Income inequality in Brazil since 1980

Question

How has income inequality in Brazil evolved since 1980, and what does peer-reviewed research attribute the changes to?

Hypothesis

Income policies, especially the rising minimum wage and targeted transfers, explain most of the fall in income inequality in Brazil from 2001 on.

Period

1976–2025

Method

Descriptive Description of inequality and poverty series, kept separate by survey (old PNAD and Continuous PNAD), compared with neighboring countries, and a review of studies that decompose the fall in inequality. There is no original causal estimate: shares of the fall attributed to each factor come from the cited studies and depend on their methods.

Claims causality

No: descriptive analysis

This visualization is descriptive and does not show that a government caused the change. See the methodology.

Result

Inequality measured by household surveys rose in the 1980s, was stable in the 1990s and fell steadily from 2001 to 2014: the old PNAD Gini went from 0.596 to 0.518. Studies attribute a relevant part of the fall to the minimum wage and transfers, but also to labor market changes, and tax data suggest top-income concentration changed little. The available evidence does not allow a confident claim that income policies explain most of the fall.

What the data show

The 1980s: high and rising inequality

According to the old PNAD survey, the Gini of household income per capita was 0.584 in 1981 and reached 0.636 in 1989, the highest value in the series. It was a decade of stagnant income and rising inflation: GDP per capita changed by −5.0% between 1980 and 1990, and consumer price inflation reached 1,430.7% in 1989.

The 1990s: less poverty, same inequality

Between 1993 and 1995, the period of the Real Plan, the poverty rate measured with Ipea's regional lines went from 43.0% to 35.1%. The Gini barely moved: 0.604 in 1993 and 0.601 in 1995. Barros, Henriques and Mendonça (2000) described this, in the authors' own normative words, as an "unacceptable stability" of inequality.

2001 to 2014: a sustained fall

The old PNAD Gini changed by −0.078 between 2001 and 2014, from 0.596 to 0.518. Over the same period, the poverty rate with regional lines went from 35.8% in 2003 to 13.3% in 2014.

Since 2012: the Continuous PNAD

IBGE's current official series starts at 0.540 in 2012. After the 2015–2016 recession, when GDP per capita changed by −8.1% in two years, the index rose to 0.545 in 2018. In 2020 it went to 0.523 and back to 0.543 in 2021; in those two years income comes from the survey's 5th visit rather than the 1st, because of the pandemic. In 2024 the value was 0.504. The chart shows each year's 95% interval: small differences between adjacent years may not be statistically significant.

Brazil and its neighbors

The fall of the 2000s was not unique to Brazil. Between 2002 and 2014, on the World Bank Gini (0 to 100 scale), Brazil changed by −6.0, Mexico by −1.7 and Colombia by −2.4; urban Argentina changed by −9.2 between 2003 and 2014. When several countries move together, factors common to the region probably matter, alongside national policies.

Inequality beyond income

In 2023 Brazil's HDI was 0.786 and its inequality-adjusted HDI 0.594. The gap is the human development lost to inequality in health, education and income.

What the research attributes

Peer-reviewed studies point to several factors, and none explains the fall alone. The rising minimum wage, targeted transfers such as Bolsa Família and labor market changes, such as lower returns to experience, carry different weights depending on the period, the income measure and the method. Evidence for and against is listed below, each item with its source paper.

Evidence for the hypothesis

• According to a structural model of the labor market, the 128% real increase in the minimum wage between 1996 and 2018 accounts for 45% of the fall in labor earnings inequality, with muted effects on employment and output. The share depends on the model's assumptions and refers to labor earnings, not total household income. [doi:10.1257/aer.20181506]
• Conditional cash transfers were less than 1% of total income but, being well targeted, accounted for about 21% of the fall in the Gini in Brazil and Mexico between the mid-1990s and the mid-2000s, in an accounting (not causal) decomposition. [doi:10.24201/ee.v0i0.387]
• A comparative review concludes that Bolsa Família helped reduce inequality and extreme poverty, with no negative impact on labor force participation. [doi:10.1017/S0023879100009390]

Contrary or disputed evidence

• In a decomposition of labor earnings between 1995 and 2012, the main driver of the fall in the Gini was lower returns to labor market experience, accounting for 53% of the fall, more than the falling education premium. This factor is not an income policy. [doi:10.1093/wber/lhab005]
• With income tax data, top-income concentration is substantially higher than in household surveys and remained broadly stable between 2006 and 2012. The fall measured by surveys may overstate the reduction in overall inequality, and with it the role attributed to any policy. [doi:10.1590/00115258201537]

Limitations

• The old PNAD and Continuous PNAD series are not level-comparable and were not spliced. Comparisons across periods should be made within each series.
• Household surveys underestimate top incomes; the level and possibly the trend of overall inequality differ with tax data.
• Regional comparisons use the World Bank Gini, which for Argentina covers only the urban population, and whose underlying surveys change over time in several countries.
• Shares of the fall attributed to the minimum wage, transfers and the labor market come from studies with different periods, income measures and methods; they do not add up.
• Temporal coincidence between a policy and the fall in inequality does not show that the policy caused the fall (see the site methodology).

Related policies

• Progresa/Oportunidades (Mexico)

Where each number comes from

References

1. Engbom, N., & Moser, C. (2022). Earnings inequality and the minimum wage: evidence from Brazil. American Economic Review, 112(12), 3803–3847. doi:10.1257/aer.20181506
2. Soares, S., Osório, R. G., Soares, F. V., Medeiros, M., & Zepeda, E. (2009). Conditional cash transfers in Brazil, Chile and Mexico: impacts upon inequality. Estudios Económicos, 207–224. doi:10.24201/ee.v0i0.387
3. Soares, F. V., Ribas, R. P., & Osório, R. G. (2010). Evaluating the impact of Brazil's Bolsa Família: cash transfer programs in comparative perspective. Latin American Research Review, 45(2), 173–190. doi:10.1017/S0023879100009390
4. Ferreira, F. H. G., Firpo, S. P., & Messina, J. (2021). Labor market experience and falling earnings inequality in Brazil: 1995–2012. The World Bank Economic Review, 36(1), 37–67. doi:10.1093/wber/lhab005
5. Medeiros, M., Souza, P. H. G. F., & Castro, F. A. (2015). O topo da distribuição de renda no Brasil: primeiras estimativas com dados tributários e comparação com pesquisas domiciliares (2006–2012). Dados, 58(1), 7–36. doi:10.1590/00115258201537
6. Barros, R. P., Henriques, R., & Mendonça, R. (2000). Desigualdade e pobreza no Brasil: retrato de uma estabilidade inaceitável. Revista Brasileira de Ciências Sociais, 15(42), 123–142. doi:10.1590/S0102-69092000000100009