Dear readers,
In 2020, I started publishing monthly guest articles written by some of the world’s most renowned antitrust scholars. The series continues in 2021. The one for February is authored by A. Douglas Melamed, Professor of the Practice of Law at Stanford University. In it, Doug explains that the (multiple) acquisitions of nascent competitors by monopolists might violate Section 2. He goes on to show that how to assess them in practice is less clear. I am confident that you will enjoy reading it as much as I did. I am confident that you will enjoy reading it as much as I did. Doug, thank you very much!
All the best, Thibault Schrepel
****
Acquisitions of Nascent Competitors under Section 2 of the Sherman Act
The Federal Trade Commission’s case against Facebook is potentially very important, not just for Facebook, but also for antitrust law. The Commission alleges that Facebook unlawfully maintained its social network monopoly in violation of Section 2 of the Sherman Act by a series of actions, the most important of which were its acquisitions of Instagram in 2012 and What’s App in 2014. That claim raises the important question of whether acquisitions that might not violate Section 7 of the Clayton Act might nevertheless violate Section 2.
Both acquisitions were reviewed but not challenged by the Commission before they were consummated. It is possible that they would be found to violate Section 7 today, perhaps in light of information that was not available to the Commission before they were consummated. That is, however, far from clear.
Before it was acquired by Facebook, Instagram was a relatively small but rapidly growing, differentiated social network that competed with Facebook; in today’s parlance, it would be regarded as a “nascent” competitor. Before its acquisition, What’s App was a messaging service that was thought to be uniquely able to compete with Facebook by extending its messaging service into a broader form of social network. It might be thought of as a nascent competitor, but because the competition jeopardized by the acquisition would have required What’s App to develop fundamentally new service offerings, it is more likely to be regarded as a “potential” competitor. Regardless of the label, case law suggests the possibility that neither acquisition would violate Section 7 unless the court determines that, absent the acquisition, it was more likely than not that the acquired firm would grow into a significant competitor of Facebook in the social network market.1As to existing competitors, see, e.g., FTC v. Arch Coal, Inc., 329 F. Supp. 2d 109, 115 (D.D.C. 2004) (“likely substantially to lessen competition”). As to potential competitors, see, e.g., United States v. Marine Bancorp., 418 U.S. 602, 633 (1974) (“substantial likelihood”); Tenneco, Inc. v. FTC, 689 F.2d 346, 352 (2d Cir. 1982) (“would likely have entered”). It is not clear that a plaintiff could prove that with respect to either acquisition.
This note addresses the question of whether the acquisitions might be found to violate Section 2, even if they do not violate Section 7. There is no question that acquisitions can violate Section 2, but some commentators have argued that the Clayton Act was enacted to strengthen the antitrust laws and that the earlier-enacted Sherman Act therefore cannot be construed to prohibit mergers that would be lawful under Section 7.2E.g., Douglas H. Ginsburg & Koren Wong-Ervin Challenging Consummated Mergers Under Section 2, CPI (2020) That seems wrong for two reasons. First, it is widely understood that the antitrust laws evolve in a common law-like process, so even if the Clayton Act was intended in 1914 to be more aggressive than the Sherman Act as it had been construed by the courts before then, the meaning of the Sherman Act might well have evolved since then to be more far-reaching in some respects than the Clayton Act. Second, even if the Clayton Act treats mergers more aggressively than the Sherman Act in general, the Sherman Act might be more aggressive with respect to acquisitions used as part of a scheme to maintain monopoly power.
The argument that the acquisitions violate Section 2 is straightforward. It starts with the assumption that Facebook has monopoly power in the social network market protected by substantial barriers to entry. The acquisitions prevented Instagram and What’s App from evolving into important competitors that might have materially reduced Facebook’s monopoly power in that market. While it is uncertain whether those firms would have evolved that way, it is clear that the acquisitions eliminated the possibility that one or both would have done so.
I.
