Republican Senate Majority Hangs by a Single Thread—the State of Georgia

Thanks to strong Democratic support in rural areas and Atlanta, Georgia was a long-time Democratic state that only began to consistently elect Republicans in the early 2000s. However, as Democrats became more progressive and Republicans more conservative, the Peach State proceeded to give majorities to the Republican presidential nominees in each election between 2000 and 2016. Now, because of an influx of transplants predominantly residing in the Atlanta metropolitan area, Georgia is a purple state. Georgians elected Republican Governor Brian Kemp by only 55,000 votes (1.39%) in 2018, and appear to have opted for presumptive President-elect Joe Biden by roughly 12,000 votes (0.24%) on November 3, 2020.

In accordance with Georgia’s new status as a battleground state, incumbent United States Senator David Perdue, a Republican, failed to obtain 50% of the vote on November 3, 2020. He was thus forced to compete in a runoff against Democratic challenger Jon Ossoff, a documentary filmmaker who ran unsuccessfully for the House from Georgia’s sixth district in 2017.

Georgia’s other seat is also being subjected to a runoff, as neither Democrat Raphael Warnock, the pastor of Ebenezer Baptist Church, nor Republican incumbent Kelly Loeffler, received 50% of the vote in November.

Both runoff elections will be held tomorrow, January 5th, and the outcome will determine which party controls the Senate for the next two years. Perdue’s race (which will subsequently be called Georgia’s regular election) tilts Republican, whereas Loeffler’s race (i.e., Georgia’s special election) is a toss-up. This assessment is based primarily on my analysis of early voting statistics and, to a much lesser extent, public-opinion polling.

Assessing Georgia’s Regular Election:

Incumbent David Perdue was first elected in 2014, dispatching Democrat Michelle Nunn by 7.9 points in a race that was described as a toss-up. In 2014, Perdue received 4.8 more points of support than was anticipated by public-opinion polling and, in the first round of the 2020 race, Perdue’s support was underestimated by 3.4 points.

Conversely, support for Perdue’s Democratic challenger, Jon Ossoff, was overestimated by 0.7 point when he ran for Congress in 2017. Perdue’s campaign thus has no reason to worry about his current standing in polling: 48.5% compared to 49.3% for Ossoff.

As of mid-morning on January 3rd, more than three million early votes have been cast in Georgia. The state provides age and demographic breakdowns for these votes, and based on exit-poll data for the November 3, 2020 Perdue vs. Ossoff race with adjustments for accuracy, I estimate that Perdue received 1,444,911 early votes (48.13%) compared to 1,557,189 (51.87%) for Ossoff. My calculations and assumptions can be accessed through the attached link.

Based on the proportion of votes registered on or after the final day of early voting in 2020, I also anticipate that there are roughly 102,240 early votes yet to be counted.* If my earlier assumptions about the partisan breakdown of the early votes prove correct, Perdue can win the overall race, even if Ossoff wins the early votes I expect will be registered on or after January 4th by 50 points—given (1) an Election Day turnout of at least 719,183 and (2) the same breakdown of Election-Day votes between Perdue and Ossoff as in the first round (61.36% Perdue and 38.64% Ossoff, by my calculations).

Provided that the number of early votes cast relative to the same point in the 2020 race (77.35%) is any indicator, Perdue can expect an Election-Day turnout of 736,103. Therefore, turning out 719,183 voters—which would be necessary only under the worst-case scenario for Perdue—is a feasible proposition. And polls, which have underestimated Perdue in the past, look solid for him. Georgia’s regular election accordingly tilts Republican.

Contemplating the Special Election:

Kelly Loeffler’s biggest liability is that she was appointed to replace outgoing Republican Senator Johnny Isakson, and has never won an election in her own right. FiveThirtyEight’s Nate Silver reports that roughly the same proportion of elected senators and appointees opted to run for reelection between 1990 and 2008, but appointees won their primary and the general election just 49% of the time—compared to 88% for senators who were popularly elected. This high failure rate is either the result of appointed senators having no experience winning a race in their own right or their incompatibility with the electorate.

Additionally, Loeffler’s level of support is unpredictable because of the process by which she became the Republican nominee. For perspective, Perdue effectively ran in a head-to-head race against Ossoff, with the Libertarian candidate receiving about 115,000 votes. In Loeffler’s case, she ran in a three-way, jungle primary-type race against Warnock and then-Republican Congressman Doug Collins, coming in second place with just 25.9% of the vote. Collins performed well in North Georgia and some patches of South Georgia. And the 2019 Louisiana gubernatorial race shows that candidates do not always perform as well in runoffs as the two-party vote share in the first round would indicate, especially if one candidate has regional appeal.

The awkward nature of Loeffler’s ascendancy is compounded by the fact that exit polls for the special election—which I relied on heavily for my calculations—overstated support for Democratic candidate Matt Lieberman. As a result, my estimation required a sizable adjustment to reflect the 2020 first round results, and even this tweak does not account for the fact that Loeffler will not necessarily receive the support of everyone who voted for a Republican in the jungle primary.

That said, my data show that Loeffler is in a worse position than her Republican counterpart Perdue. Based on partisan identity alone, I predict that Loeffler has received 1,432,302 votes (47.71%) thus far, meaning that she is running 4.58 points behind Warnock’s forecasted 1,623,259 votes (52.29%)—a calculation that may overstate Loeffler’s position, for the reasons I listed above.

In order to win, Loeffler needs an Election-Day turnout of at least 662,527, assuming that (1) the Election-Day votes break down the same way they did on November 3rd (60.73% Republican and 39.27% Democratic), (2) that there are 102,240 ballots inputted on January 4th, and (3) that they go for Warnock 52.29% to 47.71%. If the January 4th ballots go for Warnock by, say, 50 points, Loeffler would need an overwhelming Election-Day turnout of 878,919 voters. This is a worse-case scenario for Loeffler, but it indicates the precariousness of her standing.

In short, Loeffler’s race is a toss-up. Consider that if roughly 736,103 voters turn out on Election Day, as trends in early voting indicate, she could lose even if Perdue wins. Loeffler’s marginal underperformance relative to Perdue is also illustrated by public-opinion polling, which has her vote share at 48.0% versus 48.5% for Perdue.

Conclusion:

The Republicans are marginally favored to maintain control of the United States Senate, as David Perdue is slightly favored in his race against Jon Ossoff and Kelly Loeffler is neck-and-neck with Raphael Warnock. However, Republican competitiveness in both races is contingent on an Election-Day turnout of more than 500,000 people, assuming that those who cast ballots on January 5th are of the same political persuasion as the November 3rd voters. Otherwise, the Democrats’ lead in the early vote will make Chuck Schumer (D., New York) the Senate majority leader.

* 4,013,155 early votes were reportedly cast in Georgia for the 2020 general election. 3,880,984 were reported by the morning of November 1st. 4,013,155 divided by 3,880,984 is roughly 1.03405. This fraction times the current number of early votes for 2021 (3,002,100) is 3,104,322. 3,104,322 subtracted by the current number of votes is 102,240.

*The views expressed in this article solely represent the views of the author, not the views of the Chicago Thinker.