October 18, 2026 | Fieldwork October 9–13; national survey, not an election result
Two findings from an ABC News/Washington Post/Ipsos survey have drawn attention ahead of the U.S. midterm elections. Among U.S. adults, 33% approve of President Donald Trump's job performance and 66% disapprove. Among likely voters, 52% say they prefer the Democratic candidate in their congressional district and 44% favor the Republican candidate.
The findings are related but answer different questions about different groups of people. Combining them as though they measure the same thing would mislead readers. Neither figure guarantees which party will control the U.S. House of Representatives after the November election.
What the poll actually measured
The survey was conducted October 9–13, 2026, among 2,545 U.S. adults through Ipsos' KnowledgePanel. People are recruited to the panel through probability-based address sampling, rather than simply choosing to answer an open website poll. Respondents were interviewed in English and Spanish, and the results were weighted to reflect characteristics of the adult population and relevant election measures.
The reported overall margin of sampling error is approximately ±2 percentage points for adults and ±2.2 points for results among registered and likely voters. Those margins represent only one component of survey uncertainty; wording, nonresponse, turnout modeling and opinion shifts can also matter.
The research was sponsored by ABC News, The Washington Post and Ipsos. Their methodology and questions are publicly documented.
Presidential job approval: 33% among adults
Asked whether they approve of Trump's performance, 33% of U.S. adults said yes and 66% said no. In the registered-voter subset, the poll reported 37% approving and 63% disapproving. These are not contradictory figures; registered voters are a narrower group than all adults.
The poll also found substantial differences by political identification. Among respondents identifying as Republicans, 79% approved, while among independents 22% approved. Those contrasts help explain why the national figure alone says little about the strength of support within the president's own party.
House vote preference: an eight-point national difference
Among likely voters, the poll measured 52% preference for Democratic House candidates and 44% for Republican candidates. Among registered voters more broadly, the comparison was 50% to 42%. Such a question is often called the generic congressional ballot: it asks which party's candidate someone would support in their own district, without necessarily naming a local candidate.
An eight-point national difference can be an important indicator of the political climate. But the U.S. House is chosen across hundreds of individual districts, not by a single nationwide popular vote. Candidate quality, incumbency, local events, redistricting, campaign strategy and turnout influence the outcome.
It also is not appropriate to take the eight-point gap and predict an exact number of seats. That requires district-level data and a modeling approach the poll itself does not provide.
Economic dissatisfaction and possible influences
The survey found widespread concern about the cost of essentials. Seventy-one percent of adults said gasoline was unaffordable, and 21% approved of Trump's handling of the cost of living. The same poll reported dissatisfaction on inflation, healthcare affordability and household budgets.
Those results describe how participants perceive conditions. They do not, by themselves, establish the precise causes of price changes or prove that any one issue will decide voters' choices.
ABC's analysis also showed differences in enthusiasm: 72% of Democratic voters and 65% of Republican voters said they were absolutely certain to vote. This is a self-reported intention, not verified turnout, and actual Election Day participation could differ.
What competing political interpretations miss
Democrats can reasonably point to the lead in national House preference as evidence of public dissatisfaction. Republicans can point to continued high Trump approval within their own party, unresolved voters, campaign differences across states and the uncertainty of turnout. Both are relevant to understanding the survey, but neither makes the November result certain.
A fair reading also avoids claiming the survey represents every registered voter equally. The approval figure is from adults; the House figure cited in the headline is from a likely-voter model. That distinction belongs next to the numbers, not hidden in a methodology note.
What to watch before Election Day
Other high-quality polls, especially consistent trends over multiple field periods; district-level surveys; early-voting and final turnout data; and material changes in economic or political events will all add useful information. Individual poll results become more informative when viewed alongside methods and wider evidence.
Reader question
If national public opinion and actual House seats can diverge, which measure should headline coverage emphasize—and why?
