Inequality in Latin America and the Caribbean: a wide-ranging review

Scholarly research in the social sciences, generally for good reasons, is often built upon a myriad of incremental advances: relatively narrow contributions that seek to obtain clear answers to narrow and specific questions, typically written by small groups of authors. At the same time, opportunities to pause and take stock exist—typically in survey articles, handbook chapters or monographs—but these too tend to be written by one or two authors and inevitably represent their own views and perspectives.

Although there is much to be said for these well-established ways of producing knowledge in the social sciences, some broader questions—such as how and why a high-inequality equilibrium persists over centuries across many nations—would seem to require a collective effort involving many perspectives, disciplines and national experiences. This was why, inspired by the 2019 launch of Deaton Review of inequality in the United Kingdom, a group of us set up the Latin American and Caribbean Inequality Review (LACIR) in 2020.¹

Our objective, as stated at the outset, was to ‘go beyond a description of the region’s high inequality levels and seek an understanding of why Latin America’s inequality exceptionalism has persisted for probably the last 70 years, despite major structural economic and social change. […] Latin American inequality has persisted despite clear social pressures for its reduction and the shared conviction that excessive inequality is detrimental to economic progress. Does this mean that deep economic, social, or political factors are responsible for a long-run high-inequality equilibrium? If so, what are these factors and how is it possible to modify them to move economies and societies towards more equitable equilibria?’²

To guide the effort to address these questions, we convened a Panel of scholars consisting of Facundo Alvaredo (Paris School of Economics and London School of Economics), Orazio Attanasio (Yale University), Richard Blundell (University College London), François Bourguignon (Paris School of Economics), Ana De La O (Yale University), Marcela Eslava (Universidad de los Andes), Raquel Fernández (New York University), Francisco Ferreira (London School of Economics), Ana María Ibañez (Inter-American Development Bank), Santiago Levy (Brookings Institution), Nora Lustig (Tulane University), Julian Messina (Universidad de Alicante), Florencia Torche (Princeton University) and Andres Velasco (London School of Economics).³ The Panel was extremely ably assisted by Valentina Contreras (London School of Economics), who served as Secretary to the Review and Managing Editor of this Supplement.^4,5

During a series of initial meetings in 2020–2021—held online because of the COVID pandemic—this interdisciplinary group sought to organize the broad thematic space for the Review in a tractable set of topics, ultimately arriving at five themes, namely: (i) Levels and trends of inequality: establishing the facts; (ii) Inequality of Opportunity: nature and determinants; (iii) Inequality and markets: the roles of labor, capital and goods markets; (iv) Taxation and redistribution: are governments levelling the playing field? and (v): Inequality and political power: structures and struggles.

In 2021 the Panel commissioned twenty-seven papers—five or six for each theme—from a total of 74 authors from a wide range of disciplines and from many different institutions in Latin America and elsewhere. In keeping with the collective and reflective nature of the Review, these authors first discussed their initial thoughts and outlines in online thematic workshops. First drafts of these papers were presented and discussed at an in-person conference at the Inter-American Development Bank headquarters in Washington, DC in August 2022, and advanced drafts were submitted to the panel before a second conference, held at the Universidad de Los Andes Caribbean campus in Cartagena, Colombia, in 2023. Revised versions of those 27 papers were ultimately submitted for consideration for this Supplement of Oxford Open Economics, where they underwent a final round of anonymous refereeing.

In this Editorial, we briefly discuss what we learned about the five original themes, summarizing the main findings from each of the papers that were commissioned. These summaries represent our own reading of the papers and should not substitute for a careful reading of each paper in this Supplement. A final session offers a few closing remarks and outlines the second and final phase of the Review, currently underway in 2024/25.

Theme 1. Levels and trends of inequality: establishing the facts

The five papers in this theme seek to summarize the available evidence on inequality levels and dynamics in the region across five different domains: income, wealth, land ownership, educational attainment and achievement, and health outcomes. The emphasis in this block of papers is on measurement and on seeking to establish the basic facts, often—but not always—in the face of imperfect and sometimes contradictory data. The need for better and comparable data is a cross-cutting theme that has emerged in many of the contributions in this volume.

