versión On-line ISSN 0717-6821
Cuad. econ. v.38 n.114 Santiago ago. 2001
* Sector Manager, Poverty, Latin America and Caribbean Region, The World Bank, Washington D.C.
** Formerly with the World Bank. Currently consultant in Santiago. Vía Verde 9771, Vitacura, Santiago, Chile.
Keywords: Poverty, Income Distribution, Latin America.
Classification JEL: H5 y 13
The Bible says succinctly that "the poor you will always have with you". In Latin America's case that certainly seems to be true. Sluggish growth, an income distribution that favors the rich, and repeated economic and political crises seem to dominate the hemisphere and prevent a real progress in reducing poverty. Can anything be done? What have we learned in the past decade that might change the way we attack the problem? The papers in this volume are an attempt at a partial answer, drawing on experience in the Latin America and Caribbean region (LAC) in general, and for some representative countries: Mexico, Brazil, Chile. They represent both standard methods of analysis as well as some of the new thinking on poverty concepts and measurement.
Poverty: How Much? And Why?
"In general it is easier to identify people who are poor than it is to bring relevant parts of economics to bear on their circumstances" T.W.Schultz
Measuring poverty is, at best, an inexact science. Money income is only one dimension of poverty; other important dimensions can include a range of things including having access to social services, having a job, health and nutrition status, life expectancy, ability to participate in the political process, being able to live in a community free from violence. According to Sen (1983), the definition may not border so much on having things and access, but on having the "capability" to improve you life, and the freedom to make that choice. The recent World Development Report 2000/2001 of the World Bank places new emphasis on "empowerment" as determining poverty status _ the ability of poor people to participate in the planning, organization and operation of institutions that affect their lives.
Despite the new thinking on poverty measurement, the old techniques of poverty lines remains dominant as a simple metric that can be used over time, and with care, across countries. Generally, poverty lines done by countries vary considerably, however, in their definition of poverty. The normal basis for these lines is some estimate of minimum calorie consumption needed to sustain a person, and some crude estimate of non-food expenditures. The choice of the calorie level and non-food ratio (inverse Engel coefficient) can vary between countries, as well as the techniques for adjusting households for their composition (adult equivalents) or scale economies.
Poverty in Latin America: Trends and Determinants
In the study by Wodon et al. (in this volume), the authors overcome some of the problems of making inter-country comparisons of poverty by applying a uniform definition of poverty to different country household surveys. This study develops consistent estimates of the evolution of poverty during the last 15 years for 18 Latin American countries, and presents an analysis on the influence of various determinants of poverty, including household variables (education, age, gender), employment, ethnicity, location, and migration for a sample of nine countries. Based on aggregates figures for the total population of the region weighted by population size, they find that the share of the population in poverty has decreased in the 1990s and it is now back to its level of the mid-1980s. Nevertheless, due to population growth the number of poor has increased in the last 15 years. The authors estimate that about one-third of the population of the region lives in poverty, and one-sixth lives in extreme poverty, meaning that the latter lack sufficient income even to purchase the basic food basket. In absolute numbers, this means about 180 million poor people in LAC (in 1998), which because of population growth, means there are now about 30 to 40 million more poor people than there were in the region in 1986. Furthermore, they find that the rural poor are more apt to be extremely poor, and to lie further below the poverty line than their urban cousins.
It is notable, and especially relevant for comparisons of policy regimes across governments and through time, that Wodon et al. find more countries with poverty rate reductions than countries with deteriorations. Weighting all countries equally, there has been a more consistent reduction in poverty during the period of study. Although their poverty levels differ from detailed studies for individual countries, the trends are consistent with those in other studies. On a country-by- country basis, the analysis shows a strong link between growth and poverty reduction, with an estimated elasticity of -1.0. The study illustrates well the extent to which large macroeconomic shocks has undermined progress towards poverty reduction, such as in Mexico during 1995, and Argentina and Brazil during the late 1980s.
