Poverty and inequality in Latin America: a story of two decades.

Gasparini, Leonardo ; Cruces, Guillermo

Latin American countries have succeeded in reducing poverty and income inequality over the last decade thanks in part to both economic growth and deliberate social policy measures. This study provides an overview of the available evidence of the changes in income distribution that have occurred in Latin America over the past two decades and their causes. While some attribute the improvements in distribution to changes in the international economy and the positive trend in the Latin American countries" terms of trade, others highlight the influence of changes in public policy. Both of these two sets of factors may have played an important role and may have interacted with one another in various ways.

INTRODUCTION

Both poverty and income inequality decreased significantly in Latin America during the first decade of the twenty-first century. At the start of the century, 25 out of every 100 people in the region were living on less than $2.50 per day; today, only 14 out of every 100 are in that situation. (1) Other indicators of income poverty and of various types of material deprivation are consistent with this result. The region's income inequality has also declined considerably, with the mean Gini coefficient falling from 0.534 in 2002 to 0.499 in 2010. (2)

These improvements in social indicators have been linked to at least two factors: on the one hand, most of the region's economies have been experiencing robust growth together with upswings in employment and labor income; on the other, a majority of the countries have boosted social spending and have put wide-ranging social protection systems in place or have greatly expanded the scope of their existing systems.

These inroads into poverty alleviation notwithstanding, the social situation in Latin America remains a cause of concern. Over 150 million Latin Americans live on less than $4.00 a day at purchasing power parity, and 80 million of them are below the poverty line of $2.50 per day. (3) In addition, despite the progress made in the last decade, the degree of inequality is far from negligible. In fact, the available empirical evidence suggests that the only countries outside of Latin America that have comparable levels of income inequality are several sub-Saharan African nations and some countries in Southeast Asia. (4)

In the rest of this paper we document the changes that took place in the levels of poverty and inequality in Latin America during the 1990s and the 2000s and analyze some of the main arguments about the determinants of those changes. (5)

POVERTY AND INEQUALITY: THE EVIDENCE

Following the disappointing economic performances of the 1980s and the turbulent 1990s, Latin American economies expanded and became more equal in the first decade of the twenty-first century. The combination of rapid growth and improvements in income distribution triggered a steep reduction in income poverty that marks a stark contrast with the region's performance during the preceding decades. (6) Figure 1 shows the aggregate poverty rates for Latin America based on the poverty lines of $2.50 and $4.00 per day. Income poverty declined slightly in the 1990s in the wake of the 1994 Mexican crisis--known as the "Tequila crisis." Thereafter, the net result of opposing trends in different countries held nearly steady until 2003 and then began to fall sharply and continued to do so until the end of that decade, despite the global financial crisis from 2008 to 2009.

[FIGURE 1 OMITTED]

As in the case of poverty trends, the region's performance in terms of income distribution has changed a great deal in recent decades, with income inequality climbing in the 1980s and 1990s before falling off sharply in the 2000s. The break in this trend appears to have occurred around 2002 (Figure 2). Although some countries had begun to exhibit a more equal distributional pattern in the late 1990s (e.g., Brazil and Mexico), others did not begin to do so until the early 2000s (e.g., Argentina, Peru, and Venezuela). By 2002, distribution patterns in all the countries of the region were becoming more equal. While these distributional changes were not on a large scale, the downtrend in inequality certainly stands in contrast to the trend seen in the preceding decade.

It is also interesting to note that while inequality was declining in Latin America, it remained unchanged or was on the rise in other parts of the world. The

[FIGURE 2 OMITTED]

region's positive performance in this respect--both in absolute and relative terms-is particularly remarkable when viewed against the backdrop of its long-standing difficulties in achieving fairer distributional patterns. The "excess inequality" that has typified Latin America as compared to other world regions is now somewhat less marked than it was a decade ago. (7) The narrowing of the inequality gap between Latin America, Asia, and Eastern Europe is especially noticeable, as the sweeping economic changes occurring in these latter two regions have tended to heighten existing income inequalities in recent decades.

