← Back to scans

Will the Democratic party win the House, the Democratic party win the Senate, and the Democratic party win the Presidency? — Democratic Sweep

POWER-28-DH-DS-DP · Elections · 2026-04-14
34%
Agent
42%
Market Price
-8.0%
Edge
30%
Confidence
Volume: 57,701
Spread: 3.0c
Days to resolution: 1303
Markets in event: 8
Final Rationale
My independent estimate of 22% contained internal contradictions: 11 favorable evidence signals yet a sub-base-rate conclusion, and inconsistent conditional probabilities. Correcting these per the critique pushes the estimate to ~30-33%. The Kalshi market at 41% reflects genuine information but with only $162K volume and a 13-point spread versus Polymarket (54%), it deserves moderate but not full deference. The 2.5-year horizon remains a real source of uncertainty — Democratic candidates are unknown, economic conditions could shift dramatically — which justifies remaining below market price. Modern declining split-ticket voting supports higher conditional sweep probabilities than the full post-WWII era suggests.
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 $t{=}0$.
$n = 58$ follow-ups
1. Planner Opus · Fermi decomposition + tool plan
Tools chosen: kalshi_data article_search web_search wikipedia_lookup code_execution fred_data
Sub-questions (Fermi decomposition)
  1. Will the Democratic party win the Presidency in 2028? (w=0.40)
  2. Will the Democratic party win the House in 2028? (w=0.30)
  3. Will the Democratic party win the Senate in 2028? (w=0.30)
Planner reasoning
This question requires estimating the joint probability of three simultaneous Democratic victories in 2028. The combination logic is effectively a minimum or product of correlated probabilities. Key analytical angles: (1) Base rate of Democratic trifectas post-WWII is roughly 25% (5 of 20 elections), but conditional on winning the presidency it rises to ~56%. (2) Presidential probability: Republicans hold the White House in 2028 as an open seat (Trump is term-limited), and historically the incumbent party loses open-seat elections ~62% of the time. Economic conditions under Trump by 2028 will heavily influence this. (3) Senate map: Class II senators elected in 2022 are up in 2028 - need to assess whether this map favors Democrats or Republicans. (4) House: Coattail effects from a presidential win historically help the same party in House races, but gerrymandering and structural factors matter. (5) Correlation structure: All three outcomes are positively correlated - a strong Democratic presidential wave tends to sweep congressional seats too. For YES to occur, the sequence is: strong Democratic primary candidate emerges (2027-early 2028), Trump-era economic or political environment creates backlash, Democratic presidential win in Nov 2028 with coattail effects sufficient to flip both chambers. The 1303-day horizon means the 2026 midterms (180 days away) will provide the first major signal about the political environment heading into 2028.
2. Research Parallel tool calls · legacy v2
ToolStatusTimeItemsSummary
kalshi_data OK 2.2s -
article_search OK 0.3s -
web_search OK 44.6s -
wikipedia_lookup OK 2.0s -
code_execution OK 0.0s -
fred_data OK 1.1s -
3. Evidence Synthesis Sonnet · 18 items · legacy v2
#StrengthCredibilityDirectionSourceClaimPriced In
1 MODERATE 72 ↑ UP kalshi_data Kalshi's Democratic Sweep market is priced at 41%, up from ~22% at its low and up 6% over the past 30 days, suggesting rising Democratic presidential prospects are driving overall sweep odds. Yes
2 STRONG 68 ↑ UP web_search Polymarket prices Democrats at ~60.5% to win the 2028 presidency, with William Hill offering 4/6 (implying ~60%), driven by fallout from U.S. military action against Iran in late March 2026. No
3 MODERATE 65 NEUTRAL web_search Polymarket has $519M in total volume on the 2028 presidential winner market; JD Vance leads at 18.8%, Gavin Newsom at 16.6%, with no single dominant favorite on either side. Yes
4 STRONG 90 ↑ UP wikipedia_lookup Trump is term-limited and ineligible for 2028, making this an open-seat election — historically the incumbent party loses open-seat presidential elections roughly 62.5% of the time post-WWII. Yes
