Amid a spate of brutal hate crimes against people of color – with Muslim women shedding their hijabs to avoid random attacks, and the word "nigger" making an ugly resurgence in our discourse – the political press appears to have coalesced around the idea that we really need to understand the pain felt by the white people who elected Donald Trump.
It's clear that white working-class voters in the Rust Belt provided Trump with a razor-thin margin of victory in the Electoral College, despite losing the popular vote by historic margins. The data show that Trump won a number of Midwestern counties with lots of blue-collar whites that went for Obama in 2012, in some cases by large margins.
But how we interpret that data has important ramifications for how the Democratic Party moves forward. If, as a New York Times headline blares, Trump's win was in large part a result of non-college educated white voters who supported Obama in 2012 defecting to the Republicans – perhaps for good – then the logical conclusion is that Democrats have to reach out to this group specifically or face the prospect of future losses. And that means speaking not only to their economic anxiety, but also appealing to their cultural and social grievances. It might mean, for example, moderating the party's support for gun safety measures, which are an important wedge issue for many rural white people in those key states Trump flipped. The last time the party decided to chase blue-collar "Reagan Democrats," it resulted in Bill Clinton's push for welfare reform.
If, on the other hand, Trump energized just enough Republican-leaners who stayed home in 2012, and Hillary Clinton failed to turn out just enough Democratic partisans, then we can attribute this disaster to factors that aren't specific to this group. It may be that she was an unpopular candidate who faced a perfect storm of media coverage tainted by a tendency toward false equivalence, hackers releasing her campaign's internal emails, a clumsy intervention by FBI Director James Comey and latent misogyny – all of that while running against a celebrity who dominated nearly every news cycle. If that's the case, then the solution, whatever it is, should be the same for blue-collar white Democrats as it is for Democrats in general – running a better candidate who's more focused on a progressive economic agenda, for instance – and we shouldn't indulge in a lot of handwringing over this one group of white people.
Based on what we now know, there's good reason to believe this last analysis is the correct one. According to the exit polls, Clinton underperformed Barack Obama's 2012 results among not only non-college educated whites, but also white men; black men and women; Hispanic men and women; Asian men and women; men and women of other races; every age group except voters over 65; liberals, moderates and conservatives; Protestants, Catholics, adherents of other religions and those who claim no religious affiliation; married men and unmarried men and women; union and non-union households; self-identified Democrats; straight people; people who think undocumented immigrants should be given legal status; and people who think the country is going in the right direction. In that sense, the commentariat's intense focus on non-college whites already seems a bit odd.
It's true that in 2016, non-college whites swung to the GOP by a 15-point margin relative to 2012. But Clinton underperformed Obama among voters of all races who make less than $30,000 per year by an identical margin. If the pundits were churning out hundreds of columns about how the Dems need to win back low-income voters, it would likely have a more salutary effect on Democratic policy.
As for whether these were mostly former Obama voters, it's important to understand that we're talking about a group that tends to turn out in low numbers. In 2012, only one in four high school grads voted. That was true of fewer than 10 percent of those who didn't complete high school, and around four in 10 of those with some college experience. So in every one of those counties that flipped from Obama to Trump, it's safe to assume that there was a much larger pool of non-voters and Romney supporters for Trump to draw from than Obama voters. Occam's Razor would suggest that some combination of Trump's celebrity, his image as a populist outsider and his naked appeal to white racial grievances caused a significant number of people who had given up on the system and not bothered to vote in past elections to give Trump a shot.
The polls provide some circumstantial evidence for this conclusion. The consensus view is that public opinion surveys missed these voters. And those polls relied on "likely voter screens" – a series of questions most pollsters use to determine who is likely to turn out on Election Day. Obama 2012 voters switching to Trump would slip through these screens and be counted in the polls; those who didn't vote in recent elections would not.
The conventional wisdom that non-college whites switched parties in large numbers also flies in the face of what we know about partisanship in our increasingly polarized society. While lots of voters claim to be independents, research conducted by Michigan State University political scientist Corwin Smidt found that the number of true "floating voters" – people who can see themselves voting for a candidate from either party – is at an all-time low. According to Smidt, a low-information voter today is going to be as aware of the differences between the two major parties as a well-informed voter was in the 1970s, and those who identify as independents are more stable in their support for one party or the other than "strong partisans" were at that time. It's another reason to believe that Trump got Republican-leaning voters to the polls, including those who sat out 2012, rather than winning over white working-class Democrats. Nobody assumes Trump's gains among Latinos or blacks mark a permanent realignment, and there's no obvious reason to believe that's true for white working-class Dems.
While there are significant regional differences in white working-class voters' partisan leanings, Democrats have long struggled to win a majority of this group nationwide. That's not new. What was new in 2016 was a massive education gap opening up among voters of all ethnicities. In 2012, there was virtually no difference between the preferences of voters according to their levels of educational attainment. This year, those with a college degree went for Clinton by a nine-point margin, while voters without one favored Trump by eight points. This was by far the largest gap recorded in exit polls since 1980.
That education gap was wider among whites than voters of other races, but that can probably be explained by the fact that a college education makes white people less responsive to racialized, "dog-whistle" politics. (Michigan State University researchers Adam Enders and Steven Smallpage found that racial prejudice, rather than economic populism or desire for an authoritarian strong man, "predicted support for Trump over Clinton.")
Any effort to court voters who are animated by racial grievance wouldn't just be morally dubious, but would also risk alienating the fastest-growing groups within the Democratic coalition. Non-college-educated whites represent a demographic that's in decline as college graduation rates rise and the electorate becomes more diverse. They may have broken heavily for Trump, but their overall share of the electorate was down this year compared to 2012.
Democrats should certainly focus on a progressive economic agenda – both because it's good policy in our new Gilded Age, and because there's evidence that it spurs enthusiasm among their entire coalition, including with non-college whites.
And even if more granular data than the exit polls can provide eventually supports the notion that lots of blue-collar whites who voted for Obama in 2012 backed Trump this time, the party still shouldn't do anything differently to court them. After all, Trump's victory was a black swan event. Hopefully, we'll never have another presidential contest pitting a totally unqualified, right-populist TV star against an unpopular Democrat running under the weight of an FBI investigation while hackers publish her campaign's internal emails. And if we do, we should probably all just pack up and move to Canada anyway.
Protests against Donald Trump broke out in major cities across the country after Election Day. Watch here.