House Control Odds: Implied Probability Vs. Election Models

The House control odds are not close. On Kalshi's 2026 House winner market, the Democratic side trades at an 85-cent bid against an 86-cent ask, an implied probability of about 85%, with the Republican side at 14 to 15 cents ahead of the November 3, 2026 midterm elections. That is a heavy favorite in a chamber Republicans actually hold, and the price rests on one of the most reliable base rates in American politics rather than on any single poll. This page runs the market's number against the fundamentals the way a DFS player runs a projection against a slate, and the strangest number on the board is not the price at all. It is where the trading concentrates, and that board is below.

The Quick Answer

Kalshi prices Democrats at about 85% to win control of the U.S. House in the 2026 midterms, with Republicans near 15%, on a board that has traded about 22.3 million contracts. The price leans on midterm history (the president's party has lost House seats in all but two midterms since World War II) and on Republicans defending a majority so thin that a net gain of three seats flips it. The seat math, the polling range, the volume board where the cheap side out-trades the favorite, and the fee-adjusted worked example are all below.

The Market At A Glance

One mechanical note before the argument: the book here is deep and tight. The Democratic side shows a one-cent spread with about 346,000 contracts resting at the ask on the August 12 pull, so the screen price is a real price, not a thin-market artifact you have to discount before trusting. That matters because everything below starts from taking 85 cents seriously. So where does a number that lopsided come from in a chamber the other party controls? Start with the base rate.

The Base Rate: What Midterms Charge The President's Party

Strip away the 2026 particulars and price the situation alone, the way you would price a park factor before looking at the hitter. The president's party has lost House seats in all but two midterms since World War II (1998 and 2002 are the only exceptions), a 90% hit rate across the 20 postwar editions, though a 20-election sample is a historical tendency to weigh, not a law to bet on. The recent versions of the pattern have not been gentle:

The row that matters most is 2022, the smallest loss on the table, because even that mild edition would be far more than enough this cycle. The 2024 elections left Republicans, the party holding the White House under President Trump, defending a certified 220-215 House majority, one of the thinnest in modern history, so Democrats need a net gain of just three seats to reach the 218 that controls the chamber (resignations and vacancies have nudged the live count since, but the flip math runs off the elected baseline). Now put the two numbers together: in the 18 postwar midterms where the president's party lost ground, it never lost fewer than four seats, more than the three Democrats need. A base rate that strong, applied to a majority with almost no cushion, produces a big Democratic number before anyone reads a single district poll. That is the skeleton of the 85-cent price. The flesh on it is the environment.

The Environment: What Polling Adds And What Models Subtract

Public generic-ballot polling averages in late July and early August 2026 mostly sit between roughly Democrats plus-4 and plus-6, with at least one likely-voter-adjusted average running closer to plus-8; the exact figure depends on the aggregator and its screen. For calibration, a national edge in that range sits between 2022, when Republicans won the House vote by about 3 points and squeaked out a majority, and 2018, when Democrats won it by about 8.6 and gained about 40 seats. The polling, in other words, points the same direction as the base rate, and the market is charging accordingly.

So why would anyone price this below 85? Because seat models have to digest something a national average cannot see: the district map itself moved. Texas redrew its congressional lines in 2025 with the stated aim of adding Republican-leaning seats, California voters answered that November by passing Proposition 50 to add Democratic-leaning ones, and the litigation over those maps and their imitators is still running. A redrawn map changes how efficiently a national margin converts into seats, which is exactly the kind of input that makes model probabilities and market prices diverge even when they read the same polls; district-by-district raters like Cook Political Report and Inside Elections still score only a few dozen of the 435 seats as genuinely competitive, so a modest shift in how that short list breaks moves the seat count more than another point of national margin does. We are not going to invent a model probability to set against the price; no seat-level model we would put our name on is on this page yet, and when our model panel prices this market, this page will carry that verdict next to the live number. My own read, stated as a read and nothing more: the base rate has earned most of that 85 cents, but a mid-decade map shuffle is exactly the kind of variance the postwar table has never had to price, and I would want wider tails than a one-cent spread implies. Our guide to what transfers from DFS to prediction markets calls this the matchup layer, and on this board the matchup layer was literally redrawn mid-cycle.

The Volume Board: The Action Lives On The Cheap Side

The strange number teased in the opening sits in this table, and it is the reverse of what the price would lead you to expect.

