Scan: 2026 Balance of Power: R Senate, R House

1. Planner Fable 5.1 · Fermi decomposition + tool plan

Tools chosen: polymarket_direct polymarket_related kalshi_related article_search claude_news wikipedia code_execution

Sub-questions (Fermi decomposition)

What is the current Polymarket price for 'R Senate, R House' and how does it compare to the implied product of the separate Senate-control and House-control markets on Polymarket and Kalshi?
What is Trump's current net approval rating and the generic congressional ballot margin, and how do these compare historically to midterms where the president's party held the House?
How many competitive Senate seats are Democrats realistically contesting (e.g., NC, ME, OH, TX, IA, AK) versus Republican pickup opportunities (GA, MI, NH), and what is the probability Democrats net +4?
How has mid-decade redistricting (Texas, Ohio, Missouri, Florida, California's Prop 50 response, Virginia, court rulings on VRA/Louisiana) shifted the House seat baseline for 2026?
What do special election results and off-year 2025 elections (VA, NJ governors, state legislative specials) indicate about the 2026 environment and Democratic overperformance?
What is the current House margin and how many Republican-held seats in Harris-won or close districts plus GOP retirements are in play?
Historically, how often has the president's party retained both chambers in a midterm when starting with a narrow House majority (<10 seats) and net-negative approval?

Planner reasoning

This is a joint outcome: Republicans hold both the Senate and the House in the 2026 midterms. The Senate map favors Republicans structurally (Democrats need net +4), while the House is the swing chamber where the president's party historically loses seats. The key is the House probability conditional on the Senate staying R, so we anchor on Polymarket, cross-check Kalshi and individual-chamber markets, and gather generic ballot, presidential approval, retirement, and redistricting news.

2. Research Parallel tool calls

tool	status	time	items	summary
polymarket_direct	OK	0.3s	1	## This Market's Polymarket Data 2026 Balance of Power: R Senate, R House - Current price (probability): 13.50% - 7-day price change: +2.00% - 30-day price change: +2.00% - Total volume: $2,936,397 (USD notional) - Price range: 10.50% - 19.50% - Data points: 91 days
polymarket_related	OK	0.2s	0	Scanned 100 active Polymarket markets, kept 0 matches. keyword 'House control 2026': 0 markets \| keyword 'Senate control 2026': 0 markets \| keyword 'balance of power 2026': 0 markets \| keyword 'midterms': 0 markets
kalshi_related	OK	6.8s	3	3 related markets / summaries. series KXHOUSE: 0 markets (skipped 0 no-signal) \| series KXSENATE: 0 markets (skipped 0 no-signal) \| series KXCONTROL: 0 markets (skipped 0 no-signal) \| keyword 'House control 2026': ok \| keyword 'Senate control 2026': ok \| keyword 'balance of power': ok
article_search	OK	7.8s	18	Fetched 18 unique articles across 5 queries (lookback=45d). '2026 midterm House forecast Republicans majority': 10 articles \| '2026 Senate map Democrats path to majority': 10 articles \| 'generic ballot poll 2026 midterms': 10 articles \| 'redistricting 2026 House seats Texas California court ruling':
claude_news	OK	38.0s	21	## 2026 Midterm Outlook Summary Generic Ballot & National Environment - Democrats hold a substantial generic ballot lead; Silver Bulletin's likely-voter adjusted average shows Democrats' lead at D+6.6 to D+7.6 as of late August 2026 (natesilver.net). Separately, Nate Silver's model puts the H
wikipedia	OK	6.8s	3	Fetched 3 Wikipedia entries (0 missing pages).
code_execution	OK	151.9s	0	## Key Findings Historical base rates (US midterms, 1946–2022, n=20 elections): - Full retention (both chambers) after midterm, all 20 elections as denominator: 4/20 = 20% ended with the president's party controlling both House and Senate (1950, 1962, 1966, 2002-equivalent unified outc