The principal legal authority for the argument is the D. C. Circuit’s unanimous en banc decision in United States v. Microsoft.3253 U.S. 34 (D.C. Cir. 2001) (en banc). The court held in that case that Microsoft violated Section 2 by a course of conduct that harmed Netscape’s browser application and thus extinguished a small but realistic likelihood that it would evolve into, or become a key complement for, a competing computer operating system that would reduce Microsoft’s monopoly in the operating system market. Even though there was only an uncertain, multi-step connection between the harm to competing browsers and the maintenance of Microsoft’s operating system monopoly, that connection was sufficient because Microsoft’s conduct “reasonably appear[ed] capable of making a significant contribution to . . . maintaining [its] monopoly power.”4Id. at 79. The requisite causal connection between the conduct at issue and harm to competition can be inferred “when exclusionary conduct is aimed at producers of nascent competitive technologies as well as when it is aimed at producers of established substitutes.”5Id.
This aspect of the Microsoft case seems to mean that a monopolist violates Section 2 when it engages in anticompetitive conduct that reduces the likelihood that its monopoly power will be reduced, regardless of whether a reduction in its market power absent the conduct was more likely than not. If so, one implication might be that a monopolist may not acquire a potential competitor in its monopoly market unless the monopolist can demonstrate (i) that the acquired firm is one of several potential competitors and the acquisition will thus not materially reduce the likelihood that its monopoly power will be reduced or (ii) that there are substantial merger-specific efficiencies sufficient to offset the possible reduction in competition.
Commentators, however, have suggested a number of reasons why the Microsoft case might not lend such support to antitrust challenges to nascent acquisitions. Broadly speaking, the reasons fall into two categories: (1) those regarding the nature of the injury to competition and (2) those regarding the defendant’s conduct.
(1) Commentators have made two related arguments about injury to competition. They note that the court found that the conduct at issue in the Microsoft case did in fact foreclose Netscape from market opportunities and thus prevented it from gaining the critical mass of users necessary in order for it to evolve into a viable operating system or to help others to do so.6Douglas H. Ginsburg & Koren Wong-Ervin, supra n. 2. This argument is buttressed by the D.C. Circuit’s subsequent decision in Rambus v. Federal Trade Commission,7522 F.3d 456 (D.C. Cir. 2008). which the commentators argue holds that Section 2 is violated only when actual harm to competition in the relevant market is proven and that it is not enough to prove only anticompetitive conduct that had the potential to harm competition.8Douglas H. Ginsburg & Koren Wong-Ervin, supra n. 2.
These arguments seem wrong for two reasons. First, Instagram was actually competing in the social network market at the time of the acquisition, and Facebook’s acquisitions of Instagram and What’s App completely eliminated the possibility that either would become a significant competitor of Facebook. The harm to competition in the monopolized, relevant market caused by Facebook’s acquisitions is thus more immediate and more direct than the entirely speculative harm regarding possible new entry caused by Microsoft’s conduct aimed at Netscape. While Microsoft’s conduct was shown to have harmed Netscape in the separate browser market, the court made clear in its later discussion of the attempted monopolization and tying claims that the government had not proven harm to competition in the market as a whole, even in the browser market.9United States v. Microsoft, supra n. 4 at 81-84 (attempted monopolization), 95 (tying).
Second, the Federal Trade Commission sought rehearing en banc of the Rambus decision on the ground that the decision was inconsistent with Microsoft. The petition was denied, and for good reason. Rambus involved alleged misrepresentations by a potential entrant, with no market power, that had the potential to distort a specific decision by a standard-setting body many years earlier. The court ruled for the defendant on the ground that the Commission explicitly did not find that the conduct actually distorted the decision. In substance, the court declined to adopt the unprecedented principle that ordinary business torts that did not harm competition might be found to violate Section 2 on the ground that the conduct might have harmed competition. The Microsoft case and the Federal Trade Commission’s case against Facebook are very different because they involve maintenance of an existing monopoly and because the conduct in those cases, unlike the conduct in Rambus, reduced the likelihood that the monopoly power would be eroded in the future including after the antitrust litigation. The court knew in Rambus that the conduct was not shown to have harmed competition in the past and could not do so in the future. The court could not know that in Microsoft or in the Facebook case.
(2) Some have suggested that the Microsoft case does not support the case against Facebook on the ground that the conduct found to be unlawful in the Microsoft case provided no procompetitive benefits.10E.g., Timothy J. Muris & Jonathan E. Nuechterlein, First Principles for Review of Long- Consummated Mergers, Criterion J. on Innovation 29, 39 (2020). Facebook’s acquisitions, by contrast, combined complementary assets and might be found to have created at least some procompetitive benefits. The problem with this argument is that nowhere in its lengthy opinion did the Microsoft court suggest that its discussion of injury to competition and causation applied only to conduct found to have no benefits at all. And the court made clear elsewhere in the opinion that conduct can both provide efficiency benefits and violate the antitrust laws if the harms outweigh the benefits.11United States v. Microsoft, supra n. 4, at 59.
