Wednesday I mentioned Michael McDonald, the professor at the Univeristy of Florida who is an election data guru. Something I noticed in the first days of early voting was that most of the swing states that surfaced gender breakdowns for early voting showed around a ten point spread between men (~45) and woman (~55). There are more women than men and women vote more than men. So the difference didn’t surprise me. But that spread still seemed pretty big. So I asked McDonald whether that was a signal of any sort. He said, no, that’s roughly the spread you see in early voting.
But over the last couple days, both in my exchanges with him and in a few of his tweet updates, something else has come out of this. That ~10 point spread is about what we should expect from other cycles. But we’re also seeing a lot more Republican early voting. All things being equal that high rate of Republican early voting should be compressing that gender divide. But it’s not.
So what’s going on here? It’s not clear. There are a few possibilities. One is that that Republican early voting is itself skewing female and thus keeping that gender spread high. Another possibility is that Democratic women are voting early at even higher rates than in previous cycles and that’s off-setting the impact of Republican early voting. McDonald has a post discussing the different possibilities for North Carolina. But it’s behind a substack paywall.
For the moment I don’t think we can say this is good or bad for one party or the other. It’s more that there’s some additional factor which is allowing the rate of Republican early voting to expand without that 10 point gender spread contracting. It’s not clear what that is yet. But whatever it is could turn out to be significant for the final results of the election.
I should note that while McDonald’s post is about the situation in North Carolina, there are two other swing states releasing gender breakdown data for the early votes. And the numbers are comparable and actually larger. In Georgia it’s 11.2 percentage points and in Michigan it’s 13.1 percentage points. So it’s possible that other factors could be at work in those states.
