The Washington PostDemocracy Dies in Darkness

Trump can’t blame Democratic crossover voters for his Georgia failures

Georgia Secretary of State Brad Raffensperger (R) speaks at a news conference in Atlanta on Dec. 14, 2020. (Kevin D. Liles for The Washington Post)

In the aftermath of the Georgia primaries last week, an intriguing question has arisen. Former president Donald Trump’s endorsed candidates for governor and secretary state were beaten handily — surprisingly badly, in fact. In the secretary of state race, incumbent Brad Raffensperger (R) was able to eke past the 50 percent margin needed to avoid a runoff despite having run fairly close in polling with Trump’s pick for the job.

Take away a bit over 55,000 votes and Raffensperger drops below 50 percent of the votes cast. As we pointed out, far more than 55,000 2022 Republican primary voters had voted in the Democratic primary in 2020. Did Raffensperger avoid a runoff election thanks to Democrats showing up to support him?

Trump, never one to miss an opportunity to deflect blame, suggested at a rally in Wyoming that this is precisely what happened. It’s an appealing frame for him, certainly. But data released this week from the state (from Raffensperger’s own office, as it happens) shows that it’s not actually what happened.

Sign up for How To Read This Chart, a weekly data newsletter from Philip Bump

The most recent data suggest that nearly 67,000 people who cast ballots in the Republican primary last month had voted in the Democratic primary in 2020. But we can break out those numbers further, looking at how those 67,000 voters had voted in prior party primaries. And when we do that, we get a better answer to our original question.

There are essentially four groups of voters here. They are voters who voted in the Republican primary in 2022, the Democratic primary in 2020 and:

  • No other primaries (about 19,300 of the total),
  • At least one Democratic primary and no Republican primaries from 2014 to 2018 (about 17,400 voters),
  • At least one Republican primary and no Democratic primaries (about 17,000 voters), or
  • At least one Republican and one Democratic primary over those three cycles (about 12,900 voters).

Now, everyone we’re talking about here is a “swing voter” in the sense that they have voted in one primary from each party over the past three years. But within that group, we have two groups that deviated from their partisan norm only once. Those are the “consistent” partisans identified on the chart above. (Those are meant to be scare quotes.)

This includes the “consistent” Democrats, people who voted in two or more Democratic primaries from 2014 to 2020 and the Republican primary this year. This is the group that we can comfortably describe as Democrats who switched over just for this cycle.

Many of the group that voted in no primaries from 2014 to 2018 might also be Democrats who wanted to influence the Republican primary. They have no prior record, perhaps because they are new voters. But, as you’ll see, the uncertainty of this group doesn’t really matter.

After all, Raffensperger avoided a runoff by a bit more than 55,000 votes. If we assume that all of those who didn’t vote in any primary before 2020 were Democrats and add in the “consistent” Democrats, we land at fewer than 37,000 votes in total. Even if all of those people voted for secretary of state (which is itself unlikely), removing them isn’t enough to push Raffensperger below 50 percent.

All of the other 30,000 Georgians who voted in the Democratic primary in 2020 and the Republican one in 2022 (the “consistent” Republicans and the swing voters) voted in at least one Republican primary from 2014 to 2018. These are not well described as Democrats hoping to influence the other party’s primary. Even if you count anyone who voted in a Democratic primary over that period alongside the “consistent” Democrats and those without a prior primary vote, you still end up short of 55,100.

(A procedural point: Some analyses have considered the important figure here to be a bit under 28,000 — about half of 55,100. If votes flipped from Raffensperger to Trump’s endorsed candidate, Rep. Jody Hice, that would make sense. But the question here is turnout, people voting at all. Not how they voted.)

There were clearly some Democrats who voted in the Republican primary, perhaps to hand Trump a loss. But the available data suggests that they were not the difference maker in Raffensperger’s race.

Just to be completist about it, they were not the difference in Gov. Brian Kemp’s reelection bid, either. He won the Republican primary by more than 625,000 votes.