Read an Excerpt
Declining Inequality in Latin America
A Decade of Progress?
By Luis Felipe Lopez-Calva Nora Claudia Lustig
Brookings Institution Press
Copyright © 2010 Luis Felipe Lopez-Calva and Nora Claudia Lustig
All right reserved.
Chapter One Explaining the Decline in Inequality in Latin America: Technological Change, Educational Upgrading, and Democracy LUIS F. LÓPEZ-CALVA AND NORA LUSTIG
Latin America often is singled out because of its high and persistent income inequality. With a Gini coefficient of 0.53 in the mid-2000s, Latin America was 18 percent more unequal than Sub-Saharan Africa, 36 percent more unequal than East Asia and the Pacific, and 65 percent more unequal than the high-income countries (figure 1-1). However, after rising in the 1990s, inequality in Latin America declined between 2000 and 2007. Of the seventeen countries for which comparable data are available, twelve experienced a decline in their Gini coefficient (figure 1-2). The average decline for the twelve countries was 1.1 percent a year.
The decline in inequality was quite widespread. Inequality declined in high-inequality countries (Brazil) and low-inequality-by Latin American standards-countries (Argentina); fast-growing countries (Chile and Peru) and slow-growing countries (Brazil and Mexico); macroeconomically stable countries (Chile and Peru) and countries recovering from economic crisis (Argentina and Venezuela); countries with a large share of indigenous groups (Bolivia, Ecuador, and Peru) and countries with a low share (Argentina); in countries governed by leftist regimes (Brazil and Chile) and in countries governed by non-leftist regimes (Mexico and Peru); in countries with a universalistic social policy (Argentina and Chile) and in countries with a historically exclusionary state (Bolivia and El Salvador). Inequality in Latin America is the result of state capture by elites, capital market imperfections, inequality of opportunity (in particular, of access to good-quality education), labor market segmentation, and discrimination against women and nonwhites. Hence, the observed fall in inequality is good news.
This book is among the first attempts to address the question of why inequality has declined in Latin America during the last decade, through in-depth analyses of Argentina, Brazil, Mexico, and Peru. In all four cases, the data come from country-based household surveys and the analyses focus primarily on changes in labor income inequality and changes in the size and distribution of government transfers (and remittances when relevant).
The four countries analyzed here can be considered a representative sample of middle-income countries in Latin America. The sample includes one of the five most unequal countries in Latin America (Brazil) (figure 1-3); a traditionally low-inequality country, which witnessed the largest increase in inequality of the region in the past three decades (Argentina); three of the largest countries in the region in terms of population and GDP (Argentina, Brazil, and Mexico); two countries where innovative, large-scale conditional cash transfers have been implemented (Brazil and Mexico); and one country with a large indigenous population (Peru).
All four countries experienced substantial market-oriented reforms in the 1990s (in the case of Mexico, since the 1980s). In particular, trade and foreign investment were liberalized, many state-owned enterprises were privatized, and, more generally, markets were deregulated. The four countries also faced significant macroeconomic crises between 1995 and 2006 and, except for Argentina, have pursued broadly prudent fiscal and monetary policies in particular since 2000. In 2003, following the boom in commodity prices, Argentina and Peru began to benefit from very favorable terms of trade; as a result, both countries enjoyed high per capita growth rates between 2003 and 2006 (7.8 and 5.2 percent a year, respectively). In contrast, GDP per capita growth was modest in Brazil and Mexico (2.7 and 2.8 percent a year, respectively).
Two leading factors seem to account for the decline in inequality in Argentina, Brazil, Mexico, and Peru during the last decade: a decrease in the earnings gap between skilled and low-skilled workers and an increase in government transfers to the poor. The decrease in the earnings gap, in turn, seems to be mainly the result of the expansion of basic education during the last couple of decades; it might also be a consequence of the petering out of the one-time unequalizing effect of skill-biased technical change in the 1990s associated with the opening up of trade and investment. In any case, in the race between skill-biased technical change and educational upgrading, in the past ten years the latter has taken the lead. The equalizing contribution of government transfers seems to be associated with the implementation or expansion of large-scale conditional cash transfer programs in Argentina (Jefes y Jefas de Hogar), Brazil (Bolsa Escola/Bolsa Familia and BPC), and Mexico (Progresa/Oportunidades) and with in-kind transfers in Peru.
In this chapter, we discuss the evolution of inequality and its determinants and present a synthesis of the main findings of the chapters included in this book.
Rising Inequality: The 1980s and Early 1990s
Income inequality increased in most Latin American countries during the so-called "lost decade" of the 1980s and structural reforms of the early 1990s. Although data availability constrains comprehensive comparison, the evidence suggests that the effects of the debt crisis during the 1980s were unequalizing. In particular, because the poor were less able to protect themselves from high and runaway inflation and orthodox adjustment programs frequently resulted in overkill, those in the poor and the middle-income ranges were hurt disproportionately while the income share of the top 10 percent rose. The unequalizing effect of the crisis was compounded because safety nets for the poor and vulnerable were conspicuously absent (or poorly designed and inadequate) in the Washington-led structural adjustment programs in the 1980s.
