After two weeks of early voting, Georgia’s early voters still look more Democratic than they did at this point in 2022.
If they vote like similar voters did in 2022, Jon Ossoff would get roughly 51% to 54% of the vote for the two major parties’ Senate candidates, up from Raphael Warnock’s 50.5% in 2022.
More than a quarter of early voters had not voted in a Democratic or Republican primary from 2016 through mid-2022. Based on their age, race, sex and county, we assume they lean about two to one Democratic, and that assumption drives much of our estimate of the Democratic lead among people who have voted.
Election Day could erase much of the early edge, as it did in 2022.
This is the longer update we promised after the second week of early voting, with ballots through Saturday, Oct. 24. It follows our Oct. 21 analysis and Oct. 22 update. For daily turnout by county, see “Counties outside the Metro Atlanta Core outgrew the core on Saturday, and voters 18 to 34 gained share” and the Georgia Early Voting Dashboard. Our next update will be published Wednesday, Oct. 28. Ballot records show who has voted, not how. No votes are counted until Election Day.
What it could add up to
These are scenarios for the Senate race, not a forecast. We give each candidate’s share of the votes cast for the Democrat and the Republican, leaving out third-party candidates; we call this the two-party vote.
In 2022, Warnock won 50.5% of the two-party vote. If this year’s voters vote like similar voters did in 2022, the change in who is voting early would put Ossoff at roughly 51% to 54%, about 1 to 3 points better than Warnock. If voters are 2 points more Republican than in 2022, Ossoff would get roughly 49% to 52%. If they are 2 points more Democratic, roughly 53% to 56%.
- Who votes early. Each scenario raises the Democratic share of the 2022 early vote by how much more Democratic this year’s early voters look than 2022’s: 1.1 to 3.2 points, depending on how we measure it.
- Who votes on Election Day. We can’t see this year’s Election Day voters yet, so we tried two cases: the same as in 2022, or shifted as much as early voters. Election Day keeps its 2022 share of the two-party vote, 35.6%.
- How people vote. We assumed voters vote like similar voters did in 2022, then tried 2 points in each direction.
If voters vote as in 2022, Ossoff would do about 1 to 3 points better than Warnock
Democratic share of the two-party Senate vote in each scenario. Each range covers our different measures of this year’s early voters and the two Election Day cases. Election Day includes provisional ballots. Bars reach the middle of each range; the whiskers mark its ends; the gray line is the 2022 result.
| Democrat’s share of the two-party vote | 2022 actual (Warnock) | Voters 2 points more Republican than in 2022 | Voters vote as in 2022 | Voters 2 points more Democratic than in 2022 |
|---|---|---|---|---|
| Early vote | 55.1% | 54.2% to 56.3% | 56.2% to 58.3% | 58.2% to 60.3% |
| Election Day | 42.1% | 40.1% to 43.3% | 42.1% to 45.3% | 44.1% to 47.3% |
| Overall | 50.5% | 49.2% to 51.7% | 51.2% to 53.7% | 53.2% to 55.7% |
Source: Georgia Secretary of State ballot records and 2022 certified results; Civic Forge Solutions analysis.
Our estimates for each voter are set to match how Georgia’s early voters actually voted in the 2022 Senate race. Applied to everyone who has voted so far, including in Gwinnett, they suggest that about 280,000 more would vote for Ossoff than for his Republican opponent, if they vote like similar voters did in 2022. This is an estimate, not a count, and it covers only the ballots cast so far.
The estimate and the scenarios are separate calculations and should not be added together. The estimate covers only people who have voted so far. The scenarios apply the change we measure to the whole 2022 early vote and then add Election Day.
The governor’s race could look different
In 2022, Warnock won 50.5% of the two-party vote for Senate while Stacey Abrams won 46.2% for governor, on the same ballots. Abrams ran 4.0 points behind Warnock among early voters and 4.7 points behind on Election Day. If voters split their tickets the same way this year, the Democratic candidate for governor would get roughly 46.9% to 49.4%, short of a majority across that range. But 2022’s split reflected that year’s candidates, including a sitting governor. This year’s race for governor has no incumbent, so voters may split differently, and the same goes for races further down the ballot.
