Income mobility over a lifetime might be on the decline
Ask five people about economic mobility and you might end up with six definitions of the term. It’s a knotty subject because the term really does have a number of definitions. Lots of discussions about mobility center on the intergenerational kind—about how much of a parent’s economic situation gets passed onto their children. But there’s also intragenerational mobility—a lesser-discussed kind of mobility that looks at how much a person changes their position on the income ladder within their own lifetime.
Research over the past couple of years indicates that intergenerational mobility in the United States has stayed relatively constant since the early 1980s. In contrast, a new paper shows that intragenerational mobility has been on the decline amid rising income inequality.
The new paper, released today as part of Equitable Growth’s working paper series, is by Michael D. Carr and Emily E. Wiemers, both of the University of Massachusetts-Boston. It looks at how mobility over a career has changed since the early 1980s. Previous research on intragenerational mobility using data from the Social Security Administration found the overall level hasn’t changed that much. Carr and Wiemers also use data from the Social Security Administration, but they include administrative data from the Internal Revenue Service as well. Both of these datasets are linked to data from the Survey of Income and Program Participation, or SIPP, which follows households over time and has information on demographics and education levels.
The two economists find that a person’s initial place in the income distribution mattered more in the early 1990s than it did in the early 1980s. Specifically, they look at the mobility of workers over 15-year periods, the first spanning 1981 to 1996 and the second from 1993 to 2008. In the more recent cohort they find a stronger correlation between a person’s rank when they start a career and that person’s position later in life. A stronger correlation means that a person’s position earlier in life is more predictive of their later position, and therefore there are lower levels of movement up and down the distribution.
What’s particularly interesting is that the inclusion of the SIPP data lets the authors look at how trends in mobility have differed by educational levels. For workers with less than a high school diploma, intragenerational mobility hasn’t changed that much. The difference isn’t statistically significant in Carr and Wiemers’s analysis. But mobility has declined (as shown by a stronger correlation) for workers with a college degree. In fact, for the 1993-2008 cohort, college-educated workers didn’t experience more mobility than workers with less education. The results indicate that this may be because college-educated workers now have a higher starting point than in the past.
These results all come with the usual caveats of new working papers. But if the results stand, they require some careful thought from both researchers and policymakers.