The 2026 US Midterms: A Data-Driven Study of Record-Low Approval and a Looming Democratic Wave

Executive Summary

Ten days before the November 3 midterm elections, the empirical record describes one of the most lopsided pre-election environments in modern American polling history. Donald Trump's national approval has fallen to 31 percent in the Reuters/Ipsos survey released October 19 -- the lowest reading of either of his presidential terms -- while the New York Times/Siena College poll (34 percent, October 21) and the CNN/SSRS survey (32 percent) converge on the same verdict. Polling averages put roughly six in ten Americans against him: RealClearPolitics 38.4 percent approve to 60.4 disapprove, the New York Times average 36 to 61, and the Silver Bulletin 36.5 to 60.5. The Times notes it is the lowest pre-midterm standing for any president since Harry Truman's 33 percent in 1946.

The congressional ballot tells the same story. Democrats lead Republicans by 8 to 12 points across pollsters -- Reuters/Ipsos 47 to 35, the NYT/Siena 52 to 41 (a record 11-point margin), Pollsmax 50.6 to 42.6 -- and prediction markets now price a 91 percent chance of a Democratic House and a 68 percent chance of a Democratic Senate. The drivers are measurable and overlapping: a cost-of-living crisis that voters say has worsened on Trump's watch, a war with Iran that most voters call a mistake, and an immigration crackdown a majority say has gone too far. This study assembles the polling record, the economic data and the market signals into a single evidence-based assessment of what the numbers can -- and cannot -- tell us about November 3.

Methodology

This study rests on public polling and economic data collected between October 8 and October 23, 2026, with every load-bearing figure corroborated across at least two independent pollsters or wire reports. The core survey inputs are the Reuters/Ipsos national poll (1,679 U.S. adults, margin of error +/-2 percentage points, October 19), the New York Times/Siena College poll (October 21), the CNN/SSRS survey, the Washington Post-ABC News-Ipsos poll, and the Strength In Numbers/Verasight poll -- supplemented by the Civiqs state-by-state tracker (more than 137,000 responses) and the standard polling averages (RealClearPolitics, the New York Times, Silver Bulletin). Prediction-market data from the Kalshi election hub provides a real-money cross-check on the polling consensus. Figures are reported exactly as published by each pollster; where pollsters disagree on method or sample, the range is given rather than a false precision.

Two limits shape what follows. First, polls are snapshots of sentiment, not votes: a double-digit generic-ballot lead does not translate mechanically into seats, particularly in a Senate map where the Democratic margin narrows sharply when only the 35 states holding Senate races are counted. Second, this study treats polling error as a standing risk -- 2016, 2020 and 2022 each carried misses -- and weighs the findings against it rather than assuming the consensus is exact.

Empirical Findings

1. The approval collapse. Four major pollsters now place Trump's approval at 31 to 34 percent: Reuters/Ipsos at 31 percent (its lowest of either term, one point below the previous low two weeks earlier), CNN/SSRS at 32 percent (a record low, down one point from late September), the Washington Post-ABC News-Ipsos poll at 33 percent with 66 percent disapproving, and the NYT/Siena at 34 percent (the lowest in that poll's tracking). The state picture is equally stark: the Civiqs tracker shows Trump's net approval underwater in most states, including Texas (-21), Michigan (-27) and even historically Republican Kansas (-1). Only eight states still give him a positive net rating -- Alabama, Arkansas, Idaho, North Dakota, Oklahoma, Utah, West Virginia and Wyoming, all states he carried in three consecutive presidential elections -- with Wyoming (+14) his strongest and Montana and South Dakota dead even at 46-46.

2. The generic ballot. Democrats hold a national advantage of 8 to 12 points depending on the pollster: Reuters/Ipsos 47 to 35 among registered voters (the widest lead of the cycle, up from 44 to 37 two weeks earlier), the NYT/Siena 52 to 41 (an 11-point record margin), Pollsmax 50.6 to 42.6, and the 538-style average at roughly 8.6 points. The caveat is geographic: when the NYT/Siena asked only voters in the 35 states with Senate races, the Democratic lead narrowed from 11 points to five. The House, fought entirely on national terrain, is the chamber the numbers endanger most.

3. The economy as the decisive front. Only 28 percent of Americans approve of Trump's handling of the economy, down from 41 percent in February, and 48 percent say their financial situation has worsened since his return to office -- the highest reading in Washington Post-ABC surveys since 1986. The Verasight poll found 63 percent of respondents naming food prices and inflation among their top three problems, ahead of jobs, health care, democracy and immigration. The Associated Press reports annual inflation at 3.7 percent, up from 3 percent when Trump returned to office in January 2025, with unemployment at 4.2 percent and growth a little above 2 percent -- figures the White House cites as resilience, but which do not match household experience. Most strikingly, the Washington Post-ABC News-Ipsos poll found 71 percent of Americans regard petrol as unaffordable, compared with 71 percent who called it affordable in February, before the war with Iran began. On October 3 in Ohio, Trump declared "we have almost no inflation" -- the polling suggests voters have reached a different conclusion.

