On the Character and Causes of Inequality in Latin America

Abstract

Diego Sánchez-Ancochea, The Costs of Inequality in Latin America: Lessons and Warnings for the Rest of the World. London: I.B. Taurus, 2020. 224 pp. £ 17.00 hardcover. Amy C. Offner, Sorting Out the Mixed Economy: The Rise and Fall of Welfare and Developmental States in the Americas. Princeton, NJ: Princeton University Press, 2019. 400 pp. £ 25.00 hardcover. Inequality is one of the most striking features of most Latin American societies, interacting both with people's everyday lived experiences and with deeply embedded social structures. As in all middle- and low-income countries, high levels of income inequality necessarily imply high levels of poverty, and this poverty is visible in rural backwaters, in urban shanty towns and in the streets of major cities. It implies spatial segregation as the rich attempt to keep the poor out of their communities, often employing armies of private security guards to do so (Jayadev and Bowles, 2006). It allows small economic elites to undermine democracy, both constitutionally through their domination of the media and the legal system, and unconstitutionally through military interventions. And it underpins extreme social inequalities, where the rich feel disgust towards the poor, and the poor see the rich as inhabiting another world. These inequalities have deep roots in Latin America's history. But it is not an immutable characteristic, and inequality has varied over time to a much greater extent than can be accounted for by any appeal to either geography or the colonial origins of the region's republics. This essay builds an argument for a historical approach to inequality based in particular on two recent books: The Costs of Inequality in Latin America: Lessons and Warnings for the Rest of the World by Diego Sánchez-Ancochea (2020) and Sorting Out the Mixed Economy: The Rise and Fall of Welfare and Developmental States in the Americas by Amy C. Offner (2019). Sánchez-Ancochea provides a wide-reaching exploration of the nature and consequences of inequality in Latin America, and insights into its proximate causes. Offner's account of the Colombian economy in the mid-20th century illuminates the role of partnerships between domestic capitalists and the US establishment in key institutional developments. While the two books have very different focuses, they are complementary for the purpose of building a picture of the deeper determinants of inequality. The first part of the article presents Sánchez-Ancochea's multidimensional analysis of inequality in Latin America. High income inequality matters directly for social welfare and social relations, but also has pernicious effects on other aspects of the economy such as growth and innovation. I then turn to Offner's institutional history of Colombia and widen the argument to the history of inequality in Latin America. Following this, I show how these two books and related literatures support an institutional-historical approach to analysing inequality. Both books describe barriers to inequality reduction in the region, which I synthesize into what I refer to as the internal and external constraints on progressive redistribution. These are caused by the reaction of domestic elites, the behaviour of international financial markets, and intervention by the USA. Sánchez-Ancochea's account of inequality in Latin America displays the depth and range of harmful effects that inequality has throughout society, working through multidimensional feedback loops. At the macroeconomic level it contributes to poor economic performance as uncompetitive and stagnant elites have little incentive to invest in more advanced economic sectors. This may explain why Latin America spends a much smaller share of GDP on research and development than Asia or the developed countries, leading to relatively little indigenous technical progress. Moreover, the rich have been successful in keeping taxes low, leading to historically weak investment in education and other public goods; what public investment there is has disproportionately been targeted at services for the rich. Personal income taxes account for less than 10 per cent of total tax revenues in the region compared with an OECD average of nearly 25 per cent. In eight Latin American countries the top 10 per cent of the income distribution pay less than 5 per cent of their incomes in income taxes. Tax loopholes that primarily benefit the rich further allow them to avoid an estimated 4 per cent of GDP in taxation. Due in large part to the low levels of taxation, low investments in education for the majority have been a long-running feature of most Latin American countries. The rich send their children to private schools, so they have little incentive to invest in the public sector. This not only contributes to low economic growth but also feeds back to high inequality, as educational levels in most of the population remain relatively low and are strongly correlated with lifetime incomes. High inequality in a middle- or low-income capitalist economy necessarily implies low wages for the majority, and Sánchez-Ancochea, like many observers, links these low wages with labour market informality. However, while there is no doubt that informality and inequality go hand in hand, recent research casts doubt on the claim that one causes the other. It is more likely, instead, that they have a shared cause in the lack of good jobs — itself a result of under-investment at the macroeconomic level.11 See Bleynat et al. (2021) for a historical application of this argument to Mexico building on the Lewis model of development. This means that formal work can also be very unrewarding. For Mexico, Maloney (1999) finds that workers move between formal and informal work without expressing a clear preference for either, while Bleynat and Segal (2021) find that many low-paid workers complain vocally about ill-treatment in formal employment, often