Betting on election results is now legal in the US and is as easy as betting on a sports team to win. But experts caution against mistaking betting markets for polls.
A handful of companies, including Kalshi and Interactive Brokers, allow users to bet on Donald Trump or Kamala Harris to win the election. Days before the election, even stock trading platform Robinhood joined the party by offering its own betting market. Companies set their own betting limits, Kalshi allows users to bet up to $100 million.
“We believe event contracts give people a means to participate in real-time decision-making, opening up a new asset class that democratizes access to events as they happen,” Robinhood said in a statement on Monday.
Election betting was legalized a few weeks ago as the 2024 race heads into its home sprint. The Commodity Futures Trading Commission, which regulates financial contracts betting on events, has consistently blocked applications to open election betting under both Republican and Democratic administrations.
But that changed last month when a U.S. district court sided with Kalshi in the company’s lawsuit against the CFTC, arguing that it should be allowed to offer the contracts. The decision cleared the way for Kalshi to open election betting markets while the CFTC appealed. The ongoing legal dispute could still shut down US election betting markets, but after the 2024 vote.
Meanwhile, “it puts us, the CFTC, in charge of being the election police,” said Chairman Rostin Behnam he said last weekarguing that the agency will respect the court order when election bettors place their bets. Since then, millions of dollars have flooded the markets. Kalshi said bets placed on the general election reached $30 million in the first three weeks and rose to $139 million on Friday.
Abroad, where betting on American elections is nothing new, billions of dollars were in circulation.
Crypto based betting platform is Polymarket It hosts more than $2 billion in bets, probably from foreign traders. Users are technically banned in the US, where the CFTC slapped the company $1.4 million in fines In January 2022, it issued a cease and desist order effectively banning it from the domestic market for treating election betting as an unregistered platform. But with private, encrypted internet networks readily available, industry watchers say Americans can still access foreign platforms.
PredictIt, a project of New Zealand’s Victoria University of Wellington, offers unique contracts. The platform is available for Americans, but the betting limit is $850. Having “skin in the game” forces users to remove “spin and emotion” from their expectations of election results, says founder John Aristotle Phillips.
But Phillips says the public should interpret prediction markets carefully when it comes to politics.
“The market will give you the probability of an event happening,” he said. “It has nothing to do with and it doesn’t give you an idea of what the margin of victory or loss will be, in terms of what the poll is supposed to show.”
In many of these markets, bets are settled by the dollar wagered. So if a general election bet says 52 cents for a candidate to win, then a successful 52 cents bet on that candidate will return $1 – a payout of 48 cents.
But if the scores show one candidate 52 cents to 48 cents over the other, that doesn’t mean the former will win 52% of the popular or electoral vote.
Polls ask prospective voters who they will vote for, so if 52% of voters tell the pollster they support one candidate and 48% support the other candidate, there can be a difference of 4 percentage points between that group of voters.
NBC News does not refer to betting odds as a proxy for political voting. Betting and gambling platforms do not use the methodologies used in traditional political polling and therefore do not replace political polls.
Nevertheless, public debate around betting market odds has blurred the lines, with prominent voices sometimes confusing the two.
In early October, the multi-billionaire Trump donor Elon Musk cited Polymarket speculates and published on X that the former president “leads” Harris by 3%.
But Polymarket only found that bettors were 3 percentage points more likely to win Trump than the opposite outcome, not that Trump had a 3-point lead over Harris among actual voters. Musk claimed that the Polymarket coefficients are “more accurate than polls”, despite the fact that the users of the Platform are completely foreigners who cannot vote in US elections.
Polymarket also attracted attention after four accounts placed large bets at the end of October. as This was first reported by The Wall Street Journalbettors appeared to increase the odds for a Trump victory. Polymarket then said that all these bets made by a single French citizen. The company did not respond to a request for comment.
The CFTC only allows US residents to use the newly opened election betting markets, meaning many bettors on the platforms are also potential voters.
Kalshi allows bettors to bet up to $100 million on the outcome of the general election, but the company says any attempt to scale back by placing large bets should be washed away by the market.
“If the truth is actually 50-50 and the market is priced 60-40, there’s a huge incentive, many tens of millions of dollars, to bring prices back. This happens in these markets as well,” said Tarek Mansur, the founder of Kalshi.
He added that unlike Polymarket, Kalshi is required to comply with laws such as Know Your Customer, so users must provide personal information to prevent potential financial crimes. “We know if they are bad actors and we enforce the rules,” Mansoor said.
Still, Adam Cochran, founder of venture capital fund Cinneamhain Ventures, says he’s concerned about how voters perceive election betting odds. Prediction markets should simply be “a data point of how we rate the election,” he said, “just a slice” of where the momentum is going in the campaign.
“Politics in America, and increasingly around the world, is a lot more like rooting for your favorite sports team,” Cochran said.
Mansour acknowledged that it was a “fair concern” that the general public might mistakenly see odds as a measure of a candidate’s advantage.
“This is different from a survey. “There is a little misconception,” he said.