Essential Facilities in Tech: A Poor Match
Many tech commentators have, whether they realize it or not, invoked the “essential facilities doctrine” as a foundational argument for why we need to be concerned about companies with high market share. You can spot these arguments when a critic uses language such as a company “needs” something from another company to do business. While it might be helpful to think about broad questions of the tech ecosystem, my previous post explains that essential facilities means something specific when it comes to competition laws.
Essential facilities is basically a narrow exception to the rule that a company has no duty to deal with its rivals. If the activity described does not fit into this exception, then no duty exists － meaning companies are free to deal or not deal with other companies as they choose. This of course is not an antitrust immunity, it just means that the mandatory access remedy is unavailable to plaintiffs if a true antitrust violation has occurred. For example, if the Terminal Railroad case, which involved the acquisition of a monopoly of the shipping infrastructure around St. Louis, was decided without reaching a mandatory access remedy, then the court would simply have ordered a divestiture of vital infrastructure.
Defining “Essential” as a Traditional Two Step Test
The essential facilities doctrine is somewhat unusual in that, although it describes the conditions by which a mandatory access remedy may be granted, it is also closely intertwined with the conduct at issue. In other words, a broad reading of the essential facilities doctrine would suggest that owning an essential facility is the sin that justifies a court to abridge the ordinary freedom of dealing. This kind of reading encourages wide application, but as the Supreme Court states, forced sharing is an “uncertain virtue.” It is essentially the seizure of a private asset for the public good, and the Supreme Court has “been very cautious in recognizing such exceptions” to the right to deal freely.
The Trinko court seems to ground the mandatory access remedy in traditional antitrust terms by framing it not as a one step process, that the facility is essential, but a more traditional two step process, that the facility is essential and there is conduct that can only be explained by a desire to achieve anticompetitive ends. Looking at alleged violations through the traditional monopoly power plus anticompetitive conduct structure is a helpful way for courts to approach unusual circumstances. For example, the Second Circuit in New York v. Actavis followed this structure in setting a remedy that forced Actavis to continue to offer its original Alzheimer’s drug so that patients would have the opportunity to substitute the lower cost generic version once that hit the market. The Second Circuit focused on the conduct: Actavis had planned to force patients to switch from the original version to a slightly different version without a legitimate business justification, and the most convincing explanation was that they were doing it to prevent generic competition. This conduct is called product hopping.
Therefore, the first step is determining the essentiality of the asset. The classic example of an essential facility is infrastructure, like utilities and transportation routes. These are the facilities in the Supreme Court cases of Terminal Railroad and Otter Tail and the Seventh Circuit’s MCI Communications decision. These assets are not only expensive to replicate, they also require a continuous line of easement or other property rights to build. This often means they require government assistance in putting together these rights. Additionally, replicating this infrastructure creates little public good unless the infrastructure is already being used at capacity. It will often instead have negative public externalities, requiring more space and causing disruption during construction. Courts, including the Supreme Court, have found other facilities to have met the essential prong. But the further one moves away from the archetypical case, the weaker the essential argument becomes. And even infrastructure has resisted the application of a mandatory access remedy, as found in Trinko.
Tech platforms do not share many characteristics with infrastructure. It is not overly expensive to replicate the software (although most platforms do face a challenge in determining the best way to attract a user base). It is also easy in many cases for companies and consumers to bypass platforms. For example, an internet company can use marketing to convince consumers to go directly to a website or it can build an app to encourage consumers to bypass the browser altogether. There also isn’t any government participation in the creation of the asset, like there is when a company needs to acquire rights to build infrastructure. Finally, there aren’t really any negative externalities in either creating a competing asset or bypassing the asset.
The second step is identifying anticompetitive conduct. A lack of a legitimate business justification for demonstrably anticompetitive conduct can help a court decide to use a mandatory access remedy in cases where essentiality might be more difficult to assess, as in product hopping cases like New York v. Actavis. But tech complainants are not likely to get very far here, as even using prime real estate for sponsored results is a legitimate business justification because it generates profits necessary to continue to invest in the platform. Other justifications to change results include responding to consumer demand, improving the quality of results, and to head off gaming by third parties.
