Prepared remarks: The Enduring Promise of Antitrust
In this essay, I will address the challenges related to antitrust remedies in high technology industries. But before doing so, I want to say a few words about the role of states—and State AGs in particular—in antitrust enforcement and the state of antitrust law.
I. The Role of the States in Antitrust Enforcement
During the 1970s, Congress began to develop a range of “cooperative federalism” regulatory programs. Under such programs, Congress authorizes state enforcement of federal law, allowing the federal government to set a floor for enforcement, but giving states additional authority to tailor standards as well as pick up any slack in enforcement (and it is critical to note that, given constrained federal resources, that slack may be the product of workload). By instituting such a model, Congress adopted a hedging strategy—ensuring a base level of uniformity while allowing for appropriate experimentation.
The environmental laws provide the classic example of cooperative federalism in action, with the Clean Air Act being a clear case in point. Under the Clean Air Act’s model, the EPA authorizes state agencies to address air pollution using a variety of tools, provided that they ensure a basic level of air quality. Where state agencies decide to go above the level specified by the EPA, they are permitted to do so. Following this precedent, both telecommunications regulation and health care policy later adopted a cooperative federalism architecture, blending state and federal authority and calling on state agencies to develop and enforce federal regulatory standards.
Antitrust law operates in a functionally similar manner. In 1976, in adopting the Hart Scott Rodino Antitrust Improvements Act, Congress embraced the ability of State AGs to enforce federal antitrust law on behalf of their States, using what is called “parens patriae” authority. The theory of this delegation of authority, like other cooperative federalism programs, is two-fold: (1) states may be better positioned to know of competitive issues in their jurisdictions; and (2) states may have a greater willingness to take action and have the ability to collect damages on behalf of their citizens, thereby further advancing the goals of antitrust law. As the Supreme Court put it, the role of states in antitrust enforcement “was in no sense an afterthought; it was an integral part of the congressional plan for protecting competition.”
One of the questions inherent in a cooperative federalism framework is whether the federal government has the authority to prevent states from going further than the federal government where, in their view, local conditions warrant. In the environmental arena, the EPA has the authority to ensure a minimum level of enforcement, but not to prevent states from taking additional action. In the antirust arena, the situation is similar: the federal government can take action to ensure a basic level of enforcement, but it does not have the power to prevent states from going further, under federal or state law, to stop anti-competitive conduct.
For an example of parallel federal and state action, consider the Microsoft case. In that case, the federal government ultimately decided—after a remand on the remedies issue by the D.C. Circuit Court of Appeals—on a regulatory remedy and declined to pursue structural relief. A number of states who were part of this litigation took a different view and proceeded to challenge the absence of divestiture. As this case was litigated and ultimately decided, the D.C. Circuit made clear that the States have the requisite authority to pursue a different view from the federal government if they choose to do so.
The opposite approach—empowering the federal government to bar states from antitrust enforcement whenever it so chooses—would undermine the architecture of cooperative federalism and hurt consumers in states where State AGs pick up the slack in federal enforcement by bringing additional resources to bear and by applying their local expertise. Consider, for example, the case of a recent merger in Colorado Springs between DaVita’s clinical network and UnitedHealth Group’s Medicare Advantage insurance product. In this case, UnitedHealth consummated this merger after its market share declined from around 75% to around 50%, owing to the emergence of a disruptive entrant, Humana. Humana emerged as a rival after building its relationship with DaVita’s clinics, which referred patients to it. In the wake of the merger, however, Humana faced the prospect of losing access to a critical resource.
The Federal Trade Commission reviewed the UnitedHealth/DaVita merger and declined to take action in the Colorado Springs market. In Colorado, however, we were concerned about the prospect of UnitedHealth using control over DaVita’s clinics to re-establish its dominant position in the Medicare Advantage market, leading to higher prices, less choice, and lower quality offerings to patients. By taking action in this case, separate and apart from the FTC, we were able to protect Colorado consumers. And rather than protest our action, the FTC respected our authority. Indeed, two Commissioners wrote separately to highlight the valuable role State AGs play in enforcing antitrust law.
