In this interview, Bhash Mazumder discusses his work on intergenerational mobility, why the subject is so difficult to study, and how race factors into mobility and inequality.
Bhash Mazumder is a senior economist and research advisor at the Federal Reserve Bank of Chicago, where he focuses on labor economics, education, and health. He is also currently an Affiliated Scholar at the Stone Center.
Mazumder recently spoke to us about his paper, “The Decline in Intergenerational Mobility After 1980,” available through the Stone Center Working Paper Series, as well as about how mobility changed for both whites and Blacks around this key inflection point.
What is the trend over time in intergenerational mobility in the United States? Has it increased as was expected along with rising inequality, or has it not, as some research has indicated?
We look at two different concepts of intergenerational mobility in this paper. We focus mainly on relative mobility, which is what most of the economics literature has looked at until more recently. And we find that there has been a decline in relative mobility that has pretty much coincided with the major change in inequality in the labor market, where there was a key inflection point around 1980.
We also look at absolute mobility, and find a decline in that as well — maybe a little bit of a smaller decline than what was found in the paper by Chetty and coauthors published in 2017 in Science, but consistent with their findings.
Your paper looked at two cohorts: one born between 1948 and 1953 that entered the labor market during the 1960s and 1970s, and a second born between 1961 and 1964 that largely entered the labor market after the pronounced rise in inequality around 1980. Has intergenerational mobility continued to decline for later cohorts?
My sense is that it has been more stable since that inflection point. But we don’t really have great data on that as more recent cohorts are still relatively young. One reason why you might expect it to be flat in more recent periods is that part of what’s driving intergenerational persistence has to do with the returns to schooling — for example, how much you get in the labor market for having a college degree versus not — that appears to have been relatively flat in the last couple of decades. There’s a good paper on this by David Autor showing that the return to a college degree has been relatively flat in recent years. And in some other work I’ve done, I’ve seen that relative intergenerational mobility has been relatively flat since that big rise in inequality after 1980. That’s my sense of where things are, but I don’t think we know conclusively.
Why is intergenerational mobility such a difficult question to study, in terms of data availability and methods?
One of the real challenges for studying intergenerational mobility in general is that it’s really a lifetime concept. Ideally what you want is income for parents covering a big part of their lifecycle, and for kids as well. That makes the data requirements really high.
And in the U.S., we’ve only really created what we call panel data — where we track in surveys the same people for very long periods of time — starting in the middle of the last century, and mostly toward the second half of the 20th century. The best-known survey for doing this is the PSID, which started in 1968. That’s not really well suited for capturing patterns around this inflection point. The reason is that if you thought that people who entered the labor market during the 1970s had higher mobility because inequality hadn’t yet risen, you would really want to study people who were born during the 1940s to early 1950s. And there’s very little of that available in the PSID sample.
Another paper by Chetty et al published in the American Economic Association Papers and Proceedings in 2014 investigated trends using administrative income data, but that data only covered recent cohorts born since 1970. They found that trends in relative mobility were flat which is consistent with previous studies such as the work by Autor on the returns to schooling I mentioned earlier.
So we turn to another data source that hasn’t been used as much by economists in the last couple of decades, but had been back in the ’70s and ’80s. These are the National Longitudinal Surveys. And what’s nice is the first National Longitudinal Surveys started in 1966 and allowed us to look at children who were born in the 1940s and really home in on the cohorts that would have been born before inequality rose. That’s how we’re able to overcome this dataset challenge.
And how do racial differences in intergenerational mobility factor into this discussion?
In previous work, I’ve found that there are stark difference in intergenerational mobility by race. Blacks have much less upward mobility and much greater downward mobility than whites. In the current Stone Center paper we did not analyze trends by race, but ahead of this interview we did produce some new results. What we found is that the very large gaps in mobility by race have stayed about the same. This is shown in the figure below.
[Figure 1]
Related Research: The Decline in Intergenerational Mobility After 1980