POVERTY, INEQUALITY, AND GROWTH

In order to work out what factors are behind these changes in poverty levels, we must first assess the changes that have occurred in terms of economic growth and income inequality. Figure 3 traces trends in growth, inequality, and poverty. During the 1990s, Latin American economies expanded at a moderate pace, providing a driving force behind a modest decrease in aggregate poverty, although inequality increased in a number of countries. The years around the turn of the century saw stagnant per capita GDP together with shifts toward greater inequality, which translated into higher levels of poverty. The 2000s, on the other hand, were a time of strong GDP growth, at least until 2008, coupled with improvements in wealth distribution. The combination of these factors, in addition to the global financial crisis having a limited impact on the region, contributed to a sizeable decrease in poverty in Latin America.

[FIGURE 3 OMITTED]

A decomposition exercise can be carried out to measure the quantitative influence of the "growth effect" and the "distribution effect" on changes in poverty? The former is the change in poverty that would have occurred if incomes had grown at the same rate, and hence the distribution had remained unaltered. The distribution effect is the change in poverty that would have occurred if only distributive changes had taken place, with no growth in mean income. In the 1990s, the growth effect translated into, on average, a 5 percent drop in poverty, while the redistributive effect translated, on average, into a one point increase in poverty. In the 2000s, trends became much more positive and homogeneous, with both growth and redistribution trends helping to bring poverty levels down in all countries. On average, the growth factor and redistribution factor contributed six and four points respectively to the decline in poverty. (9)

These results suggest that, in order to understand why the trends in poverty reduction in these two decades differ so sharply, it is important to look at the determinants of economic growth. However, it is even more informative to delve into the reasons for the substantial differences between the two periods in terms of income distribution.

The reasons for the decrease in inequality seen in the 2000s versus the increase that occurred in the 1990s are the subject of a heated debate in academic and political circles throughout the region. This research effort has focused mainly on the labor market and on public and private transfers while largely setting aside the role played by other sources of income such as capital, land rents, and business profits. The neglect of these other factors is essentially due to the fact that household surveys fail to capture these income sources properly. Estimates of functional income distribution based on national accounts information suffer from a number of shortcomings and are therefore no more than a minor palliative for the problems involved in estimating non-labor income. (10)

THE LABOR MARKET

Since labor income accounts for over 70 percent of the total income captured in Latin American household surveys, changes in the labor market have a strong impact on income distribution. This fact is corroborated when the sources of changes in wealth distribution are decomposed. (11) In particular, there is a close correlation between changes in the wage gap between skilled and unskilled workers, known as the "wage premium," and income inequality. Controlling for other observable factors, Gasparini et al. report uneven trends in this wage gap over time--an average annual increase of 1.8 percent in the 1990s versus an annual reduction of 2.8 percent in the 2000s (Figure 4). (12)

[FIGURE 4 OMITTED]

In what follows we will discuss changes in the wage premium driven by three different factors: the labor supply, the demand for labor, and labor-market policies and institutions.

Labor Supply

Latin America has been making great progress in terms of formal education. The average number of years of formal education for adults and school enrollment rates for children and young people have been climbing in every country of the region for decades. This has been reflected in the expansion of the relative supply of skilled labor. Figure 4 indicates that for the region as a whole, the rate of increase in this supply has been fairly even over the past two decades with the mean annual growth rate coming to 3.4 percent. If it were not for the changes that have occurred in labor demand and labor institutions, this rising level of education would have been reflected in a narrowing of the wage gap by a similar extent in each of these last two decades. The fact that this is not what actually occurred suggests that changes in relative labor demand, labor policies, or both, have played a crucial role. (13)

Labor Demand

Figure 4 illustrates the change in relative demand that would be consistent with the changes seen in wages and in the labor supply under certain elasticity-of-substitution assumptions. While this is a conjectural exercise, it serves to highlight the need to seek factors that may have had differing effects on the relative demand for skilled and unskilled labor during the past two decades. In the following sections we briefly review a number of these factors.

