chile's experience is that in some cases it probably pays to confront the vested interests of teachers and health workers up front. The number of days lost in strikes, even the political costs to the government, will be rapidly compensated by immense welfare gains, as more flexible, decentralized, incentive and user oriented social services are provided. From Extended Coverage to Focus on Vulnerabilities The previous story is, however, incomplete. As the Brunner and Elacqua study demonstrates, 60 percent of the variance in school performance is explained by the socio-cultural and economic milieu in which a student's family life develops . Only 40 percent is caused by the quality of the 13 school. Thus, it is a bit unfair to put the blame exclusively on bad teacher performance, for bad learning scores at the poorer schools in the country. A large share is caused by a very adverse family environment which, in turn is a consequence of large social inequalities as they exist in Chile. An adverse socio-cultural milieu and huge income disparities reduce the effectiveness of traditional social policies as hard core extreme poverty is tackled. The problem is compounded by the adverse impact of unexpected negative income shocks on poor families. These detonate loss 12J. J. Brunner and G. Elacqua (2003) 13J. Brunner and G. Elacqua (2003) 11 CASE STUDIES IN SCALING UP POVERTY REDUCTION of jobs, serious health problems, the break up of families or the impossibility of meeting school financial needs for the children of very poor families14. These shocks install a permanent economic insecurity that reduces family cohesion, and debilitates motivation for better performance at the school or job seeking or retraining. When this point is reached, traditional sector-specific social policies lose effectiveness in a rather dramatic way. The World Bank has identified this problem as poor social risk management vis a vis poor families. A new, holistic approach for social policies is suggested15. Governments in Chile in the last ten years have explored different venues to solve this problem. Early efforts consisted in developing experimental programs focused on groups of very poor people willing to organize themselves around specific projects that would allow them to improve their condition. The focus was on credit access for microenterprises, upgrading skills, small irrigation projects in rural areas, youth training programs and housing and environmental improvement in squatters and slums. Several institutions and programs were established to deal with these problems16. The programs represented a step forward and away from the traditional centralized, sectoral actions on health, education and social infrastructure. The new experimental programs explicitly included, as a requisite, the active involvement of individuals or families belonging to the extreme poor category. Those reached by the programs got enthusiastically involved. But soon the weak aspects were also evident. The project approach means that when the project is done, the initial favourable impact evaporates. And either new projects are designed for the same "clientele" or you hope that in the short time span of a year, people have become more self reliant. In other words, there is no cumulative effect. A second shortcoming is, obviously, the reach of the programs. By its very nature, an experimental program has a scale-up problem. It is seldom the case that an experimental program evolves into a more systematic effort that would change the way public institutions approach poverty eradication. No integral approach, emerges. It is usually the case that each agency continues doing its own thing. The "experiment" does not alter their generally burocratic organizational culture. An additional and more serious problem is that the percentage of the very poor reached by these programs is small. In fact, as shown in Table 7, decreasing the number of indigent population has proved to be much more intractable than taking people up the poverty line. The concern with this result is that what public opinion sees is a very unequal society unable to deal seriously with the hard core of extreme poverty. It has a demoralizing effect and builds a feeling of frustration and futility. This mood easily translates into a discredit of democratic politics and of "neoliberal" economic policies. It is intellectualized in the form of radical postures against 14D. Raczynski, C. Serrano and M. Valle (2002). 