II.
The precedents thus seem to permit finding acquisitions of nascent competitors by a monopolist to violate Section 2, even if they do not violate Section 7. That leaves unaddressed, however, the questions whether that is sound as a matter of antitrust policy and what must be required to prove such a violation of Section 2. These are complex questions, and in this brief note, I will suggest only some preliminary thoughts.
Monopoly maintenance cases are rare. They require, among other things, proof that a firm with an existing monopoly engaged in anticompetitive conduct to extinguish or diminish a competitive threat to that monopoly. Perpetuation of a monopoly is at least potentially more harmful than the kind of conduct that is involved in more ordinary antitrust violations and creates lesser or more temporary forms of market power. And, while other forms of anticompetitive conduct reduce the likelihood that or timeframe in which a third party might be a competitive threat or, as in the case of Netscape, a complement to unidentified possible future competitors, acquisitions eliminate the possibility of future competition between the merging firms. Further, while general merger law is based on the implicit assumption that mergers usually create efficiencies, there is reason to be skeptical about horizontal acquisitions by monopolists because monopolists are especially likely to acquire nascent competitors in order to eliminate a competitive threat. Evidence that the monopolist coerced the acquired firm to sell might support such an inference. For these reasons, the expected value for economic welfare of these acquisitions might well be negative, even if the likelihood of harm is low.12See, e.g., C. Scott Hemphill & Tim Wu, Nascent Competitors, 168 U. Penn. L. Rev 1879, 1891 (2020). It thus seems reasonable to conclude that acquisitions of nascent or potential competitors by existing monopolies might be found to violate Section 2, even if they would not violate the more lenient standards of Section 7 applicable to all mergers.
Some of the elements of a monopoly-maintenance-by-acquisition claim seem easy to identify. The acquiring firm must be an existing monopoly, and its market position must in general seem secure in light of its resources, the nature of competition in the market, entry barriers, and similar considerations. Because acquisitions by monopolies of smaller firms might generate valuable efficiencies and provide important exit opportunities for investors in the acquired firms, not all acquisitions of firms that might conceivably become or facilitate a competitive threat to the monopoly should be illegal. Instead, there must be a reasonable possibility – more than di minimis but not necessarily more likely than not – that the acquired firm or firms will, if not acquired by the monopolist, develop into a substantial competitor of the monopolist. And because the acquisition of one or a few among many firms that are roughly equally likely to pose a competitive threat is not likely to harm competition, the acquired firms must either have some unique, differentiated path to becoming a substantial competitor or be among a relatively small number of firms reasonably able to become a substantial competitor. Proof of these elements should be enough to create a presumption that the acquisitions violate Section 2.
That leaves two more difficult questions: (1) Does a series of nascent acquisitions affect the analysis, other than by giving the plaintiff multiple bites at the unlawful acquisition apple? (2) How might nascent acquisitions be justified by efficiencies or other procompetitive benefits?
(1) Multiple acquisitions, each of which has only a small likelihood of injuring competition, might in aggregate have a substantial risk of harming competition. For example, one acquisition might have a 5 percent chance of harming competition; four acquisitions, each of which has an independent 5 percent chance of harming competition, would in aggregate have almost a 20 percent chance of harming competition. It is tempting to say that, if the aggregate likelihood of harm is big enough, the group of acquisitions should be unlawful.
Antitrust violations require anticompetitive conduct. Courts have long made clear that multiple instances of aggressive but procompetitive conduct do not violate the antitrust laws because the aggregation does not change the character of the conduct and thus cannot justify deeming any of the conduct to be anticompetitive. That principle would seem to make a series of acquisitions immaterial, except perhaps as evidence of anticompetitive intent.
There are two reasons, however, why a series of acquisitions might appropriately change the outcome other than just by creating additional evidence. First, the earlier acquisitions might change the market circumstances by, for example, reducing the number of possible new entrants or by affecting the evolution of the monopolist’s business and thus the commercial opportunities available to others. Those changes might make a later acquisition more likely to harm competition than if it had not been preceded by the others.13See, e.g., United States v. Microsoft, supra n. 4, at 72.
