Election 2024: OK, Here's Something That Could Affect the Outcome

For each presidential election since 2012, I have:

1. Issued a state-by-state prediction of outcomes; and
2. Stood by that prediction instead of changing it at the last minute.

I usually try to do that fairly early (this year I did it on September 21).

And, without exception in the three elections I've predicted for, I've been correct as to the overall winner and as to the winner in (exactly, all three times) 48 of 50 states.

My prediction this year, in aggregate, is 319 electoral votes for Kamala Harris, 219 electoral votes for Donald Trump.

Will this be the year I fail to correctly predict the outcome in 47 or fewer states and/or the overall winner?

Well, it's possible. And here's a new data point that might contribute to that.

Presidential elections that are even a little bit close always boil down to turnout -- the candidate/campaign that's better at getting its voters to the polls wins.

And until recently, Trump's campaign was sucking at that, having handed off its Get Out The Vote operation to an amateurish outfit with little experience.

But now Elon Musk's America PAC is spending big money ($87 million as of October 9) on good old-fashioned (although very tech-assisted) door-knocking/direct mail GOTV work in swing states.

And they seem to be doing it right. Canvassers, the Post reports, "were told to focus their remaining time on getting confirmed Trump supporters to vote early, instead of spending precious time trying to change the minds of undecided or Harris voters."

I said before that Harris is in the driver's seat and that there's nothing Trump himself can do to change that -- unless she publicly shits the bed, she's got the edge on "number of supporters." Every time he runs his yap, he loses a few more, and that's going to remain true.

But if Musk can get Trump's vote out better than the Democrats can get Harris's vote out, it could still go the other way.