If voters split their tickets as in 2022, the Democrat for governor would trail even where Ossoff leads
Democratic share of the two-party vote for governor in each Senate scenario, after applying the 2022 gap between Warnock and Abrams separately to early and Election Day ballots. Election Day includes provisional ballots. Bars reach the middle of each range; the whiskers mark its ends; the gray line is the 2022 result.
| Democrat’s share of the two-party vote for governor | 2022 actual (Abrams) | Voters 2 points more Republican than in 2022 | Voters vote as in 2022 | Voters 2 points more Democratic than in 2022 |
|---|---|---|---|---|
| Early vote | 51.1% | 50.2% to 52.3% | 52.2% to 54.3% | 54.2% to 56.3% |
| Election Day | 37.4% | 35.4% to 38.6% | 37.4% to 40.6% | 39.4% to 42.6% |
| Overall | 46.2% | 44.9% to 47.4% | 46.9% to 49.4% | 48.9% to 51.4% |
Source: Georgia Secretary of State 2022 certified results; Civic Forge Solutions analysis.
What changed since Oct. 21
Georgians cast about 418,000 more early ballots from Thursday through Saturday, nearly a third more than our first analysis counted through Wednesday, Oct. 21. None of the scenarios moved by more than a tenth of a point. Our estimated Democratic lead grew mainly because more people voted. The newest voters were a little less Democratic than earlier ones: we estimate 57.0% of them would vote for the Democrat in a two-way race, compared with 58.6% of those who voted earlier.
Three more days of voting barely moved the scenarios
Each column compares this year with the same number of days before Election Day in 2022: Oct. 26, 2022 for Wednesday and Oct. 29, 2022 for Saturday. Figures through Wednesday are recalculated from the latest ballot records, which include ballots cast earlier but accepted later, so they can differ slightly from our earlier analysis. The estimated lead covers all early voters, including Gwinnett; the other figures leave out Gwinnett.
| Through Wednesday, Oct. 21 | Through Saturday, Oct. 24 | |
|---|---|---|
| Senate, if voters are 2 points more Republican than in 2022 | 49.3% to 51.8% | 49.2% to 51.7% |
| Senate, if voters vote as in 2022 | 51.3% to 53.8% | 51.2% to 53.7% |
| Senate, if voters are 2 points more Democratic than in 2022 | 53.3% to 55.8% | 53.2% to 55.7% |
| Governor, if voters split their tickets as in 2022 | 47.0% to 49.5% | 46.9% to 49.4% |
| Estimated Democratic lead among people who have voted early | about 223,000 votes | about 280,000 votes |
| How much more Democratic early voters look than in 2022, voter-profile measure | +3.3 points | +3.2 points |
| Same, neighborhood measure | +1.2 points | +1.1 points |
| Early voters with no primary history | 25.7% (2022: 15.2%) | 28.1% (2022: 17.2%) |
Source: Georgia Secretary of State ballot records, voter files, voter history and 2022 certified results; Civic Forge Solutions analysis.
What we assume about voters with no primary history
More than a quarter of this year’s early voters, 452,550 people or 28.1% outside Gwinnett, had not voted in a Democratic or Republican primary from 2016 through June 2022. At this point in 2022, that share was 17.2%. Ballot records cannot tell us how they will vote, so we assume they will vote like other Georgians of the same age, race, sex and county.
On that assumption, they lean about two to one Democratic, 66%, up from 63% for the same group in 2022, because more of them are young or people of color. Outside Gwinnett, they account for about 135,000 votes of our estimated Democratic lead, up from 65,000 for the same group in 2022, more than half of the lead’s growth.
Of the voters in this group who have voted in a party primary since June 2022, 68% chose a Democratic ballot, close to our assumption. But that is only 41% of the group, and a primary ballot is not a November vote.
