Start to In-Person Early Voting Suggests Enthusiasm for Georgia Runoff
By the Socio-Economic Research & Tech Team; editing: Jim Flood.
Methodology: Analysis of Georgia Secretary of State data
With less than three weeks until Election Day, here’s the current state of the race of the Georgia Senate runoffs. Runoff data is through Tuesday, December 15.
- Thus far, “runoff dropoff” has not occurred. The number of early votes cast in the runoff represents 96% of early votes cast in the general election at the same point in time. While voting by mail is lower this time around, early voting in person was higher in its first two days relative to the general election.
- As expected, there are wide variations in voting patterns across the state. In the counties around Atlanta, an underperformance in voting by mail is being offset by higher turnout in in-person early voting.
- The overwhelming majority of people who have voted in the runoff also voted early in the general election.
- Unsurprisingly, older voters are outperforming younger voters in the runoff, a dynamic also evident at the same point in time in the general election. Additionally, white voters are outperforming their share of the electorate in voting by mail, whereas Black voters are outperforming their share of the electorate when it comes to early voting in person.
To begin, we compare overall ballot request and return stats from the Georgia Secretary of State (SOS) for the runoff election with the equivalent stats from the general election at the same point in time (PIT). Meaning, the same number of days out from Election Day.
With less than three weeks to go before Election Day, the number of early votes cast in the runoff represents 96% of early votes cast in the general election at the same point in time. However, across the two elections, those votes break down differently based on voting methods. From a point-in-time comparison, mail-in ballot requests and returns in the runoff election are lagging their counterparts in the general election, almost 25% lower for both categories (Table 1). On the other hand, early voting in person — which the Georgia SOS also calls “advanced voting” — is 38% higher in the runoff than at the same point in time during the general election.
Note: Although Georgia SOS uses the terms absentee-by-mail and advanced voting, we will use vote-by-mail (VBM) and early voting in person (EVIP) for the remainder of this piece.
Table 1: Overall request and return stats for runoff and general election PIT
Of course, these overall stats potentially mask substantial voting differences (in terms of amount and method) across counties.
Note: In the tables below, we cite general election vote shares for Biden and Ossoff. Because Warnock ran against 19 other candidates, his share of the vote in the November election is likely not representative of his support in the runoff.
To understand differences in VBM across counties, we examine return rates. We define return rate as the number of accepted ballots divided by the total number of requested ballots minus those canceled by the requester (because those ballots have already been “dealt” with). Thus far, the top five counties with the highest return rates are on the small side (Table 2).
Table 2: Top five counties with the highest VBM return rate
When we analyzed the top 20 counties with the highest return rates in the runoff — each with a return rate higher than 50% — Biden and Ossoff both received more than 50% of the VBM votes in the general election in half of them.
But how are the counties around Atlanta doing? As many commentators have pointed out, it was the Atlanta suburbs and exurbs that propelled Biden to victory, and turnout in these same counties will make or break Democrats’ chances of winning the runoffs. Based on ballot return rates, the state of the runoff in those counties is mixed: while counties like Rockdale and Gwinnett were beating their corresponding general election PIT return rates as of December 15, the remaining five Atlanta metro area counties were not (Table 3). Moreover, the return rates for the majority of these counties are only 50 to 60 percent of the return rate of the counties highlighted in Table 2.
Table 3: VBM return rates of Atlanta area counties
Early Voting In Person
The first two days of in-person voting in the runoff tell a more encouraging story. Because the concept of return rates doesn’t apply to EVIP, we ask a simpler question: is the number of early votes cast during the runoff greater than the number of early votes cast at the same point in time during the general election? For 132 out of Georgia’s 159 counties, the answer after the first two days of early voting was yes.
Once again, the counties with the highest percentages of PIT voting — meaning, the highest levels of voting in the runoff relative to how many votes had been cast in the general election at the same point in time — are smaller counties. However, unlike in the case of VBM, key Atlanta metro counties are also overperforming, with six out of the seven highlighted counties (Table 4) above 100% of general election early in-person PIT.
Table 4: Percentage of general PIT for counties around Atlanta
2021 voters, meet 2020 voters
Are the runoffs bringing out voters who skipped or were ineligible for the general election? Thus far, no. The people who have voted in the runoff — whether by mail or in person — are overwhelmingly those who voted in the 2020 general election (Table 5). Additionally, there do not appear to be any major shifts in voting method: the vast majority of people voting in the runoff not only voted early in the 2020 general election, but also used the same voting method as before.
Table 5: Characteristics of runoff voters
To campaigns, of course, it’s important to know not only how many people are voting, but also who is voting. When we compare the people who have voted thus far in the runoff to the entire Georgia electorate and to those who voted in the 2020 general election, some noteworthy patterns emerge:
- Gender: In the runoff, female voters are outperforming their share of the electorate when it comes to VBM; however, their share of EVIP in the runoff is about equal to their share of their electorate. At this point in time during the general election, female voters were outperforming their share of the electorate in both early voting methods (Table 6).
Table 6: Vote share of various universes by gender
- Generation: Younger Georgians (Millennials and younger) are making up a significantly lower share of runoff voters than their share of the electorate at large. For example, although younger Georgians represent 40% of the electorate, they only make up 11% of VBM runoff voters and 16% of EVIP runoff voters. This pattern was also present at the same point in time in the general election (Table 7).
Table 7: Vote share of various universes by generation
- Race: Vote share by race varies somewhat dramatically across voting method. White voters make up a larger share of VBM runoff voters compared to their share of the electorate. On the other hand, Black voters make up a larger percentage of EVIP runoff voters compared to their share of the electorate (Table 8).
Table 8: Vote share of various universes by gender
Thus far, the feared dropoff in voter enthusiasm for the Georgia runoff has not materialized. As of December 15, the runoff is fewer than 30K votes behind the general election at the same point in time. However, until early voting in person started — with higher-than-anticipated turnout — lower rates of voting by mail relative to the general election (PIT) had raised concerns of a potential decline in voter interest. Of course, with only two days of in-person early voting observed, it’s unclear whether this enthusiasm will continue.