The Georgia Senate Runoffs Are Today…
My first instinct is to say that the GOP will win both seats in the Senate Runoff. Generally in America, 42% of voters want a Democratic Trifecta, 40% would like a Republican Trifecta, and the rest want either third party/ambiguous views or a split Executive/Legislature because they prioritize “compromise” over “actual policies that would help people”. I preface the blog with this because it's a good baseline to have in any American Election.
Georgia is to the right of the national partisanship by around 4.2%, and that doesn't even mention that Ossoff ran behind Biden by about 2% margin 6 months ago.
But, this time the Senate will be at the top of the ticket and there will only be the two candidates listed. That makes a huge difference in the very marginal votes. If you want to be mad for a second, there will be thousands of Georgia Ballots that will vote for the Ossoff/Perdue Race and leave the Warnock/Loeffler Race blank and vice versa. When elections are close, some of it comes down to random chance. This could be a super close election, and every small subset of people may decide the race.
Another important thing to remember is that Trump is also not on the ballot. This has a possibility of not bringing the GOP low social trust (Jeffrey Epstein didn’t Kill Himself type, not lie to pollsters type) voters. There’s this basic theory I sometimes believe that Trump outperforms polls by statistically significant amounts in the overwhelming white, suburban/exurban states and districts (e.g. Wisconsin, Michigan, Pennsylvania, KY06, IN05, AR02) and that Trump was able to bring out voters who don’t normally vote and generally love conspiracy theories. This relies on the assumption that low trust voters are currently an undersampled group of people. Trump is great at convincing these people to vote for him, but only when he’s on the ballot as shown in 2018 and 2019 when polling was spot on.
But realistically, it’s likely a mix of 100 things and not one simple explanation. Also, polling is really difficult! (Here’s a fun poll that shows Bernie probably would’ve been able to cut into this group of voters)
Some Quick Maths:
Not going to reinvent the wheel and say that Atlanta will be super important in a Georgia Election! Biden had 51,270 more votes than Ossoff in Atlanta and the surrounding counties (Fulton, DeKalb, Gwinnett, Newton, Henry, Clayton, Cobb, and Douglas). That’s about 60% of the margin Perdue won by (if you assume those 51,270 votes voted for Perdue then it more than makes up the whole margin plus 20,000 votes).
Screenshot from Garrett Harin’s website: https://gisanddata.maps.arcgis.com/apps/MapSeries/index.html?appid=5d930a3281ad403391b1d9b3f7f9d145
Here is a map of Biden vs Ossoff (Credit to Garrett Herin, Link). You can see that Biden dominated this area vs Ossoff, but Ossoff was able to make up ground in rural areas.
A somewhat ubiquitous trend was that people voted their topline preference, and hedged downballot. In rural areas around 3-5% vote was Trump/Democrat, and in urban areas around 3-5% vote was Biden/GOP (even with Biden underrunning Hillary Clinton in cities).
We will probably see the Ossoff and Warnock coalitions look more like Biden than Ossoff Part 2. (This is in the weeds and a difference of 3-5% in certain counties and we won’t know until a couple of weeks after the election but it’s all super important for winning elections).
I will be super interested to see if the difference between Ossoff vs Warnock looks like Biden vs Ossoff. This would mean there is a small subset of people in both parties that just really like splitting their ballot, even if it’s for the same office.
I've skipped over turnout so far. If Atlanta+Suburbs goes from 41% of the Electorate, to 43% of the electorate, you create some room to do worse in the other areas. Stacey Abrams did a good job of this in 2018, Georgia Suburbs were just a bit too red still. (I also want to point out that the additional 2% of the electorate could be disproportionately Republican and lower Democrat’s vote share in these counties, but even so, if that additional 2% is further left than Georgia in general, it still helps Democrats. This is what happened in Philadelphia, Detroit, and other heavy Biden counties that went more Republican than 2016).
Unfortunately, we won’t really know the full information until weeks after tomorrow when we are able to comb through all the statistics.
So with all this information, here’s what I’ll be using to watch the election results tomorrow. The Washington Post has made a really sophisticated precinct-level model that is way better than anything I could dream of making. And I’ll link to it as soon as I find it. And I’ve created a spreadsheet that compares how Warnock and Ossoff are doing compared to Joe Biden by Region with some graphs linked here. I’ll also plan on watching Kornacki do his maps!
A couple of last things is that it’s a lot easier to predict Ossoff's race than Warnock's. Luckily, we do have some empirical knowledge on runoffs using past data, and it’s a bit optimistic for Warnock.
Another reason for optimism for Democrats are turnout numbers are good so far. I put this far down because it’s very possible that the GOP has a high turnout and surpasses the Democratic Turnout so far.
You can read about how black turnout is giving Democrats hope in this thread.
That’s about it, everything has gone right for the Democrats and it may not be enough. This will be the most important election in Joe Biden’s tenure until he is up for re-election. We will see what happens soon enough.