Alvaredo et al. (2025) review the evidence on income inequality in the region over the last 75 years or so. They try to make sense of a landscape of multiple data sources that suffer from serious comparability problems, with the aim of summarizing ‘what can be said with confidence about the levels and dynamics of income inequality in Latin America.’ They assemble a metadata set of more than 5600 Gini coefficient estimates (as well as an even greater number of quantile shares) for 34 countries over the 1948–2021 period. They find several different estimates for the same country/year combinations, reflecting differences in income concepts; in the treatment of the data for the same concepts and, most of all, in data sources. In particular, different combinations of data from household surveys, tax records and other administrative data sets, and imputations from national accounts lead to wide ranges (or ‘bands’) of inequality estimates for the same country years—often in the ten to twenty Gini points range. The authors argue that these ranges reflect genuine uncertainty about inequality levels because, while household surveys alone are almost certain to lead to underestimates, methods to combine tax data, surveys and national accounts lead to different outcomes among themselves, with no single dominant approach. Fortunately, there is much less disagreement about trends, with almost every country in the region displaying some version of an inverted-U pattern. Inequality generally rose in the 1970s and 1980s, reached a peak in the 1990s or 2000s (depending on the country) and fell in the 2000s until around 2015. While the timing and magnitude of the cycles varied markedly, the broad pattern is nonetheless common.

Turning to wealth, Carranza et al. (2025) are faced with much scarcer information. They scan a data constellation consisting of some household wealth surveys, a small number of datasets with wealth or estate tax records, some ‘rich lists’, and data on capital incomes from other surveys. They group their findings into three categories of decreasing confidence. In the highest-confidence group, they find that (i) aggregate wealth/income ratios have been rising steadily since 2000, perhaps from around 2.0 to 4.0 in market values; and (ii) for the three or four countries for which there is data on wealth distribution, concentration is high, with the top 1% of households owning around 40% of total wealth in Colombia, Chile and Uruguay.

In an intermediate confidence category, which they describe as ‘conjectures from indirect evidence’, they report evidence that capital incomes are also very unequally distributed, although the cross-country variation is high enough to be of concern. They also find that for the ‘bottom’ ninety percent of the population (or more) across all countries, any and all wealth is held as housing. In the lowest confidence category, which they describe as ‘speculation from partial evidence’, they see some suggestion that wealth inequality was largely stable in the 2000s, with possible increases at the very top after 2010.

Zooming in from total wealth to holdings of agricultural land, Gáfaro et al. (2025) use rich agricultural census data for nine countries in the region to document very high levels of land inequality in Latin America when compared to other developing regions: the Gini coefficient for land in LAC averages 0.84, as opposed to 0.73 in the Middle-East and North Africa, 0.56 in Asia and 0.51 in sub-Saharan Africa. Even though agriculture is no longer the dominant economic sector in any of the LAC countries considered, the authors also find evidence that agricultural—non-agricultural wage gaps, as well as wage gaps within agriculture, contribute a considerable share of overall income inequality in the household surveys: more than a third for the average country in their sample. This contribution seems to arise from very large differences in worker productivity across the region’s farms, leading to similarly substantial differences in land yields (which do not always vary monotonically with farm size) and consistently increasing output per worker in larger farms.

Although land is a historically important asset that clearly still matters for the overall picture of inequality in the region, the most important asset for the vast majority of Latin Americans is their human capital, which plays a critical role in determining their earnings in the labor market. In a wide-ranging review of patterns of educational attainment and achievement, Fernández et al. (2025) document a number of common trends across the region. They find that attainment, measured by years of schooling completed, has been rising substantially across cohorts in every country for which they have data, with averages for 25–34-year-olds circa 2019 being sometimes double those for 65–74 year-olds. These attainments are distributed neither uniformly nor independently across populations, with parental background being a powerful predictor of individual years of schooling. In Brazil, for example, 85% of people with at least one parent that attained tertiary education complete it themselves, as compared to 15% of those neither of whose parents completed secondary education.

There are also large spatial gaps, generally in favour of urban areas, and gender differences that are largely in favour of women. In terms of achievement in test scores – which are a thought to be a better measure of actual learning – the authors find that family background is also a strong predictor of better outcomes, but no more so than in comparator countries outside the region, such as Korea, Turkey, or the United States. The largest difference vis-à-vis those countries are in (PISA) test score levels, with the region reaching much lower average scores. These averages also hide substantial differences in performance across school types, with private school students doing much better than their counterparts in public schools—a theme to which we return below, when the region’s social contract is discussed. Highlighting the importance of education for an understanding of inequality in Latin America, the region is found to have considerably higher returns to both secondary and tertiary education compared to a broad set of comparator countries.

Just like education, health is also both a key component of human capital and an important dimension of wellbeing in its own right. Again as education, health is itself multidimensional, with multiple different outcomes contributing to overall health status. In a comprehensive assessment, Bancalari et al. (2025a) investigate health gradients for several indicators of maternal and child health; reproductive health; non-communicable diseases; and mental health, with respect to education; gender; urban–rural status; and a measure of household ‘wealth’. Their description is informed and framed by the fact that the region is in the midst of an incomplete—but sometimes advanced—epidemiological transition, where the main causes of morbidity are shifting towards chronic, non-communicable diseases. Although the authors uncover complex, and sometimes surprising patterns, the overarching picture is one of interconnected, overlapping inequalities, with better educated and richer individuals having higher rates of access to and use of health services (from antenatal visits to access to contraceptives), as well as better outcomes (from lower stunting to less prevalent teenage pregnancy). Patterns are more varied across countries for certain non-communicable diseases such as hypertension, depending on the country’s position along the epidemiological transition.

