Welcome to the Antitrust Antidote—a quarterly publication analyzing significant U.S. antitrust decisions from legal and economic perspectives. Authored by former Federal Trade Commission (FTC) enforcer Koren W. Wong-Ervin with former FTC economist co-authors Jeremy Sandford and Nathan Wilson, alternating each quarter. The title of this series, “Antitrust Antidote,” while mostly meant to be humorous (perhaps limited to those who have heard Koren’s “let’s talk economics” as a cure for a bad day), also refers to the practical guidance we aim to provide throughout the series. We hope you enjoy it![1]
***
There were a number of decisions from March through June 2025 including: (1) a Ninth Circuit win by Microsoft and Activision in the FTC’s suit challenging the vertical integration between the two companies; (2) two district court decisions (U.S. et al. v. Google Ad Tech and U.S. et al. v. Live Nation) in which courts refused to apply the more defense-friendly refusal-to-deal standard, instead applying tying law; (3) a district court decision on the applicability of the Supreme Court’s landmark decision in Ohio v. American Express; (4) the denial of X Corp’s motion to dismiss antitrust counterclaims brought by a data scraper; and (5) the dismissal of a hub-and-spoke and data sharing putative class action involving competitors’ use of a software platform.
FTC v. Microsoft/Activision (9th Cir. May 7, 2025)
On May 7, the Court of Appeals for the Ninth Circuit affirmed the district court’s denial of the FTC’s motion for preliminary injunctive relief against Microsoft’s acquisition of the video game developer Activision Blizzard. Following the decision, the FTC dismissed its complaint, ending the case. The FTC alleged that the vertical integration between a content-platform operator (Microsoft) and a content producer (Activision) would substantially lessen competition in three U.S.-based markets: the markets for (1) gaming console devices, (2) gaming subscription services, and (3) gaming cloud-streaming services. The Ninth Circuit did not consider market definition.
As a threshold issue, the Ninth Circuit held that the district court applied the correct legal standard, rejecting the FTC’s argument that, under FTC Act Section 13(b)’s “more lenient standards” for preliminary injunctions, every factual dispute should be resolved in its favor. The court held that the FTC’s position “ignores the settled principle that a preliminary injunction remains ‘an extraordinary and drastic remedy’ that must be affirmatively justified by the FTC.”
With respect to the market for gaming console devices, the Ninth Circuit affirmed the district court’s finding that the FTC did not sufficiently show that Microsoft had the incentive to foreclose rivals either by making the popular Call of Duty game exclusive to its Xbox console or by releasing only an inferior version of the game for rival consoles. The court declined (on the grounds that it was unnecessary) to address the FTC’s argument that Microsoft’s binding commitments not to deny access should not be considered when assessing the FTC’s likelihood of success on the merits.
For the full foreclosure analysis, the Ninth Circuit pointed to the lower court’s findings that: (a) “Call of Duty’s cross-platform play is critical to [Microsoft’s] financial success,” e.g., Activision’s CEO testified that the company’s Call of Duty revenues from PlayStation “are probably twice the Xbox revenues”; (b) Microsoft “would be expected to experience serious ‘reputational harm’ if it pulled Call of Duty from PlayStation and thereby blocked millions of PlayStation gamers’ access to the game”; (c) the FTC had “not identified any instance in which an established multiplayer, multi-platform game with cross-play . . . has been withdrawn from millions of gamers and made exclusive”; and (d) “despite exhaustive discovery involving nearly 1 million documents and 30 depositions, the FTC ha[d] not identified a single document which contradicts Microsoft’s publicly stated commitment to make Call of Duty available on PlayStation (and Nintendo Switch).”
For the partial foreclosure analysis, the Ninth Circuit agreed with the FTC that the district court erred in concluding that “[i]f the FTC has not shown a financial incentive to engage in full foreclosure, then it has not shown a financial incentive to engage in partial foreclosure.” The court, however, went on to say that the district court “also separately held, in addition, that the FTC presented insufficient evidence to support its partial foreclosure theory.” The court pointed to the record evidence that no game developer had ever “intentionally develop[ed] a ‘subpar game for one platform versus another,’ because it would lead to a significant loss of goodwill among gamers,” as well as the lack of evidence that Microsoft had engaged in such conduct in the past.
