A new study by Antony Davies and John Pulito, from the Mercatus Center, find that tax rates do influence the migration of people across state and county lines (pdf). They found that:
This paper explores the relationship between high-income tax rates and the interstate migration of high-income households. Controlling for property-tax rates, sales-tax rates, high-income tax brackets, unemployment, and state/county specific and time-specific effects, we find that higher state income-tax rates cause a net out-migration not only of higher-income residents, but of residents in general. We also find that changes in the income levels to which the tax rates apply similarly affect out-migration. For county-level data, we find that high-income households react to a lowering of income levels to which higher tax rates apply in the same way that they react to increases in the tax rates themselves. This behavior suggests that the tendency to lower the threshold for “high income” or “millionaire” households to capture households and flee to more tax-friendly environs. Finally, for state-level data, we find that the effect of property taxes on migration is significantly stronger than the effect of high-income tax rates on migration. For example, a one percentage point increase in the property-tax differential between two states has almost three times the effect on migration as does a one percentage point increase in the difference in high-income tax rates. All of these data suggest a recipe for population depletion. States lose households to more tax-friendly states by (1) lowering the “high-income” threshold so as to capture more households, (2) increasing high-income tax rates, and (3) increasing property-tax rates.
This is good stuff, however, I noticed two little issues in the study relating to their tax data:
- First, the study only identifies Tennessee as taxing only dividends and interest. Yet, New Hampshire also only taxes dividends and interest. Also, it is unclear if they excluded Tennessee and New Hampshire or not from the study.
- In footnote 3 they state that “. . . while North Dakota, Rhode Island, and Vermont now have a progressive income-tax system, this was not the case in 2000. As of 2000, these states charges an income tax equal to a certain percentage of the federal income tax liability.” This is not true. Since the federal income tax code is progressive, then taking a fixed percentage of the federal liability will “import” the progressiveness of the federal tax code to the state.