15R. Holzmann and S. Jorgensen (2000) 16Fosis, Chile Barrio and Pro-Rural are the examples. See D. Raczynski and C. Serrano (2002) 12 SUCCESSES AND FAILURES IN POVERTY ERADICATION: CHILE globalization, or in a new version, of old-style populism, to which Latin America has so often felt fatally attracted to in the past. These problems were recognized early on in Chile. Already in 1994, President Frei implemented a first attempt at scaling-up experimental programs by proposing an integral, decentralized approach that would deal more effectively with extreme forms of poverty. A "National Plan of Poverty Eradication "was designed . Eighty municipal districts were selected as policy targets. They were supposed to concentrate on the largest proportion of extremely poor families. The effort was a step in the right direction and it did produce some progress in some of the poor areas of the country. But the new program did not solve three critical problems: first, although municipalities were supposed to coordinate and intensify efforts towards poverty eradication, no additional resources were provided. Second, how could it be explained to a particular local authority that the next door municipality was included as one of the 80 districts targeted in the Plan, and his or her's was not?. Thirdly, programs at the local level continued to focus on the individual and no integral family-focused approach to the poor families was developed, and no real participation of potential beneficiaries occurred. By the beginning of the following decade, in 2002, it was already obvious that a new approach was needed. President Lagos proposed "Chile Solidario", a family-centered, integral program designed to reduce the vulnerability associated to the condition of extreme poverty17 The approach draws heavily from a World Bank proposal. It focuses on the family as a unit. It involves the head of the family in a mutual commitment with the government agency in charge. They are supposed to jointly identify the sources of vulnerability and risks of the family and then design an integral, coordinated action destined to cover the main deficits, in terms of minimum income needed for subsistence, access to basic social services, to schooling for the family's children, and training and empowerment for the adults in the family unit. Progress is monitored by a public official in charge of accompanying the family throughout the process. After basic needs are covered, the next step is to facilitate access to stable jobs, networking with neighbours and community organizations. After a 2 year period, people are expected to be empowered with a new acquired capacity to self sustain the family and thus break away from the cultural and economic trap that surrounds very poor families. The Chile Solidario program scales up from the experimental programs in the 1990s ­ FOSIS, Chile Barrio, Pro Rural- and focuses on the total universe of the extremely poor, some 220 thousand families. It has already been implemented in 80 percent of municipal districts. It engages national ministries, sectoral agencies and municipalities. A preliminary evaluation of the program stresses both strong positive aspects and shortcomings18. A field study shows that a more effective coordination of national and local agencies dealing with social policies has been achieved. The program has produced a strong 17For an evaluation of "Chile Solidario" see C. Serrano and D. Raczynski (2003). 18C. Serrano and D. Raczynski, (2003). 13 CASE STUDIES IN SCALING UP POVERTY REDUCTION motivational impact on the women involved. They do get tangible benefits: knowledge about access to existing subsidies, supplementary income, access to school tuition for children and to emergency but transitory employment programs. Some of the shortcomings are also obvious. The program tends to create a paternalistic, clientelistic relationship with those benefited. Income support and subsidies are obtained, but the empowerment aspect does not seem to be working. Families are visited by the monitor, but are not motivated to network with community organizations. Two years as the intended time-frame for each family to participate in the program, seems like a short period of time to break from the vicious circle of poverty. Chile Solidario does represent the most serious effort thus far to scale up a systemic, participatory response to the problem of extreme poverty. It advances well beyond the experimental stage. It is innovative and takes into account the right variables. However, it certainly needs close monitoring, just-in-time corrections and improvements, in order to effectively solve the key problem in poverty eradication: the transition from dependency and income-assistance to self-reliance on the part of the most deprived members of society19. Back to Macroeconomics Chile Solidario was launched at a time when the growth rate of the Chilean economy had slowed down substantially, as a consequence of recessionary tendencies in the world