Before it dropped out of the bidding for Warner Bros. Discovery, Netflix sought to portray the relevant market for assessing its proposed merger broadly, including both paid streaming services and advertiser-supported services such as YouTube and TikTok. Live Nation has adopted a similar strategy, attempting to depict a vast market for tickets to live events at any venue, whether at a major concert venue, stadium, smaller amphitheaters, or a large theater, to undermine the Justice Department’s monopolization claims.
There’s a reason that companies love to make expansive, sometimes absurd, claims about market definition: as the Supreme Court has narrowed the scope of per se illegality in antitrust law, plaintiffs are required to demonstrate that defendants possess market power. And, as I explain in the following post, one way to make this requirement more burdensome is to turn the trial into a costly, complex fight over market definition. Don’t look at the anticompetitive effects of the conduct, exclaims the defendant; focus instead on whether monopolization was theoretically possible given the assets under our control. Either plaintiffs lose the market-definition fight and lose the case outright, or they win the fight but burn through so much of the court’s goodwill that there is little left for the ultimate showing of anticompetitive effects.
Courts should deemphasize the role of market definition and focus instead on whether the challenged conduct injured workers, consumers, or some other counterparty. And if the caselaw prevents such a focus, Congress should provide fresh guidance in the form of antitrust reforms.
The Plaintiffs’ Heavy Burden
Antitrust cases are adjudicated pursuant to one of two evidentiary standards—the per se rule and the rule of reason. If the challenged conduct is considered a per se offense, such as a hotel-handshake price-fixing agreement among rivals, plaintiffs must establish that the conduct occurred (as well as a causal nexus between the conduct and some anticompetitive effect). Most conduct, however, is evaluated under the rule of reason, a much more exacting standard, which asks plaintiffs to demonstrate that a firm’s behavior generates more anticompetitive harm than procompetitive benefits. The Supreme Court, in line with 30 years of pro-business antitrust decisions, has been funneling more cases into this category, reclassifying formerly per se illegal conduct such as resale price maintenance into the rule-of-reason framework, as potentially motivated for procompetitive reasons.
Because plaintiffs often don’t know which standard will apply until after expert reports are due, their experts—your essayist included—must prepare a proof of market power as an expensive insurance policy. Ironically, while the courts recognize “raising rivals’ costs” as a legitimate economic theory of harm in antitrust cases, they remain blind to the cost-raising litigation strategies deployed while adjudicating these same cases.
Under the rule of reason, the plaintiff bears the burden of proving that the defendant possesses market power and has inflicted antitrust injury on a counterparty via some cognizable restraint. If that burden is met, the defendant is then permitted to offer efficiency justifications, at which point the plaintiff must rebut those defenses. This structure imposes a significantly heavier evidentiary burden on plaintiffs than defendants. Unsurprisingly, defendants often press for rule of reason, where disputes over market definition and market power can become dispositive long before the merits of competitive harm are meaningfully examined.
At the center of this costly and cumbersome evidentiary structure lies the concept of market power, which can be established in one of two ways. Direct evidence of market power entails showing that the defendant has raised prices above competitive levels (or suppressed wages below competitive levels) or excluded rivals. For example, in the market power section of my report in Le et al. v. Zuffa—an antitrust class action brought by professional mixed martial arts fighters against UFC’s parent company—I pointed the court to my impact model, which showed that as more ranked fighters were foreclosed by UFC’s exclusive contracts, the UFC was able to reduce its fighters’ pay. How could UFC have reduced wages via the exclusive contracts, I asked, unless it possessed market power? Indirect evidence, by comparison, requires proving a relevant antitrust market, and then showing that the defendant has a high share of the market, and said share is protected by entry barriers.
Market power is often the whole ballgame. According to law professor Michael Carrier, between 1999 and 2009, courts disposed of 96.8 percent of rule-of-reason cases because plaintiffs failed to meet their burden of proving market power or demonstrating anticompetitive effects. Carrier found that of 215 cases in which anticompetitive effects were not proven, the court disposed of 66 cases (30.7 percent) after finding a lack of market power without discussing anticompetitive effects. When direct evidence is a mixed bag or otherwise not dispositive, the locus of attention shifts to the market-definition debate. What is the relevant market in which the accused firm is exercising market power?
One method by which economists delineate the boundaries of an antitrust market is the set of “practical indicia” identified in Brown Shoe Co. v. United States (1962). The Brown Shoe factors assist courts in determining whether a proposed market is economically meaningful, and are considered by the Department of Justice and the Federal Trade Commission (FTC) when determining the relevant market during merger review. They ask whether buyers and sellers perceive the market as distinct in ways that would allow a firm controlling it to exercise pricing power. Examples of such factors include industry or public recognition of the market, peculiar product characteristics and uses, or distinct prices. When met, each of these factors supports the notion that the evaluated product exists within the boundaries of a relevant market.