5 STRONG 88 ↑ UP fred_data CPI is up ~10.5 index points YoY (as of March 2026), unemployment is 4.3% (up 0.1 YoY), and consumer sentiment is 56.6 — a historically low level suggesting negative economic conditions under Republican incumbency. Yes
6 STRONG 85 ↑ UP fred_data Consumer sentiment at 56.6 is near historically depressed levels, which has consistently preceded incumbent-party losses in presidential elections. Yes
7 MODERATE 60 ↓ DOWN code_execution Historical base rate modeling estimates Democratic sweep probability at ~16.4%, using post-WWII trifecta rates (25%) adjusted for open-seat dynamics (62.5% opposition win rate) and conditional trifecta rates given a Dem presidency (55.6%). Yes
8 MODERATE 62 NEUTRAL web_search Gavin Newsom leads Democratic presidential nomination odds at 27.5% on Polymarket; Kamala Harris re-entering at 6.2% suggests Democratic primary field is unsettled with no clear frontrunner. Yes
9 MODERATE 63 ↑ UP web_search Polymarket prices Democrats at >85% to win the House in the 2026 midterms, signaling strong anti-Trump/anti-incumbent sentiment that, if sustained, could carry into 2028 House elections. Yes
10 MODERATE 65 ↑ UP code_execution Historical data shows Democrats have won the House in 5 of 9 elections when they won the presidency, suggesting strong coattail correlation between presidential and House outcomes. Yes
11 MODERATE 70 ↑ UP kalshi_data The Democratic Sweep Kalshi market (which requires House win) rose from ~22% to 41%, implying collective market belief in Democratic House prospects has nearly doubled over the market's history. Yes
12 WEAK 52 ↑ UP web_search A Newsweek-cited traditional bookmaker stated Democratic trifecta is 'currently the most likely scenario,' citing both presidential and congressional races. Yes
13 MODERATE 82 NEUTRAL wikipedia_lookup The 2028 Senate elections feature Class 3 senators (last elected in 2022), meaning the map reflects a class that last ran in a Democratic-leaning midterm year; seat composition will determine whether Democrats face a favorable or unfavorable map. Yes
14 WEAK 48 NEUTRAL kalshi_data Pennsylvania 2028 Democratic Senate nominee Brendan Boyle is at 32% on Kalshi (up 11% in 7 days and 30 days), indicating a competitive but uncertain Democratic recruitment picture in a key swing state. No
15 WEAK 45 NEUTRAL kalshi_data New York Senate 2028 Democratic nominee Chuck Schumer is priced at 42% on Kalshi (down 5% over 30 days), suggesting uncertainty about the New York Democratic Senate candidate even in a safely blue state. Yes
16 MODERATE 62 NEUTRAL web_search Polymarket has Senate control near even odds for 2026 midterms, implying the current political environment is competitive but not yet a wave for either party in the Senate. Yes
17 MODERATE 58 ↑ UP web_search The Iran military operation by the Trump administration in late March 2026 eroded Republican favorability significantly, which could affect Senate race dynamics if the geopolitical backlash persists to 2028. No
18 WEAK 55 ↑ UP code_execution In all 5 post-WWII instances of Democratic trifectas, Democrats won the Senate when they won the presidency, but this is from a small sample and reflects favorable maps at those times. Yes
Information Gaps
  • No detailed breakdown of 2028 Senate Class 3 seat composition — how many seats are held by Republicans vs. Democrats and in which states (swing vs. safe).
  • No data on 2026 midterm outcomes, which will be the primary signal for 2028 political environment and congressional control.
  • No direct Kalshi or Polymarket market specifically for Democratic presidential nominee odds (only VP nominee AOC at 9% is available).
  • No analysis of gerrymandering effects on 2028 House district maps after potential post-2030 census redistricting.
  • No economic forecasts for 2027-2028 GDP, inflation, or unemployment to assess incumbent-party headwinds.
  • No polling data on Trump approval ratings or generic ballot spreads for 2026-2028.
  • No data on Republican presidential field strength beyond individual candidate odds on Polymarket.
