The essential facilities doctrine is back. Yet, despite its recent endorsements, the doctrine’s criticisms linger. They range from allegations that monopolies lack incentives to monopolize adjacent markets to doubts about the doctrine’s administrability and fears of error costs. They also include claims of entrenching monopoly power and decry severe limitations of the antitrust remedy. Many of these objections to the essential facilities doctrine are fueled by persistent myths and misconceptions. It is time to move on and leave the debunked myths and disproven misconceptions behind.
By Nikolas Guggenberger[1]
The “essential facilities” doctrine is on the cusp of a reawakening in American antitrust law. The doctrine grants competitors the right to access essential facilities of monopolists to the extent that these competitors depend on the facilities and cannot reasonably duplicate them. This approach forced railroad companies and utility providers to share their infrastructure, for example. After a generation of decline, the doctrine is gaining momentum, as it has received prominent endorsements across the political spectrum: The 2020 Democratic House Majority, “Investigation of Competition in Digital Markets,” included the approach in its recommended toolkit to reign in Big Tech, as it its Republican counterpart, the brief report titled “Third Way.” In the same vein, several bills have just been introduced as part of a bipartisan “Anti-Monopoly Agenda for A Stronger Online Economy: Opportunity, Innovation, Choice,” most notably the “American Innovation and Choice Online Act,” sponsored by Rep. David Cicilline (D-RI). Several of the legislative proposals contained elements of access rights as they are core to the original essential facilities doctrine. Even the traditionally antitrust-skeptic business community seems out of lockstep, with smaller enterprises and their interest groups favoring more vigorous enforcement against dominant digital platforms.
This recent shift is remarkable, given skepticism with which U.S. courts have treated the essential facilities doctrine since the early 1990s. In light of the evolving debate about the merits of the essential facilities doctrine, I have recently addressed[2] some of the myths that brought about the doctrine’s decline and whose perpetuation still stand in the way of its revival. Among these are the exception-turned-rule one-monopoly rent theorem, the purely theoretical internalizing complementary efficiencies theorem, the overblown concerns about administrability, information aggregation, and error costs, and, finally, the misguided allegation of entrenching monopolies.
Before diving into the persisting myths, however, let me briefly lay out the case for the essential facilities doctrine in the digital economy.[3] Economic concentration has increased across industries in the U.S. over the past decades. This development has become especially evident in the digital economy. Just a few platforms dominate the landscape: in e-commerce, there is Amazon; there are two relevant app stores, Apple’s and Google’s; Facebook has a firm grip on social media; and Google has become synonymous with online search. Where the market does not provide reasonable alternatives to essential facilities, goods, or services, access rights can open markets and enable follow-on innovation. This is the logic behind the 1912 Supreme Court ruling in Terminal Railroad Association. On the internet, large platforms occupy positions in the economy comparable to the railroads of the late nineteenth and early twentieth century.
For the application of the essential facilities doctrine, let us focus on app stores. Apple would need to be considered a monopolist controlling the app store as an essential facility to fulfill the first prong of the essential facilities doctrine, as originally defined in MCI Communications. Many developers do not have sufficient substitutes because users tend to single home on one of the OS ecosystems, either Apple’s or Google’s. And these ecosystems are not necessarily interchangeable, because of the different user characteristics. Also consider that certain apps need to have the ability to reach all smartphone users to provide sufficient utility. To fulfill the second prong of the essential facilities doctrine, app developers must be practically or reasonably unable to duplicate the app store’s infrastructure. That inability stands to reason not only because of the incumbents’ size advantages but also because of the connection of the app stores with the operating systems and Apple’s hardware or licensing and tying arrangements in the case of Google.
The third condition for an essential facilities-claim requires that the platform denies a competitor the use of the facility. The distinction between an outright denial of use and inappropriate conditions frequently is impossible to draw. Take Epic’s (ultimately unsuccessful) complaint against Apple, for example. The deplatforming of a developer or their apps, as with Epic and its game Fortnite, can be seen as an outright denial of use. At the same time, Fortnite’s deplatforming is directly tied to Epic’s decision not to comply with the conditions for access that Apple demanded, namely, to refrain from offering alternative direct methods of payment for in-app purchases. Either way, the essential facilities doctrine remains applicable because “[a]greeing to deal on unreasonable terms is merely a type of refusal to deal.”
