Georgetown Americas Institute

Measuring inequality accurately is essential for Latin America. The region has experienced numerous fluctuation cycles of wealth distribution in recent decades, creating both high levels of income and wealth inequality. A recent IDB working paper titled “Seventy-five Years of Measuring Income Inequality in Latin America” analyzes the history of the efforts to measure inequality and tackles the uncertainties that arise from disparities in estimates and measurement systems across countries.  GAI welcomed one of its authors, Nora Lustig—a resident fellow with GAI, Samuel Z. Stone Professor of Latin American Economics at Tulane University, and the founding director of the Commitment to Equity Institute (CEQ)—for a discussion of the paper’s findings.

How Inequality Is Measured Matters

Lustig argued that understanding the different ways of measuring inequality levels is critical. She explained that depending on the source of the information, inequality levels—and particularly those before the 1990s—can differ markedly. Similarly, household survey-based measurements are more likely to underestimate true inequality numbers. 

“Statements are usually made on inequality that make improper use of information. Since this is such a politically-charged phenomenon, I think it is important to understand both how we measure it and what we can say with certainty, as well as point to where the uncertainties lie.” -Nora Lustig

There is some consensus, however, on the trend of inequality, and most measurements show that while it grew in the 1980s and 1990s, it fell through the 2000s due in large part to expansions in education and increases in cash transfers and the minimum wage. To get a good picture of the evolution of inequality, Lustig underscored the importance of using a multiplicity of sources, including surveys and tax data, as well as applying different statistical methodologies. 

Methods and data matter. For instance, the numbers for Brazil in the 1960s differ greatly according to sources and methods. Meanwhile, Mexico has seen inequality either rise or drop in recent years depending on different measurements. That is why, Lustig argued, income inequality should be measured through a combination of data sources and methods. For example, she proposed that it include nationally representative household surveys as well as household pre-fiscal and post-fiscal income. It should also account for the impact of rent on income, and inequality calculations should include consumption expenditures for cross-regional comparisons. 

How Much Do We Actually Know about Inequality?

Lustig recognized four clear periods that emerge from disparate analyses of regional inequality data. The first, ranging between the 1950s and 1970s, was marked by scarce and idiosyncratic information. The second and third, which range from the 1980s up to around 2015, constitute two halves of an inverted “U” pattern: rising inequality in the 1980s and 1990s, only to then decline in the 2000s and 2010s. The fourth and ongoing period is defined by regional divergence as inequality rose, fell, and remained constant in different countries. The problem is that most of the historical data is based on household surveys alone, which are likely to underestimate real inequality. 

“Recent progress in the measurement of income inequality in the region is real, incomplete. We now know we were wrong in the past, but we do not yet know ‘the truth.’” -Nora Lustig 

The challenge for Lustig is figuring out what comes next. Inequality is clearly sensitive to different combination and imputation methods, but it is not clear that recent synthetic estimates represent credible and robust estimates of true inequality levels. The good news is that trends are much more robust than specific level estimates, which lends certainty to findings like the inverted “U” pattern of earlier years. In fact, the decline of inequality at the turn of the millennium was a generalized pattern across the region. Expanded access to education played a central role in reducing the wage gap, and certainty that education works as a driver to reduce inequality is high. 

“One thing that we are certain about is that when we look at the decline of inequality in the 2000s, this was clearly associated with expansions in education, wealth transfers, and remittances.” -Nora Lustig

GAI Founding Director Alejandro Werner provided introductory remarks. A full recording of the event is available on GAI’s YouTube channel. The report can also be found online on the IDB’s website.