preferring the independence of small-scale and own-account work. That is, a low-paid and low-status formal job is often not preferred to informality. And from the perspective of inequality reduction, consider the case of Bolivia, which reduced its Gini coefficient by 20 points in the period 2000‒19 — not just the greatest inequality reduction in the region, but one of the most dramatic ever seen globally outside of full social revolution. It was achieved not through any reduction in informality, but in considerable part through a rise in informal sector incomes (Mauricio Vargas and Garriga, 2015). Informality should probably be seen as a characteristic, not a cause, of high inequality.22 The concept of ‘informality’ has a complex history and numerous possible definitions. See Portes and Haller (2010) for discussion. Income inequality goes hand in hand with social inequality, as famously captured by the Brazilian photographer Tuca Vieira in his photograph of a Sao Paulo favela just next to an elegant apartment building sporting a swimming pool on every balcony.33 For Vieira's 2004 photograph, see www.commondreams.org/views/2020/12/10/sixteen-years-later-helicopter-returns-site-worlds-most-famous-photograph While the coexistence of these extremes is an endemic feature of many major Latin American cities, the proximity portrayed in this photograph is potentially misleading, as the rich minority make ever-greater efforts to segregate themselves away from the poor majority. Gated communities, protected by private security guards and often accessible only by car, have mushroomed. Sánchez-Ancochea describes a 10 km wall of concrete and wire that separates the luxury neighbourhood of Las Casuarinas in Lima from the poverty of the neighbouring Pamplona Alta. The low levels of investment in public education compound social distances, with around one in five Latin American children attending private school. And spatial inequalities exacerbate educational inequalities as poorer neighbourhoods receive poorer public services, not just in education and health but also in basic infrastructure from street lighting to water supplies. Qualitative research on Mexico reveals the effects of this segregation on people's perceptions of one another. Sánchez-Ancochea cites Bayón and Saraví’s (2017) findings on the fear and ignorance that residents of wealthy areas display towards poorer areas. Bleynat and Segal (2021) find social distance manifesting at both ends of the income distribution, as the rich express horror and disgust when they come into contact with the everyday lives of the poor majority, while both the poor and the middle classes describe the rich as living ‘in the clouds’ or ‘on another planet’, and as having ‘contemptuous’ attitudes. Social distance, or ‘othering’, is particularly acute when it interacts with racial differences. Latin American countries vary widely in their racial composition, but racism and discrimination are ubiquitous. Sánchez-Ancochea shows that people of indigenous and African descent are far more likely to be in poverty than those of European descent. Advertisers systematically use white models even where their primary audience is non-white because white people are perceived as having higher social status. Economic and social inequalities like those just described have been noted in Latin America for as long as social scientists have been investigating these issues, but as new research has extended our knowledge of inequality, it is becoming apparent that Latin America has not always been exceptionally unequal. The new economic history of inequality has shown that in the 19th century, Latin America was not more unequal than parts of north-western Europe (Williamson, 2010), although inequality probably grew during the period of globalization leading up to World War I. After World War II the region did not experience the substantial decline in inequality that many countries experienced, but most Latin American countries did achieve what Rodríguez Weber (2018) calls a ‘little levelling’.44 FitzGerald (2008) also demonstrates the variability of inequality over time in five Latin American countries, though his data are less precise than more recent country studies. There were significant reductions in inequality in Mexico from the late 1950s to the early 1970s, in Argentina and Brazil for at least two decades from the early 1940s, in Chile from the mid-1930s to the early 1970s, and in Uruguay from 1940 to the late 1960s. Colombia — to which we will turn in more detail below — is one of the few countries with historical data that did not experience a significant episode of declining inequality in the period. However, this was indeed only a ‘little levelling’ because the declines were neither very substantial nor sustained. In part, this was due to the debt crisis of the early 1980s, which led to the slashing of real minimum wages and social spending across the region. This was ostensibly to reduce current account deficits by cutting consumption and hence imports, but it was compounded by a rising tide of neoliberalism among policy makers and capitalist elites who had become disillusioned with state involvement in the economy. In several countries the rise in inequality began earlier. The 1973 military coup in Chile led by Augusto Pinochet led to a rise in the Gini coefficient of more than 7 percentage points within a decade. In the case of Argentina, the Gini coefficient jumped from a remarkably low 0.36 in 1974 to 0.43 to 1980, driven by the wage policy of the military dictatorship that took power in 1976 (Altimir et al., 2002). The Gini then continued to rise, reaching around 0.50 in the late 1990s. Inequality rose under a statist military government from 1976 to 1983, a centrist democratic government overburdened by debt and still threatened by military revanchists from 1983 to 1989, and a neoliberal democratic government from 1989 to 1999. Most strikingly, the rise in inequality was driven