economy. This explains in part the difficult transition for Chile Solidario families from income-support status to stable income generating jobs. However, what is surprising in the Chilean case, is that the slowdown from a 6.6 percent growth rate in 1997 to negative growth in 1999 and a meagre 2.2 growth in 2000, did not detonate an increase in the number of poor. As is shown in Table 4, poverty decreased from 23.2 percent in 1996 to 21.7 percent in 1998, and again to 20.6 percent in year 2000 in spite of a recession, although indigence increased very slightly between 1998 and 2000. How can this be explained? It is useful to look at Table 10, where rates of growth in the economy are compared with the annual increase in social expenditures, during the economic downturn. It is clear from Table 10 that social expenditures did play a countercyclical role: they expanded while the economy was contracting as in 1999, and they doubled the rate of growth of the economy, when the latter went down sharply as in 1988 and 2000. 19Another highly innovative program, with a similar focus, is Brazil's Comunidad Solidaria, implemented by the Cardoso administration since 1995. 14 SUCCESSES AND FAILURES IN POVERTY ERADICATION: CHILE Table 10. Countercyclical Impact of Social Expenditures, 1997­2000 1997 1998 1999 2000 GDP, annual growth rate 6.6 3.2 -1.1 4.2 Social expenditures, annual growth rate 5.6 7.6 7.8 6.1 Social expenditures, percent of total 67 68 69 71 Government expenditures Source: Banco Central (2001) op. cit. Ministerio de Hacienda (2003) op. cit. The lesson to be drawn from this, is that a sound macroeconomic policy pays when hard times come for the disadvantaged. Chile could sustain countercyclical social policies because the government accumulated public savings in the good years. It could spend more in the social sectors in the bad years, without retreating from economic stabilization, budget balance and low country risk. This runs counter to a repeated perverse cycle in Latin American economies: during bad years, the first government programs to be curtailed are social programs and public works. These cuts have a negative impact on the most vulnerable groups. Why, then, does the number of indigents increase?. Between 1998 and 2000 the employment of wage earners fell by 2.7 percent. But the number of self-employed increased by 6 percent. This means that some people in the lowest quintile lost a permanent job consequently suffering an income loss. This was only partly compensated by occasional odd jobs. This outcome in predictable. Temporary, informal sector jobs are a standard component of survival strategies for the poorest families. The conclusion is that vulnerable groups cannot be totally isolated from the negative impact of a recession. There is still a lot to learn about how to design a social safety net that reduces the impact of adverse income shocks at the lowest income levels. This is a key issue, given the recurrent nature of adverse external shocks on emerging country economies. The Political Economy of Poverty Eradication It is quite clear from the previous pages that early success of poverty eradication policies in Chile was the consequence of a high degree of political consensus around 2 basic concepts: one, a tax increase was a necessary condition for reducing poverty. The reason for this was a generalized perception of inequity in the previous free-market model that translated into widespread political support for what would otherwise have been a very unpopular way of starting a government: increasing taxes across the board. Why did the private sector end up supporting such increase in taxation? The business community put a high value on stability and permanence of the open market model. They were also given explicit guarantees by the government that tax changes would be a once and for all proposition, at the very beginning of the new Administration, and that stable rules would follow. 15 CASE STUDIES IN SCALING UP POVERTY REDUCTION The credibility of this proposition was enhanced by the decision to proceed, in 1991, with a unilateral reduction of external tariffs. This was read by the private sector as a clear proof of the commitment by the government to an open market economy and free trade. What followed was a spectacular increase in private investment, explaining growth rates around 8 percent a year for the period up to 1997. This was only interrupted by the Asian crisis. However, the slowdown in poverty reduction in the second part of the nineties cannot be solely explained by a slowdown in growth. As more and more resources were devoted to increase the coverage and quality in health and education, it became clear that efficiency issues took center stage. The