Billions of dollars were spent this election season trying to shape what Americans believed.
Political advertising in the 2026 midterms topped $12.8 billion, making it the most expensive election cycle in U.S. history. In Texas, early estimates put spending on the Senate race at nearly $573 million, among the highest totals ever for a non-presidential contest.
For all the differences between the candidates and races, the familiar claims kept coming. Campaigns said they would win. Analysts stood by their numbers. Super PACs argued their candidate was best on both policy and character. And television coverage often seemed to cut from one accusation-filled ad to the next.
There is a huge industry devoted to nudging people toward the version of reality that benefits the politicians seeking office. Americans need ways to make sense of all that noise. It’s not surprising, then, that prediction markets are attracting attention.
Like any new technology, they’ve also drawn plenty of criticism. Senators have warned about election-related contracts, and some states have tried to sue or ban the markets. Meanwhile, somewhere, a podcaster is probably still explaining how the whole thing is rigged.
Election Day offered a useful test. Across the federal races being tracked, Kalshi’s markets had the eventual winner leading in 98% of contests. They reflected not only the likelihood of safe incumbents winning, but also the close odds in the major battlegrounds.
That result wasn’t necessarily a fluke. Kalshi reviewed 2,243,741 resolved markets, covering the period from its 2021 launch through the middle of 2026. Its analysis found that prices became well calibrated as events got closer. Put simply, an event priced at 30 cents happened roughly 30% of the time, while one priced at 90 cents occurred about 90% of the time.
Outside analyses have found similar patterns. Brier.fyi, an independent project that evaluates prediction-market accuracy, saw comparable results across nearly 74,000 Kalshi markets. A study published last month in the Proceedings of the National Academy of Sciences examined more than a million resolved forecasts on four platforms and found that events occurred about as often as their prices implied. A review of this year’s primaries also found that candidates won at roughly the rates predicted by their market prices.
Critics may point to an upset such as Democrat Annie Andrews defeating Republican Sen. Darline Graham in South Carolina and argue that prediction markets are useless. But Andrews was considered the underdog, with roughly a one-in-three chance going into Election Day. If underdogs never won, a one-in-three estimate wouldn’t make sense in the first place.
Prediction markets don’t need to be perfect to be useful. They can add something valuable to the polling, forecasting and reporting already available, and they may help improve that work.
Good data journalism takes time and money. As newsrooms have reduced staff and budgets, specialist data teams and deeper investigations have often been among the first to go. Prediction markets could give this kind of work a fresh incentive: when information has value, people have a reason to seek it out. That might mean more research, more detailed reporting and, yes, more work for data specialists. Reliable information is increasingly valuable, and markets reward people who find it, assess it and act on it.
Ask a profitable trader what they’re reading, and there’s a fair chance it’s a precinct-by-precinct breakdown of Hinds County, Mississippi, rather than the latest television series. I know the type.
Understanding the world isn’t just something to bring up at a dinner party anymore. Careful attention to the facts might also improve your finances—and help the rest of us stay better informed.
For people who feel ignored or shut out by mainstream media, prediction markets may be another useful development. They’re part of a longer effort to widen the range of voices and information available to the public.
For decades, a small number of networks and prominent anchors had an outsize influence over what counted as acceptable opinion. Talk radio challenged that arrangement, followed by cable news, blogs and podcasts. Those formats offered alternatives and sometimes brought attention to stories that might otherwise have been overlooked.
With billions of dollars still being spent to spin the truth, another way to distinguish evidence from persuasion could be welcome.
Prediction markets work by making people back up their views with money. A campaign insider can confidently claim a race is already decided without much personal risk. A trader has to put money behind that claim, and gets paid only if the prediction is right.
Even people who never trade may benefit. A market can compress a great deal of available information into one number, which offers a quick way to gauge what people collectively expect.
Depending on the candidates you supported, Tuesday’s results may have felt thrilling or deeply disappointing. Elections involve real consequences, and emotions can run high. That’s one reason it helps to have a way to step back and look at a contest as clearly as possible.
After more than a decade trading in these markets, I’ve noticed that successful traders tend not to invest with their hearts. I’ve come out ahead in several election cycles, including some when the results went against what I wanted. That’s the discipline involved: forecasting, not wishful thinking. We can hope for the world we want while investing based on the world we have.
If putting money on election outcomes seems odd, remember that politicians have been making money from the public for a long time. Perhaps it’s time for the public to get something back.
