It’s a long way down if the glass breaks.

Will a Tax Hike in Illinois Make Migration Worse?

Yes! No! Maybe! Both! It Depends!

Illinois looks set to raise its taxes, pending a veto-override vote in the lower house of the state legislature on Thursday.

The politics are boring and I won’t discuss them. Likewise, I have no opinion on the budget itself, having done approximately zero thorough research on it. I don’t know details on what’s being debated other than a few features that seem salient for migration. However, since it’s Illinois, there’s going to be debate about migration-related impacts. I was asked to blog it:

So here’s the question. How will income tax changes impact Illinois’ outflow of people?

Well, for this, we can look to the academic literature. My preferred study for this is here, because it has the richest data for estimating the elasticity of migration with respect to taxes, using, as it does, longitudinal data following people across tax, geographic, income, and employment changes. But I’ve written at some length about the research on taxes-and-migration, even specifically in Illinois, as I’ll return to later on.

Basically, I’m using a pretty standard estimate of the sensitivity of migration to tax changes. For different population groups, elasticities rangte from -0.02 to -0.05, but for most groups, it’s around -0.025. I should note the implication here: I am using estimates of migration sensitivity to taxation near the low end of the academically justifiable range. Estimates 2+ times higher exist within the academic literature and are not crazy.

I will base all of my estimates on the ACS, since it gives sufficient age/race breakouts to compute migration rates directly comparable to the groups identified in the paper I’m using as my elasticity benchmark. I should also note I truncate the elasticity range at the top; i.e. I assume the paper’s highest group-specific elasticities are implausibly high given other groups. This is another down-bias on my estimate.

Furthermore, for my estimate of the tax shock to plug into my group-specific elasticities, I bias downwards. Technically, tax burdens rise by 32% (4.95%/3.75%). However, I assume that expanded tax credits and clever tax avoidance will cause Illinoisans’ experienced taxes to only rise 20–30%.

In other words, at every turn, I am seeking ways to bias my estimates downwards to give a minimum plausible threshold for the effect. Nobody should be seriously advancing effect sizes smaller than these.

Under these down-biased, consensus-range estimates, gross outflows could be expected to rise by between 2000–2500 people per year.

For reference, gross outflows in 2015 according to the ACS were about 304,000 in 2015. So this is a 0.6% to 0.9% increase in gross outflows. It amounts to nudging Illinois’ gross outflow rate from 2.396% to 2.414%… which is definitely within the normal margin of error.

The correct adjectives to use when describing a one-year change of this magnitude is “marginal,” because the change is within the normal margin of error for ACS estimates.

Now, that is just outflows. I do not have estimated inflow elasticities of the same detail, but generally speaking inflows are more sensitive than outflows, especially in the short-run, to any given shock. Let’s assume elasticities would be 25% greater for inflows. This means that gross inflows change from 1.632% to 1.617%. Again, these are changes much smaller than the margin of error.

But the effect of a tax change is fairly persistent, only offset by future tax changes in Illinois or other states. If we assume that this tax-impact only declines by, say, 5% per year due to other-state changes, and no change is made to Illinois taxes, then the 10-year impact is in the 15,000 to 22,000 range for higher outflows, and -14,000 to -19,000 for lower inflows.

Over ten years, a plausible, consensus-based estimate for the direct impact of the proposed tax change on Illinois population via migration is -30,000 to -39,000, or about -0.24% to -0.3%.

Is that small or large? I dunno. By definition, it is “marginal,” i.e. a small Census revision could have a bigger observed impact than the tax change, so that seems pretty small. On the other hand, a loss of 30,000 extra people is the loss of an entire town or county in many cases. That’s big. I won’t enter the fray about whether this change is “big” or “small,” in a politically meaningful sense. But I will mention some caveats:

1. This change is not large enough to offset larger demographic trends. That is, if Illinois got good international migration or fertility, this specific tax increase wouldn’t stop it from growing.
2. I biased my estimates downwards; if I adopt estimates near the top-end of the academically justifiable range, population losses due to tax-associated net migration come to more like 70,000 to 100,000 over 10 years.
3. These effects are much smaller than those estimated for the progressive tax proposal a while back, which makes sense given that proposal had a much higher top rate.
4. Population losses due to outflows could exceed the raw net migration losses. Most papers suggest that youngish people have higher migratory elasticities; ergo, additional losses will also reduce the high fertility/low mortality population, worsening natural increases. Over ten years, this could worsen estimates by several thousand people.
5. Ten years is a long time; take my estimates with a grain of salt! Lots can change between now and then!
6. I have made no explicit assumption about how revenues are used. The implicit assumption from the paper I’m drawing on suggests that this is the impact of taxes if revenues are spent in the “normal” way, i.e. an average share of marginal revenues being dedicated to education, pensions, roads, etc. If more money goes to current services like education and transit, I could be overstating losses from migration. If more money goes to debt service, pensions, or welfare, I could be understating losses from migration. I don’t know enough about the current debates to know how this money will be used.
7. There is some evidence that people are forward-looking in terms of migration; i.e. they engage in some degree of Ricardian equilibration. So states with poor credit lose migrants even if they have good tax/service mixtures, because people believe they will eventually face higher taxes or worse services. To the extent a tax increase changes peoples’ expectations about Illinois’ future fiscal performance, it could alter migration. The more stability a change introduces to expectations, the better; the more a change in current policy raises estimates of future service spending and lowers estimates of future taxes, the better.
8. You can’t have everything at once. You cannot raise services, cut taxes, improve creditworthiness, and change expectations-channels all at once without some remarkable model parameters about the shape of peoples’ expectation-forming mechanisms.

I won’t say whether proposed changes are good or bad, or how they fit into the above heuristics. I just hope people in Illinois are putting some serious brain-work into planning for these questions, and I especially hope they will work together in a spirit of community while they tackle these problems, because the attitude they bring to the table matters in itself for the migration response.

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.

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.