RONDA CHURCHILL/AFP via Getty Images. NATALIE BEHRING/AFP via Getty Images.
When trying to figure out the outcome of an election, most look at polling, expert takes, or even the stock market.
However, prediction markets, which allow bettors to wager on the outcome of events, may be the most accurate way of calling the result ahead of November 5. They've successfully forecast the last few presidential cycles and even predicted President Joe Biden would drop out weeks in advance.
While they're not without their misses — they were notably wrong during the 2022 midterm elections, and after Biden won the 2020 election — Rutgers University statistics professor Harry Crane says prediction markets are historically more accurate than traditional polling.
In the past month, a series of election-related events—an attempted assassination against Trump, President Joe Biden stepping down and endorsing Kamala Harris, and both candidates picking their running mates—have adjusted polling considerably.
Yet Crane says these adjustments are often reflected in prediction markets more quickly.
Polymarket, a crypto-based betting market, predicted a 70% chance President Joe Biden would drop out of the race this year as far back as July 4, two weeks before he actually bowed out.
"It's an event-driven market. So the past month has been very active because events have occurred that are of extreme significance," Crane said.
And the markets will likely get even quicker at factoring in new information on election night in November.
As vote totals, exit polls, and absentee ballots come in, prediction markets "move pretty significantly and pretty quickly in response to that information," he said.
There's a range of participants in betting markets, and most are making highly informed wagers, Crane said.
There are recreational bettors who may just have an interest in whatever topic is up for debate, as well as professional bettors and betting syndicates whose livelihoods are at stake.
There are also companies looking to hedge risk if they may be affected by candidates' proposed policies.
With $606.5 million riding on the election outcome on Polymarket alone, participants watch the news closely, Crane said. Even if they aren't based in the US, they are just as clued into election updates since their money is on the line.
He likens it to sports betting: "I don't need to be a professional football player to know how to understand what's going to happen in a football game. And in fact, I could be better at it than players in the game from a betting or informational perspective," he said.
Crane says that prediction markets could become even more accurate— and avoid any potential risk of manipulation by political groups —if they are fully legalized in the US.
The most popular prediction markets are largely unavailable to US participants. US-based users can't bet on Polymarket, and Betfair, one of the longest-running events markets, is only available in some US states.
PredictIt, a project founded by researchers in New Zealand, has operated in the US under a no-action letter from the Commodity Futures Trading Commission since 2014, but limits American bettors to up to $850 in wagers, and only 5,000 betters can participate in each market.
PredictIt is facing headwinds from the CFTC, which is looking to outlaw all election betting. If passed, Americans wouldn't be legally allowed to participate in any prediction markets, though some may try using VPNs to bypass regulations, which likely occurs now.
Crane also said more participation would make manipulation less likely.
While there's no hard evidence, he guesses that political bets may be especially prone to attempts at manipulation or influence.
"Everything kind of feeds off of everything else, and by trying to limit one aspect of it, or trying to police one aspect of it, it can often ruin the entire market function," Crane said.
"I think that no limitations is actually the way to go if we want to have better, cleaner, more accurate markets."