We assume early voters with no primary history lean about two to one Democratic, and they add more to the estimated lead than in 2022
Early voters outside Gwinnett with no Democratic or Republican primary from 2016 through June 2022. The assumed share is our estimate, based on age, race, sex and county, of the Democratic share of their two-party vote; blue cells lean Democratic and red cells Republican, darker the further from an even split. The estimated lead is estimated Democratic minus Republican votes, rounded to the nearest thousand; bars show its size, blue for a Democratic lead and red for a Republican one. These are estimates, not counts.
| Early voters with no party primary through 2022 | Voters, 2022, through Oct. 29 | Voters, 2026, through Oct. 24 | Assumed Democratic share, 2026 | Estimated Democratic lead, 2022 | Estimated Democratic lead, 2026 |
|---|---|---|---|---|---|
| All voters without a primary history | 262,015 | 452,550 | 66% | +65,000 | +135,000 |
| Ages 18 to 34 | 62,118 | 127,788 | 75% | +25,000 | +61,000 |
| 35 to 54 | 89,080 | 132,911 | 68% | +24,000 | +45,000 |
| 55 to 64 | 55,219 | 84,437 | 60% | +9,000 | +16,000 |
| 65 and older | 55,555 | 106,522 | 56% | +7,000 | +13,000 |
| Black | 78,424 | 148,710 | 94% | +68,000 | +128,000 |
| White | 140,765 | 208,787 | 43% | −18,000 | −28,000 |
| Hispanic | 8,030 | 27,007 | 73% | +3,000 | +12,000 |
| Asian | 9,305 | 19,582 | 70% | +4,000 | +7,000 |
| Other or unknown race | 25,491 | 48,464 | 68% | +9,000 | +16,000 |
Source: Georgia Secretary of State ballot records, voter files, voter history and 2022 certified results; Civic Forge Solutions analysis.
The second week’s voters
The scenarios held because the second week’s voters stood out from the same week’s voters in 2022 about as much as the first week’s did. In both years, second-week voters were less Democratic than the first ones, and more of them had voted only in Republican primaries than only in Democratic ones. But the mix was still 5.3 points more Democratic than in the same week of 2022, compared with 6.7 points in the first week.
By neighborhood, second-week voters came from precincts that were, on average, 1.1 points more Democratic than the same week’s voters in 2022, compared with 1.3 points in the first week. On Saturday, though, voters came from precincts 0.5 points less Democratic, on average, than voters on the same Saturday in 2022. It was the first day with voting across the state when that happened, though by primary history Saturday’s voters were still more Democratic.
Every day of the second week brought more voters with no primary history than in 2022, and fewer who had voted only in Republican primaries
Early voters statewide by the party primaries they voted in from 2016 through June 2022. “Both” means the voter took part in both parties’ primaries in different years. Days are matched by the number of days before Election Day, and each voter is counted once, on the day they voted. The top row also includes mail ballots returned before early voting began and, for 2022, one more day of in-person voting; the first-week row covers matched days only. Gwinnett is left out.
Source: Georgia Secretary of State ballot records and voter history; Civic Forge Solutions analysis.
By neighborhood: more ballots from Democratic areas, fewer from the most Republican
We sorted every early voter by how their precinct voted for president in 2024 and compared each group with the same point in 2022. Through Saturday, Oct. 24, Georgia had accepted 1,751,150 early ballots. Outside Gwinnett, early ballots are up 5.7% from 2022. They are up 13.3% from precincts where Harris won 55% to 70% of the vote and 10.6% where she won 70% or more. They are down 1.6% where she got under 30%.
Early ballots are down from the most Republican precincts and up most from Democratic ones
Change in accepted early ballots, Oct. 24, 2026 compared with Oct. 29, 2022, by the Democratic share of the 2024 presidential vote in each voter’s precinct. Gwinnett is left out.
Source: Georgia Secretary of State ballot records, voter files and 2024 precinct results; Civic Forge Solutions analysis.
