Latin American Inequality in the Long Run

This special issue of the Revista de Historia Económica — Journal of Iberian and Latin American Economic History contains a selection of papers originally presented at the conference «A Comparative Approach to Inequality and Development: Latin America and Europe,» organized by the three authors of this introduction. The conference was supported by the Instituto Figuerola (Universidad Carlos III de Madrid), and it met at the Fundación Ramón Areces on May 8-9, 2009. The conference was sponsored by the GlobalEuroNet-Research Networking Programme of the European Science Foundation, the Ministerio de Ciencia e Innovación (Acciones Complementarias), Fundación Ramón Areces and the HI-POD Project1. Each of the papers published here has undergone the journal’s usual blind peer review process.

Inequality has been a topic of intensive research and debate over the last decade, especially in Latin America. The aim of the conference was to put the recent Latin American experience with inequality in the long-term perspective of many centuries, and to promote a comparative analysis. Compared with Europe and other parts of the periphery, has inequality really been a distinctive aspect of Latin American development since the arrival of the Iberian conquerors? What form has it taken? Has it made an important contribution to the region’s disappointing growth performance over the 20^th century?

1 Background

Mainstream economics has discussed inequality mainly in terms of wealth and income distribution. Following Simon Kuznets’s lead, the mainstream tradition has mainly asked how modern economic growth has an impact on inequality. Kuznets himself stressed the role of structural change in income distribution, although he also paid attention to demographic change, social change, the evolution of political power and more.

Research on within-country and between-county inequality has also been assessed using standard economic tools. Seeing economic development as a process of changing demands for various factors of production, the basic approach has been to document how relative factor prices — land rents, wages, skill premia and financial returns — react in response to economic growth. The link between globalization and inequality has also attracted increasing attention as a force explaining within-country inequality change, between-country divergence and, more generally, world income distribution (Bourguignon and Morrisson 2002; Milanovic 2005; Williamson 2006). Given different initial conditions and endowments, global forces produce different distribution outcomes, as, for example, between mainly resource-abundant Latin America and labor-abundant East Asia.

While the search for the determinants of long-term inequality change is still ongoing, a new literature has developed, which explores, instead, the impact of inequality on growth. This shift in emphasis from the determinants of inequality to its impact has accompanied the development of endogenous growth theory, that is, the idea that advance in productivity must be explained endogenously in any plausible growth model. Inequality, human capital formation and productivity growth are thought to be closely connected: the more egalitarian the economy, the higher the educational attainment and the faster economic growth.

Other inequality-growth links include access to financial capital markets and political instability. If entrepreneurial capability is equally distributed but access to financial assets is not, then the growth rate will be lower in more unequal societies. If income and wealth are unfairly distributed, the poor and alienated have less to lose by revolt, violence and lawlessness, all having a negative impact on growth (Ponzio 2005; Bates et al. 2007). Even technical change has been linked directly to inequality. To the extent that the elite spends heavily on labor-intensive services in low-income societies while the abundant poor consume only foodstuffs and low-quality manufactures, there is little stimulus for the mass production of skill- and capital-intensive industrial goods where technical change is likely to be most rapid. These factors will, in turn, inhibit the chances for low-income countries to be competitive in world (or even domestic) markets for industrial goods.

Theory has developed further to encompass the role of institutions in fostering inequality. Even more, new institutional thinking emphasizes that differences in long-term performance between countries can be explained by differences in the institutions that impact on the distribution of income, wealth and political power (North et al. 2009; Acemoglu et al. 2005).

This discussion in the literature on inequality has also gone beyond economic growth. The human development approach pioneered by Amartya Sen and adopted by the United Nations links development to the capabilities of the population. In this sense, long and healthy lives, access to education and political and civil freedom are not merely or even mainly just inputs to a production function with economic growth as an outcome, but themselves are expressions of development (Dasgupta and Weale 1992). The way these capabilities are distributed among the population matters for development and living standards.

2 Proxies and Sources

Either because inequality has many different dimensions or because data on income have been hard to find in the distant past, researchers have increasingly explored inequality proxies. Thus, land rent–wage ratio trends have been quite successful in approximating inequality trends in pre-industrial societies in which land and labor are factors of production that matter. As the relative importance of land decreases with modern economic growth, the wage/GDP per economically active population (per capita if data are lacking) ratio has been shown to be another effective proxy for change in inequality, a proxy especially effective for early and middle stages of modern development. Its rationale is that while the denominator captures the returns to all factors of production, the numerator captures the returns to raw labor. This proxy works well until the mid-20^th century, after which the distribution of earnings becomes increasingly important. However, this proxy can pose measurement problems even for earlier periods. Historical GDP estimates are typically based on volume indices — implying some implicit GDP price deflator — while wages are deflated by consumer price indices; thus, if consumer goods prices evolve differently than do GDP deflators, the index may misinform. Another problem faced by these inequality proxies is that they do not allow us to make any international comparisons since they speak to changes over time, not levels. Nevertheless, when these proxies are linked to modern inequality estimates, long-run pictures of inequality trends can be obtained (Williamson 1999, 2002; Prados de la Escosura 2007a).

The Kuznets tradition has spent most of its energy in trying to see whether his famous inverted-U could be found in modern economic growth experience. One of the problems with this approach, however, is that it says nothing about international inequality, that is, about inequality between countries. International inequality had been tackled in the convergence-divergence debate that asks why growth rates differ between countries, and whether the prediction of the Solow model is confirmed (e.g. whether low-income countries catch up with high-income countries). However, the convergence-divergence debate does not speak to within-country inequality. To fill this void, Bourguignon and Morrisson (2002) offered some quantitative conjectures about the evolution of world income distribution, and by doing so made it possible to discuss between-county inequality, within-country inequality and world inequality all at the same time.