Poor families have a certain number of characteristics that define their poverty. In terms of poverty determinants, Wodon et al. identify poor households are those which:
¨ have a larger number of babies and children;
¨ have younger household heads, and/or are female headed;
¨ have heads with limited educationnormally less than secondary education
¨ have a head or spouse that is unemployed, or underemployed.
Education is clearly a key factor that consistently emerges as a correlate of families who have escaped poverty. According to the estimates of Wodon, et al., a secondary education will increase earnings of a household head by 50% over a similar person with no education. However, an educated household head is not normally sufficient for a family to escape poverty, if only one household member is working. What is required is that their be employment and education for the household head and his/her spouse as well. Thus, programs that improve the education, training and employment opportunities for women remain essential, not only to increase earnings, but also as a means to reduce fertility and family size, another direct correlate of poverty.
Public sector social programs can have a major impact on both poverty and income distribution. In the past, few government officials worried excessively about who benefited from government expenditures, and those that did worry about it had little or no data at hand to study the problem. A generation of better, more detailed household surveys now reveal a similar pattern across most Latin American countries: social spending generally does benefit the poor, and helps improve income distribution overall. However, some parts of social spending are notoriously pro-rich, particularly programs of higher education which generally benefit non-poor families, while recovering little or nothing in terms of student fees. Primary health and primary education are generally very "pro-poor", while housing expenditures, electricity subsidies, and tranport susbsidies generally are not. In recent years, more and more governments have focused on developing systems to target social programs to poor families, often using a variety of economic and social indicators as a form of proxy for income.
Family Factors in the Intergenerational Transmission of Poverty
Much like an inherited trait, poverty tends to pass from parent to child. How prevalent is this "curse of the poor," why do some escape it, and how can we help improve the odds? The study in this volume by Enrique Aldaz-Carroll and Ricardo Morán sets out to gauge the extent of this "Intergenerational Transmission of Poverty" (ITP) in 16 Latin American countries, analyze certain factors affecting it, and raise policy considerations. Among the various a priori determinants of ITP, the study focuses on "family factors" - those more closely related to characteristics of the household, such as mother's schooling, than to its economic and social environment. The empirical results indicate that the prevalence of ITP in Latin America is strong and that family factors play an important role in the educational achievement of poor children - and hence on their expected lifetime income. Regressions for 16 Latin American countries show that children in poverty with fewer siblings, more educated parents, higher household income, and living in urban areas are significantly more likely to complete secondary education. Completion of secondary education is taken as the threshold level of schooling at or above which a child of poverty should have a fair chance of escaping the poverty cycle in the 21st century. Results for the subset of countries with the required data further show that children of the poor born to single adolescent mothers, or did not attend a preschool program, or were undernourished, are less likely to complete secondary education than children of the poor without the corresponding attributes. Moreover, supplementary data reviewed for this study, but that could not be adequately indexed for the regressions, tend to support findings by other researchers pointing to two additional family factors affecting educational performance among children from poor households: domestic violence and ethnicity. Study results suggest that poverty-reduction strategies take into account family factors much more than is commonly the case, especially in complement to the supply of education and other basic social services that are so greatly emphasized today. As corollary, we recommend that such social services should, whenever possible, focus on the households in poverty, rather than on their members individually, to improve their children's education outcomes and thereby increase their chances of breaking out of the cycle of intergenerational poverty.
The Problem of Income Distribution
Why has there been so little progress on reducing poverty? Economic growth during the 1990s averaged about 3.2% per year or about 1.5% per capita (1991-99,World Bank, 2001). While not outstanding, it was about average for the developing world over the same period; lower than East and South Asia, but higher than Africa and Central Asia. Not surprisingly, most research on developing countries has found a clear, but not perfect, inverse relationship between poverty rates and economic growth (Ganuza and Taylor, 1998; Morley, 1998). Poverty rates are also negatively affected by changing income distribution, and during the past decade, there has been a slight tendency for income distribution to worsen, although this also varies among countries. For the 1980s, the Gini coefficient of income distribution seems to have worsened (Londoño and Szekely, 1997). For the 1990s, Szekely and Hilgert (2000) find that for a sample of 15 countries, 8 had statistically signficant worsening, and only one had an improvement (as measured by the Gini coefficient). In some countries, the worsening seems to have started generally after 1995. A worsening income distribution with rising income implies that the gains from growth have gone disproportionately to the rich. However, for the region as a whole, the tendency for worsening seems to be consistent during the decade of both the 1980s and 1990s, in a region which started the period already having one of the worst income distributions in the world (Deininger and Squire, 1996 ). According to estimates of the Interamerican Development Bank (1998), the top 10% of the region's population receive 40% of total national income, while the bottom 30% receive only 7.5%.