The conclusion that “the mere fact that a company does not have a financial incentive to engage in full foreclosure does not, without more, establish that it similarly lacks an incentive to engage in partial foreclosure” is correct as a matter of economics. The profitability of any foreclosure strategy depends on the particulars of supply and demand in the related markets, as well as the form that the foreclosure strategy takes.
In general, economists accept that finding total foreclosure is unprofitable does not rule out that a partial strategy could be profitable, especially when implemented via modest increases in the price of the input. That is because partial foreclosure through the raising of an input’s price to a rival may benefit the integrated firm in two ways. First, it tends to increase the profits earned on all continued sales to the downstream rival. Second, it may lead consumers to switch to the integrated firm’s own downstream product rather than pay the higher price needed to cover the increased input cost. In contrast, if the integrated firm engages in a foreclosure strategy by simply refusing to sell to its downstream rival, it can only benefit from the strategy through the recapture of sales.
By contrast, if the form of partial foreclosure simply takes the form of selling only a fraction of the quantity of the input previously provided, the conditions for profitability may be the same as for total foreclosure. That is because the route to profitability would be more likely to turn just on the fraction of foreclosed units that would shift to the integrated firm.
Fundamentally, however, economists recognize that there are no hard-and-fast rules when it comes to assessing the competitive effects of a non-horizontal transaction. Simple tools such as vertical arithmetic may provide initial intuition and possibly serve as a screen, but there is no substitute for thinking hard about how supply and demand work in the related markets, not just in a static sense but also dynamically. For example, one problem with strict reliance on static vertical arithmetic models is that they are limited to calculating diverted sales today and do not take into account dynamic effects such as an acquiring firm’s plans to grow the business and how that might affect the profitability analysis.
With respect to the library subscription services market, the Ninth Circuit held that the district court did not abuse its discretion by holding that the FTC had not made an adequate showing that the merger would substantially lessen competition. The court noted that, given that Activision had long opposed putting its content on library subscription services, the merger’s effect of making such content available for the first time in the subscription market, even if exclusive to Microsoft, would not substantially lessen competition.
Finally, with respect to the cloud-streaming market, the Ninth Circuit held that the district court did not abuse its discretion in finding an insufficient likelihood of success given that the FTC failed to show that Activision content would be available to this market in the absence of the merger.
U.S. et al. v. Google (E.D. Va. Apr. 17, 2025)
On April 17, a Virginia federal district court issued a decision ruling in part for Google and in part for the DOJ and State AG Plaintiffs. In Google’s favor, the court rejected Plaintiffs’ attempt to challenge Google’s long-ago consummated mergers, concluding that “Plaintiffs have failed to show” that Google’s acquisitions of DoubleClick and Admeld were anticompetitive. The court also rejected Plaintiffs’ claim that Google unlawfully monopolized a so-called “advertiser ad network market,” concluding that “Plaintiffs have failed to show” that such a market constitutes a relevant antitrust market. The court did not find that Google has monopoly power—much less acquired or maintained monopoly power via anticompetitive acts—in any market for advertiser buying tools. Instead, the court concluded that advertisers have many choices outside of Google.
In Plaintiffs’ favor, the court found that Google unlawfully acquired and/or maintained monopoly power in two separate markets—one for publisher ad servers and the other for ad exchanges—through a “series of anticompetitive acts” consisting of three product design decisions (“First Look,” “Last Look,” and the Unified Pricing Rules) related to how Google’s auctions work. All three involve Google’s unilateral decision not to deal with and aid rivals. The court also found that Google engaged in unlawful tying in violation of Sherman Act Section 1 through “technical and policy restrictions that prohibited publishers from receiving real-time bids from AdX [Google’s ad exchange] (the tying product) unless they also used DFP [Google’s publisher ad server] (the tied product).” In so holding, the court erroneously rejected Google’s argument that the conduct amounted to a lawful refusal to deal with rivals.
