By Douglas H. Ginsburg & Koren W. Wong-Ervin*
The Supreme Court’s 2018 decision in Ohio v. American Express (AmEx) sparked a number of debates over its meaning and scope. Those debates have been rekindled by recent decisions in FTC v. Surescripts and U.S. v. Sabre. The debates focus on (1) whether AmEx is limited to two-sided transaction platforms, (2) the proper way to measure competitive effects, and (3) whether AmEx is limited to claims under Section 1 of the Sherman Act. This Article provides legal and economic analysis of these issues.
The AmEx Decision
The Court in AmEx applied the rule of reason under Section 1 of the Sherman Act to analyze the competitive effects of vertical restraints that AmEx imposed upon merchants, which prevented them from steering customers toward other payment media that were less costly to the merchant. The Court held that the Department of Justice (“DOJ”) failed to satisfy the “first step” of the three-step, burden-shifting framework because it focused on only one side of the two-sided market at issue. Specifically, the Court held that, while the DOJ showed AmEx’s allegedly anticompetitive conduct increased the fees AmEx charged to merchants, because it did not take account of offsetting benefits to cardholders, that did not suffice to prove the conduct had a net anticompetitive effect in the market for credit card transactions. The Court instructed that “[t]o demonstrate anticompetitive effects on the two-sided credit-card market as a whole, the plaintiffs must prove that AmEx’s anti-steering provisions increased the cost of credit-card transactions above a competitive level, reduced the number of credit-card transactions, or otherwise stifled competition in the credit-card market.”1 In other words, the Court required an integrated competitive-effects analysis, i.e. one that takes account of all sides of a platform, and it rejected the separate-effects assumption that harm to consumers on one side of the platform means there has been harm to competition.
The Court also held that, “[i]n two-sided transaction markets, only one market should be defined.”2 The Court defined “transaction platforms” as two-sided platforms that “facilitate a single, simultaneous transaction between participants.”3 The Court went on to say that “[t]he key feature of transaction platforms is that they cannot make a sale to one side of the platform without simultaneously making a sale to the other.”4
AmEx Beyond “Transaction Platforms”
There has been much debate about whether AmEx is limited to transaction platforms, with some economists criticizing the Court’s distinction between transaction and non-transaction platforms.5 As the Global Antitrust Institute has explained, even with non-transaction platforms, it may be necessary to consider both sides of the platform in order to assess accurately the competitive effects of particular conduct.6 We believe the decision provides legal support for this economic conclusion. Indeed, 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. Indirect network effects occur when the value obtained by a customer on one side of the platform increases with the number of customers on the another side of the platform; e.g. video game developers value video game consoles more when there are more game users, whilst game users value consoles that have more games.7 Interdependent demand refers to a situation in which profit maximization requires taking all sides of the platform into account.
The Court deemed “transaction platforms” a “special type” (or subcategory) of two-sided platforms.8 The Court defined a “two-sided platform” as one that “offers different products or services to two different groups who both depend on the platform to intermediate between them.”9 According to the Court, all two-sided platforms (the umbrella term), not just the subset of transaction-platforms, differ from “traditional markets” in that, among other things, they exhibit indirect network effects. The Court went on to explain that “[t]wo-sided platforms, therefore, must take these indirect network effects into account before making a change in price on either side.” As such, “[p]rice increases on one side of the platform . . . do not suggest anticompetitive effects without some evidence that they have increased the overall cost of the platform services.”10
The Court then said “it is not always necessary to consider both sides of a two-sided platform. A market should be treated as one sided when the impacts of indirect network effects and relative pricing in that market are minor.” “Not always necessary” obviously implies it is sometimes necessary, even for non-transaction markets, to consider effects on both (all) sides of a platform. As the Court explained, whether it is necessary turns upon the strength of the network effects. The Court’s further statement that two-sided transaction markets are a special type of two-sided platform because they “exhibit more pronounced indirect network effects and interconnected pricing demand” does not imply that AmEx applies only to this transactional subset of two-sided markets. Simply put, an integrated competitive-effects analysis is in order whenever a two-sided platform has pronounced indirect network effects, and the Court identified the subset of “transactional platforms” as one example in this subset that can be assumed to have meaningful indirect network effects. The Sabre decision, in which the court held that, under AmEx, “it is always necessary to consider both sides of the subset group of two-sided transaction platforms,” is consistent with this insight. Given that the Sabre court concluded the platform at issue constituted a transaction platform, it did not reach the issue whether, or when, the analysis may be necessary for non-transaction platforms.11
The crucial question then becomes how significant must the indirect network effects be to trigger the requirement of an integrated competitive-effects analysis. The Court was imprecise on this issue: “Indirect network effects exist where the value of the two-sided platform to one group of participants depends on how many members of a different group participate.”12 It also suggested that unidirectional effects are less important than are bidirectional effects, distinguishing newspapers from the credit-card market on the ground that “in the newspaper-advertisement market, the indirect network effects operate in only one direction; newspaper readers are largely indifferent to the amount of advertising that a newspaper contains. Because of these weak indirect network effects, the market for newspaper advertising behaves much like a one-sided market and should be analyzed as such.”13 Ultimately, the Court’s rationale for its holding on competitive effects is that, with transaction platforms, “[t]o optimize sales, the network must find the balance of pricing that encourages the greatest number of matches between cardholders and merchants.” In other words, the Court recognized the importance of interdependencies of demand, which imply a profit-maximizing firm will take into account effects on both sides of the platform.
