Pivot Extra: From ‘Trump Curious’ to ‘Trump Furious’: Inside the Midterm Odds

Kara Swisher opens by noting that analysts at the Cook Political Report see Democrats with a chance to win the House, while the Senate has moved from strong Republican confidence to a competitive contest. Kristen Soltis Anderson says the biggest surprise is not a particular Democratic gain but the absence of the normal Republican recovery. In the past two midterms, Democratic numbers were strong during the summer and then narrowed toward Election Day. Republicans have become somewhat more enthusiastic than they were over the summer, but Democrats have been energized for more than a year, and the expected GOP return to the race has only partly happened. David Wasserman says the competitive map has expanded beyond the familiar narrow band of districts. Democrats are now polling competitively in places that Trump carried by 15 or even 20 points in 2024, including Kentucky’s Sixth District, New York’s 21st, western North Carolina, Nevada’s Second, and possible Senate races in Kansas and Nebraska. Terrance Woodbury says enthusiasm can understate turnout among young voters and voters of color. His firm tracks whether people believe their vote has power; that figure had fallen below 50% among young voters during 2025 but has risen toward 60%, alongside stronger early voting among young voters of color in Georgia and North Carolina.

Asked why Cook moved 11 more House races toward Democrats, Wasserman says Republican candidates without strong personal brands are being dragged down by Trump’s low approval. Republicans such as Susan Collins, Ashley Hinson, and Brian Fitzpatrick remain competitive because voters know them independently of Trump, but lesser-known candidates are struggling. The Democratic candidate Jamie Ager, a farmer, is ahead in a western North Carolina district Trump carried by roughly 10 points. Wasserman says Republicans’ redistricting efforts may produce only three or four additional seats, while more than 20 Republican-held seats are now toss-ups or worse. His median projection is a Democratic gain of 15 to 20 House seats. For the Senate, Democrats first need to hold their own seats, with Michigan the most vulnerable because Abdul El-Sayed must broaden his appeal after a divisive primary. North Carolina is described as the best Democratic pickup opportunity. Ohio and Alaska have become the next-best targets, followed by Texas and Maine; Iowa, Kansas, and Nebraska are longer shots. Wasserman says the outcome could range from Democrats winning 48 or 49 seats to reaching 54 on an excellent night. Anderson cautions that undecided voters could still change individual races, while Wasserman notes that roughly 8–9% of battleground voters dislike Trump’s job performance but have not committed to Democrats.

Anderson defends polling by distinguishing Trump presidential elections from midterms. Polls were off in 2016 and 2024 partly because Trump mobilizes voters who are difficult to survey, while polls naturally overrepresent politically engaged people. That bias can be useful in a midterm, because politically engaged respondents are more likely to vote. She says the current polls generally have the expected balance of education, party identification, and likely voters, but worries that Democrats are often the first and most eager people to answer surveys. Undecided voters also matter: in Virginia, Echelon modeled the electorate correctly but underestimated Democrat Jay Jones because undecideds ultimately broke for him. Woodbury adds that the “Trump effect” changed both turnout and vote choice. From 2020 to 2024, men of color shifted 16 points toward Trump, and Hispanic men under 50 supported him at majority levels for the first time in Woodbury’s experience. Many of these “Trump-curious” voters are now disillusioned because Trump’s approval has collapsed. The panel discusses prediction markets such as Kalshi and Polymarket, which have attracted more than $220 million in 2026 midterm bets. Woodbury is suspicious of their regulation and accuracy; Anderson says they serve a different purpose from polls, since campaigns use polls to decide how to change outcomes, not merely predict them. Wasserman says professionals also see private party and super-PAC data that ordinary bettors may not have.

Ben Wickler asks which messengers and media voters trust in an environment where people distrust institutions and tune out political communication. Woodbury says lower-propensity midterm voters trust people like themselves—friends, family, peers, and selected influencers—more than politicians, celebrities, clergy, or pollsters. He argues that campaigns should equip trusted influencers and young people to become their own messengers, because following, liking, and resharing content demonstrates a preexisting relationship of trust. Anderson agrees that there is no single messenger. Older voters still consume local television news, while other voters use influencers and online personalities, but information often works by confirming an existing belief rather than because the source has earned universal trust. Wasserman says Democrats benefit from being the out party: their candidates can adapt messages to local districts, whereas the national Republican message that the economy is “great” conflicts with what voters see in grocery stores and at gas pumps. He describes the Democratic brand as weak—roughly 30% favorable—but says Trump keeps shifting attention back to himself. In his tracker, Trump is underwater 37% to 61%, while the generic congressional ballot has Democrats ahead about 50% to 41%.

