We are seeing an outpouring of new empirical work on inequality, led by the economists Raj Chetty of Harvard University and Emmanuel Saez of the University of California at Berkeley. The findings in their two latest papers, written with several co-authors, are casting a fresh light on contemporary political debates.
Perhaps their most striking finding is that in the U.S., intergenerational mobility has remained essentially static for about 50 years. Their own evidence focuses on people born between 1971 and 1993. Consider, for example, the probability that children who are born in the nation's bottom fifth will end up (by their late 20s) in its top fifth. For those born in 1971, the probability is about 8.4 percent; for those born in 1986, it is about 9 percent.
In the early 1970s, there was a pretty strong correlation between people's place in the income distribution and that occupied by their parents. That correlation has neither grown nor decreased over time.
All this is nothing to celebrate. If a nation -- say, Russia -- discovered that it had relatively low intergenerational mobility in 1970, and that the mobility level has stayed constant since then, it wouldn't have a new reason for national pride.
But Chetty, Saez, and their co-authors also emphasize that over the last four decades, the U.S. has seen a significant increase in income inequality as well. This increase has important implications for intergenerational mobility.
As the rungs of the economic ladder move further apart, a stable level of mobility becomes a more serious problem, because it matters a lot more whether your parents are rich or poor. The "birth lottery," as the authors call it, ends up having bigger consequences.
It's one thing if children have an 8 percent chance to get into the top 20 percent, when the top 20 percent isn't so far ahead of the rest of the pack. It's quite different if children have that same 8 percent chance, but the top 20 percent is way ahead of the rest. In the latter case, a child's ultimate prospects depend far more on the wealth of his or her parents. To put it another way, it is much worse for a child to have only a small chance to escape the bottom if the bottom is very low and the top is very high.
In a separate paper, also just posted, Chetty, Saez and their co-authors offer a more detailed account of their earlier findings of significant regional differences in the U.S. with respect to intergenerational mobility. Their striking conclusion is that the nation can be "described as a collection of societies, some of which are 'lands of opportunities' with high rates of mobility across generations, and others in which few people escape poverty." It turns out that the birth lottery is accompanied by a geographical lottery.
The researchers find especially low levels of economic mobility for children in the Southeast, where poor people are relatively stuck. In Charlotte, North Carolina, for example, a child born into the bottom fifth of the income distribution has just a 4.4 percent chance of getting into the top fifth. Disturbingly, Americans in such regions have lower levels of mobility than in any developed country for which data are available.
But in the Mountain West, the rural Midwest and some other regions, Americans have a lot of mobility -- comparable to that found in the nations with the highest intergenerational mobility (such as Denmark). In Salt Lake City, for example, those born into the bottom fifth have about an 11 percent chance of making it into the top fifth, and the figures are comparable in Pittsburgh, San Francisco and San Diego.
While the researchers can't establish causation, they find that certain factors are highly correlated with increased mobility. Areas with low percentages of single parents show higher mobility. There are also strong correlations between upward mobility and high-quality K-12 school systems. Higher mobility is highly correlated with indices of "social capital," meant to capture community involvement and social networks (including the fraction of religious people).
Levels of mobility are unusually low in regions with larger African-American populations. But whites show similarly low levels of mobility in such areas. The researchers suggest a possible explanation: Areas with large African-American populations are highly segregated, and segregation might have adverse effects on mobility for both groups.
In the diverse findings, it's important to keep the two broadest conclusions in mind.
The first is that while intergenerational mobility has been constant for decades, the effects of those constant levels on children at the bottom are getting worse in terms of the "birth lottery," simply because of growing inequality. The second is that there are significant geographic disparities in mobility, which is potentially good news, because regions that are doing poorly might be able to learn a lot from those that are doing better.