Source Notes｜2026年米中間選挙・候補者／団体／世論調査の比較補助資料
［初出日未確認・表示日は移行日］ This article is a pre-result study of the November 3, 2026 U.S. midterm elections. It begins with Joseph Ax's Reuters election-day report, but does not endorse a par
This article is a pre-result study of the November 3, 2026 U.S. midterm elections. It begins with Joseph Ax's Reuters election-day report, but does not endorse a party or convert polling into a declared result. It asks what different political actors believe is at stake, what empirical evidence supports their claims, and what remains uncertain.
The Election Is Not a Single Referendum
The House's 435 seats, 35 Senate seats, and many state offices and ballot measures are at stake. Democrats need net gains of at least three House seats and four Senate seats to control those chambers, according to Reuters. The White House frames the contest as a defense against candidates it calls radical. Democrats frame it as accountability for Trump's presidency, cost of living, healthcare and institutional checks. These are actors' competing frames, not neutral descriptions of the candidates.
The Reuters/Ipsos survey conducted October 29–November 2 reported 32% presidential job approval. Its likely-voter generic ballot was 52% Democratic/lean Democratic versus 44% Republican/lean Republican. These are a national survey and a likely-voter subsample, not constituency-by-constituency results. Pew's September 28–October 4 registered-voter survey gave Democrats 44% and Republicans 36%, with another 20% choosing someone else or unsure. The two studies measure different populations and cannot be averaged mechanically.
What Each Side Calls the Election
| Actor or group | Observable position or priority | Evidence / caveat |
|---|---|---|
| Donald Trump and Republican campaign | Portray many Democratic nominees as politically extreme and rally supporters to defend Republican control | Reuters records Trump's November 2 remarks; characterizations are campaign rhetoric, not factual verdicts |
| Democratic leaders | Cast the election as a check on the administration, affordability and oversight | Reuters reports prospective congressional oversight; allegations of wrongdoing require separate proof |
| Moderate or cross-party voters | Judge local candidates, practical costs and institutional checks rather than party identity alone | No single uniform independent-voter position; Reuters/Ipsos highlights anti-Trump motivation among independents |
| Progressive Democrats | In Michigan, Abdul El-Sayed campaigns on universal healthcare and criticism of Israel | Reuters; intraparty differences persist |
| Republican candidate Mike Rogers | Competes against El-Sayed in Michigan, supported unexpectedly by Democrat Haley Stevens | Reuters; cross-party endorsement is not representative of all Democrats |
| Texas Democrat James Talarico and Republican Ken Paxton | Competing visions of Texas representation; Paxton supported by Trump and heavily financed pro-Republican outside efforts | Reuters; expenditure and endorsement are distinct from voter preference |
| Susan B. Anthony Pro-Life America / Women Speak Out PAC | Oppose abortion-rights expansion and prioritize anti-abortion candidates | Organization's November 2 announcement reports an \$80 million electoral program and 10.5 million contacts; self-reported outreach does not prove persuasion |
| Planned Parenthood Action Fund | Supports reproductive-health access and endorsed candidates including Talarico and El-Sayed | Organization's public endorsement list; advocacy source, not independent polling |
| Human Rights Campaign | Prioritizes LGBTQ rights and endorses candidates through its electoral program | Its 2026 election reporting; not synonymous with all LGBTQ voters |
| Economic and labor-oriented constituencies | Affordability, jobs and healthcare are salient; positions vary across labor organizations and sectors | Pew, Gallup; avoid treating all workers as a political bloc |
What Voters Actually Say
In Pew's survey, the economy was very important to 84% of registered voters: 82% of Republican-candidate supporters and 88% of Democratic-candidate supporters. Immigration mattered very much to 72% versus 53%, respectively; healthcare to 53% versus 82%; violent crime to 64% versus 37%; climate and environmental policy to 15% versus 66%. Both sides name the economy but emphasize different threats.
Gallup, using a different threshold, found 50% rated the economy and 50% healthcare extremely important; 47% said Iran. These measures are not interchangeable. A Pew breakdown showed Democratic congressional preference 45% among women versus 31% Republican, 42%–42% among men; Hispanic voters 51%–24%, Black voters 70%–7%, and ages 18–29 51%–23%. Demography is not individual destiny and the remainder may be undecided or third-party supporters.
Forecasts Versus Outcomes
The October 30 Cook Political Report assessment shifted a dozen House ratings toward Democrats and projected a net Democratic gain of 12 to 25 seats, while describing seven Senate contests inside polling margins of error and Democrats as slight Senate favorites. Reuters reported a House advantage for Democrats and an uncertain Senate. These are probabilistic assessments. Turnout, district-specific races, model error, legal disputes and timing of counts can change the outcome.
Several Senate races reveal the limits of national narratives: Michigan's El-Sayed–Rogers contest and Haley Stevens' crossover endorsement; Alaska's Peltola–Sullivan race, including contested allegations about a same-name third candidate, which Peltola denied; Texas's Talarico–Paxton contest; Kansas's Adam Hamilton–Roger Marshall race; and Maine's Susan Collins–Troy Jackson contest. An allegation, an investigation, a judicial dismissal and an established violation are different states of evidence.
Questions to Carry Past Election Night
Did the national generic ballot translate into district seats? Which voters turned out rather than merely expressed enthusiasm? Did affordability override identity and cultural issues? Did campaigns produce persuasion, mobilization, or only loud visibility? Did each party recognize its internal disagreement? Which legal claims were substantiated in court? What changed in legislative oversight, fiscal expectations, trade policy and energy geopolitics after the results?
The purpose is not to award virtue by affiliation. It is to place identical evidentiary demands on opposing claims, acknowledge asymmetries where documented facts warrant them, and preserve the difference between a political statement and an empirical finding.
Sources: Reuters election-day overview; Reuters/Ipsos survey; Pew issue priorities; Pew demographics; Gallup issue priorities; Cook Political Report; SBA Pro-Life America; Planned Parenthood Action Fund.
