Antitrust and policy circles are abuzz with calls against the power, monopoly and other, held by large firms. If one is worried about a firm being too large, then the solution seems obvious: break up the firm and surely that will solve the problem. Any attempt to break up large firms is likely to face many practical and legal challenges. In this article I ask whether breakup is the correct answer, even if these challenges can be overcome.

By Aviv Nevo1


In many policy circles big tech companies, usually meant to refer to Google, Amazon, Apple, and Facebook, are viewed as having too much power.2 Despite such concerns, consumers continue to shop on Amazon, search using Google, connect with their friends using Facebook and Instagram, and purchase Apple devices and services. Consumers’ continued support of these companies could be an indication that the products and services offered by big tech have benefited consumers and are superior to alternatives, or they could be an indication that consumers are locked into choices and effectively denied alternative options and the benefits of healthy competition.

Critics of big tech clearly think it is the latter and have proposed solving the problem by making big tech somewhat smaller. This involves limiting future acquisitions as well as potentially breaking up existing business units. For example, some have called for separating Amazon the retailer from Amazon the marketplace, to undo Facebook’s acquisition of Instagram and WhatsApp, and to separate Google’s search engine from Google’s other lines of business. At some level, we can think of this as taking the “big” out of big tech. Seems like an obvious way to proceed. But is it?

For one thing any attempt at breaking up big tech will likely face legal headwinds. There is some historical precedent in breaking up large firms (and trusts) – consider Standard Oil and the Bell System. Yet more modern attempts of breaking up large companies have been less successful. The lengthy investigation and litigation of Microsoft in the 1990s and early 2000s, first by the FTC and then the DOJ, sought to break up the company. Despite the case generally having been considered a win for the antitrust agencies (and for the economy), Microsoft was not broken up. Interestingly, some have claimed that the win in that case is what allowed the current generation of big tech firms to thrive.3 Hearing these claims, one cannot avoid linking tech to Hydra, the serpent-like monster in Greek mythology that grew two new heads every time one of its heads was cut off.

Given these difficulties, we should ask if breaking up companies is solving the real problems we are concerned with. For example, some of the concerns raised with Facebook is that it is used to spread misinformation. Similarly, concerns have been raised with Google’s and Amazon’s control (and alleged abuse) of consumer data. The hope is that breaking up big tech will lead to more competition and that competition will solve these problems. Neither of these is guaranteed.

To have the hope of increasing competition one must break up a firm into viable business units. In the breakup of Standard Oil and AT&T there seemed to be natural ways to do so. This need not always be the case. For example, some of the discussions of breaking up Microsoft involved separating the operating system from other parts of the business, such as the web browser. From the consumer’s point-of-view such separation might make sense, but it is less clear how to deal with various implementation issues. For example, who gets to keep what assets? Where does top management go? How do you ensure that the pieces into which a company is broken up will constitute viable business units?

In some cases, the goal of antitrust policy is to ensure that a dominant firm does not abuse its position. For example, the Microsoft case aimed to ensure that Microsoft does not use its position in operating systems to harm competition in the browser market. With the benefit of hindsight, it seems like generating competition for Microsoft’s browser did not require breaking up Microsoft. Interestingly, Google, the firm that has come to lead in this browser market, is now a target of antitrust concern.

Even if breakup leads to additional competition, it is not clear that this competition will have the desired effect. Competitive markets are generally good at reducing prices and increasing variety, quality, and output. This is the case because firms compete to attract consumers through lower prices, higher quality, and better service. However, competition for consumers can sometimes lead to unexpected outcomes. Suppose, for example, that we are worried about privacy and the misuse of consumer data. One can imagine a world in which competition improves privacy policies, especially if consumers view privacy as a desired attribute. If consumers are more likely to adopt a platform that has more stringent privacy policies, then competition will encourage firms to improve their policy standards.

However, consider a world in which consumers do not care about privacy, or care about it very little. Privacy and misuse of data could get worse with competition. Since consumers care little about privacy, the platforms will have little incentive to improve it to attract consumers. Furthermore, suppose the platforms compete for advertisers and therefore when faced with greater competitive pressure the firms might have an incentive to collect more detailed consumer data and find new and innovative ways to attract advertising dollars. In this world, it is the competition for advertising that drives the incentive for collecting, and potentially abusing, consumer data. Therefore, breaking up the platforms is unlikely to address concerns regarding privacy and indeed might make things worse.

Similarly, it is not clear that having multiple platforms will create less misinformation. Generally, monopolists reduce output relative to more competitive outcomes. Following this logic, competition can lead to more output, and if output is considered “bad” then competition can potentially lead to adverse outcomes. Therefore, one might claim that less competition is better when trying to restrict “bad” output. Taking the argument seriously – that a loss of competition is good when the firms create a product that is “bad” – has some unintended implications, and illustrates that one needs to be careful when applying the logic of competition and antitrust to somewhat unique non-traditional aspects For example, should we allow cigarette manufacturers to merge so they raise prices and reduce output? Or should we allow two polluting plants to merge so they reduce output and reduce pollution? In both these cases the answer depends on the goal. If the focus is enforcing competition laws the answer seems clear: we should not allow these mergers if they involve a “loss of competition.” Policy makers that want to reduce smoking and pollution, have other policy tools at their disposal. Competition policy should focus on competition issues. In other words, if we break up a company, we should be clear that it is for the right reasons. Breaking up is not a catch all solution, and indeed as the above discussion shows might not be a solution at all.

