Self-Preferencing and Antitrust: Harmful Solutions for an Improbable Problem

Platforms, including vertically integrated platforms, are ubiquitous — both in the modern economy, and throughout history. While some such firms have managed to capture the popular imagination (and ire), there is good reason to expect self-preferencing by vertically integrated platforms is far more likely to benefit competition than to harm it. While there are limited cases in which self-preferencing could be harmful, existing antitrust laws are more than capable of addressing these situations. Finally, the proposed cures for a largely illusory problem are likely to inflict a great deal of harm, for little apparent gain. Antitrust enforcement involving platform self-preferencing should continue to follow the better course established by the existing legal and economic framework.

By D. Bruce Hoffman & Garrett D. Shinn¹

 

Over the last few years, antitrust law has surged to an unusually prominent position in popular and political discourse — largely due to a chorus of claims that antitrust enforcement has been overly lax for close to half a century.² Some of the loudest complaints target situations where a firm that offers a place for others to sell their products (a “platform”) decides that it’s going to sell its own products there as well — and sometimes gives its products an advantage.

This practice, frequently called “self-preferencing,” is often taken as a self-evident evil in need of a cure. Thus, Senator Elizabeth Warren colorfully used a baseball metaphor to disparage self-preferencing, tweeting that “You can be an umpire, or you can be a player — but you can’t be both.”³ Even popular media such as Wired magazine has echoed these sentiments.⁴

But this proposition is more an article of faith than the result of sound economic theory or evidence. In reality, self-preferencing is likely to be beneficial to consumers and competition, or at worst harmless, except in certain particular circumstances. And existing antitrust law is quite capable of appropriately handling those circumstances. Further, the proposed cures for self-preferencing are likely to do more harm than good, as well as being difficult if not impossible to administer coherently.

 

I. WHAT IS “SELF-PREFERENCING BY PLATFORMS?” SOMETHING PRETTY NORMAL, AS IT TURNS OUT

For the purposes of this article, we’ll use the term self-preferencing to refer to a situation that meets three criteria: when a (1) platform (2) vertically integrates and (3) gives its downstream products an advantage of some kind.

First, what is a platform? A platform is just a firm whose business involves providing a place for other firms to sell their products or services. Instead of a traditional vertical supply chain where a firm purchases inputs from an upstream supplier and transforms them into a product or service to sell to downstream consumers, a platform provides a meeting place for buyers and sellers to transact more or less directly. The platform is not selling products so much as it is selling a means of connection, and it is compensated, usually through rents or commissions, for providing this means of connection.

Some of the companies most closely associated with the modern economy are platforms. For example, Amazon is a platform that connects sellers and buyers for everyday products from books to silverware to televisions. Uber is also a platform, connecting drivers and their cars with individuals who want to go from Point A to Point B. But platforms are not modern inventions. Farmers’ markets, flea markets, strip malls and shopping malls are also platforms, connecting merchants with goods to sell with consumers who want a convenient place to find merchants.

Second, the platforms we’re considering may decide to also sell their own products in their platform. So, for example, suppose someone who owns a large tract of land conveniently located near a town hosts a farmers’ market there every Saturday, and charges rent to the farmers who sell their produce at her market. Now let’s suppose that this market owner also decides to grow some produce of her own, and sell it in her market. This is also called vertical integration — our farmer is both providing the platform and selling on it.

Platform vertical integration is commonplace. When a consumer buys an Amazon Basics product on Amazon.com, this is the result of Amazon’s vertical integration. When a consumer downloads the iTunes app from the Apple App Store, this is similarly an instance where a platform makes its own product available on its platform. When a user visits the Group page for Oculus VR on Facebook, this could also be viewed as a form of vertical integration. Such examples generally go unnoticed, in large part because these situations are innocuous, ordinary, and in many cases simply reflect common sense. After all, it’s hard to imagine anyone complaining that the farmer we mentioned earlier also has a booth at her own market to sell her peppers and tomatoes alongside those of the other farmers (though, as we’ll see shortly, some critics seem to have recently discovered this millennia-old practice, and apparently think it should be illegal).⁵

For our third element, let’s take this a step further. Let’s suppose that our platform — at least arguably — gives its own products an advantage. So, for example, our market-owning farmer puts her own booth right by the main entrance to her market, or maybe next to the coffee stand or beer tent (depending on the time of day). This practice — a platform owner vertically integrating and giving her own product an advantage on the platform — is “self-preferencing.”