Poverty and Income Distribution Dynamics. The Case of Chile
The dynamics of income distribution and poverty are clearly seen in the case of Chile (Contreras, Larrañaga, Litchfield and Valdés, this volume). Between 1987 and 1998 the rate of poverty fell from 40% to 17%, as a result of both rapid growth and an extensive system of social programs directed at the poor. However, there has been no improvement in income distribution; the top 10% received 44% of total income in 1998, virtually unchanged from the level of 1987. While income distribution seems to have improved between 1987 and 1994, it worsened between 1994 and 1998, so the net effect was no change in distribution (the Gini coefficient in 1998 is virtually the same as in 1987). Thus, it is difficult to know if one should worry about the recent upswing in the Gini, or treat it as part of a cyclical movement that will even out in the long term. There is evidence that in many countries, technological change and shifting comparative advantage (loosely known as "globalization") has meant rising demand for skilled workers, with the commensurate impact on widening the gap between the skilled and the unskilled. But is this a permanent adverse impact of globalization, or will it be corrected by an eventual supply response of more skilled and educated workers being produced by the education system? Only time will tell.
This study on Chile develops and applies a methodology for the estimation of the implicit income transfer from government subsidies in education, health, and housing. Most measures on poverty and inequality in Latin America, including Chile, do not adjust for this in-kind transfers, and thus tend to overestate poverty and inequality. For 1998, the adjustment for such transfers lowers the Gini coefficient from 0.56 (unadjusted) to 0.50 (adjusted), which suggests that social policies in Chile have had a significant impact in reducing income inequality.
Brazil: Explaining the Changes in Income Inequality
On average, Brazil has one of the highest levels of inequality in Latin America, however during the period 1981 to 1995 for which there is reasonably good household level data for the country as a whole, significant fluctuations in inequality are observed. The paper by Ferreira and Litchfield (in this volume) investigates the increase in inequality observed in Brazil during the 1980s as well as the decline in the first half of the 1990s. In their analysis the micro-determinants of inequality (education, demographics, regional location, spatial variables) perform well in explaining inequality levels. The large differences between households with different education suggests that educational attainment is 3 to 4 times more influential than any other structural or demographic factor considered. Age and gender have negligible explanatory power. Nevertheless, race, region, and the urban-rural location of the households are also important determinants of overall inequality. The increase in inequality during the 1980s appears to have been driven by an increase in educational attainment and by a high and accelerating inflation. In the 1990s, the fall is associated by an increasing equality between urban and rural households, falling inflation, and declining returns to education.
The Role of the Public Sector in Fighting Poverty
Public sector social programs can have a major impact on both poverty and income distribution, as discussed in the study on Chile in this volume. In recent years, more and more governments have focused on developing systems to target social programs to poor families, often using a variety of economic and social indicators as a form of proxy for income.
The rise of "globalization" has also generated more insecurity and uncertainty about the future, as economic shocks are transmitted more rapidly than ever, and employment and income prospects seem more uncertain, particularly for the poor. While it is commonly believed that globalization has increased instability, and that this instability impacts the poor more adversely, this does not seem to be supported by the evidence. A recent World Bank study (World Bank, 2000) shows that instability has not increased, when comparing the 1990s to the 1980s, and while LAC is more unstable than OECD countries, its level of instability is equal or less than other developing regions. Furthermore, there is no evidence that poorer groups suffer from greater instability of incomes. However, even in countries with robust growth rates (Argentina, Colombia, Mexico), there is evidence of higher unemployment duration, increased turnover, and growing levels of informality and insecure jobs (see also Maloney, 1999).1 Finally, from a welfare standpoint, it seems clear that an equal percentage decline in income for all households will more severely affect those at or below the poverty line, than those well above it.
