A Closer Look at IRS Migration Data

Back to Blog

Earlier this year, the conservative Center of the American Experiment (CAE) released a report based on Internal Revenue Service Statistics of Income (SOI) that purported to show significant migration of high income households from Minnesota due to the state’s tax climate. However, a deeper dive into the IRS data finds that the CAE report omitted or failed to emphasize trends that contradict their premise that income is fleeing Minnesota due to taxes.

From the outset, the CAE report was hamstrung by methodological problems. First of all, the data they used showed only movement of tax filers, not of income. This distinction is critical. Former Tax Foundation economist and migration blogger Lyman Stone summarizes the situation:

It is common for migration commentators to treat the AGI [adjusted gross income] of IRS SOI migrants as “migration of money.” This is an egregiously wrong use of the data. The IRS SOI user guide makes clear that this is not a viable interpretation of the data, and thus those who read the data this way have either failed to perform the most basic due diligence by looking at the manual, or else actively mislead their readers.

Stone gives several reasons why the IRS SOI data does not track migration of income, contrary to what is assumed in the CAE report. Two of these reasons are described below.

  • According to Stone, “When a migrant moves from one state to another as a result of a job transition, it is likely that they move from a job that will still exist after their departure, to a job that existed before their arrival.” For example, if a person with a job that pays $100,000 a year leaves Minnesota, this person’s income generally does not leave with him. The job is likely taken by a Minnesota resident, in which case the income associated with that job stays in Minnesota.
  • The SOI data measures the income of migrants after they arrive in a new state, which is not necessarily the same as their income prior to leaving. For example, if an unemployed person with no income leaves Minnesota for a job in another state, Minnesota loses no income.

In both of these instances, the assumptions used in the CAE report would result in an erroneous report of income migration where none (or a smaller amount than assumed) occurred.

Another problem with the CAE report is that it conflates the address from which a return is submitted with the residence of the filer, which is the critical factor in determining the state to which taxes are paid. A tax return may list an address of the tax preparer, a business address, or a P.O. box that is different from the residence of the taxpayer. To the extent that the state from which the tax is filed is different than the state of residence, a false attribution of income migration would occur. Such situations are likely uncommon for simple wage earners, but could be significant for high-income households with complex business relationships and tax structures. It is precisely these high-income households that are the focus of much of the fretting over tax-driven income migration.

Apart from these methodological problems, a March 2016 North Star article noted several oversights within the CAE report. Foremost among these was the fact that most of the alleged income migration that the CAE bemoans occurred when anti-tax, anti-investment conservatives dominated state fiscal policy. The CAE report failed to recognize the concurrence between the policies they endorse and the outcomes they decry.

As if these problems weren’t enough, much of the remaining wind came out of the sails of the CAE report with the release of information from the Minnesota Department of Revenue (DOR), which revealed a significant increase in the number of very high-income taxpayers and the income taxes paid by them in the year following enactment of a new “top tier” income tax rate affecting households with taxable income in excess of $250,000 for married joint filters, $200,000 for heads of households, and $150,000 for single filers. The premise of the CAE report was that these tax increases should have caused massive tax flight among high income households.

Even if we accept the premise that the IRS SOI data measures income migration (which it does not), a closer look at this data reveals patterns that are inconsistent with the premise that this alleged migration is driven by taxes. For example:

  • Minnesota’s alleged “net income migration” (i.e., SOI income outflow from Minnesota minus SOI income inflow to Minnesota) has indeed grown in recent years, but the rate of growth has actually slowed since passage of the 2013 tax act. Minnesota’s alleged net income migration increased by $277 million from 2012 to 2013, but only by $190 million from 2013 to 2014—the first year that the 2013 tax increases were in effect. If net income migration was driven by the desire to avoid taxes, the rate of income loss should have accelerated—not slowed—with the implementation of the 2013 income tax increases.


  • Given that the net tax increase from the 2013 tax act fell almost exclusively on high-income households,* one would expect the highest rate of growth in income migration among these households—provided that the premise that taxes drive income migration is correct; however, the exact opposite is observed in the IRS SOI data. The rate of growth in alleged net income migration among households with adjusted gross income (AGI) of $200,000 or more slowed dramatically from 2013 to 2014 relative to from 2012 to 2013.
  • The CAE report uses the Tax Foundation’s State Business Tax Climate Index to assess the extent to which a state’s tax climate is conducive to business.† Less than one-twentieth of the variation in Minnesota’s alleged “net income migration” from 2013 to 2014 can be explained in terms of this index—a correlation that is not statistically significant. The absence of a significant relationship is consistent with the findings of a 2014 DOR report, which notes, “Most rigorous and peer-reviewed studies of tax migration fail to find any statistically significant effects of tax variables. This is true when studies are limited to seniors, and it is true for both estate taxes and income taxes.”
  • From 2013 to 2014, 85 percent of Minnesota’s alleged “net income migration” went to the twelve states with an average winter temperature above 40ᵒ F (compared to Minnesota’s frigid 12ᵒ F). This supports the premise that migration is fueled largely by retirees seeking to move to warm weather states. In fact, the growth in “income migration” in recent years—which does not coincide with changes in tax policy—does coincide with the increasing number of baby boomers reaching retirement age.
  • If we examine the period prior to 2011 (when the IRS SOI data was produced by the U.S. Census Bureau using a less complete set of returns), the years of the largest alleged net migration of income from Minnesota do not coincide with major tax increases. In fact, the second largest annual net “income migration” occurred from 1999 to 2000, immediately after the largest income tax rate reductions in over a decade. Furthermore, during the first decade of the millennium—when Minnesota taxes and total state and local government spending fell significantly relative to the national average—the degree of alleged income migration changed little.

Even if we accept the premise that the IRS SOI data is measuring income migration—which is clearly not the case—there are numerous patterns within this data set which contradict the assertion that the alleged income migration is driven by the level of state and local taxes. These facts—combined with new data from DOR showing an increase in high income filers in the year following the 2013 tax increase—should be sufficient to cast strong doubt upon, if not openly disprove, the assertion that income is migrating from Minnesota as a result of taxes.



*An analysis of DOR data summarized in a February 2016 North Star article showed that the 2013 tax act likely reduced state and local taxes for a majority of middle-income households. (Additional analysis of the 2014 tax acts shows that the combined effects of the 2013 and 2014 tax acts were a likely reduction in taxes for the majority of all Minnesotans with incomes below $202,000.) This being the case, all of the alleged tax flight resulting from the 2013 tax act should have occurred among high income households if the CAE’s premise that migration is driven largely by taxes is correct.

 TF’s State Business Tax Climate Index (SBTCI) is skewed against states that rely heavily on progressive sources of revenue, such as the individual income tax. In Grading Places, economist Peter Fisher contends that the SBTCI authors provide no evidence that the indicators they choose actually measure business competitiveness and further argues that the SBTCI is largely arbitrary and bears little relationship to what businesses actually pay in taxes. However, insofar as the SBTCI was chosen in the CAE report to gauge states that are “creating the jobs and opportunities that attract residents,” it is fair to assume that the SBTCI is an indicator of tax policies the CAE considers desirable.