 

II. WHO’S AFRAID OF SELF-PREFERENCING? SURPRISINGLY, A LOT OF PEOPLE

It might seem that no one should be very surprised, or very alarmed, about the kind of practice we’ve described. After all, it’s probably been going on for as long as humans have been engaging in commerce. But, as we noted at the outset, self-preferencing has become a favorite boogeyman of the chorus of antitrust critics.

The European Commission was early to the assault on self-preferencing:

• In 2017 the Commission levied a 2.42 billion euro fine against Google for allegedly abusing its dominance as a search engine to benefit its comparison shopping service.⁶ The Commission first found Google “dominant” in multiple national markets for general online search services, and next found that Google had abused this dominance by giving its own comparison shopping service more favorable positioning and display on its general search results pages than it provided to competing third-party comparison shopping services. The Commission found that this behavior was capable or likely of having an anticompetitive effect because Google’s conduct decreased traffic to competing comparison shopping services, which conceivably could have led these competing services to exit the market, reduced their incentives to innovate, or reduced Google’s incentive to improve its own product; and because it may have impeded consumer access to the most relevant information.⁷ The Commission did not appear to be bothered that these theories of harm depended mostly on speculation about future harms that could not be observed in the market.⁸
• More recently, the European Commission proposed draft legislation on this topic in the form of the Digital Markets Act, which it introduced in December 2020. This Act would define certain platforms as “gatekeepers” based exclusively on their size and impose additional obligations on them for activities on their platform.⁹ It would, in essence, treat platforms that achieve a certain level of success as quasi-public utilities.

The U.S., while not originally on board with the attack on self-preferencing, now seems determined to make up for lost time:

• In December 2020, a coalition of states led by Texas sued Google over the way in which Google uses its position operating an auction-based advertising exchange to discriminate against rivals. Texas and its co-plaintiffs channel Senator Warren in saying that Google’s model was operating as “pitcher, batter, and umpire, all at the same time.”¹⁰
• The U.S. House Subcommittee on Antitrust, Commercial and Administrative Law of the Committee on the Judiciary held seven hearings and published a 400-plus page majority staff report on Competition in Digital Markets.¹¹ One of the key targets of the report was the extent to which large digital platforms supposedly use their power as platforms to benefit other aspects of their operations. The Committee included as recommendations “reduce conflicts of interest through structural separations and line of business restrictions” and “implement rules to prevent discrimination, favoritism, and self-preferencing.”
• Senator Klobuchar has introduced a bill that would dramatically impact antitrust enforcement.¹² While much of the bill is focused on mergers, it would also define a new antitrust violation for conduct that “materially disadvantages 1 or more actual or potential competitors.” The legislation specifically states that it would not be a defense for a platform to take an action that benefits one category of platform users if it presents an appreciable risk of harming a second category of platform users.
• The House Judiciary Subcommittee on Antitrust, Commercial, and Administrative Law has begun to hold hearings “to consider legislative proposals to address the rise and abuse of market power online and to modernize the antitrust laws.”¹³ The first hearing focused on “proposals to address gatekeeper power.”

This political and popular attention is not aimed solely at self-preferencing. Broader efforts to prohibit platforms from vertically integrating, and to break up those that have, are also underway.¹⁴

 

III. WHAT’S SO BAD ABOUT SELF-PREFERENCING? AS IT TURNS OUT, NOT MUCH

So we see that modern political and popular discourse treats self-preferencing as nearly a per se evil. But, is there any reason to think that is right? Consider the example of that most quotidian of platforms: the shopping mall.¹⁵

A shopping mall is a platform. Its value is as a gathering place for stores on the one hand and shoppers on the other. It is compensated for this role by rent collected from its tenants. Let’s suppose that a particular mall decides to open its own pretzel store in its food court.

