After months of campaign work, it’s fun to sit around a play with numbers for a little bit. (But don’t forget MS-Sen!) Here’s updates on two close races that haven’t been covered much here.
In NY-22, D was up ~1300 votes but with ~13,000 absentees and several thousand “affidavit” (provisional) votes to count. There was some reason for nervousness because more of the absentees were registered R than D. Two D-leaning counties have finished their counts. In both, the new votes tilted significantly more D than the election-night ones. A third county, R-leaning, reported slightly incomplete results. It also tilted more D than on election night. At this point, there’s very little reason to doubt that it will be a D win. If you disagree, you can buy NY-22 R shares at 3 cents on the dollar at Predictit.
In UT-04, the R has been steadily gaining on the D as more rural votes come in. D is currently up 1002, but the remaining regular votes from Utah County will leave the race nearly tied before the provisionals come in, maybe slight D lead. That’s where it gets interesting. There are expected to be ~9000 provisionals from Salt Lake, and ~3000 from Utah. If they go just like the average from those counties (8% D, 50% R), R wins by up to ~1000. Provisionals, however, seem to consistently lean ~10% more D than the other votes, at least from what I’ve read about CA, FL, and OH. If UT is anything like those other places, that adds ~1200 to the D margin. So I think a D win is slightly more likely, but with low confidence. The gamblers at Predictit think there’s ~68% probability of an R win, but I think they’ve neglected prior knowledge about provisionals. Or maybe I’m missing some specific knowledge about Utah and its provisionals. The R’s may see it my way, since they’re suing to try to interfere with counting provisionals in Salt Lake.
Meanwhile GA-07 looks like a heartbreaker, although there’s always a chance a hand recount can find 400 more votes.
CA-39 was a done D long ago, well before the count went D. It’s fun to look at the charts, but nothing much new to learn.
The sleeper CA-21 (wrongly called R by AP etc.) and its uncertainties is discussed elsewhere. My eccentric view is that it’s slightly more likely D than R.
Does anybody know anything about the long-shots TX-23 and NY-27?