Pigs fly. Hell freezes over. The Electoral College favors ... Democrats?

This scenario is not very likely — but probably worth a bet at 300:1

Here at Silver Bulletin, we’ve repeatedly emphasized the idea that the Electoral College is much more likely to hurt Democrats than to help them — as it did in 2000 and 2016. This is a conclusion borne directly from our model. As of Thursday, our forecast is that Kamala Harris is a 3:1 favorite in the popular vote — but the Electoral College is basically still a toss-up. That’s because the model figures there’s a 20 percent chance that Harris wins the popular but loses the Electoral College, but only a 0.3 percent chance of the same thing happening to Donald Trump.

At the New York Times yesterday, however, Nate Cohn offered a dissenting view. Cohn isn’t predicting an Electoral College split favoring Democrats, but he thinks the penalty Harris faces will likely be smaller this year. On that point, the Nates agree — though we differ on the extent of the difference.

In 2016, there was nearly a 3-point gap between Hillary Clinton’s popular vote margin and the result in the tipping-point state, Pennsylvania. The gap grew even wider in 2020, to nearly 4 points, although Joe Biden won the popular vote by enough that he still just barely carried the tipping-point state, Wisconsin. This time around, with Pennsylvania projected to be the tipping point state again, the gap is only 0.7 points instead, according to the Times’s polling averages — that’s the difference between Harris’s small lead in Pennsylvania in their numbers and her slightly larger margin in national polls.

We have the gap a bit wider. In our averages, Harris leads by 2.8 points nationally but by 1.5 points in Pennsylvania. So we’re splitting hairs here, but that’s a 1.3-point difference, not 0.7.

And if you scroll down and look at our forecast, the difference gets just a bit wider again. We project Harris to win the popular vote by 2.5 points, but Pennsylvania by 1.0, a 1.5-point difference.

Let’s pause here to note that Harris would probably take a mere 1.5-point Electoral College penalty if she could lock it in today. Democrats have won the popular vote by that amount or more in the past four elections. Still, I’m going to take some time to explain why our model thinks this scenario is not so robust for Harris — but also how there are a couple of exceptions that might arise.