In this edition of CPI Talks we have the pleasure of speaking with Catherine Tucker, the Sloan Distinguished Professor of Management and a Professor of Marketing at MIT Sloan. She is also Chair of the MIT Sloan PhD Program.
Thank you, Professor Tucker, for taking this time to talk to CPI.
1. Is the breaking up of a company ever viable as an antitrust remedy? What has the historical experience taught us thus far?
As an undergraduate in economics, I was reared on stories of breaking up natural monopolies. Indeed, the industrial organization of the 1980s and 1990s was focused on questions of whether you could take a natural monopoly and make it more efficient by breaking it up.
The economics underlying breakup remedies, when it came to the apparent natural monopolies of the 1980s, was reasonably straightforward. The argument was that there were natural economies of scale in erecting and maintaining phone lines, for example. This implied that the existing incumbent would always have a sustained competitive advantage over an entrant.
Typically, the rationale now given for the breakup of platforms is that there are network effects that lead to a natural monopoly. I have written extensively why this viewpoint does not reflect the empirical literature in economics on the nature and scope of network effects.
However, even if one ignores this empirical literature, and tries to make the case that network effects do lead to natural monopoly, the underlying nature of the economics of that argument means that breakup is not an attractive remedy.
2. What about the welfare effects of breakup remedies? Are breakup remedies effective at improving consumer welfare?
The key difference between network effects and economies of scale is that at their essence, network effects are a demand-side phenomenon. Network effects enter directly into a prospective platform users’ utility function. Indeed, when I teach entrepreneurs how to build platforms, I get them to mentally model the likelihood of their acquiring a customer as being a function of the network effects they can create.
By contrast, economies of scale do not directly enter into a consumer’s utility function. Instead, they allow a producer to supply a good at a cheaper cost than its rivals. Much of antitrust debate is focused on the question of how much of that cost advantage benefits the consumer in terms of lower prices, relative to benefiting the firm in terms of higher profit margins. However, this economic ambiguity of how the gains of lower costs are split between consumer and firm does allow for debate. It also means that a regulator intent on breaking up a firm to limit economies of scale could perhaps argue that consumers could in theory still benefit, because if competition leads to lower profit margins, then this could counteract naturally higher costs and still lead to lower prices.
By contrast, because network effects directly enter into a platform’s utility function, there is no ambiguity that a breakup focused on curtailing network effects would diminish platform users’ utility directly. And there is no clear argument about lower prices that would help temper this direct reduction in utility. First, many users of platforms pay a zero price or are subsidized. Second, the reason a user-group on one side of the platform is typically subsidized is because users on the other side of the platform benefit asymmetrically from their presence. If we restrict a platform’s ability to construct network effects, then the rationale for a cross-subsidy disappears. So that means that not only do you have a reduction in welfare from the direct eradication of network effects, but you extinguish one rationale for a mutually beneficial cross-subsidy. In other words, if a regulator seriously believes in the argument that network effects lead users to be persistently attracted to a particular platform, then any breakup remedy designed to curtail platforms’ network effects will negatively affect consumer welfare.
3. Could breakup remedies be effective at increasing competition between so-called “digital platforms”?
There are many reasons to think that network effects are localized and fragile in nature and unlikely to give sustained competitive advantage. But let us imagine a two-sided platform where regulators have managed convincingly to demonstrate empirically that the network effects of that platform are both global and persistent in a way which led to a winner-takes-all market. What would that imply for the likelihood of a breakup improving competition?
Well, in some sense, it seems almost a trite question to ask. If you seriously believe that network effects are persistent and global enough to lead to a winner-takes-all market, then the effects of any breakup will be temporary. All that will happen is you take the ‘Winning’ platform, and temporarily split it into two platforms that compete. However, the same network effects that drove the decision to break up the platform will also lead to the re- emergence of a single winner, and the stronger the network effects are that the regulator is trying to counteract, the quicker that will happen.