The pattern of inequality for the four countries analyzed here is shown in figure 1-4. Both Argentina and Mexico show a clear inverted U. That is not the case for Brazil, and comparable data for Peru do not go far enough (Ginis for 1984 and 1991 are not strictly comparable). However, as Jaramillo and Saavedra argue in chapter 8 of this volume, there are indications that income inequality increased during the period of reforms in the early 1990s (1991-93 in figure 1-4).
In the early 1990s, as governments turned to market-oriented reforms to pull their economies out of crisis, inequality continued to increase, driven in part by a significant increase in the relative returns to tertiary education (figure 1-5). What was behind the sharp increase in returns to education? Figure 1-6 shows that the supply of skilled and semi-skilled workers rose in the 1980s and 1990s; therefore the increase in returns to education must have been driven by skill-biased changes in the composition of demand for labor. There is evidence that both the sectoral reallocation of production and employment and the skill intensity within sectors changed in favor of skilled workers, in particular college graduates. Results, therefore, are consistent with the presence of skill-biased technological change, in particular after the opening up of the economies in the 1980s and 1990s. While for Argentina (Gasparini and Cruces, chapter 5 in this volume), Mexico (Cragg and Eppelbaum 1996 and Esquivel and Rodríguez-López 2003), and Peru (Jaramillo and Saavedra, chapter 8 in this volume) there is evidence that the direct effect of trade liberalization on wage inequality seems to have been small, the indirect effect of trade and of capital account liberalization through their impact on adoption of new skill-intensive technologies of production and organization might have been substantial. That is not the case for Brazil, where trade liberalization seems to have caused a reduction in skill premiums and wage inequality, as suggested by Ferreira, Leite, and Wai-Poi (2007). That may be the main reason why inequality in Brazil did not increase during the reform period and did not show the inverted U found in other countries.
Declining Inequality: The Mid-1990s Onward
The rising trend in inequality came to a halt in the second half of the 1990s or in the early 2000s, depending on the country (figure 1-4). From then until the global financial crisis in 2008-09 (for which data on income distribution are not yet available), inequality declined in most countries in Latin America. In particular, inequality declined in the four countries analyzed here, beginning in 1994 in Mexico, 1997 in Brazil, 1999 in Peru, and 2002 in Argentina (figure 1-4). Income inequality as measured by the Gini coefficient fell by 5.9 percentage points in Mexico (1994-2006), 5.4 percentage points in urban Argentina (2002-06), 5 percentage points in Peru (1999-2006), and 4.8 percentage points in Brazil (1997-2007).
Why did inequality decline in these four countries during the last decade? Have the changes in inequality been driven by market forces, such as the demand for and supply of labor with different skills? Have labor market institutions such as the strength of unions or minimum wages changed? Or have governments redistributed more income than they used to? We attempt to answer those questions here.
Specifically, the studies for Argentina, Brazil, Mexico, and Peru ask what the contribution was of demographic factors (changes in the proportion of adults in the household, for example) to the observed change in inequality in household per capita income. Were changes in the distribution of labor income an important equalizing factor? If so, were those changes, in turn, driven by changes in the distribution of personal characteristics (in particular, in the distribution of educational attainment), changes in returns to personal characteristics (returns to education, in particular), or changes in employment, hours worked, or occupational choice (wage labor or self-employment, for example)? If changes in all three were relevant, what caused them to change in turn? Was it increased coverage of basic education, the mix of production skills generated by technological change, macroeconomic conditions, or stronger labor unions? What has been the role of changes in the distribution of nonlabor income? Do changes in government transfers account for a significant part of the change in inequality in nonlabor and overall income inequality?
In each chapter, the authors estimate the contribution of proximate causes, relying on parametric and nonparametric methods to decompose changes in household income inequality. The empirical analysis is combined with circumstantial (that is, indirect) evidence and historical narratives to put together the multidimensional "jigsaw puzzle" of the fundamental determinants of inequality over time.
Several patterns recur throughout the four case studies. First, in all four countries changes in the distribution of the dependency ratio were equalizing. The proportion of dependents fell more in poorer households, but the contribution of that factor was far less important than the contributions of the reductions in labor income inequality and nonlabor income inequality. Also, the equalizing contribution of demographic changes was already under way in the 1990s, reflecting the reduction in fertility rates that has characterized the region in the past two or three decades. It is not a new phenomenon. The two most important differences between the 2000s and the 1990s (and 1980s too, depending on the country) are the observed declines in both labor income inequality and nonlabor income inequality.
