FOR IMMEDIATE RELEASE: October 29, 2026

Contact: Michael Clauw, mclauw@campaignforaccountability.org, 202.780.5750

WASHINGTON, D.C. – Kalshi’s election odds aren’t as reliable as the company claims. A new Campaign for Accountability (CfA) analysis found that underdogs in “competitive” 2026 U.S. primary elections won 30% more often than Kalshi forecast. These markets also grew less accurate in the final week before election day as bettors piled onto perceived favorites.

As election conspiracists point to prediction market misses as supposed evidence to back up vote-rigging claims—and 38% of voters say such a mismatch would make them question official results—CfA’s analysis suggests election odds on prediction markets are not always as accurate as their proponents claim.

Read CfA’s report. 

CfA executive director Michelle Kuppersmith said, “Prediction markets sell their election odds as truth machines immune to bias, but the collective guesses of online gamblers are not facts.”

CfA reviewed day-before-election Kalshi odds for 645 primary contests in all 50 states and Washington, D.C. In “competitive” races, defined as races where the frontrunner’s odds fell between 50% and 90%, Kalshi was, on average, 7.3 percentage points overconfident in the perceived favorite’s chances of winning. Across those 238 “competitive” forecasts, Kalshi markets predicted 57 upsets, but there were 74 (30% more).

The report also examined candidates one week before their election to see how frontrunner odds shifted in the closing days. At the one-week-out point, these “competitive” markets were nearly perfectly calibrated, but confidence in the frontrunners kept climbing through the final week, ending about six points higher than the share that actually won. In short, Kalshi forecasts for these races were less accurate on election eve than a week earlier.

CfA compared these findings with Kalshi’s own data, including an August 2026 working paper that appears to confirm CfA’s finding showing election markets fail to improve in the final week.  The report also discusses a June 2026 analysis by The Washington Post that shows certain subsets of Kalshi’s 2026 primary markets with near-perfect calibration. Kalshi promoted a single data point from the Post’s analysis as outside proof of its markets’ accuracy, but failed to mention other parts of the analysis that suggest less accurate forecasts. CfA’s report also weighs the actionability of prediction market odds against information provided by conventional election experts and modelers—which generate similar information without social costs.

After the Los Angeles mayoral primary in June, paid prediction market influencers suggested Spencer Pratt’s falling odds on prediction platforms were evidence of election fraud. Election officials around the country have called the resulting disinformation “destructive,” and a challenge “to the confidence of our election system.”

As gambling addiction rates rise, addiction experts warn that prediction markets, just like other forms of gambling, can lead to addiction. The prediction market boom allows online betting in states like California and Texas that have not legalized it and effectively lowered the gambling age from 21 to 18, regardless of state age limits. Kalshi is also courting women with pop culture-tinged influencer campaigns and boasts that the percentage of female traders on the platform recently doubled from 13% to 26%.

For a few, the prediction market boom has been lucrative. Kalshi is valued at $40 billion and Polymarket at $21 billion, turning their founders into billionaires. As the Trump administration backs off regulation, Donald Trump Jr. has advised both Kalshi and Polymarket, and his firm, 1789 Capital, has reportedly invested hundreds of millions of dollars in Polymarket.

The companies’ founders have donated at least $5 million to political groups and politicians so far this election cycle, and have hired dozens of lobbyists to preserve favorable regulatory conditions. Meanwhile, a Wall Street Journal analysis found that over 70% of Polymarket accounts lose money, and 67% of profits go to just 0.1% of accounts.

“Executives, lobbyists, investors, media partners and celebrity spokespeople all have a stake in keeping election betting alive,” Ms. Kuppersmith added. “For everyone else, these markets aren’t worth the harm.”

Campaign for Accountability is a nonpartisan, nonprofit watchdog organization that uses research, litigation, and aggressive communications to expose misconduct and malfeasance in public life and hold those who act at the expense of the public good accountable for their actions.