Source: United States Department of Justice News
Remarks as Prepared for Delivery
Thank you for that kind introduction. Before I begin, I want to thank the team at Keystone for organizing this inspiring conference. The group of leaders, enforcers, advocates, and thinkers here speaks volumes about the global reach and significance of the issues that we confront. And a special “thank you” to the superlative and brilliant Cristina Caffarra for bringing us together today for this exceptional exchange of ideas.
As a preview, I want to talk today about the role of platform power. But, before I do that, I want to pause and reflect on some of the recent and extraordinary work of my talented and dedicated colleagues at the United States Department of Justice, Antitrust Division.
Competition Enforcement Is Working To Deliver a More Dynamic Economy
First, we have sought to prioritize bringing the right cases for the right reasons and I am pleased that our efforts have generated tangible results. We have blocked or forced the abandonment of numerous major mergers, including in industries as varied as ocean shipping, publishing, industrials, professional services, government contracting, enterprise technology, and numerous others. These merger challenges have preserved competitive markets for citizens and stakeholders in all corners of economy. And, let me be clear, we’re just getting started.
It’s not just merger enforcement—we have advanced our effort to revitalize Section 2 of the Sherman Antitrust Act to address monopolization. We have brought more cases in the last year than were brought in the previous 25! These cases involve markets touching the lives of millions of Americans, such as digital advertising, highway repair, and cross-border shipments. And, again, we are just getting started. And for the first time in nearly 50 years, we are bringing monopolization cases as both civil and criminal actions as the statute set forth when it was enacted by our Congress in 1890.
We have also launched the broadest enforcement program in the history of Section 8 of the Clayton Act, which prohibits interlocking directorates on corporate boards. Interlocking directorates have been a key driver of concentration for the last forty years. To address this pervasive threat, we currently have 17 active Section 8 investigation and, to date, we have caused more than 14 directors to resign from corporate boards.
Finally, we have added to our incredible team at the Division—over 100 new hires since I arrived, with more on the way. This includes attorneys, economists, and paralegals, but also includes new types of essential talent: technologists, data scientists, industry experts (e.g., banking experts) and others. We are developing the expertise to match the needs of a modern economy. I often remark that we are building our version of a business school faculty, which we are doing under the leadership of our Business School Dean — the incomparable Susan Athey.
And this is just the tip of the iceberg. The work we are doing across the program in everything from tech, to supply chain, to transportation, to labor competition, to procurement fraud and beyond is groundbreaking and a testament to the dedication, fortitude, and raw talent of the hardworking people at the Antitrust Division. I would be remiss if I did not mention that we have the wind at our backs. The Attorney General — a brilliant antitrust mind in his own right — has offered critical leadership and support for our efforts to reinvigorate antitrust enforcement. And President Biden’s Executive Order on Promoting Competition in the American Economy has provided a definitive mandate to revisit and revitalize antitrust enforcement and competition policy across the entire U.S. government.
I say all this not to spike the American Football, as the American saying goes, or to rest on our laurels. Far from it. I emphasize our progress so far to demonstrate how far we have come in such a short period of time, but also to emphasize how much work we have left to do and how we must think critically and move purposefully as we advance.
To that end, I want to continue the conversation I started with you last year about digital platforms and platform power. How do they work? How do they compete? How does market power play out? And what can and should be done to curb antitrust abuses?
Analysis of Platform Industries in Antitrust Enforcement
In any investigation, the first set of questions we need to ask is: “how does competition present itself in this market and what are the potential threats to competition and the competitive process?” This starting point—understanding how competition presents itself in a specific market—is critical. But, unfortunately, it is not always where antitrust analysis begins. Sometimes—as a function of comfort, habit, or risk-aversion—we have fallen into the trap of skipping this important foundational step.
There can be a tendency at times to fall back on the comfort of a particular analytical framing or “box” most familiar to us, that loosely describes the problematic conduct or merger-related concerns under investigation. But by failing to take a step back and understand fully how competition is actually playing out in a given market, we miss the opportunity to consider whether the analytical tools we are applying make sense in context. Perhaps more problematically, we may fail to recognize and act on anticompetitive market conduct because it does not fall neatly within one of our frequently used boxes. But the forms of competition are as varied and dynamic as our economy itself.
So, let’s talk about platforms, and how they compete.
