Methodological Notes for “How State Taxes and Migration Impact Inequality”
Because Nobody Wants a 2,300 Word End Note
Some Specific Taxes and Their Effects
Income Taxes
Generally speaking, income taxes will have the most impact on high-income migration, which means middle-aged migration. Middle-aged people are less likely to be migrants, so while income taxes matter a lot for the people they affect, the people they affect aren’t the most common migrants.
The exception is what we might call the “expectational channel.” Migrants, especially educated migrants, may have low income when they migrate, but may reasonably expect higher income in the future. Thus even low income migrants may care about taxes they don’t pay, if they are optimistic about their own future. This means that even “millionaire taxes” may discourage migration by lower-income people, though we should expect a fairly strong “skill bias” in changed migration among the lower income.
One group has a very binary relationship with income taxes: retirees. For retirees, what matters is less the tax rate and more the exemptions. Some states exempt major portions of retirement and social security income, while others do not. This essentially binary effect should impact retirees, though some curious academic research has found that it has very little effect (though the same research also found that migrants move away from locations that provide generous public services and amenities: which is a very strange finding).
Sales Taxes
As I mentioned in Isabelle’s story, sales taxes will disproportionately impact people who consume taxed goods. This tends to mean lower-income individuals, who have higher consumption as a share of income, and lower savings. But this can also just mean young people.
How states tax online sales may be particularly important to young people, who are more likely to shop online. If living in one state or another changes whether online purchases are taxed, it could influence where online shoppers live.
Finally, sales tax exemptions, such as for groceries, services, or clothes, can radically alter the burden of the sales tax. States that tax more services may see less “income bias” in their sales taxes, and thus less distortion of migration flows, whereas states with very narrow sales taxes will likely see more biased migration flows, all else being equal.
Younger and lower-income people are more likely to migrate, so even if sales taxes have a more muted effect than income taxes, the groups they effect are much more likely to migrate. Thus sales taxes can be expected to impact migration.
Estate, Inheritance, and Gift Taxes
These taxes are all uniquely important to retirees and people near the end of their lives. Academic research has shown a strong tie between such taxes and retiree migration (though, again, there’s not universal agreement). However, retirees migrate less than many other groups, so even a large effect on retirees may have a small effect on total migration.
Excise Taxes
Excise taxes, such as on alcohol, cell phones, cigarettes, or other products, are rarely large enough to drive migration. Most migrants probably aren’t even aware of these taxes. However, the main reason for this isn’t that people don’t care, but that these taxes tend to be small overall and effect very targeted groups. Not everyone pays every excise tax.
All the same, for migrants who have uniquely strong preferences for taxed products, excise taxes could, theoretically, impact migration. There’s very little evidence on this, however, and usually such connoisseurs have sufficient income to cover the taxes, or else they like to live in geographic areas known for the product in question. In such areas, usually the industry (like wine in California) successfully lobbies for lower taxes.
Thus, while theoretically excise taxes could have big effects on narrow groups, in reality they don’t seem to have a huge impact.
Property Taxes
The early literature on taxes and migration in the 1950s and 1960s emphasized how people migrating around cities responded to property tax burdens. Indeed, property taxes can have a huge impact on short range migration. People may move to a new house, neighborhood, or even county to “shop” for lower taxes. And property taxes do show huge variation across local areas.
However, long-range migration isn’t likely to be as influenced by property taxes. First of all, while property taxes vary widely around localities, virtually every urban area has high-tax districts and lower-tax districts, with just a few exceptions. Certainly almost every state (except maybe New York and New Jersey) has some low and some high tax areas.
Second, property taxes are primarily paid by property owners. Even if renters bear some of the incidence, it’s indirect, and much of it falls on the owners. But homeowners are far less likely to migrate than renters. And when people do buy a house, while they do consider property taxes, those taxes usually pay for local services people value: schools, parks, police, etc. Most people get their money’s worth for property taxes, so the net effect on long-range migration is probably small. People may move because they prefer a smaller or a larger government, but it’s not likely to be a long-range move, and it’s almost certainly a smaller effect than other taxes, especially given how unlikely to migrate homeowners are.
Review of the Literature
Some of the earliest academic work exploring the link between taxes and migration was carried out in the 1950s by economists like Charlies Tiebout, NAME Tullock, and Christopher Stewart. Stewart was particularly focused on sales tax disparities, and also developed a theoretical framework for secondary migration caused by business relocation. On the other hand, Tiebout and Tullock focused on property taxes and public service disparities as drivers of migration among individuals. For various reasons, the “Tiebout-Tullock” hypothesis proved to be the most lasting and popular paradigm, but I’ll return to both frameworks regularly in this post.
In 1975, Richard Cebula and Christopher Curran extended the Tiebout-Tullock hypothesis to show that, at least in the 1960s, higher property taxes had a demonstrable effect on urbanizing migrants. Then, in 1984, work by Seong Woo Lee and Curtis Roseman including property and income taxes, and exploring migration differences by race, further found a significant and negative role for taxes. In 2009, Cebula updated his study with more recent and more detailed data, and found that the inclusion of income taxes alongside property taxes also confirmed the results of his original regressions.
In the 1990s, a new round of writing on interstate tax competition begin. One of the most notable early publications came from Robert Preuhs, who initiated the formal study of how state political ideology might drive migration, and also greatly expanded which taxation and spending variables were included in tax-migration studies. Preuhs found, like most academics before him, that high taxes negatively impacted migration (as did high welfare spending, although high education and infrastructure spending attracted migrants, as did a more liberal political ideology).
But another important paper, by Karen Conway and Andrew Houtenville, began to question the taxes-migration consensus. Their paper found some limited support for the Tiebout-Tullock hypothesis, but also some interesting quirks in the data suggesting that the story wasn’t quite so simple.
