With the polls being as good as they have been of late, Trump’s commentariat has been out in force, declaring that either the polls are all rigged (with literally dozens of different firms all saying the same thing, that’s unlikely), or they’re somehow not weighting their demographics right–insisting that Trump is electrifying white people, particularly white men, so much that turnout among those groups is going to be a lot higher than it was four years ago. Leaving aside the fact that non-white voters are probably equally motivated this year–fighting for your very life will do that–let’s take a look and see if it really is possible for a higher percentage of white turnout to make a difference.
Most pollsters are using 2012 numbers to weight their results, assuming that the electorate is going to be roughly the same: 72% white and 53% female (among other variables.) Overall demographic change being what it is, chances are pretty good that the numbers this year will probably be even less white. But just to humor folks, I crunched some numbers to see what things might look like if white turnout was indeed higher than it has been lately.
For the polling percentages, I used the breakdown from the Oct. 18 Fox Poll, which with Hillary at +6, is pretty close to the average of the most recent reliable polls (throwing out the outliers at both the top and bottom.)
The last time a Republican won was 2004, when white turnout was 77%. But even if turnout rolled back to those numbers, it still wouldn’t be enough, because his support is sharply down among ALL demographic groups–even white men.
Looking at the breakdown for the four-way race, Trump is doing absolutely awfully with every group compared to 2004:
White men: Bush 62%, Trump 51%
White women: Bush 55%, Trump 44%
Non-white: Bush 35%, Trump 17%
Moreover, these lower numbers aren’t just an artifact of having a robust third-party vote this year. Check out the Kerry/HRC comparison (again, this is the four-way):
White men: Kerry 37%, HRC 32%
White women: Kerry 44%, HRC 41%
Non-white: Kerry 72%, HRC 68%
She’s doing about as well as Kerry did, but Trump’s doing SO much worse than Bush that there’s simply no chance he can win, even if white turnout was at the same level. The final result in that case would be HRC 44%, Trump 40% (with the rest going to the third parties.)
OK, but what if white turnout is MORE than 77%? And what if men turn out at a higher rate than women, which hasn’t happened in a generation? Fine. Let’s run the numbers with turnout levels for every election since 1980.
The only point at which Trump comes close to winning is with white turnout at 84% or more, which hasn’t been seen since 1992.
For funsies, I also projected what things might look like this year if white turnout was instead at 70% (with a 48/52 split for men/women): 46% vs. 38%. Ouch.
Trump therefore has to count on three things:
1. White turnout percent being up to a level we haven’t seen in a generation.
2. White men turning out equal to or greater than women
3. Doing a LOT better with both of those groups than he is now.
In fact, doing a little more playing with the numbers, if all else is the same (i.e. if Hillary doesn’t suddenly lose a bunch of voters to third parties, which is unlikely at this point), the only way he wins is if he gets at least 52% of white men and 47% of white women, and white turnout is more than 85%. With a (very generous) white turnout of 80%, he still loses by a point and a half unless his numbers start going up more. He might be able to win very white and very conservative states such as North Dakota, but once the non-white population starts dipping below 80%, and the % of whites who are Republican starts going down, he gets in trouble. There simply aren’t enough states with that high of a white population that have enough electoral votes to give him a win.
Bottom line: Turnout isn’t Trump’s ace in the hole. Improving his performance with actual voters is the only chance he has, and seeing as how he’s managed to alienate virtually everyone except cishet white racists, that’s very unlikely to change in the next 16 days.
*Note: This post has been edited to reflect more-recent polling results.