Henry Olsen is a senior fellow at the Ethics and Public Policy Center.
Eight months before the midterms, it’s clear that President Donald Trump is unpopular with Americans. Much less understood, however, are consistent indications he’s not as unpopular with the Americans who are going to matter in November.
Right now, the gap isn’t large — but if it holds, even a small difference could make all the difference in the reshaped political landscape that will emerge for the last two years of the president’s term.
The big picture isn’t in dispute. Political analysts are correct to argue that Trump’s low job approval ratings, if they persist till November, probably will hurt the Republican Party in the midterms. A president’s job approval is a strong predictor of midterm results, which usually means bad things for the president’s party when it’s as low as Trump’s is. But this year, a problem arises because virtually all analysts use an approach that understates Trump’s true standing among voters.
Most commentators and strategists look to one of the many good aggregators that take publicly available polls and use them to construct a job approval index. That’s the right way to do it, since any single poll can easily return an outlier finding.
However, overall job approval, though commonly and prominently tracked, is not the figure to watch ahead of the midterms. The better indicator is job approval among registered and likely voters. Usually, these two aren’t all that far apart, but right now there is separation. And because of that, the aggregators are currently overestimating Trump’s political weakness.
We can see this by constructing our own polling index. If we start by taking the long-standing RealClearPolitics index, but remove polls that only sample all adults, an altered picture emerges. The traditional average showed Trump with a 43.4 percent job approval rating as of Tuesday. But strip out three polls of all adults, which only give him a 39.7 percent rating, and he rises to 44.3 percent.
That change isn’t large, but exit polling shows that upward of 90 percent of voters who approve of a president’s job performance vote for that person’s party in congressional elections. On Election Day, a one-point increase in job approval could easily mean the difference for a few House seats or even the crucial race that determines control of the Senate.
Going back to 2012, the difference between the traditional RealClearPolitics polling average and mine has never been more than 1.5 points on Election Day.
But the gap between Trump’s approval rating in polls sampling all adults and those surveying registered or likely voters has persisted since it emerged last summer.
It can be found, in fact, even within a single survey when the pollster tracks both measures. The Economist/YouGov poll, for example, does that. Since September, this poll has consistently found Trump to be polling three to four percentage points higher among registered voters than among the entire adult sample.
The upshot is that Trump’s registered-voter approval has been consistently higher than his overall approval. Pollsters don’t know for sure why this is, but it’s not hard to come up with a plausible theory. Trump’s unprecedented public crackdown on illegal immigration has probably soured many noncitizens on him. The Pew Research Center estimates there were about 28 million noncitizen immigrants in the United States as of 2023, a figure that’s likely even higher today.
Noncitizens cannot and do not vote, but they are part of any survey sampling all adults. The nearly 30 million noncitizen residents comprise nearly 10 percent of the country’s population. So, a large drop in Trump’s popularity among them would have a significant impact on his overall rating, even though that drop would have no effect on the election.
It’s also worth noting that registration and voting rates are much lower among Hispanic and Asian citizens than they are among White and Black ones. That means that some of the groups most likely to feel directly affected by Trump’s immigration enforcement policies are also the likeliest to be screened out of polls sampling only registered or likely voters.
For all these reasons, I plan to focus heavily on polls and indexes that measure Trump’s job approval rating among voter samples as the midterms approach. If the gap widens between those results and his overall rating, it could mean a better night than generally expected for the GOP. If it narrows, and his overall rating is flat or gets worse, then the chances of a blue wave probably increase.
A lot could change between now and Election Day. Will this effect fade if outgoing homeland security secretary Kristi Noem’s aggressive public enforcement approach is replaced by a quieter one? Could some other major developments jumble the numbers in an unexpected way? I’ll be watching closely.
The post The hidden reason Republicans could outperform midterm expectations appeared first on Washington Post.