Unfortunately, federal antitrust authorities don’t always show respect for state antitrust enforcement. In late 2019, for example, the Department of Justice filed a brief taking an unfortunate and unjustified position in the Sprint-TMobile merger. (Colorado had earlier dropped out of that litigation in favor of a settlement that benefited Coloradans, so we were not involved in the litigation itself.) In particular, the DOJ asserted in its brief that the states’ “role does not permit states to override the sovereign interests of the United States.” In essence, the DOJ argued that it is the supreme arbiter of antitrust law and once it takes a view on a matter, the states are foreclosed from any enforcement that would be contrary to the selected path of the federal agency. That view runs contrary to what Congress intended in framing broad enforcement of the antitrust laws that recognizes the critical role of the states. Because this position would up-end 45 years of antitrust practice and jurisprudence, the litigating states properly responded that the “States are independent enforcers of the antitrust laws, and it is the role of the Court—not any federal agency—to decide the lawfulness of the merger.”
The DOJ’s flawed argument in the T-Mobile case was based on the dissenting opinion in Georgia v. Pennsylvania Railroad Co. Notably, the majority in that case held that states are authorized to seek injunctive relief under the antitrust laws. Moreover, the DOJ’s suggestion—that the DOJ and FCC merger review in the T-Mobile case constitutes a federal regulatory proceeding that bars contrary state action—contradicts Congress’ empowerment of State AGs and is a dangerous idea as well. Under DOJ’s theory, if a federal antitrust agency approves a merger, antitrust enforcement against the merger would be precluded as long as the merger was subject to review and approval by the federal authority. That would diminish antitrust enforcement and would, in effect, permit the federal antitrust agencies to extinguish right of states and private parties to enforce, and seek remedies for harm caused by violations of, the antitrust laws.
The DOJ’s rationale in the T-Mobile case would easily justify exempting transactions from antitrust review because a federal regulatory agency approved such a matter. Notably, the DOJ invoked the FCC’s regulatory action in that case as a basis for stripping States of authority to challenge the merger. This is a dangerous claim and goes flatly against federal merger law, which has consistently rejected rejecting antitrust claims for non-antitrust reasons since the Supreme Court’s 1963 decision the Philadelphia National Bank case.
In a later speech, DOJ Antitrust Division chief Makan Delrahim defended the DOJ’s position. He argued that allowing states to bring antitrust actions of their own “creates the risk that a small subset of states, or even perhaps just one, could undermine beneficial transactions and settlements nationwide.” Moreover, he suggested that states should not be authorized to seek any “relief that is incompatible with relief secured by the federal government.” This concept of federal supremacy is incorrect and ignores the fact that states can enforce the federal antitrust laws only by bringing cases in federal court. If states advance claims that are unfounded and would “undermine beneficial transactions,” the courts will reject them. And the courts can, of course, take into account any action or decision by the federal antitrust agencies in assessing the state’s claims, just as the Court in Philadelphia National Bank took into account the actions there of federal bank regulatory agencies. But there is no basis in the statutes, the cases, or sound policy for a decision by a federal agency to preclude the states from exercising their rights under the antitrust laws by asking a federal court to prevent or provide remedies for a violation of those laws.
Although the Court ruled against the States in the T-Mobile case, Judge Marrero declined to adopt Delrahim’s proposed limit on States’ role. Rather than reject the States’ authority to bring the action, the court evaluated the case on the merits, noting that the views of federal regulators can be informative, but are not conclusive. To be sure, the presence of a remedy—a fix to the harm occasioned by the merger, as it were—is a fact of life that the litigating states and the court rightly had to address. Similarly, the DOJ would also need “to litigate the fix” if another federal regulatory agency (say, the Federal Communications Commission) adopted a remedy in the face of a DOJ merger challenge. But to face challenges in litigation is a far cry from being barred from the courtroom.
