Polymarket vs Kalshi 2026 Midterms: Two Rulebooks
Published 29 October 2026 · Prices frozen 29 October 2026, 21:05 UTC
Key takeaways
- Four separate books price a Democratic Senate between 67.05% and 67.72%. Polymarket’s direct board 67.50¢, Kalshi 67.50%, the Balance of Power legs 67.05% and the Republican seat ladder 67.72%. Five days out, the spread across every venue and every structure is 0.67 points.
- They are not the same contract. Kalshi can settle on the identity of one Senate officer in February. Its fallback is the party of the President pro tempore on 1 February 2027. Polymarket counts seats by the party under which each member was elected, so a post-election party switch changes one venue and not the other.
- The clause that splits them has a 42.6% chance of binding. Four Polymarket state boards carry an independent leg — Nebraska 30.5%, Montana 10.0%, South Dakota 4.5%, Idaho 3.9% — and three of those candidates have publicly refused to caucus with either party. Polymarket may then hang until 4 January 2027.
- The two chambers are priced as very nearly one bet. Conditional on a Democratic Senate the House is 98.9% Democratic; conditional on a Republican Senate, 80.3%. Run it backwards and a Republican House puts the Democratic Senate at 10.3%.
- Nothing is tradeable in the gap. The Senate quotes are identical at 67 bid and 68 ask on both venues. The House leaves two tenths of a cent in either direction, which Kalshi’s trading fee absorbs. The disagreement here is about terms, not price.
Put the Polymarket vs Kalshi 2026 midterms odds side by side tonight and you get the least interesting answer available: 67.5% and 67.5%. Same number, same bid, same ask. Every tracker that shows the two venues in adjacent columns is quietly telling you there is nothing to see here, and on price it is right.
It is also comparing two contracts that resolve on different events. Polymarket’s three midterm control boards carry $85.4m between them and settle on a seat count. Kalshi’s Senate book has traded 9.11m contracts and settles, if the networks do not hand it an answer first, on which party the President pro tempore of the Senate belongs to on 1 February 2027. Those are not the same question, and there is a 42.6% chance this election is the one that notices.
We read both rulebooks in full rather than the summary text, pulled every leg of all four Polymarket boards and both Kalshi chambers in one pass, divided the Balance of Power branches out by hand to recover the conditionals the leg prices hide, and bounded each gap at the bid and the ask instead of mid to mid. All prices are frozen at 29 October 2026, 21:05 UTC; the live boards are linked throughout.
Where the 2026 midterms odds stand on both venues
Start with the four numbers that matter. Polymarket’s Which party will win the Senate in 2026? board is the one that gets quoted, and at $10.90m it is the deepest single Senate contract on either venue. Kalshi’s CONTROLS-2026-D sits on precisely the same quote. Then there are two more Polymarket structures that answer the same question sideways: the Balance of Power board, which prices both chambers jointly, and the Republican Senate seat ladder, which prices the count rung by rung.
| Book | Implied D Senate | Book sum | Volume |
|---|---|---|---|
| Polymarket direct (32224) | 67.50% | 100.00¢ | $10.90m |
| Kalshi CONTROLS-2026-D | 67.50% | 100.00¢ | 9.11m |
| Balance of Power (32228) | 67.05% | 100.30¢ | $43.95m |
| R seat ladder (106193) | 67.72% | 101.15¢ | $4.17m |
That is a remarkable amount of agreement. Four books, two exchanges, three completely different contract shapes, and the widest pair is 0.67 points apart. Compare it with what we found on the same chamber three days ago, when Polymarket’s five conditional Senate boards were scattered across two full points and left 1.3¢ on the table. The boards people actually trade are coherent. The ones they do not, are not.
The House is nearly as tight. Polymarket’s House control board pays 92.5¢ against Kalshi’s 91.75% midpoint — three quarters of a point, on $30.56m and 24.15m contracts. So the honest headline is that the two venues agree, and the interesting question is what they agree about.
The same price on two different questions
Kalshi’s Senate contract says it “may be determined early based on a consensus of media calls projecting control of the U.S. Senate” and that otherwise “victory will be determined by the party identification of the President pro tempore of the Senate on February 1, 2027”. Its House contract says the same with the Speaker. Both books list a close time of 1 February 2027, not election night.
