Tools chosen: polymarket_direct kalshi_related polymarket_related claude_news gdelt_news claude_news code_execution

Sub-questions (Fermi decomposition)

Planner reasoning

This is a Polymarket question on 2026 House control, so the Polymarket price is the primary anchor and Kalshi's equivalent market is an independent crowd check. Key drivers are the historical midterm penalty against the president's party, the generic ballot margin, redistricting changes (mid-decade gerrymanders in TX, CA, MO, etc.), special election overperformance, and Trump's approval rating. News and forecaster commentary will capture redistricting and retirement developments that historical base rates miss.

# Current state Republicans currently hold a narrow House majority (roughly 217-220 to Democrats' 213-215, depending on vacancies), meaning Democrats need a net gain of only ~3 seats to win control in the 2026 midterms. Generic ballot polling shows a mid-single-digit Democratic lead, Trump's approval is near record midterm lows, and redistricting fights (TX/MO/NC/OH GOP gains vs. CA Prop 50 Democratic response) are roughly offsetting nationally, though final maps and legal status remain partially unresolved. # Timeline of key events - 2025 (various): Texas, Missouri, North Carolina, Ohio, Indiana Republicans initiate mid-decade redistricting aimed at net GOP seat gains (reported, Brookings/NBC). - 2025-11 (approx.): California voters approve Proposition 50 (64.4% support), replacing independent-commission maps with a Democratic-favorable map through 2030, intended to offset GOP gains elsewhere (confirmed, Wikipedia/PBS/Al Jazeera). - 2025-12-05: U.S. Supreme Court allows Texas's new GOP-favorable map to be used for 2026, pausing a lower-court block on racial-gerrymandering grounds; long-term validity beyond 2026 unresolved (confirmed, Votebeat/Houston Public Media). - 2026-08 (ongoing): Generic ballot averages show a stable Democratic lead (Silver Bulletin D+7.6 likely-voter-adjusted; RealClearPolling D+6.4; Decision Desk HQ D+6; Ballotpedia D+6 in April) (reported, multiple aggregators). - 2026-08: Trump approval sits at 33-38% with net approval around -18 to -24, among the worst pre-midterm readings since Carter; brief improvement in July reversed (reported, NYT/Silver Bulletin/Reuters-Ipsos/DDHQ). - 2026-08-26/27: Maryland court blocks a Democratic-backed redistricting ballot amendment from appearing in November, a setback to further Democratic seat-gain efforts (confirmed, multiple outlets). - 2026-08-28: Fox News reports Trump's redistricting push has "netted Republicans 10 seats" — a partisan framing that conflicts with more neutral analyses (e.g., Brookings) estimating net effects closer to a wash; treat as rumored/contested, not established. - 2026-08-24: Reports on House retirements ("members leaving after 2026") circulate but specific partisan breakdown not detailed in available research (reported). # Event Will Democrats win a majority of U.S. House seats (or control via Speaker election) after the 2026 midterms? # Outcomes to forecast - Yes (Democratic control) - No (Republican control) # Kalshi market anchor Direct Kalshi YES price for this exact ticker was not returned in raw research (kalshi_direct tool output absent). Related Kalshi proxies: Speaker-of-the-House market has Jeffries (Dem) at 79% (down 4pts in 7 days); 2028 Democratic House control contract at 76% (implying continued/expanded Dem control baseline); Polymarket's identical-question contract trades at 89.5% (up from 80.5% low, +3pts over 30 days), $5.68M volume — this is the closest available cross-venue anchor and should be treated as the primary consensus proxy in absence of direct Kalshi data. # Sub-question answers 1. **Generic ballot trend** — Democrats lead by ~6-7.6 points across major aggregators (Silver Bulletin D+7.6 adj., RealClearPolling D+6.4, DDHQ D+6, Ballotpedia D+6), a lead that dipped to ~D+4 in July before recovering by August 2026 [claude_news]. 2. **Redistricting net effect** — GOP maps in TX (~+2 realistic net per Brookings, despite +5 target), MO (+1), NC (+1), OH (potentially +2) are roughly offset by CA Prop 50 (~+5 Dem seats); net national effect is close to a wash, though contested (Fox News claims GOP nets +10, likely partisan overstatement) [claude_news, Brookings, gdelt]. 