Mexico: Do Agricultural Support Programs Alleviate Poverty?
One constant problem with all public sector social programs is one of evaluation. Do these programs really work, and do they effectively and efficiently reduce poverty? How do we really know what would have happened without the project or program? Evaluation techniques for measuring poverty impact are still being developed (see Baker, 2000), and many of these problems cannot be resolved completely. The study by Cord and Wodon in this volume deals with the evaluation of two important programs in Mexico's rural areas. As shown in López and Valdés (2000), in Wodon et al., and Ferreira and Litchfield in this volume, in most countries the incidence of poverty -an particularly extreme poverty- is higher in rural areas than in urban areas. This study is very relevant today both because Mexico has a unique experience with several years of massive government programs for the rural areas, and also due to its methodological rigor. Using a panel data approach based on household surveys in 1994 and 1997 and considering the potential endogeneity of program participation, the study by Louise Cord and Quentin Wodon examines the impact of two large government programs, PROCAMPO and ALIANZA. The former is a "de-coupled" income support program for producers of basic crops ( a fixed payment per hectare delinked from current production), covering 3 million farmers in approximately 90% of Mexico's cultivated area. The PROCAMPO program is a substitute for indirect subsidies formerly tied to inputs, and price supports for products, which were abolished as part of the NAFTA agreement. On the other hand, the program ALIANZA para el Campo is aimed at fostering agricultural productivity through a matching grants scheme, and the provision of support services (research, extension, information, training).
The authors find that ALIANZA appears to have had no impact on reducing poverty. PROCAMPO, on the other hand, has had a significant impact in raising household income of farmers which are very poor. In fact the income transfer from the program provided about 40% of household income of the poorest decile, after considering both its direct income effect and its indirect multiplier effect in reducing liquidity constraints for crop intensification and reducing the risk aversion. As part of the process of trade liberalization programs in agriculture several countries are considering programs that would provide temporary compensation to low income farmers producing import-competing products. PROCAMPO is perhaps the only program of such kind with several years in operation in the whole region. The analysis on why ALIANZA failed in raising income has important lessons for the design of rural development programs in the region.
The study by Cord and Wodon makes an important methodological point. Using simple multiple regression, participation in PROCAMPO reduces the probability of being poor by 10%. However, a more thorough two stage estimation raises the impact to 42%. The difference between these two estimates reflects the impact of the endogeneity of program participation; since the program is not uniformly available in all regions of the country, the results from simple multiple regression are biased. This example illustrates a more general problem: it is very easy to come to misleading conclusions in evaluations if the techniques used do not adequately control of exogenous or endogenous factors.
For a region of considerable wealth, Latin America remains a region of much poverty. The articles in this volume try to address some of the salient issues related to that poverty in a quest for finding causal factors. While exact answers are always elusive, some key factors in explaining this poverty are clear:
¨ growth has not been robust, and so the numbers in poverty have grown even as the poverty rate for the region has remained, more or less, the same;
¨ income distribution remains highly skewed toward the rich _ wage gaps between skilled and unskilled seem to have widened in recent years in many countries, and overall income distribution seems to have become more unequal in the last 20 years;
¨ poor families have workers with low levels of education, and the education received by poor families in poor neighborhoods is of lower quality;
¨ there is a strong negative correlation between per capita income and family size; poor families are large, and hence policies that could induce an adjustment in family size should not be neglected;
¨ poverty is self-perpetuating _ children of poor families tend to repeat the cycle of low income and large family size;
¨ poor families live in areas that do not receive social and infrastructure services.