Polymarket does not mention either officer as its primary test. It counts seats: a party wins control by taking a majority of the chamber’s voting seats, and critically, “seats are attributed to the party under which the member was elected”. If a member-elect switches party or vacates between election day and the organisation of Congress, Polymarket still books the seat to the party that elected them. Kalshi’s fallback is the opposite: it looks at who holds the gavel in February, which is exactly the thing a switch changes.
| Term | Polymarket | Kalshi |
|---|---|---|
| Primary test | Majority of voting seats | Consensus of media calls |
| Fallback | Half the seats + Vice Presidency | Pro tempore / Speaker party, 1 Feb 2027 |
| Seat attribution | Party that elected the member | Not seat-based |
| Post-election switch | Ignored | Can decide it |
| Independents | Counted on caucus announcement | Not addressed |
| Listed close | 4 Nov 2026 04:59Z | 1 Feb 2027 15:00Z |
| Can hang until | 4 Jan 2027 23:59 ET | 1 Feb 2027 |
Read that table as two different bets and the identical price starts to look less like consensus and more like coincidence. In the overwhelming majority of outcomes it does not matter: one party clears 51 seats, the networks call it, both contracts pay the same side within hours. The divergence needs an election that is ambiguous in a specific way — and this is the year that ambiguity has a name.
The 4 January clause and the four independents
Polymarket’s Senate rules carry a paragraph Kalshi has no equivalent of. Members elected as independents count toward a party’s total only once they “formally caucus with, or, for newly elected members, who announce their intent to caucus with” it. If that decision cannot be determined on election day, the market “may remain open until the earlier of (i) an announcement from the relevant members of their intents to caucus, or (ii) January 4th, 11:59 PM ET of the following calendar year”. And if no announcement comes by then, the member “will not count towards the seat count of any party”.
That clause is usually decorative. In 2026 it is not, because four of the Senate boards on Polymarket carry a live independent leg, and the candidates behind three of them have spent the autumn saying they will not caucus with anybody.
| Race | Candidate | Independent leg | Board volume |
|---|---|---|---|
| Nebraska (59204) | Dan Osborn | 30.50% | $1.13m |
| Montana (57660) | Seth Bodnar | 10.00% | $294,503 |
| South Dakota (57670) | Brian Bengs | 4.50% | $74,572 |
| Idaho (57643) | Todd Achilles | 3.90% | $440,511 |
| P(at least one wins) | — | 42.59% | — |
The Nebraska board is the one to watch, and it is worth seeing how strangely it reads. Polymarket prices the Democratic leg at 0.05¢ and the independent leg at 30.5¢, because its state rules say candidates running as independents “will not be encompassed by the ‘Democrat’ or ‘Republican’ options regardless of any affiliation they may have with the party”. The Nebraska Democratic Party has endorsed Osborn and declined to field a nominee of its own. So the board showing Democrats at a twentieth of a cent is not a forecast about which caucus gains the seat; it is a statement about ballot lines.
Osborn has been explicit that he “will not caucus with either of the two major parties if elected”, has floated a separate independent caucus, and has put it as “I’m going to caucus with everybody”. Bengs in South Dakota has said he would caucus with Osborn, which is not a party. Bodnar in Montana has run on the same pitch. If one of them wins a seat that decides the chamber and holds that line into January, Polymarket’s board sits open while Kalshi’s waits for a pro tempore vote — and the two can land on opposite sides.
A 50-50 Senate goes Republican, priced at 10.4%
The arithmetic everyone gets wrong on the 2026 midterms odds is the threshold. Republicans hold 53 seats going into 3 November, Democrats 45 plus the two independents who caucus with them. A majority is 51, so Democrats need a net gain of four, not three: at 50-50 Vice President JD Vance breaks the tie, and Polymarket’s rules say so in advance by awarding a tied chamber to the party holding half the seats and the Vice Presidency.
| Republican seats | Raw | Normalised | Meaning |
|---|---|---|---|
| 47 or fewer | 42.50¢ | 42.02% | Clear D Senate |
| 48 or 49 | 26.00¢ | 25.70% | Narrow D Senate |
| Exactly 50 | 10.50¢ | 10.38% | Tie, VP holds it for R |
| 51 or more | 21.90¢ | 21.65% | R majority outright |
The ladder is the cleanest independent read available, and it lands at 67.72% for the first two bands against the direct board’s 67.50¢. But it is worth being precise about why that is not quite a check on the same number. “Republicans 49 or fewer” implies Democrats reach 51 only if every non-Republican seat is Democratic. Put an uncommitted independent in the chamber and you can have Republicans at 49, Democrats at 50 and nobody at a majority. The ladder and the control board agree to within a quarter of a point on two questions that come apart in precisely the 42.6% of worlds discussed above.