The fourth and final condition for liability under the essential facilities doctrine opens the assessment to an array of potential justifications, including legitimate business reasons that go beyond the mere profit maximization or desire to exclude rivals. There might be security or privacy concerns as well as objections to the content and features of the applications, for example. Security and privacy concerns may relate to technical deficiencies or business practices. Objections to the content and features of applications may be based on the lack of precautions against hate crimes, incitements of violence, the (mental) health impacts on users, and the protection of children, to name just a few. And in fact, Apple and Google cited Parler’s insufficient content moderation as cause for the app’s delisting in January 2021, following the violent storming of the Capitol building in Washington D.C.
As applied, the essential facilities doctrine does not require incumbents to open their facility free of charge. Also, the operator of the facility may use admission and governance criteria beyond strict technical neutrality. In many instances, strict technical neutrality — even if such standard were attainable — would diminish the usefulness of the service. Just think of a search engine. Instead, the essential facilities doctrine builds on a normative understanding of non-discrimination, focused on avoiding anticompetitive exclusions.
Finally, the essential facilities doctrine should not be conceptualized as a stand-alone policy response to concentration in the digital economy. In fact, there is often good reason to embrace structural remedies instead of the essential facilities doctrine or alongside the definition of access rights. These structural remedies may focus on horizontal concentration or vertical integration.
I. THE SINGLE MONOPOLY THEOREM
The most prominent critique of the essential facilities doctrine commonly goes by the name single monopoly rent theorem. incumbent to open their facility free of charge. In fact, the obligation to do so would be counterproductive, as it would undermine the ability of the operator to continue the service. Robert Bork, one of the most prominent figures of the Chicago School movement, argued that a firm could only reap monopoly profits once — hence the name of the theorem. The theorem is intuitive and, stated in the abstract and the simplest setting, true: where there is only a single monopoly rent, it can indeed only be captured once. The relevant question is whether the theorem’s assumptions broadly hold in the real world or whether they apply only in limited circumstances.
The hypothesis rests on several assumptions. First, the monopoly power must be absolute; as Johnathan Baker writes, the theorem only holds if a monopolist “has literally no rivals and faces no potential entrants, and if buyers have literally no alternative to the monopolist’s products.” Second, complementary products or services in adjacent markets must enter the consumer market in fixed ratios. Third, the adjacent market must be perfectly competitive. Even proponents of restrained approaches to antitrust enforcement acknowledge these assumptions as conditions for the validity of the single monopoly rent theorem. Contention arises over the typicality of conditions that define the boundaries between exceptional circumstances and policy guiding regularity. Newer evidence-based thinking rightly characterizes the theorem’s insights as an outlier, applicable in only a small handful of empirical settings. Its qualifying assumptions and exceptions have been shown to be so attenuated that the model represents the exception, not the rule, when describing monopolist incentives to monopolize adjacent markets.
Notably, the digital goods often do not need to be consumed together. To illustrate, consider Amazon. It mainly facilitates the sale of physical and virtual goods. For both types of products, the platform service is not strictly necessary, even if it is deemed essential for the purpose of the essential facilities doctrine. Amazon’s (own and third-party) inventory of physical merchandise could also be sold through a variety of traditional brick and mortar stores, from grocery stores to sports outlets and home improvement retailers. And while Amazon sells e-books in its proprietary Kindle format, creating a necessary nexus between the platform and this version of the e-book, the same content can be read in a different format on other e-book readers. Thus, Amazon and similarly situated platforms have an incentive to monopolize secondary markets like those for retail goods and e-book content because they cannot extract full monopoly rents from their market power over the platform services alone.
Revenues from outside sources constitute a further exception. And online, advertising funding is ubiquitous — an argument on which Barbara van Schewick bases her analogue analysis of telecommunication networks. Additionally, the characteristics of data — and specifically data exhaust, generated by observing incidental user behavior — provide an independent basis for the profitability of exclusionary conduct in adjacent markets.
