Data show that a 'whitelash' did not elect Trump
The morning after the election, Van Jones offered his thoughts on the outcome:
"This was a whitelash against a changing country. It was whitelash against a black president in part. And that's the part where the pain comes."
Jones spoke honestly, from a place of sincere emotion, and I have a great deal of respect for that. His thesis certainly appealed to a lot of people. Given that Trump began his campaign with racist rhetoric and never really stopped, it also makes quite a bit of sense on the surface. There’s one problem: The numbers say it doesn’t hold water.
Yes, Hillary Clinton won the popular vote by a margin now approaching two million. That’s very important to remember going forward, but not especially helpful in terms of exploring why Trump did as well as he did. Let’s also leave aside the impact of James Comey’s completely inappropriate actions. Even if, as it appears true, Comey (the FBI director, --ed)) was enough to swing the election to Trump because of the tight margins in three key states, it doesn’t change what we can learn from the election in terms of race, income, and education.
From the national exit polls, here are the numbers that disprove the whitelash thesis: Trump did a slim 1 percent better among whites than Mitt Romney did four years ago. Were some whites drawn to Trump’s side by racism? Absolutely. But he appears to have lost pretty much an equal amount among those whites disgusted by it.
Furthermore, Trump improved over Romney by much more among every non-white ethno-racial group large enough to measure. He improved by 7% among blacks, 8% among Latinos, and 11% among Asian voters. Along similar lines, an exit poll conducted by the Council on American-Islamic Relations (CAIR) found that Trump received 13% of the Muslim vote. That doesn't sound like much, but it represents almost twice the percentage Romney won. No whitelash there.
Overall, turnout looks like it will come in at just about the same percentage of the eligible voter population as we saw in 2012. However, as Northern Ohio University political science professor Robert Alexander explained, “You saw turnout spike in more rural counties. If you take a look at a lot of the larger cities you did see depressed turnout there. It certainly was more consequential for Hillary Clinton than it was for Trump.”
Despite the more heavily rural voting population compared to 2012, Trump didn’t do significantly better than did Romney among whites overall. Of course, given that whites are about two-thirds of the voting population, gaining 1 percent among whites is important, but the gain of about 8 percent overall in the one-third of the voter population that is not white adds up to more votes.
Yes, these exit polls could be off by a couple of percent, but remember, the national polls weren’t off by that much. For example, Five-Thirty-Eight predicted a popular vote win for Hillary of 3.6 percent, and it looks like she’ll win the popular vote by close to 2 percent. That’s actually a better performance than the average polling miss of 2.0 percent in the twelve presidential elections before this one. In 2012, for example, the national polls were off by 2.7 percent, but no one noticed because all that happened was that Obama won by more than expected. So, if you reject exit polls this time, you have to always reject them, which would mean we’d know very little about demographics and voting. Either way, they’d have to have been off by a ton for this election to represent a whitelash.
On education, Trump gained significantly over Romney among all voters without a college degree, and Clinton gained significantly over Obama among voters with a degree. Looking at race and education combined tells the fuller story. Trump improved over Romney by 14 points among whites without a college degree, while Hillary improved over Obama in 2012 by 10 points among whites with a degree. Overall, Trump did 16-17 points better among whites without a degree than those with one. Among voters of color, however, non-degree holders were actually a bit stronger for Hillary than degree holders. So education mattered, but much more so among whites.
How about income? Trump improved over Romney by the biggest amount among the people helped most by Democratic policies, i.e., the poorest Americans: by 16 percent among those earning less than $30,000, and by 6 percent among those making $30,000-$50,000. Clinton, on the other hand, improved over Obama by 2 percent among those making $50-100K, and by 9 percent among those making $100-200K. This is clear and striking evidence that the election results were determined much more by class than by race for white voters taken as a whole.
Finally, although my focus is whether the whitelash theory was accurate, let’s talk about gender as well. Trump ran five points stronger among men than Romney did, whereas Clinton improved over Obama’s performance among women four years ago by only one point. Such a result, despite women having the opportunity to vote for the first woman president—not to mention against an opponent who bragged that he could get away with committing sexual assault because he’s a “star”—has to count as a colossal disappointment.