  • The Iran conflict's lasting impact on public opinion is unknown and could reverse quickly with a ceasefire.
Key Uncertainties
  • Whether the Iran military operation creates sustained Republican vulnerability or dissipates before 2028 (most critical near-term swing factor).
  • 2026 midterm results — a Democratic wave would dramatically increase sweep probability; a Republican hold would decrease it.
  • Identity of the Democratic presidential nominee and their electoral coalition strength.
  • Identity of Republican nominee (post-Trump) and their relative strength vs. Trump's 2024 coalition.
  • 2028 Senate map specifics: which Republican senators from the 2022 class are in swing states and vulnerable.
  • Whether inflation/economic conditions improve or worsen under Republican governance by 2028.
  • Significant divergence between Kalshi (41%) and base-rate model (~16%) — the ~25 percentage point gap reflects either a market overreaction to current events or a systematic underweighting of current political conditions in the base-rate model.
  • The Polymarket Democratic Sweep at 54% vs. Kalshi at 41% represents a 13-point spread requiring explanation — possible differences in market composition (retail vs. institutional, U.S. vs. international traders).
Market Context
Price: 0.41 · 7d: +0.02 · 30d: +0.06
Volume: LOW-MODERATE — $162K total traded on Kalshi sweep market; 251 avg daily contracts is thin for a multi-year political market · Depth: UNKNOWN — depth data not provided
Reliability: MEDIUM — CFTC-regulated Kalshi is credible but the $162K total volume is low relative to the $519M on Polymarket's presidential winner market alone; the large spread between Kalshi (41%) and Polymarket (54%) sweep odds warrants caution and suggests potential mispricing on one or both platforms
4. Independent Forecaster Opus · No market price · legacy v2 Did NOT see market price
Independent estimate: 22% (33% confidence)
base rate
23%
35% conf
evidence driven
22%
35% conf
contrarian
21%
30% conf
Spread
2pp
Low disagreement
sq1: Will the Democratic party win the Presidency in 2028?
58%
Base rate: Incumbent party loses open-seat elections ~62.5% of the time post-WWII, giving Democrats a starting point of ~62.5%. However, this base rate covers varied conditions. Current evidence: (1) Consumer sentiment at 56.6 is historically depressed — this pushes UP, consistent with incumbent-party loss. (2) Trump is term-limited, making this a true open seat — the Republican nominee won't have presidential incumbency advantage. (3) Polymarket at ~60.5% for Democrats aligns with the base rate. (4) Iran military action and inflation are creating anti-Republican sentiment now, but 2028 is 2.5 years away — much can change. Pushing DOWN slightly: (a) No clear Democratic frontrunner yet — an unsettled primary could produce a weaker nominee. (b) Economic conditions could improve by 2028. (c) The 62.5% base rate includes some elections where conditions were much worse for the incumbent party. I'll moderate slightly below the base rate given the long time horizon and uncertainty: 58%.
sq2: Will the Democratic party win the House in 2028?
52%
Base rate: Democrats have won the House in approximately 50% of post-WWII elections overall. Conditional on winning the presidency, the rate rises to about 56% (5/9 instances). Evidence updates: (1) Strong anti-incumbent sentiment in 2026 midterms (Polymarket >85% Dem House) pushes UP — if Democrats win the House in 2026, they start from a position of incumbency in 2028. However, presidential-year House outcomes don't always mirror midterm results. (2) Coattail effects: if Democrats win the presidency, House win probability increases substantially due to correlation. (3) Gerrymandering and structural Republican advantages in House districts push DOWN slightly. (4) The long time horizon means current sentiment may not persist. Unconditional estimate (not conditional on Dem presidency): ~52%. The House is structurally competitive but Democrats face some geographic disadvantage.
sq3: Will the Democratic party win the Senate in 2028?