With the new millennium came a new round of academic contention. Agrowing number of academics have begun see the link between taxes and migration as weak, suggesting that high-earners may not be as responsive to tax-hikes as previously believed and may only have a very limited impact on the distribution of population. Karen Conway in particular has forcefully argued that taxes have little impact on elderly migrants in particular. Other studies have also questioned traditional assumptions that the possibility of migration reduced potential for state-level redistribution, and that income taxes influence migration.
So is the jig up for taxes influencing migration? Actually, no, far from it. Research demonstrating the link between taxes and migration has become far more precise and detailed.
Research carried out by Canadian researcher Kirk Collins in 2008 likewise that interpovincial and interstate tax differences can seriously alter returns to education, thereby contributing to out-migration and “brain drain.” A 2013 NBER working paper by a team of economists including Emmanuel Saez found that tax preferences for the very wealthy provided in Denmark succeeded in attracting in-migration by high-earners (though there were some less rosy effects as well). This wasn’t Saez and Co’s first time demonstrating that taxes impact where people live: in a 2010 NBER working paper, he and his co-authors demonstrated that European football players’ migrations were influenced by local tax rates.
Research focused just on cross-border metro areas (which I’ve noted before) shows how taxes not only impact where people live, but traffic flows and commuting times, thus infrastructure demand (and thus the allocation of state and federal infrastructure spending, independent of local taxation). These studies find that the largest differences don’t materialize except in cities with very big tax wedges, and don’t find the effect to be extremely large, but it does exist.
Perhaps most convincingly of all, in 2011 economist Mark Gius used longitudinal data from the National Longitudinal Survey of Youth to track individuals across many years of their lives. He assessed their migration and income, and looked at the tax rates they would face as they moved. In this analysis of unprecedented detail, he found significant evidence that taxes influence migration. However, he also found that the effect was small.
In sum, this debate is still raging. There’s no “consensus” on taxes and migration. However, most studies do find significant negative effects of higher taxes, even as there is wide variation in assessments of which taxes matter, how large the effect is, and whether or not that effect “matters.” I’ve adopted what I regard as a fairly middle of the road position: taxes may not cause migration, but they do alter destination of migration, and do define some of the structural effects of migration.
So is Texas getting migrants because it has no income tax? Somewhat. But oil matters too, and booming cities, and immigration, and natural population growth, and education, and geographic location, and many other variables.
Endogeneity Biases: What if Taxes Impact Growth? What if Taxes Interact With Growth?
The vast majority of migration studies include some kind of control for economic growth. They aim to determine how much migration is due to just taxes, not economic growth.
But there’s a problem: many economists believe that taxes impact growth. I’m not going to argue one way or the other here, as I’m not a macroeconomist and far greater minds than mine have hashed out this debate ad nauseum. But this argument, as it relates to migration, amounts to an endogenous variable bias. Maybe we don’t detect an effect of taxes because the effect of taxes is captured in economic growth. If nobody went tax-shopping, but taxes had a huge impact on migration, then migration regression estimates for taxes would be low or zero. But that wouldn’t indicate taxes didn’t impact migration: if low taxes boosted growth, they did impact that migration.
No study of which I am aware has sought to decompose economic growth at the state or city level into tax-induced and non-tax-induced components, and then recalibrate migration regressions based on such a decomposition. Indeed, such a study would be phenomenally difficult to appropriately control and carry out, and would doubtless be subject to numerous criticisms. Thus, we can only speculate.
Notably, most advocates of the low taxes-high migration connection also believe that low taxes increase economic growth, and believe that high migration increases economic growth. If migration also impacts growth, then we need a second decomposition: how much growth is driven by taxes, how much by tax-induced migration, how much by non-tax migration, and how much by “other.” This decomposition can continue as an infinite regress to some asymptote of taxes and migration.
I fall in the camp that believes migration probably does impact economic growth, and taxes probably also impact economic growth (all else being equal). That said, drastic tax reductions can cause governance issues that repel migration, or reductions in public services that may also repel migrants.
I present no resolution here to the endogeneity problem in taxes and migration. But the very existence of the problem supports the idea that taxes impact migration. The regressions that currently exist measure only one kind of taxes-migration relationship, and it’s a kind that is likely very marginal. There are other channels that may be much larger about which we lack information. Overall, current academic literature suggests that higher taxes may slightly reduce migration, but that’s a lowest possible estimate given other possible channels for taxes to impact migration.
Furthermore, low taxes may not create migration on their own. Low taxes don’t matter if there are no jobs. Thus taxes interact with available employment. The availability of jobs operates as a “first test” for a migrant, and taxes come second. So studies of migration may need to differentiate the effect of taxes by the period of the business cycle, state or local unemployment rates, or data from the JOLTS survey.
Are You Saying Rising Inequality is Because of Migration and State Taxes?
NO.
I am arguing that the existence of decentralized taxation in a Federal economy with freedom of movement creates a base level of higher inequality. Numerous other factors have contributed to rising inequality. Falling migration actually may have dampened rising inequality in the short run (though reduced mobility and thus increased inequality in the long run).
My argument is not that we can explain all inequality through migration. It is that certain structural features of migration and federalized taxation boost inequality.
When states change their income tax codes, it can alter migration patterns or alter inequality, or some mixture of the two. That’s all that I am arguing. Taxes obviously have other effects as well, and inequality and migration both have other determinants as well.
Start the series from the beginning!
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. But what’s just as valuable for me is if you click the recommend button at the bottom of the page. Thanks!
Follow me on Twitter. Follow my Medium Collection at In a State of Migration. I’m a grad student in International Trade and Investment Policy at the George Washington University’s Elliott School. I like to write and tweet about migration, airplanes, trade, space, and other new and interesting research. Cover photo from Unsplash.