A question we should ask right off the bat is: why would the mall open its own pretzel store? Is it likely due to some nefarious motive — some secret desire, say, to squelch the businesses of the pretzel and other snack stores already inhabiting its food court, and to take their customers — and the money they spend — for itself? The answer is probably not, for a couple of reasons.

First, if the mall owner wants to make more money, opening its own pretzel store is a complicated, expensive, risky, and inefficient way to achieve that goal. It has much better tools at its disposal, such as raising the rent a little, or doing a bit more advertising to drive additional traffic to its mall. Occam’s razor thus cuts this explanation off pretty neatly.¹⁶

Second, and perhaps more importantly, platforms don’t really care what items businesses are selling on the platform; a mall has no particular reason to care whether a given store sells pretzels, shoes, or jewelry.¹⁷ Nor do platforms have a strong interest in dictating what particular products customers are buying; again, the mall doesn’t really care if you’re buying burgers, shirts, or perfume. What platforms want is more businesses successfully selling more stuff to more customers. They care about the mix of those businesses only insofar as that mix is more likely to be attractive to more customers, and — in turn — more attractive to other businesses. The mall wants stores to want to be in its mall, and customers to want to shop there.¹⁸

What all this means is that (1) because the platform has other, easier ways to make more money than by vertically integrating, and (2) because the platform’s motivation is to make both sellers and buyers happier, the most likely reason a platform would vertically integrate is to make buyers happier. That will make them more likely to come to the mall, spend more time there, and spend more money in all the stores at the mall. In turn, this will make sellers more likely to want to be in the mall, and able and willing to pay higher rent because of the higher sales they’ll make. In more technical antitrust terms, the most likely reason for platform vertical integration is to increase output and consumer (and overall) welfare.

So let’s consider our new pretzel store. As we’ve shown above, the most likely reason for our mall owner to open its own pretzel store is that it perceives a gap in the products and services being offered to consumers in the mall, and thinks it’s best able to fill that gap. Perhaps the mall owner noticed a shocking dearth of high-quality pretzels, or has a particularly good pretzel recipe, or has some other reason to think it’s in a uniquely good position to provide mall shoppers a really attractive pretzel offering. Whatever the specific reason, when the mall owner vertically integrates, it’s probably trying to improve the mall’s consumer offering. As a result of our mall opening the pretzel store, hungry shoppers have additional options for food and competition between food court shops increases, which may even lead to lower prices at all stores in the food court. And, shoppers who might otherwise have left the mall to buy food somewhere else now spend a little more time there, shopping at other stores and increasing those merchants’ sales.

Of course, the mall owner might get this wrong — the pretzel shop might be low-quality or unpopular, maybe even driving shoppers out of the mall. But if that happens, the mall owner suffers as much as (or more than) the other businesses, because the mall owner’s main goal is to have more shoppers spending more time and money in the mall, not fewer shoppers spending less time and money.

But what if the mall owner does more than open its own pretzel shop — what if it “favors” that shop, by, say, not charging itself rent, and so offering lower prices for pretzels than other tenants, or perhaps by putting its shop in a prime location, or surrounding it with the best tables, nicest fountains and seating areas, and so forth? In other words, what inferences should we glean from self-preferencing? The answer turns out to be the same — the most likely reason for self-preferencing is to improve the overall quality of the mall (the platform) to consumers, and thus to increase output and welfare.

It’s actually not difficult to understand why self-preferencing is probably good for consumers and competition in most circumstances if we just recall why the platform is likely to vertically integrate in the first place — to increase traffic and sales in the mall. To trace this point through, consider how self-preferencing would play out for our mall owner and pretzel store.

Let’s start by considering the possibility that the mall owner actually just wants to hurt the competing pretzel and snack shops, or even drive them out of the mall — could that explain self-preferencing? It should be obvious that the answer is probably not, for a simple reason: the mall owner has much better tools for that purpose. It could, for example, just kick the competing snack stores out of the mall, or raise their rents to unsustainable levels. Making its own pretzel offering particularly compelling for customers is a costly and roundabout way to drive out rivals, if that’s what the mall owner really wants to do.