The court accepted the DOJ’s position that Google was imposing a tie by refusing to allow its ad exchange to submit real-time bids from any ad server besides its own. In concluding that the refusal to deal doctrine from the Supreme Court’s decision in Trinko does not apply to Google’s conduct, the court reasoned that the conduct at issue is not an outright, unilateral refusal to deal with a rival but rather a constraint on customers because it “effectively … compelled” customers to use the tied product, and had the “effect of limiting Google’s publisher customers’ choice of publisher ad server for reasons other than competition on the merits.” The court accepted the DOJ’s theory that Google “effectively” ties its products together via a technical tie between its products and the economic infeasibility of using more than one ad server, even if Google does not expressly condition access to its products on a customer agreeing not to deal with rivals. In so holding, the court ignored the practical reality that the “tie” boiled down to a refusal to deal with rivals and that, unlike a typical tying claim in which the remedy is prohibiting the tie, here, the remedy would require a compulsory dealing obligation.
The court also suggested that Trinko is, or should be, limited to highly regulated industries “in which state and federal regulators require the leading firms to share access to capital-intensive infrastructure.” One of the (many) problems with this position is that it ignores the underlying economic logic of Trinko, which applies beyond regulated industries. Trinko was all about protecting the incentives to innovate of both the monopolist and its rivals; avoiding courts as central planners having to set the proper price, quantity, and other terms of dealing; and avoiding collusion—all of which apply equally to ad tech. Subsequent decisions applying Trinko—including then Judge (now Justice) Gorsuch’s decision in Novell v. Microsoft—are not limited to regulated markets.
As a general matter, economic theory does not tend to support an antitrust duty to deal. As the Court noted in Trinko, the virtue of forced sharing is “uncertain.” This recognition is in line with modern economic learnings. A compulsory sharing obligation “has two main and opposing effects on welfare.” See Padilla, Ginsburg, & Wong-Ervin (2019). On one hand, it “reduces the incentives to innovate both in the first place and in creating competing alternative technologies. Indeed, those who advocate forced sharing often underestimate the abilities of rivals to create workarounds or other competing products.” “Working in the other direction, compulsory [sharing] may increase [static] competition in the short term.” While consumers gain from increases in static efficiency, “economics teaches us that the gains from dynamic efficiency, including innovation . . . are an even greater driver of consumer welfare.” Indeed, Robert Solow won the Nobel Prize in economics for demonstrating that gains in wealth are due primarily to innovation—not to marginal improvements in the efficiency of what already exists.
Judge Brinkema also refused to treat the ad tech stack as a single two-sided market under the Supreme Court’s decision in AmEx, in which the Court required an integrated competitive-effects analysis, i.e., one that takes account of all sides of a platform, and rejects the separate-effects assumption that harm to consumers on one side of the platform means there has been harm to competition. Judge Brinkema reasoned that “[d]istinct products should not be grouped into a single omnibus market simply because they work together to achieve the same overarching purpose.” As Judge Douglas Ginsburg and Koren W. Wong-Ervin explained in a 2020 article, “AmEx is best understood to imply that an integrated competitive-effects analysis should be conducted for any platform that exhibits pronounced indirect network effects or interdependent demand.” In other words, whether to conduct an integrated competitive-effects analysis should not turn on whether a platform qualifies as a single, simultaneous transaction platform. Instead, the key determinant is the degree of interdependencies of demand.
The remedies trial is scheduled to begin on September 22, 2025. Plaintiffs’ proposed remedies include:
- Divesture of AdX (Google’s ad exchange).
- Phased Divestiture of DFP (Google’s publisher ad server). Phase 1 requires Google to build an integration for rivals that would allow DFP to integrate with and receive bids from an open-source solution (known as “Prebid header-bidding wrapper”) in the same manner and on the same terms as DFP integrates and solicits bids from AdX. Phase 2 would require Google to modify DFP to separate out the code that performs the final auction and then provide this auction-logic code under an open-source license. Phase 3 would require Google to divest the entirety of the remainder of DFP.
- A series of behavioral remedies. These include: (a) prohibiting Google “for a period of ten years from operating an ad exchange, or any product with similar functionality that transacts any open web display advertising”; (b) data-sharing commitments, including ones aimed at categories of data that were not the subject of Plaintiffs’ theory of harm at the liabilities stage; and (c) requirements that Google’s buyside tools deal with all third-party ad tech tools on non-discriminatory terms with respect to bidding, matching, placement of ads, or provision of information, except at the express instruction of an advertiser.