Voters from precincts Harris won with 55% or more are 41.5% of early voters this year, up from 39.3% in 2022. Voters from precincts where she got under 30% are 25.3%, down from 27.2%. Averaged over all early voters, that makes them about 1.1 points more Democratic than in 2022, the smallest of our three measures.
Three ways of measuring, one direction
A precinct gives everyone on the same street the same value, whatever their own politics. So we also estimated how Democratic each early voter is in three ways, and compared this year’s average with the average at the same point in 2022, 10 days before Election Day. This year’s early voters look more Democratic all three ways: by 3.2 points on our voter-profile measure, which uses age, race, sex, county and past primaries; 2.1 points on age, race, sex and county alone; and 1.1 points by neighborhood. None moved more than two tenths of a point since Wednesday.
Early voters look more Democratic than in 2022 all three ways, about as much as on Oct. 21
Change in early voters’ average Democratic leaning from the same number of days before Election Day in 2022. Gwinnett is left out.
Source: Georgia Secretary of State ballot records and voter history; Civic Forge Solutions analysis.
When we tested it on 2022, the voter-profile measure, the most detailed of the three, came closer than the other two to each county’s early vote, on average weighted by ballots, and it moves the most. The neighborhood measure misses differences between neighbors and moves the least, but it rests on actual votes rather than estimates. The demographic measure falls in between.
Fewer early voters with a Republican primary history
The voter-profile measure moves the most because it knows which party primaries each voter has taken part in. Through 2022, 32.4% of this year’s early voters had voted only in Republican primaries, against 40.0% of early voters at this point in 2022. The share who had voted only in Democratic primaries fell less, from 38.0% to 35.2%. In 2022 the two groups were about even (0.95 to 1); this year it is 1.08 to 1.
The share of early voters who had voted only in Republican primaries is 7.5 points lower than in 2022
Primary voting through June 2022, the same cutoff for both years. Gwinnett is left out.
Source: Georgia Secretary of State ballot records and voter history; Civic Forge Solutions analysis.
Who is voting early
Compared with this point in 2022, this year’s early voters include more Black, Hispanic and Asian voters and fewer white voters. Both older and younger voters make up more of the total: voters 65 and older and voters under 45 each have a larger share. Nearly everyone has voted in person: mail and electronic ballots are 4.8% of early ballots, against 9.5% in 2022.
White voters are a smaller share of early voters than in 2022
Shares of accepted early ballots, Gwinnett left out. Race and age are from the October 2026 voter file for both years.
| Share of early voters | 2022, through Oct. 29 | 2026, through Oct. 24 | Change, points |
|---|---|---|---|
| Black | 30.7% | 32.7% | +2.0 |
| White | 59.1% | 54.5% | −4.5 |
| Hispanic | 1.5% | 2.8% | +1.2 |
| Asian | 1.5% | 2.1% | +0.6 |
| Women | 55.2% | 55.6% | +0.5 |
| Under 45 | 19.8% | 21.4% | +1.5 |
| 65 and older | 41.3% | 43.7% | +2.3 |
| Voted in person | 90.5% | 95.2% | +4.8 |
| Voted by mail or electronically | 9.5% | 4.8% | −4.8 |
Source: Georgia Secretary of State ballot records and voter files; Civic Forge Solutions analysis.
Mail voters look less Democratic than in 2022, by 3.4 points on our voter-profile measure, while in-person voters look 4.3 points more Democratic. Mail is a much smaller share of early ballots this year, so the in-person change outweighs it.
Metro Atlanta and other counties
Early voters in every metro Atlanta county look more Democratic than in 2022, by both the voter-profile and neighborhood measures. The heavily Democratic core barely moved: DeKalb by less than a tenth of a point and Clayton by 0.6. Among metro counties, the biggest changes are in the suburbs: Paulding (+7.9), Henry (+6.3) and Douglas (+6.3). Which counties are voting more makes almost no difference statewide (it is worth 0.1 points toward Republicans), so the change comes from who is voting within each county.