Theme 2. Inequality of opportunity: nature and determinants

It is well established that, in the process of human development and skill formation, what happens in the first few years of life plays a very important role in determining life cycle trajectories and outcomes, including many of those described in the Theme 1 papers above. This consensus has important implications for the distinction between inequality of outcomes, which some commentators perceive as not necessarily unfair or undesirable, and inequality of opportunities, which, on the contrary, is widely considered unjust and to be avoided. When characterizing inequality and its evolution it is therefore important to establish when inequality among individuals appears in the life cycle and what are the determinants of its appearance and evolution.

Inequality of opportunities is also linked to social mobility: a society where parents at the bottom of the income or wealth distribution find it difficult to provide their children with access to a stimulating environment or a quality education is likely to be characterized by low social mobility with inequality and poverty being transmitted across generations. Finally, if inequality in human development arises early in life, it is also important to establish the causes of the observed differences and whether early disadvantage can (and should) be addressed by policy interventions. Overall, it is clear that a good understanding of inequality needs the characterization of inequality over the life course, throughout which different factors, the family, the education system, neighborhoods, the marriage and labor markets, play an important role.

In a region as unequal as Latin America, these questions are extremely salient. For this reason, several papers in this Supplement address issues related to the general theme of the inequality of opportunity and the characterization of the dynamics of inequality over the life cycle.

The first paper in this Theme, by Brunori et al. (2025) looks at inequality of opportunity and social mobility in different outcomes, such as income and education. Inequality of opportunity is measured in terms of the predictive power of a set of pre-determined or inherited circumstances on those outcomes. The paper provides much new evidence and distinguishes ex-ante and ex post inequality of opportunities, with the former referring to the expected value of a given outcome for a given type and the latter the entire distribution of the outcome. The approach proposed lets the data speak about what is the best definition of ‘circumstances’ with respect to a given outcome. The paper finds that LAC intergenerational mobility (mainly in education) is low by international standards, although recent trends have shown improvements. Having said that, the paper also documents considerable heterogeneity across countries. However, in general, inequality of opportunities is clearly high in Latin America.

The heterogeneity observed in social mobility across different countries in LAC, is also present in another important dimension: inequality in early years development, which is studied in a second paper on this theme. The paper by Attanasio et al. (2025) first documents that inequality starts early in the life cycle and, to a large extent, these inequalities are driven by differences in parenting practices and other environmental factors that affect child development. The paper, which looks at five countries for which data on early child development is available, documents that countries in the region with relatively high inequality in income and wealth are largely also countries that exhibit large levels of inequality in early childhood development. The paper also stresses the extreme paucity of high-quality data on child development in the region and the importance of a longitudinal analysis that, following individuals over the life course, can provide much better evidence.

The role that parenting practices have in determining young children’s outcomes is only one of the reasons for looking at the dynamics of the nature of families in the region, which has changed considerably in recent decades. The paper by Esteve et al. (2025) describes these trends and documents an important relationship between family forms and both children schooling and female labor force participation. As in other contexts, the authors show that growing up in stable families is conducive to better adult outcomes for children. Surprisingly, they find that children of single mothers perform better than children of cohabitating (and maybe unstable) families. These patterns are evident in a context of significant social change, including a considerable decline in fertility, substantial increases in cohabitation, and rising rates of union dissolutions and single motherhood. At the same time, these changes in family structure have social-class specific patterns in most countries.

Another important dimension of inequality in the region is gender inequality. The paper by Berniell et al. (2025) offers a comprehensive view of gender gaps in different dimensions, including education and labor markets. This paper also documents a considerable amount of heterogeneity across countries, which are divided into ‘lower-middle’, ‘upper-middle’ and ‘high’ income ones. The main results are that gender gaps are still important in many dimensions, although some progress has been made in others. Looking at how these gaps evolve over the life cycle is important. The paper contains many original findings on attitudes and aspirations.

The paper by Telles et al. (2025) looks at racial and ethnic inequality in Latin America, using a new and more consistent measure of race and ethnicity. The paper starts by exploring the region’s ethno-racial demography with a particular emphasis on census categories, national survey categories, and skin color. The paper finds that black and indigenous populations and those with darker skin color experience educational, income and occupational disadvantages, even after controlling for their social origins. The paper has a special emphasis on Brazil but also documents substantial heterogeneity across countries.