48%
Base rate: Democrats have controlled the Senate in roughly 45-50% of post-WWII election cycles. The 2028 Class 3 senators were last elected in 2022, which was a relatively favorable Democratic year. This means Republicans defending seats won in less favorable conditions could be vulnerable — but the specific map composition is unclear. Key considerations: (1) If Democrats win the presidency, Senate probability rises due to coattail effects and VP tiebreaker — historically when Dems win presidency, they've usually won the Senate too (small sample). (2) Senate near-even odds for 2026 suggests competitive but not wave environment. (3) The Senate has structural challenges for Democrats due to small-state bias favoring Republicans. (4) Iran backlash could help in swing-state Senate races but may fade. (5) The Class 3 map details are a major information gap. Unconditional estimate: 48%. The Senate is the hardest chamber for Democrats to win and hold.
Show full reasoning trace
base rate: Post-WWII, Democratic trifectas have occurred in ~25% of elections (1948, 1960, 1964, 1976, 2008 — 5 of 20). Open-seat elections favor the opposition party ~62.5% of the time. When Democrats win the presidency, they've achieved a trifecta 5/9 times (~56%).
evidence updates: Current anti-Republican sentiment (consumer sentiment 56.6, Iran backlash, inflation) pushes probability UP from the 25% base rate. However, the 2.5-year time horizon is a major moderating factor — sentiment can shift dramatically. The unsettled Democratic primary field and unknown 2028 economic conditions add uncertainty. Net effect: modest upward adjustment from 25% base to ~23% (actually slightly below because I'm being cautious about the long horizon despite favorable current conditions).
combination method: Joint probability with conditional dependence: P(Pres) * P(House|Pres) * P(Senate|Pres,House) = 0.58 * 0.65 * 0.60 ≈ 0.226. Strong positive correlation between all three outcomes means conditional probabilities are substantially higher than unconditional ones.
final: I estimate a 23% probability of a Democratic sweep in 2028. This is close to the historical base rate of 25%, modestly reflecting that current conditions are somewhat favorable for Democrats (open seat, economic headwinds for Republicans) but tempered by the very long time horizon and significant uncertainties. I note that prediction markets are pricing this substantially higher (41-54%), which I believe reflects overweighting of current anti-Republican sentiment that may not persist to November 2028.
ensemble: {'ensemble_count': 3, 'perspectives': {'base_rate': {'p_yes': 0.23, 'confidence': 0.35}, 'evidence_driven': {'p_yes': 0.22, 'confidence': 0.35}, 'contrarian': {'p_yes': 0.21, 'confidence': 0.3}}, 'spread': 0.02, 'individual_p_yes': {'base_rate': 0.23, 'evidence_driven': 0.22, 'contrarian': 0.21}}
5. Devil's Advocate Sonnet · legacy v2 Market price revealed
Math Error Asymmetric Evidence Reasoning Flaw
Challenges
  1. The forecaster's conditional probability chain (0.58 * 0.65 * 0.60 = 0.226) uses conditional probabilities of P(House|Pres) = 0.65 and P(Senate|Pres,House) = 0.60, but the unconditional estimates given are P(House) = 0.52 and P(Senate) = 0.48. The implied conditional boost from winning the presidency seems somewhat arbitrary — the forecaster states the historical rate of House wins conditional on Dem presidency is 56% (5/9), yet uses 65% in the calculation. Similarly, the Senate conditional probability of 60% lacks clear justification. The forecaster is inconsistent between their sub-question estimates and their combination method.
  2. The forecaster acknowledges strong positive correlation between the three outcomes but may be underestimating the degree of correlation. In modern elections, split-ticket voting has declined dramatically. When one party wins the presidency, the probability of sweeping both chambers has increased in recent decades compared to the full post-WWII average. The base rate of 5/20 (25%) includes many elections from an era of much more common split-ticket voting. In the modern era (2000-2024), sweeps have occurred in 4 of 7 elections (2000R partial, 2004R partial, 2008D, 2016R partial, 2020D, so roughly 2-3 full sweeps out of 7). The declining split-ticket rate should push conditional probabilities higher.
