Top incomes and the ruling class in Latin American history. Some theoretical and methodological challenges

ABSTRACT

Recent studies on income inequality have some characteristics that differentiate them from their earlier counterparts. The spotlight on high incomes has illuminated a new angle from which to view income inequality. Because estimates of top income shares can be used as a proxy for power inequality, they can enrich our comprehension of the role of the elite in Latin America’s economic development. However, scholars interested in studying the history of economic inequality in Latin America face certain methodological and theoretical problems of their own: (1) because food and other commodities such as minerals represent the lion’s share of exported goods in Latin America, cycles in commodity prices have shaped the region’s economic history. Thus, the crux of income inequality in Latin America is who becomes richer and who becomes poorer when exports prices rise and fall; and (2) the sort of fiscal statistics typically used capture only a few countries and sometimes only limited periods. Thus, as I argue, scholars should use dynamic social tables to produce new information. I exemplify both points with a historical analysis of three Latin American countries: Chile, Colombia, and Argentina.

KEYWORDS:

• Inequality
• Latin America
• ruling class
• top incomes
• social tables
• ground rent inflows

Acknowledgements

This paper was presented at the Second World Inequality Conference at the Paris School of Economics (December 2021) and at the 15th ‘Jornadas de Historia Económica’ organized by the Uruguayan Economic History Association (Montevideo, December 2022). I acknowledge those who attended both conferences, particularly Pablo Astorga and Camilo Martínez. I also express my gratitude to María Gómez León, Branko Milanovic and two anonymous referees for their thorough reading and insightful comments on earlier versions of this paper.

Disclosure statement

No potential conflict of interest was reported by the author.

Notes

1 Paradoxically, as Klein (2017) and others have argued, most historians, who emphasize the humanistic roots of history, have abandoned their commitment to fully integrate it into the social sciences. However, there is no good reason why historians should deny history’s humanistic origin and renounce its legacy in order to recognize it as a social science. In fact, the political economy approach offers a good opportunity to reconcile the two traditions.

2 The trend also applies to economic historians inspired by Marxism and the Annales school (Bértola and Rodríguez Weber 2016).

3 ECLAC economists are known because they promoted a development strategy centred on industrialization, which proved to be unsustainable. I would like to underline the way they approached the problem of Latin American development and their focus on the historical specificities which characterized its variety of capitalism, not the solutions they provided.

4 In his study of Brazil, Gómez León (2021) estimated the Gini index to be 0.29 in 1870, 0.23 in 1900, and 0.31 in 1920. However, these estimates are likely to underestimate the level of income inequality because they do not take into account capital concentration. The study only considers one category of ‘owner’ for each economic sector (agriculture, industry, and services), and fails to consider the significant level of inequality among owners within each sector, such as landowners, industrialists, and merchants. As a result, these estimates might present an unrealistic picture of a relatively equal slave-owning and post-slavery society, rather than reflecting the actual level of income inequality. From 1960 onwards, Gómez León used household surveys (from the All the Ginis dataset) to complete her picture of Brazilian inequality history. This caused a sudden jump in the Gini index from 0.35 in 1950, which was obtained from social tables, to 0.53 in 1960. Gómez León interpreted this change as the upward side of a Kuznets curve. I believe instead that this change is more likely an illusion created by the use of different methodologies. However, other scholars before Gómez-León have also argued in favor of a Brazilian Kuznets Curve too; see Milanovic (2016, p. 82). I thank María Gómez León to raise my attention to this reference.

5 See also section 4.

6 In any case, does there exist a methodology which does not have them?