More importantly, consider what will happen if the mall owner’s actions hurt other snack stores and the net result — the changes in the other snack stores plus the effect of the addition of the new pretzel store — is bad for customers. In other words, the other snack stores close, or raise prices or cut the quality of their offerings because they’re being squeezed on rent, and whatever the mall owner’s own pretzel store offers isn’t good enough to outweigh the negative effects of the degradation of the other snack stores. So now there are long lines at the pretzel store, and shoppers are upset because they don’t have the options they appreciated before. If that happens, the mall will become less attractive to shoppers. Some won’t come, some will leave sooner, and in total, they’ll spend less. As shopper spending declines, the mall becomes less attractive to merchants; some close, or cut quality, or reduce hours, and total sales fall, which in turn makes the mall even less attractive to shoppers, and so on.¹⁹ And as all of this happens, the mall owner gets less rent and is worse off, which plainly wasn’t its goal.

So if the mall owner doesn’t want to hurt rival snack shops by setting up and favoring its pretzel store, what is it trying to do? The answer is obvious — the most likely reason for a mall to do this is because it thinks the net result of its actions will benefit shoppers and make the mall more attractive (both to shoppers and merchants). In other words, our little thought experiment demonstrates that the most likely explanation for platform self-preferencing is procompetitive, benefiting both buyers and sellers on the platform and increasing overall output.

The mall example may seem quaint in a world of multibillion-dollar digital platforms, but the principles it illuminates are no less applicable to those platforms than to shopping malls or farmers’ markets. Self-preferencing should generally be expected to be efficient and pro-competitive. This is because a platform is agnostic to its source of revenue. Its primary interest is in maximizing traffic on the platform to drive revenue, and it will not likely take actions that endanger this interest. Because the platform’s highest interest is in maximizing traffic, the platform’s interests are a good proxy for consumer interests and welfare. The conclusion then is that if a platform vertically integrates, it is because it expects that it can improve its platform by doing so. If a platform self-preferences, it is because it expects that consumers will be well-served as a result.

One underlying premise of this argument that deserves further reinforcement is that it is important to define what one means when one refers to anticompetitive conduct or outcome. “Unfairness” is not anticompetitive. To adopt a truism of U.S. antitrust law, the antitrust laws protect competition, not competitors. While baseball has proven an irresistible metaphor for some commentators, the metaphor is fundamentally flawed. The economy is not a sport or a schoolyard, and behavior that may be perceived as unfair by a competitor — for example, lowering prices when the competitor would have preferred market-wide prices to stay high — is often the pinnacle of pro-competitive behavior.²⁰ Rather, an “anticompetitive” outcome means that one of several possibilities is occurring in the market: overall, prices are rising, output is falling, or innovation or quality is decreasing.²¹ Behavior that harms competitors but doesn’t harm consumers or the competitive market is by definition not anticompetitive.

 

IV. BUT SHOULD WE WORRY ABOUT A MALL MONOPOLIST?

Well, the reader may be thinking, that’s all well and good for shopping malls, which face lots of competition. But what if the self-preferencing platform is a monopolist? It turns out, though, that the basic intuition we describe above should generally hold even where the platform has market or monopoly power. This is the case for two reasons.

First, if the platform’s vertical integration and self-preferencing is actually bad for consumers, the result is simply that the monopoly platform makes more profit in the short run, but undermines its monopoly in the long run. The effects are no different than a monopolist charging a monopoly price (using our mall example, raising rents sky-high because no other malls are around); the monopolist makes more in the short run, but at the expense of the longevity of its monopoly.²²

Second, in the U.S. at least, it’s not illegal for a monopolist to reap the rewards of its monopoly, because those rewards help drive innovation and competition.²³ Section 2 of the Sherman Act doesn’t prohibit monopolies; it assumes that a competitive economy can result in monopolies, because some competitor might just do things better than all of its rivals.^24] Rather, Section 2 prohibits anticompetitive conduct — conduct that harms the competitive process — that results in a monopoly.²⁵ As the Supreme Court has explained, even so-called monopoly leveraging (using a monopoly in one market to gain an advantage in another) does not violate the antitrust laws so long as this leveraging does not create or maintain monopoly power in a second market.²⁶ Antitrust law isn’t price regulation.

