Intragenerational Income Mobility in Latin America
In lieu of an abstract, here is a brief excerpt of the content:

Latin American countries have experienced substantial macroeconomic instability over the last few decades. While the region as a whole experienced economic growth during most of the 1990s and 2000s, there were also years of stagnation and economic decline. Furthermore, most of the countries in the region experienced quite varied episodes of growth, crisis, and recession. Economists have traditionally assessed the welfare impact of these fluctuations on the population by studying the evolution of economic inequality and poverty. Questions regarding who benefits from economic growth and who is hurt by economic decline have been answered by examining the changes in cross-sectional inequality or poverty associated with these episodes. While it is important to know the evolution of inequality or poverty per se, this type of analysis fails to measure one important aspect of welfare, namely, the evolution of the well-being of given economic units through time. The goal of mobility analysis is precisely to study this dynamic evolution of well-being for units identified through time.

The following example demonstrates the difference between cross-sectional analyses of inequality (or poverty) and mobility analysis. Take an imaginary economy with two individuals whose initial incomes are $1 and $3. Suppose the economy grows and the new incomes are $1 and $5. Inequality has clearly increased in the course of economic growth, but what has happened to the destinies of specific individuals? Anonymous data provide no indication. Panel data, however, reveal two underlying possibilities. The result of adopting the notational convention that individuals (denoted by Greek letters) are ordered [End Page 101] from lowest initial income to highest initial income in the initial income vector was either

Case I:inline graphic


Case II:inline graphic

In the first case, the income of the poorest individual remained unchanged, while the income of the initially richer individual grew. In the second case, the initially poor individual experienced a substantial income gain, while the other individual recorded an income loss.

As this brief example illustrates, simply comparing anonymous distributions of income across time cannot answer questions like whether the (initially) poor are getting poorer and the (initially) rich richer or whether economic growth is benefiting individuals who were initially poor. Mobility analysis addresses such issues by tracking the evolution of individual incomes over time and identifying the winners and losers of the growth process. In other words, the crucial difference between mobility studies and dynamic comparisons of cross-sectional measures of inequality and poverty is that mobility studies can unveil the intertemporal anonymity that accompanies cross-sectional studies.

Economic mobility has not been widely studied in developing countries until very recently owing to the lack of suitable data. Studying mobility requires longitudinal data tracking economic units (that is, individuals, households, or firms) over time. Collecting this type of data is expensive, and historically few Latin American countries carried it out. Now, however, such data sets are available for a number of Latin American and Caribbean countries; table A-1 in the appendix provides a list of available panel data sets that can be used for income mobility studies for these countries. In this paper, we discuss how the knowledge gleaned from mobility studies differs from comparable cross-sectional analysis.

The structure of the paper is as follows. The next section discusses what mobility is, how it can be measured, and how it differs from inequality. The subsequent section reviews previous mobility studies in Latin American countries. The paper then summarizes the contribution of our own recent work, and [End Page 102] the final section discusses what lies ahead in mobility research for Latin American economies.

As indicated by the title, this paper deals only with the study of intragenerational mobility. Intergenerational mobility is an important area of research, but we do not address it for the sake of brevity. Readers interested in this literature applied to Latin America should refer to the authoritative paper by Behrman, Gaviria, and Székely.1

Defining Mobility and How It Differs from Inequality

As used in this paper, an income distribution is the entire vector of...