House 91%, Senate 64%: why the midterm odds are so far apart

Last checked 2026-10-26

Short answer

The House and Senate odds differ because the two chambers are won in different ways. On 25 October, DeFi Rate's tracker of Kalshi and Polymarket gave Democrats 91% for the House and 64% for the Senate. The House turns on a national swing across 435 seats. The Senate needs a net gain of four from a short list of close states, and Democrats must win most of them at once.

• What happened On 25 Oct, DeFi Rate's tracker put Democrats at 91% for the House and 64% for the Senate. Neither figure moved over the week.
• Why the gap Senate control needs a net gain of four from a fixed set of mostly Republican-held states. Several of those races are priced between 40% and 68%.
• Why it matters abroad Only the Senate confirms Fed nominees. Tariff bills and Ukraine aid need both chambers.
• What to watch Election day is 3 November 2026. Ohio, Texas and Kansas are the close races where markets and polls have diverged or moved most this month.

Key facts

House odds, 25 Oct (DeFi Rate)	Democrats 91% (Kalshi 90.4¢, Polymarket 93¢)
Senate odds, 25 Oct (DeFi Rate)	Democrats 64% (Kalshi 64.6¢, Polymarket 65¢)
Current Senate	53 Republicans, 47 Democrats; Democrats need a net gain of four
Current House	Republicans 218, Democrats 214, one independent, two vacancies; 218 is a majority
Silver model, 8 Oct	Democrats 74.8% Senate, 93.1% House
Election day	3 November 2026

The number: a 27-point gap between two chambers

Start with 91 and 64. Those are the chances DeFi Rate's tracker gave Democrats of winning the House and the Senate on 25 October, combining prices from Kalshi and Polymarket. Neither number had moved in a week. Kalshi's own newsroom put the Senate at 62% on 15 October, so check the timestamp and source of any figure you see.

The two platforms do not quote exactly the same prices. For the House, Kalshi stood at 90.4¢ and Polymarket at 93¢. For the Senate, the figures were 64.6¢ and 65¢. A 93¢ contract pays $1 if Democrats win, so the price works as the market's implied probability. These are prices set by traders' bets at a stated time. They are not polls and they are not forecasts.

Money is flowing in. DeFi Rate counted $9.4M placed on election markets in a single 24-hour period. Polymarket's balance-of-power market had traded more than $12 million by 21 September.

Why is the House so much safer for Democrats than the Senate?

The House is a big, broad target. All 435 seats are on the ballot, and Republicans hold 218 to Democrats' 214, according to Kalshi. A small net gain hands Democrats the 218 they need. When a national mood moves across hundreds of districts at once, a few upsets in either direction matter less.

The Senate is a short, fixed list. Thirty-five seats are up this year, 22 held by Republicans and 13 by Democrats, Kalshi says. Democrats start at 47 seats to 53 and need a net gain of four. Cook Political Report, as summarised by 270toWin, lists Kansas, Ohio, Texas, Maine, Alaska, Iowa and Michigan as competitive or toss-up.

This is where the arithmetic bites. DeFi Rate warns that being favoured in several close races one at a time is not the same as winning all of them together. A party can be the favourite in each of four races and still more likely than not to lose at least one. The races are also linked: a polling error that hurts Democrats in one state could well show up in others too.

For illustration only: take two races priced at 68% and 65%, like Ohio and Texas on DeFi Rate's tracker. If they were independent, the chance of winning both would be the two prices multiplied together, which comes out below even odds. Real races are correlated, so the true joint figure differs, which is why the chamber market is the number to watch.

Polymarket's balance-of-power market makes the combinations visible. It offers four outcomes: a Democratic sweep, a Republican Senate with a Democratic House, a Republican sweep, and a Democratic Senate with a Republican House.

Which Senate races carry the most weight?

