Editor's note: This article was also published in The Surveyor, YouGov America's email newsletter. Subscribe to The Surveyor for regular updates on YouGov's polling.

More Americans disapprove than approve of Donald Trump's job performance by a 24-percentage-point margin, and Trump's net job approval has been below -10 in every Economist / YouGov Poll since June 2025. But over that time, those same polls have asked how people expect to vote in the 2026 congressional elections if faced with unnamed Democratic and Republican candidates, and consistently found that Democrats are leading Republicans nationally by single digits — and often the low single digits. Even as Trump's job approval has declined, the race for Congress has barely budged.

There's a strong historical correlation between presidential job approval and congressional election results. So why has the polling about the U.S. House election diverged so much from Trump's job approval this election cycle? The discrepancy appears to be driven largely by three different groups, for their own reasons.

Before getting into these three groups of voters, it is worth noting one other source of the discrepancy. YouGov generally tracks job approval among all U.S. adult citizens, but vote intent among the smaller population of registered voters. These two populations have slightly different views of Trump: Registered voters are slightly less likely than all citizens to disapprove of Trump's job performance.

But Trump's approval among registered voters has declined at the same time as his approval among all citizens has fallen, even though the decline among registered voters has been slightly less steep. There's still a lack of correlation between the change in Trump's net approval among registered voters and their 2026 vote intent.

The Economist / YouGov Poll has been asking about registered voters' 2026 vote intention since February 2025. Since then, Trump's approval has fallen dramatically, but the Republican share of expected 2026 vote is down only slightly. That's because an increasing share of voters disapprove of Trump but still say they'll vote for Republicans for Congress. In February 2025, 40% of registered voters said they approved of Trump and would vote Republican, while 1% disapproved of Trump and would vote Republican. This month, 34% of registered voters say they approve of Trump and will vote Republican, while 6% disapprove of Trump and will vote Republican. The share of registered voters who approve of Trump but plan to vote for a Democrat has changed little, to 2% from 3%.

To better understand this phenomenon, we aggregated several recent internal YouGov surveys, in order to have a larger sample to examine subgroups. These combined surveys, fielded July 9 - 16 with 4,379 respondents, found that 58% of registered voters strongly or somewhat disapprove of Trump's job performance, but only 44% of registered voters both disapprove of Trump and say they'd vote for a Democrat for Congress. Another 5% of registered voters disapprove of Trump but would vote Republican, while 9% disapprove of Trump and say they would either vote for a third-party candidate, aren't sure how they'd vote, or wouldn't vote.

This 9% of registered voters who disapprove of Trump but aren't committed to either party in the congressional elections could be a crucial group of swing voters — one with several subgroups of swing voters. YouGov worked with an AI tool to conduct a cluster analysis of these voters and found that they fall into three groups, which could be called the following:

  • Blue leaners: Otherwise reliable Democrats who haven't decided to vote for a Democrat for Congress (about 40% of the larger group of swing voters who disapprove of Trump but aren't planning to vote for either party in the House elections, or about 3.5% of registered voters)
  • Disengaged voters: Though left-leaning in their politics and hostile to Republicans, this group rarely votes or even follows news about current events (about 20% of the group, or about 2% of registered voters)
  • Red leaners: Members of this group generally vote for Republicans and identify as conservative or moderate, but disapprove of Trump and remain on the fence about 2026 (about 40% of the group, or about 3.5% of registered voters)

The margin of error for these combined polls is around ±2 points for the whole sample, and larger for small subgroups, so the numbers for these small subgroups are approximate.

Among these voters who disapprove of Trump but are on the fence about 2026, the Blue and Red leaners are relatively similar in age, education, and interest in public affairs. The disengaged voters, in contrast, are much more likely to be young, to lack college degrees, and to not follow the news very closely.

So why are Democrats not doing better despite Trump's low approval ratings? With regular polling of these questions and the cluster analysis, we can say that, about 40% of it is anti-Trump Republicans who haven't said they will vote for Democrats. About 20% of it are disengaged voters who dislike Trump but rarely show up to vote. And the final 40% are otherwise Democratic-leaning voters who for whatever reason are not saying they'll vote for Democrats for Congress.

As both parties build their fall campaigns, these three groups seem likely to make their decisions about whether to vote, and who to vote for, for different reasons.

There are a lot of unanswered questions about these groups, what they believe, and how they'll change as November nears. What questions do you want to know about these registered voters who disapprove of Trump but aren't on board with Democrats? Let us know in the comments and we'll consider them when drafting future surveys.

Methodology: This article includes results from a series of online surveys conducted July 9 - 16, 2026 among 4,379 U.S. adult citizens. Respondents were selected from YouGov’s opt-in panel to be representative of adult U.S. citizens. The sample was weighted according to gender, age, race, education, 2024 presidential vote, 2020 election turnout and presidential vote, baseline party identification, and current voter registration status. 2024 presidential vote, at time of weighting, was estimated to be 48% Harris and 50% Trump. Demographic weighting targets come from the 2019 American Community Survey. Baseline party identification is the respondent’s most recent answer given around November 8, 2024, and is weighted to the estimated distribution at that time (31% Democratic, 33% Republican). The margin of error for the overall sample is 2 percentage points.

Image: Getty (Kevin Dietsch)

What do you think about the election, American politics in general, and everything else? Have your say, join the YouGov panel, and get paid to share your thoughts. Sign up here.

Subscribe to the YouGov newsletter