← Back to scans

Will the Republican Party win 193 seats in the 120th Congress? — Below 193

KXRHOUSESEATS-27-193 · Elections · 2026-04-09
40%
Agent
40%
Market Price
+0.0%
Edge
42%
Confidence
Volume: 191,893
Spread: 1.0c
Days to resolution: 298
Markets in event: 11
Final Rationale
My independent sub-question decomposition actually yields ~0.44 when properly combined (P(sq1)≈0.42 + P(sq2)≈0.01 + P(sq3)≈0.03 minus overlaps), but was artificially deflated to 0.307 by an incoherent averaging method. I ACCEPT the Devil's Advocate's Challenge 1 (math error in combination) and Challenge 2 (evidence-driven ensemble member at 0.20 was unjustifiably low). I PARTIALLY ACCEPT Challenge 3 — redistricting effects may be somewhat double-counted but still provide a real buffer. I ACCEPT Challenge 4 that large information gaps warrant staying closer to market consensus. The corrected independent estimate of ~0.42-0.44 is very close to the high-reliability market price of 0.40. Republicans losing 30+ seats requires a significant wave, which is historically plausible (~35-40% base rate for midterm losses of this magnitude) and supported by Democratic special election overperformance and high GOP retirements, but redistricting gains provide meaningful buffer. I settle at 0.40, deferring slightly to the well-traded market given remaining uncertainty about approval ratings, generic ballot polls, and economic conditions.
Price vs. prediction
How the market moved after the agent's call. Solid blue line is the actual market price; dashed line is the agent's static forecast at $t{=}0$.
$n = 54$ follow-ups
1. Planner Opus · Fermi decomposition + tool plan
Tools chosen: kalshi_data kalshi_orderbook article_search web_search wikipedia_lookup code_execution
Sub-questions (Fermi decomposition)
  1. Will Republicans enter the 120th Congress (January 2027) with fewer than 193 House seats based on 2026 midterm results? (w=0.50)
  2. Is the current Republican House seat count (April 2026) already below 193? (w=0.25)
  3. Will special elections or vacancies between April 2026 and February 2027 cause Republicans to drop below 193 seats even if midterm results alone would not? (w=0.25)
Planner reasoning
This question resolves YES if Republicans hold fewer than 193 House seats on February 1, 2027, shortly after the 120th Congress is seated on January 3, 2027. The key driver is the November 2026 midterm elections. Republicans entered the 119th Congress with approximately 220 seats. To fall below 193, they would need a net loss of 27+ seats. Historical base rates are informative: in 2018 (Trump's first midterm), Republicans lost ~40 seats from ~241, which would be analogous. In 2006 (Bush second term), Republicans lost ~30 seats. Both scenarios would push Republicans below 193 from a 220-seat baseline. However, 2022 and 2014 showed smaller losses of 9 and 13 respectively, which would keep Republicans well above 193. The political environment in 2026 matters: Trump's approval ratings, economic conditions, and structural factors like district composition all matter. Special elections between now and November 2026 could also slightly erode the baseline. The threshold of 193 is notably low - it would represent Republicans being in the minority (Democrats would need 218+ for majority), suggesting this resolves YES only if Democrats win a significant House majority. Current forecasts for 2026 midterms and any existing prediction market prices on adjacent brackets will be critical evidence.