Economists may also use a hypothetical monopoly test (HMT) to establish relevant markets in antitrust cases. The HMT market definition exercise examines “customers’ ability and willingness to substitute away from one product to another in response to a price increase” by a hypothetical monopolist. To use an earlier example, if a monopoly seller of subscription-based streaming services could profitably raise prices over competitive levels, then subscription-based streaming service constitutes a relevant market; if not (because, say, too many customers would switch to advertiser-supported video), then the next closest services (i.e., advertiser-supported video) would be added to the monopolist’s portfolio, with the test repeated until the price hike is profitable. The two methods of defining a market are cumbersome and prone to judgment calls.
If these market definition tests sound complicated, that’s because they are. And because of the expansion of the rule-of-reason evidentiary requirement, plaintiffs often need to trudge through these complex exercises to establish market power—a process that can easily zap both a plaintiff’s resources and a judge’s goodwill.
Winning the Market-Definition Battle, Losing the Case
To see how this process works in practice, a quick case study is in order. I was the FTC’s economic expert in a recent challenge of Meta’s acquisition of Within, the maker of Supernatural, the most popular virtual-reality (VR) fitness application. Per the FTC’s theory of harm, Meta was buying Supernatural in lieu of entering the VR-based fitness segment with its own new product (a variant of its Beat Saber app in partnership with Peloton); thus, the merger represented a loss of potential competition. Basically, Meta was trying to buy the whole market to get rid of a rival, instead of expanding consumers’ options.
As explained above, the rule of reason requires that, before it could explore the theory of harm, the FTC needed to establish that VR-based fitness apps were a relevant antitrust market. Meta and its experts asserted that the relevant market contained, among other things, fitness apps on gaming consoles and non-VR-based fitness products and services such as Peloton and Tonal. The proof of market definition consumed roughly half of my (very long) expert report. In addition to the Brown Shoe factors, I also offered evidence to support the notion that a hypothetical monopoly provider of VR-based fitness apps could profitably raise prices over competitive levels because consumers would not substitute for products like Tonal, as Meta suggested (that is, an HMT).
There was one hitch: because Supernatural had never changed its subscription price, we could not estimate the price sensitivity (formally, the demand elasticity) among its customers directly using econometric analysis. Instead, we had to survey existing and potential VR customers as to their sensitivities about a hypothetical price hike for Supernatural. In response, Meta hired a survey expert to criticize my methodology. Nearly the entirety of my deposition and my cross-examination concerned the reliability of my survey technique. In the end, the attack on the survey was unsuccessful, as the court decided (correctly in my view) that the FTC’s “Brown Shoe factors [were] sufficient to inform the Court’s understanding of the ‘business reality’ of the VR dedicated fitness app market.” The evaluation of said proof by the court consumed roughly 22 pages of its decision, a significant investment of resources.
Was Meta’s attack on my survey methodology really wasted? Perhaps not. While the court ruled in the FTC’s favor on market definition, it was a tough slog, and lots of goodwill had been burned up before consideration of the anticompetitive effects even began. After reviewing the evidence, the court (wrongly in my view) decided that the FTC failed to prove that Meta would have entered the VR-based fitness market with its own new product, absent the acquisition.
The court seized on the fact that Meta largely abandoned internal planning for the Beat Saber-Peloton app after it initiated its pursuit of Supernatural in March 2021. But this should indicate that Meta perceived the two options as mutually exclusive. The relevant question was, what is the probability that Meta would have entered the VR-based fitness space de novo with the Beat Saber-Peloton app conditional on not having acquired Supernatural? By examining the likelihood of entry conditional on the anticompetitive acquisition already having taken place, the court answered the wrong question. In January 2026, Meta elected to quietly kill off Supernatural, ceasing all updates on the app, a reduction in output more extreme than our models even contemplated.
This outcome underscores a broader lesson: the market-definition burden chewed up considerable resources and the court’s goodwill. Having awarded a “victory” to the plaintiffs on market definition, the court was more skeptical in considering the evidence of Meta’s de novo entry.
Two-Sided Markets Raise the Evidentiary Bar Even Higher
If the expense and court-exhaustion required of plaintiffs by the market-definition exercise weren’t enough, courts have now added another wrinkle to further intensify plaintiffs’ market definition burden. Traditionally, under the rule of reason, a defendant’s efficiency defenses must generally accrue to consumers in the same relevant market in which the plaintiff alleged competitive harm. Courts have been reluctant to permit defendants to justify anticompetitive effects by pointing to broader social or economic benefits, or beneficiaries in other markets. In Philadelphia National Bank (1963), the Supreme Court determined that benefits from economic stimulus to the city of Philadelphia could not justify harms to the bank’s customers: “We are clear, however, that a merger the effect of which ‘may be substantially to lessen competition’ is not saved because, on some ultimate reckoning of social or economic debits and credits, it may be deemed beneficial. A value choice of such magnitude is beyond the ordinary limits of judicial competence, and in any event has been made for us already, by Congress when it enacted the amended § 7 [of the Clayton Act].”