Another recent research direction has been to estimate incomes at the top using tax data, an especially useful enterprise for filling in trends across the 20^th century (Atkinson and Piketty 2007, 2010). For earlier times, the so-called social tables have been used to construct benchmark estimates for pre-industrial societies in which inequality data are especially scarce (Lindert and Williamson 1982, 1983; Milanovic et al. 2007).

Thus far, we have been talking mostly about the distribution of income. What about the distribution of wealth? Most historical studies of wealth inequality have dealt only with land, as it is the most easily available source of information. These studies have had difficulty in overcoming two severe shortcomings: no information on the majority of the population that does not own land; and the absence of information on other forms of wealth, like financial assets and education, which increase in importance over time.

Recent research has also explored other measures of well-being and inequality, including anthropometrics, or the so-called biological standard of living (Fogel et al. 1982; Komlos 1995; Steckel 1995). Anthropometric research has now been extended to the discussion of inequality. The basic idea is to measure the extent to which different socio-economic groups reveal different anthropometric trends (Baten 2000; López-Alonso 2007). Another alternative measure is the Human Development Index (HDI), which has also been the subject of historical analysis, even in Latin America (Astorga et al. 2005; Prados de la Escosura 2010). Indeed, recent work has added inequality measures to the HDI (Bértola et al. 2010). Since the HDI is constructed by weighting education by a third, it is hardly surprising that studies of schooling have flourished. And since education in the HDI is itself estimated by weighting literacy rates by two-thirds and school enrollment rates by a third, it is hardly surprising that these two have gotten most of the attention. Of course, many criticisms have been directed towards this index. For example, it has been observed that international differences in the HDI tend to vanish once most of the population is literate and has attained primary education. When higher levels of education are given a greater weight in the HDI, divergence trends reappear or at least convergence trends look much less robust (Bértola et al. 2010). Another problem posed by the estimates of school enrollment is the quality bias. Frankema (2009) has used new techniques to estimate the quality of primary education, showing that an improvement in school enrollment rates are often offset by diminished quality of schooling, falling attendance rates and rising drop-out rates. Similarly, we now have better estimates of average years of education (Morrison and Murtin 2009) as well as of inequality in the access to education (Bértola et al. 2010).

Since information on school enrollment is seldom available for the 19^th century and earlier, scholars have been exploring new ways to fill the human capital measurement gap. Numeracy studies estimate human capital by using information on the accuracy with which people reveal their age in censuses. Age heaping reveals the numeric capabilities of the population, and it correlates well with other human capital indices in which both are available for modern societies (Baten et al. 2010). This technique has also been used to study inequality (Mokyr 1985; Crayen and Baten 2009).

3 Latin America

Latin America has attracted much of the attention on inequality and for good reason. The continent has today the highest within-country inequality (López and Perry 2008). It has also failed to catch up with world leaders during the 20^th century, especially during the last few decades (Prados de la Escosura 2007b). This failure was particularly notable since, at the same time, other regions experienced fast growth and catching up. So the question quite naturally arose: Was there any connection between the lack of catch-up and high inequality?

Most of the research carried out in and on Latin America from the 1950s to the 1970s, emphasized the oligarchic features of the region’s development. These features appeared in colonial times as a high concentration of political power, wealth and income in the hands of the elite, where they owned most of the land and had control over trade and labor. While independence in the early 19^th century was followed by disruption, disorder, loosely established property rights and badly defined national borders, during the belle époque the first globalization boom brought with it a great strengthening of the elite’s power. It also happened while labor markets were subject to strict control. Still, these belle époque features had their roots in the colonial past.

The new institutional economic history recovered this research tradition, arguing that the institutions established shortly after colonization were responsible for a long-term equilibrium of political and economic inequality, low human capital formation and slow growth, and in which rent-seeking dominated. Even though their emphasis on centuries-old cultural and political heritage, resource endowments and socio-political arrangements varies, the different representatives of the new institutional economic history seem to agree on the basic ideas (Engerman and Sokoloff 1997; North et al. 2000; Sokoloff and Engerman 2000; Acemoglu et al. 2002; Robinson 2006).

This new literature has had an important impact on our thinking, not only in academia, but also in important international organizations (De Ferranti et al., 2004, Ch. 4). Yet, the views of the new institutional economic history have not remained uncontested. Coatsworth (2008) recently raised a formidable challenge by arguing that the roots of Latin American backwardness and inequality are to be found in the period 1770-1870, when the region missed the Industrial Revolution and the modernization it embodied2. He argues that the empowerment of the local elites did not take place in the early colonial times, but later on. Even more, he argues that this process of empowerment yielded the belle époque liberal reforms that made fast post-1870 growth possible. Although Coatsworth did not make the point, the new institutional economic history can also be criticized for almost completely abandoning the role of international relations as an important force by which domestic inequalities are strengthened and sometimes overturned (Bértola 2010; Williamson 2010), a central pillar of the earlier tradition.

4 This Special Issue

The present volume contains six articles, the collection of which speaks to the issues raised above. We are confident that it will make a significant contribution to what has become one of the central debates about today’s Latin American economies, and to illustrate the importance of history to contemporary analysts.