  3. The forecaster's 'ensemble' approach adds no real value — all three perspectives (base_rate: 0.23, evidence_driven: 0.22, contrarian: 0.21) are essentially identical with a spread of only 0.02, suggesting the forecaster isn't genuinely considering alternative framings. A true contrarian perspective, given the strong evidence of anti-incumbent sentiment, should arguably be higher (not lower) than the base rate perspective.
  4. The forecaster systematically discounts all favorable evidence by citing '2.5 years is a long time' but doesn't apply the same temporal discount symmetrically. If current conditions could improve for Republicans, they could also worsen. The net effect of long time horizons should be to push toward base rates, not systematically below them. Yet the forecaster arrives at 22-23%, which is actually BELOW the 25% historical base rate despite acknowledging that current conditions are somewhat favorable for Democrats.
  5. The evidence direction count is 11 UP, 1 DOWN, 6 NEUTRAL, yet the forecast ends up below the historical base rate. This is a significant inconsistency. The forecaster claims to be making a 'modest upward adjustment from 25% base' but actually adjusted downward to 22-23%. This contradicts the stated evidence balance.
  6. The forecaster dismisses prediction markets (Kalshi 41%, Polymarket 54%) as 'overweighting current anti-Republican sentiment' but doesn't provide a rigorous framework for how much weight current sentiment deserves 2.5 years out. Markets aggregate many participants with diverse information sets. A 19-percentage-point divergence from even the lower market price requires stronger justification than 'things could change.'
Suggested adjustment: +8pp
The forecaster's own evidence balance (11 UP vs 1 DOWN) and stated favorable conditions should place the estimate at or above the 25% base rate, not below it. The conditional probability chain contains unjustified numbers that don't match the sub-question estimates. The asymmetric application of time-horizon discounting systematically biases downward. Correcting for these flaws — particularly the contradiction between evidence direction and final estimate, and using more defensible conditional probabilities — suggests moving upward by roughly 8 percentage points to approximately 30%. This is still well below market prices but corrects the internal inconsistencies.
Market comparison: Forecaster: 22%, Market: 41%. The 19-point divergence is partially justified by the forecaster's legitimate concern about long time horizons and uncertainty, but the forecaster's internal reasoning contradicts itself — the evidence balance and stated conditions favor Democrats yet the final estimate falls below the historical base rate. The market volume is low-moderate ($162K), providing some reason for skepticism about market efficiency, but the forecaster's systematic downward bias in evidence processing is a more concerning issue than any market mispricing.
6. Aggregator Opus · final calibrated forecast
34% (30% confidence)
My independent estimate of 22% contained internal contradictions: 11 favorable evidence signals yet a sub-base-rate conclusion, and inconsistent conditional probabilities. Correcting these per the critique pushes the estimate to ~30-33%. The Kalshi market at 41% reflects genuine information but with only $162K volume and a 13-point spread versus Polymarket (54%), it deserves moderate but not full deference. The 2.5-year horizon remains a real source of uncertainty — Democratic candidates are unknown, economic conditions could shift dramatically — which justifies remaining below market price. Modern declining split-ticket voting supports higher conditional sweep probabilities than the full post-WWII era suggests.
7. Entry Decision Opus Trader
Side
NO
Position Size
$500
Contracts
862
Trader Confidence
30%
Trade Rationale
The forecast estimates 34% probability versus the market price of 42%, giving an 8% edge on the NO side. This is a meaningful edge but comes with significant caveats: the forecaster confidence is very low at 0.30, the Devil's Advocate flagged math errors and reasoning flaws suggesting the true probability could be higher (critic recommended +8% adjustment), and the 1303-day horizon introduces enormous uncertainty. The sub-question forecasts (58% presidency, 52% House, 48% Senate) actually multiply to roughly 14.5% if independent, but correlation effects push this higher — the forecaster's 34% already accounts for this. The portfolio already has substantial election exposure with multiple 2028-related positions, so I want to keep this small.
Allocation Logic
Minimum position size of $500 given: low forecaster confidence (0.30), borderline 8% edge that the critic argues may be smaller, extremely long time horizon, low liquidity, and existing election concentration in the portfolio.
Entry price: $0.58
Current: $0.51
Status: OPEN
P&L: -$60.34