Antitrust analysis, of course, requires an understanding of firm behavior. Here, there is no reason for the AmEx platform to exist if not to facilitate transactions between two sides. As David Evans says of platforms generally, “The platform reduces transaction costs by acting as an intermediary between the two sides and thereby organizing and facilitating trade between different types of customers. This trade increases economic efficiency and surplus for the participants, since, by its nature, trade makes both parties better off.”14
In a recent article, former Federal Trade Commission (“FTC”) Commissioner Joshua Wright and former FTC economist John Yun explain why, as a matter of economics, the question whether an integrated competitive-effects analysis should be conducted turns on whether there are meaningful interdependencies of demand.15 As the Global Antitrust Institute has explained, “[w]hile payment card platforms are distinct for having a fixed proportion of consumption on the two sides, the importance of this distinction can be overstated.”16 Even with respect to non-transaction platforms, to assess the relative participation level on one side of the platform it is necessary to consider the participation level of the other side. “Profit maximization still depends on a joint assessment of the pricing and volume on multiple sides. Whether the volume on each side is a precise 1-to-1 matching or something highly correlated does not change this fundamental fact.”17
Take operating systems for example. An operating system provides an interface between hardware devices, such as a computer or a smartphone, and software applications, such as a word processor. A successful operating system requires attracting both users and application developers. Users value an operating system more when more applications are written for it, and software developers value it more when more consumers are offered their product. While there is no single, simultaneous transaction or shared output between the two sides, the attractiveness of the platform to users will affect revenues to applications developers, and revenues to applications developers will affect the supply of applications to users. As Rochet and Tirole have explained, this is “why prices on various sides are economically related and do not amount to a comparison of the ‘apples’ on one side of the market with the ‘oranges’ of the other. The ‘link’ between the prices are the externalities—i.e., the cross-group effects between the various groups.”18
In his AmEx dissent, Justice Stephen Breyer suggests that this interrelationship is no different than that of complements in setting prices in a single-sided market.19 Rochet and Tirole address how the different sides of a platform differ from complements: “The interaction between the two sides gives rise to strong complementarities, but the corresponding externalities are not internalized by end users, unlike in the multiproduct literature (the same consumer buys the razor and the razor blade).”20 While there can be interrelated demand for complements, the issue is one of degree.
All this is to say that two-sided platforms form a continuum, with the degree of interdependent demand being the key determinant of whether an integrated competitive-effects analysis should be conducted. An integrated analysis is particularly appropriate when the alleged practice is inherently two-sided, such as the anti-steering rule in AmEx, which was designed to balance the interests on both sides of the platform.
Market-Wide Output as the Proper Measure of Competitive Effects
AmEx supports the (economically sound) position that, at least for markets involving transaction platforms, changes in market wide output, rather than in price, should be the indicator of changes in consumer welfare. In analyzing competitive effects, the Court quoted its earlier decision in Brooke Group to say: “Where . . . output is expanding at the same time prices are increasing, rising prices are equally consistent with growing product demand.”21 The Court also repeated its understanding that “[m]arket power is the ability to raise price profitably by restricting output.”22 The dissent took issue with this point, saying “the majority retreats to saying that even net price increases do not matter after all, absent a showing of lower output.”23 As the Global Antitrust Institute has explained, the Court’s point is particularly important in a platform market because “price is a significantly noisier signal of consumer welfare [there] than it is in a single-sided market. Prices might appear simultaneously as predatory on one side . . . and supra-competitive on the other side . . . . In contrast, the output levels on both sides are either identical or at least highly correlated.”24
AmEx Beyond Section 1
There is also a debate over whether AmEx is limited to claims under Section 1 of the Sherman Act. In Sabre, the court held that “for purposes of understanding AmEx, there is no meaningful distinction between Section 1 Sherman Act claims and a Section 7 Clayton Act merger challenge.”25 The court then applied AmEx’s two-sided-markets framework to the DOJ’s Section 7 challenge of Sabre’s proposed acquisition of Farelogix, rejecting the DOJ’s suggestion that AmEx applies only to the credit-card industry.26 In possible contrast, some interpret FTC v. Surescripts as limiting AmEx to Section 1. A close reading, however, reveals that the court in Surescripts distinguished AmEx not on the ground that it does not apply outside Section 1, but rather on the ground that AmEx involved a non-monopolist, whereas it was undisputed that Surescripts had monopoly power.27 Indeed, the court applied AmEx in the respect relevant here, as had the FTC in its complaint, which alleged the defendant’s exclusionary practices had raised “the net price [of its services] (taking into account both sides of the network).”28 In any event, limiting the two-sided market framework of AmEx to Section 1 cases would make no sense given that the economic underpinnings of the case apply to all two-sided transaction platforms (as well as non-transaction platforms that exhibit meaningful indirect network effects), particularly when the conduct at issue is itself two sided. For example, in a predatory pricing case, a platform’s decision to price in a way that subsidizes the more elastic side is two-sided conduct given the interdependent demand.
