The Washington PostDemocracy Dies in Darkness

Opinion What you can — and can’t — expect to learn from the North Carolina special election

President Trump speaks at a campaign rally in Manchester, N.H., on Aug. 15. (Patrick Semansky/AP)

On Wednesday morning, one of the major parties is going to be irrationally exuberant and the other will be despondent once the results are in from the special election in North Carolina’s 9th Congressional District. But trying to read the tea leaves is almost always a useless exercise. Historical data shows that individual special elections don’t have much predictive power. We can get a decent read of public opinion by aggregating the results of many special elections. But we don’t need to go through that math when numerous polls are already telling us that the GOP is in bad shape.

The narratives are easy to predict. If Republican Dan Bishop handily wins, Republicans will be ecstatic. They’ll claim that the polls are wrong about President Trump, and that both his brand and the GOP’s are in good shape. Trump will write a tweet that simultaneously congratulates Bishop and takes credit for his win. And Democrats will worry that the polls are systematically underestimating the president’s strength.

And if Democrat Dan McCready wins or comes close, the blue team will see it as yet another sign that Trump is destined for a 2020 loss. Some Republicans will worry that they made a mistake by running a proponent of the state’s infamous “bathroom bill.” And others will panic about the party’s ability to win even light red states like North Carolina next November.

But we don’t have to rewind very far to find an example of a hyped-up special election that didn’t tell us anything useful about the political future. In June 2017, the entire political world was laser-focused on Georgia’s 6th District — a suburban Atlanta seat vacated by Tom Price, Trump’s nominee for health and human services secretary. The matchup between Democrat Jon Ossoff and Republican Karen Handel was supposed to be a perfect early test of Trump’s electoral strength. Romney had won the district by more than 20 points and Trump had won by less than two. An Ossoff win would signal that Democrats were permanently taking the suburbs, and a Handel win was supposed to show that Trump could expand his coalition and win back Romney-Clinton voters.

Follow David Byler's opinionsFollow

But the race turned out to be a completely unreliable signal of the nation’s mood. Handel outperformed Trump’s margin and won the race, but Trump wasn’t able to expand his coalition in 2018. It turns out that Ossoff was a bad candidate, and Democrats took the seat when they ran gun control activist Lucy McBath in 2018. More importantly, Democrats won control of the House in November 2018 mostly by flipping Republican-held suburban seats.

That’s not to say that the special in Georgia’s 6th was all noise and no signal — the results did tell us something when they were placed in the right context. When analysts compared the Handel vs. Ossoff performance to the state’s partisan baseline — the measurement of how red or blue the district is in a year when the parties are roughly equally popular — and averaged that with a similar statistic for other special elections in Pennsylvania’s 18th District, Montana’s at-large district, South Carolina’s 5th and other races, we got useful results. Democrats were overperforming each district’s partisan lean by a wide margin on average, and that presaged the Democratic landslide of 2018.

So far, special elections haven’t told us too much about 2020. There’s only been one other House special election this cycle, and we can’t conclude much from just one race. Democrats have been outperforming their baseline in state legislative special elections by about five to six points. And as Nathaniel Rakich has pointed out, that’s not so different from their overall lead in the generic ballot — a poll that essentially asks voters if they’ll vote for Democrats or Republicans in the upcoming House elections.

And we don’t really need to look at special elections to get a good read of the national political mood — the polls are all basically saying the same thing. Approval polls show that Trump is highly unpopular, and generic ballot polls (which correctly forecast the 2018 House results) show that down-ballot Republicans aren’t performing particularly well, either. Trump is behind most of his potential Democratic opponents in national and swing state polls of the 2020 race. He’s just not that popular.

Trump could still win reelection. Maybe his Democratic opponent will be too far to the left for swing voters, or maybe Trump will be able to damage them through personal and policy attacks. There’s a lot we won’t know until the fall of 2020. But the special election in North Carolina’s 9th District — no matter how many ways we dissect it — won’t tell us what’s going to happen.

Read more:

Max Boot: A new poll shows voters aren’t buying what Trump is selling

Jennifer Rubin: Eight tips for Thursday’s Democratic debate

Arnold Schwarzenegger: Trump can’t erase a decade of clean air progress with a Sharpie

Greg Sargent: Americans are growing more accepting of immigrants. Take note, Democrats.

E.J. Dionne Jr.: Like it or not, electability matters

David Byler: The DNC had a smart plan for shrinking the Democratic field. What happened?