In short, the states are partners in antitrust enforcement, reflecting the cooperative federalism architecture adopted by Congress. In effect, Congress has empowered states to act as a check on federal enforcement or, more precisely, on instances of federal underenforcement; as such, it declined to allow federal inaction or preference for particular remedies to remove the states from antitrust enforcement. In this sense, the issue is not—as the DOJ suggests—that states might “displace the federal government’s role as the nation’s federal antitrust enforcer,” but rather that the states are positioned to pick up any slack and ensure that important issues are raised before the courts whether or not the federal agencies are inclined or able to do so.
II. The State of Antitrust Law As To Dominant Platforms
One great challenge for antitrust law is to address concerns that dominant platforms are able to exclude upstart, or nascent, rivals before those firms can emerge as competitive threats. In some cases, the concern is that dominant firms can acquire such upstarts before they emerge as a significant threat. In others, the concern is that the firm will respond to an upstart rival with actions that protect its dominance, say, exclusive dealing, tying, loyalty rebates, and most-favored nation provisions. Under either concern, dominant firms can undermine these upstart and nascent competitors to protect their position in the market. An expert report commissioned by the UK Competition and Markets Authority recently summarized the concern; “it is the threat of being overtaken by rivals that provides the spur to companies to innovate and produce new products that companies want.”
In the wake of the Chicago School revolution, a number of commentators—myself included—have explained that antitrust doctrine has, on occasion, mistakenly bent over backwards to avoid false positives, effectively ignoring false negatives. Consider, for example, predatory pricing cases, where courts excused dominant airlines’ actions focused on excluding a rival. The American Airlines case is a notable case in point. In that case, the Tenth Circuit affirmed a summary judgment ruling on the ground that the government had not demonstrated pricing of airplane flights below marginal cost. With the benefit of hindsight on this case (and higher prices and more concentration in the airline industry), however, the failure to focus more on the costs of false negatives bears close attention.
Appreciating the risk of under-enforcement, we can examine critically two potential limits in the reach of monopolization law. First, courts must consider whether enforcers must demonstrate consumer harm as a specific element of Section 2. In some cases, courts have suggested such a showing is necessary whereas others have made plain that demonstrating harm to the competitive process (including to nascent rivals) suffices.
Second, courts have, since the Trinko case, evaluated whether a change in a prior course of dealing is sufficient to give rise to antitrust liability. In Trinko, the Supreme Court declined to overrule the Aspen Skiing decision and the duty to deal doctrine, but it suggested that it lay at the outer limits of antitrust liability. In Aspen Skiing, the decision of a firm that owned three mountains to refuse to continue cooperating with a firm that owned one mountain through a four-mountain pass was the basis of antitrust liability for monopolization. As Judge Posner later put it, Aspen Skiing can be viewed as justifying antitrust liability where “cooperation is indispensable to effective competition.” Despite Trinko’s skepticism, the duty to deal doctrine set out in Aspen Skiing continues to have vitality.
The Chicago School project continues to perform an important role in calling for greater economic rigor in antitrust analysis. But the trend to downplay or overlook false negatives is a dangerous one. Going forward, I would like to see a greater re-examination of efforts to limit antitrust liability (as in the predatory pricing example) rather than to further restrict it (as in the case of duty to deal doctrine or requiring specific proof of consumer harm).
III. Reflections on Antitrust Remedies
Let me now share a few reflections on antitrust remedies. For starters, I want to underscore that, by contrast to substantive antitrust doctrine, too little scholarly analysis informs how to approach antitrust remedies. I won’t attempt to offer a comprehensive look at antitrust remedies here, and all potential remedies are worthy of serious consideration, but I will mention three broad categories of remedies that, among others, merit further discussion and analysis, specifically with regard to access and interoperability issues in high technology industries: (1) relying on a regulatory agency; (2) identifying a prior course of dealing as a benchmark; or (3) establishing a set of requirements to be overseen by an antitrust court and/or other entity.