by a decline in real wages, not just relative to the incomes of the rich but also in absolute terms: by 2000 real wages in Argentina were a full one-third lower than their level in 1974. Mexico does not have comparable time series data for the Gini coefficient until 1984, but Bleynat et al. (2021) find a similar timeline using alternative data: both median wages and unskilled urban wages in Mexico collapsed from the end of the 1970s, staged a modest recovery in the 1990s, but in 2015 still remained below their level of the early 1970s, and a third below their 1978 peak. GDP per worker stagnated over the period but did not fall with wages, meaning income became more concentrated at the top of the distribution, and inequality rose. A great turnaround began around 2000. Since then, Gini coefficients have dropped substantially. This was driven primarily by rises in real minimum wages and government transfers, particularly non-contributory pensions and conditional cash transfers.55 See Lustig et al. (2016) for an analysis of proximate causes of this inequality decline. According to household surveys reported by the World Bank, Gini coefficients declined by an average of 9.3 percentage points from 2000 to 2019, ranging from a decline of 5 in Brazil to an extraordinary drop of 20 in Bolivia.66 The figure of 9.3 is the unweighted average of declines in Argentina (8.2), Bolivia (20.0), Brazil (5.0), Chile (8.4), Colombia (7.4), Ecuador (10.7), Mexico (7.2) and Peru (7.6), from 2000/2001 to the latest year available, which is 2019 for most countries (https://databank.worldbank.org/reports.aspx?source=world-development-indicators; accessed 8 December 2021). While inequality declined under both left- and right-wing governments, Feierherd et al. (2021) show that declines were significantly higher under left-wing governments, which raised taxes, minimum wages and non-contributory pensions more substantially. However, it should be noted that while the findings of improved living standards for the majority are uncontested, there is some doubt over whether the elite have really loosened their relative control of resources: administrative and tax data in some countries show that the income shares of the top 1 per cent have not declined in the way that household survey data suggest. For instance, while survey-based Gini coefficients have declined substantially since 2000 for both Brazil and Chile, the income share of the top 1 per cent estimated from administrative data increased.77 See Morgan (2017), de Rosa et al. (2020) and the World Inequality Database (https://wid.world). The above historical sketch demonstrates that inequality in Latin America has experienced substantial fluctuations. How can we account for this variability? The distribution of incomes depends on a wide range of factors including exogenous technological change, global price fluctuations, geography and demographics. But these do not by themselves explain the volatility we observe in inequality. Drawing on Offner's account of the historical development of institutions in Colombia and related literature, I will focus instead on the institutions and policies which, I suggest, are among the deep determinants of inequality. These institutions, in turn, are determined by conflicts and negotiations within civil society. In particular, I will interpret recent historical literature as suggesting that rapid and large changes in inequality are due to movements of reform and reaction. To understand the interplay between civil society groups and governmental institutions, let us go back to the mid-20th century era of state-led development. This was a period of rapid growth for the region, and it was also the period in which institutions of state were being constructed in order to manage and support this growth. For the case of Mexico, Bleynat (2021) shows how pressure from civil society in Mexico City, consisting of capitalists, labour unions and organizations of market vendors and other low-income groups, forced the official party to develop new institutions to manage conflicts both between and within sectors of society. While the government did not shy away from repression, it was also forced to build a developmental state and make major public investments that brought about significant increases in living standards in the 1960s. Expansion and modernization of political institutions was necessary to manage the expansion and modernization of the capitalist economy. This developmentalist social contract that underlay Mexico's ‘little levelling’ weakened as the economy encountered balance of payments problems in the late 1970s and was dismantled during the debt crisis from 1982. Not only were real minimum wages and social spending cut, but public investment was also slashed, in what Lustig (1998) describes as ‘overkill’ even for its macroeconomic goals. In a neoliberal outburst, much of the public sector was privatized, the financial sector was liberalized, and the economy was radically opened to international trade and investment. The following 40 years were marked by income stagnation for the majority of the population, and by 2015 the income share of the top 1 per cent had risen to nearly 30 per cent, one of the highest recorded in the world.88 Bleynat et al. (2021) provide data on median and urban unskilled incomes over the period. Data on top incomes (pre-tax national income, share of top 1 per cent) are available only from 2002 (https://wid.world/data/; accessed 8 December 2021). Unlike Mexico, Colombia did not participate in the ‘little levelling’ described above, and Offner's account of the institutions that comprised the mid-20th century mixed economy gives us important clues as to why not. Through a fascinating case study of the Cauca Valley Corporation (CVC) she shows that in a country with a small and weak state, capitalists took the lead in developing institutions to manage economic development. The CVC was established by a group of Colombian