3. **Historical base rate** — President's party has lost House seats in 18/20 midterms since 1946 (90%), averaging a 25.6-seat loss; P(losing ≥3 seats, the GOP's threshold) ≈ 85-90% historically, though the narrow-majority subset (3 comparable cases) shows a more mixed 33% flip rate [code_execution]. 4. **2025 special/off-year election overperformance** — Not found in available research; explicit data gap [claude_news]. 5. **Trump approval** — 33-38% approval, net approval -18 to -24, among worst pre-midterm ratings since Carter; comparable to or worse than 2018 (Dem +40 seats) and better than only Bush 2006 (Dem takeover with 11.4pt generic ballot lead) [claude_news, NYT/Silver Bulletin/Reuters-Ipsos]. 6. **Forecaster ratings (Cook/Sabato/Split Ticket)** — Not retrieved in this research pass; explicit data gap, though secondary headline states "Democrats Favored to Win Both House and Senate" (Political Wire, 2026-08-11) [gdelt]. 7. **Polymarket vs Kalshi divergence** — Polymarket YES = 89.5%; direct Kalshi price unavailable in raw data, but related Kalshi Dem-House proxies (76-79%) suggest Kalshi may be pricing somewhat lower than Polymarket, indicating possible cross-venue disagreement of roughly 10-13 points. # Key facts (high-confidence, factual) 1. [code_execution] Since 1946, president's party lost House seats in 18/20 midterms (90%); GOP needs to lose only 3 seats to lose majority. 2. [claude_news] Trump approval ~33-38%, net -18 to -24 (Aug-Sept 2026), among worst pre-midterm approvals post-WWII. 3. [claude_news] Generic ballot: Democrats lead by ~6-7.6 points across major trackers, as of Aug 2026. 4. [claude_news/Wikipedia] CA Prop 50 passed 64.4%, effective through 2030, redraws map to favor Democrats by up to 5 seats. 5. [Votebeat] SCOTUS allowed TX GOP map to stand for 2026 (Dec 2025 ruling), pausing lower-court block. 6. [Polymarket] Polymarket YES price 89.5%, up from 80.5% low over 91 days, $5.68M volume. # Cross-market signals - Kalshi related: Speaker Jeffries (Dem) contract at 79% (down 4pts/week); 2028 Dem House control contract at 76%. - Polymarket: Identical-question market at 89.5% YES, rising trend (+3pts/30d), high volume/liquidity. - Sportsbook implied: None available. - Note: Polymarket (89.5%) noticeably higher than Kalshi-adjacent proxies (76-79%) — potential venue divergence or proxy-market noise; direct Kalshi price for this ticker unavailable. # Analyst opinions and speculation - Political Wire (Aug 2026): "Democrats Favored to Win Both House and Senate." - The Conversation: Trump's near-record-low approval paired with favorable Dem generic ballot signals Congress flip likely. - Brookings: skeptical that TX redistricting alone secures GOP seats; could backfire in bad environment. - Fox News (partisan-leaning): claims redistricting nets Republicans 10 seats — likely overstated per neutral analysts. # Directional lean per outcome - **Yes (Dem control)**: Strong support — historical base rate (90% out-party gains), sizable generic ballot lead (D+6-7.6), historically weak Trump approval, redistricting roughly neutral, both prediction markets pricing Dem control as strong favorite (76-89.5%). - **No (GOP control)**: Support — extremely thin margin needed to defend (only need to hold ~3 net seats), GOP redistricting gains in multiple states, close-majority historical subset shows more mixed 33% flip rate, some approval recovery in July 2026, legal/redistricting uncertainty could still shift maps. # Gaps / unknowns - No direct Kalshi YES price for this specific ticker was retrieved. - No 2025 special-election/off-year overperformance data (VA, NJ) obtained. - No current Cook Political/Sabato/Split Ticket seat-by-seat ratings or retirement counts by party. - Final legal status of multiple redistricting maps (TX post-2026, Maryland blocked) still evolving. # Calibration anchors - Polymarket YES (same question): 89.5%, trending up. - Kalshi proxy markets: Speaker Jeffries 79%, 2028 Dem House control 76%. - Historical base rate: ~85-90% probability out-party (Dems) gains ≥3 seats in midterms; narrower-majority precedent shows more uncertainty (33% flip in 3 historical analogs). - 2018 precedent (comparable approval/ballot environment): Democrats gained 40 seats with D+7.6 generic ballot lead under a similarly unpopular Trump.