 

V. IS PLATFORM SELF-PREFERENCING EVER A PROBLEM? SOMETIMES, YES — BUT WE CAN DEAL WITH THAT

By this point, it might seem that platform self-preferencing can never be a problem, and should be per se legal under the antitrust laws. But that’s not correct. There are at least two scenarios under which platform self-preferencing could harm competition and consumers, and should therefore be subject to enforcement. Both, however, require specific, unusual facts. And both can be readily dealt with under existing antitrust law.

First, a platform with an existing monopoly in the platform market may anticompetitively maintain its monopoly by identifying and eliminating nascent threats to that monopoly. Vertical integration with self-preferencing can be one tool in the monopolist’s anticompetitive kit, if the monopolist sees a service offered on its platform as a nascent threat of this kind. But note that this requires that the vertical offering — the snack shops, in our mall example — has to threaten the platform monopoly — the mall, in our example. That fact pattern isn’t likely to be common, though it does occur from time to time.

The classic example of this form of monopoly maintenance by a self-preferencing platform is the United States’ turn-of-the-century case against Microsoft. In United States v. Microsoft, the government challenged a variety of Microsoft business and licensing practices related to its Windows operating system and Internet Explorer Internet browser.²⁷ The District Court, upheld by the D.C. Circuit, agreed with the government that Microsoft’s behavior constituted illegal monopolization.

At the time of the Microsoft case, Microsoft’s Windows operating system had greater than 95 percent share of the operating system market. But so-called “middleware” products like the Netscape Navigator browser threatened Microsoft’s position by providing an alternative way for computer programmers and users to accomplish the same functions that they relied on Windows to perform. As the court explained, “a middleware product written for Windows could take over some or all of Windows’s valuable platform functions — that is, developers might begin to rely upon APIs exposed by the middleware for basic routines rather than relying upon the API set included in Windows. If middleware were written for multiple operating systems, its impact could be even greater.” If this occurred, “a consumer could have access to the applications he desired — regardless of the operating system he uses — simply by installing a particular browser on his computer, then he would no longer feel compelled to select Windows in order to have access to those applications; he could select an operating system other than Windows based solely upon its quality and price. In other words, the market for operating systems would be competitive.”

To combat this threat, Microsoft took steps to promote its own Internet browser at the expense of Netscape Navigator. In particular, Microsoft designed its product such that Explorer came pre-installed with Windows and could not easily be removed, and it imposed licensing restrictions on computer OEMs meant to give Explorer an advantage over Navigator or other browsers. The court found that such actions had exclusionary effects, were not justified by any procompetitive rationale, served to maintain Microsoft’s monopoly in the operating system market, and were thus illegal.

The second set of circumstances under which self-preferencing could be subject to antitrust enforcement is where a platform is using self-preferencing to monopolize a downstream market. To go back to our mall, imagine that the mall owner thinks, for whatever reason, that by vertically integrating into pretzels and favoring its own pretzel shop, it could monopolize the downstream pretzel market — not just in its mall, but more broadly, in a well-defined antitrust market. That could violate Section 2 (though how often such downstream monopolization is actually plausible is an open question). Once again, though, an attempt by a platform monopolist to use self-preferencing to monopolize a downstream market is squarely in the wheelhouse of current antitrust law; no radical legal changes are necessary to address this problem if and when it arises.

One last point on this issue bears mentioning, though: self-preferencing where the monopolist makes its own product better, as opposed to degrading its rivals’ products, is unlikely to constitute anticompetitive conduct.²⁸ Monopolists don’t have to compete with one hand tied behind their backs.²⁹ If their monopoly has put them in a position to offer a particularly good product to consumers, that should be celebrated, not condemned. However, a monopolist that self-preferences by degrading or harming its rivals’ products could face enforcement under Section 2 (if the other conditions we describe are satisfied).

