Susie Kim

In March 2019, Senator Warren unveiled an ambitious antitrust proposal: breaking up Amazon, Google, and Facebook.[1] She focused on the most controversial mergers such as Amazon’s acquisition of Whole Foods, Facebook’s acquisition of WhatsApp and Instagram, and Google’s purchase of Waze and DoubleClick. This post will analyze separately the two prongs of Senator Warren’s proposal: (1) designating large tech platforms as “Platform Utilities” and (2) appointing regulators to unwind mergers.

Platform Utilities as a Firewall

Senator Warren’s proposal to designate big tech platforms as platform utilities is a new take on an old model. Previously applied to the railroads, telecommunications, and investment banks, the utility model is a firewall of sorts which holds that the entity (and its power over the respective technology) is so paramount that they must operate under a higher standard relative to other monopolists. The genesis of this model begins with AT&T and the Roosevelt administration, where AT&T became a legal monopoly in the telecommunication sector and agreed to supervision by government regulators.[2]

Firewalls make sense when there is a need for the entity to dominate a market. In economic terms, such a firm could be considered a natural monopoly whereby the market demands that a single firm supply the high fixed or startup cost for the industry. Both the railroad and telecommunications industry required high startup costs such as laying down tracks and setting up wiring respectively. Tech platforms could be considered as the 21st century example of natural monopoly industries given that they exhibit similar characteristics such as network effects. Interestingly, the startup cost to creating a platform is miniscule so it is not a natural monopoly in the traditional economic sense. However it makes sense that there are synergies associated with one platform catering to all users because it makes the experience better through more reviews, more product choice, and increases utility to each member. To consider the last point, many users utilize tech platforms through word of mouth as opposed to traditional advertising, because for example Facebook is more useful the more of your friends are on the platform.

The biggest question on firewalls is their long-term viability. Under the Glass-Steagall Act, investment and commercial banks were separate entities.[3] Despite this, investment banks were able to circumvent the firewall and invest in traditional commercial lending through the rise of shadow banking and securities.[4] Limiting the ability of tech platforms to compete directly with their suppliers will require a finesse in outlining how much, if any, ownership tech platforms are allowed in subsidiaries that may compete on the platform. Additionally, a well-reasoned plan would keep a close eye on enforceability and the relative ease or hardship in regulating the platforms.

The last consideration on platform utilities is the $25 billion revenue line defining which tech companies will be considered as platform utilities.[5] Although having a straight-forward measurement makes it easy to categorize tech companies, it runs the risk of being too simple. There is a discussion around whether other indicators may be more useful, or at least helpful, in defining a platform such as the number of users, network with other third parties and subsidiaries, control of available advertising dollars, and control of key labor supply.

Unwinding Mergers

Senator Warren’s plan to create a new enforcement agency that would focus specifically on unwinding mergers is interesting to say the least. The practice of retroactively unwinding mergers will spark caution and uncertainty among businesses and presumably dim merger activity. To this end, any prospective enforcement agency or authority should take care to issue specific guidelines around which mergers will be up for evaluation. One way would be to cooperate with the other antitrust enforcement agencies such as the FTC and DOJ to issue a “conditional merger” for mergers that are suspect, which would outline a period of time within which the unwinding enforcement agency would monitor the behavior of the merged entity. This would predictably increase the cost of mergers intended for an anticompetitive purpose as they would not be allowed to scramble the eggs throughout that period.

Senator Warren’s plan also includes, fascinatingly, the creation of a separate new agency that will focus on retroactively undoing previous mergers. The need for such a separate entity is questionable. Creating a new regulatory body makes sense if one is concerned about the ability of the existing enforcement agencies (DOJ, FTC, and the State AGs) to effectively untangle previously approved mergers. First, an agency could be in the awkward position to break up a merger that it has already approved of, and may wish to avoid doing so in order to protect the prior decision. Second, even if it is not the same agency that both approved and then broke up a merger, there could still be the issue where peer enforcement agencies are reluctant to step on each other’s toes – again leading to a situation of under enforcement. A theoretical split between the agencies who approve of mergers and an agency that breaks up previously ordained mergers may make for clearer incentives. On the downside, the new agency would require resources and potentially end up as another agency with overlapping responsibilities in the antitrust context.


The relationship between antitrust laws and politics are not a recent phenomenon. While Senator Warren may be motivated, in part, to send a strong signal to her supporters of her progressive agenda, her aggressive proposal does bring a new flavor to the antitrust debate.

The desire to regulate winners under a utility model is not a new idea in the United States. In fact, AT&T was the first telecommunications monopolist to willingly subject itself to government regulation (in return for a legally protected monopoly). The question is not if Senator Warren’s plan is feasible, but whether it makes sense from an effectiveness standpoint. The regulation of monopolies under a utility model will require a well-informed and sharply defined regulatory body that will resist regulatory capture.

Following the Hart-Scott-Rodino (HSR) Act, there was a definite move towards a preemptive merger-notification regime. It is unclear, if Senator Warren’s proposal to create a retroactive enforcement agency were to succeed in the legislature, how much the proposal would impact the pre-merger notification regime. A narrowly tailored scope for specific problematic mergers may help in ensuring there is a bite to behavioral remedies. This balance would have to be carefully defined to ensure a fair and clear world that would allow for merger activity while protecting against anticompetitive consolidation.

[1] Elizabeth Warren, Here’s how we can break up Big Tech, Medium (Mar. 8, 2019),

[2] Joseph D. Kearney & Thomas W. Merrill, The Great Transformation of Regulated Industries Law, 98 Colum. L. Rev. 1323, 1329-30 (1998).

[3] Arthur E. J. Wilmarth, The Road to Repeal of the Glass-Steagall Act, 17 Wake Forest J. Bus.& Intell. Prop. L. 441, 449 (2017).

[4] Id. at 458-60.

[5] Supra note 1.