At their core, platforms typically function as intermediaries. They can offer different products or services to different groups, both or all of whom depend on the platform to intermediate between them. Some platforms are two-sided—for instance, one that connects users with developers of software applications. But most are multi-sided, connecting individual people with one another, even as it also connects them to various products and services. The opportunities for platforms are almost boundless, but so are the opportunities for anticompetitive behavior and harm.
Here is where it is important to understand and focus upon market realities. Platforms typically differ from more traditional markets in numerous fundamental ways.
First, platforms often benefit from economies of scale as they grow. They are not private highways, but instead grow in strength and value by inviting participation from others. They have an incentive to start open to induce participation and gain scale, because scale and related effects are particularly strong.
What does that mean for competition? It means that these effects can often drive concentration in platform markets or cause markets to tip in favor of a single, dominant firm. Platforms can seek to achieve scale through a variety of both competitive and anticompetitive means in order to create a moat that leaves new entrants scrambling to keep up. For example, dominant platforms can build their market power and deepen competitive moats through strategic mergers and acquisitions (to block entry or the growth of rival) or through conduct that deprives rivals of the ability to achieve scale through competition on the merits. In these markets, the cumulative impact of a series of anticompetitive acts can be supercharged by network effects that cause a market to tip. The whole of these interlocking and reinforcing effects can be much greater than the sum of the parts.
Second, firms that own dominant platforms often get to set the rules of the road in unique ways. Many platform decisions—how to match users; the prioritization of users, products, and content; the allocation of surplus; and information and data accessibility—influence the behavior of platform users and participants. But where a platform is dominant, these decisions may affect not only participants on the platform but the very structure of the industry across a wide array of related markets. Critically, dominant platforms may be able to set those rules free from the discipline of a properly functioning competitive market.
So how do these special characteristics of platform markets play out in our day-to-day work? In the merger context, this means that we need to be particularly mindful of transactions that impede the ability of market participants to meaningfully choose the platforms they want to use based on the quality, price, privacy, security, innovation and other competitive elements of the platform itself. Where a merger reduces or eliminates competition on one side of a platform market, limiting other platforms’ ability to offer a competitive product on the merits, competition can suffer in direct or excess proportion across the entire platform. Where a merger impedes the ability of users to multi-home across platforms, the effect can be to drive out essential competition between platforms and create a deep moat that insulates the platform from competition.
As we think about our approach to mergers, platforms often do not fit neatly within the traditional horizontal and vertical geometries of competition. We must recognize the distinctive nature of platforms and apply analytical frameworks that reflect market realities. The geometry of platform markets more closely resembles a complex gemstone than a two-dimensional drawing on a piece of paper. The competitive significance of a platform market often depends on the district perspective of a subset of market participants. As a result, it is often easier to measure competition in the metaphorical distance between monopoly platform and a director competitor or competitive threat than attempting to define the outer boundaries of the market.
A similar story often plays out in our conduct investigations, where dominant firms distort the competitive process in their favor to impede rival’s access to users on one or both sides of the platform.
Just two and a half years ago, our Congress’ House Subcommittee on Antitrust released a report observing that “dominant platforms in many cases have also integrated into adjacent lines of business . . . operat[ing] both as key intermediaries for third-party companies as well as direct competitors to them.”[1] It went on to observe that “[i]n recent years, significant reporting has documented how the dominant platforms can exploit this dual role, through data exploitation, self-preferencing, appropriation of key technologies, and abrupt changes to a platform’s policies.”[2] The Subcommittee’s investigation—independent of any investigations by the Division—“uncovered numerous examples of this exploitative conduct, suggesting that these are increasingly systemic, rather than isolated, business practices.”[3]
This report is a historic document and has done so much to bring sophistication and advance awareness of these important issues. I would like to address directly today one particular category of conduct that is often referred to as “self-preferencing.”
Exclusionary Discrimination by Dominant Platforms
To jump ahead to the conclusion, let me say that the phrase “self-preferencing” often understates the competitive problem. In the United States, the phrase is often misappropriated, to lump anticompetitive behavior with otherwise competitively benign conduct.
Let’s back up, though, and explain what some people mean when they use this term. Some use the term self-preferencing to refer to conduct they perceive as ordinary. This might include a firm showcasing its various products to existing customers or the same firm offering basic discounts for multi-product purchases.
