Platform Antitrust.

AuthorHovenkamp, Erik
  1. INTRODUCTION

    About two decades ago, economists began to recognize that many firms face a complex "chicken-and-egg" problem that is not accounted for in conventional theories of competition and pricing. (1) The firm, acting as an intermediary, caters to two or more distinct "sides" of users that derive value from interacting with one another. But, the firm's service will not attract one side unless there are participating users on the other. Commenters had already recognized that the credit card industry is fundamentally "two-sided" in this way, with both merchants and consumers valuing the payment network only if the other side is actively participating. (2) But it has now become clear that such firms--known as two-sided (or multi-sided) platforms--are broadly scattered throughout the modern economy. (3) This has been fueled in part by advancements in technology and interconnectivity, (4) which provide the infrastructure for many prominent tech platforms like Facebook, Google Search, and Microsoft Windows. The prevalence of this business model, once deemed an idiosyncrasy of payment card networks, has provoked an outpouring of economic research on platforms and platform competition. (5)

    Consider an example. Airbnb lets homeowners list their spare bedrooms or entire homes for potential short-term rentals by prospective guests who subscribe to the service. For each stay booked over the platform, Airbnb takes a separate fee from each party. Further, the two sides of users--hosts and guests--use the service on a purely voluntary basis. For instance, Airbnb cannot necessarily guarantee a prospective guest that she will be able to book a home in a particular city on a particular night; there has to be a host in that city who has listed his home as being available. As such, each side's demand for the platform depends not only on the fees it is charged by the platform, but also on the extent of active participation by users on the other side: guests value the service only if there are reasonably many homes available for booking; and it is only worthwhile to be an active host if there are reasonably many guests who might submit booking requests.

    As such, the platform faces a significant coordination challenge. It must maintain an adequate balance of participation levels within the two sides in order for anyone to derive value from its service. Such balancing is a tangled and dynamic process: anything affecting participation on one side (such as the fees charged to its users) will necessarily influence the platform's appeal to users on the other side, which, in turn, affects the latter side's participation as well. (6)

    Two-sidedness adds new layers of complexity to the analysis of pricing and competition, and the results sometimes run counter to conventional economic intuition. As a consequence, platform economics has important implications for antitrust. However, this article's central policy claim is that such considerations can (and should) be accounted for within the established structure of antitrust's "rule of reason" analysis, which is a multi-stage burden shifting framework. (7) In particular, the courts should address two-sided commerce by continuing to rely on the rule of reason as a tool for sorting through complex cases incrementally--not by forcing all of the added complexities, including all conceivable efficiencies that might justify a platform's conduct, into the plaintiff's initial burden.

    A hallmark of platform commerce is the acute need to garner (and maintain) a sufficient balance of active participants across the two sides of the market. In some cases, this may require that the platform impose restrictions or coordinate transactions in ways that facially intervene in the competitive process. For example, Uber sets ride prices on drivers' behalf, which necessarily prevents them from competing with each other on price. (8) But if such competition were permitted, the resulting decline in prices would diminish drivers' earnings, potentially leading many fewer drivers to sign up in the first place. This makes the platform less attractive to riders, as their ability to book a trip on relatively short notice requires widespread availability of drivers. The ensuing decline in the rider base further diminishes drivers' interest in joining the platform, which further diminishes riders' interest, and so on. As in this example, when a platform's restraint is procompetitive (for reasons relating specifically to platform commerce), it is almost always because it helps to establish cross-platform participation, or else to prevent such participation from unraveling.

    However, it is critical to distinguish cases in which restrictive practices are plausibly necessary from those in which such explanations are merely pretextual. Platforms are profit-seeking firms, just like more conventional antitrust defendants. They may attempt to restrain trade not because it is commercially necessary, but simply because it is profitable. (9) For example, in Microsoft, the defendant tied the Windows operating system (which is a platform) to its own web browser, among other challenged practices. (10) Whatever its potential motivations, it seems unlikely that Microsoft's conduct was necessary to maintain demand for the Windows platform, which was already dominant. Instead, such conduct was likely an effort to monopolize the browser market by excluding smaller rivals.

    Antitrust commenters have increasingly emphasized the need to account for the distinctive features of platform competition, though there is no consensus on precisely how this should be undertaken in practice. (11) By contrast, until very recently, antitrust case law had given almost no substantive attention to these issues. While a few opinions had recognized the relevant commercial environment as involving distinct but highly-interrelated strands of transactions, (12) this typically amounted to little more than a passing observation.

    That is no longer the case, however, as the Supreme Court recently confronted platform commerce head-on in AmEx III (.13) In June of 2018, the Court issued its first decision on how antitrust's rule of reason (14) is to be applied in cases involving platform defendants. (15) It was superficially a question of how to define the "relevant market" for purposes of an antitrust adjudication. (16) In particular, the question was whether the market definition must include both groups of users, which would require a plaintiff to prove a net injury to competition across both user groups--not just to win on the merits, but simply to carry its initial burden. The Supreme Court held that it does. (17)

    Most of the important complexities arising under two-sided competition center on the juxtaposition of countervailing effects--that is, pro and anticompetitive effects--arising within the separate sides of the market. In fact, even outside the platform context, such a juxtaposition of plausible effects is very common in antitrust disputes. And the rule of reason ordinarily divides the burdens of establishing them; it bifurcates them into separate stages, delaying the need for potential balancing or "netting out" of the effects (which is notoriously difficult) until the final stage of the adjudication. By evaluating the effects carefully and independently, a court is better equipped to determine whether such balancing is genuinely necessary; and, if so, the court is at least in a better position to compare the relevant effects. However, the Court's AmEx III decision largely abandoned this burden-shifting framework, effectively collapsing the entire rule of reason analysis--and all of its intermediate inquiries--into the plaintiff's initial burden.

    Whether or not one agrees with its holding, the AmEx III decision is inarguably a watershed moment for platform antitrust. Against this backdrop, this Article considers how antitrust ought to accommodate the distinctive features of platforms and platform competition. It focuses principally on conduct evaluated under the rule of reason, (18) with emphasis on vertical restraints and unilateral conduct. (19) The analysis is organized as follows: I begin by providing an overview of the distinctive features of platforms and platform competition, as reflected within the platform economics literature. (20) Part III then explains how such factors may bear on the analysis of various restrictive practices that are already familiar within antitrust, but whose effects may become more or less concerning when undertaken by two-sided defendants. In Part IV, I address the economic effects of an important category of restraints that are unique to platform markets. Finally, Part V turns to the broad question of law that was at issue in AmEx III.

    One of the important competitive dynamics arising in platform markets is known as

    "steering." (21) This refers to any efforts aimed at inducing users to opt for one platform over another. The restraint at issue in AmEx III was an example of this: it prohibits its merchants from offering AmEx cardholders a better price at checkout if they agree to switch to an alternative card (e.g. Visa), since competing cards generally charge lower network usage fees to merchants. (22) But, more generally, steering restraints take many different forms, and arise in many platform markets. (23) In general, steering strategies are usually procompetitive, as they typically act as a vehicle for price competition among rival platforms. Restraints on steering should therefore be regarded as a potential source of serious antitrust concerns.

    However, as discussed in detail in Part III, many research articles suggest that such restraints may be necessary to maintain adequate participation, and thus regard their welfare effects as highly ambiguous. (24) The AmEx III opinion cites these commentaries copiously. Importantly, however, these arguments stem primarily from economic models involving a...

To continue reading

Request your trial

VLEX uses login cookies to provide you with a better browsing experience. If you click on 'Accept' or continue browsing this site we consider that you accept our cookie policy. ACCEPT