Prediction Markets Put Democrats at 94-95% for House on Election Day

Prediction markets put Democrats' chances of retaking the U.S. House at 94% to 95% on Election Day morning, according to Kalshi and Polymarket respectively. The same platforms priced a Democratic Senate takeover at 77% on Kalshi and 78% on Polymarket. Polymarket maintains a data partnership with MarketWatch's publisher, Dow Jones.

The numbers describe a sweep scenario. Yet the specialists watching these markets are not convinced the odds tell the whole story.

Are the markets too bullish on Democrats?

Brian Schaffner, a professor of political science at Tufts University who specializes in public opinion, argues the markets have drifted ahead of the evidence. "Both the House and Senate predictions seem a bit more bullish on Democrats than the prominent forecast models out there," Schaffner told MarketWatch in an email.

He went further on the Senate. "In particular, I see the Senate outcome as a bit more of a toss-up than the markets do," he said. "In general, the markets seem to be following the polls and forecast models, but a bit more decisively, which runs the risk of suggesting too much confidence in a good night for Democrats."

The distinction matters for anyone reading the percentages as certainties. A 77% probability still leaves roughly a one-in-four chance of the opposite outcome — the sort of gap that decides Senate control. Schaffner's caveat applies to both chambers, not just the closer Senate race.

Wall Street's Washington analysts share some of that skepticism. Henrietta Treyz at Veda Partners wrote in a Monday note: "contrary to conventional wisdom, I still expect Republicans to narrowly hang on to the U.S. Senate with a 50-50 split and Vice President J.D. Vance breaking any ties."

Takeaway: the headline odds favor Democrats, but at least two named analysts think the markets have overshot the underlying data.

Which races will decide the Senate?

The Senate fight likely turns on six contests rated as "toss-ups" by Cook Political Report analysts:

• Alaska
• Iowa
• Kansas
• Maine
• Ohio
• Texas

Michigan's contest also belongs on the watch list. Cook Political Report had rated it a "toss-up" until Monday, when it shifted the race to "leans Democratic."

Republicans currently hold a 53-47 Senate majority, meaning Democrats need a net gain of four seats if they do not control the tiebreaking vote, or a more modest shift if they do. In the House, Republicans defend a slim 219-214 edge.

Takeaway: seven races, six of them coin flips on paper, will determine whether the Senate's 77-78% implied probability pays off.

Can prediction markets be wrong?

They have been, twice, in recent memory. In the 2022 midterms, markets signaled an approaching red wave; voters delivered what analysts called a red ripple instead, with Republicans winning only a slim House majority and failing to take the Senate.

The miss is not confined to U.S. politics. A week before this Election Day, betting markets made Brazilian politician Flávio Bolsonaro a strong favorite to win his country's presidential election. Incumbent Luiz Inácio Lula da Silva won.

Takeaway: today's percentages are market prices, not forecasts with a track record that earns blind trust.

How long will results take?

Investors may not get clean answers tonight. The 2024 House contest and the 2022 fight for the chamber both required days of vote-counting. In 2020, the presidential election took days to call, and Senate control remained undetermined for two months.

Many analysts have predicted GOP losses because of voter frustration with elevated prices for gasoline and other essentials. Midterm elections have also been bruising for most modern presidents — a historical pattern working against President Donald Trump's Republican Party.

Voters are not choosing a president today, but the stakes for portfolios are concrete. The outcome affects fuel prices, taxes including tariffs, healthcare expenses, housing costs and the artificial-intelligence boom. One governor's race, MarketWatch noted, could even derail the AI trade that has powered the stock market.

Takeaway: the practical question for markets is not just who wins, but how quickly the count produces a verdict both parties accept.