Georgia Senate Runoff: Preview and Predictions

It may be over already

First, Some Context

In the nationwide 2022 Senate map, Georgia’s race was seen as one of the “core four” competitive contests that would decide control of the chamber. The other three were two Dem-held seats in Nevada and Arizona, and a GOP-held open seat in Pennsylvania. Like Georgia, they also faced a political environment shaped by a concurrent gubernatorial election. In Nevada and Arizona, the gubernatorial contests were essentially even contests, providing Laxalt and Masters (already poor candidates themselves) almost no margin for error to win. Pennsylvania was another story entirely: there, Josh Shapiro won the governorship by an absolutely crushing margin, giving Fetterman a massive lift in his race against Oz.

In this context, Georgia stands out as the place Republicans were the best positioned to succeed, by far. The eight-point lift provided by Kemp would have been enough for any of the other frontline Republican candidates, even lowly Blake Masters, to come out ahead. And Georgia, as is typical in the Deep South, historically has very little ticket splitting. Everything was set up for a Republican to win the senate race outright. It shouldn’t have been possible not to.

Herschel Walker made the impossible happen. He underperformed Brian Kemp by nine points, an almost unimaginably large split for an inelastic state where politics have always been polarized around race. He fell behind incumbent Senator Raphael Warnock by around a point, roughly 35,000 votes, on the same ballot where Kemp won by nearly 300,000 votes. It was a cataclysmic performance. No other Republican senate candidate in a competitive state had anywhere close to this level of drop-off.

Compared to the same map from the governor’s race, the swing map from 2020 for senate looks like a result from a different timeline. Walker was utterly incapable of matching any of Kemp’s gains in the Atlanta metro area, or north Georgia in general. He put up worse numbers than Trump both inside and outside the perimeter, beyond into the exurbs, and even into the mountains. His performance in Cobb County, once a Republican stronghold, was the worst of any Republican statewide candidate in recent history. He reached his nadir in the fast-growing suburbs south of Atlanta: counties like Henry, Rockdale and Newton that have become popular spots for Atlanta’s budding Black middle class. In these counties, Walker underran Trump’s support in 2020 significantly, with Democratic gains verging on double digits.

Walker’s one saving grace were modest gains he received over Trump in south Georgia’s rural counties, both in but mostly around the Black belt. These areas are generally more racially polarized and less educated than north Georgia, so it is likely that gains made by Kemp at the top of the ticket managed to stick there more than they did in Buckhead or East Cobb. This, coupled with a modest third-party vote share, is why Walker did not lose outright, despite his slippage across north Georgia.

Georgia Early Voting Numbers And How To Read Them

With the rise of early voting across the country, a cottage industry has developed every October leading up to elections, where experts and amateurs alike parse through data, trying to figure out where each party stands before election day. If done correctly, it can be very insightful, most famously in Nevada but also in states like North Carolina and Florida. It also almost exclusively done for states where voters can register by party, for obvious reasons. Since Georgia does not provide party registration to voters, it is not possible to provide such a direct analysis here.

But Georgia does provide the race of its voters, and since Georgia politics are incredibly polarized along racial lines, this measure can serve as a surprisingly good substitute to party registration. Simply put, Democrats win almost all Black voters and do well with Asian, Hispanic, and “other” voters. Republicans win the bulk of white voters. If the turnout for Black and other minority voters is higher, it means Democrats are in a stronger position. The same is true for Republicans and white voters.

This measure will never be as precise as party registration numbers, and in elections without an existing baseline, the implications of the splits can be hard to measure with confidence. But runoff elections, where the previous general election does provide a baseline, is where this measure really starts to shine. With an actual election now corresponding to the finalized early vote splits, determining which party is doing better, or worse, is easy: simply check which racial groups are making up larger, or smaller, shares of the vote compared to November.

What The Numbers Say Now

First, some level of caution must be exercised when making comparisons between the 2022 runoff and the ones that happened in 2021. As a result of the SB202 election bill, runoff elections in Georgia were reworked, with the election date for Senate races moved from early January to early December. The principal effect of this was to drastically reduce the early voting period from several weeks in the general election to only five mandatory days for the runoff. It’s not hard to tell what the intention of this was. Almost everywhere in the country, Democrats predominantly vote early, and Republicans vote on election day. By drastically reducing time for early voters to vote while election day remained unchanged, the new election design seemed to be set up to make Republican-leaning election day voters compose a larger share of the electorate.

