WASHINGTON, UNITED STATES — President Donald Trump's approval rating stands at 35% as registered voters favor Democratic congressional candidates over Republicans by 44% to 37%, a new Reuters/Ipsos poll finds, with the cost of living dominating voters' priorities less than two months before the November election.
President Donald Trump's national approval rating has recovered from the low recorded by Reuters/Ipsos in August, but the same polling series shows Democrats entering the final stretch of the 2026 midterm campaign with an advantage among registered voters and with economic concerns weighing heavily on the electorate.
The Reuters/Ipsos survey conducted September 11-14 found that 35% of U.S. adults approved of Trump's performance as president and 62% disapproved. Among registered voters, approval was 39% and disapproval 61%. Ipsos interviewed 1,143 adults, including 896 registered voters, through its probability-based KnowledgePanel.
The congressional numbers point in a different direction from the modest improvement in Trump's personal rating. Asked how they would vote if the election for Congress were held now, 44% of registered voters chose the Democratic candidate in their district and 37% chose the Republican. Another 12% said they did not know, 4% said they did not plan to vote and 2% selected another party.
That produces a seven-percentage-point Democratic advantage among registered voters. It should not, however, be read as a forecast that Democrats will win either chamber. The question measures national preferences, while control of Congress will be decided through hundreds of individual House races and the Senate contests being held this cycle.
The federal general election is scheduled for November 3. All 435 House seats are contested in midterm elections, while roughly one-third of the Senate is elected every two years.
Cost of living is the top voting issue for nearly half of respondents
The poll provides a clearer indication of what is driving political pressure than the presidential approval number alone.
Forty-eight percent of adults identified the cost of living as the single most important factor in deciding their congressional vote, far ahead of democratic values and norms at 22%, immigration at 12% and crime at 5%. Among registered voters, the cost-of-living figure was 47%.
Trump's ratings on those pocketbook issues were substantially below his overall 35% job approval. Only 27% approved of his handling of the U.S. economy, while 65% disapproved. Approval fell to 22% on inflation and rising prices, against 70% disapproval, and to 21% on respondents' own cost of living, where 71% disapproved.
A newsroom comparison based on the Ipsos topline therefore puts Trump's approval on the cost of living 14 percentage points below his overall job approval, while his inflation rating trails it by 13 points.
Those differences matter because economic conditions are not simply one issue among many in this poll: respondents themselves rank living costs as the dominant consideration in their congressional vote.
The pattern is not new. In an earlier Reuters/Ipsos survey conducted July 29-August 3, 48% of adults also selected the cost of living as their most important midterm consideration. At that point Democrats led Republicans 35% to 27% when respondents were asked which party had the better approach to living costs.
Democrats edge Republicans on the economy but GOP retains advantages on immigration and crime
The September poll also tests party reputations issue by issue, showing a political landscape more complicated than the seven-point generic-ballot advantage.
Among registered voters, 38% said Democrats had the better plan, policy or approach to the U.S. economy, compared with 37% for Republicans — effectively a one-point difference and well within the poll's sampling error. Among all adults, Democrats led 35% to 33%.
The Democratic advantage was much wider on the cost of living: 39% of registered voters preferred the Democratic approach and 32% the Republican one. Democrats also led 43% to 31% on healthcare and 43% to 33% on democratic values and norms.
Republicans, meanwhile, retained advantages on two issues where Trump's own approval was stronger. Registered voters preferred the Republican Party to Democrats on immigration by 43% to 36%, and on crime by 43% to 29%.
Trump's issue ratings followed a similar hierarchy: immigration was his strongest measured issue at 39% approval among all adults, followed by crime at 37%. His ratings were markedly lower on foreign policy, Iran, the economy, inflation and cost of living.
The result is an electorate that does not give either party an across-the-board advantage. Democrats perform better on several economic and social-policy measures, while Republicans continue to lead on immigration and crime.
Trump's $5,000 payment proposal draws 63% opposition in the Ipsos survey
The poll also offers the first direct measure in the survey of public reaction to Trump's newly announced proposal to provide a $5,000 payment to adult Americans if Republicans retain both the House and Senate.
Seventy-four percent of respondents said they had heard at least a little about the proposal, including 28% who had heard a lot. When Ipsos described Trump's promise and asked whether it was appropriate for the president to propose the payment, 63% called it inappropriate, 17% appropriate and 18% were unsure.
Opposition crossed party lines, although by very different margins. The topline shows 89% of Democrats calling the proposal inappropriate, as did 61% of independents or respondents identifying with another party and 37% of Republicans. Among Republicans, 33% considered it appropriate and 30% were unsure.
The proposal is not an enacted federal benefit. Congressional approval would be required before federal money could be distributed, a distinction that is central to evaluating the promise. Reuters has estimated that paying $5,000 to roughly 240 million adults would cost about $1.2 trillion and reported that Trump has pointed to tariff revenue as a potential funding source.
That fiscal estimate remains separate from the Ipsos polling findings: the survey measures public reaction to the proposal, not whether Congress would enact it or how a final program would be financed.
Trump's 35% approval is up two points from his August low
The September result also provides evidence of a modest rebound in Trump's standing, though the size of the movement warrants caution.
Ipsos reported Trump's approval at 33% in its August 14-17 Reuters/Ipsos poll, which the company described as the lowest reading of either his first or second presidential terms. The September survey places him at 35%, a two-point increase.
A two-point difference is smaller than the September survey's plus-or-minus three-percentage-point sampling margin of error for all adults. It therefore should not be interpreted by itself as evidence of a statistically definitive shift in national opinion.
The September survey also used a sample considerably smaller than some earlier Reuters/Ipsos polling. Ipsos says its 1,143-person national probability sample was weighted for demographic characteristics, 2024 presidential vote and party identification. The sampling margin of error is ±3.0 points for all adults and ±3.4 points for the 896 registered voters. Margins are larger for partisan subgroups.
That methodological distinction is especially important when reading small differences such as the 38%-37% Democratic edge among registered voters on economic stewardship.
Democrats show higher voting enthusiasm, but national polling cannot allocate House seats
The survey contains another potential warning sign for Republicans: Democratic respondents reported greater enthusiasm about participating in November.
Among Democrats, 59% described themselves as very or somewhat enthusiastic about voting in the congressional elections, compared with 45% of Republicans and 27% of independents or others. Among registered voters overall, 51% were enthusiastic.
On a separate turnout question, however, the partisan difference was narrower. Seventy percent of Democratic respondents said they were almost certain or very likely to vote, compared with 67% of Republicans.
Those measures help describe the electorate but cannot determine turnout. Nor can the national generic ballot translate directly into House seats. Congressional elections are decided district by district, while Senate control depends on the particular states holding elections and the candidates competing in them.
The FEC confirms November 3 as the regularly scheduled 2026 federal general-election date. Its election calendar also shows that the campaign has already entered federally defined pre-election periods governing electioneering and coordinated communications.
With seven weeks remaining, the Reuters/Ipsos poll establishes a national baseline rather than an election result: Trump remains substantially underwater in job approval, Democrats hold a measurable generic-ballot advantage among registered voters, living costs dominate voter priorities, and neither party has displaced the other across every major issue.
The unanswered question is whether those national attitudes will persist — and, more importantly, how they will translate into turnout and individual contests on November 3.