A final paper on the theme of inequality of opportunities, by Gomez-Lobo and Oviedo (2025), considers spatial inequalities. The paper tries to map aggregate inequalities to micro level disparities. Considerable attention is paid to inequality within and across urban and rural areas, and within the former, between small and large urban areas. The paper finds that LAC is very different from developed countries. Inequality between urban and rural is important, but inequality within urban areas also plays a big role.

Overall, the papers in this theme suggest that inequality of opportunities in LAC is large, which leads to low social mobility. The dimensions of inequality of opportunities are many and of different importance. Indeed, the paper on inequality of opportunities uses a methodology that identifies the important sets of ‘circumstances’. There is much heterogeneity in the region for many dimensions of inequality: social mobility, early childhood development and family dynamics, gender and spatial inequality. Progress can be observed in some dimensions but not in others.

Theme 3. Inequality and markets: the roles of labor, capital and goods markets

Genetics, families and communities help shape much in the early life of individuals – from their personal characteristics to their initial health and wealth, to the quality and extent of the education and skills they acquire. In many cases, those attributes matter in and of themselves, but they are also the endowments people bring to markets—of labor, capital and goods and services—which help determine their market incomes.

The nature, structure and imperfections in those markets play a critical role in shaping economic inequality everywhere. Imperfections are particularly important in regions like Latin America, where non-competitive markets are the norm rather than the exception. Instead of efficiently allocating resources and rewarding workers proportionally to their contributions, imperfect markets exacerbate disparities by offering unequal wages to equally productive individuals and diminishing labor’s share of overall revenue. This dynamic is further intensified by concentrated firm ownership. Larger firms wield greater market power, and since these firms are disproportionately owned by society’s higher echelons, product and labor market imperfections further skew wealth distribution.

The paper by Eslava et al. (2025b) shows that market structure in Latin America is marked by excessive fragmentation, commonly referred to as the ‘missing middle.’ Around 60% of workers are either self-employed or employed by micro-enterprises with fewer than five employees. These micro firms dominate the economic landscape but operate with very low productivity and limited capacity to generate high-quality, well-paying jobs. In contrast, larger firms, which are more productive and have greater access to resources, employ only about 15% of the workforce. The remaining workers are employed by small firms that rarely scale up and grow. This highly fragmented structure is very different to that observed in developed countries, where a healthy ‘middle class’ of firms is observed, but also contrasts with countries from other middle-income regions in its bias towards own account work and micro firms.

The work by Eslava and coauthors further shows that the fragmented employment structure exacerbates inequality through both direct and indirect effects. Directly, the over-representation of workers in self-owned enterprises and micro firms means most low-wage workers are trapped in low-productivity environments, earning wages that reflect the limited output of these firms. Indirectly, the abundance of micro firms fails to challenge the dominance of large incumbents.

It is thus no surprise that one of the main findings in the paper by Eslava et al. (2025a) is the widespread market power among larger firms. In the region, large formal firms can charge prices significantly above their marginal costs. At the same time, they pay workers’ wages that are substantially below their marginal revenue product—their contribution to production. Due to firms’ market power in both product and labor markets, the share of income allocated to labor is reduced. This reduction in the labor share further exacerbates inequality. But perhaps surprisingly, the degree of market power in Latin America does not stand out compared to that observed in a selection of middle-income countries from other regions. While market power in product and labor markets contributes to overall inequality, it does not appear to be at the core of Latin America’s inequality exceptionalism.

The fragmentation of production is compounded by significant barriers to worker mobility, which further amplify the market power of large incumbent firms. When workers cannot easily transition between jobs or relocate to other regions, firms face less competition for labor, allowing them to suppress wages below workers’ marginal productivity. The limited mobility of workers in the region is well illustrated by the adjustments that follow trade liberalization events, as documented by Dix-Carneiro and Kovak (2025).

Dix-Carneiro and Kovak (2025) provide a comprehensive review of the role of globalisation in shaping inequality dynamics in the region. A key finding is that it is difficult to establish general conclusions regarding the impact of trade liberalisation on inequality. Research outcomes depend heavily on the dimension of inequality analyzed—whether across workers, sectors, or regions—as well as the data and methods used. However, a common thread across recent trade literature has emphasised that adjustments to trade shocks are sluggish, as illustrated by the case study of Brazil discussed in the paper. Even 20 years after a negative trade shock, the most affected regions have yet to return to their initial levels of formal employment and wages.

