The Exit Polls

With respect to the U.S. presidential election results, most of my social circle is still at the anger and denial stages of grief. Me, I’m a historian by training, so I tend to bury myself in trying to understand how and why it happened.

It’s familiar ground for me: I have considerable background in the rise of extreme right-wing movements in mid-20th-century Europe; understanding why and how the extreme right came to power is an important area of study in my former field.

I appreciate that activists on the ground right now won’t have much patience for that sort of analysis — they’re not interested in understanding the motives of people who voted for a racist, misogynist bigot — but I think that’s a mistake. If you want to fix the problem, and, you know, prevent it from happening again, you have to understand it first.

Let’s begin with the surprising and inexplicable fact that Trump performed far better among constituencies that he had absolutely no business doing so well in.

According to the exit polls — the New York Times has a marvellous infographic here — Trump held on to 42 percent of the women’s vote, only one point down from Romney in 2012. He actually won 53 percent of the white women’s vote (white women with a college education supported Clinton 51 to 45 percent). How the hell did that happen?

It gets weirder. Trump got 29 percent of the Latino vote — that’s up eight points from Romney in 2012. (Though Latino Decisions questions that number, pointing out that a large portion of the Latino electorate voted before Election Day.) He also got 8 percent of the African American vote (up 7) and 29 percent of the Asian American vote (up 11).

If the exit polls are accurate — and I don’t know how big an if that is — then Trump held his own among women voters despite repeated allegations of sexual assault and harassment (plus, you know, the pussy-grabbing), and improved his standing with Latino voters despite the build-the-wall rhetoric and calling Mexicans rapists.

This is, on its face, insane.

And if you think that’s crazy, dig this: a substantial number of Trump’s voters didn’t think very much of him.

Trustworthiness was a wash: about equal numbers said Clinton and Trump were untrustworthy (61 and 63 percent), and of those voters, about one in five voted for them anyway. We’re used to not trusting our politicians very much. But 63 percent of respondents also said that Trump didn’t have the temperament to serve effectively as president, and one in five of those respondents — 20 percent — voted for him anyway.1

One gets the impression that the Clinton campaign strategy to portray him — quite correctly, in my view — as temperamentally unfit for the presidency didn’t work very well. Not if 12.6 percent of the entire electorate agreed with their premise, and voted for him anyway.

That 12.6 percent was roughly six times what Clinton would have needed to win. Moving one or two percentage points from the Trump column to Clinton would have flipped Florida, Michigan, North Carolina, Pennsylvania and Wisconsin to Clinton, and we wouldn’t be having this discussion about how she could win the popular vote by so much and still lose in the electoral college.

Many commentators point to racism, or at least the lack of discomfort in voting for a racist candidate. They’re almost certainly right. But that’s at least partially problematized by Trump’s small-but-not-insignificant minority support, plus the fact that the regions that swung to Trump this time had no trouble voting for Barack Obama in 2008 and 2012. Glenn Greenwald points to economic anxiety, which he describes as “inextricably linked” with racism, sexism and xenophobia:

[O]ne must confront the fact that not only was Barack Obama elected twice, but he is poised to leave office as a highly popular president: now viewed more positively than Reagan. America wasn’t any less racist and xenophobic in 2008 and 2012 than it is now. Even stalwart Democrats fond of casually branding their opponents as bigots are acknowledging that a far more complicated analysis is required to understand last night’s results. As the New York Times’s Nate Cohn put it: “Clinton suffered her biggest losses in the places where Obama was strongest among white voters. It’s not a simple racism story.” Matt Yglesias acknowledged that Obama’s high approval rating is inconsistent with depictions of the U.S. as a country “besotted with racism.”

People often talk about “racism/sexism/xenophobia” vs. “economic suffering” as if they are totally distinct dichotomies. Of course there are substantial elements of both in Trump’s voting base, but the two categories are inextricably linked: The more economic suffering people endure, the angrier and more bitter they get, the easier it is to direct their anger to scapegoats. Economic suffering often fuels ugly bigotry. It is true that many Trump voters are relatively well-off and many of the nation’s poorest voted for Clinton, but, as Michael Moore quite presciently warned, those portions of the country that have been most ravaged by free trade orgies and globalism — Pennsylvania, Ohio, Michigan, Iowa — were filled with rage and “see [Trump] as a chance to be the human Molotov cocktail that they’d like to throw into the system to blow it up.” Those are the places that were decisive in Trump’s victory.

Countering the economic argument (and in support of the racism argument), people point to exit poll data that show Clinton winning voters making less than $50,000 per year. (It’s actually pretty close above $50K — Trump never does better than 50 percent.)

But that’s not how economic anxiety works. Economic anxiety is greatest among the people who are doing better than that — a bit better, but they think they’re not doing better enough, and they’re worried that they could lose what they have. It’s not a paradox for people making a bit more money to be afraid for their economic circumstances.

And indeed, we find that among respondents who say that their family financial situation is worse today than it was four years ago (27 percent of respondents), Trump won bigly — 78 to 19 percent. Among respondents who thought that the next generation of Americans would have it worse than today (34 percent of respondents), Trump won 63 to 31 percent. If you were doing well, better off than you were four years ago, or thought things would get better in the future, you almost certainly voted for Clinton. Trump won the pessimist vote.

Small wonder then that of the 39 percent of respondents who said that the most important candidate quality was that they “can bring needed change,” 83 percent voted for Trump.

So. Economic anxiety facilitates bigotry and makes a Trump victory possible.

Some voters — just enough voters to make the difference — made the decision to accept the bigotry and vote their economic anxiety.

This is not to excuse or explain away the bigotry. But if you think the rise of the Nazis had nothing to do with economic anxiety, you don’t know your history.

John Scalzi makes the analogy of subscribing to a cable package: you really want HBO, but you also get Cinemax. Regardless of whether you really only wanted HBO, you’re still a Cinemax subscriber. In voting their economic anxiety, they also voted for racism, sexism, homophobia, and harassment.

That’s not the sort of thing that makes Clinton supporters feel very favourably toward them at the moment.

That’s understandable in the moment, but a mistake in the long run, because some of these voters — some — are voters the Democrats had before, and could get back. And they’ll need them back, if they want to regain the White House. You can’t get the racist vote. You don’t want the racist vote. But you only need a few percentage points — the gettable voters who broke for Trump in the end.

(I’m leaving aside the question of depressed voter turnout and minority voter suppression, mostly because it’s not something exit poll data can deal with.)

Writing in Slate, Helaine Olen argues that Clinton blew the campaign in precisely the same way her husband did not in 1992: by making the campaign about Trump’s character and not the economy — whose improved fortunes are not evenly distributed. What may have done her in was not her email server, but her private speeches and Wall Street ties.

Trump, on the other hand, was incredibly canny. He said four words that I suspect resonated with this part of the electorate — “The system is rigged” — and four more words that neutralized his billionaire status and, incredibly, suggested he empathized with them: “Believe me, I know.”

And that’s one way of explaining how we got here. How the pussy-grabber and the wall-builder managed to get support from women and minority voters. Because this campaign was more about the economy than the Clinton campaign realized, to its — and our — chagrin.

  1. Forty-three percent said the same thing about Clinton; only five percent of those voted for her anyway.