I am generally in favor of blunt conclusions. But I’ve presented each explanation here as a theory because we’re talking about one or two cycles and I’m uncomfortable reading too much into the poll numbers.
The images and data are from the exit poll results conducted for the national election pool (form the NYT). The sample is over 30,000 voters. Like all polls, take with a suitable amount of salt.
Gender, Race and Age
The little arrows in the chart above compare the 2016 spread with 2012. So for instance in 2016, Trump won men overall by 12%. That was 5% better than Romney, who garnered 7% more. Clinton gained 1% of women over Obama’s numbers, but the margin widened by 5% among men. Clinton’s narrower popular vote victory (0.9% vs 3.9%) and electoral college loss can be explained by underperformance among male voters. Clinton did not outperform enough among women to compensate.
Though it is impossible to get into people’s hearts, on its face, this suggests that the male voters were less persuaded by Clinton than her Democratic predecessor. If we accept other things as equal, this does suggest sexism played a role in the result.
Clinton lost ground compared to Obama ‘12 with each and every racial/ethnic group. The smallest loss (1%) was among white voters. One possible interpretation is that with two white candidates on the ballot, minorities did not vote based on identity and reverted to ideology/policy.
Taken together, you could say that identity politics “works” (or at least worked for Obama) along racial/ethnic lines, but does not work (or did not work for Clinton), along gender lines.
There was an interesting generational divide. Trump improved on Romney’s margin among 18-29 (5%) and 45-64 year olds (5%). Hillary’s gains among 65+ (4%) and 30-44 year olds (1%) could not make up the losses in other age cohorts. I really can’t come up with a good explanation for this. I’ve thought about younger people being more liberal, party alignment when they first came of age, differing impact of recent financial crisis.
Party Affiliation & Marital Status
Clinton did improve on Obama’s margin with Republicans by 4%. Which suggests the attempt to turn Republicans off from Trump did work. But Clinton also lost 5% more Democrats to Trump. As a whole, I think this should be taken to mean that party loyalty held. Roughly 90% of affiliated voters cast their ballots along party lines, even with two very polarizing candidates. Republicans were not turned off by Trump in sufficient numbers. Neither were independent voters, who shifted very slightly towards Trump.
Clinton lost ground with unmarried voters to the tune of 10%. This is a large move and seems important, but I’m not sure how to interpret it. Perhaps it’s related to income, income insecurity for married vs. unmarried persons, or and desire/willingness for “change”. 44% of adults in the US are unmarried
Income and Class
Income seems to have been a very big differentiator. Voters who make under 50K consistently break for Democrats. Trump made very significant inroads into this portion of our coalition. That is to be expected with a populist economic message, even if there are doubts about ability to deliver (“at least he’s saying he will try”).
Though Clinton did improve significantly with voters whose income exceeded 100K, Trump’s larger gains among the larger group making <50K
Income |
<30K |
30K-50K |
50K-100K |
100K-200K |
% of Electorate |
17% |
19% |
31% |
24% |
Dem Gain/Loss |
-16% |
-6% |
2% |
9% |
DEM Total Vote Gain/Loss |
-2.72% |
-1.14% |
0.62% |
2.16% |
I can see only one way to read this result. Class/Income based politics can move the needle a lot (16% margin improvements!). This would argue for more class based politics and economic arguments. A second conclusion is that the Democratic party and the Clinton campaign allowed our coalition to be breached. The GOP always wins with wealthier voters. Clinton made some inroads there. But that was overcome by the extent to which Trump improved on Romney’s result with under 50k voters. He breached the Democratic coalition (in the right geographies), we only grazed his.
But we won the Popular Vote
Okay, back to being blunt. Yes, we’ve won the popular vote 6 out of 7 past elections. And we’ve lost the presidency three out of seven times. Plus we’ve seen our state/local footprint narrow.
Maybe we need to ask ourselves whether Republicans are just better at politics than we are.
But, but, but Fox News
At this point, someone will interject, “but FOX News”. Okay, Fox news.
Why don’t we have our own FOX News? And I don’t mean to gin up our coastal bases which is a total waste of our time. Instead, why not produce something that appeals to the center of the country. You know, those states where 2 senators add to electoral votes and give them outsized representation. Why don’t we have news-ertainment produced for those markets, with their cultural cues, designed to undermine Republican candidates and ideology?
Maybe we need to ask ourselves whether Republicans are just better at propaganda than we are.