The biggest moves toward Democrats are in the suburbs
Change in early voters’ average Democratic leaning, Oct. 24, 2026 compared with Oct. 29, 2022, in percentage points. *Gwinnett is compared with all of its 2022 early voters.
Source: Georgia Secretary of State ballot records and voter history; Civic Forge Solutions analysis.
- Cobb’s early voters look 3.8 points more Democratic than in 2022, down from 4.4 on Wednesday but still more than the statewide 3.2. If voters there vote as in 2022, Ossoff would get 59.0% to 62.2% in Cobb, against Warnock’s 58.4%.
- Newton, Cherokee and Cobb have added early ballots fastest compared with 2022. In Cherokee and Forsyth, among the region’s most Republican counties, early voters look 2.9 and 4.5 points more Democratic than in 2022.
- Paulding is the only county here with fewer early ballots than in 2022, and it has the biggest move toward Democrats of these counties.
| County | Early ballots through Oct. 24 | Change from 2022 | More Democratic than 2022, voter profile (points) | More Democratic than 2022, neighborhood (points) | Warnock, 2022 | Senate scenario if voters vote as in 2022 |
|---|---|---|---|---|---|---|
| Fulton | 210,543 | +4.8% | +1.6 | +0.8 | 75.0% | 75.5% to 76.6% |
| DeKalb | 146,779 | +6.2% | +0.1 | +0.2 | 85.7% | 85.7% to 85.9% |
| Cobb | 146,746 | +16.4% | +3.8 | +1.0 | 58.4% | 59.0% to 62.2% |
| Gwinnett* | 140,622 | n/a | +3.6 | +1.2 | 60.4% | 61.2% to 64.0% |
| Cherokee | 54,017 | +17.1% | +2.9 | +0.8 | 30.2% | 30.7% to 33.1% |
| Henry | 48,291 | +12.9% | +6.3 | +1.3 | 64.9% | 65.9% to 71.2% |
| Forsyth | 46,673 | +6.2% | +4.5 | +0.9 | 33.2% | 33.8% to 37.7% |
| Clayton | 43,888 | +8.2% | +0.6 | +0.6 | 88.6% | 89.0% to 89.2% |
| Hall | 33,614 | +5.8% | +4.6 | +1.1 | 27.1% | 27.8% to 31.7% |
| Fayette | 30,030 | +5.6% | +4.2 | +0.9 | 48.4% | 49.0% to 52.6% |
| Coweta | 27,638 | +10.3% | +4.4 | +0.9 | 32.7% | 33.2% to 37.1% |
| Douglas | 27,216 | +10.8% | +6.3 | +1.7 | 66.0% | 67.2% to 72.3% |
| Paulding | 26,318 | −5.0% | +7.9 | +1.2 | 37.5% | 38.3% to 45.4% |
| Newton | 19,305 | +19.1% | +4.8 | +1.4 | 58.1% | 58.9% to 62.9% |
| Rockdale | 18,850 | +10.3% | +4.1 | +1.1 | 74.5% | 75.3% to 78.6% |
*Gwinnett: the state’s 2022 ballot records are missing most of the county’s in-person early ballots from Oct. 24, 2022 on, so its change in ballots is not shown and it is compared with all of its 2022 early voters. County scenarios use each county’s 2022 Senate result by how ballots were cast and its change on our two measures; every caution about the statewide scenarios applies.
Outside metro Atlanta, Chatham (Savannah), Richmond (Augusta), Athens-Clarke and Baldwin (Milledgeville) are still the only counties with at least 5,000 early ballots where early voters look less Democratic than in 2022 by both measures. By the voter-profile measure, all four moved back toward their 2022 levels this week. If voters vote as in 2022, Ossoff would get 71.9% to 72.4% in Athens-Clarke, against Warnock’s 72.5%, and 59.4% to 60.1% in Chatham, against 60.5%. The University of Georgia’s Tate Student Center opens as an early voting site in Athens Oct. 27 to 29, according to the county’s schedule.
