For some reason, we seem to have gotten a fresh round of “this is why she lost” blather of late, so please allow me to weigh in–with some actual data, pulled from CNN’s exit polls. Tl;dr version: Dudes. Primarily white. Primarily young. Primarily without a college degree. Primarily voting for Johnson.
First off, let me be perfectly clear: The biggest factor in this and every other election is race. White folks, about 70% of the electorate, vote Republican by about a 60/38 split, whereas people of color vote Democrat by even more lopsided percentages. This has been the case pretty much forever, and there can be no discussion of WTF happened without acknowledging that reality. Religion is also a big deal: White Christians, who make up ~57% of the electorate, split about 65/35 for Republicans, but everyone else votes Democrat. Lastly, education also factors in, though again race tells part of the story: Whites with less education tend to be more conservative, while people of color don’t change their votes all that much relative to education.
Gender, however, is also an enormous factor, and as I’m going to illustrate here, it was a bigger factor this time than it has been in the recent past–bigger than anything else.
The question I set out to answer: Why did we elect a Black man by decent margins in 2008 and 2012, but the Democrat candidate lost the EC this year even with whites dropping their percentage of the electorate by two points from 2012? What changed so drastically in just four years that made people who voted for Obama decide not to vote for Clinton?
The easy answer, of course, is that it was just a quirk of the electoral college–which it was, let’s be honest. Winning the popular vote by ~3 million should have delivered the EC, and it ultimately came down to only about 80,000 votes, spread across three states (Michigan, Wisconsin, and Pennsylvania) previously thought to be comfortably Democrat. Most conventional-wisdom think pieces have blamed this on the “white working class” being fed up and liking the way Trump talked, but that’s actually not what happened. The white working class has trended Republican for decades. This is nothing new. Moreover, even though people making <$50k did drop about seven points in their support for the Democrat, they made up less of a percentage of the electorate, and people making >$100k increased in both electorate percentage and support for the Democrat. So while income may have been a slight factor, it definitely wasn’t the deciding one.
If you listen to some folks, they’ll argue that Clinton was simply less popular overall than Obama. And that is true. She dropped a couple of points relative to him across virtually all demographic categories. But guess what? Trump also dropped points relative to Romney. Both candidates were less popular than their party counterparts from four years ago, handing off a bit to third parties, so if we’re going to ignore all demographic breakdowns, this shouldn’t have been a factor.
How about age? Did that make any difference? To a degree, yes. People <40 dropped their support for the Democrat by a larger margin than they did for the Republican, and this was one of the places race actually didn’t help Clinton: White millennials dropped less than non-whites of that age. However, it should be pointed out that because the group in question is only about 13% of the electorate nationally, and considerably less than that in the white-heavy states in question, their overall impact on the results was negligible. Far more critical a factor related to age is gender, and I’ll tie those two together momentarily.
Let’s have a look at one of the spreadsheets I’ve been using to crunch these numbers:
In case looking at that gives you a headache, let me cover the highlights: The national numbers do show Clinton losing points across all groups, with most of those points ending up with third-party candidates. But it’s the state breakdowns that make the picture much, much clearer: The biggest drop among women of any race in the four states listed here is four points, and she actually gained points among some. The average drop for all women in these states: 1 point–the same as the national average. The average drop for all men? 8 points. Among those three states, men of all races who supported Obama in 2012 ditched her by significantly higher margins than did women. But where did those men run to? About half went to Trump, but another half went to third-party candidates, primarily Johnson, 60% of whose national support came from men (a wider gender split than even Trump had.)
This drop among men in these states is higher than the average drop among millennials (about 4.7 points) and the drop among non-college grads (6 points) and those making <$50k (7 points). Moreover, it should also be noted that the education difference also has a major gender factor. While we don’t have a 2012 breakdown of education by gender to compare, it’s interesting to note that white college-grad women favored Clinton by seven points, and only 5% voted third party. White college-grad men favored Trump by 14 points, and 8% voted third party. Among non-college grad whites, women favored Trump by 27 points, but men favored him by 48 points (their third-party support was roughly equal.) So again, education does make a difference, but the gender split is a far greater factor. Alas, we don’t have an age-by-gender report, but it’s safe to assume the percentage of the electorate and vote splits by gender were probably about the same as they were for all voters as a whole. Which, given the sharp declines in Democrat support among both men and millennials probably means that the lion’s share of those losses was coming from younger men.
Again, I want to reiterate that it’s unlikely that men of color were the tipping point, even though they did have significant drops, because they’re a much smaller percentage of the electorate, especially in the three upper-Midwest states that decided the election. It is possible that men of color had some effect in states with larger non-white populations: I also did the numbers in the above spreadsheet for NC and FL to get a better idea of what happened there, and there were indeed some pretty major drops, especially among Black men: 10 and 13 points, respectively. Black women also had larger drops there, too. But even at larger percentages of voters than you’d find in, say, Wisconsin, those groups are overall so much smaller, and so much more Democrat-heavy to begin with, that we simply can’t pin those losses there. A massive drop among a group with a much higher percentage of the electorate–white men–makes a far bigger difference.
So, now that we’ve determined that young white men dropping their support for the Democrat is what flipped these key states, we can start to ask why. And, going by the split in where those votes went, I think it’s a pretty easy call. Did some get swayed by Trump’s rhetoric? Sure. But half or more of those lost votes went to third parties, overwhelmingly Johnson, who did twice as well among millennials than he did with older folks. What, then, made a bunch of young, white guys vote for him over Clinton? One might argue that it was his stance on pot legalization, but given that Johnson was also in the 2012 race, that couldn’t have been the deciding factor. So what was different this year? Well . . . the Democrat was a woman, and one whose primary opponent made a point of reaching out to young white men and more or less ignoring everyone else. One last bit of evidence in favor of this: The education breakdown. While everyone without a college degree dropped their support for the Democrat by about six points, those with at least some college–perhaps, for instance, those happening to be active students–doubled their support for third parties between 2012 and 2016. Bernie was beloved by college students of all demographics, but white college men basically turned him into a religion. When he lost the primary, they decided he’d been done wrong by the evil woman and the DNC as a whole, and set about to taking their revenge. They knew very well they were unlikely to come to any real harm from who knows how many years of Republicans controlling all three branches of government, so that hell was something they were willing to risk just to pitch a tantrum about their cult leader not winning the nomination.
All this being the case, I think we can finally dispense with the think pieces browbeating educated city folk for not being more thoughtful about rural racists. They’re a pain in the ass in every election, but they’re not what lost us this one. That honor goes almost entirely to Bernie Bros. And if we’re to have any hope at all of winning the next election, they’re the ones we need to figure out how to handle. Capitulating to them, as with capitulating to already-conservative racists, isn’t an option. Working on ensuring voting access and motivation for other groups? That, we can do.