Is Election Betting Bad for Democracy?

Politics nerds rejoice: You can now legally bet on elections. 

After a protracted legal battle, earlier this month the financial exchange Kalshi secured a court ruling that allowed it to accept American wagers on political outcomes. The decision is a major victory for “prediction markets,” online platforms that let users bet on all kinds of future outcomes. On Kalshi, for example, users can now gamble not just on whether Donald Trump will win over Kamala Harris in next week’s presidential election but on who will be nominated at the Oscars, what mortgage rates will be next quarter, and when George R. R. Martin will finally release the next Game of Thrones book. 

Even before the ruling, prediction markets had become big business, exploiting legal gray areas to operate. On Polymarket, a Kalshi competitor that runs on crypto, users have laid down the equivalent of over $2.6 billion on the race. Kalshi’s market, much smaller, still comes in at a cool $102 million. 

But strange things are happening in the markets. While Trump and Harris are polling neck and neck, and most election models have the race at close to a coin toss, the numbers on Polymarket recently showed Trump breaking well away from Harris, giving him a 66.7 percent chance to win. But is this the wisdom of the market, or something more nefarious? Trump’s lead has been driven by bets placed by just four big-spending accounts—who Polymarket says are all from the same person, a French trader. 

While it’s not obvious why someone is betting $46 million on Trump, the spike could change voting behavior—and potentially alter the outcome of the race. If a voter sees that one candidate is heavily favored by the market, they may change their vote, or just stay home. That’s particularly troubling because U.S. residents aren’t yet allowed to make political bets on Polymarket. 

Given this, many are reasonably wary of prediction markets. There’s a reason federal regulators tried to fight their growth, after all. But the benefits that prediction markets offer far outweigh worries that one whale is attempting to influence the election. 

I’ve written that legal sports gambling, and the rise of online gambling generally, were a socially harmful mistake. But I see prediction markets as fundamentally positive. Unlike a bet on the outcome of a roulette wheel or a horse race, prediction markets generate socially valuable information—a reason to be more supportive of prediction markets than other forms of gambling.  

Instead of expert opinions or complex models, prediction markets use something simpler to predict the future: profits. A market offers a contract on the outcome of an election, and buyers bid the price up and down. If players believe the contract on, say, the winner of the presidential election is too bullish for one candidate, people can make money by betting on the other, resetting the price. Academic research supports this approach—markets are as or more accurate than polls, since they force people to put their money where their mouth is.

Who will win elections is of enormous import to millions of people and businesses, affecting decisions about hiring and firing, about local government spending on schools and parks, and about where people choose to live and work. And it’s not just elections. Being able to know the likelihood of different outcomes in advance can help facilitate these decisions, making society run more smoothly for everyone.

Let’s say that you’re a business owner, planning to scale up. You could do your own research on coming interest and unemployment rates. Or you could check a prediction market, where people compete to make money on these topics—thereby aggregating all the information you need in one price. On some platforms, you can set up your own markets, and you could ask users to bet on how well your products will sell next quarter—in effect soliciting a market forecast which you can use to plan out your inventory more efficiently, or devise new offerings.  

Plus, betting on future political and cultural events isn’t new. There’s been organized betting on presidential elections since at least 1868. Tech companies have used markets to predict whether projects will stay on schedule, drugmakers to predict infectious disease activity, and the Department of Defense once controversially proposed to use them to help forecast terror attacks. Sites like Kalshi, Polymarket, and Manifold offer the same services, but open to the public, with the data generated by users’ bets easily accessible.  

In short, it’s good news that prediction markets finally have the freedom to run, but they need to start answering tough questions about their potential misuse, and soon. 

Most obviously, gambling is addictive. Some people will do it compulsively, even at substantial cost to themselves and others. It may be hard to imagine someone addicted to prediction markets. But politics, like sports, evokes tribalistic passions that make people do irrational things—like put millions behind Trump or Harris.  

The bigger risk, though, is in the way Americans perceive prediction markets. For better or worse, people are often wary of putting money on consequential events. That’s why the DoD was forced to put the kibosh on its experiment with using prediction markets to forecast terrorism—they were taken as making light of serious events.

Most alarming is that putting big money on elections can exacerbate preexisting tensions surrounding election legitimacy. Neither Trump nor Harris is going to throw the election to make a buck. As is apparently happening with Polymarket and Trump, a bettor with enough capital could temporarily move a market decisively toward his preferred candidate, in turn affecting voters’ perceptions of the odds—and changing their behavior. In other words, prediction markets can influence the very thing they’re predicting.  

But over the long run, distortions should be self-correcting. If someone has bet too much on Trump, then other bettors have the opportunity to buy up Harris at an artificially low price. This should bring prices back in line with reality. Indeed, the fact that other prediction markets are leaning toward Trump suggests the spike is not entirely down to manipulation. The consensus reflects the narrowness of the swing state vote, the GOP’s electoral college advantage, and the possibility of yet another big polling error in Trump’s favor.

But even if the risks of manipulation are overstated, it’s important that prediction markets hold themselves to a high standard. It’s illegal to manipulate the stock market; prediction markets should try to enforce the same rules. Operators should be willing to pause trading or ban traders when they believe someone is trying to deliberately skew an important market. They should do this precisely because they want prediction markets to succeed. After all, wouldn’t we all like to know how our vote, or our company, or our relationship will play out?  

Unlike other forms of gambling, prediction markets have the potential for huge social value. But that doesn’t mean they have social trust—that has to be earned.

Charles Fain Lehman is a fellow at the Manhattan Institute and a contributing editor of City Journal. Follow him on X @CharlesFLehman.

To support more of our work, become a Free Press subscriber today:

Subscribe now