1. Both forecasters land at ~86-88%, essentially splitting the difference between Polymarket (89.5%) and Kalshi proxies (76-79%), but neither seriously grapples with why Kalshi proxies (which are not identical questions—Speaker election and 2028 control differ mechanically from "House majority after 2026") should be weighted so heavily; this risks anchoring on noisy, non-equivalent markets rather than the fundamentals-based estimate. 2. Both cite the "narrow-majority historical analog" showing a 33% flip rate from only 3 comparable cases, yet neither adjusts much for it — a reference class this thin (n=3) is nearly uninformative and could be given more explicit weight-discounting language, or alternatively flagged as too small to matter, but they hedge in between without justification. 3. Neither forecast meaningfully incorporates the missing special-election/off-year overperformance data (explicitly flagged as a gap in the brief) or Cook/Sabato/Split Ticket seat-level ratings — these are typically the most predictive inputs for House control and their absence should widen uncertainty bands more than a few points, yet both converge tightly around 86-88%. 4. The brief notes Jeffries Speaker contract is "down 4pts in 7 days" and generic ballot dipped to D+4 in July before recovering — both are recent-momentum signals suggesting possible erosion in the Democratic position that neither forecaster discusses as a risk to the trend continuing through November 2026 (the event is over a year away from the "current" data point). 5. Both treat Fox News's "+10 GOP seats" redistricting claim as dismissible partisan spin, but neither considers that unresolved litigation (TX map "long-term validity beyond 2026 unresolved," Maryland amendment blocked) could shift net redistricting effects unfavorably for Democrats between now and the election, understating tail risk on the "No" side. 6. Given the base rate (85-90%) and generic ballot/approval fundamentals point to a very similar range as the market prices, there's a risk of circular reasoning: both forecasters use the fundamentals to validate the market price and the market price to validate the fundamentals, without an independent structural check (e.g., seat-by-seat Cook ratings) that could reveal the "No" case is stronger than either acknowledges given only ~3 seats needed and multiple competitive tossups.

87% (70% confidence)

Yes: 87% No: 13%

Both forecasts converge at 86-88%, consistent with the fundamentals: D+6-7.6 generic ballot lead, Trump approval near record midterm lows (33-38%), a ~90% historical base rate of president's-party seat losses, and Democrats needing only ~3 net seats with redistricting roughly a wash. The closest cross-venue anchor (Polymarket, identical question) sits at 89.5% and trending up, while the lower Kalshi proxies (76-79%) are mechanically different questions and warrant only modest downward weight. The critique correctly notes missing seat-level ratings, special-election data, and unresolved map litigation, which justifies shading slightly below Polymarket rather than matching it — but these gaps cut both ways and don't reveal a shared directional flaw. Final: 0.87 Yes, blending the market anchor with the strong fundamentals while preserving meaningful downside risk from the narrow-majority precedent and time remaining until November 2026.

Total pipeline time: 199.4s