You may have missed it in the overwhelming swirl of information in the month leading up to the election, but back in mid-October, I wrote a whole article highlighting a minor detail from one poll’s crosstabs (from Public Religion Research Institute) that I found very compelling. And it’s one that, in retrospect, may have had more explanatory power than any of us thought, at the time.
White voters who still live in the community in which they were raised are supporting Trump over Clinton by 26 percentage points (57% vs. 31%, respectively). Trump also has an advantage over Clinton among white voters who live within a 2-hour drive from their hometown (50% vs. 41%, respectively). However, among white voters who live farther away from their hometown, Clinton leads Trump (46% vs. 40%, respectively).
Now that we’ve seen the post-election map, those places where people are most likeliest to have stayed in the same place where they were raised, are also the places that seemed to trend the hardest in Trump’s direction: rural areas across the Midwest where the population is slowly falling, populated almost exclusively by white people, most of whom aren't college-educated.
Many pundits have focused on the white working-class aspect of these places. But there’s another dimension that goes beyond the racialized economic concerns of the residents of these places, that instead wonders about the larger mindset of those who grew up in and then left those places, versus the mindset of those who chose to stay. There’s, of course, a chicken-and-egg problem here, in that education plays a key role in mobility; if you seek higher education, you’re likely to need to leave your small town, and then to get a job that takes advantage of your higher education, you’re likely to end up in a metro area.
But as Josh Barro pointed out, there may simply be a big personality difference between those who stay and those who leave, whether it’s optimism vs. pessimism, or openness vs. intolerance, or having some agency over one’s life instead of simply stewing in your resentments—which, if you think about it, is really what the big themes of this year's election were, more so than any specific set of policies.
You’re probably puzzling over the red/blue map at the top of this article; it’s not really intended to represent a likely outcome of the next presidential election. It could, however, reflect an exaggeration of future trends, in terms of which states are moving more toward the Democrats or away from them. It’s a map of which states have above-average rates of mobility (in blue) and below-average rates (in red). (While the Census Bureau doesn’t ask people whether they live in the "community where they were raised," it does ask if they still live in the state in which they were born.) In other words, whether you think of it as simply a map of where people are moving to take advantage of economic opportunities or a map of a broader optimism, you can think of it as map of dynamic America vs. stagnant America.
What’s most striking is if you list all the states in order of what percentage of residents were born in the same state, and look at the top of the list (i.e. the states where the most people haven’t left home), you see two things: one is some Deep South states where the electorate is already so racially polarized that there’s little swinging from one election to the next. The other is the Midwestern/Rust Belt states with large blue-collar white populations that turned against Hillary Clinton this year (some where we had some advance warning, some that surprised us at the very end).
Louisiana 78.3 percent
Michigan 76.6
Ohio 75.2
Pennsylvania 72.9
Mississippi 71.5
Wisconsin 71.4
Iowa 71.1
Alabama 70.1
West Virginia 69.7
Kentucky 69.6
By contrast, the states with the lowest rates are ones that have decidedly moved in the Democrats’ direction in recent years: Nevada (25.8), Florida (35.9), Arizona (39.2), Alaska (41.8), New Hampshire (42.1), and Colorado (42.7), for instance.
The few states on the map that surprised me were Texas and New York; these are states with booming cities and lots of immigrants that you’d probably think of as “dynamic” on the whole. However, they’re only slightly above the national average of 58.5 percent (Texas at 59.7 and New York at 63.1), and you have to remember that there are still white working-class areas in, say, east Texas and upstate New York.
Along those same lines, there’s a big variation even within those Rust Belt states. The tale of the county-level election results makes clear that while Clinton cleaned up in the urban, suburban, and college-town counties in those states, it wasn’t with either an overwhelming-enough percentage of the vote or with large-enough turnout to overcome the surge of unexpected Trump votes in the rural, blue-collar parts of those same states. But, for comparison purposes, look at the rates of “people living in the state where they were born” in some of Clinton’s best counties in those states …
Washtenaw Co., Michigan (Ann Arbor): 60.1 percent
Franklin Co., Ohio (Columbus): 66.6 percent
Montgomery Co., Pennsylvania (Philly suburbs): 69.7 percent
… which are still higher than the national average, but much lower than the statewide averages, and compare them to some of the less populous counties that most notoriously surged in Trump’s direction:
Bay County, Michigan (Bay City): 88.6 percent
Marion County, Ohio (Marion): 83.5 percent
Fayette County, Pennsylvania (Pittsburgh exurbs): 85.6 percent