What the balance-of-power board says
The Balance of Power board is the biggest midterm market on either exchange at $43.95m, and because its five legs partition every outcome you can divide one branch by itself to recover conditional probabilities neither single-chamber board publishes. This is the same technique we used on the conditional oil boards, with one important difference: here the volume is on the board doing the work rather than on a side bet.
| Quantity | Value | Compare | Gap |
|---|---|---|---|
| P(D House | D Senate) | 98.88% | 92.50% direct | +6.38 pts |
| P(D House | R Senate) | 80.30% | 92.50% direct | −12.20 pts |
| P(D Senate | D House) | 71.51% | 67.50% direct | +4.01 pts |
| P(D Senate | R House) | 10.34% | 67.50% direct | −57.16 pts |
| D sweep, if independent | 62.44% | 66.30% priced | +3.86 pts |
Two things fall out. First, the chambers are priced as close to a single bet: a Democratic Senate drags the House to 98.9%, and a Republican House drags the Democratic Senate down to 10.3%, a 57-point move from the unconditional price. Second, the sweep is worth 3.86 points more than multiplying the two single-chamber prices together, which is the market saying out loud that it does not treat these as separate events. A national swing big enough to take the Senate takes the House by definition.
That last row is also a warning about the kind of comparison this article started with: build a Senate price by adding two balance-of-power legs on one exchange and read a direct contract on the other, and you are comparing a joint probability with a marginal one. Our methodology page sets out how we normalise before quoting anything side by side.
Where the venues differ, at the bid and the ask
Mid-to-mid gaps are the standard way to compare two exchanges and they overstate everything, because nobody trades at the mid. So here is every cross-venue midterm gap at the prices actually showing.
| Contract | Polymarket | Kalshi | Cross |
|---|---|---|---|
| Senate D | 67.00 / 68.00 | 67.00 / 68.00 | none |
| Senate R | 32.00 / 33.00 | 32.00 / 33.00 | none |
| House D | 92.00 / 93.00 | 91.70 / 91.80 | 0.20¢ |
| House R | 7.00 / 8.00 | 8.20 / 8.40 | 0.20¢ |
The Senate is a dead heat to the cent on both sides of the book, which almost never happens between two venues with different user bases, fee models and funding currencies. The House leaves two tenths of a cent twice over, and in opposite directions, which is the signature of Kalshi’s tighter tick rather than a disagreement: it quotes the House in tenths of a cent where Polymarket quotes in whole cents, so its 91.70/91.80 sits inside Polymarket’s 92.00/93.00 rather than against it.
One note on depth, since most comparisons get it backwards. Kalshi’s Republican House contract has traded 56.50m contracts against 24.15m on the Democratic side, and its Republican Senate book shows 12.88m against 9.11m. On both chambers the bigger book is on the losing side, which is what you expect where hedging flows dominate.
How the Senate price moved into the final week
The agreement in the first table is a snapshot, and the snapshot is less stable than it looks. Polymarket’s Democratic Senate leg has covered a five-point range in thirty days and a three-point range in the last four days alone.
| Date | D Senate | D House | Note |
|---|---|---|---|
| 30 Sep | 62.50¢ | 92.50¢ | Month open |
| 10 Oct | 62.50¢ | 90.50¢ | Senate low |
| 22 Oct | 67.50¢ | 92.50¢ | First touch of 67.5 |
| 26 Oct | 64.50¢ | 91.50¢ | Our last Senate piece |
| 29 Oct | 67.50¢ | 92.50¢ | Tonight |
Kalshi tells the same story: CONTROLS-2026-D shows a previous price of 65.00 against a last trade of 68.00, three points in a session. Which matters, because every competing page we checked while researching this one was quoting prices from 13 to 22 October, when the Senate sat between 63.5¢ and 64.5¢. Four points of stale data is more than the entire cross-venue gap anyone is trying to measure.
What would move this before 3 November
Five days out, the things that can still move a 67.5¢ Senate are narrow. Six Republican-held seats are the whole ballgame — Maine, Ohio, Alaska, Texas, Iowa and Kansas — against Michigan as the one genuinely contested Democratic hold. Democrats need four of them. Polymarket runs a board on each, and the ones to watch for a late repricing are the two where the market has already moved furthest from the fundamentals: Alaska, where it prices a Democratic flip at 71%, and the independent races that cannot resolve into either column.
For the rulebook divergence specifically, there is exactly one thing to watch, and it is not a poll. If Nebraska is called for Osborn on the night and the Senate is anywhere near 50-50, the two venues stop being interchangeable immediately: Kalshi’s contract starts waiting on February and Polymarket’s starts waiting on a sentence from one man in Omaha. A 30.5% leg on a $1.13m board is carrying a lot of structural weight for something no price tracker mentions.