resistance of teachers unions and health care workers unions to decentralization in wage bargaining, to performance-based incentives and to flexibility in hiring and firing practices, inaugurated a period of decreasing returns to additional social expenditures. In order to understand why the government chose not to confront public sector unions, two factors have to be considered. The first one is that the three governments that followed the Pinochet authoritarian era represented a Center ­ Left coalition, supported by the labor union movement, very much like socialdemocratic governments in Europe. It was not politically obvious that confronting those who had fought against Pinochet, side by side with the Center ­ Left politicians now in government, was a wise course to pursue at the very beginning of a new democratic government. The second factor is of a historical nature. Chile developed, since the 1950s, an imperfect welfare state modeled after the European social institutions of the post-war period. As demonstrated in the literature, such a model is built around the notion of providing social protection to the "insiders", that is to those members of the workforce who have a stable permanent job and are represented by strong centralized union organizations20. These include public sector employees and members of the large, modern enterprises in the public and private sectors. The social protection for "insiders" in Chile as in most of Europe developed specific institutions: centralized wage bargaining, lifelong employment guarantees in the public sector, and high costs of lay-offs to firms who need to make adjustments in its work force. Free access to public health programs, and guaranteed social security coverage, plus unemployment insurance, were also part of the social protection package for "insiders". The problem is that a large majority of the work force in Chile consists of "outsiders", who hold temporary, lower-quality jobs or are chronically unemployed. The fight against poverty has to be understood, then, as an attempt to bring the "outsiders" into the social protection network, that historically has benefited mainly the insiders, well represented by nation-wide labor organizations. Thus, the conflict is not only one of management practices in the public sector. It has also to do with redressing the scope and reach of institutions that were originally designed for the 20G. Esping-Andersen (2000, 2002); N. Gilbert and R. Von Voorkins (eds), (2003). 16 SUCCESSES AND FAILURES IN POVERTY ERADICATION: CHILE organized, modern segment of the labor force, towards the less organized members of the labor force. Of course this problem is not unique to Chile. It is at the core of the current debate in Europe about reshaping their welfare states. Thus, describing these conflicts and constraints in trying to design a poverty-eradication strategy, does not represent a neo-liberal anti labor bias in social policies, but a recognition that the movement towards a more equitable society requires also a new fresh approach to the role that the labor movement will or will not play in building such a more inclusive and more cohesive social system21. When labor takes only a defensive position vis a vis the pressure of "outsiders", labor unions lose membership and relevance. This has been the case in Chile during the last decade. As the more enlightened labor leaders realize the reason for this, they open up to change, as the Chilean case demonstrates. But it took more that a decade, for this to start happening as public health and education workers are concerned. The focus on "outsiders" in the labor force is not exempt from constraints and conflicts. These are particularly severe when dealing with those trapped in extreme forms of poverty. The fight against indigence, requires an integrated multisectoral approach to be implemented simultaneously by a large number of government agencies, both at the national, sectoral and local levels. The problems here are the entrenched, bureaucratic practices. The only serious response is to fight the bureaucratic culture by tackling head-on the modernization of the State apparatus. This is more easily said that done. Bureaucrats defend their turf and quotas of power. Users of public services do not pressure the government to modernize, because they do not see the benefits of such a policy in terms of better quality services, in the short term. In the absence of a deep public sector reform, governments try to persuade the higher echelons of public officials to respond to those policy goals that require a high level of cooperation, as in the case of the fight against poverty. But even cabinet members end up suffering a "contagion effect" from the bureaucratic