However, Georgia voters proved much more adaptable to the new schedule than the authors of SB202 may have assumed. As a result of the smaller period, the daily volume of voters exploded to never-before-seen levels. In the general election, a good day of early voting may have seen a little over 100,000 voters coming to the polls. In the runoff, the daily average was close to three times that. With this record-setting pace, the total number of early voters surged over the weekday period, finishing at about 75% of the total November early vote. And with early voting ending last Friday, we finally have the (mostly) completed early voting splits.

They tell a clear story. The early voting electorate for the runoff is one indisputably more favorable for Democrats than the one that existed at the end of early voting in November. The headline is the Black vote share, which is up by 3%: from 29.1% in the general to 31.9% by the end of Friday. This has come mostly at the expense of the white vote share, when has fallen from 57.2% to 55.1%. This is a devastating development for Herschel Walker. The electorate that had the smaller, 29.1% Black early voting share was already one he was incapable of winning. That it moved further to the left puts him in a very tough position, to say the least.

With that said, these kinds of shifts between a runoff and general election aren’t necessarily fatal. Sometimes there are just external factors, even as simple as bad weather, that can create noticeable changes in turnout on paper without necessarily indicating reduced enthusiasm. An example of this is how prohibitively long lines and a small number of voting stations almost certainly led to artificially reduced turnout in Fulton county compared to peer counties. Because of this, there is a credible argument that Fulton may see increased election day turnout. Unfortunately for Walker, Fulton is the country that contains most of the city of Atlanta, and such a development would benefit Democrats exclusively. In the larger picture, the fact that Democratic numbers have improved by so much compared to November, even with one of their two largest counties metaphorically fighting with one hand behind its back, speaks to how potent Democratic turnout has been across the board.

What’s Caused Low Republican Turnout?

Like Fulton, Republicans also have major counties that have lagged in voting. Three of their most important exurban strongholds—Cherokee, Coweta, and Spalding counties—have some of the worst turnout in the state. The extent to which this may be the result of Fulton-esque issues with ease of voting is not clear. Precinct-level turnout maps show a stark difference in turnout beginning and ending at county borders. This is especially noticeable for Cherokee. At first, this would seem to indicate issues at the administrative, rather than the voter enthusiasm, level: elections are run at the county level, after all, so a problem that appears localized to one county may just mean it is a local issue, specific to that one county in that one week and not to Republican voters overall. The problem for the GOP is that, at least in Cherokee, it is very difficult to find what this problem might have been.

It was not as if Cherokee voters had significantly fewer opportunities to vote than voters in other counties with higher turnout. During the early voting period, Cherokee county offered three early voting locations, across two cities, for the five mandatory weekdays of early voting, with one spot open on Sunday during the optional weekend period. In comparison, similarly sized Forsyth county, which had no issues with turnout, provided four early voting locations, all located in one city, without offering any Sunday voting. That was the extent of the difference. Three locations versus four. Extra polling stations are always convenient, but a single additional location is hardly a sufficient explanation for such a massive gap between the two similar counties.

Of course, Cherokee is just one county. But is emblematic of the most pressing issue facing Republicans heading into Tuesday: poor turnout rates that have few credible explanations other than just low enthusiasm among Republican voters. This phenomenon is evident not just by county, but within the counties themselves. Precinct-level measurements across many of Georgia’s largest counties has shown a strong connection between Warnock support and voter turnout during early voting. This occurred from urban Fulton, to suburban Cobb and Henry, to even exurban Fayette, Hall, and Oconee. Only in Forsyth and Gwinnett, which both have large, low-propensity Asian populations, does this correlation not occur. What makes this so concerning for the GOP is that there is no explanation for it other than low enthusiasm among their voters. If a country has bad early turnout, it can at least be theoretically excused as just an issue with the county, since counties run elections in Georgia. But within a county, all voters are operating under the same regime. Administrative impacts are, in effect, controlled. That GOP turnout is so weak, in so many instances, across so many different types of counties, with all voters being provided with the same level of opportunity to vote, makes it impossible to credibly claim that it is happening for any reason other than Republicans just not being interested in voting for Herschel Walker.

Conclusion

In theory, this race is not impossible for Walker to win. Election day, where all Republican candidates gain the bulk of their votes, is still a wildcard. With the right splits, or the right turnout, he could possibly eek out a victory. The issue is that the “right splits” or the “right turnout” would mean unprecedentedly strong levels of Republican support and turnout for an election day. The only way this could be said to be plausible is if there were clear factors that led to artificially reduced Republican turnout during the early voting period. These are no such factors.

This is the crux of why Walker’s position is so poor. Voting trends at both the county and precinct level strongly suggest that low enthusiasm is the principal factor behind poor Republican turnout. Republican voters had every opportunity this week to come out and vote for Herschel Walker. They just didn’t. Maybe this changes on election day. But right now, there is no good reason to think so.

Rating: Likely Democratic