The special election in Georgia’s 6th Congressional District has moved into extra innings. In the primary on Tuesday, Democrat Jon Ossoff won the most votes, with 48 percent, but fell just short of the 50 percent necessary to win outright. He’ll face Republican Karen Handel, who finished second with 20 percent of the vote, in a runoff election on June 20. So what should we make of Round 1? What do Tuesday’s results tell us about the coming runoff and the 2018 midterms? Here are five takeaways.
1. The polls were good
Ossoff earned 46 percent, on average,1 in polls taken in the three weeks leading up to the election. That’s just 2 percentage points off the final result. Given that special election polling has tended to be pretty error prone, the polls in Tuesday’s race were nothing short of amazingly accurate. That doesn’t guarantee that polls for the June runoff will be nearly as predictive. But for a first high-profile test of polling since the 2016 election, when surveys underestimated Donald Trump’s support, Georgia 6 is not consistent with the idea that polling is broken, or that polls are systematically missing Republican support.
Polls are imperfect tools and have a margin of error, but the Georgia 6 surveys were good. (Even the 2016 polls were much better than the conventional wisdom suggests.)
2. The early vote was not predictive
While no poll showed Ossoff getting to 50 percent (even with undecided voters allocated), he was dominating the early vote. This led to a lot of excitement among Democrats. Indeed, Ossoff ended up winning over 60 percent of the early vote that was reported Tuesday night, and there was speculation on Twitter for a bit that he would win Round 1 outright. The election day vote, however, was far less favorable for Ossoff. The 11 GOP candidates on the ballot in Georgia 6, combined, took around 60 percent of it. That pattern mirrored what we saw in 2016, when Hillary Clinton cleaned up among early voters in key swing states that she nevertheless went on to lose. The fact remains that a vote cast on election day is worth the same as a vote cast early. So add the Georgia 6 Round 1 results to our ever-growing pile of reasons to take early vote results with a salt mine’s worth of salt. It’s not predictive.
3. The result is consistent with a pro-Democratic national environment
There’s some chatter out there that Ossoff’s showing is a bad sign for Democrats. He didn’t clear 50 percent, they say, and he barely improved on Clinton’s performance in Georgia 6.
I think that’s a flawed argument.
For one, Clinton had already greatly improved on previous Democrats’ performance in Georgia 6. She lost to Trump there by only 1.5 percentage points. Former President Barack Obama lost the district by 23 points in 2012, as did Democratic congressional candidate Rodney Stooksbury in 2016.
So if you’re just looking at the 2016 presidential result as your benchmark you’re probably missing something. Instead, our best estimate of the partisan lean of a district is to take a weighted average2 of its past two presidential election results. By that measure, a Democrat would be expected to lose Georgia 6 by 9.5 percentage points in a neutral national environment (one in which the two parties fought to a tie nationally). Democrats did far better than that on Tuesday, losing by 2 points. The Democratic candidates combined took 49 percent to the Republicans’ 51 percent.
Related: Politics Podcast
4. What this means for the 2018 midterm is less clear
The Republican +2 aggregate margin in Georgia 6 implies a national environment in which Democrats are competitive in a bunch of GOP-held House seats in 2018. According to the weighted average of the past two presidential elections, there are 48 House districts that were won by GOP candidates in 2016 that are bluer than Georgia 6. The district’s Round 1 results suggest Republicans could lose a good portion of those 48 seats. And Democrats need to win just 24 Republican-held seats for control of the House.
That’s clearly a good sign for Democrats.
Of course, the national political environment could change between now and November 2018. Moreover, the Georgia 6 result isn’t anywhere near as strong for Democrats as last week’s result in the special election in Kansas’s 4th Congressional District. The Georgia 6 Democrats outperformed the weighted average by 7.5 percentage points. In Kansas 4, Democrat James Thompson beat it by 22 points.
Still, that’s the difference between a good Democratic year in 2018, with the House in play, and a crazily, ridiculously good Democratic year, with the House a foregone conclusion to flip to Democratic control. (Again, that’s if the national political winds don’t shift between now and then — an unlikely proposition.)
The truth is we need a larger sample size of special election results before understanding what Kansas 4 and Georgia 6 tell us about the midterms.
5. The runoff looks close right now
Polls conducted before Tuesday suggested that Handel and Ossoff were running in a near tie in a potential runoff. In an average of five polls conducted since mid-March, Ossoff held a scant 0.4-point lead over Handel. A formula created by my colleague Nate Silver based off previous runoff elections3 also points to a tight runoff. Combining the lean of Georgia 6 on the presidential level over the last two elections (9.5 percentage points more Republican than the nation), Ossoff’s margin over Handel (28 points) and the aggregate margin of the Republican candidates over the Democratic candidates (2 points), Handel is favored in the runoff by less than a point. With the relatively wide margin of error on this calculation, this is the equivalent of a tossup.
In other words, get ready for another two months of exciting Georgia 6 coverage! Tuesday was just the beginning.