Source: Prof. Michael McDonald (http://www.electproject.org/home/voter-turnout/demographics)
I am among those who feel the cake has been baked on this election since before the conventions, and possibly since before the first primary votes were cast; since the conventions we’ve been struggling over how thick the icing will be. I have a feeling that white vs non-white sentiment toward HRC, combined with relative predicted turnout, would be all that has been required to estimate the final outcome. I understand from an academic perspective why even the aggregators’ vote totals were bouncing all over the place but I can’t accept that actual people in large numbers were waffling between these two candidates on a daily basis.
During the primaries, 538 had an interesting chart tracking Hillary’s and Bernie’s performance relative to goals for each state to order to accumulate sufficient delegates to win the nomination. Their projected goals were suitably 538-esque in their complexity (and accuracy!), and a primary predictor and differentiator was the white-ness, the Euro-ness, dare I say, the Caucasionality, of each state. It kind of looks like today’s result will take a very similar shape.
I have thought about this post for a while, and have had two reasons for hesitation. First, I’m sure folks much smarter and astute and skilled than I am have presented a model for today’s electorate along similar lines (if you know of one, please do link in the comments), but I don’t see one on 538. It makes sense they might not do such a model, as it makes more sense for a rolling process like a primary with relatively little reliable public polling, and it also is baked into the much more comprehensive model made with proper, timely polls. In other words, I don’t really know what I’m talking about or how statistics work.
Second, despite my status as a privileged American white human, despite all the effing racists and sexists I know and love, it’s very hard for me to accept that this whole thing might just come down to skin color. Half or more of each candidate’s support comes from blind party affiliation, and the other half comes from folks who believe they are voting for the person who will be best for his or her pocketbook. A large portion of the other side’s voters ARE deplorable racists, naturally, but a large number on both sides are folks who have been squeezed by The System — it just so happens that enough of the sqeeze-ees in this country don’t know and/or don’t care that their candidate is an unfit charlatan, and they are just looking to “send a message” and “blow things up” to get (or get back) something they feel has been unjustly taken away. They’re not wrong about their symptoms, they are however sickeningly wrongwrongwrong about the cause and the cure. In short, I don’t agree HRC needs to consider “Trump voters’ concerns,” but I believe she has campaigned on solutions to real problems which are fueling the deplorables’ pain and response to that pain. Racism and ignorance is the deciding factor whether a given deplorable can give credit where it is due for any relief, or can accurately place blame for its continuation or increase.
Anyway, a table:
Overall electorate by white vs non-white, with a wild guess at vote share for each candidate & group
I should note that the 2016 electorate number of 71.8% white is based entirely on a trend of the previous four presidential election turnouts, and the HRC vote shares are guesses on my part. This is to say, the values in this matrix were numbers for which we could have trended and guessed any time in the last four years based on Obama’s performance and Hillary’s various strengths and challenges with the electorate.
So a broad, race-only estimate says we should expect HRC to land in excess of 50% of the vote. Taking out third-party votes, something in the 50% to 53% range. The difference between the actual tally and the latest polling averages (which HRC just under 50% with a 2 to 5 point lead) may accrue based on just how much browner (and more non-white-male!). Possibly more importantly, Trump’s share is remarkably consistent with his magic 40%-ish ceiling in the polls.
I have made a similar study on a state-by-state basis, using only the overall white/non-white balance of the overall population — which are very different numbers, obviously, from the electorate percentages for each. Strikingly, the overall vote total predicted by that state-by-state comes up with the same spread — and an EV outcome of 317 EV’s (using that somewhat silly aggregators’ math where HRC wins 25.4 of Texas’s 38 EV’s) to 351 EV’s.
I will follow up with an analysis of the state-by-state when I have the actuals to compare. We’re going to win tonight. There’s still a lot to do not only to change hearts and minds, but to make America healthier, more just, and more fair.
Stronger Together.
Peace.