The answers to the questions we have addressed about the meaning and applicability of AmEx are of critical importance to getting the analysis right in future cases. Our examination of both the reasoning in AmEx and the relevant economic principles supports the conclusions that: (1) AmEx is not limited to transaction platforms, but instead applies to all two-sided platforms that exhibit meaningful indirect network effects; (2) changes in market-wide output are the best indicator of competitive effects in two-sided markets; and (3) the economic principles underlying AmEx hold for two-sided platforms, especially when the conduct at issue is two-sided, regardless of the antitrust cause of action in suit, whether Section 1 or another.
* Douglas H. Ginsburg is a judge on the U.S. Court of Appeals for the District of Columbia Circuit, Professor of Law and Chairman of the International Board of Advisors to the Global Antitrust Institute at Antonin Scalia Law School, and a former Assistant Attorney General in charge of the Antitrust Division of the U.S. Department of Justice. Koren W. Wong-Ervin is a Partner at Axinn, Veltrop & Harkrider LLP and former Counsel for IP & International Antitrust and Attorney Advisor to Commissioner Joshua D. Wright at the U.S. Federal Trade Commission. The authors thank Joshua Wright and John Yun for their insightful comments and Max Fischer-Zernin for his research assistance. The views expressed here are the authors’ alone and do not necessarily represent the views of Axinn or any of its clients. Ginsburg owns shares in Google and Axinn represents a number of clients, including Google, that may have an interest in the subject matter of this article.
1 Ohio v. Am. Express Co., 138 S. Ct. 2274, 2287 (2018).
3 Id. at 2286.
4 Id. at 2277.
5 For an account of the debate, see, e.g. Joshua D. Wright & John M. Yun, Burdens and Balancing in Multisided Markets: The First Principles Approach of Ohio v. American Express, 54 Rᴇᴠ. Iɴᴅ. Oʀɢ. 4:6 717-40 (2019) [hereinafter “Wright & Yun”]; The Federal Trade Commission’s Hearings on Competition and Consumer Protection in the 21st Century, Platforms, Comment of the Global Antitrust Institute, Antonin Scalia Law School, Geo. Mason. L. & Econ. Res. Paper Series 18-39 (2018), available at https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3267489 [hereinafter “GAI Comment”].
6 GAI Comment, supra note 5; see also Joshua D. Wright & John M. Yun, Ohio v. American Express: Implications for Non-Transaction Multisided Platforms, Competition Policy International Antitrust Chronicle (Jun. 25, 2019).
7 David S. Evans, The Antitrust Economics of Multi-Sided Platform Markets, 20 Yale J. on Reg. 325, 332 (2003).
8 AmEx, 138 S. Ct. at 2277 (“Because the interaction between the two groups is a transaction, credit-card networks are a special type of two-sided platform known as a ‘transaction’ platform.”).
9 Id. at 2280.
10 Id. at 2286.
11 U.S. v. Sabre Corp., 1:19-cv-01548, 2020 WL 1855433 (D. Del. Apr. 8, 2020) (emphasis added).
12 AmEx, 138 S. Ct. at 2280.
13 Id. at 2286 (citation omitted).
14 David S. Evans & Richard Schmalensee, Two-Side Red Herrings 1-11 (2018) (unpublished manuscript), available at https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3261802.
15 Wright & Yun, supra note 5.
16 GAI Comment, supra note 5, at 13.
18 Jean-Charles Rochet & Jean Tirole, Platform Competition in Two-Sided Markets, 1 J. Eur. Econ. Ass’n, 990-1029 (2003) [hereinafter “Rochet & Tirole”]; see also Jean-Charles Rochet & Jean Tirole, Two-Sided Markets: A Progress Report, 37 Rand J. Econ., 645-667 (2006).
19 AmEx, 138 S. Ct. at 2296 (Breyer, J. dissenting).
20 Rochet & Tirole, supra note 18, at 991.
21 AmEx, 138 S. Ct. at 2288.
23 Id. at 2302.
24 GAI Comment, supra note 5, at 10; see also Wright & Yun, supra note 5, at 33-35.
25 Sabre, 2020 WL 1855433, at *34..
26 Id. at *33 (“DOJ argues that Amex does not govern here, but its arguments are unavailing. At trial DOJ suggested that Amex may be limited just to the credit card industry, but DOJ points to nothing in Amex to support such a conclusion.”).
27 F.T.C. v. Surescripts, LLC, No. CV 19-1080 (JDB), 2020 WL 264147 (D.D.C. Jan. 17, 2020).
28 Id. at *9.