Second, and more fundamentally, the principle that aggregation of multiple instances of lawful conduct does not make any of the conduct unlawful makes sense when the conduct is thought to be lawful because it is procompetitive. It is less clearly sensible with respect to acquisitions that are thought to be lawful, not because they create procompetitive benefits, but because the risk of harm from any of them is too small. In that case, competition could be protected by finding a violation when the aggregate risk of harm becomes great enough.
There are, however, arguments for not finding a violation on the basis of the aggregate risk of harm. Antitrust law requires predictability. Firms need to know in real-time whether their conduct will be lawful or unlawful. That objective is undermined if firms have to imagine how a creative antitrust plaintiff might combine the pending acquisition with lawful acquisitions in the past to argue that the combination is unlawful. Also, is hard to see how all the acquisitions could be deemed to be illegal because the last one pushed the defendant over the limit, yet prohibiting only the last acquisition would enable the monopolist to sequence its acquisitions in order to make the least problematic acquisitions the most likely to be found to be unlawful. Maybe the solution is to find the set of acquisitions to be unlawful but to restrict the remedies to those appropriate to compensate for the harms caused by the acquisition(s) that pushed the set across the line to illegality.
(2) Section 7 law is based in part on the implicit assumption that most mergers have procompetitive benefits and thus requires plaintiffs to prove a likelihood of substantial harm. Once that proof is made, efficiency defenses are rarely successful. Monopoly maintenance under Section 2, however, is not based on any such assumption. A monopoly maintenance case can thus appropriately impose a lighter burden on the plaintiff to prove harm to competition, as in the Microsoft case, while permitting the defendant to defend its conduct by proving procompetitive benefits. The hard question is determining what the defendant needs to prove.
In form, nascent horizontal acquisitions are collusion cases, in which the firms agree not to compete. The usual test in such a case is, to oversimplify, whether merger-specific efficiencies enable the output of the merged firms to exceed the combined output of the two firms absent the merger. Thus, for example, Facebook might justify the Instagram acquisition by showing that the merger would enable improvements to one or both products that would increase total output.
But some nascent acquisitions might create merger-specific efficiencies that will not increase output in the overlap market. The acquisitions might, for example, generate R&D synergies or permit product improvements or cost reductions in adjacent or complementary markets. Section 2 law is unsettled about how to handle such efficiencies.
The cases and commentary suggest 5 alternatives. (i) Ignore the efficiencies on the ground that they do not benefit trading partners in the relevant market; that might be undesirable where the harm to competition is speculative. (ii) Find the acquisition to be lawful on the ground that it has procompetitive benefits; that would permit a risk of substantial harm in the monopolized market for what might be an insubstantial efficiency benefit. (iii) As suggested by the Microsoft case, “balance” the benefits and harms; courts rarely attempt such balancing, probably because of uncertainty about how to measure and compare benefits and harms. (iv) Find the acquisition to be lawful if the benefits are “disproportionate” to the harm; this would permit the court to paint with a broader brush and would thus ameliorate the difficult problems of measurement and comparison, but it would forego efficiencies that are greater than but not disproportionate to the harm. (v) Find the acquisition to be lawful if it would have been profitable for the monopolist even absent the competitive harm; this would make Facebook’s acquisition of Instagram lawful if the purchase price and integration costs were less than the sum of the value to Facebook of the merger-specific efficiencies plus the profits it could earn from operating Instagram without charging supra-competitive prices. This alternative would often be difficult to apply and would permit a risk of substantial harm in the monopolized market for what might be relatively insubstantial efficiency benefits. The latter possibility would depend in part on the extent to which the acquired firm was able to negotiate a price that reflected the full value of the acquisition to the monopolist; this alternative might thus induce monopolists to coerce involuntary sales of nascent competitors or to conceal the full purchase price of the acquisition.
*
In sum, it seems reasonably clear that nascent acquisitions by monopolists might violate Section 2, even if they would not violate Section 7 and even if harm to competition is less likely than not. How a series of multiple acquisitions by the monopolist and merger-specific efficiencies should affect the analysis is less clear.
A. Douglas Melamed
Professor of the Practice of Law, Stanford Law School
***
Citation: A. Douglas Melamed, Acquisitions of Nascent Competitors under Section 2 of the Sherman Act, CONCURRENTIALISTE (Feb. 8, 2021)
Read the other guest articles over here: link