Election Day voters are different
In each of Georgia’s last three general elections, early voters were more Democratic than Election Day voters. What changed was the size of the gap. In 2020 and 2022 it was wide. In 2024 it nearly closed: mail voting, which had been heavily Democratic, fell sharply, and Election Day voters were much less Republican than in 2020 and 2022.
The early vote was far more Democratic than Election Day in 2020 and 2022, but not in 2024
Democratic candidate’s share of the two-party vote in the top race on the ballot, by how the ballot was cast. Early ballots include in person and by mail. Election Day here leaves out provisional ballots; the scenario tables include them, so 2022’s Election Day share appears there as 42.1%.
| Election | Democrat’s share, early ballots | Democrat’s share, Election Day | Gap, points | Election Day share of all ballots |
|---|---|---|---|---|
| 2020, Senate (Ossoff) | 51.6% | 38.4% | 13.2 | 19.7% |
| 2022, Senate (Warnock) | 55.1% | 42.0% | 13.1 | 35.9% |
| 2024, President (Harris) | 49.7% | 46.3% | 3.3 | 23.3% |
Source: Georgia Secretary of State certified results; Civic Forge Solutions analysis.
In 2022, the last midterm, Warnock won 55.1% of the two-party early vote and 42.0% on Election Day, a gap of 13.1 points, and Election Day was 35.9% of all ballots. In 2024 the gap was 3.3 points and Election Day was 23.3% of ballots.
Will the gap look more like 2022 or 2024?
Signs it may be closer to 2024:
- Some voters who used to wait for Election Day, many with Republican primary histories, are now voting early. 190,775 of this year’s early voters outside Gwinnett, 11.8%, voted in the 2022 general election but not early, and more of them have Republican primary histories than Democratic ones (39.5% against 27.9%).
- Mail voting, the most Democratic way to vote in 2022, has shrunk to 4.8% of early ballots, from 9.5%.
Signs it may be closer to 2022:
- This is a midterm, and in the last midterm the gap was 13.1 points.
- Early voters are more Democratic than in 2022 even with those voters included. That may mean many Republicans are still waiting for Election Day, or that more Democrats are voting early than in 2022; the records cannot yet tell which.
If the gap is as wide as in 2022, Republicans can make up a large early deficit on Election Day. If it is as narrow as in 2024, the early vote is a much better guide to the final result.
In 2022 the gap was widest in Newton and narrowest in Forsyth. In 2024 it shrank everywhere except Clayton, and in Cherokee, Forsyth, Paulding and Hall, Election Day voters were more Democratic than early voters.
| County | Gap, 2022 (points) | Gap, 2024 (points) | Election Day share, 2022 | Election Day share, 2024 |
|---|---|---|---|---|
| Newton | 21.9 | 9.2 | 40.9% | 25.9% |
| Rockdale | 13.8 | 9.9 | 27.6% | 20.9% |
| Henry | 12.4 | 3.3 | 28.0% | 18.8% |
| Fayette | 12.0 | 3.5 | 31.5% | 16.8% |
| Gwinnett | 10.5 | 4.6 | 35.8% | 23.5% |
| Cobb | 10.2 | 4.3 | 33.9% | 21.1% |
| Coweta | 10.2 | 0.4 | 42.3% | 22.8% |
| DeKalb | 9.0 | 8.5 | 28.6% | 20.0% |
| Douglas | 7.0 | 1.8 | 29.5% | 21.5% |
| Paulding | 6.7 | -2.0 | 33.9% | 22.6% |
| Cherokee | 6.7 | -0.5 | 38.3% | 22.0% |
| Fulton | 5.8 | 1.5 | 27.0% | 17.2% |
| Hall | 4.2 | -4.5 | 34.1% | 20.4% |
| Clayton | 3.7 | 3.9 | 27.5% | 23.4% |
| Forsyth | 2.6 | -1.7 | 31.8% | 16.0% |