4. The war drag. The Strait of Hormuz crisis has moved from a foreign-policy file to a household-budget issue: 65 percent of respondents in the Washington Post-ABC News-Ipsos poll called the decision to take military action against Iran a mistake, and with Brent crude above $110 a barrel the fuel-price consequences are priced into the election. Trump campaigned as a president who would keep America out of costly foreign conflicts; the collision between that promise and the affordability pledge is now the central tension of his midterm case, as US politics coverage has tracked all autumn.

5. The immigration backlash. Immigration was supposed to be Trump's signature strength, but the methods have cost him the argument: 56 percent say he has gone too far in deporting undocumented immigrants. Reuters reported preliminary Department of Homeland Security data showing a 95 percent decline in migrants apprehended at the southwestern border versus the Biden administration -- a measurable result the White House cites -- but fatal encounters involving federal immigration agents in Maine, Minnesota and Texas, including the deaths of two U.S. citizens in Minnesota, have intensified scrutiny. Voters now prefer Democrats on immigration and deportations by a narrow one-point margin, per Verasight; the GOP retains a slim edge on crime and border security.

6. Intensity and intent. A Pew Research Center survey found 44 percent of registered voters characterize their midterm vote as a vote "against" Trump, and 78 percent of Democratic voters say it really matters which party wins, against 69 percent of Republicans -- the enthusiasm gap favors the opposition. Fifty-eight percent of registered voters want a Democratic-controlled Congress that can restrain Trump (Washington Post-ABC News-Ipsos), a demand for a check on the presidency rather than an endorsement of the Democratic platform, which remains unpopular even among Democrats themselves. The NYT/Siena poll's warning for the opposition is explicit: voters want a Congress that reins Trump in, but they have not decided the Democrats deserve their trust.

7. The money's verdict. Prediction markets, where traders risk capital on outcomes, broadly confirm the polling consensus. Kalshi's election hub prices a 91 percent chance of a Democratic House and a 68 percent chance of a Democratic Senate. The House map: 208 races deemed safe Democratic against 144 safe Republican, 18 likely Democratic to 34 likely Republican, 9 lean Democratic to 17 lean Republican, with five toss-ups. The Senate: 12 safe Democratic plus 2 likely Democratic against 13 safe Republican plus 1 likely Republican, with 4 lean Democratic, 2 lean Republican and a single toss-up -- Iowa, where Ashley Hinson faces Josh Turek. The most-traded Senate contract is Texas (James Talarico against Ken Paxton, more than $37 million in sales, traders favoring Talarico 68 to 32); the most-traded House contract is also Texas (Bobby Pulido against Monice De La Cruz), with Minnesota's Ilhan Omar against John Nagel the second most-traded -- a race shadowed by the federal investigation into Representative Omar. Democrats need a net gain of four seats in each chamber to take control, with all 435 House seats and 35 Senate seats on the ballot.

Outlook

The base case the data supports is a Democratic House majority of meaningful size and a Senate that leans Democratic but remains genuinely contested -- the chamber where the national polling margin narrows from eleven points to five when the map is restricted to its 35 races. Three variables could still move the outcome. The first is turnout: a 44-percent "vote against Trump" electorate that turns out at presidential-year intensity would overshoot the polls, while a disengaged Democratic base would undershoot them. The second is the price channel: petrol and grocery costs are the transmission belt between the Iran war, the Hormuz shipping crisis and the ballot box, and any sharp move in fuel prices in the final ten days would register directly. The third is polling error itself -- the industry's misses in 2016, 2020 and 2022 run in both directions, and a uniform error of even three points would redraw the Senate map while leaving the House forecast intact.

What the numbers cannot do is resolve the strategic question the Democrats face: how to convert a referendum on Trump into a mandate of their own. The polls are unambiguous that voters want a Congress capable of checking the president; they are equally clear that the Democratic Party has not earned voters' confidence on immigration, spending or foreign policy. The opposition's task in the final stretch is to answer the question the data keeps posing -- disapproval of the president is not a policy program -- in the competitive districts where House control will actually be decided. On November 3, the test is whether Trump's own closing argument, delivered at rallies like his Saturday appearance in Perry, Georgia, mobilizes his supporters or mobilizes the larger number of Americans who have already decided they want him checked.