When the Market Watches the Court

Introduction

In November of the 2025 term, the Supreme Court heard argument in a case set to reshape the national economy. Learning Resources, Inc. v. Trump (U.S. 2026) asked whether the President may impose sweeping, cross-industry tariffs under the International Emergency Economic Powers Act, which authorizes the President to “regulate . . . importation.”

As argument wound down, court-watchers and journalists rushed to forecast the outcome. The New York Times sensed skepticism from the Court’s conservative wing. The Washington Post concurred. Other commentators hesitated, stressing that a victory for the administration was still very much in play.

This is the traditional legal commentariat at work: experts and practitioners opining on the goings-on of the Supreme Court. But in Learning Resources, this commentariat was not alone. Watching alongside the legal world, thousands of people expressed their opinions with their wallets, wagering over $10 million on the outcome of the case.

These wagers took place on prediction markets, where users trade contracts tied to the likelihood of a future event. Similar markets have become commonplace in sports and election forecasting, where consumers can bet on sprawling gray-market sites or trade small-dollar shares on user-friendly platforms like PredictIt. Now, they are turning their attention to the law.

This Essay explores the future of legal prediction markets. Part I explains how markets work and what makes them hard to beat. Part II then turns to the largest legal prediction market to date: the outcome of Learning Resources. Finally, Part III considers whether markets are well-suited to forecasting legal outcomes, both in principle and in practice.

I.  What Are Prediction Markets?

Prediction markets are platforms where users can buy and sell contracts. Each contract pays depending on the resolution of a future event. For example, a contract might pay \$1 if a Republican candidate wins the 2028 presidential election and \$0 otherwise. The contract’s price reflects the implied probability of the event. If the 2028 contract trades at 40¢, the market projects a 40% chance of Republican victory.

From there, the mechanics are straightforward. Users trade contracts based on their own estimates of an event’s likelihood. All else equal, a trader buys when the contract price is below her subjective probability. A trader who believes a Republican win is a 50-50 call, for instance, would snap up any contract at 49¢ or less. If her estimate is correct, she profits in expectation. As millions of contracts trade back and forth, the price moves to reflect the market’s aggregated prediction.

A.  Beating Prediction Markets

Prediction markets “produce remarkably accurate predictions.” Each trader is under financial pressure to buy or sell carefully. If she is not confident in either side of a contract, she has no incentive to participate. The result is a market that drives out uninformed traders and rewards experts with insider knowledge. As economist Alex Tabarrok once quipped, “A bet is a tax on bull****.”

Beating a prediction market is no easy task. Economics makes consistent outperformance almost impossible. Every trader sees the same price and has the same opportunity to buy. And any delta between the posted price and true probability is an invitation to close the gap. In other words, if a contract seems obviously mispriced, think again: Odds are the market already incorporated everything you know, plus more that you don’t.

That is not to say markets are infallible. So-called “whales” can dominate markets with thin liquidity, moving prices in ways that don’t reflect consensus. Manipulators might distort the price by betting to drive narratives rather than to profit. And regulatory barriers or other constraints on access can thin the pool of informed participants. Markets are also dependent on clear resolution criteria, which can pose problems for sensitive predictions. Every contract must resolve to a “yes” or “no,” with no room for finesse. Disagreements about resolution often turn fractious, and prolonged disputes can even undermine a platform’s credibility.

But those limitations are the exception, not the rule. Well-designed markets excel at turning private information into public probabilities. Whales and manipulators create opportunities to profit by correcting a distorted price. And access barriers are disappearing as platforms like Polymarket and Kalshi explode onto the mainstream.

B.  Prediction Markets About the Law

The idea of a market that predicts Supreme Court decisions is not new. In 2006, Professor Miriam Cherry and Robert Rogers proposed “Tiresias,” a market that would allow users to do exactly that. They bemoaned the state of legal prognosticating, where scattered “newspaper stories and op-eds” batted around predictions with little accountability. A market, they argued, could aggregate individualized information to create a more robust forecast of how the Court would rule. The payoff was obvious. The “ability to know a probable Supreme Court outcome in advance” would be of great interest to practitioners, if not the Court itself.

Three years later, Professor Josh Blackman launched FantasySCOTUS, an online league that invited users to register predictions for upcoming Supreme Court cases. Users predicted votes down to the justice and received points for accuracy. During the Court’s 2009 term, participants made over 11,000 predictions for 81 cases. Collectively, FantasySCOTUS outperformed chance, and a high-volume group of “power predictors” correctly predicted almost 65% of cases. Subsequent work fed FantasySCOTUS predictions into a machine learning model, which achieved “a peak performance of over 80% accuracy.”

These projects demonstrate the power of legal crowdsourcing. But they are just a demonstration. The Tiresias market was never created. And FantasySCOTUS “is not a traditional prediction market.” It is closer to a “prediction aggregation mechanism,” with a free-to-play interface and no exchange between buyers and sellers. In any event, after eclipsing 20,000 users in 2014, FantasySCOTUS appears considerably less active today.

II.  Predicting Learning Resources

Now familiar with the mechanics of crowd-sourced predictions, turn back to Learning Resources. Described by many as the blockbuster case of the term, Learning Resources commanded significant attention in the court-watching world. It was also watched by the market.

Polymarket and Kalshi—the two largest prediction markets by volume—each opened a market for Learning Resources on September 2, 2025. The markets asked a simple question: “Will the Supreme Court rule in favor of Trump’s tariffs?” Beneath the headline, each platform specified resolution criteria for what “rule in favor” would actually mean. And just like that, traders went to work, sinking a cool $10 million into their predictions about a thorny question of statutory interpretation and constitutional law.

