Income Distribution under Latin America’s New Left Regimes

Abstract

This paper reviews the decline in income inequality that has taken place over 2002–2007 in most Latin American countries against the background of its steady increase over 1980–2002. The paper then analyzes the factors that could explain this trend reversal. It focuses in particular on favorable external conditions, cyclical factors, improvements in the distribution of educational achievements and the subsequent drop in skill‐premium, and changes in macro‐economic and social policies introduced in several countries, particularly by a growing number of left‐of‐center governments that have come to power during the past decade. An econometric test for the years 1990–2007 using a sample of countries covering the majority of the population in the region indicates that, in addition to a favorable business cycle and external conditions, a decline in skill premium and the new policy model of fiscally prudent social‐democracy that is emerging this decade in much of Latin America impacted favorably the distribution of income. If this approach will survive the current crisis, much of the recent inequality decline is likely to become permanent.

Key words:

• Income inequality
• Terms of trade
• Policy regimes
• Latin America

Acknowledgements

I would like to acknowledge the comments on prior versions of this paper received from an anonymous referee, as well as from Francois Bourguignon, Guillermo Cruces, Michael Grimm, Rafael de los Hoyos, Saul Keifman, Stefan Klasen, Juan Carlos Moreno‐Brid, Abijit Sen, Rolph van der Hoeven and participants to conferences or seminars held at ISS‐Den Haag (18 September 2008), University of Montevideo (30 September 2008), UN House in Asuncion (2 October 2008), CERDI‐University of Auvergne (5 December 2008), TISS‐Bombay (13–15 March 2009), EcIneq conference in Buenos Aires (18–19 July 2009), and PegNet conference of Den Haag (3–4 September 2009). I would also like to thank Bruno Martorano for outstanding research assistance. Of course, all remaining errors are only mine.

Notes

1 CEPAL’s Economic and Social Survey 2008–2009 predicts a 1.9% fall in GDP in 2009 and a recovery of 3.1% in 2010, with a cumulative loss of 10 points of GDP growth in relation to the 2003–2008 trend.

2 The coefficient of variation of the national Gini indexes fell from 0.10 to 0.07 over 1992–2006 (Gasparini et al., 2009).

3 An important part of the gains in terms of trade left the region in the form of profit remittances, as the exploitation of natural resources in Latin America is often in the hands of TNCs. Chile and Peru account for over one‐half of the regional outflow of profit remittances, although they account for only 8% of the region’s GDP.

4 For instance, in Argentina, agriculture accounts for a modest 8% of the total labor force.

5 However, in Mexico most of the migrants come from low‐income groups (communication from Rafael de los Hoyos).

6 However, during the same period, the gap between rich and poor in accessing tertiary education widened.

7 Corporación Latinobarómetro is a non‐profit non‐governmental organization based in Santiago, Chile. Since 1995 it has carried out polls on various political topics by surveying 19 000 households from 18 countries of the region [http://www.latinobarometro.org].

8 Governments developed a variety of fiscal mechanisms for appropriating part of the increase in commodity prices (CEPAL, 2007a, p. 31). Argentina financed part of its spending from resources generated by export duties. In turn, Venezuela, Bolivia and Chile created new taxes to increase the revenue generated from their non‐renewable resources. As a result, the share of fiscal resources represented by such revenue in Bolivia, Chile, Colombia and Mexico rose from of 27.8%, 7.6%, 9.9% and 29.4% in the 1990s to 34.8%, 20%, 14.2% and 37.5% in 2006–2007.

9 Such policy requires that the build‐up of international reserves during upturns be matched by measures to sterilize their monetary impact. Sterilization of this type is easier when there is a fiscal surplus. Otherwise it is necessary to sterilize via a mix of traditional open market operations, sales of central bank bonds in the market, or higher reserve requirements. For this reason, a fiscal surplus is an essential complement to the policy aiming at maintaining a stable and competitive real exchange rate.

10 Minimum wages varied between 20 and 143% of low‐skilled wages, with the number of beneficiaries varying between 1 and 20% of the labor force.

11 In 2005, Cuba, Uruguay, Brazil, Argentina, Bolivia, Costarica and Panama had social expenditure/GDP ratios of 15–20% (near the OECD level), but most Central American and Andean countries had ratios below 10%.

12 Since 1996–1997, Bolivia, Honduras and Nicaragua enjoyed debt cancellations equal to 5%, 6% and 2% of their GDP.

13 Of the (18 × 18) 324 observations of the Gini coefficient of income inequality, 175 derive from the SEDLAC database [www.depeco.econo.unlp.edu.ar/sedlac/eng/statistics.php] and are obtained through a standard procedure using household surveys data, 11 are taken from WIDER’s WIID2c [www.wider.unu.edu/wiid/wiid.htm], three from Badeinso‐Eclac 2008 [www.eclac.cl/estadisticas/bases/], 13 from WDI 2007, and one (Argentina in 2007) from national sources. Ninety‐eight data‐points were interpolated by filling gaps of one or two years in time series with stable trends. In three cases, interpolation was used to fill gaps of three years, and in three cases of four years; that is, for a total of 21 data‐points referring mostly to the early 1990s. Finally, 23 cells (for Ecuador, Guatemala, Nicaragua, and Paraguay in the early 1990s) remain blank. A successful check was carried out to ensure that the data filled in by interpolation replicated the trend of other income concepts with available data. In most cases, the data refer to disposable household income per capita. In a few cases, it was not possible to find out the income concept used for computing the Gini coefficients, as this information was not included in the survey questionnaires. This might introduce some error in the measurement of the dependent variable. However, as it was possible to verify a strong co‐variance between trends of Gini coefficients based on different income concepts, this bias may affect the value of the country intercepts in the fixed effect estimation, without affecting the parameters of the explanatory variables. All data cover the nation as a whole, except for Argentina (where surveys initially covered the Greater Buenos Aires, then the 15 main cities, and later the 28 main cities), Bolivia (where between 1990 and 1995 the coverage was only urban), and Uruguay (urban coverage only).

14 In this regard, it must be noted that reverse causality makes no sense in the majority of the relations in Table 9. For instance, it is not plausible that changes in domestic inequality affect the real exchange rate, or can affect lagged, exogenous or policy variables (such as Gini income 1990, migrant remittances, terms of trade, ratio of direct/indirect taxes, ratio of pension coverage Q1/Q5, and minimum wage). Also, a fall/increase in Gini income may affect the Gini of years of education only after a considerable lag. It is also implausible that a decrease in inequality will affect the expenditure on social insurance/GDP, which depends on the coverage of formal employment as far as pensions are concerned, and on tax revenue and public expenditure allocation for conditional cash transfers. The only relation in which reverse causation may be plausible is that between the Gini inequality and the growth rate of GDP per capita. In this case, however, this relation would be characterized by time lags, thus excluding the possibility of reverse causation on synchronous data. Furthermore, the literature on the impact of higher inequality on GDP per capita growth is not unanimous. Neokeynesian and neoclassical models postulate a positive relation between these two variables, while ‘political economy’ and ‘incentives’ models assume a negative one. On the whole, reverse causality does not seem plausible.