Many of us didn’t believe Donald Trump could win the presidency, because we didn’t think he could win Michigan, Pennsylvania, and Wisconsin.
Yet Trump swung these states red in 2016, largely on the strength of outperforming among blue-collar workers. Trump, according to exit polls, got about 500,000 votes from union households in Michigan compared to Romney’s 430,000 4 years earlier. That difference is far greater than his 12,000-vote win in Michigan.
Many observers hailed this as a realignment: Trump had swung the working class into the Republican column. The counterweight was that he had pushed the highly educated suburbs into the Democratic column.
Is this analysis true? A closer look at election results tells a more complicated story.
For one thing, the shift of the suburbs towards the Democrats — and the working class towards the Republicans — predates Trump by decades. Montgomery County, Maryland, Westchester County, New York, the collar counties around Philadelphia were all turning bluer for years.
Meanwhile, Republicans were steadily picking up workers without college degrees. Republicans’ share of high school dropouts climbed from 28 percent in 1992 to 49 percent in 2004 before Obama took back a big chunk. Those who graduated high school didn’t go to college went from 36 percent GOP 1992 to majority Republican in 2004.
Secondly, there’s good evidence that the Rust Belt voters who swung to Trump didn’t actually become Republicans. They merely became Trump voters.
Look at the 2018 results: Michigan and Wisconsin both replaced their Republican governors with Democratic governors. Both states reelected Democratic senators. Pennsylvania reelected Democrats as both governor and senator. Ohio re-elected a Democrat to the U.S. Senate.
What happened? One explanation is that a lot of the voters who swung from Obama to Trump then swung back to the Democrats.
In Michigan, Democratic governor-elect Gretchen Whitmer outperformed Hillary Clinton in every single county. In some counties, the Clinton-to-Whitmer swing was larger than in others. What best predicted bigger swings to Whitmer in 2018? Bigger swings to Trump in 2016.
Whitmer’s largest swing was in rural Kalkaska County in the northern end of Michigan’s lower peninsula. Kalkaska is a working-class white county. It’s 96.8 percent white, and less than 1 in 7 adults over age 25 have finished college. It swung 24 points from Romney to Trump, and then a full 24 points back to Whitmer. Mathematically, then, every new Republican that Trump picked up in Kalkaska went back to the Democrats in the midterm elections.
Twelve Michigan counties flipped from voting for Obama in 2012 to voting for Trump in 2016. Call them the “pivot counties.” Half of those pivot counties flipped back to Democrats in 2018. Almost all of them gave back a majority of the GOP gains.
Macomb County — the most populous pivot county — swung 16 points to Trump, and then 15 points to Whitmer. Bay County swung 19 points to Trump, and then 16 back to Whitmer.
One way to read these numbers is that many of the new voters Trump picked up were not converted to the religion of Republicanism. This isn’t hard to imagine. Trump won blue-collar voters by running on a platform very different from that of the Republican establishment. Critical of free trade, opposed to most immigration, lukewarm towards free-enterprise, and unfamiliar with conservative thought, Trump was perfectly positioned to win over voters who aren’t ideologically aligned with, say, Paul Ryan.
But we don’t have quite enough data to reach that conclusion. For one thing, there are no real constants in these metrics. Obama, as the exit poll data on educational attainment suggest, was better than most Democrats at winning over blue-collar workers. Romney was particularly ill-suited for the working class. Who’s the exception, and who’s the rule is hard to judge.
Then there’s the fact that turnout constantly grows, shrinks, and shifts. The swings in Macomb and Bay Counties may be less about voters voting Obama-Trump-Whitmer, or Obama-Trump-Tony Evers in Wisconsin, and more about liberal voters showing up, then staying home, then showing up again.
In Wisconsin, for instance, turnout percentage fell in 69 of 72 counties. In some counties it fell more rapidly. Given Hillary’s famous coolness towards Wisconsin, the Trump swing might best be explained by Democrats who stayed home.
If we look back at Michigan, the swinging behavior might best be explained this way:
Obama inspired working-class liberals. Hillary didn’t. Trump won a decent number of working-class independents. Trump has fired up the Left again, and so the working-class liberals turned out in 2018 in numbers closer to how they turned out for Obama. At the same time, the working-class independents inspired by Trump to vote in 2016 stayed home in 2018 because they had no interest in backing regular old Republicans.
That’s just a hypothesis, but it fits the available data. The upshot: Trump has tapped into something real in the Rust Belt, but he hasn’t brought about a positive realignment for the Republican Party.
