Perhaps because of an indoctrination in identity politics, many in the media are obsessed with viewing the results of the 2016 election through a racial lens. Exit polls (I know, I know) suggest Donald Trump improved on Mitt Romney’s margins with all ethnic groups. In fact, he seems to have received a smaller absolute percentage of the “white” vote than Mitt Romney.
Even as pundits fixate on racial dynamics, the changes in income and education for partisan preferences might be far more revealing. According to national exit polls, there was a decisive swing of voters without a college degree toward the Republican nominee and an appreciable movement away from him by college-degreed voters. In 2012, Mitt Romney lost high-school dropouts by 29 points and high-school graduates by 3; those with some college but without a college degree he lost by a single point. Trump, however, won those who had a high school degree or less by 6 points, and those with some college he won by 9 points. While Trump substantially improved with those without a college degree, he lost ground with college graduates, with a margin that was 8 points worse than Romney’s with college graduates and those who held postgraduate degrees (he lost the latter group by 21 points).
Income tells a similar story: Trump still lost those making under $50,000 a year, and he only fought Clinton to an essential draw among the middle and upper classes. However, compared to Mitt Romney, he did much better with the poor and working class and worse with the wealthy. Romney lost those making under $30,000 by 28 points; Trump lost them only by 12. Trump did 6 points better than Romney among those making between $30,000 and $50,000, losing them by only 9 instead of 15. Conversely, he lost about 9 points relative to Romney among those making over $100,000 annually.
Polling has had some errors this year, and the early rounds of exit polling presented a mistaken picture of the electorate. Still, county-level electoral data confirm these trends. Working-class counties across the country--especially in the Midwest--swung toward the GOP relative to 2012. Meanwhile, many wealthy inner-ring suburban counties trended more Democratic. This trend was noticeable both around coastal metropolises like Washington, DC and in middle America; Romney won Johnson County, Kansas, the wealthiest county in the state and part of the Kansas City metropolitan area, by 17 points in 2012, but Trump only eked it out by 2. The New York Times has a handy map showing the swing of each county in the Democratic or Republican direction; most counties across the country swung more Republican, but, often, the counties that swung more Democratic are among the wealthiest in a given state. In Minnesota, for example, the only counties that swung more Democratic were those in Minneapolis-St. Paul area.
In many of the counties on the southern border, Trump received more votes than Romney. For example, he improved upon Romney’s margins and absolute vote total (getting 19 percent of the vote rather than 13) in Texas’s Starr County, where Hispanics constitute over 95 percent of the population. That may be evidence that Trump did indeed improve upon Romney’s numbers with non-”white” voters.
In many of the counties on the southern border, Trump received more votes than Romney. For example, he improved upon Romney’s margins and absolute vote total (getting 19 percent of the vote rather than 13) in Texas’s Starr County, where Hispanics constitute over 95 percent of the population. That may be evidence that Trump did indeed improve upon Romney’s numbers with non-”white” voters.
State results accentuate and at times exaggerate this national trend. The Rust Belt was crucial for Trump’s victory, and huge swings can be seen in these states. In Wisconsin, Trump did over 20 points better than Romney’s margin with those who had a high school degree or less and 15 points better with those with some college, but his margin with postgraduate voters fell by 18 points compared to Romney; he lost that demographic 26-69. He did much better than Romney with workers making under $50,000 annually, losing them only by 4 points instead of 25. However, he lost those making over $100,000 annually by 2 points rather than win them by 20, as Romney did. (See this interesting thread by Jeff Blehar on the county-level results of Wisconsin. That county-level analysis tracks in many ways with what the exit polls suggest.)
In Pennsylvania in 2012, Romney lost high-school graduates by 21 points, and those with some college by 4 points. Trump won them by about 13 points and 3 points, respectively. College graduates and those with graduate degrees swung against Trump relative to Romney, who won college grads by 16 points and lost postgraduates by 8 points. Trump, however, only won college graduates by 4 points and lost postgraduates by 22 points. By massively improving his standing with voters without a college degree, Trump more than made up for his slippage with college grads and advanced degree holders. Trump also considerably improved upon Romney’s performance with mid- and lower-income voters. While Romney lost voters making under $50,000 annually by 36 points, Trump only lost them by 12 points.