The market opened with relatively stable odds. On Polymarket, the government’s chances of victory fluctuated between roughly 40% and 50% from September 2 until the end of October. Kalshi reported similar odds, dipping into the mid- and high-30s. Notably, the market proved largely insensitive to briefing. There were no significant oscillations on September 19 or October 20, the respective filing dates for the government and challengers’ briefs.

Then came oral argument. On the morning of November 5, the government’s odds stood at roughly 40%. As Solicitor General John Sauer began to speak, its odds began to plummet. Before the challengers’ advocate could even make his case, the probability of a pro-Trump ruling had “slid[ ] to a fresh low of 18%.” After all was said and done, Polymarket and Kalshi each settled around 25%.

Media outlets seized on the tumble. The Daily Beast noted a “[t]errible [s]tart” for the administration, linking tumbling odds to dissatisfaction with Sauer’s arguments. Another analysis suggested Sauer “had a rough day at the office.”

These critiques may well be right. But they do not follow inexorably from the market’s movement. That movement only reveals the new price at which trades clear. It does not reveal why prices shifted the way they did. Observers can draw inferences, but they are working with imperfect and limited information—which is why they so rarely outperform the market in the first place. Perhaps the government’s odds suffered because the challengers outperformed General Sauer. Or perhaps argument merely revealed what the justices already thought.

Markets can tell us, though, that the commentariat’s consensus looks to be well-calibrated. The State’s odds of prevailing stabilized at around 25% after argument. As the Court held off on its opinion through the first few weeks of 2026, traders updated their estimation of a pro-tariffs ruling, reaching the mid-30s before dropping back down.

That estimation proved sound. In February 2026, the Supreme Court decided Learning Resources, holding that the International Emergency Economic Powers Act does not authorize the President to impose tariffs. Legal prognosticators declared victory. The commentariat rolled on. And, this time, so did the market.

Will Trump defy the Court by imposing fresh tariffs? Will he refund businesses for unlawful ones? Will the Court hear another case on tariffs? These are tough questions, inspiring debate across analysts and major outlets. They are also inspiring markets. There are now dozens of legal prediction markets, running the gamut from Tariffs 2.0 to new blockbusters and judicial appointments.

III.  Law in the Year of Prediction Markets

Prediction markets are on the horizon. Polymarket and Kalshi skyrocketed to prominence after outperforming experts in the 2024 election. Today, markets exist for “everything,” from concert cancellations to Congressional legislation. Learning Resources might be the biggest Supreme Court market as of March 2026, but it will not stay the biggest for long. Lawyers and court-watchers would do well to familiarize themselves with this latest method of legal forecasting.

A. Predicting the Law

Is the law suited to prediction markets? On the one hand, legal ecosystems are information rich. Traders can read statutes and opinions, follow along with briefing, and listen to oral argument. They can compare the present case to analogous predecessors or research political science models that forecast decisions by ideology, all of which are publicly accessible.

Cases are consequential, too. Blockbuster Supreme Court opinions can reshape whole sectors of the economy, unsettle regulatory regimes, or tilt the playing field for the next election. That makes their outcomes salient to a wide range of actors, from lawyers to businesses to the general public—all of whom have reasons to form and act on predictions. When those predictions are channeled into a market, the result is a relatively thick, information-rich exchange.

On the other hand, legal questions are ill-suited for clear resolution criteria. Many questions presented are not binary propositions but bundles of procedural and substantive thickets, each of which might come out differently. A Supreme Court decision can splinter into a plurality. A pending case can be dismissed as improvidently granted, mooted, or remanded on procedural grounds. Doctrinally, those outcomes are very different. But an event contract must resolve to “yes” or “no” come payout day. Drafters are forced to navigate this doctrinal maze, writing resolution criteria that are easy enough to parse without butchering the law.

B. Legal Status of Prediction Markets

Regulators have begun to turn their attention to prediction markets. In 2022, the Commodities Futures Trading Commission (CFTC) brought an enforcement action against Polymarket, finding that many of its event contracts were “swaps” under the Commodity Exchange Act.

“Swaps” are transactions that transfer financial risk between participants. Because wagers on Polymarket look a lot like binary contract options—all-or-nothing payouts that are often classified as swaps—the CFTC accused it of operating without registration. Polymarket was ordered to pay $1.4 million and wind down certain U.S. markets.

Today, Polymarket has won approval to offer contracts to Americans through brokerage firms. And Kalshi, its primary rival, operates legally as a federally regulated exchange. It is not clear how the law of prediction markets will develop. But chasing money with regulation has always been a losing strategy. So far, new markets have managed to “outpace[ ] the regulatory response.”

Conclusion

Confidence is cheap. It costs nothing to announce that a judicial result was obvious all along. It is even cheaper to dress up that confidence as insight, weaving a few choice quotations into a narrative about why the Court ruled the way it did. Prediction markets are different. Instead of hiding behind vague, open-ended predictions, traders put their wallets on the line. In that environment, it pays to be honest about uncertainty.

Learning Resources is case in point. As analysts rushed to forecast the outcome, thousands of traders quietly registered their beliefs. In a matter of minutes, the market moved from a contested coin flip to a long shot. That price movement did not decide the case. But it did provide something that cannot be found in the legal commentariat: a real-time, probabilistic assessment of how the Supreme Court was likely to rule.

Of course, law is not a casino. Trading contracts on Supreme Court cases will not replace learning the doctrine or reading an opinion with care. But when thousands of interested parties are groping through the same fog, court-watchers could do worse than paying attention to the numbers.