This month, Georgia voters sent two Democrats to the U.S. Senate after runoff elections, breaking with a decades-long history of electing Republicans. Raphael G. Warnock and Jon Ossoff’s victories were particularly surprising, since Republicans usually have an advantage in off-cycle elections, which usually bring out an older, whiter and more conservative electorate (not just in Georgia but nationwide) than those who show up in November.

How did both Democratic candidates overcome that disadvantage? Our analysis finds two notable patterns that could account for the wins: high turnout among Black voters and low turnout in the predominantly White and rural precincts that had supported incumbent President Donald Trump — where voters were more likely to believe Trump’s baseless allegations of electoral fraud.

How we did our research

The Georgia secretary of state’s office provides detailed data on the characteristics of who turns out to vote. We used Georgia’s recently updated data, which includes comprehensive information about who voted in the Jan. 5 runoff election, including self-reported information on the race and ethnicity of nearly every registrant.

Georgia does not have voters register by party. To examine the likely partisanship of those who did and did not vote, we examined the smallest geographic unit that reports election results: the precinct. We looked at each of Georgia’s 2,656 precincts’ November and January results, examining what proportion of the precinct’s registered voters cast a ballot in each election. To measure each precinct’s attitudes toward Trump, we looked at the proportion of its ballots cast for Trump and other Republican candidates in November.

Black turnout usually drops more than White turnout from the general election to the runoff

As was true in every state, record numbers of voters in Georgia cast ballots in November. Of the over 5 million Georgians who voted, 63.1 percent were non-Hispanic White, 30 percent were Black, 3.5 percent were Hispanic/Latino, and 3 percent were Asian.

Historically, turnout drops dramatically in Georgia statewide runoffs — particularly for minority voters. This hurts Democrats: In 8 out of 9 previous statewide runoff elections in Georgia, Democrats did worse than in the general.

In 2018, Georgia had a runoff election to decide whether Republican Brad Raffensperger or Democrat John Barrow would replace now-Gov. Brian Kemp as Georgia secretary of state.

The November election was close, with Raffensperger winning 49.1 percent to 48.7 percent for the Democrat. In that runoff, voter turnout dropped across the board: Only 37 percent of people who voted in the 2018 general showed up for the runoff. But minority turnout was hit hardest: The share of voters who were Black dropped from 30.5 percent in the general to 28.6 percent in the runoff. With White voters increasing their share from 62.6 percent to 66.3 percent, Raffensperger’s vote share grew nearly two points, putting him into office.

But this year, Black turnout dropped less than White turnout

This year, the opposite happened. Record-breaking spending and grass-roots, Black-, Latino-, Native American- and Asian American-led efforts to mobilize voters meant that early voting alone exceeded total turnout for any previous runoff election. Runoff turnout was 89.6 percent of November’s showing, with over 4.4 million Georgians casting ballots.

Black turnout was 91.8 percent of that in November; White turnout was lower, at 89.5 percent of the November total. The share of voters that were Black in the runoff thus increased to 30.7 percent. If Black voters had shown up in roughly the same proportions as they had in the 2018 runoffs, Ossoff’s 55,000 vote victory would have been a roughly 30,000 vote loss; the Warnock-Loeffler race would likely have been mired in a controversial recount; and Republicans would control the Senate.

Some of Trump’s strongest supporters stayed home in the runoff

For months, Trump refused to concede the presidential election to Democrat Joe Biden. In particular, Trump and his allies claimed that Georgia’s results were fraudulent.

Did this discourage Trump’s supporters from voting? Precincts that had supported Trump at high rates in the general election reported more dramatic turnout declines than precincts in which Trump was less popular. For example, in precincts where Trump received 75 percent of the November vote, turnout dropped 8.3 percentage points from the general election to the runoff. Precincts where Trump received 25 percent of the vote showed a drop of only 5.9 percentage points from November to January.

When we used other ways to measure Trump’s popularity in each precinct, we found similar results. For instance, in the November elections, then-Rep. Douglas A. Collins (R-Ga.), a fierce Trump supporter, was a third candidate running in the special election against incumbent Sen. Kelly Loeffler (R) and Warnock. In precincts where Collins won a higher percentage of the vote, people turned out at lower rates in the runoff.

In contrast, precincts where Loeffler, the more traditional candidate, had performed better against Collins in November, recorded smaller drops in turnout. Similarly, if we look at the November Senate race among then-Sen. David Perdue (R), Democrat Jon Ossoff and Libertarian Shane Hazel, there were also large declines in turnout in the precincts where Trump had received more Republican votes against Joe Biden than Perdue had gotten against his opponents.

Taken together, these patterns suggest that Trump’s strongest backers failed to vote in the runoff.

In elections decided by narrow margins, small differences in turnout by race and political attitudes can be enormously consequential. That was true in Georgia this month. Strong Black turnout, motivated by years of organizing by groups such as the New Georgia Project, and Trump’s discouraging rhetoric about voter fraud made the difference in handing both Georgia Senate runoffs — and therefore control of the U.S. Senate — to the Democrats.

Bernard L. Fraga is an associate professor of political science at Emory University and author of “The Turnout Gap: Race, Ethnicity, and Political Inequality in a Diversifying America

Zachary Peskowitz is an associate professor of political science at Emory University and a W. Glenn Campbell and Rita Ricardo-Campbell National Fellow and Starr Foundation Fellow at the Hoover Institution.

James Szewczyk received his PhD in political science from Emory University.

More on the Georgia elections: