Does Migration Have an Educational Bias?

Evidence from Available Surveys

I didn’t discuss education in yesterday’s discussion of migration data because the IRS Statistics of Income do not have any educational microdata attached to them. But today I want to look at CPS and ACS data on educational migration and see if it tells us anything interesting.

To start with, let’s look at some graphs of migration rates by education. Here’s CPS data from 1964–2017, with missing years linearly extrapolated. Also, 1999–2005 won’t match what you can download from CPS because I make corrections for a known mistake in CPS’ imputation methods for nonresponsive surveys.

I’m going to use consolidated all-cross-county migrations as my key variable; this includes any move across state lines within the U.S., whether within a state or across state lines.

The first thing to take note of is that more educated people have always had higher migration rates, from 1964 to the present day. What interests me, however, is the compression of difference: up to 1990, non-HS grads, HS grads, some-college folks, and then BA-or-higher folks had about evenly spaced migration rates. But then they compress, such that since 2000 or so, there have basically been two groups: folks who did at least some college, and folks who did not do at least some college.

Let’s show this a different way: gross migration rates as a percentage of unweighted average rate across these 5 classifications.

This chart essentially eliminates the trend component in migration. And it shows a fascinating change in relationships. Migration has greatly converged because, relative to society-wide migration rates, low-education peoples’ migration rates rose significantly from the 1960s to the late 2000s, though that trend has since reversed. Meanwhile, migration rates for the best-educated people actually had the worst performance from the 1960s to the 2000s, and, relative to grou-average, have been stable since.

Far from less-educated people being chronically stuck in place, declining migration through the 2000s primarily impacted well-educated people!

Since the 2000s, that may have changed some, as well-educated people have seen some migration recovery, while less-educated people have not. But less-educated people today remain relatively more migration-prone, vs. society on the whole, than they were in the 1960s. And on the whole, the 2000s were a time of uniquely “compressed” migration rates, meaning people of all educational levels moved at roughly similar rates.

But of course, that’s relative rates, vs. societal norms. What happens if we compare the pace of decline for each group? This exercise will depend heavily on what time period we chain to, and if we use absolute decline or relative decline. I’ll produce four graphs showing all possible metrics.

As you can see, in absolute terms, the steepest decline in migration has been among the most-educated groups. Their decline also began earlier, in 1985 vs. 1990. Thus, in terms of which group has changed the most versus their migration rates from 1965 to the late 1980s, the answer would have to be educated people.

We have experienced a decline in migration led by an increasingly geographically stagnant upper-class.

Let’s zoom in to the more recent period, however.

I show pre-2010 difference vs. 2010 as well but want to focus on post-2010 changes. No group’s migration has worsened since 2010 in this data. But some have risen more than others. Migration of graduate degree-holders has essentially returned to pre-recession levels… while folks with a high school degree or less are stock at or near the recession-period bottoms. Folks with any amount of college are in between.

While the 1980s-200s decline was most severe for the most educated, the post-recession migration recovery seems to be almost exclusively among educational elites.

This is a big change in the relationship between categories of migrants. Now, selection may have a role here… but I don’t think so. The reality is that these relationships changed sufficiently quickly that I really don’t think the underlying-migration-propensity of individuals who select into graduate vs. just a BA changed that much between 2005 and 2017, nor at the lower end either. And the long-run trend suggests that if anything selection tended to move lower migration-propensity people into higher educational categories, which is surprising.

But let’s cut the data a new way. Instead of absolute decline, let’s look at relative decline.

You might have been saying to yourself, “But Lyman, the decline in migration being greatest among the most education is just a statistical artifact of their higher initial-period gross rates. If you index, then…”

No.

When we index to 1964, it turns out that migration has still fallen the most among the most educated, and the least among the least educated.

This does not lend credibility to the idea that people who had pre-existing potential to be geographically mobile filtered up educationally, as rising population share among the best-educated groups coincided with falling gross migration, while the decline was far less severe among the shrinking groups. So the story of the 20th century at least through the 2000s was pretty simply one of plummeting migration by elites, while less-educated people tried to hang in there and keep moving somewhat more.

But what about the recent period?

Okay, there’s the boom for better-educated folks. But guys. The truth is that trends are fairly for anybody below graduate degrees. College folks are better than non-college, but it’s not a huge difference, and completion does not seem hugely important relative to non-completion or an Associate’s Degree.

So there is a skills-biased migration story, but mostly at the very tippy-top-end of skills.

Let me do this another way. Here’s the same data from 1990 to present, but I’ve broken out PhDs, Masters’, and other professional degrees separately.

As you can see, PhDs fell less than other groups, but have posted real gains. Professional degree-holders are an odd lot and have lower migration than other advanced degree-holders, but have seen more gains since 2010 than BA holders, though their current rates are similar. And MA holders also saw a very good recovery. But the point is, even among graduate degree-holders, the highest-degree-holders have done best.

Since the late 2000s, migration has increasingly biased in favor of the very highest-educated groups.

Their migration rates are still very low compared to even the early 2000s, and these trends may not last, but they do seem to reflect the “spiky” world of talent.

Comparing to the ACS

But this is all CPS data. How does ACS data stack up? Due to some differences in the data, I can only compare interstate migration rates for the ACS and CPS, 2001–2016, by major educational groups.

As you can see, the ACS shows a more consistent recovery for graduate-degree holders and BA-holders, and indeed smoother lines overall, and also higher total estimates. But broadly speaking, the core trends are the same: no recovery for low-education migrations, significant recover for educated migrants, most significant for the highest-education migrants.

So this trend looks pretty solid and verifiable. Migration today is becoming more educationally biased.

Conclusion

Throughout the 20th century, migration fell primarily because the most-educated people moved less. This is contrary to the story you often here, where it’s working-class people who stopped moving. It is true that more educated people migrate more, but this cross-sectional truth should not be mistaken for a time-series truth. But things have changed since 2010. Today, migration is rising for the best-educated, but stagnant or falling for the least educated. This is a reversal of trend from the 20th century. However, this reversal might simply be a return to pre-1990s norms, whereby better-educated people had vastly higher migration rates than today.

We’ll see how the future develops!

Check out my Podcast about the history of American migration.

If you like this post and want to see more research like it, I’d love for you to share it on Twitter or Facebook. Or, just as valuable for me, you can click the recommend button at the bottom of the page. Thanks!

Follow me on Twitter to keep up with what I’m writing and reading. Follow my Medium Collection at In a State of Migration if you want updates when I write new posts. And if you’re writing about migration too, feel free to submit a post to the collection!

I’m a native of Wilmore, Kentucky, a graduate of Transylvania University, and also the George Washington University’s Elliott School. My real job is as an economist at USDA’s Foreign Agricultural Service, where I analyze and forecast cotton market conditions. I’m married to a kickass Kentucky woman named Ruth.

DISCLAIMER: My posts are not endorsed by and do not in any way represent the opinions of the United States government or any branch, department, agency, or division of it. My writing represents exclusively my own opinions. I did not receive any financial support or remuneration from any party for this research.