2. Research Parallel tool calls · legacy v2
ToolStatusTimeItemsSummary
kalshi_data OK 2.4s -
kalshi_orderbook OK 0.1s -
article_search OK 0.8s -
web_search OK 58.7s -
wikipedia_lookup OK 2.2s -
code_execution OK 0.0s -
3. Evidence Synthesis Sonnet · 16 items · legacy v2
#StrengthCredibilityDirectionSourceClaimPriced In
1 STRONG 75 ↑ UP kalshi_data Kalshi market 'Below 193' (KXRHOUSESEATS-27-193) is priced at 40%, up +6% over 30 days but down -1% over 7 days, with high liquidity (depth 27,313 contracts) and average daily volume of 1,465 contracts. Yes
2 STRONG 72 NEUTRAL code_execution Historical base rate analysis shows Republicans lost 27+ seats in 4 of 8 modern midterms (50%), and starting from ~220 seats, a 2006-like loss (-30) would yield 190 seats (below 193) while a 2022-like loss (-9) would yield 211 seats (above 193). Yes
3 STRONG 85 ↑ UP article_search Democrat Emily Gregory won a special election in a district including Mar-a-Lago (Palm Beach) in March 2026, flipping a seat Republicans held by 19 points in 2024, described as 'part of a broader trend of the party overperforming in elections.' Yes
4 MODERATE 78 ↓ DOWN article_search Republicans pursued aggressive mid-cycle redistricting in Texas (5 new GOP-leaning districts), Ohio, Indiana, South Carolina, Missouri, and Florida, which could offset some seat losses in the 2026 midterms. Yes
5 MODERATE 80 NEUTRAL article_search Ballotpedia tracked 42 battleground House districts for 2026, with Democrats holding 22 and Republicans holding 20; Democrats need a net gain of only 3 seats for majority control. Yes
6 MODERATE 78 ↑ UP article_search As of April 2026, 57 representatives (36 Republicans, 22 Democrats) announced retirement, creating a significant number of open seats and potentially unfavorable conditions for Republicans in marginal districts. Yes
7 MODERATE 68 NEUTRAL kalshi_data The related market 'Democrats hold above 249 seats' is priced at only 24%, suggesting the median market expectation is that Democrats will win significantly but likely land between 218-249 seats, which would correspond to Republicans holding 186-217 seats. Yes
8 MODERATE 78 NEUTRAL article_search There are 14 Democratic-held House districts that Trump won in 2024 and 9 Republican-held districts that Kamala Harris won in 2024, creating an asymmetric seat defense picture slightly favoring Democrats in net competitive seats. Yes
9 MODERATE 62 ↑ UP article_search Democrats won Virginia's 10 House seats market at 73% probability (up 21% in 30 days), suggesting strong Democratic performance expected in Virginia, a bellwether state. No
10 MODERATE 72 ↑ UP article_search Government shutdowns and Trump's war with Iran without congressional approval are creating significant political headwinds for Republicans heading into the 2026 midterms. Yes
11 STRONG 92 ↓ DOWN web_search As of April 2026, Republicans hold approximately 218 House seats (with some vacancies), far above the 193 threshold; Republican Clay Fuller won the GA-14 special election on April 7, 2026, maintaining the GOP seat. Yes
12 STRONG 88 ↓ DOWN web_search As of early 2026, Republicans had 218-214 majority with 3 vacancies; NJ-11 special election scheduled for April 16, 2026 was still pending at time of reporting. Yes
13 STRONG 88 ↑ UP article_search Democrat Emily Gregory flipped a Trump-endorsed Republican's seat in a special election in Palm Beach (Mar-a-Lago district) in March 2026, demonstrating Democrats are winning special elections in historically safe Republican territory. Yes
14 MODERATE 82 NEUTRAL web_search Multiple Republican vacancies occurred in late 2025/early 2026 (MTG resigned, LaMalfa died, Sherrill resigned from NJ), creating a pattern of special elections; the current narrow margin means individual seat changes have outsized impact. Yes
15 STRONG 85 ↓ DOWN article_search Special elections between now and November 2026 are unlikely alone to drive Republicans below 193 seats since they currently hold ~218 seats, requiring a net loss of 25+ seats via special elections alone which is historically unprecedented. Yes
16 MODERATE 75 ↑ UP article_search Special elections in 2025-2026 show a consistent Democratic overperformance trend (Mar-a-Lago district flip, NJ special election pending), which could erode the Republican baseline before November 2026 midterms. No
Information Gaps
  • No direct forecaster model outputs (Cook Political Report, Sabato's Crystal Ball, FiveThirtyEight) for Republican seat projections in the 2026 midterms — would be critical for calibrating the distribution of outcomes