Thus, while self-preferencing is in most cases likely to be beneficial, there are two situations where it could be anticompetitive. Both situations, however, can be addressed by current law (though both require very specific facts that seem unlikely to arise with great frequency).

 

VI. A WORD ABOUT CURES THAT ARE WORSE THAN DISEASES

Medieval doctors often killed their patients with ill-advised treatments.³⁰ Similarly, there is a serious risk that the current zealotry against self-preferencing will inflict considerable economic damage in an attempt to cure problems that, to the extent they exist, could be remedied by less-drastic measures. To explain the danger, we briefly touch on the practical consequences of some of the remedies being bandied about for self-preferencing.

Some anti-self-preferencing advocates are calling for root-and-branch structural relief, i.e. preventing platforms from self-preferencing by preventing them from vertically integrating in the first place, and carving up platforms that are already integrated.³¹ After all, you can’t favor your own product if you’re not allowed to offer it in the first place. But empirical evidence indicates that vertical integration is generally good — it reduces prices, reduces costs, increases competition, and benefits consumers — and is not usually harmful.³² And, as we’ve shown above, this is equally true for platform vertical integration. In fact, because of platforms’ incentives and the consequences of multisided indirect network effects, vertical integration by platforms seems particularly likely to be beneficial, and particularly unlikely to be harmful. Indeed, many of the complaints leveled at vertically integrated platforms are, on their face, complaints about practices that benefit consumers by providing better products (though competitors, of course, don’t like those products). Broadly prohibiting platform vertical integration to avoid the risk that in some rare cases vertical integration might be anticompetitive would be like amputating your leg to treat a mosquito bite.

Some commentators suggest that rather than prohibiting platform vertical integration, platforms — or some platforms — should be barred from self-preferencing.³³ That’s certainly better than barring vertical integration outright. But prohibiting self-preferencing is also easier said than done in any coherent and administrable way, and would likely result in harming consumers and competition for little or no gain.

To illustrate, let’s return to the mall and the pretzel shop. Should the mall owner be prohibited from putting its pretzel shop in a high-traffic location — even though shoppers would be delighted to have the shop there? Or precluded from improving the lighting, or the seating areas, around the pretzel shop? What if the mall owner happens to learn, by looking at shopping and traffic data, that a particular corner of the mall — one that no store has shown interest in — happens to be highly-trafficked by customers who are particularly likely to want pretzels. Should the mall owner be prohibited from using this information to get those customers the pretzels they want? Or, suppose the mall owner happens to have invented a uniquely good pretzel-frying technique — would using that technique be illegal? Is even not charging itself rent illegal self-preferencing?

One response to this quandary might be to try to prohibit self-preferencing that degrades rivals, rather than self-preferencing that improves the platform’s own product. That’s clearly a better starting point. But, once again it turns out to be difficult to define in practice — and it creates a high risk of opportunistic behavior.

If, for example, the mall owner launches a series of renovations, snack stores could complain if their area gets a face-lift later than the spaces around the pretzel store, because they’re being put at a relative disadvantage. But they could just as readily complain if their area gets renovated first, because the dust and disruption from the construction could drive shoppers away from their stores — and maybe some of those shoppers would end up at the pretzel shop. Ultimately, a prohibition of this nature would become a “Swiss Army knife” tool store owners could use to extract costly concessions from the mall without any connection to true competitive issues. And the ever-present threat of legal sanctions for routine business activities could have a chilling effect on the mall, deterring it from making sensible choices or eventually forcing it to give up on the pretzel shop — even if doing so hurts customers.

There are other problems as well. For one, how should enforcers decide which platforms should be subjected to remedies? Clearly, prohibiting all platforms from self-preferencing would be absurd, impossible to enforce, and a massive government overreach that would no doubt spark an equally massive backlash; our farmer should be able to put her vegetable stand by her market’s entrance without worrying that antitrust enforcers will descend on her, brandishing subpoenas and dense economic treatises, early one Saturday morning.