Michigan in many ways follows the trends of Pennsylvania and Wisconsin: Trump gained with lower-income voters and with voters with less formal schooling but also lost ground with upper-income and degreed voters. In Ohio, he did better overall than Romney in many categories. Tracking union voters sheds light on this dynamic. In Michigan, Ohio, and Wisconsin, Trump did at least 20 points better than Romney among union voters, which helped him secure those states.
These trends could also be seen in states that Trump did not win. For instance, in Virginia, Trump did better in 2016 than Romney did among lower-income voters, but slipped with the middle and upper classes. The educational swing in New Hampshire caused the Granite State to become more educationally polarized than racially polarized; the gap between “white” and “non-white” voters was smaller than the gap between high-school-only voters, who backed Trump, and postgraduates, who overwhelmingly didn’t.
Again, exit polls are imperfect vehicles, but the magnitude of those swings suggests that there was likely some shift in the electorate based on income and formal schooling. Peggy Noonan has proposed that one way to view the 2016 election is a clash between the “protected” and the “unprotected.” In Coming Apart, the social scientist Charles Murray noted a similar divide between the secure Belmont, a fictional embodiment of white-collar wealth, and the more precarious Fishtown, an imagined locus of blue-collar struggles. It seems as though Trump drew some of the traditionally Democratic “unprotected” to him while repelling some “protected” voters who usually lean Republican.
These results suggest great opportunities and equally great dangers for Republicans. If Donald Trump and the GOP Congress are able to govern well (by avoiding international crises, behaving ethically, and successfully responding to economic anxiety), it would seem possible that they could expand their majorities in 2018 and 2020 by winning back the college-educated and upper-income voters who used to favor them more. Trump’s 47 percent, then, could be a foundation for a majority in 2020. In addition to locking down the states he has already won, a marginally better performance with Belmont voters could help Maine, New Hampshire, Minnesota (where Trump won voters making under $100,000 a year), and even potentially Colorado and Nevada tip into the Republican column.
If, however, Republicans disappoint the working class and prove the fears of many white-collar voters to be well-founded, they could be in for a massive electoral repudiation. Blue-collar voters could desert them while college-educated voters continue to shake their heads in disgust. Trump’s Electoral College majority seems robust, but, absent continued working-class enthusiasm, it could also prove tenuous.
A further difficulty might face Republicans hoping to build on Trump’s coalition: There might be such tensions between Belmont and Fishtown that any gains with one group would be offset by losses with the other. If that’s the case, expanding the current Republican majority could be troublesome. However, while there are conflicting interests (and ideologies) here, statesman-like behavior in the White House combined with prudence in legislation, which meets at least some of the demands of both groups, could establish some common ground.
With a base that is more economically anxious and more blue-collar than in the past, Republicans will need to keep worker-oriented policies at the front and center. Trade could prove an especially tricky topic; many embattled Republican senators campaigned against or were conflicted about TPP, and a majority or plurality of voters in key Rust Belt states agreed with the idea that international trade “takes away U.S. jobs.” The GOP will need to be very cautious about introducing new trade compacts and might need to embrace a more proactive case for trade reform. On immigration, Republicans could increase federal enforcement while working to implement changes to the legal immigration system so that it better encourages opportunity and fosters national belonging. Tax reform would be wise to focus on improving the condition of working families. The Affordable Care Act’s many deleterious effects could be rolled back and a more inclusive and flexible health-care reform could be passed. Regulatory reform could provide more opportunity for workers while also improving business vitality. As president, Donald Trump would be able to put a halt to the administrative culture war launched by the federal bureaucracy during the Obama years. Trump’s judicial appointments could also shore up his support with the traditional Republican base and with voters who view governing through judicial fiats as another embodiment of unaccountable political power.
In the aftermath of 2012, it seemed clear that Republicans needed to improve their standing with the working class. Trump's election may be a sign that they have made some headway in that goal. But it will take success in governing to solidify and to expand that coalition.
