More on Minnesota’s out-migration

My op ed in Monday’s Star Tribune, ‘Migration out of Minnesota is on the rise’, generated more comment than I had expected. A former MPR journalist branded it racist, a point an anonymous phone caller also made in somewhat more colorful language. The Star Tribune also carried a response on its letters page. Sadly, my respondent evidently hadn’t read the article all the way through before responding.

Yes, I did provide evidence that taxes and migration are linked

I’ll reply to the first paragraph more fully tomorrow. The second reads:

John Phelan, an economist with the Center of the American Experiment, believes that our taxes are too high and provides evidence that Minnesotans earning $50,000 or more are leaving the state (“Migration out of our state has begun again,” Opinion Exchange, Feb. 3). Unfortunately for his argument, he provides zero evidence that the two issues are connected. In fact, studies show that state income tax rates have little correlation to migration patterns. People move for a variety of reasons including jobs, education, housing and quality of life.

In fact, the penultimate paragraph of my oped contained the following:

The large and growing body of evidence on the effects of taxation on migration was summarized recently by economists Henrik Kleven, Camille Landais, Mathilde Muñoz and Stefanie Stantcheva. They found “there is growing evidence that taxes can affect the geographic location of people both within and across countries.”

That paper – titled ‘Taxation and Migration: Evidence and Policy Implications’ – is freely and publicly available on the National Bureau of Economic Research’s website. My respondent might not like its findings, but to say that I provided “zero evidence” is simply false. It is also false to say that “studies show that state income tax rates have little correlation to migration patterns”

The IRS data does track income

The next objections reads:

Phelan also assumes that when a person leaves Minnesota, he or she takes the job and the income with them. In actuality, most jobs stay here and are filled by someone else. One would think that an economist would know that.

My respondent is not the first person to make this argument. My former colleague, Peter Nelson, dealt with it back in 2016, when we did our first report using IRS data:

Objection #1: IRS data do not measure income migration.  It is not accurate to compare the total incomes of people who moved in and out of Minnesota in a given year and describe the net difference as the loss of income from the state.

The IRS has itself used the difference between the inflow and outflow of income to show a net change in income for a region.  When they first introduced the income data they wrote a paper demonstrating how the data can be used with various examples.  Looking at migration between 1992 and 1993 for the Northeast region, they found “a net loss of income to the region of, nearly $5 billion.” (Kozeilec 1995).

Also, academic studies I’ve reviewed that use the IRS income data for migration research purposes discusses the income flows in terms of “gains” and “losses” of income to a state or county. (See references below.)  These studies do find the difference between the inflow and the outflow of income very meaningful.  Studies specifically refer to this difference as “net migration” just as the IRS does.  In fact, a couple studies use this net migration figure to estimate the impact of migration on per capita income.  (Plane 1999; and Cromartie and Nord 1997.)  Moreover, the U.S. Census Bureau, according to the IRS, “relies heavily on these data to estimate the population between censuses and to estimate per capita income.” (Kozielec 1995.)

Objection #2: IRS data do not accurately measure the migration of income because when someone moves from one state to another the income they earned does not necessarily go with them.

It’s true that the income from a particular job usually does not move with someone as people usually move to take a new job.  But this does not mean the IRS data don’t provide a meaningful and accurate measure of income migration.

Some studies do issue a caveat explaining how the income tied to a job itself doesn’t move, rather it’s the income earner and their capacity to earn income that moves. (Plane 1999; Manson and Groop 2000)  No one is claiming the income generated by specific job moves.  Rather, it’s the income generated by a specific person working that moves.

The usage of the income flow data by the IRS, the Census and academics to assess net income flows and estimate per capita income changes creates a strong consensus that the IRS income flows provide an accurate measure of the change in the aggregate income base of a region from migration.

My article was about domestic migration, not refugees

The final two paragraphs are, possibly, the most curious of all:

When I looked at the website for the Center of the American Experiment, I found that it is a right-wing organization that advertises itself as “Minnesota’s Think Tank.” I find that to be a curious combination of arrogance and ignorance.

Perhaps that is why Phelan makes the casual but offensive assertion that folks earning $50,000 or more are Minnesota’s “highly productive citizens.” Minnesota has a history of accepting refugees from all over the world. These people have a history of being hardworking, law-abiding citizens who educate their kids. Usually within a generation, their offspring are teachers, law enforcement personnel, doctors, engineers and scientists. In a list of highly productive citizens, I would rank them near the top.

First, note that my respondent says that “Minnesota has a history of accepting refugees from all over the world”. No doubt, but my oped was specifically and solely about domestic migration – that from elsewhere in the United States – not international migration, which is the category refugees would fall into. My respondent is responding to a point I didn’t make. Along with the false point about providing “zero evidence”, this proves that his reading of the op ed was not what it should be.

Second, in economics ‘productivity’ has a reasonably clear meaning: it is the amount of output (goods or services, GDP) a worker produces in a given period. And, as I explained yesterday, the evidence shows that pay and productivity are linked: the higher your productivity, on average, the higher your pay. That is what I mean when I say that Minnesota’s higher paid workers – like “engineers and scientists” – are its higher productivity workers. If we are losing these folks, whatever their parentage or place of origin, that is bad economic news for the state. You can only find this “offensive” if you are ignorant of basic economics.

You always want people to engage with your research. You hope, when they do, that they will, at the very least, have actually read it before doing so. Sadly, my respondent evidently hadn’t.