Comments

Carol Graham: This is an excellent paper on a topic that is important to Latin America's future, in general, and the sustainability of its reforms, in particular. Alejandro Gaviria makes nice use of empirical data from both Latin America and the United States, and he uses sound methodology. I agree with the general direction of the findings, and much of our own work on inequality supports that general direction. However, the story is more complex than the one that Gaviria tells, particularly with regard to preferences for redistribution. In this latter area, our findings depart quite markedly from his.

The paper lacks a discussion of what mobility indicator is most important to attitudes about redistribution, future behavior, and so on. There are many different views on this issue (as well as some empirical results), and a discussion would have enriched the paper. I personally think that attitudes about longer-term trends—and children's future—are the most important. Here I am not so sure that Latin Americans are as far from the United States as the paper suggests. While 56 percent of U.S. citizens in the General Social Survey (GSS) think that their children will live better than they, 55 percent of Latin Americans think so. That is a surprisingly small and insignificant difference. To some extent, this reflects hope and optimism as much as anything else (in that happier people tend to have higher prospects of upward mobility, and the correlation is stronger for more speculative questions about the future). Yet it also suggests that Latin Americans retain similar hope for the future mobility of their children, despite more difficult objective constraints than people in the United States.

The paper notes that almost half of Latin Americans think that their socioeconomic status is the same as that of their parents, while a remarkably high 36 percent of Americans think that their status is the same as or worse than that of their parents. These differences are not that great, given the wildly different economic contexts and differences in macroeconomic stability. The two regions also seem to hold relatively similar views of the causes of poverty. In Latin America, the paper reports that 36 percent of respondents think that poverty is caused by circumstances other than skills and personal efforts. In the U.S. GSS, 46 percent of respondents think that insufficient effort is the reason for poverty. This is different, but it not as far off as one might have guessed. Moreover, almost 80 percent of U.S. respondents think that the lack of jobs is an explanation that is somewhat or very important to poverty.

In terms of actual mobility differences between Latin America and the United States, the paper notes differences in intergenerational educational mobility. The links between parents' and children's education are strongest at the top end of the distribution in Latin America. This is not surprising, not only because of the limited supply of higher education that the paper notes, but because of all of the other barriers that members of poor households face in trying to reach university levels of education in the region. The general concavity of the distribution for the region may also reflect the previously strong incentives for completing secondary school (such as a middle-class lifestyle, stable job in the public sector, and so on), which now have changed. The kinds of jobs that used to be available to someone with just secondary education are far fewer and less desirable than they were before; the bubble in the distribution may be explained by these earlier and more generalized investments in secondary education.

Income mobility is a trickier story to tell because of data problems. Peru provides some anecdotal, but provocative, evidence. An important caveat here is that these data address intragenerational rather than intergenerational mobility, which is different from the focus of the paper although not orthogonal to the broader discussion. My coauthor and I compare mobility rates over a ten-year period for Peru and the United States, and we find more relative mobility in Peru.¹ Some of this is explained by macroeconomic...