Gindling and Ronconi (2025) discuss the limitations of traditional remedies to excessive labor market power in the region. While the minimum wage has proven effective in reducing wage disparities and countering market power in many developed countries, its efficacy in the region is severely undermined by widespread non-compliance. Although minimum wage levels in Latin America are relatively high compared to median wages, weak and inconsistent enforcement mechanisms prevent many workers from benefiting from these protections. Moreover, the pervasive nature of non-compliance undermines workers’ bargaining power, as the implicit threat of unenforced minimum wage laws severely limits the potential leverage the policy might provide.

The paper by Narita and Menezes-Filho (2025) addresses a fundamental feature of Latin American labor markets: their exceptionally high rates of job turnover. In countries like Argentina, Brazil, Ecuador, Mexico, Paraguay and Peru, 24–44% of workers change jobs annually, and a significant portion (30–50%) exits the labor force entirely. This trend is largely driven by the unregulated informal sector, which employs over 40% of the workforce and is characterised by shorter job durations and a lack of benefits.

The relationship between high turnover rates and inequality is complex. Job transitions initially reduce inequality by benefiting low-wage earners and younger workers through immediate wage gains, especially when moving from informality to formality. However, over time, the wage stability and human capital accumulation associated with staying in a job outweigh the short-term benefits of switching. Because frequent transitions are more common among unskilled and informal workers, they ultimately contribute to rising wage inequality in the long run. The persistence of informality exacerbates this dynamic, limiting access to more productive, higher-quality jobs.

Theme 4. Taxation and redistribution: governments levelling the playing field?

In Latin America, as elsewhere, the incomes determined in markets are not necessarily the final incomes available to households to finance the consumption and saving choices that ultimately determine well-being. The State can and does intervene, through taxes, subsidies, transfers and regulations which, taken together, can cause the income distribution to differ substantially from that which would be determined by markets alone.

Indeed, redistributive policies in Latin America and the Caribbean are generally equalizing; however, their ability to substantially reduce inequality remains constrained (Lustig et al., 2025; Pessino et al., 2025). This limitation stems, in part, from poorly designed fiscal systems, compounded by weak state capacity, which results in low tax revenues, a slightly equalizing tax structure and insufficient government expenditure on social programs (Lustig et al., 2025; Pessino et al., 2025). Moreover, the design of fiscal systems and social protection programs often disregards the region’s pervasive informality, which diminishes the effectiveness of both. High informality erodes tax revenues, creates incentives to remain informal and undermines the progressive intent of social protection programs, particularly pension systems in some countries (Altamirano et al., 2025).

Despite these limitations, fiscal systems in the region do play a role in reducing income inequality. Using fiscal incidence analysis across 18 countries, Lustig et al. (2025) find that fiscal systems, when taking into account both taxes and spending, are equalizing in all cases. The average reduction in the Gini coefficient, comparing pre- and post-fiscal income, ranges from 9.3 to 10.1 points, depending on whether pensions are treated as deferred income or government transfers. However, the redistributive capacity of fiscal systems has remained virtually stagnant over time and, in half of the countries, fiscal policies are poverty-increasing, even while reducing overall inequality. While direct taxes and transfers contribute to the equalizing effect, the majority of the effect stems from spending on education and health, which is largely pro-poor. The fiscal capacity to reduce inequality varies widely across the 18 countries.

The composition of the tax structure further contributes to the limited equalizing effect. According to a statutory incidence analysis for ten countries in LAC by Pessino et al. (2025), overall redistribution through taxes is equalizing, yet small. Indirect taxes—primarily the regressive Value-Added Tax (VAT)—constitute nearly half of total tax revenue and are unequalising. In contrast, personal income taxes are equalizing, but generally characterized by low rates, narrow bases—largely due to high non-taxable income thresholds—and significant pro-rich tax expenditures. As a result, the tax burden is higher for the lowest income deciles (17% for the 10% lowest income decile) compared to the highest income ones (10% for the 10% the highest income decile).

Pessino et al. (2025) stress repeatedly the importance to consider not only the redistributive role that governments and states play via taxation, but also the expenditure side. The importance of considering how governments in the region spend the resources they collect (and those that they do not), which is already discussed in Pessino et al. (2025) leads naturally to the next three papers, that consider cash transfers and subsidies, pension systems, and health systems.

Cash transfers, which expanded rapidly across the region during the 1990s, have made only modest contributions to reducing inequality. Stampini et al. (2025) analyze 167 programs—including conditional cash transfers, non-contributory pensions and other initiatives—implemented in 17 countries, collectively covering 166 million individuals (~30% of the population). While these programs have had positive impacts on direct beneficiaries, particularly through poverty reduction and other socioeconomic improvements, their effect on inequality remains limited.

Two key factors explain this limited impact. First, the value of the transfers is relatively small. On average, transfers represent only 32% of the aggregate poverty gap, with significant variation across countries, ranging from as low as 4% to as high as 88%. Second, poor targeting reduces program efficiency, with only 55% of the poor receiving transfers, leading to substantial under-coverage in many countries.