3. Evidence Brief Sonnet 5 · 7942 chars

# Current state As of research date (~Sept 2026), the 2026 midterms have not occurred; Election Day is Nov 3, 2026. Polling and forecaster models currently favor Democrats retaking the House (Silver Bulletin: "likely D") and give Democrats a slight edge in the Senate (~56.8% per Silver's FLIPR model), meaning the joint outcome "R Senate AND R House" is a minority scenario requiring both an upset House hold AND a Senate hold. # Timeline of key events - 2025-12-05: SCOTUS allows Texas's new GOP-favorable congressional map to stand for 2026 pending litigation (confirmed, votebeat.org). - 2025-12-08: NPR reports redistricting battle nationally could net GOP ~10-14 seats vs Dem ~9 seats, contingent on litigation (reported). - 2026 (date unspecified, pre-Sept): Two incumbent GOP senators (Cassidy-LA, Cornyn-TX) lose renomination primaries; nominee swaps occur in ME (Platner replaces someone) and SC (special primary replaces Lindsey Graham) (confirmed, Wikipedia). - 2026-08-03: CNN reports Democratic Senate targets (TX, GA, NC, ME, OH, IA, AK) showing competitive polling, e.g., Talarico (D) near-tied with Paxton (R) in TX (reported). - 2026-08-05: CBS reports House Democrats expanding target list to 12 new GOP-held seats; CBS poll shows 47% want Dem Congress vs 40% GOP (reported). - 2026-08-25: Cook Political Report shifts FL-14, MI-10 from Lean R to Toss-up; MI-08 to Likely D (reported). - 2026-08-30/31: CNN reports GOP's new maps could add up to 10 House seats; Trump approval reported 32-34 points underwater in CNN/SSRS poll (reported). - 2026-09-08/09: Cook moves TX Senate and House ratings toward Democrats amid worsening GOP environment; polling averages suggest Dems favored to retake House (reported). - Missouri Supreme Court (date unspecified, ~Sept 2026): blocks new GOP-favorable map, reverting to 2024 map — a setback for GOP redistricting gains (confirmed per Silver Bulletin summary). # Event Kalshi market resolves YES if Republicans win control of BOTH the US Senate and US House in the 2026 midterms; NO otherwise. # Outcomes to forecast - Yes (R Senate + R House) - No (any other combination) # Kalshi market anchor No direct Kalshi price provided in research (kalshi_direct not returned in raw data); only Polymarket data available for this exact contract. **Polymarket price for the identical event: 13.5% YES**, up +2% over both 7d and 30d, trading range 10.5%-19.5% over 91 days, $2.9M volume — treat this as the best available cross-market anchor for the Kalshi price. # Sub-question answers 1. **Polymarket price vs. implied product** — Polymarket direct price is 13.5%. No separate Polymarket/Kalshi Senate-only or House-only control markets were found (0 matches on searches), so a direct product comparison is unavailable. Code-execution modeling estimates a copula-adjusted joint probability of ~19-35% (vs naive independent product ~15-30%) using assumed marginals P(R Senate)≈0.75-0.85, P(R House)≈0.20-0.35 — suggesting Polymarket's 13.5% may be somewhat below model-implied ranges, or those marginal assumptions are too generous to Republicans. 2. **Trump approval / generic ballot** — Trump net approval ≈ -19.5 to -20.6 (Silver Bulletin average, Sept 2026), with CNN/SSRS showing -32. Generic ballot: Democrats lead by D+6.6 to D+7.6 (Silver Bulletin likely-voter), D+7.2 (Silver House popular vote forecast), D+4 (Morning Consult RV). Historically, a president's party rarely holds the House under such conditions (see code_execution base rates below). 3. **Senate seats in play** — Democrats need net +4 of 35 seats up (23 GOP-held). Competitive Dem targets: TX (Talarico vs Paxton, ~60.7% Dem win prob per Silver), GA, NC, ME, OH, IA, AK. Silver's FLIPR model (Aug 24) gives Democrats 56.8% Senate control probability vs Republicans 43.2%. 4. **Redistricting** — TX map upheld by SCOTUS (Dec 2025), could yield GOP +5 seats there; national redistricting could net GOP ~10-14 seats vs Dem ~9, per NPR — but Missouri's new map was blocked by state Supreme Court, reverting to 2024 lines, partially offsetting GOP gains. 