In late July, as Americans were preparing for another round of high-stakes midterm primaries, top elections officials were on a webinar, warning about the dangers of prediction markets—specifically how both Kalshi and Polymarket are ripe for manipulation by bad actors looking to influence public opinion to turn a profit.
“Elections administration right now is very much affected by perception and misinformation,” Los Angeles County clerk Dean Logan said, “and, in our field, we have to be prepared to respond to that.”
Median Strategies, in a statement posted to its website, denied that the “experiment” was undertaken for financial gain or to influence prediction markets. It still does not bode well for future elections at a time when there are already efforts underway to undermine public trust in the democratic process.
Gallup recently found that 67% of Americans are worried election officials will be pressured to change the outcomes of races and another 57% fear their ballots will not be counted properly—perhaps unsurprisingly given President Donald Trump’s frequent lies about “stolen” elections and mail-in ballots.
Adding fake polls to the mix certainly will not help. The impact of Median Strategies’ two initial “polls” was arguably modest.
The first, which purported to look at Nevada’s high-profile races for governor and lieutenant governor, was picked up by an aggregator account on X and then shared widely. The second, supposedly examining Wisconsin’s contentious Democratic gubernatorial primary, appeared during the final days of the race, but was mostly ignored in the Midwestern state, according to the Milwaukee Journal Sentinel.
It was Median Strategies’ third “poll,” about the Los Angeles mayoral race, in which incumbent Democrat Karen Bass is facing democratic socialist City Council member Nithya Raman, that had the biggest impact.
One reason is that there has been very little polling so far in what is expected to be a tight, general election contest to lead the nation’s second largest city. Therefore, the purported findings, which were released last week, seemed revelatory.
Median Strategies claimed Bass was in the lead, picking up voters who had supported Republican Spencer Pratt, of MTV’s The Hills fame, in the June primary election. “Faced with a choice between the incumbent and a Democratic Socialists of America-aligned challenger, most Pratt voters appear to be treating Bass as the more familiar, more moderate option rather than sitting out the race,” the company wrote in its ‘analysis.’
The made-up findings spread quickly. Political strategists posted about them on social media. So did the Bass campaign, writing on X in a since-deleted post: “Doing the work, showing up, and gaining momentum. Let’s do this, LA!” The California Post—the right-leaning, relatively new outpost of the New York Post—quickly published a story about Bass supposedly getting Pratt’s voters.
Once Median Strategies fessed up, pressed by a reporter at the Los Angeles Times, even the backtracking became an exercise in scoring political points.
Bass’ campaign defended getting duped, blaming “multiple news outlets” for reporting on the fake figures, and declaring that “any bad faith attempts to influence elections should be investigated and prosecuted to the fullest extent of the law.”
Raman’s campaign blamed the mayor for promoting the “poll” without proper vetting. “It is the latest example of an administration that too often struggles to get the basics right,” a spokesperson told KTLA.
This is the stuff of nightmares for Logan and other election officials. If one company can get away with this, others are bound to try. After all, there’s tremendous incentive to do so, given the potential to make money in prediction markets.
A recent NBC News analysis found that users of Kalshi and Polymarket have already racked up roughly $200 million in trades on the outcomes of midterm election races this year.
In its statement, Median Strategies claimed that no one involved benefited financially from its fake polls. However, the identity of those behind the company remain a mystery, and Bass’ odds of winning the mayoral race in November did improve after the phony polling data was released.
Meanwhile, traditional polling is going through its own crisis of confidence. The past decade is littered with failures. More recently, progressive Democrat Abdul El-Sayed had been predicted to win this month’s Senate primary in Michigan by double digits. The same was true for democratic socialist Francesca Hong in Wisconsin’s gubernatorial primary. But El-Sayed ended up winning narrowly and Hong lost.
Indeed, prediction markets are booming at least in part because of growing distrust in traditional polling. Kalshi and Polymarket insist what they do is more accurate because their odds are based on the “wisdom of crowds.”
At the same time, a poll—humour me—commissioned by the non-partisan Partnership for Large Election Jurisdictions, which represents election administrators, found that 39% of likely midterm voters said they would have less confidence in the outcome of an election if the official result differed from the odds in a prediction market.
A majority of those surveyed also could not correctly identify what the odds even represent, with many believing they illustrate the number of votes cast for each candidate.
“Prediction markets are a form of speculation, but they are increasingly being presented, and in some cases interpreted, as indicators of likely election outcomes,” Logan recently told Wired. “The challenge is that many people don't distinguish between a market reflecting the views of participants and the official election process.”
Caught in the act, Median Strategies says it has shut down its operation. I won’t make any bets about how prepared Americans will be when the next mysterious polling company pops up. ©Bloomberg
The author is a politics and policy columnist for Bloomberg Opinion.