The Challenge of Decimating Poverty in Latin America by 2010

Contenido principal del artículo

Albert Berry


Probablemente sea aún muy pronto para llegar a cualquier conclusión definitiva sobre el efecto de los ajustes estructurales a la distribución de ingresos y su efecto sobre la evolución de la pobreza en América Latina; sin embargo, este trabajo intenta evaluar las tendencias arriba mencionadas en siete países lationamericanos durante la década de los ochenta. La evidencia que aquí se presenta apunta hacia una correlación entre los procesos de cambio acelerado en los regímenes de comercio de tales países y una tendencia hacia la concentración de los ingresos. A la vez, este artículo subraya que los países con un proceso gradual de reformas comerciales han mostrado un mejoramiento en su distribución del ingreso.ABSTRACTIt is probably too early to draw any definite conclusions on the effect of structural adjustment in the distribution of income, and the evolution of poverty in Latin America. This paper, however, represents an attempt to assess income distribution trends and the evolution of poverty in seven Latin American countries during the 1980's. The evidence presented here points at a correlation between processes of rapid change in the countries' trade regimes, and a tendency toward the concentration of income. At the same time, this article emphasizes how countries with a gradual process of trade reform have shown an improvement in their distribution of income.

Detalles del artículo

Cómo citar
Berry, A. (2017). The Challenge of Decimating Poverty in Latin America by 2010. Frontera Norte, 6(1e), 101–119.


As of the 1960s and early 1970s all of the Latin American countries had very high levels of inequality by the standards of other less developed countries, with the exceptions of Cuba (by then a centrally planned socialist economy), Argentina, and Uruguay. Somewhat less inegalitarian than those but still better than the regional average were Chile, Costa Rica, and probably Venezuela. The most common explanations of the lower inequality in the Southern Cone included their higher level of development (e.g., farther along in the Kuznets cycle) with associated development of social security systems, wage protection, etc., and their greater racial homogeneity.

Oscar Altimir, The Extent of Poverty in Latin America, World Bank Staff Working Paper (Washington, D.C.: The World Bank, 1982). Data were not available for all countries, but those excluded had only 12 percent of the region's population and were not obviously atypical in terms of degree of inequality. Since the data relate (in all or nearly all cases) to the distribution of households ranked by household income, the share of people below the poverty lines might be somewhat different from what these figures show, though it is not clear in which direction they may be biased.

Assuming the distribution of income for the region as a whole was not dissimilar to that observed for Colombia in 1970, Colombia's Gini coefficient was in the middle of the pack at that time.

If this extra period of growth brought with it a significant tightening of the labor market, then it would be realistic to expect the income share of the bottom few deciles to rise (though perhaps not the bottom decile or so), in which case the improvements cited might even understate the case.

UNCTAD, Trade Liberalization in Chile, Trade Policy Series, no. 1 (New York: United Nations, 1992): 44.

Enrique A. Gelbard, "Changes in Industrial Structure and Performance under Trade Liberalization: The Case of Argentina," Ph.D. dissertation, University of Toronto, 1990, p. 46.

Guy Pfefferman and Richard Webb, "Poverty and Income Distribution in Brazil," Review of Income and Wealth 29:2 (1983); Albert Berry, "The Effects of Stabilization and Adjustment on Poverty and Income Distribution: Aspects of the Latin American Experience," 1992, mimeo.