Determinants of the Decline in Labor Income Inequality: The Race between Education and Technology
Declines in labor income inequality appear to be associated with the educational upgrading of the labor force, which resulted in a more equal distribution of schooling attainment in the four countries, above all in Brazil, Mexico, and Peru. Figure 1-7 shows the Gini coefficients for years of schooling of the population between 25 and 55 years of age. The Gini for educational attainment declined by 5 percentage points in Brazil (1998-2007), 7 percentage points in Mexico (1996-2006), and 4 percentage points in Peru (2001-07). In Argentina, the decline was almost negligible for the period in which earnings inequality began to fall (2003-06). However, that should come as no surprise since the period was much shorter and Argentina's population had more years of schooling to begin with.
Thus, the quantity effect of education on labor income inequality resulting from a more equal distribution of the stock of education (years of schooling) was an equalizing factor. However, the Gini for years of schooling had been falling for quite some time before labor earnings inequality started to decline. In fact, in Argentina and Mexico previous studies observed that the Gini for educational attainment declined while earnings inequality increased (!). As discussed in Bourguignon, Ferreira, and Lustig (2005), that apparently paradoxical result is a consequence of the fact that the returns to education curve exhibits increasing returns: that is, additional years or levels of education command a proportionately higher return. Did returns to schooling become less steep during the period in which inequality declined? The answer is yes: the returns to tertiary education as a ratio of the returns to incomplete (and in some cases complete) primary schooling or no schooling declined (figure 1-5)
That is big news because it signals a reversal of a trend. As can be observed in figure 1-5 also, in the previous decade returns to skill had been on the rise. Why did the reversal take place-did the relative supply of unskilled labor shrink or did the demand for skilled workers subside? The chapters do not analyze that question in the context of a full model of labor demand and labor supply. However, as shown in figure 1-8, in the period of declining (increasing) inequality, an increase in the relative supply of workers with tertiary education was accompanied by a decline (increase) in the relative returns to tertiary education (relative to primary levels), except for in Mexico, where the relative returns to schooling for tertiary levels continued to rise during the period of declining inequality, albeit at a slower pace than during the period of rising inequality. Thus the data suggest that while during the 1990s the demand for skills dominated the effect, in the last ten years the growth in the supply of skills outpaced demand and the college premium consequently shrank. Or, to use Tinbergen's language, in the race between skill-biased technological change and educational upgrading, the latter took the lead.
Did the returns to schooling change because educational upgrading caught up with the increase in the demand for skilled labor or because the demand for skilled labor subsided as the effects of technological change petered off? As Jaime Kahhat shows in chapter 2, in theory the presence of either factor can result in a decline in wage inequality. A review of existing models of exogenous and endogenous technological change reveals that more often than not, the effects of technological change are unequalizing at first but not in the long run. For example, after the learning phase is over and workers become fully efficient in using the new technology, firms substitute relatively expensive skilled labor with more economical unskilled labor. These models assume perfect capital markets, allowing the supply of skills to adjust to increases in demand. Even if the demand for skilled labor did not decline, wage inequality would fall if the supply of skilled workers caught up with demand. But what would happen if capital market imperfections impede or slow down the acquisition of skills?
Using a stylized model of capital market imperfections, Kahhat shows that when unskilled workers cannot borrow all that they want to invest in acquiring more skills, the share of skilled workers in the steady state is suboptimal: income per capita is lower than it would be if capital markets were perfect and no credit constraints existed. More important for our analysis, in the steady state with capital market imperfections, an economy that starts with an unequal distribution of wealth (and binding credit constraints) will have wealth inequality in the long-run equilibrium. In such a world, a redistributive policy that increases the share of skilled workers (for example, a substantial expansion in educational access that leads to an increase in the supply of skills) would increase the relative wage for unskilled labor. In turn, the higher wage for unskilled labor would provide greater opportunity for unskilled workers to invest in education. Consequently, both the wage premium and the inequality in educational attainment fall, reducing labor earnings inequality. Therefore, policies that enhance equality of opportunity in the present by extending subsidized educational services to underserved areas (rural areas and urban slums), improving the quality of education, and/or establishing conditional cash transfer programs (such as Brazil's Bolsa Familia or Mexico's Progresa/Oportunidades) allow larger fraction of the population to accumulate enough wealth to invest in upgrading its skills and improve its earnings in the future. That process can generate a virtuous circle leading to greater equity and growth in the long run.
Excerpted from Declining Inequality in Latin America by Luis Felipe Lopez-Calva Nora Claudia Lustig Copyright © 2010 by Luis Felipe Lopez-Calva and Nora Claudia Lustig. Excerpted by permission.
All rights reserved. No part of this excerpt may be reproduced or reprinted without permission in writing from the publisher.
Excerpts are provided by Dial-A-Book Inc. solely for the personal use of visitors to this web site.