5. **Special/off-year elections** — Raw research did not directly detail VA/NJ 2025 governor results or state legislative specials; no explicit data found (gap). 6. **House margin/GOP exposure** — Republicans hold a "razor-thin" majority and can lose only two seats before losing control (CBS News, Aug 2026); Cook Political Report lists 18 competitive toss-ups, with ratings trending toward Democrats through Aug-Sept 2026. 7. **Historical base rate (full retention)** — Per code_execution: conditional on already holding both chambers pre-midterm (n=11 cycles 1946-2022), full retention occurred in 4/11 (36.4%); under "stress" conditions (low approval or thin House margin), retention rate drops to 25% (2/8), while 50% of stress cycles saw total loss of both chambers. # Key facts (high-confidence, factual) 1. [Polymarket] Identical-event contract trades at 13.5% YES, up 2% over 7d/30d. 2. [Wikipedia] 33 of 100 Senate seats up in regular 2026 elections plus special elections in OH and FL; 435 House seats up. 3. [Silver Bulletin] Generic ballot ~D+6.6-7.6; Trump net approval ~-19.5 to -20.6. 4. [Cook Political Report] House rated "likely D" trending; TX Senate/House ratings moved toward Democrats. 5. [Votebeat/NPR] TX GOP map upheld by SCOTUS; Missouri GOP map blocked, reverted to old lines. 6. [code_execution] Historical full-retention rate under stress conditions ≈25%, not directly the "R Senate + R House" scenario but analogous unified-control base rate. # Cross-market signals - Kalshi related: KXNEXTSPEAKER shows Hakeem Jeffries (D) at 87% to be next Speaker — implying Kalshi markets also lean strongly toward Democratic House control, consistent with a low probability for this Yes outcome. - Polymarket: 13.5% YES, gradually rising (+2%/mo) but still low. - No sportsbook data available. # Analyst opinions and speculation - Silver Bulletin: House "likely D," Senate near-toss-up with Dem edge (56.8%). - CNN/Cook: GOP redistricting gains partly offset by adverse environment and court losses (Missouri). - Trump himself disputes midterm-loss narrative, calling it "bogus" — but data/analysts contradict this (CNN, Aug 31). - NRCC spokesperson argues Democrats are "operating on our turf," suggesting some GOP confidence despite headwinds. # Directional lean per outcome - **Yes (R Senate + R House)**: Supporting — thin GOP House majority still standing, redistricting (TX) adds GOP seats, Senate map structurally favors GOP (23 of 35 seats GOP-held), Polymarket shows modest upward drift. Opposing — deeply negative Trump approval (-20 to -32), strong Dem generic ballot lead (D+4 to D+7.6), Cook/Silver models favor Dem House takeover, Missouri map reversal, Kalshi-adjacent Speaker market at 87% Jeffries. - **No**: Favored by nearly all model-based indicators (approval, generic ballot, seat ratings, Speaker market) — currently the consensus outcome. # Gaps / unknowns - No direct Kalshi YES price for this ticker was returned (only Polymarket used as anchor). - No separate Kalshi/Polymarket Senate-only or House-only control markets found to compute implied product. - No explicit 2025 VA/NJ gubernatorial or special election results provided in raw research. - Outstanding litigation (Texas VRA/Louisiana, other maps) could still shift House baseline before Nov 2026. # Calibration anchors - Polymarket YES price (best direct anchor): 13.5%. - Kalshi-adjacent signal: Jeffries (D) Speaker market at 87%, implying House likely D. - Historical base rate: president's party fully retains both chambers in ~36% of cycles when starting unified control; drops to ~25% under stress (low approval/thin margin) — current cycle exhibits severe stress (approval -20 to -32, thin GOP House margin), suggesting the low end of historical retention rates or below is appropriate, broadly consistent with the ~13-25% market-implied range.

