While the FTC’s lawsuit against Facebook is unlikely to lead to a “breakup,” it could force Facebook to undo some mergers. Breaking things up is easy, but breaking them up without harming consumers, labor, or other suppliers is not. How does Facebook divest Instagram and WhatsApp in a way that does not harm consumers and others?
The US Federal Trade Commission has filed a major antitrust action against Facebook. The lawsuit has been characterized as seeking a “breakup,” but the FTC is actually requesting a divestiture of firms that Facebook has acquired within the last decade. These include Instagram and WhatsApp, which Facebook has partially integrated into its platform.
The FTC also wants an injunction against Facebook policies that prohibit third-party app developers for Facebook from either providing the same core functions that Facebook offers or interconnecting with Facebook’s competitors. This part of the complaint is reminiscent of the complaint against Microsoft twenty years earlier, where Microsoft required developers of Windows applications to favor Microsoft’s own browser, Internet Explorer, over rival Netscape. If this case is appealed, it will go to the same court that decided Microsoft.
Antitrust’s purpose is to promote practices that lead to the highest output consistent with sustainable competition. High output benefits consumers by providing low prices. It also benefits labor and suppliers, who are better off as markets produce more. An antitrust remedy should be consistent with this principle. Breaking things up is easy, but breaking them up without harming consumers, labor, or other suppliers is not.
The requested injunction against Facebook’s policies limiting developer competition is well within the range of conservative antitrust remedies. Two issues likely to be litigated are market power and justifications for Facebook’s exclusion rules. The Microsoft decision readily condemned many of Microsoft’s exclusionary provisions where Microsoft controlled 90 percent of the relevant market. Here, the FTC alleges that Facebook controls around 60 percent of a market for “personal social networking” services. This number is at the lower end of the range where the courts condemn monopolization.
Interestingly, a 60 percent market share is more than sufficient to condemn contracts that contain restrictive agreements and Facebook has apparently imposed many of them. But the FTC did not challenge any agreements. That share is also sufficient for the statutory offense of “attempt to monopolize,” and the facts in this case seem to justify that claim as well. However, the FTC did not claim an attempt to monopolize either. The FTC can amend its complaint at least once, so it may add these claims later on. In any event, a strong case can be made that market power can be found on lower market shares in networked industries that become more valuable as the number of users increases—a phenomenon referred to as positive network externalities. Facebook’s current size gives it a significant advantage over smaller rivals.
Another likely market power issue will be whether the market for “personal social networking” is too gerrymandered to describe a grouping of sales realistically capable of being monopolized. This definition excludes more specialized networking services, such as LinkedIn, and also services that offer a narrower range of features, such as YouTube, Tik Tok, and dedicated messaging services. Digital platform services are highly differentiated, and this raises a problem in any antitrust case where a relevant market must be defined. Market definition is always “wrong” in differentiated markets. For example, should a percolator and a French press be put into the same market for “coffee makers”? Putting them into the same market treats them as perfect competitors, such that customers are indifferent between them. Clearly that is not true. But putting them in different markets treats them as if they do not compete at all, which is also not true. You can say the same thing about Facebook vs. YouTube; or Facebook vs. LinkedIn.
All of this is to say that proof of market power through use of a “relevant market” works very poorly in a differentiated market such as social networking. However, an empirical renaissance in the study of differentiated markets has enabled economists to develop more reliable methods for assessing power. These involve direct determination of changes in output or demand in response to a given price change. While these methods are more accurate, they are also more technical and somewhat less intuitive. They are also highly dependent on the availability of data. In this case, however, the relevant transactions are digitized and obtaining the relevant data should be easy.
The other question about the exclusionary contract provisions concerns reasonable justifications. Are they legitimate attempts to limit free riding or intellectual property violations by rivals? While Facebook has a recognized interest in protecting its IP rights, it cannot suppress all competition simply by labeling it “free-riding.” It is one thing for Facebook to sue an app developer or rival for patent or copyright infringement; it is quite another to exclude all competition on the rationale that some undetermined part of it may be infringing.
The FTC also wants Facebook to spin off Instagram and WhatsApp. The challenge is based on the theory that these mergers were exclusion strategies, intended to prevent the emergence of rivals. That strategy seems well documented by Facebook emails that the complaint quotes. This differs from the typical rationale for condemning mergers, which is fear of higher prices. While the FTC has the authority to enforce the Clayton Act’s merger provision, this complaint relies exclusively on the Sherman act’s prohibition of monopolistic practices. In general, evidence of intent is not the best kind in such cases, but here the evidence is particularly strong that Facebook’s purpose in these acquisitions was to prevent the rise of new competitors.
The merger claims do face some challenges, however. First, I do not believe that the FTC will be hindered by the fact that these mergers occurred in 2012 (Instagram) and 2014 (WhatsApp) without government objection. There was no earlier proceeding in which they were parties. As a result, the FTC is not technically bound by an earlier determination. While there is a doctrine of “laches” that precludes stale challenges, it ordinarily does not apply to government actions.
In any event, to the extent that either Facebook or these two firms are bigger than they were at the time of the acquisitions, or pose more significant threats to competition, the FTC should be permitted to go ahead. The courts have been pretty clear that anticompetitive effects that are apparent at the time of the lawsuit can be challenged even if they were not yet apparent previously, when the merger occurred. That is particularly true in this case because neither Instagram nor WhatsApp has been fully integrated into Facebook. They still maintain separate platforms and subscription portals, although they do share some data.
The technical questions associated with spinoffs are more difficult. How does Facebook divest these two firms in a way that does not harm consumers and others? According to the complaint, one purpose of these particular acquisitions was to facilitate Facebook’s expansion from a desktop/laptop platform into mobile devices. Creating a regime in which Facebook is optimized on large devices and Instagram and WhatsApp are separate programs that operate on small ones will not benefit anyone. Users have come to expect an integrated Facebook on which their accounts synch and migrate seamlessly between computers and mobile phones. And what about customers? The court cannot order customers to move from one platform to another.
Eventually after a breakup, each platform may start to compete with the others. Nevertheless, any breakup of this kind will reduce the range of network externalities, which is what makes social networking platforms so valuable. For example, currently Facebook shares some interconnection and data with both Instagram and Facebook. Severing these links will harm users. The more people that are on a platform, the greater its value, and the more services it aggregates and synchs across all available platforms, the more people it will attract. What remedy will avoid destroying the value of these firms to consumers?
A solution may actually be within reach. If we force divestiture of Instagram and WhatsApp, each of the three firms should receive non-exclusive licenses to one another’s technology, along with shared access to the customer database, although with customer rights to opt out. For customers this would mean that each could communicate seamlessly, including the sharing of all data, with customers from the other two. Postings on one would be searchable and accessible on the others. This is, in fact, how the AT&T phone company breakup was structured forty years ago, including mandatory cross-licensing of IP rights and directory access. Today, the phone system we have is a unitary network that maximizes access by more-or-less enabling everyone to call everyone else. Even so, hundreds of firms compete on this network. Email works in fundamentally the same way. Although there are dozens of providers, or clients, users on one can communicate with users on the other, while making individual messages as secure or as open as they wish them to be.
Future acquisitions are easier to address than completed acquisitions. First, they should require the FTC’s approval, as the complaint requests. Second, in most cases that raise issues similar to those with Instagram or WhatsApp, the acquisition should be limited to a non-exclusive license and, where appropriate, information sharing. That would enable Facebook to take advantage of all of the technology that the smaller firm has to offer, but without the power to exclude others.
For an expanded version of the arguments offered here, see Antitrust and Platform Monopoly, 130 Yale L.J. (2020) (forthcoming), available here.