Usually the most useful and reliable information comes from household income surveys, but their main defect is the systematically weak reporting of non-labor Incomes. When there is no reason to believe that the labor share has changed markedly or that the distribution of capital income has been altered, this underreporting is unlikely to greatly bias the estimated trends. During the 1980s, however, there is some reason to believe that the capital share has risen, as the result of higher interest rates, on government domestic debt among other things (David Felix and John P. Caskey, Baker to Brady to Chance? Tinkering with the Latin American Debt Crisis, Working Paper No. 140 ([St. Louis, Mo.: Washington University Economics Department, 1989]). During the crises themselves, a common pattern was government borrowing abroad or locally to shore up the exchange rate. This facilitated massive capital flight. Governments (e.g., those of Chile and Ecuador) essentially socialized private foreign liabilities, which are the domain of the rich; the Chilean Central Bank, pushed by the international banks to act as guarantor of private nonguaranteed foreign loans, subsidized debtors to the tune of about 4 percent of GDP over the period 1982-85 (Patricio Meller, Adjustment and Equity in Chile [Paris: OECD Development Centre, 1992]: 60). Later, when the crises had passed and structural adjustment begun, high interest rates remained the order of the day as part of the new financial orthodoxy. Our understanding of the net effects of the various impacts on capital incomes during this period is not adequate to say with certainty that the capital share has risen by enough to imply an overall trend to worsening since the onset of the crises, but that possibility must be borne in mind.

Data on the distribution among households in this same greater Buenos Aires region and among income earners in the country as a whole seem to move in parallel with those just cited for those time periods when they are available, which does not in either case include much beyond 1980. As a result it has been necessary to use the Buenos Aires earner data, but with considerable confidence that they do not misrepresent the trends that actually occurred among households in the nation as a whole (Albert Berry, "The Effects of Stabilization and Adjustment on Poverty and Income Distribution: Aspects of the Latin American Experience," 1990, mimeo.).

Berry, ibid., p. 31.

The relative price of tradeables to non-tradeables appears to have fallen between 1973-75 and 1986-87. Though the real exchange rate, defined as the nominal rate adjusted for relative inflation, rose somewhat over this period as a whole, inequality rose between subperiods when this rate did not rise (Berry, ibid., p. 39).

Gelbard, op. cit., p. 54.

Paradoxically, the data on distribution among income recipients, while showing the same cycle as for the household distribution, do not indicate that the level of inequality was greater in the late 1970s than in 1970. This anomaly, still to be fully explained, does not greatly diminish the likelihood that household distribution did worsen significantly. A problem with the Chilean information, as with that for Argentina, is that distribution data over time are only available for greater Santiago, not for the country as a whole. But Santiago is probably fairly representative of the country, as suggested by the similarity of measured inequality for the few years for which both city and national data are available. There is no automatic inconsistency in the different trends shown for the income recipient and the household distributions, since the relationship between the two can change with family composition or with the participation of secondary workers. Still, of course, it would be possible to have more confidence in the conclusions suggested here if this difference were already satisfactorily explained. Another inadequacy of the available calculations is their failure to take account of changes in the relative prices of the consumption items purchased by different income classes. Over the course of the 1980s the increase in the relative price of food may have made the distribution trends worse than the figures on nominal distribution of income make them out to be. Note that the suddenness of the increase in recorded inequality between 1975 and 1976 may be related to the severe inflation at the time, which can produce volatility in the estimates.

Over that period private consumption per person fell by about 13 percent and the share of the bottom quintile by 32 percent.

Meller, op. cit., p. 23.

The high incidence of television sets (over 70 percent), refrigerators (49 percent), radios

Meller, op. cit., p. 27. Note that the direct effects of this concentration might be felt almost entirely within the top 10 percent of the income distribution.

Economic Commission for Latin America and the Caribbean, Preliminary Overview of the Economy of Latin America and the Caribbean 1992(ECLAC, 1992): 42.

Ricardo Ffrench-Davis, "Economic Development and Equity in Chile: Legacies and Challenges in the Return to Democracy," paper presented at the conference "The New Europe and the New World: Latin America and Europe 1992," Oxford University, September 1992, p. 15. If the series cited by Ffrench-Davis (the source of the wage data is INE) is consistent with that reported by ECLAC (1992, ibid., p. 44), which shows an increase of 11.7 percent over 1989-92, then the 1992 figure is 3 percent above that of 1970.

Donald J. Robbins, "Relative Wage Structure in Chile, 1957-1992: Changes in the Structure of Demand for Schooling" (Harvard University, manuscript, 1994).

José Francisco Escandón and Albert Berry, "Colombia's Small and Medium Exporters and Their Support Systems" (Washington, D.C.: The World Bank, 1993, mimeo).