4. Forecasters Fable 5.1 + GPT-5.6 Sol · in parallel

claude-fable-5-1

62% conf

Yes	13%
No	87%

The best available anchor is Polymarket's identical contract at 13.5% YES, corroborated by Kalshi's KXNEXTSPEAKER market pricing Jeffries at 87% (implying ~13% R House). Since the Senate map structurally favors Republicans (23 of 35 seats GOP-held) and outcomes are positively correlated, P(R Senate | R House) is likely ~85-90%, so the joint probability is roughly 0.13 × 0.87 ≈ 0.11-0.12 — very close to the market. Fundamentals (Trump net approval -20 to -32, generic ballot D+4 to D+7.6, Cook/Silver 'likely D' House, Missouri map reversal) all point against full GOP retention, while TX redistricting and the thin-but-standing majority keep a modest tail alive. Historical stress-cycle full-retention base rate (~25%) is a loose upper bound given the current environment is worse than typical.

gpt-5.6-sol

76% conf

Yes	13%
No	87%

The best direct anchor is Polymarket’s 13.5% YES price, supported by the related Kalshi market pricing Hakeem Jeffries at 87% to become Speaker. A D+6.6 to D+7.6 generic-ballot environment, Trump net approval near -20, a razor-thin GOP House majority, and a “likely D” House rating make a Republican House hold the main obstacle. Republicans also have only a 43.2% Senate-control probability in Silver’s model, so retaining both chambers requires two correlated but individually difficult outcomes. Texas redistricting and the GOP-friendly Senate map preserve a meaningful upset path, but Missouri’s map reversal and worsening seat ratings keep the forecast near the market consensus.

5. Devil's Advocate Sonnet 5 · red-team critique

Both forecasts land on exactly the same number (13.0%) despite claiming independent reasoning—this looks like anchoring/herding on the Polymarket price rather than genuine independent estimation, especially suspicious given their stated confidence levels differ (0.62 vs 0.76) yet outputs are identical to the decimal.

Neither forecaster meaningfully engages with the brief's own copula-adjusted estimate of 19-35% joint probability (sub-question 1), which explicitly flags that "Polymarket's 13.5% may be somewhat below model-implied ranges" — both forecasts simply reassert the market price rather than reconciling this documented upward-pressure signal.

Forecast 2 cites Republicans at only 43.2% Senate-control probability (Silver FLIPR) and separately implies House control near 13%, but never shows how these combine to justify 13% joint — a rough independence-adjusted product would suggest something closer to 5-8% absent very high correlation, yet the correlation assumption is asserted rather than justified.

Neither forecast accounts for the ~2-month gap between the "current state" (Sept 2026) and the Nov 3 close date, during which outstanding litigation (Texas VRA, other maps per "Gaps/unknowns") and further redistricting rulings could materially shift House seat counts in either direction — this residual uncertainty is unaddressed.

Both forecasts under-weight the historical base-rate anchor of 25-36% full-retention under stress conditions (sub-question 7), treating it merely as a "loose upper bound" rather than seriously interrogating why the market-implied 13% is meaningfully below even the stressed historical rate — this gap deserves more explicit reconciliation or a modestly higher point estimate.

Neither forecast considers polling-miss tail risk (e.g., 2016/2022-style systematic underestimation of GOP performance in downballot/Senate races), which given the number of competitive toss-ups (18 per Cook) could push the true probability meaningfully above the polling-implied consensus.

6. Aggregator Fable 5.1 · final calibrated forecast

13% (69% confidence)

Yes: 13% No: 87%

Both forecasts converge on the Polymarket anchor (13.5%) and the Kalshi Speaker market (Jeffries 87%), and the critique does not overturn this. The brief's copula estimate of 19-35% rests on P(R Senate)≈0.75-0.85, which contradicts Silver's FLIPR (R Senate 43%), so that upward signal is unreliable; conversely, the proper decomposition P(R House)≈0.13 × P(R Senate | R House)≈0.85-0.90 yields ~0.11-0.12, consistent with the market. A D+7 generic ballot, Trump net approval near -20 or worse, and a razor-thin GOP House majority make the House hold the binding constraint, while polling-miss tail risk and pending redistricting litigation keep a real but modest upset path alive. The stressed historical retention rate (~25%) is a loose upper reference given this cycle's unusually severe environment. I settle at 13% YES, in line with the market and both forecasters.

Pipeline Timing

Total pipeline time: 254.1s

Per-tool research timings shown in the Research section above.