Reforms

Latin American countries undertook a series of pro-market structural reforms aimed at modernizing their economies. These reforms--although introduced by some countries in the 1970s (Chile) and 1980s (Mexico)--were a hallmark of the 1990s. They included opening economies to international trade and capital flows, boosting foreign direct investment, and increased privatization and deregulation. There is a wide-ranging consensus in the literature that, at least in the short and medium terms, these reforms heightened existing inequalities by curbing the relative demand for unskilled labor and thereby depressing wages and reducing job opportunities--particularly in the formal sector--for these workers. The impact of these reforms did differ across countries, however. For example, there is no evidence that trade liberalization had an unequalizing effect in Brazil, but there is indeed evidence that it had this kind of effect in Argentina, Colombia, Peru, and Mexico. (14)

Some of these reforms spurred investment in physical capital, which had been lagging for at least a decade. This investment shock triggered a sudden influx of new technologies such as mechanization, robotization, and, in general, the growing usage of computers, all of which initially tended to displace unskilled labor. The unequalizing impact of skill-biased technological change has been documented in developed countries as well as in Latin America. (15) In several countries these economic changes took place against the backdrop of weak social and labor institutions and therefore had a marked impact on social indicators.

Most of these reforms were launched in the 1990s but were not consolidated in the decade that followed; in fact, in a few countries some steps were taken to reverse them such as protectionist measures, re-nationalization, and tighter regulations. This differentiated pattern may have been a fundamental factor behind the differences seen in the way that relative labor demand has changed over the past two decades.

It is possible that the increased inequality caused by these changes was actually an overshoot. (16) Certain types of sudden economic changes such as the incorporation of a new technology, may initially have a strong unequalizing impact when unskilled labor is displaced. However, as time passes, it is highly likely that displaced workers will be relocated to other sectors. Even if these workers' initial wage levels are not as high as before, the extent of labor inequality will clearly decline from its peak level after the initial impact of the reform has passed. This argument could provide an explanation for at least part of the decrease in labor inequality seen in the 2000s in Latin America.

Relative Prices

The relative price structure of Latin American economies shifted during the past two decades as a result of external shocks and domestic policies. (17) In the 2000s, in particular, the prices of commodities produced by the region soared, especially in the case of agricultural goods. Interestingly, the rural-urban income gap narrowed in all the countries in the region in the 2000s after having widened or remained steady in the 1990s. This reduction in the rural-urban gap, which was partially linked to changes in international prices, was clearly an equalizing factor.

While the economic signals of the 1990s brought about changes in the production structure that were biased toward skilled-labor-intensive sectors, the modifications seen in the relative price structure in the 2000s were more neutral. Gasparini et al. illustrates this contrast by constructing an indicator for relative skilled-labor demand based on sectorial changes in production and employment. These changes had a pro-skilled-labor bias in the 1990s but were roughly neutral in the 2000s.

The improvement in the terms of trade has--in addition to altering the production structure via price signals--made resources available that have had an impact on distribution via employment and more ambitious social policies.

Employment

Thanks to more favorable external conditions and sound macroeconomic policies, the countries of the region succeeded in expanding aggregate demand and thus lowered their unemployment rates during the 2000s. This helped to reduce inequality in two ways: directly, by boosting the incomes of people who had previously not been employed, and indirectly, by putting increasing upward pressure on wages in the labor market. The influence exerted on distribution by this greater labor demand appears to have been quantitatively significant in some cases such as Argentina, but less so in others such as Brazil and Mexico. (18)

Crisis and Stabilization

Major macroeconomic crises are associated with spikes in inequality. Once the economy has begun to return to some degree of normality, the level of inequality tends to subside. A number of countries were hit by severe economic crises around the year 2000 (Argentina, Colombia, Ecuador, Paraguay, Uruguay, and Venezuela), and all of them witnessed steep drops in GDP and huge, but short-lived, spikes in poverty and inequality. Part of the decline in inequality that came in the wake of these crises can be accounted for by the reestablishment of economic links that had been disrupted by such a major shock. This rebound effect has been especially strong in the years following the economies' stabilization, with poverty and inequality falling almost as fast as they had risen during the crisis. (19)

Labor Policy

In the 2000s many countries of the region changed their approach and began to introduce more proactive labor policies. They raised the minimum wage, backed trade unions supporting their involvement in wage negotiations, and put income policies into place. Minimum wages were revised upward in many countries and helped to increase the earnings of low-income workers. The readjustment of the minimum wage was partly a result of a more interventionist policy approach, but this readjustment also benefited from a substantially stronger labor market. The impact of a higher minimum wage on distribution is not as obvious as it might seem, since it applies only to the formal sector of the economy, where a majority of workers do not come from low-income households. Nonetheless, the empirical evidence suggests that a minimum wage hike influences the pay levels of informal workers as well, and therefore ultimately has an equalizing effect. (20)

