GA Early Voting Update 10/29
First a note about polls. Republican leaning pollsters like Insider Advantage seem to be flooding the zone with junk. Insider Advantage's latest Georgia poll shows Raphael Warnock getting an incredible 36% of the white vote (5% higher than even Biden got), yet losing with 45% of the vote because it also shows Herschel Walker getting an incredible 20% of the black vote (would be about double what Trump received) and winning the non-white/black vote (Biden won this vote comfortably). So I'd ignore it. Recall that Insider Advantage's Phil Kent, a noxious character, called the election for Trump after ballot counting already showed a likely Biden win a few days after the Nov 2020 election. They aren't independent analysts, they're Republican boosters. It's entirely possible that Walker has a lead in line with the IA polling topline, but it won't be by drastically under-performing with white voters and massively over-performing with non-whites.
Voting continues at a swift pace. Up to 1.5mm votes. Republicans won yesterday's early vote by about -3,000 votes, but Democrats actually put about +1,000 new-voter net votes in the bank. 43% of the '21 advance in person voters have cast a ballot this year, 41% of the Dem rich '21 absentee by mail voters have cast a ballot, 9% of Rep rich '21 election day voters have cast a ballot, and just under 4% of '21 non-voters have cast a ballot. Democrats would win the anyone early+'18 vote (which now tops 4mm expected votes) by around +1,400 today, but as you'll see below the vote should be a lot higher than that (and probably more Republican).
There are about 126k absentee by mail votes that have yet to be returned or processed, Democrats would win them by around +28k votes if they all came back. That, plus the votes that are already in, would put us at around +80k, or just under Kemp's 2018 election day margin.
What's election day likely to look like? Kemp's -94k win from '18, Trump's -220k win from '20 or Perdue's -358k win from '21? Probably something between '18 and '20. The extreme vote differences by mode from '21 are unlikely to re-materialize. The '22 primary gives us some hints.
When we look at '21 absentee by mail voters who haven't voted yet, among those who voted in the '22 primary, 60% of Democratic primary voters and 65% of Republican primary voters returned to election day voting. When we look at '21 advance in person voters who have yet to return a ballot, a nearly identical 70% of Democratic primary voters and 69% of Republican primary voters voted on election day. And for '21 election day voters who have yet to cast a ballot - those who did in the '22 primary were >90% for election day. Election day will be bigger than the '21 runoff and much more Democratic as lots of Democrats are waiting until election day, unlike in the past two elections.
So what's it all mean? At this point I'd expect early voting to close with around 2.5mm ballots cast, or around 1mm votes in the final week. That would be about 20% higher than in 2018 and would put us on track for a total turnout of 4.5mm ballots, which is similar to the '21 runoff. Partisan-wise, my best guess remains an electorate that is about 1% more Republican than the 2020 electorate - or similar to where most Democrats ended up in the 2018 election.
Tomorrow I plan on sharing some thoughts on Warnock/Kemp voters and why I'm skeptical of some polling that shows huge Kemp leads relative to the Senate race.