And such a value choice is beyond the limits of economics, which cannot provide crisp advice to courts on how to weight the impacts across two disparate groups. The danger of this weighting exercise is evidenced in a series of horrific decisions in favor of the NCAA, which managed to effectuate a wage-fixing scheme against (largely Black) athletes for decades under the rationale that viewers with a taste for amateurism benefitted from the scheme, and those benefits served to offset the harms to the athletes. Such third-party offsets were, until recently, effectively barred from consideration.
But in Ohio v. American Express (2018), the Supreme Court re-opened the permissible harm-balancing door by requiring courts to consider effects on both sides of a two-sided transaction platform. Ohio and other states challenged American Express’s (Amex) use of anti-steering provisions, which as the name suggests, barred merchants from steering customers to lower cost platforms—that is, payment platforms (credit cards) that charged merchants a lower transaction fee than Amex. If merchants could steer customers towards using other cards, Amex would have to compete for merchants’ business on the merits and decrease its transaction fee. Under the old rules, Amex would be barred from citing benefits of the conduct to cardholders.
To bring back these forbidden third-party offsets, the Court redefined the relevant market to be “two-sided,” with merchants on one side and cardholders on the other. Because cardholders were now part of a relevant market that involved an exchange between Amex and merchants (the natural counterparty), any purported benefits to cardholders could be counted against the harms to merchants. The benefit was ill-conceived, as cardholders were also injured by the anti-steering provisions—economics shows that when markets are competitive, an ad valorem tax like Amex’s transaction fee is largely passed through to consumers in the form of higher prices. The Court insisted, however, that higher fees earned off the backs of merchants could fund more generous cardholder rewards, such as Centurion lounges in airports.
Perhaps recognizing its gift to the defense bar, the Court added a limiting principle: for a market to be considered a two-sided platform, indirect network effects must flow in both directions. In the case of Amex, that meant that more merchants made the platform more valuable to cardholders, and more cardholders made the platform more valuable to merchants. Because readers of a newspaper did not value the addition of more advertisers, the Court reasoned, newspapers would not be considered a two-sided platform under this requirement.
As if this new efficiency justification isn’t enough of a gift to defendants, by introducing the concept of a two-sided market, the Court has erected new roadblocks for plaintiffs to define markets. Recall that there are primarily two ways to define markets: the Brown Shoe factors, and the hypothetical monopoly test (HMT). Application of the Brown Shoe factors isn’t really affected under the two-sided lens. But the standard one-sided HMT arguably (at least per defendants’ experts) no longer speaks to whether a hypothetical monopolist of a two-sided platform could raise prices. The purported problem is that, in a two-sided market, prices and demand on one side cannot be analyzed independently from the other side. For instance, a price hike on one side could put downward pressure on the price of the other side of the market, via what economists called the seesaw principle, thereby affecting demand on the other side.
Economists quickly scrambled to revise the HMT to account for a two-sided test. David Evans published a paper showing how a two-sided platform made the math exceedingly complex. Evans then tried to introduce a two-sided HMT on behalf of plaintiffs in Epic v. Apple (2021), but the district court rejected it: “The Court finds Dr. Evans’ [HMT] analysis fatally flawed by several standards, including his own. Dr. Evans has acknowledged that a double-sided…test should include simultaneous testing of both sides of the market using at least 14 inputs.” In light of the intractable nature of the two-sided HMT, and with the one-sided HMT no longer dispositive, plaintiffs are left with vanishingly few options.
The Court’s foray into two-sided market expansion has made market definition even harder for plaintiffs. This new evidentiary burden undermines plaintiffs’ efforts to establish market power, a necessary condition to establishing anticompetitive effects.
***
In sum, bringing an antitrust case is not for the faint of heart. The evidentiary burdens on plaintiffs are formidable, especially under the ever-broadening purview of the rule of reason. Recent developments, such as increasingly complex market-definition exercises, and the introduction of two-sided markets, have raised the bar even further. Critics who fixate on the FTC’s losses under Lina Khan’s stewardship—while ignoring its many victories—often underestimate just how structurally difficult antitrust litigation has become.
But this does not mean that public or private enforcers should retreat. As my father used to coach in youth baseball, you can’t get a hit unless you take the bat off your shoulders. Critics also ignore the deterrent effect of bringing a case on future anticompetitive behavior. Defeat in the FTC’s challenges to Meta’s acquisitions of Within and Instagram do not prove those cases lacked merit. They illustrate something else: the game is tilted toward defendants from the outset, and it often turns on a threshold inquiry—market definition—that often predetermines the outcome. To level the playing field, courts should focus their attention on proof of anticompetitive effects. And if caselaw precludes such an approach, Congress could classify certain conduct under per se treatment (e.g., the use of a common pricing algorithm) and give fresh instructions to courts on the weight to attach (if any) to market definition.