And if you take one number from this page, take the one neither venue prints. The headline comparison is 67.5% against 67.5%, a dead heat on the deepest political market of the cycle. The probability that those two identical prices are answering the same question is about 57%.
Related analysis
- 2026 midterms odds: the Senate flipped in 30 days
- 2026 House odds: markets 90¢, models 76-93%
- 2026 Senate odds and oil — five conditional boards, five prices
- Polymarket vs Kalshi: Venezuela leader odds compared
- 2026 governor race odds: 36 boards vs one ladder
- 2026 midterms hub — every Senate and governor race we track
New to prediction markets?
New traders get a $20 bonus
Deposit $10 on Polymarket and get a $20 trading bonus to put behind your first call.
See the 2026 midterms boards on Polymarket →New Polymarket users only. Bonus is trading credit, expires 30 days after it lands, and terms are set by Polymarket — check the current offer at signup. 18+. Fee-free markets. GeoOdds earns a commission on signups through this link.
Frequently asked questions
What are the Polymarket vs Kalshi 2026 midterms odds right now?
At 21:05 UTC on 29 October 2026 both venues price a Democratic Senate at 67.5%, on the same 67/68 bid-ask. On the House, Polymarket pays 92.5¢ against Kalshi's 91.75% midpoint, a gap of 0.75 points. Polymarket's three midterm control boards carry $85.4m between them; Kalshi's Senate book has traded 9.11m contracts.
Do Polymarket and Kalshi resolve Senate control the same way?
No, and this is the part price trackers leave out. Kalshi may settle early on a consensus of media calls, and failing that on the party of the President pro tempore on 1 February 2027. Polymarket counts a majority of voting seats, attributing each seat to the party under which the member was elected, with a 50-50 tie going to the party holding the Vice Presidency. The two can pay opposite ways on the same election.
What is Polymarket's 4 January clause?
Polymarket counts a member elected as an independent toward a party only once that member caucuses with it, or announces an intent to. If that cannot be determined on election day, the market may stay open until the earlier of an announcement or 4 January 2027 at 11:59 PM ET, after which the member counts toward no party. The Republican seat ladder carries the identical clause.
Which independents could trigger it?
Four Polymarket Senate boards carry an independent leg: Nebraska 30.5%, Montana 10.0%, South Dakota 4.5% and Idaho 3.9%. Treating them as independent events, the chance at least one wins is 42.6%. Dan Osborn, Seth Bodnar and Brian Bengs have all said publicly they would not caucus with either major party, which is the condition the clause describes.
How many seats do Democrats need for the Senate in 2026?
Fifty-one. Republicans hold 53 seats going in, so Democrats need a net gain of four. A 50-50 chamber is not enough, because Vice President JD Vance casts the tiebreaking vote and Polymarket's rules award a tied chamber to the party holding half the seats plus the Vice Presidency. Polymarket's Republican seat ladder prices exactly 50 Republican seats at 10.4%.
Can you arbitrage the gap between Polymarket and Kalshi?
Not meaningfully. The Senate quotes are identical on both venues, 67 bid and 68 ask, so there is nothing to cross. The House leaves 0.2¢: buy Kalshi's Democratic side at 91.80 and sell Polymarket's at 92.00. Kalshi charges trading fees and Polymarket does not, which absorbs two tenths of a cent. Nothing here is betting advice.
Are the House and Senate priced as one bet or two?
Close to one. Polymarket's Balance of Power board puts a Democratic House at 98.9% conditional on a Democratic Senate, and 80.3% conditional on a Republican one. Conditional on Republicans holding the House, a Democratic Senate is only 10.3%. The joint sweep is priced 3.9 points above the product of the two single-chamber prices, so the market treats the chambers as correlated.
Sources: the Polymarket Gamma and CLOB public APIs (events 32224, 32225, 32228 and 106193, plus state boards 59204, 57660, 57670 and 57643) and the Kalshi public markets API, all read at 29 October 2026, 21:05 UTC. Polymarket’s resolution terms are quoted from the Senate control board and Kalshi’s from the CONTROLS-2026-D contract, in both cases the operative text the venue publishes rather than a summary. The 53-45-2 going-in split, the six competitive Republican seats and the independent candidacies are as described here. Kalshi volume and open interest are contract counts; Polymarket volume and liquidity are dollars, and the two are never added on this page. Prices here are a frozen snapshot; the live numbers are on the market pages, which refresh every 10 minutes. Nothing here is financial or betting advice.