Democrats hold a 12-point advantage over Republicans on the national congressional ballot, while independent voters favor Democratic candidates by 25 points, according to a new NPR/PBS News/Marist survey released Wednesday. The result is the latest evidence that President Donald Trump has become a drag on his party as voting begins in parts of the country.

Marist’s release found that 53% of registered voters would choose the Democratic candidate in their district if the election were held now, compared with 41% who would choose the Republican. Among independents, the split was 57% Democratic to 32% Republican.

The survey does not forecast how many seats either party will win. National preference can diverge from district-level outcomes because voters choose among individual candidates under maps that distribute each party’s support unevenly. Turnout, campaign spending and late events can also move the result. Even with those cautions, the size and composition of the advantage are politically consequential with six weeks remaining before Election Day.

Trump is a central factor

The poll suggests the midterm contest is functioning partly as a referendum on Trump. Forty-five percent of registered voters described their congressional choice as mostly a vote against the president, while 24% called it mostly a vote for him. Another 30% said Trump was not much of a factor. A majority of independents, 54%, said their vote was mainly against Trump.

The Democratic lead is not a sudden one-poll reversal. Marist reported a comparable 10-point advantage in May, meaning the party’s edge has persisted even as the campaign moved into its closing phase. What has sharpened is the evidence that the president is organizing voter sentiment: the share voting against Trump exceeds the share voting for him by 21 points.

That imbalance matters because both parties are largely holding their own supporters. The major movement is among voters who do not identify with either party. The detailed poll tables show a politically engaged electorate: 75% said they were following the midterms closely or very closely, and 86% said they cared a great deal or quite a bit about which party controls Congress.

Trump’s job approval stood at 39%, against 59% disapproval. Strong disapproval alone reached 53%, a sign that opposition is not merely tentative. His personal rating was similar: 37% favorable and 58% unfavorable. An NPR analysis noted that the strong-disapproval share was the highest Marist has recorded for Trump since it began asking the question in 2017.

Issue advantages have shifted

The generic ballot is reinforced by voters’ assessments of the parties on issues that Republicans have traditionally used to their advantage. Respondents preferred Democrats over Republicans on the economy, 42% to 34%; foreign policy, 42% to 36%; and gas prices, 40% to 27%. On immigration, voters were more closely divided, but Democrats still led 41% to 38%.

Independent voters widened those gaps. They preferred Democrats on gas prices by 26 points, foreign policy by 22 and the economy by 19. On immigration, they favored Democrats by 14 points, though 18% said neither party would do a good job.

Household sentiment remains difficult. Thirty-four percent said their family finances had worsened over the past year, while 20% said they had improved and 42% reported little change. Independents were especially negative: 41% said their finances had worsened, compared with 16% who said they had improved. That does not prove economic concerns caused the ballot preference, but the alignment helps explain why candidates are emphasizing prices and household costs.

The appetite for change extends beyond party labels. Fifty-six percent said they preferred candidates with new ideas, while 38% favored candidates with a proven record. That measure may cut in different directions from race to race—challengers can claim renewal, while incumbents can argue their experience is needed—but nationally it reinforces an unsettled, anti-incumbent mood.

A second poll points the same way

The finding is not isolated. A separate Economist/YouGov survey conducted Sept. 18–21 found Democratic candidates ahead 52% to 38% among likely voters, a 14-point margin. YouGov found independents favoring Democrats 52% to 27%, almost the same 25-point gap reported by Marist. Among all registered voters in that survey, however, the Democratic advantage was narrower at seven points, underscoring how screens and question wording can change toplines.

The Marist survey interviewed 1,280 registered voters Sept. 14–15 by live telephone interview, text and online. Its stated margin of error is plus or minus 3.8 percentage points. Subgroup estimates, including the independent-voter result, carry more uncertainty because they are based on fewer respondents. The PBS report presents the result as a significant lead, but neither the release nor the underlying data converts the national margin into a seat projection.

For Republicans, the immediate problem is broader than persuading voters to like individual candidates. The new surveys indicate that many independents view their congressional choice through Trump and currently prefer the opposition party. For Democrats, the risk is treating a national polling advantage as a guarantee in a collection of distinct state and district contests. The numbers describe a favorable environment, not a completed election.