However, a plaintiff could succeed with clear instances of unjustified anticompetitive conduct. For example, tricking customers into buying a more expensive product by making them believe it was the least expensive in the search results. Or purposefully misleading the consumer on the source of the product. Or making ads look exactly like organic results to mislead consumers. This type of conduct fits more traditional antitrust complaints, which also means that a more traditional remedy will likely be chosen over a mandatory access remedy.
The MCI Test
The MCI case provides an alternative four step test:
(1) control of the essential facility by a monopolist; (2) a competitor’s inability practically or reasonably to duplicate the essential facility; (3) the denial of the use of the facility to a competitor; and (4) the feasibility of providing the facility.
Applying this test, it seems likely that platforms are not as essential as the archetypical infrastructure facility, especially since competitors can often bypass the platform through available mechanisms like creating phone apps or doing their own marketing. Step 3 is difficult to make a case for, because often the complainant is not denied access to the platform but only denied access to the choicest real estate on the platform. This also creates problems with step 4: How is a court to determine that the complainant should have that space over all other choices? This is a difficult decision for a court to make, and they simply do not have the institutional expertise to step into the shoes of a tech business and decide how choices should be displayed on a results screen.
Based on these considerations, it is unlikely that essential facilities claims will find support using the MCI test.
Much of the literature on essential facilities － for or against － describes it as a narrow remedy, and for good reason. A mandatory access remedy is awkward, and most seem to agree that its overuse would be a net negative. The tech industry provides some good examples of the concerns expressed over the years of the essential facilities doctrine.
The Supreme Court found the concern over the administrability of the essential facilities doctrine to be the most persuasive in its Trinko decision. There the Court stated “[w]e think that Professor Areeda got it exactly right: ‘No court should impose a duty to deal that it cannot explain or adequately and reasonably supervise.’” Mandatory access remedies in tech that sound simple are actually very difficult for a court to design and supervise.
For example, take the argument that a company needs to be on the front page of a search. How is a court to decide how that space is to be allocated? Tech companies regularly update their search algorithms to increase accuracy, respond to consumer preference, deter gaming, and even based on business decisions. A court would have to insert its own judgment into that process, a process that naturally results in winners and losers. The court would also have to have access to the algorithm and continually monitor the search algorithm and make decisions on who gets that real estate. This is not something courts have experience in. The Trinko court puts it this way: “Enforced sharing also requires antitrust courts to act as central planners, identifying the proper price, quantity, and other terms of dealing—a role for which they are ill-suited.”
The other common criticism is that an overburdensome essential facilities doctrine will discourage innovation. A company is less likely to develop a search engine or a platform if they believe that a court will one day dictate how their service is to be run. This would have more of an impact on future innovation, and could have the unintended consequence of actually solidifying current market shares if startups look at what is happening to the incumbents and get discouraged from trying to topple them through heavy investment, especially if they believed that the law would soon force them to provide competitors access to their assets. Again, the Trinko court explains: “Compelling such firms to share the source of their advantage is in some tension with the underlying purpose of antitrust law, since it may lessen the incentive for the monopolist, the rival, or both to invest in those economically beneficial facilities.”
But the unintended consequences could extend beyond this issue. A mandatory access remedy in tech will discourage companies from allowing third parties to participate in their services. In other words, it might at some point make more sense for a company to release a search product that has no organic results or to remove third party sellers from their platform than it would for these companies to deal with burdensome court oversight of their services.
The risk of unintended consequences and lack of true “essentiality” makes the essential facilities doctrine a poor fit for tech. The doctrine today is a narrow exception, and there simply is not a strong enough case that tech should fit into that exception. While critics may claim space on a platform is “needed” to do business, these claims probably have little effect on competition enforcement policy.