But exclusionary practices by a firm with monopoly power can be so much more pernicious. Dominant firms facing an antitrust investigation regularly label their conduct as mere “self-preferencing,” ducking behind a perceived — but ill advised — safe-harbor they would like to attach to that term. They claim such conduct is effectively immune from antitrust scrutiny. Sometime, this is presented as a tautology with no support cited. Other times, firms rely on a twisted interpretation of case precedents — such as Trinko and linkLine — to contend incorrectly that monopolists have an unqualified right to do as they see fit regardless of the competitive impact—full stop. Of course, this is a distortion of the law. Given its amorphous nature, labelling conduct as “self-preferencing” can end up obscuring rather than shedding light on the ultimate competitive analysis.
At the Division, our analysis does not turn on whether a particular flawed label could describe the conduct at issue. Rather our focus is on whether the conduct by a firm with monopoly power has had or will likely have an exclusionary effect and cause competitive harm. This has been the heartland of antitrust caselaw and analysis for over 100 years.
In the context of platforms, as I noted earlier, competition can play out along many dimensions. There is competition for the platform, on the platform, and, in many cases, competition to reduce the importance and relevance of the platform — in effect, eroding the power of the platform.
The relevant question is most often not whether the firm has some unilateral duty to deal with customers or rivals—in most cases involving platforms, the platform has already made that choice. Indeed, the very function of a platform is to exist as an intermediary that invites dealing and participation of others. Rather, the question we must answer is whether the firm has engaged in a discriminatory course of dealing that improperly excludes competitors from the market. Such exclusionary discrimination can take a variety of pernicious forms in platform industries that share a common result: a distortion of the competitive process.
Platforms can engage in exclusionary discrimination to exclude competitors—current or nascent—that directly threaten to displace or disintermediate a core monopoly. Platforms also may engage in such conduct to exclude competitive threats to expand the scope of their existing monopoly in order to erect additional barriers to competition that in the future might challenge the core monopoly, or as a mechanism to exercise monopoly power after entrenching its position.
Whether discriminatory conduct excludes competition for the platform, on the platform, or around the platform, or deepens the moat to insulate the platform from competition, the antitrust laws provide a mechanism for intervention.
To foster a more meaningful dialogue on this issue, I want to share some of our current thinking on this issue of exclusionary discrimination by platforms and other firms.
First, courts have long refused to place form over substance when it comes to monopolization, recognizing that exclusionary conduct can come in many forms. As the Supreme Court explained in United States v. Grinnell, the Sherman Act condemns the attainment or maintenance of a monopoly through anything other than “a superior product, business acumen, or historic accident.”[4] Those words do not encapsulate some overarching carveout for exclusionary discrimination.
Indeed, the Supreme Court has rejected head-on arguments that “the promotion of self-interest . . . immunize[s] otherwise illegal conduct.”[5] In Otter Tail, the Court held the Sherman Act “assumes that an enterprise will protect itself against loss by operating with superior service, lower costs, and improved efficiency,” not by seeking “to substitute for competition anticompetitive uses of its dominant economic power.”[6] Where a dominant firm opts to wield its economic power to engage in a discriminatory course of dealing rather than compete on the merits, it has the potential to run afoul of the Sherman Act.
Second, where a dominant firm engages in exclusionary discrimination that has the effect of creating or maintaining a monopoly—whether in a dominant firm’s core market or an adjacent one—we will not hesitate to investigate and, where appropriate, take action regardless of the label firms or their defenders seek to apply.
Our courts have long recognized that “whether any particular act of a monopolist is exclusionary, rather than merely a form of vigorous competition, can be difficult to discern: the means of illicit exclusion, like the means of legitimate competition, are myriad.”[7] But instead of simply throwing up our hands at this inherent challenge our courts and antitrust enforcement agencies must dig in to understand and assess the effect of new forms of competition within the context and realities of a particular market.
