CBPP Misses the Mark on Migration
August 26, 2014
The Center on Budget and Policy Priorities’ Michael Mazerov has a blog post out today covering the same Upshot blog migration tool we covered recently. Mazerov points out that states with no income tax appear to have done a worse job holding onto migrants than other states.
Unfortunately, what Mazerov actually discovered is that some states have just have more people than others. There’s a very basic statistical problem in the data he presents, but it’s an easy problem to miss.
To see that problem clearly, it may be better to think about people who stay in a state, rather than out-migrants. A person born in a big state, like Texas or California – just like anyone else – is likely to find plentiful opportunities in Texas or California. Therefore, she is likely to live in Texas or California – not necessarily because Texas and California are superior, but because they just happen to be very large and populous.
On the other hand, Rhode Island, as readers may notice, is small. Wyoming, North Dakota, Alaska, and most of Mazerov’s high-leaver states all have low populations as well. Simple statistical gravity suggests people in those states will find jobs elsewhere. For a person born in Wyoming, most potential employers are outside of her home state. This is not because Wyoming is inferior – it is just lightly-populated. If we divided California into sixty-six “states,” each the population of Wyoming, we would almost certainly find people leaving their home “states” in this divided California much more often. The data Mazerov presents doesn’t tell us much about tax policy: it mostly tells us about population.
That explains why Florida, California, and Texas all have some of the least “leavers” despite having radically different net migration flows for over 20 years. Meanwhile, Wyoming, North Dakota, Alaska, Montana, Rhode Island, Idaho, and others all share similarly large amounts of leavers, even though their annual net migration rates are also very different.
Of course, Mazerov’s chart only shows half the data. He shows how many people left, but not how many people moved into each state. It’s hard to talk about migration when you only look at part of the story. Indeed, that’s the same point we made in our writing on migration earlier this year. To understand how migration works, it’s important to look at all the data, and consider the options actually faced by migrants. In this case, Mazerov did not fully consider the basic distribution of employment and lifestyle opportunities around the nation, which led to the production of a chart that mostly just mirrors state level population trends.