Ohio is the unusual one. It is a special election for the rest of JD Vance's term, which runs through the end of 2028. After Vance became vice president, Governor Mike DeWine signed the certificate appointing Jon Husted on 19 January 2025. So Husted is an appointed incumbent and has not yet faced voters for this seat. His Democratic opponent, Sherrod Brown, lost Ohio's other Senate seat to Bernie Moreno in 2024 by about 3.6 points. The Ohio Secretary of State certified Brown, Husted and Libertarian William Redpath for the general election.

Texas and Kansas are the states that could stretch the map. Kalshi priced Democrat James Talarico at 62% in Texas on 15 October, while an Emerson poll from August had Republican Ken Paxton at 47% to Talarico's 46%. In Kansas the pattern runs the other way: Emerson's September poll had Democrat Adam Hamilton at 45% to Republican Roger Marshall's 43%, but Kalshi gave Marshall 65%. Emerson's pollster called it Democrats' best chance to win a Kansas Senate seat since 1932.

Markets and polls do not always agree, and the polls here are weeks older than the prices. Both Kansas and Texas leads are within the stated margins of error.

• DeFi Rate, 7 days to 25 Oct: Ohio special (Brown vs. Husted) from 65% to 68%, Texas from 62% to 65%, Kansas Democrat from 38% to 40%.
• Kalshi, 5:43 p.m. ET on 15 Oct: Maine 57% Jackson (D), Iowa 58% Hinson (R), Alaska 75% Democratic, North Carolina 95% Cooper (D).

Race	Kalshi, 15 Oct	DeFi Rate, 25 Oct	Latest poll on our sheet
Ohio (special)	60% Brown (D)	Ohio special (Brown vs. Husted): 68%, up from 65%	Fox News, 15-19 Oct: Brown 52%, Husted 47%
Texas	62% Talarico (D)	Texas: 65%, up from 62%	Emerson, 9-10 Aug: Paxton 47%, Talarico 46%
Kansas	65% Marshall (R)	Kansas Democrat: 40%, up from 38%	Emerson, 15-16 Sept: Hamilton 45%, Marshall 43%

What do forecast models say compared with markets?

Nate Silver's FLIPR model was more optimistic for Democrats on the Senate than the markets are now. On 8 October it gave them 74.8% for the Senate, up from 56.2% on 5 August, and 93.1% for the House, according to Newsweek. Its average outcome was 52.2 Senate seats and 238.7 House seats for Democrats.

On the House, model and markets roughly agree. On the Senate, the model's 74.8% sits about ten points above the 64% to 65% that traders were paying on 25 October. The model figure is older, and we could not verify any later update.

Why would a House-only win matter outside the US?

Some powers belong to the Senate alone. Under the Constitution, it confirms nominees such as Fed governors and judges, and it ratifies treaties. Fed picks are the clearest recent example. CNBC reported that the Senate confirmed Kevin Warsh as a Fed governor 51-45. Al Jazeera reported on 13 May that the Senate then confirmed him as chair 54-45, almost entirely along party lines. The House had no vote.

Trade works differently. A Reuters scenario analysis, syndicated on Investing.com, says that if Democrats win only the House, Trump keeps broad tariff authority and Senate Republicans keep confirming his nominees. With both chambers, Democrats could pass tariff bills and force him to veto them.

Ukraine aid is a spending matter, so it must pass both chambers. Rep. Jim Himes, the top Democrat on the House Intelligence Committee, told Breaking Defense that Democratic control would mean aid packages move "not in nine months, but in nine days or nine weeks". He tied that to controlling both chambers and the budget. It is one partisan lawmaker's view of a scenario, not a prediction.

What is not confirmed, and what comes next?

Polymarket's own market page could not be reached, so its figures here come from DeFi Rate and other trackers.

The one scheduled date is election day, 3 November 2026. Prediction-market legality differs by country and by US state, and nothing here is a suggestion to trade.

FAQ

Sources

• Emerson College Polling – Kansas
• Kalshi News – Senate control
• Kalshi News – House control
• DeFi Rate midterms tracker
• DeFi Rate Senate odds
• Newsweek – Silver Bulletin forecast
• Al Jazeera – Warsh confirmation

Figures change. Check the primary sources above.