Old-age pensions, despite being designed with progressive intentions, often end up contributing to inequality in practice. Defined-benefit pension systems, which are prevalent in the region, are structured to favor low-income workers by offering higher replacement rates and capping benefits for high-income earners. However, high levels of labor informality, with low-income workers frequently transitioning in and out of formal employment, result in limited pension coverage for these workers. Consequently, low-wage workers effectively pay taxes into the system, while high-wage earners benefit from substantial subsidies. Altamirano et al. (2025) examine pension systems across 25 countries in LAC, finding that in defined benefit systems, the top labor income deciles receive between 70% and 95% of all subsidies. This subsidy is comparatively smaller in defined contribution systems. Non-contributory pensions, when well-targeted, have contributed to mitigating the unequal distribution of benefits within pension systems.

The provision of health services and the activities connected to that are analyzed in detail by Bancalari et al. (2025b). In this chapter, the authors acknowledge that in recent decades many LAC countries have tried to improve access to health services, which might have improved health outcomes and maybe even somewhat reduced health inequalities. The authors document a number of important issues that are relevant in the region. To do so, they provide a useful taxonomy of different health systems which are funded in different ways. They then show that different systems are prevalent in different countries. Importantly, in many contexts, different systems co-exist, generating effectively access to very unequal quality of care. Furthermore, the presence of these overlapping systems (as is in the case for pensions) may distort substantially labor market decisions and create unintended incentives to informality.

The high levels of inequality documented throughout this Review might imply that LAC should reform fiscal systems and social protection programs to strengthen the redistributive capacity of the State. However, these are contentious issues that require broad public support. To better understand public willingness to endorse redistributive policies, Busso et al. (2025) conducted online surveys and a survey experiment across eight LAC countries. The findings reveal that while 80% of respondents believe income distribution in their countries is unjust and 77% support redistribution from the wealthy to the middle class, few are willing to pay higher taxes to contribute to these efforts. Instead, there is stronger support for increasing corporate income taxes. Support for expanding social programs, furthermore, remains lukewarm. Several factors influence these preferences, but perceptions of corruption and widespread mistrust of the state and elites significantly undermine public willingness to engage in a stronger social contract. This mistrust erodes support for higher taxes overall and greater spending on social programs.

Theme 5. Inequality and political power: structures and struggles

Understanding how a high-inequality equilibrium persists over the long run, as we set out to do, inevitably requires examining political systems, and why they fail, or chose not, to use State power to challenge economic inequalities at each stage of the process—from spending on early-childhood education and health systems, to labor market policies and anti-monopoly regulations, to the design of taxes, transfers and pension programmes.

Theme 5 of the Review considers this question in a rapidly evolving context. Since the 1980s, democratic political regimes have replaced military dictatorships and other forms of authoritarianism in most countries in Latin America. As elections became the norm, more contestation of political power followed the expansion of political inclusion. Not only have established left-wing political parties governed in many countries, but new political parties, like the Movement Towards Socialism in Bolivia and the Movement of National Regeneration in Mexico, and non-elite politicians have won national and local elections, replacing the old political guard. However, these political transformations have yet to upend the region’s highly unequal wealth, income and opportunity distribution. Even left political parties have engaged in ‘easy’ forms of redistribution, layering policies over existing policies rather than building new universal programs (Holland and Schneider, 2017). What political features maintain the levels of inequality in Latin America?

In a group of countries in the region, inequality has deep roots that stretch back to colonial rule. In this Supplement, Eslava and Valencia Caicedo (2025) review the academic literature examining the historical origins of inequality, with particular attention to the role of education, land distribution, and coercive labor institutions. Then, they replicate and expand canonical studies of the relationship between political institutions and development to show that where political and economic elites adopted coercive labor institutions, such as slavery, inequality was higher in the 19^th and 20^th centuries. The mechanism connecting slavery with modern inequality is one of low investment in human capital. Political elites underinvested in public education and public services. Hence, the descendants of slaves were often denied access to education, which limited their social mobility. Extreme levels of social exclusion in the past have also left an imprint on the political culture of those affected and of their descendants. For example, social exclusion in a leper colony in Colombia made individuals more altruistic within their community and more distrustful of the state (Ramos-Toro, 2023). While the legacies of slavery and a deficient education system correlate positively with inequality, land distribution and land reform have a mixed effect on inequality depending on political regimes and underlying interests (Albertus, 2015; Albertus et al., 2020).