The 15-cent underdog market has traded more contracts than the favorite's, by about 1.4 to 1, and carries more open interest, with about 62% of the favorite's traded contracts still open against about 59% on the underdog side. One honesty stop before anyone reads that like an ownership page: it is not one. Every open contract has a Yes holder and a No holder, so open interest counts how many contracts remain outstanding on a question, never which way the crowd leans; the DFS instinct to read big numbers on the long-shot line as "the field is on the dart" is exactly the instinct to unlearn here, and catching that reflex is half the lesson of our ownership and leverage on Kalshi piece. What the table does establish is where the action is: the cheap market is the busy one, because a 15-cent contract offers a payout profile that attracts traders on both sides of the argument, the same gravitational pull a 2% ownership dart exerts on GPP lineups. Compare the North Carolina Senate board, where nearly all the trading volume sat in the heavy favorite's market on our August 11 read; here the activity flips to the long shot. Whether 15 cents is actually cheap is the only question that pays, and it is the one the GPP-versus-cash mindset exists to keep honest: a large payout never made a probability wrong-priced.

How A Contract Price Becomes A Probability

For readers newer to event contracts than to lineups: a Yes contract pays $1.00 if the outcome happens and $0 if it does not, so the price in cents reads directly as an implied probability. A 34-cent Yes is a 34% market estimate, and the Democratic side's 85.5-cent midpoint reads as about 85%. Two adjustments apply before you treat the screen as a fair number:

1. The overround. The two ask prices sum to 101 cents, one cent of juice; normalized, the pair lands at about 85.1% and 14.9%. A sportsbook two-way runs 104 to 105 cents on the same math, which is why exchange prices convert nearly at face value while sportsbook lines need a de-vig pass first.
2. Fees. Kalshi's standard taker fee is 0.07 times price times one minus price, per contract, so effective odds sit slightly worse than the screen. The fee peaks at 50-cent prices (1.75 cents per contract) and shrinks toward the extremes, which is where this market lives.

The expected-value habit from DFS transfers directly: nothing here is a bet on who wins, it is a comparison between the price paid and the probability you believe, after fees. The cleanest way to see the fee drag is to walk a real order on each side.

A Worked Example: Two 100-Contract Orders, Fees Included

At the August 12 asks: one hundred Democratic Yes contracts at 86 cents cost $86.00 plus a taker fee of 0.07 x 100 x 0.86 x 0.14, which Kalshi rounds up to $0.85, for $86.85 committed. A Democratic House win pays $100.00, a $13.15 profit, a 15.1% return over roughly three months if media calls settle it around election night and closer to six if it runs to the February 1, 2027 backstop; a loss burns the full $86.85. One hundred Republican Yes contracts at 15 cents cost $15.00 plus a rounded $0.90 in fees, $15.90 committed, and pay $100.00 on a hold, an $84.10 profit against a $15.90 loss. In bankroll terms these are the cash-game entry and the tournament dart, and the only question either order should turn on is whether the price sits below your true probability, a judgment we leave to the reader. Describing the board is not a recommendation to trade it.

What Would Move This Price Before November

The catalysts sort onto the layers this page has kept separate, and they are not equal. On the environment side: whether the generic-ballot band drifts, and what special elections keep saying about turnout. On the map side: the redistricting litigation over the Texas and Proposition 50 lines and their imitators, which can still add or subtract competitive districts and is the one input the postwar base-rate table knows nothing about. If I had to rank them, the map outranks the polls: with the flip threshold at three seats, a single court ruling that redraws three to five likely seats moves this market more than a one-point generic-ballot wobble ever could. The callback worth ending on is that busy 15-cent book from the volume board: 13 million contracts of activity on the long-shot market means that if the map litigation or the polls genuinely tighten this race, there is a deep, liquid place for that repricing to happen fast. It is the same weighting exercise as the Texas Senate market, a near-coinflip on our last read, just run at the opposite end of the probability scale: every implied probability is a blend of base rate and environment, and the weights are the whole disagreement. Every board we track lives on the Stokastic sports prediction markets hub. If you would rather run projection-versus-ownership weighting where it settles every night instead of every two years, the MLB DFS projections, ownership, and stacks in the DataHub run the identical exercise on a daily slate, and you can try the Stokastic Sims free to move the inputs yourself.

Prices, volume, and open interest fetched from the live Kalshi book on August 12, 2026, event CONTROLH-2026. Kalshi is a CFTC-regulated exchange available to adults 18+ in eligible states as of August 2026; availability varies by state, so check the platform's own eligibility screen. Polling figures reference public generic-ballot averages as of early August 2026. This page is market analysis, not a recommendation to trade any contract, and not financial advice.