What Happened to Migration in 2015–16?

IRS Statistics of Income AND Current Population Survey Edition

I’ve been so caught up in two major topics, fertility and Puerto Rico, that I sort of dropped the ball on my core trade-stock here for a while and didn’t update you on the Current Population Survey migration data, and almost missed the IRS data. But thankfully Aaron Renn reminded me of the IRS migration data coming out today. So this post will update you on both migration datasets. Woah. Crazy.

Let’s start, as always, with headline gross rates.

The 2016–2017 CPS data showed an uptick in interstate migration, continuing the improvement since the lows seen during the recession. But CPS intercounty migration, so more local moves, fell sharply.

The IRS moves are more interesting though. The last time we talked about IRS data, I complained about a mysterious methodology change that even the IRS analysts were hard-pressed to explain. Here’s what I said:

I’m very thankful the estimated migration rate bounced back up this year, because that really validates my point: that one-year decline was not meaningful, but some kind of weird fluke in the data. The IRS mover file now shows migration rates comparable to the American Community Survey, and they both agree on rates substantially higher than the CPS, because the CPS is bad at tracking migration, in some part due to seasonality effects.

Overall, the available data suggests that interstate migration is either stable or slowly increasing, while local-area migration is stable.

So let’s look at some key subgroups. Let’s start with probably the most objectively important classification: age!

Migration by Age

Comparing the red and orange lines shows change in migration by age from 2016 to 2017. The black line shows 2000, the olive line 2007. As you can see, the broad contours of migration by age are essentially the same in all years. And as you can see, migration is still lower in 2017 for every age group than it was in 2000. That is to say, we are not close to recovering those peak migration rates, and the decline in migration is not strictly age-dependent.

But what we can see is that the migration gains from 2016 to 2017 were broad-based by age: almost every age group saw increased interstate migration. This bump was not driven by one or two groups, but by a society-wide mobility increase. That’s good news for migration.

We can also develop an age-independent “total migration rate” similar to a total fertility rate. This number tells us, if age-specific migration rates are constant over the lifespan of a child born in year X, how many times would that person move across state lines in their life? This is a useful all-in estimator of age-adjusted gross migration.

This estimate shows a rise from 2016, but arguably shows a less compelling increase in migration than the age-unadjusted rate shows. This is because migration does not change linearly with age; a bigger child cohort and smaller teen cohort artificially inflate migration through age composition, just as a large 20-something cohort and smaller retiree-cohort would. In other words, some of the observed improvement in gross migration versus the recession-period low is just an age-composition question.

But what does IRS data say? We now have several years of IRS data by age.

We can’t compare to 2015 because of the method change. But a 2014 comparison shows lower migration for households headed by people under age 26 or over age 65, higher migration for people of various in-between ages. Crucially, since most kids are in households headed by people ages 26–55, this means that actual youth migration likely rose from 2014 to 2016; IRS’ under-26 figure reflects household heads, not population age.

As a fun exercise, let’s do a comparison of age-specific migration rates in 2016 for CPS, ACS, and IRS.

As you can see, IRS and ACS have some broad agreement. CPS is very different, although CPS and ACS are close for children, while my IRS figure for children is pulled more-or-less out of thin air. Again, this is yet more reason to not put too much weight on CPS numbers.

Let’s carry on to another fascinating subject: income.

Migration by Income

Here you see, as usual, that low-income folks have higher migration rates: the under 50k-crowd moves a lot more, while the 75k+ crowd moves a lot less. In 2017, migration rose for all income groups, but the most for 75–100k and 25–50k. Since the low in 2010, migration has risen most in the aggregate for the 75–100k crowd.

How does this compare to IRS data? Weeeeelll….

So again, the 2014–15 data is bogus. But we can compare to 2013–14. Migration was lower for the under-10k crowd vs. 2013–14, and highest for the 75k+ crowd. So now quite the same trend change.

But the basic truth that lower-income folks migrate more holds up again: the under 50k crowd dominates, though IRS shows relatively higher movement for the 200k+ crowd.

Just for fun, let’s compare 2016 data for income from IRS, CPS, and ACS, as we did more age data.

Once again, the IRS and ACS show similar trends: highest migration for lower-income households, then some increase for higher-income households. The CPS… is just doin’ it’s own thing down there. This is, once again, because the CPS greatly mismeasures migration. The only good thing about the CPS is that it has a long time series.

Winners and Losers

I’m not going to do other demographic cuts because I want to publish this on the same day that the IRS data released. So let’s skip to the clickbait: winners and losers! Which states gained or lost the most due to migration?

More-or-less the usual suspects… although I want to note a couple interesting tidbits. First, one of the “usual suspects” that consistently makes good gains is Delaware. Not one people usually think of. Next up, note that New Hampshire and Maine make some gains, while California posts losses: weather must be hugely important, right? Also notable is North Dakota’s loss: the oil boom turned into an oil bust, and now they’re losing people again. Aside from that, most of the map should be self-explanatory.

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.