capitalists based in the valley as a tool for rationalizing agricultural production within their region, and was supported by US businessmen, the Rockefeller Foundation and the World Bank. The CVC's charter was issued by presidential decree in 1955, and it was given remarkable powers: in the words of Offner, the board ‘could sign international loan contracts, take land through eminent domain, adjudicate the use of water and public lands, charge rates for water and electricity, by property, and manage assets … the charter established the CVC as a public enterprise formally committed to capital accumulation in order finance its own expansion’ (p. 35). The creation of the CVC was in one sense a power grab by domestic capitalists. But it was also a response to the absence of state-led development efforts. It was ‘a way of building a developmental state within the constraints of a discredited political order, a usurping capitalist class, a government with no effective reach beyond the capital, and an anticommunist fear of central planning’ (p. 47). Like Bleynat (2021), Offner shows that the institutions of the developmental state were not a top-down effort but were driven by demands from organized civil society — where the specific elements of civil society vary across countries and over time. The CVC, then, was a pseudo state institution created and run by capitalists. Not surprisingly, therefore, it was also run for capitalists. Bernardo Garcés Córdoba, the CVC's executive director and a major local business owner, was convinced that his organization could drive the reforms required to improve the ‘general public interest’ within the region. His attitude calls to mind the old dictum: ‘for everything to stay the same, everything must change’. Just like the original speaker of those words, Prince Tancredi in Giuseppe Tomasi di Lampedusa's (1960) novel The Leopard, what he aimed to preserve were his region's highly unequal social structures. Offner describes his belief that ‘Valle's richest families needed to legitimate concentrated wealth as a credible basis for economic production or face redistribution that would destroy their economic power. Paying taxes for land reclamation and putting land to productive use were the only ways to keep their property’ (p. 74). This meant an unapologetic drive for productivity over equality. It was the CVC that implemented Colombia's land reform of 1961, aimed at increasing mechanization within the sector and improving efficiency in the allocation of land across agricultural activities. As the CVC succeeded in raising productivity through the consolidation of land holdings, it dispossessed large numbers of smallholders deemed too small to be efficient. Against substantial local opposition, it used the power of the state and the police to implement its plans. Neither Bleynat nor Offner focus directly on the question of inequality. But for those of us who are concerned with its evolution over the long run, their historical and institutional approach to political economy, rooted in civil society conflicts and negotiations, is an essential complement to economic analyses. As Rodríguez Weber (2017: 29) puts it in his study of inequality in Chile, the impact that economic factors such as growth and structural change have on inequality is mediated by historical circumstances and policies that, in turn, depend on ‘relations between social actors, which often take a conflictive and even violent form’. In modern societies the primordial form of social conflict is distributional conflict, implying a direct link with inequality.99 This approach is also consistent with the political economy literature that models social conflict as the driver of institutions that affect growth (Acemoglu et al., 2005). A key methodological difference is that the historical literature eschews formal models, which are necessarily based on a small number of quantifiable variables, in favour of qualitative narratives that can accommodate more complex causal mechanisms. Figure 1 shows how this operates. First consider the standard economic approach to analysing the distribution of income, or the proximate causes of inequality that underlie Sánchez-Ancochea's analysis. We start with market income — what households are paid in the market for their economic activities. This is the result of factor endowments, and factor returns. Factor endowments include the household's marketable skills (often measured by education or training), and any holdings of productive capital including land. Factor returns are the amount that those factors are paid, so we would consider the differential returns paid to workers with different levels of education, often called the ‘skill premium’, as well as land rents and other capital returns. Market income is then modified by fiscal redistribution in the form of taxes and cash transfers from the government, to produce disposable income, or the amount of money households have to spend or save. Finally, we can add the value of public services received by each household, especially the value of any free or subsidized public schooling or health services. This is known as final income and is our best approximation to what Dalton (1920), in one of the founding texts of modern inequality studies, referred to as ‘economic welfare’.1010 The steps from market income through disposable income to final income are described in detail in Lustig (2018). Determinants of the Distribution of Income Source: Author [Colour figure can be viewed at wileyonlinelibrary.com] Figure 1 also shows how the historical and institutional approach contributes to this standard analysis. First, it helps to explain the distribution of factor endowments. The distribution of education, for instance, is determined by the amount and type of public investment in schooling — also indicated in the figure by the arrow from public services to factor endowments. As Sánchez-Ancochea points out, Latin American countries have