Three billion dollars is a lot of money to wager on whether your neighbors will be allowed to choose their government. Now we know what it bought: nothing we didn't already have.
The pitch for Kalshi, Polymarket and their cousins is that prediction markets are a higher form of knowledge. Polls are noisy, pundits are biased, and markets, where people put real money behind their beliefs, supposedly cut through it. The midterms were the test, and the product's own defenders have now graded it. Nate Silver, who advises Polymarket, says the market prices mostly drifted toward conventional forecast models rather than anticipating them. Economist Koleman Strumpf says the markets did well but did not beat the best traditional forecasts. That is the sound of a sales pitch deflating. The crowd was not wise. It was reading the same polls as everyone else and putting a dollar sign on them.
Look at where the markets were wrong. Favorites a week out won 95% of races, which is roughly what you'd expect in a country where most races are not close. Of the 23 favorites who lost, every one lost to a Democrat. Traders gave Gina Hinojosa a 33% chance against Greg Abbott, and she won by more than 700,000 votes. That is not a rounding error. I can't tell you why the errors ran one way, and Kalshi insists it underestimated Democrats less than rival models did. But a system that claims to aggregate truth and misses in a single partisan direction deserves more suspicion than applause. Who is trading, what they are reading, and what they would like to be true all shape a market's price. Money doesn't cleanse bias. It concentrates the bias of whoever has money to spare.
Then there is the question the industry would rather skip: what is the point? A forecast that merely tracks the polls is not a public service. It is a casino table with a civic costume. The people who profit from it are the platforms, whose business depends on keeping the trading volume high, and anyone with an edge, whether better data, better connections or inside knowledge. The people who lose are ordinary bettors who were told they were participating in democracy's wisdom rather than feeding a house.
It matters that these companies are not outsiders to power. Donald Trump Jr. is an adviser to both, and the Trump-era regulatory climate has been friendly to prediction markets' expansion while states fight in court to police them. Fine print about oversight is not the point. The point is that a financial product tied to the outcome of elections is now owned and advised by people with a direct stake in those outcomes. We wouldn't let a casino operator sit on a gaming commission. We should not be relaxed about this.
And consider the culture it builds. When every race has a ticker, a campaign becomes a trade. A contest over healthcare, wages and who gets to vote becomes a probability to hedge. Volume more than tripled from late October to election week, which tells you what the product really sells: the thrill of the final stretch, not accuracy. Voters deciding whether to stand in line become spectators watching a scoreboard, and any cynic with a spare account has a new reason to spread rumors, suppress turnout news or leak early information. Billions in incentives attach to an event that should be sacred precisely because money doesn't get a vote.
Democrats shouldn't take comfort in the markets underrating them this time. The lesson is not that the casino got it wrong. It is that the casino was never the point. Congress and state legislatures should draw a clear line: sports and weather may be fodder for speculation, but contracts on elections, and on who holds power, should be off the table. A democracy can survive being wrong about a forecast. It is less sure to survive treating its own elections as a product.
How it may affect me
If you trade on these platforms, remember that election prices largely echo public polls, not hidden insight, and the platforms' business depends on trading volume whether any individual trader wins or loses. Treat the odds as one input at best, and expect state and federal rules on election contracts to keep shifting.
