Election prediction markets: where to trade and what they mean
Published 2026-05-22 · Last reviewed 2026-05-22
TL;DR
- Election prediction markets price the probability of political outcomes using real-money trading.
- Kalshi (US-legal), Polymarket (non-US), and PredictIt (US-legal, capped) are the major venues.
- Prices typically — but not always — outperform polls and pundits as forecasts.
Election Twitter is loud. Election markets are louder — real money on the line, and they moved faster than every poll in 2024.
What are election prediction markets?
Election prediction markets are exchange-traded contracts that pay $1 if a defined political outcome happens and $0 if it doesn't. The market price equals the implied probability of YES. A "Will Democrats hold the Senate?" contract trading at 41¢ means the market believes there's a ~41% probability.
Where to trade them
- Kalshi. CFTC-regulated; accepts US users. Lists party-control contracts and individual races at scale. Standard 1–7% trading fees.
- Polymarket. On-chain venue with the deepest election volume globally. 0% trading fees. Blocks US users under its terms.
- PredictIt. US-legal under an academic CFTC no-action letter. $850-per-market cap. Best long-running political coverage.
- Manifold. Play-money. Excellent for testing your calibration; not real exposure.
How accurate are they?
The academic consensus: in the final weeks of a closely-watched race, prediction-market prices generally outperform polls and most pundit forecasts. The mechanism is simple — traders who think prices are wrong put real money behind their view, which moves prices toward the better-informed estimate.
Caveats: thin markets are manipulable, and markets occasionally reflect the bias of their userbase rather than the true probability. The 2024 cycle produced several well-documented examples in both directions.
How to read an election market
Three numbers matter: last price (the most recent trade), the bid-ask spread (how confident the market is — narrow spread means consensus, wide spread means uncertainty), and volume (how much real money is behind the price).
A 65¢ contract with a 1¢ spread on $500k of volume is a meaningful signal. The same 65¢ on $500 of volume is barely an opinion.
Election markets vs polls
Polls measure stated preference; markets aggregate beliefs about outcomes, including conditional logic ("if turnout is high, then..."). The two complement each other — polls inform markets, and savvy market readers correct for known poll biases.
Related
Frequently asked questions
Where can I trade US election prediction markets?
In the US: Kalshi (CFTC-regulated, broad election contracts) and PredictIt (academic exemption, $850/market cap). Outside the US: Polymarket has the deepest election market depth by volume.
Are election prediction markets accurate?
Generally more accurate than polls in the final weeks of a race, but they're not infallible. Markets aggregate public information; they're vulnerable to manipulation in thin markets and to herd behavior in thick ones.
Can I trade the 2026 midterms on a prediction market?
Yes. Kalshi and Polymarket both list contracts on individual Senate and House races, party control of each chamber, and national vote share. Kalshi has more US-listed contracts; Polymarket has higher overall volume.
How does the implied probability work?
Each contract pays $1 if YES and $0 if NO. A contract trading at 62¢ means the market's implied probability of YES is roughly 62% (minus a small spread). Sum of YES and NO prices is approximately $1.
What's the difference between Kalshi and Polymarket for elections?
Kalshi is CFTC-regulated and accepts US users; Polymarket has deeper election volume and more granular contracts but blocks US IPs. For active US traders, Kalshi is the practical choice.
Independent coverage. Some outbound links are affiliate links — see footer disclosure.