historically invested little in public schooling because elites have been reluctant to pay for it through taxation. The role of historical institutions and policies is even clearer in the matter of land distribution: in Latin America, many of the largest land holdings were the outcomes of wars of independence or domestic conflicts in the 19th century, and their evolution in the 20th century was the result of land reforms that may have reduced or increased land concentration, as Offner describes in the case of Colombia. Second, historical institutions influence factor returns. Most obviously, the systematic conflict between labour unions and capitalists in civil society affects the distribution between wages and profits. Unions and other civil society actors also have a role in determining the distribution of wages among workers, including the ‘skill premium’. Third, historical institutions influence fiscal redistribution through both taxes and transfers (affecting disposable income) and public services (affecting final income): political struggles within civil society produce policy outcomes regarding who gets taxed by how much, and who receives what benefits. The above historical accounts illuminate the social dynamics underlying the developmental state, and Figure 1 illustrates how these feed through to income inequality. The organized interest groups that form civil society both create and shape state institutions, and the configuration of these institutions determines policies and other outcomes that in turn influence the level of inequality. In Mexico we saw that workers and other low-income groups benefited from public investments in the 1960s, until a backlash against developmentalism led to dramatic increases in inequality from the late 1970s. In Argentina, workers’ unions and other civil society organizations were deeply embedded in a social democracy represented by Peronism, and were able to reduce inequality in the mid-20th century. Rodríguez Weber (2017) shows that the increasing strength of labour unions also underlay an important rise in salaries and decline in inequality in Chile in the 1960s. In Colombia, one of the few countries not to have participated in the period's ‘little levelling’, Offner shows how capitalists maintained substantial control over state institutions, allowing them to avoid even the temporary progressive redistribution achieved elsewhere in the region. The reaction of elites to these episodes of progressive redistribution has often been vicious. Sánchez-Ancochea notes that military coups against social democratic governments in Argentina, Brazil, Chile, Ecuador and Guatemala in the second half of the 20th century were all reactions by the elite against the threats to their privileges. This is not just a historical phenomenon. In the 21st century, the democratically elected presidents Zelaya in Honduras, Lugo in Paraguay, Rousseff in Brazil, and Morales in Bolivia were all progressive leaders who threatened elite privileges through redistribution of income and power, and were overthrown in manoeuvres that were legally questionable, and certainly anti-democratic, in several cases involving the military to remove them. To this list we can add Ecuador's Rafael Correa and Argentina's Cristina Fernandez de Kirchner as targets of reaction. Both oversaw dramatic declines in inequality during their presidential tenures and were subsequently brought up on corruption and other charges. While Fernandez de Kirchner has successfully remained in national politics and in 2022 is currently vice president, Correa is in exile in Belgium, after Interpol described attempts to extradite him as ‘obviously a political matter’ (Deutsche Welle, 2018). It is these endogenous social and political dynamics that I refer to as the internal constraint on progressive redistribution: domestic elites will fight against threats to their privilege. This has often been through illegal and unconstitutional means, as in the many military coups executed in the region. But it can also be pursued constitutionally through borderline or debatably legal means — a technique now known as ‘lawfare’.1111 See Martins et al. (2021) for an analysis of the case of Brazil. In many countries in the world, it can be pursued in an entirely legal manner through private ownership of the media. Sánchez-Ancochea (p. 71) cites media moguls in Brazil and Mexico declaring explicitly that they were more than willing to use their media power to intervene in politics. Given her lucid account of the interests behind Colombia's developmentalism, Offner's reference to what she calls its ‘intractable economic inequality’ (p. 7) is surprisingly off-key — and this is a common mistake in discussions of inequality. ‘Intractable’ suggests a technical, impersonal problem that is just too difficult to solve, or implies that those with the responsibility to reduce it attempted and failed. But in the case of Colombia, it is clear that inequality remained high because those with economic and political power had no interest in reducing it. In other Latin American countries, inequality has often failed to decline, or has risen, because of deliberate reaction by elites. This internal constraint is reinforced by what I will call the first external constraint on progressive redistribution: that Latin America's northern neighbour has consistently claimed the right to intervene politically and militarily in favour of its interests. Formalized in the 1904 Roosevelt Corollary to the Monroe Doctrine, the USA has long seen itself as the guarantor of international investments in its hemisphere — including, of course, investments by its own capitalists. The role of the US in military coups against numerous social democratic governments from Guatemala to Chile is well documented, and was driven both by US concerns about specific US investments (as in the case of United Fruit in Guatemala), and more generally by the notion that social democracy might represen