  • No Trump approval rating data from April 2026 to gauge the current political environment
  • No information on the NJ-11 special election result (scheduled April 16, 2026), which was still pending
  • No information on California redistricting outcome and its potential net seat impact (article was cut off mid-sentence)
  • No economic data (GDP growth, unemployment, inflation) for early 2026 — a key predictor of midterm outcomes
  • No information on how Texas redistricting maps were ultimately resolved by courts, which would affect the Republican baseline for the midterms
  • No polling data on generic congressional ballot for 2026
  • No information on whether Democratic redistricting responses (California, etc.) offset Republican gains from Texas redistricting
Key Uncertainties
  • Net effect of redistricting: Republican gains in Texas (+5 projected) and other states vs. Democratic counter-redistricting in California and other blue states could swing the baseline by 5-15 seats
  • Whether Democratic special election overperformance translates to November 2026 general election performance — historically, special election overperformance sometimes overstates general election swings
  • Trump's approval trajectory and economic conditions between April and November 2026 — government shutdowns and Iran war are headwinds but could improve
  • Whether the 'above 249 Democrats' market at 24% implies a wide distribution — Republicans could be anywhere from 186 to 217 seats in expectation under Democratic wave scenarios
  • Court rulings on Texas redistricting and mail ballot deadlines (Supreme Court case Watson v. RNC) could significantly alter competitive district counts
  • Whether high Republican retirement rate (36 announced) creates structural disadvantage in open seat races vs. incumbents
Market Context
Price: 0.4 · 7d: -0.01 · 30d: +0.06
Volume: HIGH · Depth: HIGH — 27,313 contracts depth
Reliability: HIGH — high volume (1,465 avg daily), deep orderbook, 131 days of data, price range of 8-44% shows meaningful price discovery over time
4. Independent Forecaster Opus · No market price · legacy v2 Did NOT see market price
Independent estimate: 31% (35% confidence)
base rate
35%
35% conf
evidence driven
20%
35% conf
contrarian
37%
35% conf
Spread
17pp
High disagreement
sq1: Will Republicans enter the 120th Congress (January 2027) with fewer than 193 House seats based on 2026 midterm results?
42%
Base rate of ~37-50% for 27+ seat losses in midterms. Redistricting gains push effective baseline to ~223-225, requiring ~32+ seat loss (base rate ~37%). Democratic-favorable environment indicators (special election overperformance, high GOP retirements, political headwinds) push up by ~5-8%. Final estimate: 42%.
sq2: Is the current Republican House seat count (April 2026) already below 193?
1%
Republicans currently hold ~218 seats. Essentially impossible to be below 193 currently. 1%.
sq3: Will special elections or vacancies between April 2026 and February 2027 cause Republicans to drop below 193 seats even if midterm results alone would not?
3%
This requires midterms to leave Republicans at 193-195 AND then special elections/vacancies to push them below. Very narrow scenario. 3%.
Show full reasoning trace
base rate: Historical midterm losses of 27+ seats from the president's party occur roughly 40-50% of the time. Adjusting for redistricting gains raising the effective Republican baseline to ~223-225 seats, the required loss becomes 30-32+ seats, reducing the base rate to ~35-40%.
evidence updates: Democratic special election overperformance (+5%), high GOP retirements (+3%), political headwinds (+3%), but Republican redistricting gains (-6-8%). Net effect roughly +2-3% from adjusted base.
combination method: Sq1 dominates with 42% probability. Sq2 essentially zero. Sq3 adds ~2% marginal probability. Weighted average formula gives 0.22 but this underweights the primary pathway. Logical combination gives ~0.44. Final estimate splits the difference at 0.35.
final: Republicans falling below 193 seats requires a significant but not historically unprecedented wave election. The 2026 environment shows some Democratic-favorable indicators but redistricting gains provide a meaningful buffer. Substantial uncertainty remains due to missing approval ratings, economic data, and generic ballot polls. Estimate: 35%.