While many policy makers are concerned with acquisitions of nascent competitors and contemplating difficult if not impossible breakups, U.S. courts are letting acquisitions of potential competitors move ahead. A case in point is the attempted acquisition of Farelogix, an industry disrupter, by Sabre, a dominant firm in the industry.4> The U.S. District Court found that Sabre and Farelogix view each other as competitors and have competed in the past.5 One would think that this would lead to a finding that the merger leads to a “loss of competition” and therefore violates Section 7 of the Clayton Act. This was not the case. The Court ruled that the merger can proceed, because of reading the Supreme Court Amex decision as saying that a (one-sided) firm (like Farelogix) cannot compete, as a matter of law, with a two-sided platform (such as Sabre). Luckily in this case, CMA saved the day by blocking the merger, at which point the parties abandoned the transaction.6

Some have proposed breaking up firms, or after-the-fact scrutiny as being sometimes necessary as a solution to the uncertainty inherent in the merger review process, especially in mergers involving nascent competitors.7 In extreme cases, after-the fact scrutiny might indeed be needed and justified. However, this should be limited to extreme cases and based on careful review of the evidence. Evidence that is particularly hard to evaluate is market outcomes. One might be tempted to use ex post success of an acquired firm as evidence of the potential for competition that existed pre-merger. This type of evidence is potentially problematic since the ex post success of the acquired firm could be evidence of competition that would potentially have been lost but-for the merger. Alternately, it could be evidence of the efficiencies, including synergies in investment and R&D, generated by the merger.

Punishing firms ex post by breaking them up merely because the acquired firm ends up being successful will decrease incentives to invest in acquired firms and realize potential efficiencies. This is especially problematic when at the same time we are also worried about “killer acquisitions,” namely, acquisitions where the acquiring firm acquires a potential competitor and shuts them down. Punishing firms for the acquired firm being too successful and for not being successful enough would create the wrong incentives. An active policy of breaking up firms after deals have been cleared would create unneeded uncertainty, especially if the decision to break up will vary with the ideology of the political party in charge.

The discussion above leaves a grim view on the possibility of breakup or its usefulness as a policy tool. However, the threat of breakup can be a very powerful tool. Current policy makers should take a page out of Teddy Roosevelt’s book. Teddy Roosevelt is widely credited with invigorating antitrust policy by aggressive use of the Sherman Act and energizing a progressive movement that led to the Clayton Act and the formation of the FTC. It was this movement that eventually led to Supreme Court decision to break up Standard Oil in 1911.

There is much that we can learn from Roosevelt’s approach, but I will point to a pillar of his foreign policy. While still Vice President he is believed to have said: “Speak softly and carry a big stick, and you will go far.” These are words of wisdom that current policy makes would be wise to consider. The threat of breakup is a threat that no CEO or company president will take lightly, and therefore would be willing to go to great length to avoid. This gives the antitrust agencies great leverage. The smart move on their part would be to use this leverage to get targeted settlements to well specified problems. Such settlements would not be “sexy” and would likely draw broad critique from politicians and back seat drivers. However, given the current legal landscape and the many years litigation would take, the FTC and DOJ should consider what the threat of breakup can get them at the negotiation table. In the current environment and given how (risk) adverse most CEOs are to the possibility of breakup, the threat of breakup is likely going to yield a more favorable outcome than actual litigation.

As the saying goes, “A clever man gets out of a situation which a wise man never gets into in the first place.” Policy makers and antitrust agencies need to be clever to find a way out of the current situation, but they also need to be wise and forward looking and anticipate the problems of the future. Breaking up firms should be a last resort and should be used only if one is sure that it is likely to solve the problem and not generate problems in the future. In other words, that it is both a clever and wise solution.

1 I am the George A. Weiss and Lydia Bravo Weiss Penn Integrates Knowledge Professor at the University of Pennsylvania with appointments in the Wharton School and the Economics Department in the School of Arts and Sciences. Disclosure: I testified in 2018 on behalf of the Federal Trade Commission in its successful challenge of Wilhelmsen Ship Services attempted acquisition of Drew Marine, and in 2020 on behalf of the Department of Justice in the attempt by Sabre to acquire Farelogix. I testified on behalf of Qualcomm in defense of allegations by the FTC and the Korean FTC of Qualcomm’s licensing practices and was scheduled to testify on similar issues in a matter brought by Apple that settled. In the past three years, I have provided antitrust advice to Facebook. I have also, in the past 3 years, provided advice to roughly ten other clients on matters unrelated to the topic of this article.

2 See for example claiming that “[t]oday’s big tech companies have too much power—too much power over our economy, our society, and our democracy.”

3 See for example a speech given by Makan Delrahim, at the time the Assistant Attorney General, (“Although Microsoft was not broken up into smaller companies, the government’s successful monopolization case against Microsoft may very well have paved the way for companies like Google, Yahoo, and Apple to enter with their own desktop and mobile products.”).

4 As disclosed above, I provided testimony on behalf of the DOJ in this case.

5 For example, the Court found that “a preponderance of the evidence shows that Sabre and Farelogix do view each other as competitors”; “Sabre considers Farelogix a competitor in developing NDC technology for direct connects,”; “the record reflects competition between Sabre’s and Farelogix’s direct connect solution for major airlines,”; “Farelogix identified Sabre as a ‘key competitor’ in order delivery and offer management,” US v. Sabre Corp., et al., 1:19-cv-01548-LPS, April 8, 2020, p. 31-32.

6 For further discussion see Kostis Hatzitaskos, Brad Howells & Aviv Nevo, “A Tale of Two Sides: Sabre/Farelogix in the United States and the UK,” Journal of European Competition Law & Practice, June 2021; and Lindsey Edwards & Jonathan Jacobson, “Missing the Forest for the Trees: The Application of Amex in United States v. Sabre,” The Antitrust Source, June 2021.

7 See C. Scott Hemphill & Tim Wu, “Nascent Competitors,” 168 University of Pennsylvania Law Review 1879, 2020.