SOCIAL POLICY

The reduction in inequality and poverty in the 2000s in Latin America is also associated with higher social spending and the introduction or expansion of social protection systems. While some countries have expanded or enhanced the benefits provided under their social protection programs in the formal labor market, the main innovation has been the implementation of broad-coverage, noncontributory programs. One of the most important types of initiatives in this area has been the introduction of conditional cash transfer (CCT) programs. Most of these programs are targeted to poor households with children; others provide noncontributory pensions for older adults living in poverty. (21)

The first CCT programs were launched in the mid-1990s and the number of such programs multiplied rapidly. By 2010, eighteen Latin American and Caribbean countries had a CCT program, and the coverage of these programs amounted to 19 percent of the total population of the region. An analysis of the distributional impact of the main programs of this type in the region indicates that they have achieved substantial reductions in extreme poverty levels and have had a smaller, but still significant, effect on inequality. (22) A number of studies have found that these programs have also had positive effects in terms of education and health. (23)

Most CCT programs are designed for households with children and, therefore, usually do not afford protection for the older adult population. In countries with well-developed social security systems, retirees receive pensions based on their earnings in the formal sector of the economy. However, since Latin American countries have high levels of informal employment, large segments of the population are not entitled to a pension when they retire. The pension system's coverage in Latin America is very limited: in 2008, less than 20 percent of the retiree population was receiving pension income. In light of this situation, many countries have recently begun to introduce social protection systems for poor retirees. Argentina, Brazil, Bolivia, and Chile among others, have started up noncontributory pension programs. While there is less evidence available on the impacts of this type of pension, a review of the literature indicates that there have been sizeable reductions in poverty in this age group. (24)

POLICIES OR EXTERNAL CONDITIONS?

In the preceding sections we have discussed the factors underlying the drop in poverty and inequality in Latin America in the 2000s. Most of these factors can be linked to improved external conditions, a change in the policy paradigm, or both. In fact, the current debate in the region revolves around this question, and the answer is important not only to understand the past, but also to assess the prospects for consolidation of the improvements in income distribution in the future.

The Latin American economies were favored by the steady growth of the global economy in the 2000s, which boosted demand for their products. The terms of trade of the Latin American countries, which had wandered up and down in the 1990s, began to surge in the early 2000s--rising by 30 percent, on average, between 2002 and 2010. The combination of these improved terms of trade with better financial conditions and robust inflows of remittances, gave the region a brighter economic outlook than it ever had before. This loosened the external constraints that have traditionally limited Latin America's growth, and opened the way for higher earnings, employment, and a more ample supply of fiscal resources--all of which made it possible to put more ambitious spending policies into place.

Other authors, while recognizing the role played by better external conditions, believe that the most influential factor has been the shift in public policy toward the center-left during the 2000s. (25) This policy shift has been coupled with a widespread intensification of public policy interventions, especially in the case of social and labor policies, and the rollback of some of the pro-market reforms that had been introduced earlier on by a number of countries.

The counterargument is that these social policy inroads have only been possible because of the greater abundance of resources generated by more favorable external conditions, and that improvements in distribution have been seen in all the countries of the region, regardless of their political models. Some authors point to the somewhat greater reductions in poverty achieved by more leftist governments (e.g., Venezuela and Argentina), while others contend that the gains achieved in those countries are due partly to their economies' rebound from major crises and partly to the stronger increase in their terms of trade. (26)

In point of fact, it is quite possible that the more favorable external conditions and more proactive public policies, as well as the interactions between the two, have had an impact on income distribution. But it is difficult to analyze the different channels in isolation, and it is therefore challenging to assess how influential each individual factor has been in quantitative terms. In order to answer such an important question, further research will be needed to arrive at a more accurate understanding of the ultimate causes of the improvement observed in income distribution in Latin America.

NOTES

(1) Unless explicitly mentioned, the source for all statistics cited in this paper is the Socio-Economic Database for Latin America and the Caribbean (SEDLAC), a joint project by CEDLAS and the World Bank (sedlac.econo.unlp.edu.ar).

(2) The Gini coefficient is the most popular indicator of economic inequality. The Gini coefficient ranges from 0 (perfect equality) to 1 (total inequality).