milieu. Poverty eradication is a diffuse goal for a sectoral ministry that does not in itself enhance agency identity, for it does not allow it to deliver more of its "unique product" for which it will be evaluated. Instead more and more resources will be devoted to programs that are controlled by others and are run by some diffuse inter-Agency committee. The outcome is segmented and fragmented instead of integrated policy implementation when fighting poverty. Institutional reform in the public sector is, thus, a requisite. It has to include the profesionalization of public managers, hired through head-hunting open processes, and evaluated for performance in achieving integrated goals, such as those required for a successful strategy of poverty eradiation. Decentralization for a more accountable provision of public services destined to the poor at the local level is also a requirement. These are just two of many lessons from recent 21A. Giddens (2001) 17 CASE STUDIES IN SCALING UP POVERTY REDUCTION Chile's experience with poverty eradiation that should not be overlooked. I suspect they are valid and relevant to a large number of other national cases. 18 SUCCESSES AND FAILURES IN POVERTY ERADICATION: CHILE References ARELLANO, J. P., (1995), "Fiscal Policy and Social Development" in C. Pizarro et. al. Social and Economic Policies in Chile's Transition to Democracy, book published by UNICEF ­ CIIEPLAN. ARENAS, A., and P. BENAVIDES, (2003), Protección Social en Chile, International Labor Office, Santiago, Chile. BANCO CENTRAL, (2001), Indicadores Económicos y Sociales de Chile 1960-2000", Santiago, Chile. BAYER, H., (2001), "Propuestas Necesarios para el Sector Salud", in H. Bayer and R. Vergara (eds.), Que Hacer Ahora, Propuestas para el Desarrollo, Centro de Estudios Públicos, Santiago, Chile. BRUNNER J. J., and G. ELACQUA, (2003), El Capital Humano en Chile, Universidad Adolfo Ibáñez, Santiago, Chile COWAN, K., and J. De Gregorio (1996), "Distribución y Pobreza en Chile", Estudios Públicos Nº 64. COX C., and M.J. LEMAITRE, (1999), "Market and State Principles of Reform in Chilean Education", in G. Perry and D. Leipziger, Chile, Recent Policy Lessons and Emerging Challenges, a book by the World Bank. ESPING ­ ANDERSEN G. (2000), Social Foundations of Postindustrial Economies, Oxford University Press, Oxford. ESPING ­ ANDERSEN G., D. GALLIE, A. HEMERIJCK, J. MYLES (2002), Why we Need a New Welfare State, Oxford University Press, Oxford FOXLEY, A., (2003), "Development Lessons of the 1990s: Chile", a paper prepared for the Practitioners of Development Conferences by the World Bank. GIDDENS, A. ed (2001), The Gobal Third Way Debate, Policy Press ­ Blackwell Publlishers, Oxford. GILBERT N, y R. VAN VOORHIS eds. (2003), Changing Patters of Social Protection, Transaction Publishers, London. HELD, G. and L. F. JIMENEZ, (2001) in R.Ffrench-Davis & B. Stallings (eds.) Reformas, Crecimiento y Políticas Sociales en Chile desde 1973, ECLAC, Santiago, Chile. HOLZMAN R. and S. JORGENSEN, (2000), "Manejo Social de Riesgo: un Nuevo Marco para la Protección Social". MELLER, P., (1999), "Pobreza y Distribución del Ingreso en Chile en la Década del 90", paper presented on a "Conference on Chile 1990-1999", at the University of California, San Diego, Dec. 1998. 19 CASE STUDIES IN SCALING UP POVERTY REDUCTION MIDEPLAN, (2001), Pobreza e Indigencias e Impacto del Gasto Social en la Calidad de Vida, Santiago, Chile. MINISTERIO DE EDUCACION, Chile, (2002), Resultados de las Pruebas Simce 1996 ­ 2001 MINISTERIO DE HACIENDA, (2003), Estadísticas de la Finanzas Públicas, forthcoming. MIZALA A., and P. ROMAGUERA, (2002), "La Legislación Laboral y el Mercado del Trabajo en Chile 1975-2000", in R. Ffrench Davis and B. Stallings, (eds.) Reformas, Crecimiento y Políticas Sociales en Chile desde 1973, ECLAC, Santiago, Chile. RACZYNSKI, D., (1998), "Para Combatir la Pobreza en Chile", in R. Cortázar and J. Vial (eds.) Construyendo Opciones, Ediciones Dolmen CIEPLAN. RACZYNSKI, D. and C. SERRANO, (2002), "Nuevos y Viejos Problemas de la Lucha contra la Pobreza" in El Período del Presidente Frei, Editorial Universitaria, Santiago, Chile RACZYNSKI, D. C. SERRANO and M VALLE, (2002), "Eventos de Quiebre de Ingreso y Mecanismos de Protección Social ", a paper prepared for the World Bank. RACZYNSKI, D. C. SERRANO, (2003), "Derechos Sociales Básicos, Superación de la Pobreza y Protección Social ante la Vulnerabilidad" Asesorías para el Desarrollo, Santiago, Chile RODRIGUEZ, J. and M. TOKMAN, (2000), "Resultados y Rendimiento del Gasto en el Sector Público de Salud en Chile 1990-1999", CEPAL, Serie Financiamiento del Desarrollo WORLD BANK, (2002), Chile's High Growth Economy Poverty and Income Distribution 1987- 1998. 20