Americans can now turn their predictions about which party will control Congress into a wager, but most believe they shouldn't be able to. As the midterm elections approach, 60% of U.S. adults oppose using prediction markets to bet on election results, while 13% support the practice and 24% are unsure, according to a Reuters/Ipsos poll provided to Fortune.

Many viewers will encounter the odds anyway. CBS News announced on Sept. 30 that it would incorporate real-time prices from Kalshi into its midterm coverage, displayed alongside its own polling. Kalshi has struck similar arrangements with CNN and CNBC, and it licenses vote counts and race calls from the Associated Press. CBS News and CNN now share a common owner: Skydance, which completed its acquisition of CNN's parent company, Warner Bros. Discovery, on Oct. 6.

Thin Trading Behind the Broadcast Numbers

The prices shown on national television can rest on a remarkably small pool of participants. On Polymarket's global platform, the top 1% of wallets — the individual accounts used to place trades — account for 68% of trading volume on 2026 congressional markets, and 80% of those markets have fewer than 100 wallets trading, according to an August analysis by the Anti-Corruption Data Collective, a nonprofit research and advocacy group. A price can read like broad consensus while being set by a relatively small group of participants.

Public opposition crosses party lines. Among Democrats, 69% oppose election betting, as do 61% of Republicans. Three-quarters of Americans consider prediction-market wagering a form of gambling, and only 7% of respondents said they had ever used a prediction market to bet money.

"Most view prediction markets as a form of gambling, and it's clear that they do not see elections or government decision-making as appropriate subjects for it," Alec Tyson, a senior vice president and lead pollster at Ipsos U.S. Public Affairs, told Fortune.

Ipsos surveyed 4,506 U.S. adults from Sept. 30 through Oct. 5. The margin of sampling error is plus or minus 1.5 percentage points.

Volume Tells a Different Story

Monthly trading volume and public approval, however, measure different things. Dustin Gouker, a gambling industry analyst and author of the prediction market newsletter The Event Horizon, said sports offer what elections struggle to match: frequent events and fast-changing odds that give people more reasons to trade.

A poll release or campaign development can move an election market, but a single sports game can shift odds within minutes as teams score and leads change hands.

"I don't think it's necessarily that election trading isn't popular," Gouker told Fortune. "It's more that it doesn't feature some of the things that we see in these markets that are seeing more volume."

Combined monthly trading volume on Kalshi and Polymarket more than doubled from roughly $26 billion in May to $53 billion in July, according to a September Pew Research Center analysis, with sports driving most of the increase. On these platforms, users buy and sell contracts on whether an event will occur; winning contracts pay $1 and losing contracts pay nothing.

Those headline totals count all trading and value contracts at their potential $1 payout rather than their purchase price. They do not represent company revenue or the actual cash traders put at stake.

Kalshi Challenges the Findings

Kalshi dismissed the poll as biased, objecting to how it described prediction markets and framed its questions about wagering on government events.

"It's not hard to bias a poll and this poll is biased," a Kalshi spokesperson said in a statement to Fortune. "Ask leading questions about whether people support 'betting on war,' and you'll get the answers this poll arrived at. Kalshi doesn't even war markets."

The questionnaire, however, never used the word "war." It described prediction markets as allowing people to "wager money on the outcome of real-world events," then asked separately about elections, military actions, government policy decisions, and sports. It did not measure familiarity with the platforms. A March Ipsos poll conducted for the American Institute for Boys and Men found that just 21% of Americans were very or somewhat familiar with prediction markets.

"We welcome research on our industry, but a clearer picture would come from research that measures familiarity, uses neutral wording, and reports all responses," Kalshi said.

Ipsos did not respond to Kalshi's criticism by publication time.

The exchange is not Kalshi's first controversy this year. Traders sued the company in March over its handling of a market on whether Iran's supreme leader, Ali Khamenei, would leave office, a dispute that followed his killing in U.S.-Israeli strikes. Kalshi told Reuters at the time that the market's rules excluded death-related outcomes and that it had reimbursed trading fees and net losses.

Odds Aren't Race Calls

Polymarket argues that its audience extends well beyond bettors. More than 90% of visits to the platform are to view markets rather than trade them, a spokesperson told Fortune. "Prediction markets are a way to get unbiased, accurate information about the events that impact our lives," the spokesperson said.

Gouker said election markets draw viewers by distilling polls and news into a single probability, but converting those visitors into traders is a separate business challenge. He expects election trading activity to build as the midterms approach, though he doubts it will approach the scale of sports betting. He also expects the 2028 presidential election to attract more trading than this year's midterms. Either way, the network deals will put the odds in front of a national television audience that includes the 93% of American adults who say they have never put money on one.

He cautioned against treating high odds as a call of the race. Gouker pointed to Abdul El-Sayed's Michigan primary, noting that markets showed him as an overwhelming favorite around the close of voting before incoming results suggested a far closer contest. El-Sayed led by just under one percentage point when the Associated Press called the race.

A candidate with a 70% chance of winning, after all, still has a 30% chance of losing — a gap between a heavy favorite and a certain winner that viewers watching odds appear beside live coverage this fall will need to keep in mind. To judge a market's accuracy, Gouker said, examine whether outcomes occur as often as the odds imply: "When a market says something should happen, it happens roughly the amount of times we would expect it to."

He added, "You have to look at them as probabilities and not as these crystal balls."

This story was originally featured on Fortune.com.