To that end, we must resist the urge to pigeonhole conduct into analytical boxes that fail to account for the specific circumstances of the market at issue. As the Seventh Circuit fairly recently observed, when confronted with potentially anticompetitive conduct, we cannot allow those boxes to cloud our examination of the ultimate question: whether the conduct “harm[s] the competitive process.”[8]
Remedies Must Account for Market Realities and the Future
As I mentioned earlier, dominant firms in platform industries often have a wide array of levers and dials at their disposal, each of which can be flipped or turned separately or in combination to alter competitive dynamics and harm competition. Often this conduct is subtle or hidden from the marketplace more broadly, which makes it all the more difficult to detect, investigate, and stop
If we assume platform operators behave in a manner to maximize share price, valuation, and/or profitability, then they have incentives to flip these levers that benefit their market position at the expense of rivals’. I don’t fault them for having those incentives, but where we find a monopolist has engaged in anticompetitive conduct by pulling those levers or turning those dials, we cannot simply ignore their ability and incentive to do so again in the future.
To that end, it long has been the policy of the Division — when appropriate — to consider structural remedies, dating back to the breakup of Standard Oil in 1911 and AT&T in 1982.[9] The logic behind that stance applies to platform markets just as much, if not more so, than other industries. Structural relief can break the incentive structures that otherwise might encourage platforms to shift towards closed systems that exclude competition; relief can reinvigorate incentives to interoperate and facilitate multi-homing by users across platforms.
Structural relief also provides greater certainty to enforcers and potentially obviates the needs for ongoing, invasive monitoring. If we move some of the levers or out of reach, even a profit-maximizing monopolist cannot use them to the detriment of competition and in ways that violate the Sherman Act.
Conclusion
So, where does this all leave us? In the spirit of this conference, the antitrust laws have always been part of what the scholars Joseph Fishkin and William Forbath call a “constitutional political economy.”[10] That constitutional grounding began at the state level – around the time of the Founding, North Carolina and Tennessee both had language in their state constitutions declaring that “monopolies are contrary to the genius of a free state…”[11]
Since then, many of our greatest political leaders—Frederick Douglass, Senator Sherman, Franklin Roosevelt, Senator Kefauver—understood this constitutional political economy. They knew that, as FDR put it, “freedom is no half and half affair.”[12] Freedom in the work place, the market place, and the polling place go hand-in-hand. These leaders leaned on the antitrust laws and a competitive economy as a solution, as we must today. The problems we face today, like entrenched platforms and exclusionary discrimination, would have been familiar to our revered trustbusters of yesteryear. At the time, they might not have understood what the internet was (let alone an API), but they could identity exclusionary conduct by a dominant player when they saw it. And they understood then — as we do now — that the solution should be to restore competition to markets as a means to ensure freedom from monopoly across our economy and society.
All of this is a tall order, but I have deep faith that the Antitrust Division is more than up to the task.
Thank you.
[1] Investigation of Competition in Digital Markets, Majority Staff Report and Recommendations, Subcommittee on Antitrust, Commercial, and Administrative Law of the Committee on the Judiciary at 30 (Oct. 6, 2020) (“House Report”) available at https://www.govinfo.gov/content/pkg/CPRT-117HPRT47832/pdf/CPRT-117HPRT47832.pdf.
[4] United States v. Grinnell Corp., 384 U.S. 563, 57071 (1966).
[5] Otter Tail Power Co. v. United States, 410 U.S. 366, 380, 93 S. Ct. 1022, 1031 (1973).
[7] United States v. Microsoft Corp., 346 U.S. App. D.C. 330, 253 F.3d 34, 58 (2001).
[8] Viamedia, Inc. v. Comcast Corp., 951 F.3d 429, 453 (7th Cir. 2020) (quoting Microsoft, 253 F.3d at 58); see also United States v. Microsoft Corp., 346 U.S. App. D.C. 330, 253 F.3d 34, 58 (2001).
[9] See Standard Oil Co. of New Jersey v. U.S., 221 U.S. 1 (1911); U.S. v. AT&T, 552 F. Supp. 131 (D. D.C. 1982).
[10] Joseph Fishkin and William E. Forbath, The Anti-Oligarchy Constitution: Reconstruction the Economic Foundations of American Democracy (2022).
[11] N.C. Const. Art. 1 Sec. XXIII (1776) https://avalon.law.yale.edu/18th_century/nc07.asp; TN Const. Art. 11 Sec. XXIII (1796) https://tsla.tnsosfiles.com/digital/teva/transcripts/33633.pdf.
[12] Franklin D. Roosevelt, Acceptance Speech for the Renomination for the Presidency at Philadelphia, Pa. (June 27, 1936), available at https://www.presidency.ucsb.edu/documents/acceptance-speech-for-the-renomination-for-the-presidency-philadelphia-pa.