The debate on poverty, welfare and income distribution in Latin America is a lively and complex one. It is lively because it is sensitive, touching on issues of equity with implications for efficiency and overall growth. What makes the issue complex is that while poverty and equity are usually thought of in terms of income and wealth, the debate is about welfare and human development. These are much broader concepts than what we usually capture by money-metric measures of income. Can Latin America do a better job than is doing now with respect to income distribution and poverty alleviation, and not seriously compromise the growth potential of the economy?
There is a clear role here for public policy - programs that invest in the human capital of the poor and help them raise their productivity. But public programs need to be undertaken with caution; too often programs are not thoroughly evaluated before being expanded or undertaken. In many cases, the benefits and beneficiaries are not those initially identified by government planners.
Overall, considerable progress has been made in the past two decades in our knowledge of poverty and the factors that influence poverty. This progress has been achieved thanks to the expanding availability of household surveys in the region (Kanbur and Squire) and by the economic analyses made possible by such a rich data base. Nevertheless, when it comes to translating those analyses into practical policies focusing on income distribution, it is likely, as Harberger (2001) suggests, that what any government can do, in the Latin America of today, is in reality restricted. The policy tools available are limited. The analyses in this Special Issue show that two key policy tools are available: improved education and social expenditure policies. They have a significant redistributive potential, but their impact on income distribution might take decades rather than years.
1 The question of whether employment and earnings are becoming more "precarious" was examined for Chile for the period 1962-95 in "Chile: Poverty and Income Distribution in a High Growth Economy: 1987-95", World Bank, November 1997.
Deininger, K. and L. Squire (1996), "A New Data Set Measuring Income Inequality", The World Bank Economic Review 10(3):565-592 [ Links ]
Ganuza, E. and L. Taylor (1998), "Política Macroeconómica, Pobreza y Equidad en América Latina y el Caribe" in Política Macroeconómica y Pobreza en América Latina y el Caribe, E. Ganuza, L. Taylor, S. Morley, eds. UNDP/Ediciones Mundi-Prensa: New York. [ Links ]
Harberger, A.C. (2001), "The View from the Trenches", in Frontiers of Development Economics (eds.) G. Meier and J. Stiglitz, World Bank and Oxford University Press. [ Links ]
Hicks N. and Q. Wodon (2001), "Economic Shocks, Safety Nets, and Fiscal Constraints: Social Protection for the Poor in Latin America", CEPAL Review 73 (April). [ Links ]
Inter-American Development Bank (1998), Facing Up to Inequality in Latin America, Baltimore, MD: Johns Hopkins University Press. [ Links ]
Kanbur, R. and L. Squire (2001), "The Evolution of Thinking about Poverty", in Meier and Stiglitz, op.cit. [ Links ]
Londoño, J.L. and M. Székely (1997), Persistent Poverty and Excess Inequality: Latin America 1970-1995. Office of the Chief Economist Working Paper Series Nº 357, Inter-American Development Bank, Washington DC. [ Links ]
López R. and A.Valdés (2000), "Rural Poverty in Latin America", Macmillan Press (UK) and St. Martin's Press, USA. [ Links ]
Maloney, W. (1999), Does Informality Imply Segmentation in Urban labor Markets? Evidence from Sectoral Transitions in Mexico, World Bank Economic Review, 13:275-302. [ Links ]
Morley, S. A. (1998), "La Pobreza en Tiempos de Recuperación Económica y Reforma en América Latina: 1985-1995" in Política Macroeconómica y Pobreza en América Latina y el Caribe, E. Ganuza, L. Taylor, S. Morley, eds. UNDP/Ediciones Mundi-Prensa: New York [ Links ]
Schultz, T.W. (1993), "The Economics of Being Poor", Blackwell, Cambridge, USA [ Links ]
Sen, A.K. (1983), "Poor, Relatively Speaking", Oxford Economic Papers 35:153-169. [ Links ]
World Bank (2000), World Development Report 2000/2001: Attacking Poverty. Oxford University Press/World Bank:Washington, D.C. [ Links ]