U.S. et al. v. Live Nation (S.D.N.Y. Mar. 13, 2025)
On May 13, a New York federal district court denied Live Nation’s motion to dismiss the DOJ and State Attorneys General’s Sherman Act Section 1 tying claims, rejecting Live Nation’s attempt at this early stage to recast the claim as a refusal to deal. The court did not foreclose the possibility that after discovery it could find otherwise, depending on what the evidence shows. Specifically, the court stated: “The facts may ultimately show that the tying claim here is nothing more than a refusal-to-deal claim foreclosed by Trinko …. But at this stage, the Court’s role is to determine whether the complaint states a plausible tying claim, and it does.” As the court explained, going forward, the question will be whether the evidence shows that Live Nation: (1) coerced artists (i.e., non-rival customers) into using Live Nation as their promoter (the alleged tied product) in order to gain access to Live Nation amphitheaters (the alleged tying product); or instead (2) engaged in an outright refusal to deal with concert promoters who rent venues on behalf of artists (i.e., a refusal to deal with a rival). “If the evidence shows that promoters book venues on behalf of specific artists, that artists are the driving force behind which venues to book and when, and that artists are coerced into using Live Nation as their promoter if they want access to Live Nation’s amphitheaters, plaintiffs may have a viable tying claim.”
In denying the motion to dismiss, the court relied on the DOJ’s allegations that Live Nation targeted artists, not just rival promoters, concluding that “these allegations aren’t just about a refusal to deal with a rival promoter. They are about coercion of artists.” The relied-upon allegations included that Live Nation has had “a longstanding policy going back more than a decade of preventing artists who prefer and choose third-party promoters from using its venues,” and that “if an artist wants to use a Live Nation venue as part of a tour, he or she almost always must contract with Live Nation as the tour’s concert promoter.”
Separate from whether tying or refusal to deal case law should apply, the allegations focused on in the district court’s opinion tee up interesting economic issues. Traditionally, tying theories involve a putative monopolist in one market allegedly leveraging this position to impact a related, more competitive market. However, in its Complaint, the DOJ alleged that Live Nation already possesses significant market power in both the venue and promotion markets. Thus, there is some sense in which the tying story focused on in the district court opinion alleges that connecting distinct dominant positions in complementary markets is enabling the capture of even more rents than would otherwise be possible.
In general, economic models of the vertical restraints implicated in the DOJ’s Complaint underscore the importance of understanding how demand and supply of the various products may be connected. For example, is there demand for venues independent of promotion and ticketing? Are there efficiencies to handling promotion for an artist in multiple geographic areas? Does an integrated firm have incentives to take costs out of the system? Fundamentally, the facts regarding these issues of demand complementarities, scale efficiencies, and the internalization of multiple margins likely will be key elements for the parties and their experts to develop and explain during the merits trial.
X Corp. v. Bright Data Ltd. (N.D. Cal. Apr. 18, 2025)
On April 18, a California federal district court denied X Corp’s motion to dismiss a data scraper’s Sherman Act Section 1 and 2 counterclaims alleging that, after acquiring Twitter, X changed its Terms of Service to bar X’s data buyers and users from doing business with data scrapers like the Counter-Plaintiff, who scrape publicly displayed data that X does not own. The conduct is alleged to harm competition in the market for public-square data in the United States. Both Counter-Plaintiff and X are alleged to sell such data to businesses, academics, and others who use it to discern up-to-the-minute insights about current events.
The court rejected X’s no duty-to-deal argument on the grounds that X is “doing more than merely refusing to deal with scrapers,” but rather “is forcing its own customers not to deal with scrapers, in perpetuity, a much different problem under the Sherman Act.” The court went on to say that X’s arguments that Counter-Plaintiff “must allege the relevant market more specifically to establish market power is well-taken, but it demands too much for purposes of pleading.” Similarly, X’s business justifications for the restraint (to prevent free riding), while plausible, cannot be used at the motion to dismiss stage to overcome the allegations in the complaint.