Ffrench-Davis, op. cit., p. 12.

The effects of freer trade and globalization on income distribution will be related, among other things, to where Latin America's comparative advantage falls. Given the middle-income status of most countries of the region, that advantage may rest less on unskilled than on fairly highly skilled labor, in which case the Hecksher-Ohlin effects of freer trade may be to increase the incomes of the latter group more than those of the former, especially given that the higher-skilled groups are also more mobile internationally and therefore likely to be affected directly (not just through trading patterns) by developed country wage levels.

Diana Alarcón, "Changes in the Distribution of Income in Mexico during the Period of Trade Liberalization," Ph.D. dissertation, University of California, Riverside, 1993, p. 105.

Juan Luis Londoño, Income Distribution in Colombia 1971-88: Basic Estimation (Report to the World Bank, 1989). In this Colombian case a complicating factor is the presumed failure of the figures to pick up most of the income generated in the illegal drug trade. Its inclusion would probably raise the estimated inequality somewhat.

Comisión Económica para América Latina, Antecedentes Estadísticas de la Distribución del Ingreso: Venezuela, 1957-1985 (Santiago; United Nations, 1988). The Venezuelan survey data, which begin on a systematic basis in 1976, show a gradual decrease in inequality from that point until the early 1980s, with the Gini coefficient falling from 0.44 to 0.39 by 1982-83, after which there was little change, at least up until 1987 (pp. 30-31). The data refer only to monetary income from labor or self-employment, perhaps helping to explain the relatively low levels of calculated inequality and creating somewhat greater doubts as to the validity of any conclusion one reaches as to trends over lime than would be the case with better data. (Reported income falls short of consumption as estimated in the national accounts by 15-25 percent).

Jaime Tenjo," Evolución de los retornos a la inversión en educación 1976-89," in Educación, mercado de trabajo y desarrollo en Colombia, special issue of Planeación y Desarrollo (Bogotá: Departamento Nacional de Planeación, 1993).

Ministerio de Agricultura y Departamento Nacional de Planeación, El desarrollo agropecuario en Colombia, informe final: Misión de Estudios del Sector Agropecuario (Bogotá: Editorial Presencia, 1990), p. 228.

Rodolfo Hoffmann, "Evolucao da Distribucao da Renda no Brasil, entre pessaos e entre familias, 1978-86," in Mercado de Trabalhoe e Distribucao da Renda, ed. Guiherme Luis Sedlacek and Ricardo Paes de Barros (Vina Coletanea, 1989); Rodolfo Hoffmann, "A distribucao da renda no Brasil em 1985,1986 e 1987," Revista de Economía Política 9:2 (April-June 1989): 122-126.

Fluctuations in the measured Gini coefficient have been associated with the rate of inflation and the real exchange rate, and the Gini did reach historically high levels around 1990-91 but has since returned to the normal range (see the data presented in Eliana Cardoso, "Cyclical Variations of Earnings Inequality in Brazil," Revista de Economía Política 13:4 (52) (1993).

A hypothesis communicated to me by Ricardo Paes de Barros.

Timothy H. Gindling and Albert Berry, "The Performance of the Labor Market during Recession and Structural Adjustment: Costa Rica in the 1980s," World Development 20:11 (November 1992).

No very adequate over-time comparisons have yet been made due to noncomparability among the sources of data (Gindling and Berry, ibid.).

Alarcón, op. cit., pp. 139, 148.

William McGreevey et al., Brazil: Public Spending on Social Programs: Issues and Options, World Bank Report 3, 7086-BR (Washington, D.C.: The World Bank, 1988). More precisely, if it is necessary, the reasons are bureaucratic ones rather than technical or economic ones. The 1990s are not a time when Latin American countries can afford to be gentle with the inefficient parts of their bureaucratic apparatuses.

For a useful set of estimates see ECLAC 1992, op. cit., p. 47.

Based on the conclusions of an ongoing World Bank study of the export success and support systems of small and medium-sized manufacturing firms in Korea, Indonesia, Japan, and Colombia.