For those watching the election, the more pressing question was how the rain might affect voting.
Polling indicated that Democrat Jon Ossoff had the advantage in early voting — that is, votes cast before Election Day proper. As it turned out, that was right. Ossoff won 53.3 percent of the early vote — more than predicted by a Trafalgar Group poll conducted over the weekend.
With that lead among voters who cast ballots before the rainy Election Day, it would suggest that Republican Karen Handel would need to do better on Election Day itself. Trafalgar figured she would, by a 12-point margin, enough to give her a not-huge-but-comfortable victory. Trafalgar figured Ossoff would get about 42 percent of the vote that day, with 3 percent of voters still undecided.
Handel overperformed that estimate, earning a 16.4-point victory in Tuesday’s vote. The victory Trafalgar predicted was more comfortable than expected.
In 2007, the Journal of Politics published a paper from researchers at the Universities of Georgia, Pittsburgh and California at Merced. They analyzed presidential voting in counties from every election from 1948 to 2000, comparing turnout to precipitation records from Election Day. Their conclusion?
Every inch of rain suppresses turnout by a bit under one percentage point. What’s more, since higher turnout correlates to more Democratic votes — since Republicans tend to be more reliable voters — bad weather has a bigger negative effect on the Democratic candidate than the Republican. The researchers estimated, for example, that slightly less rain in Florida on Election Day 2000 would have handed the election to Al Gore.
But that’s an extreme example. It’s not clear that the torrents in Georgia on Tuesday had a direct effect on Ossoff at all.
First, the estimated turnout decline is for the entire voting population, not just the Democrats.
Second, the era of early voting means that Ossoff was able to bank a number of votes before the weather turned south. Those who vote early tend to be more partisan and, therefore, more likely to vote anyway. But, more important, heavy early voting wasn’t a part of the weather study; clearly more votes cast beforehand would reduce the effects of bad weather on Election Day itself, to an unknown degree.
Third, Politico’s Bill Scher made an interesting observation about turnout on Tuesday. Last November, the Democrat running against incumbent Tom Price (whose elevation to President Trump’s Cabinet prompted the special election) won 124,917 votes. Ossoff got 124,983. Each of the other seven candidates in the four contested special House elections this year earned fewer votes than their party’s candidate last November. Only Ossoff was at the same level.
Fourth, Ossoff’s final vote percentage was 48.1 percent — precisely what he earned in the primary, when the Republican candidates combined for about 51 percent of the vote. Handel got 52 percent.
Now, it’s possible that Ossoff surged after the primary and held a decent lead coming into Election Day, only to see it washed away. But that theory is undercut by polling that showed Handel closing the gap over the last few weeks of the race. Oh, and by the Trafalgar Group polling cited above.
Dekalb-Peachtree Airport is in the southeastern part of the district, in Ossoff’s political stronghold, DeKalb County. In the primary, votes from DeKalb made up 22 percent of the total, with Ossoff winning 58.4 percent of the total. It contributed 27 percent of all of the votes he received.
In Tuesday’s rainstorm, DeKalb made up … 23 percent of all votes cast, with Ossoff winning 58.6 percent of the ballots. Twenty-eight percent of his votes came from the county.
In other words: Don’t blame it on the rain.