Finally, the single monopoly rent theorem does not account for inevitable agency costs and misaligned incentives within the firm. To be more precise, managers at all levels of the firm face incentives to increase the size of the firm beyond its profit-maximizing scope. Even carefully designed compensation schemes cannot neutralize the appeal of bigness. Social status and power tend to increase with the size of the operation, not only with the operation’s profits. Invitations to Davos and Jackson Hole, private audiences with high-ranking government officials, and interview requests are only the most visible perks of bigness and power.
After deducting all the exceptions from the so-called single monopoly theorem, there is not much left — especially in data-driven platform markets. Rather than guidance for default enforcement rules, the theorem describes a theoretical exception with little practical application.
II. THE INTERNALIZING COMPLEMENTARY EFFICIENCIES THEOREM
Opponents of the essential facilities doctrine argue that monopolists benefit from the existence of a functional and competitive adjacent market. Efforts to exclude competitors would be detrimental to their bottom line. This idea is called the internalizing complementary efficiencies theorem. Indeed, open ecosystems often enable more innovation than closed alternatives, and several successful platforms have chosen open architectures to spur growth in their early stages. Once the market has tipped in favor of one dominant player, the now significant relative size advantage insulates the dominant platform from competition. As a result, the incentive structures change. It can become attractive to close in and push downstream competitors out of the market.
III. ADMINISTRABILITY, INFORMATION AGGREGATION, AND ERROR COSTS
In short, concerns about the administrability of the essential facilities doctrine, about proper information aggregation by courts, and error costs in decision-making are exaggerated and ignore the costs associated with the “market-based” alternatives. To begin with arguments based on the alleged unadministrability, the essential facilities doctrine is an exception — by design. It deviates from the fundamental principle that businesses are free to deal with whomever they please or refrain from doing so. As such, the essential facilities doctrine necessarily will remain limited to discrete cases. The lack of an all-encompassing theory is unavoidable. To increase the level of certainty, the application of the doctrine could be tied to hard(er) criteria or presumptions. More generally, it appears at least curious that concerns about the vagueness of concepts seem to cause little concerns in areas that are arguable more decisive for the future of the country.
Reservations based on courts difficulties in aggregating information and, potentially related, error costs simply lack empirical evidence. Even if one were to accept the allegations of mistakes in individual decisions, these remain anecdotal and fail to account for (potential) false negatives. And again, there are ways and means address these concerns: one might increase sector specific expertise in government agencies and courts, for example. Similarly, if the remedy is perceived to come to slowly, one might accelerate the processes.
IV. ENTRENCHING MONOPOLY POWER AND THE POLITICAL ECONOMY OF REMEDIES
The essential facilities doctrine was originally conceived as a remedy that is less intrusive of property rights than divestitures. While that was undoubtedly true in Terminal Railroad Association — the Court effectively gave the defendant a choice between granting access rights and a breakup — it not necessarily always true.
It is undoubtedly true that the essential facilities doctrine does not reestablish competition at the level of the facility. Yet, the remedy does also not entrench monopoly power, as some critics suggest. Other than some regulatory regimes, it does not burden nascent competitors. It only limits the incumbent’s earning and differentiation potential relative to that of competitors.
Overall, much of the essential facilities doctrine’s critique reflects unfounded myths, based on anecdotal evidence that ignores the hypothetical alternatives. It is time to leave outdated theories behind and replace them with evidence-based insights and value-based visions for the digital economy.
[1] Assistant Professor of Law, University of Houston Law Center, with courtesy appointment in the University of Houston Cullen College of Engineering’s Electrical and Computer Engineering Department and Affiliated Fellow, Yale Information Society Project. This work is derived from an Article originally published in Yale Journal of Law & Technology. See Nikolas Guggenberger, The Essential Facilities Doctrine in the Digital Economy: Dispelling Persistent Myths, 23 Yale J. L. & Tech. 301 (2021).
[2] https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3802559.
[3] https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3703361.
