Are broken promises an antitrust violation?

That is the question that Dina Srinivasan answers in the affirmative in her paper “The Antitrust Case Against Facebook.” This is an interesting set of issues because, frankly, my observation is that Facebook, while having a dominant position in social media (which is not an antitrust violation) had not violated antitrust law getting to that position and did not appear to be violating antitrust law when it got there.

Srinivasan’s high-level argument is novel. Broadly speaking she argues that Facebook, in its infancy, made a bunch of promises with respect to privacy and other related features that it subsequently reneged on when it ‘won’ the market. As an early user of Facebook, I have to say that I didn’t read these things as promises, but let’s suppose, for the sake of argument, that they did break a clear promise. Breaking promises are covered by other aspects of the law in many countries. How might they be an antitrust violation?

The argument is that social media markets are in a class of markets characterised by network effects. That means that when a new social media platform is starting out, in order to compete, it needs to make all manner of promises to get consumers to adopt. For instance, it may promise to be free. It may promise never to have ads. Or it may promise to do various things related to privacy protection. These promises are actually quite important in being able to build up market share — and in network industries, this matters. If consumers are sophisticated, they will worry about joining a platform that will change its tune should it become dominant. So a platform that is making promises not to change in those ways, so long as it is credible, will have an advantage in attracting consumers.

A key part of this argument is that competition matters for the provision of things consumers value, like privacy. I think that as a statement of basic economics that is correct. If consumers value privacy, then the more competitive the market, the more privacy they are likely to get with products that can supply it. Should a firm become dominant, if they gain more from reducing privacy than they would lose in consumers, they may well make that choice. In the end, like most situations, consumers end up with better quality products if there is competition and that goes for privacy characteristics too. You only have to look at Apple’s positioning relative to Google and others to see that. (Although I might add here that if you buy that argument, you can’t claim at the same time that Apple is not subject to competitive pressures).

The point here is that it is entirely plausible that Facebook today is less constrained by consumers concerns for privacy than it was when it was starting out. But is that an antitrust issue. After all, before firms become monopolists, they have to price lower and then when they become monopolists, they price higher. That is the cost of monopoly but it is also the wholly expected outcome from monopoly. Wouldn’t the same apply to product quality?

Srinivasan’s answer to this is that because this is a network industry, things are different. In those industries, initial competition is for the market, which means that promises can shape that competition and consumers are harmed and have few options when those promises aren’t kept. If antitrust law were covering this, what would happen is that Facebook could be forced by that law to keep its promises today. In other words, antitrust law could be a route to ex post regulation of Facebook.

What is unusual about this argument is that it is not one that leads to a remedy that leads to more competition. I guess one argument could be that the only way for consumers to get what is promised is by breaking up Facebook. But even then, would any of the mini-Facebooks switch to a privacy outcome? It is possible. Whatsapp is pretty private but at the same time was the subject to a huge amount of election misrepresentation in Brazil so it is not at all clear that isn’t some violation of another promise.

More to the point, we have other laws to deal with broken promises. We have consumer protection laws and also contract law. For instance, I suspect that if Facebook broke its long-standing promise not to charge consumers for access, lawyers would not hesitate in pursuing a breach of contract claim. And I have to say, contract law seems much ‘easier’ to deal with than antitrust law.

There are other reasons to worry about using antitrust law in this way. Importantly, startups have to do all manner of things. They may make promises and then things change and they have to change as well. To be sure, they face market constraints on that but even Facebook has paid a market price for its behaviour — even if it isn’t as large as some might like. You do not want to do things that constrain experimentation when firms are young.

That said, antitrust law may be appropriately used when monopolists make promises and then break them as monopolies. Deceptive conduct is amplified when it is conducted by monopolists especially as an attempt to forestall or block competitors. For young firms, they cannot use promises to block competitors in this way. As Srinivasan says, it is a competitive tool and an appropriate use and outcome of competition.

My opinion remains that Facebook has been harmed on numerous fronts by the lack of competition they face. Regulation is coming to deal with broad issues such as privacy and that is likely far more preferable than using antitrust. That said, network effects are a choice for Facebook. If Facebook were to take actions that would reduce or remove them, it could then use arguments such as Srinivasan’s to make the case that other regulation is not needed. In other words, regulation is a probability because we ended up without a competitive market.