It is clear that courts and enforcers need to devote more analysis on how to approach remedies. In the Microsoft case, for example, the district court decided on a structural remedy without undertaking a hearing or much analysis. In response, the D.C. Circuit remanded this issue for more careful consideration, leading to the ultimate regulatory requirements overseen by the district court.
One important starting principle for analysis of remedies should be one of comparative institutional competence. When, for example, a court has the option of adopting different potential remedies, it will need to conclude which option is best. With the Microsoft case almost certainly in mind, Judge Posner commented that remedies should be “sufficiently clear to be judicially administrable” and should not “impose an undue administrative burden on the district court, which would have to administer the decree.” In that sense, Judge Posner might be seen as calling for humility in devising remedies of a court’s own making.
In the AT&T case, Judge Greene opted for a belt-and-suspenders strategy, empowering both the FCC and the district court to oversee the antitrust remedies. In so doing, he went beyond the Otter Tail case, where the court required the electric utility defendant to commit to providing access to its network (through wholesale wheeling) by filing a tariff deemed acceptable by the Federal Power Commission. By contrast, Judge Greene was unwilling to rely solely on the FCC, also imposing judicially enforceable requirements that the Bell Companies could not discriminate “between AT&T and other companies in their procurement activities, the establishment of technical standards, the dissemination of technical information, their use of Operating Company facilities and charges for such use, and their network planning.” This requirement, along with others that mandated equal access to Bell Company infrastructure, was designed to be flexible and adaptable, allowing for ongoing waivers and refinements. Indeed, non-discrimination as a remedy requires careful consideration in a variety of settings.
Judge Greene’s adoption of a regulatory regime overseen by a district court was controversial and ultimately ended with the Telecommunications Act of 1996. Commentators differ on the effectiveness of this model, but there is wide appreciation for the boon to competition and innovation that resulted from the AT&T case. Less commented on, however, is that Judge Greene faced a painful dilemma: had he relied solely on the Otter Tail model, he would have lacked confidence that the FCC would be able to effectively facilitate competition. After all, Judge Greene heard powerful evidence in the AT&T case that the FCC was unable to regulate AT&T previously and was institutionally flawed.
In some cases, where a firm changes the terms of dealing to create an advantage for itself over upstart rivals, courts can look to fashioning relief based on a prior course of dealing where feasible. In both the Loran Journal case and Aspen Skiing, this was indeed the case. As I stated elsewhere, both cases grounded their remedies in “a prior course of dealing or some benchmark that can provide a tractable guide as to what constitutes a reasonable type of behavior.”
In thinking about antitrust law, courts probably underappreciate the impact of having a regulatory agency as a partner in overseeing remedies. In both Otter Tail and AT&T, the courts were able to benefit from such a partnership. In the Microsoft case, by contrast, courts were forced to develop a regime of their own making. Until and unless a digital platforms agency is established, courts in the Internet platform arena are going to be left with the model used in Microsoft as their best guide.
In addressing this challenge recently, the UK’s Competition and Markets Authority touched on a potential innovation in this area. In particular, that body suggested that “an enforceable code of conduct may help address a number of concerns that we have identified in digital advertising markets.” For American audiences, this model is likely less familiar than to their European counterparts. To be sure, the Federal Trade Commission has employed such a model successfully in the past, such as in the case of false advertising claims, but what Europeans call “co-regulation” is less familiar on this side of the Atlantic. In many respects, the framework for such a model follows from the pathmarking work of Ian Ayres and John Briathwaite’s Responsive Regulation, which envisions a more dynamic and adaptive model of regulatory oversight than traditional command-and-control regulation.
Antitrust law, over its 130 years, has proved to be an adaptive tool. In the years ahead, antitrust enforcers will need to rise to the challenges of addressing a more dynamic economy and one in which there is increasing market concentration. To meet that challenge, we will need to strengthen our systems of antitrust enforcement, including developing more effective models of monitoring anticompetitive conduct, taking effective action where needed, and implementing effective remedies.