(3) The $2.50 per day line adjusted for purchasing parity power (PPP) is an international poverty line that allows poverty comparisons across countries. It coincides with the median value of the extreme poverty lines chosen by the Latin American countries. The $4.00 poverty line is similar to the median of the official moderate poverty lines.

(4) Facundo Alvaredo and Leonardo Gasparini, "Recent trends in inequality and poverty in developing countries," Handbook of Income Distribution 2 (forthcoming); The mean Gini coefficient for the distribution of consumption per capita in 2008 is 0.440 in Latin America and the Caribbean, 0.447 in SubSaharan Africa, 0.383 in East Asia and Pacific, 0.355 in South Asia, 0.365 in Middle East and North Africa, and 0.337 in Eastern Europe and Central Asia.

(5) Luis Lopez Calva and Nora Lustig, eds., Declining Inequality in Latin America: A Decade of Progress? (Virginia: UNDP and Brookings Institution Press, 2010); Darryl McLeod and Nora Lustig, "Inequality and Poverty Under Latin America's New Left Regimes" (discussion paper prepared for Fordham University, 2010); Joao Pedro Azevedo, Louise Cord, and Carolina Diaz-Bonilla, "A Break with History: Fifteen Years of Inequality Reduction in Latin America" (Washington, DC: World Bank, 2011); Leonardo Gasparini, Guillermo Cruces, and Leopoldo Tornarolli, "Recent trends in income inequality in Latin America," Economia 10, no. 2, (2011); Leonardo Gasparini and Nora Lustig, "The rise and fall of income inequality in Latin America," in The Oxford Handbook of Latin American Economics, ed. Jose Ocampo and Jaime Ros (Oxford: Oxford University Press, 2011), Chapter 7; and G.A. Cornia, "Inequality trends and their determinants: Latin America over 1990-2010" (working paper no. 2012/09, UNU-WIDER, New York: 2012).

(6) Poverty is a complex, multidimensional phenomenon. For the sake of simplicity, this analysis focuses on monetary poverty and, more specifically, on income poverty, given the scarcity of data on consumption in most of the household surveys conducted in Latin America.

(7) For evidence on Latin American excess inequality, see Juan Luis Londono and Miguel Szekely, "Persistent poverty and excess inequality: Latin America, 1970-1995," Journal of Applied Economics 3 (2000), 93-134.

(8) Guillermo Cruces and Leonardo Gasparini, "Politicas sociales para la reduccion de la desigualdad y la pobreza en America Latina y el Caribe. Diagnostico, propuesta y proyecciones en base a la experiencia reciente," Premio Functacion Vidanta, Contribuciones a la Reduccion de la Pobreza y la Desigualdad en America Latina y el Caribe, 2012.

(9) Nora Lustig, Luis Lopez Calva, and Eduardo Ortiz-Juarez, "Declining inequality in Latin America in the 2000s: The cases of Argentina, Brazil and Mexico," World Development Report, (2012) (working paper); The aforementioned report finds that the growth and redistribution effects were of similar strength in the 2000s.

(10) For instance, it is difficult for national accounts to separate out the income of the self-employed into the components of labor and non-labor income.

(11) Javier Alejo, Marcelo Bergolo, Fedora Carbajal, and Guillermo Cruces, "Cambios en la desigualdad del ingreso en America Latina: Contribucion de sus principales determinantes (1995-2006)" (working paper, CEDLAS, La Plata: 2009).

(12) Leonardo Gasparini, Sebastian Galiani, Guillermo Cruces, and Pablo Acosta, "Educational upgrading and returns to skills in Latin America: Evidence from a supply-demand framework, 19902010" (IZA Discussion Paper 6244 and Policy Research Working Paper WPS 5921, World Bank, Washington, DC: 2011).

(13) Although, on average, the rate of increase in the relative labor supply has been similar across the two decades, there have been notable differences in some countries. There are findings that in Mexico the substantially sharper increase in the supply of skilled labor in the 2000s accounts for a large share of the reduction in the wage gap during that decade; Gerardo Esquivel, Nora Lustig, and John Scott, "A Decade of Falling Inequality in Mexico: Market Forces or State Action?" in Declining inequality in Latin America: A decade of progress? ed. L. F. Lopez-Calva and N. Lustig (Washington, DC: Brookings Institution and UNDP, 2010), 175-217.

