On Election Day in the United States, the largest prediction markets have priced a Democratic sweep of Congress at around 75%, putting their own credibility on the ballot alongside every contested House and Senate seat. Polymarket, the crypto-based exchange where users trade contracts on real-world outcomes, and Kalshi, its federally regulated American rival, both show Democrats favored to capture both chambers, and tonight's count will grade those prices in public.
The test is bigger than a single election. These platforms built their reputation in 2024 by pricing Donald Trump as the likely winner over Kamala Harris while much of the polling industry hedged. They have missed since, in Brazil and in Israel, and a Republican hold on either chamber tonight would stack a third straight electoral miss on the record and file the markets alongside the pollsters they claim to have replaced.
What the markets actually price
A sweep means winning both chambers, and the markets break that into parts. Control of the House of Representatives is priced as near-certain for Democrats, around 93% on Kalshi and as high as 95% on Polymarket. The Senate is the live question: Kalshi puts Democratic control near 71%, Polymarket between 72% and 77%, depending on the hour.
Because a sweep requires both and the House is close to settled, the roughly 75% headline is effectively the Senate number wearing a different label. That is the contract to watch tonight. If it holds, the markets were right; if the Senate stays Republican, the sweep contract pays nothing and the accuracy story collapses regardless of the House result.
How the price moved
The Democratic position did not begin the autumn this strong. On October 27, a week before the vote, the sweep was priced at 63% on Kalshi and 65% on Polymarket. Over the following week it climbed into the mid-70s, with the Senate contract doing most of the moving as late polling and early-vote data firmed up.
That drift is the mechanism prediction markets sell: a price is an implied probability that updates continuously as money takes sides, rather than a single survey frozen on the day it was taken. The question tonight is whether continuous updating produced a sharper forecast or just a more confident wrong one.
The 2024 reputation they are defending
The reason anyone is watching traces to a single night in 2024. The markets priced Trump as the favorite over Harris and were vindicated at a moment when traditional surveys showed a near-tie and cable panels treated the race as a coin flip. That result became the industry's founding story: the crowd, with its own money at stake, saw what the pollsters would not say.
The story has a flaw. A single correct call is not a track record, and the platforms have leaned on 2024 to explain away everything since. Tonight is the first US national test since that night, which is why the people who run these venues, and the traders holding the Democratic contracts, have more riding on it than the dollar figures suggest.
The misses in between: Brazil and Israel
Between 2024 and tonight the record frayed. The markets misread Brazil, and they misread Israel, where Polymarket also drew scrutiny over suspicious trading on Israeli races that raised the question of whether some positions reflected forecasting or something closer to inside knowledge. For Israeli readers who followed that episode, tonight is the cleaner test the Israel market never was: a high-liquidity, closely watched contest where the price can be checked against a result within hours.
Misses in Brazil and Israel are a pattern starting to form. A third would be hard to dismiss as noise, and it would hand the pollsters, who spent a decade being told the markets had beaten them, an answer.
Nearly $1.9 billion in play
The money behind these prices is real and large. As of late Monday, close to $1.9 billion had been traded on the 2026 midterms across the two platforms, roughly $1.3 billion on Kalshi and about $600 million on Polymarket, spread across more than 7,000 distinct outcomes covering individual races, chamber control and combinations.
Large, but not a record. Polymarket alone saw more than $3.6 billion trade on the Trump-Harris contract in 2024, so even the combined midterm figure sits well below a single presidential market. The gap reflects a structural truth that analysts keep returning to: political markets draw a fraction of the volume that flows through sports contracts, yet they carry a far longer reputational tail. A missed basketball line is forgotten by morning; a missed election is cited for years.
The skeptics, including the Speaker
Not everyone accepts the prices. House Speaker Mike Johnson, a Republican, publicly dismissed the polling that points to a Democratic wave, telling reporters: "We've been from coast to coast...and I'm telling you, the polling is inaccurate." His argument is the one that has shadowed every forecast since 2016, that the visible numbers miss a quieter electorate, and tonight it meets the markets head on.
If Johnson is right and the markets are wrong, the roughly 75% Democratic price becomes the most expensive mistake these platforms have made in front of an audience this size. If he is wrong, the markets add a US national election to the 2024 result, and the claim that they beat the pollsters gets its strongest evidence yet.
What tonight settles
By the time the Senate races are called, the sweep contract will have resolved to a plain yes or no, and with it the question these platforms cannot talk their way around. Polymarket and Kalshi spent the stretch since 2024 telling the public that a price is a sharper read on reality than a poll. Tonight that claim stops being a sales pitch and becomes a score: 2024 plus the midterms, or a losing streak that runs from Brazil through Israel to the United States. The traders holding the Democratic contracts will learn their answer in dollars; the platforms will learn whether their founding story survives contact with a second data point.







