A new national survey gives Democrats an 11-point advantage over Republicans on the issue voters most often say is driving the 2026 midterms: the cost of living. But the same poll finds that roughly four in five registered voters do not see either major party as especially capable of solving important problems or focused on people like them.

The tension between those findings is the central message of the AP-NORC survey released Saturday. Among registered voters, 38% said Democrats would do a better job on the cost of living, compared with 27% who chose Republicans. Another 27% trusted neither party on the issue.

That result is a meaningful shift from February, when adults were about evenly divided between the parties. It also arrives one month before the Nov. 3 election, as control of both chambers of Congress remains contested. Still, the poll is not a forecast. It measures issue confidence among registered voters, not likely-voter turnout in individual House and Senate races.

An advantage without broad confidence

The Associated Press analysis found that Democrats have also improved their standing on the economy and hold advantages on health care and foreign policy. Yet only about two in 10 registered voters said either party was highly capable of solving important problems or focused on helping people like them.

Voters were more critical of Republicans on several measures. Fifty-five percent said the GOP was unable to solve important problems, compared with 48% who said that about Democrats. Six in 10 said Republicans were strongly focused on helping the rich and powerful, nearly twice the share who applied that description to Democrats.

But the Democratic advantage did not become a broad endorsement. Nearly half of Democratic voters said their own party was too moderate, while a similar share said it was about right. Overall, 55% of registered voters described the Republican Party as too extreme and 45% said the same of the Democratic Party. President Donald Trump was described as too extreme by 68%, including one-third of Republicans.

The findings are consistent with a separate Wall Street Journal poll that found 21% of voters held unfavorable views of both parties. That was up from 13% in the 2022 midterms and about 12% in the two most recent presidential elections. Together, the surveys suggest that movement away from Republicans may be producing reluctant Democratic support rather than durable party loyalty.

Prices remain the pressure point

Independent polling has tracked the same economic erosion. A Reuters/Ipsos survey conducted Sept. 17-20 found Trump’s approval among Republicans on handling the cost of living had fallen from 70% in January to 44%. His approval on the economy among Republicans dropped from 80% to 56%, while his overall approval among all adults stood at 32%.

The electoral stakes are visible in Ohio, where Trump is campaigning Saturday for Republican Sen. Jon Husted against Democrat Sherrod Brown. Reuters reported that gasoline prices in the state were 35% higher than a year earlier, averaging $4.16 a gallon, while Canadian retaliation against U.S. tariffs affected about $3 billion in Ohio exports based on 2025 figures. Ohio is one of eight Senate contests rated competitive.

Those conditions help explain why a national issue advantage matters even if it cannot predict a particular result. Voters experience prices locally, while candidates must defend national party records. Republican campaigns can still benefit from Trump’s ability to motivate infrequent supporters, and Democratic candidates must convert dissatisfaction into turnout.

The parties therefore face different strategic tests. Republicans need to persuade economically frustrated voters that recent price increases are temporary or outside the administration’s control. Democrats need to move beyond assigning blame and explain what a congressional majority would do about household costs. The survey’s large “neither” bloc means both arguments are still competing for voters who have not granted either side much credibility.

What the poll can—and cannot—show

The AP-NORC poll interviewed 2,140 adults from Sept. 24-28, including 1,766 registered voters, using NORC’s probability-based AmeriSpeak Panel. Interviews were conducted online and by phone. The margin of sampling error was plus or minus 3.1 percentage points for registered voters, meaning small differences can fall within statistical uncertainty. The 11-point cost-of-living gap is larger than that margin, though sampling error is not the only possible source of survey error.

Earlier Pew Research Center data showed Democrats leading the generic congressional ballot 43% to 37%, while 20% supported another candidate or were unsure. Pew cautioned that registered-voter preferences can change and that the electorate that actually votes in a midterm may differ from the broader registered population.

National polls also cannot capture the uneven geography of congressional elections. Senate control turns on a small number of statewide contests, and House outcomes depend on district boundaries, candidate quality and local turnout. A party can lead nationally while losing decisive races, particularly when its support is concentrated in already-safe districts.

That caution remains essential. The new poll does not establish that Democrats will win Congress, nor does it show that voters have accepted a Democratic governing program. It shows something narrower but consequential: Democrats currently have the clearer advantage on costs, while neither party has persuaded most voters that it can govern effectively. In the final month, the party that closes that confidence gap—not merely the party voters dislike less—may have the stronger finish.