(14) Gordon Hanson and Ann Harrison, "Trade Liberalization and Wage Inequality in Mexico," Industrial and Labor Relations Review 52, no. 2 (1999); Jere Behrman, Nancy Birdsall and Miguel Szekely, "Economic Policy and Wage Differentials in Latin America" (working Paper 29, Center for Global Development, "Washington, DC: 2003); Rob Vos, Enrique Ganuza, Samuel Morley, and Sherman Robinson, eds., 14//20 Gains from Free Trade? Export-led Growth, Inequality and Poverty in Latin America (London: Routledge, 2006); and Pinelopi Goldberg and Nina Pavcnik, "Distributional Effects of Globalization in Developing Countries," Journal of Economic Literature 45, no. 1 (2007).

(15) Robert Feenstra and Gordon Hanson, "Foreign Direct Investment and Relative Wages: Evidence from Mexico's Maquiladoras," Journal of International Economics 42, nos. 3-4 (1997); Carolina Sanchez-Paramo and Norbert Schady, "Off and Running? Technology, Trade and Rising Demand for Skilled Workers in Latin America" (working paper, World Bank, Washington, DC: 2003); Pablo Acosta and Leonardo Gasparini, "Capital Accumulation, Trade Liberalization, and Rising Wage Inequality: The Case of Argentina," Economic Development and Cultural Change 55, no. 4 (2007); and Dierk Herzer, Philipp Huhne, and Oeter Nunnenkamp, "FDI and Income Inequality: Evidence from Latin American Economies" (working paper no. 1791, Kiel, Germany: 2011).

(16) David Card and John DiNardo, "The Impact of Technological Change on Low-Wage Workers: A Review," in Working and Poor: How Economic and Policy Changes Are Affecting Low-Wage Workers, Sheldon H. Danziger and Robert F. Schoeni, ed. Rebecca M. Blank (New York: Russell Sage Foundation, 2008).

(17) See Jose Luis Machinea and Andres Lopez, "Estructura productiva, crecimiento y equidad. Una mirada sobre America Latina" (unpublished manuscript, 2012).

(18) Luis F. Lopez-Calva and Nora Lustig, eds., Declining Inequality in Latin America: A Decade of Progress? (Brookings Institution Press and UNDP, 2010).

(19) Leonardo Gasparini and Nora Lustig, "The rise and fall of income inequality in Latin America," in The Oxford Handbook of Latin American Economics, ed. Jose Antonio Ocampo. and Jaime Ros (Oxford: Oxford University Press, 2011) Chapter 7.

(20) Mariano Bosch and Marco Manacorda, "Minimum wages and earnings inequality in urban Mexico," American Economic Journal: Applied Economics 2, (2010).

(21) Ariel Fiszbein and Norbert Schady, "Conditional Cash Transfers: Reducing Present and Future Poverty" (World Bank Policy Research Report, World Bank, Washington, DC: 2009).

(22) Cruces and Gasparini.

(23) Fiszbein and Schady.

(24) Nora Lustig, Carla Pesssino, and John Scott, "Commitment to equity: An assessment of fiscal policies in Argentina and Mexico" (working paper, Commitment to Equity Project, New Orleans, LA: 2011).

(25) Cornia and the discussion in Lustig et al.

(26) Lustig et al. report that the terms of trade rose by 70 percent between 1996 and 2007 in the countries that they categorize as having populist governments (Argentina, Bolivia, and Venezuela), 17 percent in those with social democrat governments (Brazil, Chile, and Uruguay), and 9 percent in the rest.

Leonardo Gasparini is a professor of economics at the Universidad Nacional de La Plata (UNLP), and the founder and director of the Center for Distributional, Labor and Social Studies (CEDLAS) at UNLP. Guillermo Cruces is the deputy director of CEDLAS, a professor of economics at UNLP and the Universidad de San Andres in Argentina, and a researcher at the National Scientific and Technical Research Council (CONICET) of Argentina. This analysis is based on a discussion paper prepared for a meeting on poverty and inequity in Latin America convened by the Vidanta Foundation in Puerto Vallarta in November 2012. The authors are grateful for the comments made by Jose Luis Machinea, Nora Lustig, and Pablo Gerchunoff