I've written a few times about the evidence of migration in or out of Minnesota due to its rather rude treatment of high-income earners. I have been gathering data recently on the moving vans, and a student and I are at work exploring that line of research again. We will have results we hope in April.

But meanwhile, in a new study released yesterday by the Freedom Foundation of Minnesota, two researchers looked at the IRS' data on income tax filings and returned a strong conclusion:

From 2002-2009 Minnesota lost an estimated 54,113 residents to other states, according to the new report, Minnesota�s Out-Migration Compounds State Budget Woes. These out-migrants also take their incomes with them. Between 1995 and 2007, the total amount of income leaving the state was at least $3,698,692,000 on which state and local governments would have collected an estimated $423,317,000 in additional taxes.

For example, in 2007 -- the last year of their study, a net 4,428 taxpaying units left the state, and took $378,757,000 in AGI with them. By aggregating up the thirteen years of study they find that a total of $2.548 billion in additional taxes would have been collected. That's of course over 13 years, a period in which we would spend maybe 80 times that? I would have liked that number put in context, just as the income data should be put in context of state personal income ($216,436,888,000 in 2007, to put it in the same context the FFM study does.)

What caught my eye as well was their ability to identify where the taxpayers moved to. The top five destinations of out-migrants were Arizona, Florida, Colorado, California and Texas. Four of those places are very warm. We talk about the low taxes of South Dakota, though on net 1,322 taxpaying units moved TO Minnesota. But on net more AGI left than came. What was missing here was an attempt to tease out the effects of other factors they identified like weather or cost of housing. The study shows these factors as important, but doesn't get relative importance of these additional factors. That requires a regression analysis, which that study chose not to do.

But this is a very interesting and worthy study. It uses actual tax return data rather than a survey or the loose proxy of moving vans. It can measure income flows separately from people flows. And it fits the theory that people are sensitive to the cost of government.

Labels: economics, Minnesota, taxes