Close Menu

First Midterms Crack: Republicans Take the Lead Where It Actually Counts

POLITICS

For most of this cycle, the political class has repeated the same comforting law of physics: the party out of the White House gains in the midterms. It’s the closest thing American politics has to gravity. So the newest Harvard-Harris survey is worth a second look, because it shows that gravity, at least for the moment, has stopped pulling in the expected direction.

Among likely voters, the poll puts Republicans ahead of Democrats on the generic congressional ballot, 51 to 49. Two points is not a landslide, and no single survey should be mistaken for one. But the number matters less than the direction it’s moving, and the direction is unmistakable.

Track the same pollster’s likely-voter measure across the year. In March, Democrats led by four. By July, that lead had shrunk to two. In August, it flipped — Republicans by two. That’s a six-point swing over five months, and it’s reportedly the first time any pollster has shown Republicans leading likely voters since October of last year. A trendline like that, from a single high-quality outfit using consistent methodology, is more telling than any one-off result, because it filters out the noise and shows the underlying movement.

The likely-voter figure is also the one to watch. The same poll still shows Democrats up two among registered voters, and partisans will point to that. But registered-voter samples routinely overstate the side less likely to actually turn out, which is why campaigns and serious analysts weight likely voters more heavily as an election nears. Roughly seven in ten respondents said they intend to vote, with Republicans slightly more motivated than Democrats — exactly the enthusiasm profile a party wants heading into the fall.

There’s also a history-of-polling point that cuts the same way. One analysis of more than 3,000 surveys found that polling has overestimated Democrats by an average of 3.7 points across the last four cycles, with the summer error running closer to six. If that pattern holds, this poll may not be an outlier at all — it may be the leading edge of the usual autumn correction, as summer’s Democrat-friendly samples drift back toward reality.

The honest caveats belong right here, not buried at the bottom. It is one poll. It is early. Nothing about a September number guarantees a November result, and the only survey that counts is the one taken at the ballot box. Overconfidence has cost the right before, and a two-point lead is an argument for working harder, not coasting.

But note what would have to reverse for Democrats to feel good again. They’d need the trend to stop, the likely-voter screen to be wrong, the historical polling bias to vanish, and their own enthusiasm gap to close. That’s a lot of things breaking their way at once.

The party out of power is supposed to be cruising right now. Instead, on the measure that best predicts real outcomes, it’s trailing — and losing ground. That’s the story worth watching between now and November.