The panel describes the election as an affordability contest involving healthcare premiums, housing, gas, diesel, and energy. Wasserman says voters link the administration’s foreign-policy decisions to domestic inflation and that roughly 69–70% in surveyed battlegrounds believe U.S. involvement in Iran has not been worth it. Anderson says corruption is nearly as important as affordability: voters worry that wealthy insiders and people close to power receive special deals. She also sees diminishing returns from spending; while ads can introduce unknown candidates, Michigan voters have already heard an overwhelming amount about Mike Rogers, James Talarico, and other candidates. Data centers have become a proxy for broader concerns about energy bills, corporate power, and insiders receiving favorable treatment. Republican Zach Nunn—or, as the transcript identifies him, Zach L—ran for Iowa governor on a data-center moratorium even though that conflicted with the Trump administration’s support for AI infrastructure. Woodbury says young voters increasingly identify corporations and special interests as the main cause of their pain, above either party or Trump. Wasserman says data-center opposition could become a major 2028 test, especially for Democratic governors trying both to attract technology jobs and satisfy voters worried about concentrated wealth and local costs. In Iowa, Ashley Hinson appears stronger than the Republican gubernatorial nominee because she presents herself as bipartisan, despite votes supporting the reconciliation bill and opposing tariff-condemnation resolutions.

In Texas, the panel sees an unusually competitive contest between Republican Ken Paxton and Democrat James Talarico. Anderson’s focus groups with 2024 Trump voters who now reject Paxton found two objections: disappointment with Trump’s economic and immigration record, and strong personal dislike of Paxton, whom participants called a “crook” or “criminal.” Some said they would support a normal Republican but could not vote for Paxton. Woodbury says Talarico’s response to attacks—that voters are being directed toward the wrong one percent while billionaires cause their pain—is resonating. He also says Republican attacks on gender and sexuality are less sophisticated than Trump’s 2024 messaging. Woodbury describes a “Blacklash” after redistricting and the Supreme Court’s decision affecting Louisiana’s second majority-Black district and the Voting Rights Act. Black voters increased their share of early voting in Georgia and North Carolina and became more likely to post, organize, boycott, petition, and attend rallies. Black seniors prioritized defending American democracy, while fewer than half of Black Gen Z respondents said that was their main cause; younger voters emphasized Black voting rights, racial justice, and freedom. The speakers say Gen Z distrusts both parties and wants systemic change rather than ordinary policy adjustments. Wasserman says AOC could inherit Bernie Sanders’ movement; he names Georgia’s John Ossoff as another possible Democratic 2028 contender. Anderson sees JD Vance as the Republican favorite but says Marco Rubio and Ron DeSantis deserve attention.

Asked to predict the midterms, Wasserman’s base case is a Democratic gain of roughly 15–20 House seats and a 51–49 Democratic Senate, with Texas and Maine the races he finds hardest to call. Woodbury predicts 20 or more House seats and a 52-seat Democratic Senate, including pickups in Texas, North Carolina, and Ohio. Anderson also expects Democrats to take the House and thinks a narrow Senate majority is more likely than not, though she is not confident about which states produce it. For 2028, the speakers expect a crowded Democratic primary in which a candidate able to consolidate 20–30% of a passionate base could prevail. Wasserman says AOC could inherit roughly a third of the Democratic Party’s progressive movement; Woodbury says the left’s performance in the midterms and especially the result in Michigan will shape electability arguments. Anderson says Gen Z has grown up amid Trump, COVID, and permanent political conflict, making its relationship to institutions fundamentally different from millennials’ experience. She expects Vance to lead the Republican field but notes Rubio’s rise and unexpected support for DeSantis in focus groups. The conversation closes with Wasserman hoping for relatively quick results, while Woodbury warns that election disputes may continue in courts and other venues after Election Day.