The high levels of inequality in the region have distorted the operation of modern political systems. Lupu (2025) argues that one such distortion is the presence of weak political parties, which lack institutionalization, limiting the governing parties’ ability to achieve broad-based redistribution. Strong political parties are necessary to build stable, long-term coalitions that can push for progressive policies. In contrast, weak parties often lack coherent platforms, struggle to build coalitions and rely on clientelism, which undermines meaningful redistributive policies. Consistent with this argument, the author finds a negative correlation between party system institutionalization and inequality in a sample of countries in Latin America.

Political parties should aggregate and represent citizens’ preferences. Yet, a substantial gap exists between citizens’ and legislators’ preferences in most countries in the region. Lupu (2025) shows that the average legislator expressed (in surveys) an ideological orientation to the right compared to the wealthiest quintile of citizens sampled in opinion polls. The rightward bias is extreme in Chile and Paraguay. The only exception is Bolivia. The mismatch between citizens’ and lawmakers’ preferences creates a representation deficit, contributing to policies that disproportionately benefit affluent voters. Legislative malapportionment further exacerbates this issue, enabling the wealthy to block progressive reforms (Snyder and Samuels, 2004).

Fergusson et al. (2025) discuss another manifestation of the representation deficit in the region. The election of new politicians in local elections in Colombia has yet to produce any substantial change in policy outcomes. They argue that this null result is compatible with elites influencing and capturing new politicians and citizens in contexts of high inequality. They also suggest that partially new elites internalize the existing values, hierarchies and status systems, from which they benefit once they accrue political and economic influence. Hence, when inequality is high, democracy risks being a mechanism for elite circulation, not political inclusion.

Although elite domination accounts for the high and sustained inequality in many countries in the region, it does not explain all cases. Guizzo Altube et al. (2025) document that, in some countries, governments collect taxes and engage in non-negligible redistribution. Yet, they argue, redistributive efforts have been frequently combined with macroeconomic imbalances, high inflation, low growth and low-quality public services. Compared to OECD countries, most Latin American countries score lower in government efficiency and government revenues as a share of GDP. Moreover, redistribution (measured as the percent reduction in the Gini index from market to disposable income) does not increase as government revenues rise in Latin America, whereas in OECD countries, it does. What accounts for these outcomes? The authors present a comprehensive review of the academic literature and suggest that two dimensions of politics are essential: the distribution of power in a country and whether political parties are agents of their constituents. Whereas a balance of power and strong political parties create a cooperative equilibrium (like in the OECD), the domination of either the wealthy or the Left sustains a non-cooperative equilibrium in which a trade-off between investment and redistribution becomes sharp. Moreover, when there is no balance in power, politicians have incentives to employ inefficient redistribution as a political tool to maintain control.

The deficit in representation and meager responsiveness of political systems regarding growth, redistribution and provision of high-quality services have shaped voters’ political attitudes and behaviour. As mentioned before, voters in the region express little support for raising taxes and pro-poor social programs (Busso et al., 2025). At the same time, disenchantment with the social contract is widespread, and is reflected in people’s decisions to opt out of public services when they can afford to do so. In this issue, De La O et al. (2025) document that wealthy households and a substantial proportion of the middle class are out of the public education system. Opting out of public health is only prevalent among affluent households. For both policy domains, however, they find that people who use private services have worse evaluations of public services, express less support for the public provision of those services, and are less supportive of redistribution than people inside the public systems. This creates an inequality trap, as citizens with a stronger political voice might use their political power to lobby for lower taxes and less spending for the public sector. If so, the quality of the public service may fall even more, reducing social mobility opportunities (Bénabou, 2000; Ferreira et al., 2013; López-Calva et al., 2014). Finally, their evidence is consistent with the argument that opting out of services increases the social distance between income groups, which might explain why users of private services prefer governments that do less to redistribute opportunities and income.

Closing remarks

As the preceding sessions attest, trying to understand the nature of Latin America’s high-inequality equilibrium and the reasons why it has proved so persistent has led us to explore a wide range of topics, from early childhood development to wage markdowns, to the design of old-age pensions and of political parties. All of these topics seemed to the Panel to be of first-order importance for a comprehensive understanding of the broader problem. But their thematic breadth, and the wealth of detail in each of the papers, seem to necessitate a synthesis, or at least an effort in that direction. That effort, jointly undertaken by the Panel, is the subject of the last phase of the LACIR, currently ongoing. Whereas this Supplement of Oxford Open Economics, containing the full set of 27 papers, is the first major output of the review, a short, non-technical volume outlining the synthesis and providing a narrative around the themes covered by the papers here, which we hope to publish in 2026, will be the second.