ensemble: {'ensemble_count': 3, 'perspectives': {'base_rate': {'p_yes': 0.35, 'confidence': 0.35}, 'evidence_driven': {'p_yes': 0.2, 'confidence': 0.35}, 'contrarian': {'p_yes': 0.37, 'confidence': 0.35}}, 'spread': 0.17, 'individual_p_yes': {'base_rate': 0.35, 'evidence_driven': 0.2, 'contrarian': 0.37}}
5. Devil's Advocate Sonnet · legacy v2 Market price revealed
Math Error Reasoning Flaw Missing Info Asymmetric Evidence
Challenges
  1. The combination method is incoherent. The forecaster states sq1 dominates at 42%, sq3 adds ~2% marginal probability, giving a logical combination of ~0.44, but then arbitrarily 'splits the difference' with a weighted average formula giving 0.22 to arrive at 0.35. There is no principled reason to average a correct logical OR combination with an inappropriate weighted average. The sub-questions were designed so that P(below 193) ≈ P(sq1) + P(sq2) + P(sq3) - overlaps ≈ 0.42 + 0.01 + 0.03 - small overlaps ≈ 0.44. The final estimate of 0.307 is artificially deflated by this averaging error.
  2. The 'evidence-driven' perspective at 0.20 in the ensemble appears unjustifiably low. The forecaster's own evidence balance is 7 UP vs 4 DOWN, and the strong evidence pieces (special election flips, retirement wave, political headwinds) all point toward Democratic gains. An evidence-driven estimate should be at least as high as the base rate, not substantially below it. This drags down the ensemble average.
  3. The forecaster notes redistricting gains push the effective Republican baseline to 223-225, requiring 30-32+ seat losses, and claims this reduces the base rate to 35-40%. However, the question resolves on the 120th Congress which begins January 2027 — meaning 2026 midterm results determine this. The relevant comparison is president's party losses in midterms when starting from a narrow majority. The redistricting adjustment of -6 to -8% seems oversized given that redistricting effects are already partially captured in the battleground district analysis showing 42 competitive seats.
  4. The forecaster identifies critical information gaps (no approval ratings, no generic ballot polling, no economic data, no forecaster models) but doesn't adequately widen uncertainty to reflect this. With this many missing inputs and 298 days to resolution, the forecast should be closer to the base rate / market consensus rather than confidently below it.
  5. The mention of 'Trump's war with Iran without congressional approval' is stated as creating significant headwinds but is not substantiated with polling data or concrete evidence of voter backlash. This is treated as an established fact driving the forecast upward without verification of its political impact.
Suggested adjustment: +9pp
The primary issue is the combination method error. The sub-question decomposition logically yields ~0.44 (P(sq1) + P(sq2) + P(sq3) minus small overlaps). The arbitrary averaging with a 'weighted average formula' of 0.22 has no methodological basis and artificially depresses the estimate. Additionally, the evidence-driven ensemble member at 0.20 is inconsistent with the forecaster's own evidence balance (7 UP vs 4 DOWN). Correcting the combination method alone would bring the estimate to ~0.40-0.44. Adjusting upward by 0.09 to ~0.40 accounts for the math error while maintaining some conservatism given the large information gaps.
Market comparison: Forecaster: 30.7%, Market: 40.0%. The 9.3% divergence appears primarily driven by a combination method error rather than genuinely different information or analysis. The forecaster's own sub-question estimates, when properly combined, yield ~0.44 — actually slightly above the market. The market price of 40% appears well-calibrated and has high reliability indicators. The divergence is not justified by the forecaster's reasoning.
6. Aggregator Opus · final calibrated forecast
40% (42% confidence)
My independent sub-question decomposition actually yields ~0.44 when properly combined (P(sq1)≈0.42 + P(sq2)≈0.01 + P(sq3)≈0.03 minus overlaps), but was artificially deflated to 0.307 by an incoherent averaging method. I ACCEPT the Devil's Advocate's Challenge 1 (math error in combination) and Challenge 2 (evidence-driven ensemble member at 0.20 was unjustifiably low). I PARTIALLY ACCEPT Challenge 3 — redistricting effects may be somewhat double-counted but still provide a real buffer. I ACCEPT Challenge 4 that large information gaps warrant staying closer to market consensus. The corrected independent estimate of ~0.42-0.44 is very close to the high-reliability market price of 0.40. Republicans losing 30+ seats requires a significant wave, which is historically plausible (~35-40% base rate for midterm losses of this magnitude) and supported by Democratic special election overperformance and high GOP retirements, but redistricting gains provide meaningful buffer. I settle at 0.40, deferring slightly to the well-traded market given remaining uncertainty about approval ratings, generic ballot polls, and economic conditions.