Economics has long understood exclusive contracts to serve a potentially important role in preventing free-riding by aligning incentives across independent entities. For example, if a product manufacturer invests heavily in marketing its product, it may wish to have an exclusive contract with retailers so that customers attracted to the retailer by the marketing are not diverted by an unadvertised lower priced alternative they see on the shelves. The facts in the instant matter do not necessarily align neatly with seminal exclusive dealing examples. In particular, X’s conduct appears to involve shifting from a policy of sharing (at no cost) a byproduct of its communications platform business to reserving and monetizing that byproduct itself. Economic theory would tend to predict that exclusives have anticompetitive effects when the impact of losing sales to the party with the exclusive would deprive a firm operating as a competitive constraint of minimum-efficient scale and lead to an enduring loss of competition in the market. Whether or not some reduction in scale translates to reduced efficiency will be fact dependent and can vary across (or potentially even within) industries. Moreover, it must be the case that free-riding concerns are not sufficient to deter any important pro-competitive actions.
Segal v. Amadeus IT Group, S.A. (N.D. Ill. Mar. 31, 2025)
On March 31, an Illinois federal district court dismissed with leave to amend a proposed class action under Sherman Act Section 1 alleging that eight operators of luxury hotel brands and the owner of a software platform (Demand360) engaged in a hub-and-spoke conspiracy with Demand360 as the hub. Plaintiff alleged that, “[b]y virtue of exchanging proprietary, non‐public, present and forward‐looking demand data” via Demand360, “the Hotel Defendants have been able to charge increased rates for Luxury Hotel Rooms despite historically low overall demand, divorced from the market forces that drive supply and demand in a competitive environment.” The court held that the complaint failed to allege a plausible hub-and-spoke conspiracy given the absence of “any allegations that the hotel defendants, the ‘spokes’, made agreements with one another sufficient to form a ‘rim.’”
The court also rejected Plaintiff’s argument that he alleged a plausible Section 1 claim based on the sharing of non-public information. The court pointed to the fact that Plaintiff did “not allege that by sharing occupancy information, defendants conspired to stabilize or otherwise anticompetitively [a]ffect occupancy rates.” Rather, Plaintiff’s “claim appears to be rather that the defendants used occupancy data to affect pricing, a somewhat attenuated connection compared to the cases on which he relies.” The court noted that Plaintiff “seem[ed] to be asserting two contradictory theories”: First, “that if the user of [Demand360] learns from the occupancy data it receives that its competitors are close to or fully‐booked during the time frame in question, it can raise its rates without fear that the competition will undercut them, since the competition has little or nothing available to sell.” But on the other hand, Plaintiff contends that the data show “that as a result of the exchange of occupancy information, prices rose as demand fell, demonstrating that the market is not responding to normal competitive forces.” The court concluded that “[t]he problem is that one theory depends on the competition being fully‐booked and having no supply, and the other theory depends on prices increasing as occupancy declines. … Without a claim that the information‐sharing led to anticompetitive pricing, it is hard to be sure what [Plaintiff] means to allege and what pleading standards should apply.”
The court has indeed identified a key tension in Plaintiff’s argument. If the theory is that the sharing of information facilitated more accurately responding to changes in aggregate demand, then one would expect to find changes in occupancy rates positively correlated with prices. However, the data advanced in Plaintiff’s second amended complaint show the reverse. In other words, prices increased while occupancy declined. This inconsistency does not preclude the possibility that Plaintiff is correct that reduced uncertainty had some impact. Other factors may also have been changing that confound the relationship.
Separately, even if price and occupancy rate changes were positively correlated, it would not necessarily indicate that Defendants were actively colluding. Technological changes that enable efficient responses to shifts in demand would produce such a positive correlation. While this could lead to higher prices in high demand periods, consumers in low demand periods would be expected to benefit from lower prices, confounding simple welfare conclusions. Moreover, published research shows that reducing demand uncertainty can both enable collusion where it was previously not sustainable and disrupt it where it had prevailed.
Nathan Wilson & Koren W. Wong-Ervin
***
[1] The authors thank Brett Wierenga, Ellen Geyer, and Ramzie Fathy for their research assistance. The views and opinions set forth herein are the personal views or opinions of the authors; they do not necessarily reflect the views or opinions of the organizations with which they are affiliated, or those organizations’ management, affiliates, employees, or clients. Jones Day and Econic Partners represent or otherwise work with a number of clients, including Google, that may have an interest in the subject matter of this article. This publication was not funded or sponsored.