The volume will focus on three key points—one on the nature of Latin American inequality, and two on its reproduction mechanisms. The first point is that Latin American inequality is generally high, multifaceted and interconnected and characterized by important horizontal cleavages. The high levels of inequality were apparent in terms wealth and land inequality, as well as very high education premia, inequalities in access to health services and inequalities of opportunities. One important factor that emerges is that the region is not uniform. Some countries are much more unequal than others, and these different yet similar stories can teach important lessons on the origins and dynamics of inequalities. However, in income terms, despite all of the uncertainty about exact levels and of comparability challenges across regions, it is quite clear that, at least at the lower bound represented by estimates based on household surveys, Latin America is one of the world’s two most unequal regions—if not the single most unequal.⁶

Inequality is multifaceted, of course, because of its many dimensions—health, wealth, income, land, education, opportunities, and more. It is interconnected because these different outcomes tend to be positively correlated, with groups that have greater wealth and income also displaying better health and higher education, which they can systematically pass down to their children. These associations are not universal, but the exceptions are just that. And it is marked by horizontal cleavages because the inequalities in the size distributions are not blind to colour, race, ethnicity, gender, or geography. Group-based inequalities are pronounced, particularly among the races and ethnicities that comprise the continent’s populations, but also between its rural and urban spaces, and between men and women.⁷

Both of the other two key points, relating to drivers of the persistence of inequality, refer to inequality cycles. The short cycle, so to speak, takes place on a human timescale, of lifecycles and bequests. It arises from the combination of intertemporal complementarities with the opportunity gradients described earlier: the children of richer and better-educated parents develop key capabilities and skills in early childhood that later stand them in good stead to benefit from the better, often private, schools to which they are sent. They are likelier to attend university, which yields huge economic returns, in large part by matching workers to ‘better’, larger, formal-sector firms. Those advantages generally extend to old age pensions and, earlier, at a parenting age, enable these individuals to transmit more resources, both human and financial, as well as better opportunities to their own children—so the cycle restarts.

There is evidence that this short cycle is at work, in some form, in virtually every society where data is available. But the strength of the cycle, measured—for example—in terms of intergenerational persistence or inequality of opportunity, differs widely and is particularly pronounced in Latin America. In those regions, such as Europe or parts of Asia, where it is less pronounced, this typically reflects the effects of equalizing public action. So, an understanding of why the short cycle of inequality reproduction is allowed to function with so little impediment in Latin America requires understanding why our States do not systematic pursue those equalizing policies or put in place those equalizing institutions.

That is where a long cycle of interacting economic and political inequalities plays a role. This long cycle operates on a historical or civilizational timescale. It accounts for why the distant heirs of the large landholders and mine owners in colonial times are still disproportionately well-represented among the region’s economic and political elites, while the descendants of slaves and indentured indigenous workers centuries ago are still disproportionately represented among today’s poor and dispossessed. But while these disproportionalities are facts, the mechanisms from which they result are complex, and not always clear. They operate despite the advance of democracy and of a sizeable middle class. Despite the rise of new elites and the formation of political parties and popular movements that seek to represent the less privileged, as discussed in various papers under Theme 5. They involve elements of a social contract that evolves with the times, whilst at the same time preserving some of their fundamental exclusionary features. These mechanisms, it is fair to say, are less well understood than those underpinning the short cycle, and we look forward to returning to them in Phase 2 of this Review.

Funding

LACIR was funded primarily by a grant from the Inter-American Development Bank, as well as by funds from the International Inequalities Institute (III) at the London School of Economics and the Economic Growth Center at Yale University.

Footnotes

1

The Deaton Review, led by Nobel laureate Sir Angus Deaton, was funded by the Nuffield Foundation and run out of the Institute for Fiscal Studies in London. Its main findings were published in another Supplement of this journal (IFS Deaton Review, 2024). Two members of the Deaton Review Panel, Orazio Attanasio and Richard Blundell, are also members of our Panel.

2

Alvaredo et al. (2021), p.2.

3

We were also initially joined by Sonya Krutikova (Institute of Fiscal Studies), who contributed a great deal to the initial phase of the Review.

4

Valentina Martínez-Pabón and Valentina Papu also kindly provided very valuable support and assistance at different points of the process.

5

LACIR was funded primarily by a generous grant from the Inter-American Development Bank, as well as by funds from the International Inequalities Institute (III) at the London School of Economics and the Economic Growth Center at Yale University. Its secretariat was based at the III/LSE.

6

The uncertainty about which is the single most unequal arises because, whereas inequality in typically measured for incomes in Latin America, it is measured for consumption expenditures in Africa. Estimates are therefore hard to compare, and attempts at conversion, albeit imperfect, suggest a tight race for that inglorious title.

7

Although the Review was unable to look at inequalities by sexual orientation, recent work suggests that there are important inequalities along that dimension too, though self-reporting and data issues remain a challenge (Muñoz et al., 2024).