But drawing finer lines isn’t much more promising. For starters, using ad hoc or arbitrary classifications — such as the type of business involved, or the size of the platform in transaction value³⁴ — has no apparent relationship to competitive harm. Such capricious line-drawing would likely lack a rational basis, and certainly would be unlikely to advance any legitimate economic end.

Some proposals suggest using market power as a yardstick, a measure that at least finds its foundation in antitrust law principles.³⁵ But despite its appeal, that proposal fares little better. Platforms are ubiquitous, and because “market power” can often be creatively defined,³⁶ limiting enforcement to platforms with market power is a fairly hollow safeguard. For example, if a mall has numerous local retail competitors, but the next closest mall is two towns over, does it have market power? Is the market for retail, or malls? Thinking back to our farmer — does she have market power because she owns a piece of land that’s convenient to a particular town? Disgruntled or opportunistic competitors will have a strong incentive to push the boundaries of any rule to gain leverage and extract rents from platforms — but that’s not likely a procompetitive result.^37]

 

VII. CONCLUSION

Platforms, including vertically integrated platforms, are ubiquitous — both in the modern economy, and throughout history. While some such firms have managed to capture the popular imagination (and politicians’ ire), there is good reason to expect self-preferencing by vertically integrated platforms is more likely to benefit competition than to harm it. Further, existing antitrust laws are more than capable of addressing the limited cases in which self-preferencing could be harmful. Finally, proposed cures that reach beyond current antitrust law are more likely to inflict harm than they are to effectively address any problem that needs a solution. Antitrust enforcement involving platform self-preferencing should continue to follow the course established by the existing legal and economic framework.

---

1 Hoffman is a partner and Shinn an associate at Cleary Gottlieb Steen & Hamilton LLP. This article elaborates on points made by Hoffman in various speeches during his tenure as Director of the FTC’s Bureau of Competition; the views expressed are the authors’ alone, and don’t necessarily reflect those of the firm, or any of its clients, or anyone else. Cleary represents numerous clients who operate digital and other platforms, buy on platforms, sell on platforms, supply platforms, compete with platforms, and engage in various combinations of those activities.

2 E.g., Senator Amy Klobuchar, Press Release, Senator Klobuchar Introduces Sweeping Bill to Promote Competition and Improve Antitrust Enforcement (Feb. 4, 2021) (“While the United States once had some of the most effective antitrust laws in the world, our economy today faces a massive competition problem. We can no longer sweep this issue under the rug and hope our existing laws are adequate.”); Commissioners Rohit Chopra & Rebecca Kelly Slaughter, Joint Dissenting Statement Regarding the Vertical Merger Commentary (Dec. 22, 2020) (“[The Vertical Merger Guidelines and Commentary] reflect the same status quo thinking that has allowed decades of vertical consolidation to go uninvestigated and unchallenged. . . . We look forward to turning the page on the era of lax oversight and to beginning to investigate, analyze, and enforce the antitrust laws against vertical mergers with vigor.”); Lina M. Khan, Amazon’s Antitrust Paradox, 126 Yale L.J. 710 (2017) (“This analysis reveals that the current framework in antitrust—specifically its equating competition with ‘consumer welfare,’ typically measured through short-term effects on price and output—fails to capture the architecture of market power in the twenty-first century marketplace.”); Small Business Rising, Our Goals, (“America’s antimonopoly laws were enacted to safeguard small businesses, farmers, and working people from coercion and exploitation by preventing large corporations from amassing too much power. But beginning in the 1980s, the FTC and DOJ, as well as the courts, reinterpreted these laws. They abandoned concerns about outsized power and instead oriented antitrust enforcement around the idea of maximizing efficiency. Under this framework, enforcers embraced consolidation and adopted a permissive approach to predatory behavior by big corporations.”); Robert H. Lande & Sandeep Vaheesan, The Atlantic, Ban All Big Mergers. Period. (Feb. 25, 2021) (“Sadly, that tradition [of aggressive antitrust enforcement] gave way in the 1970s and ’80s, as federal judges, the Justice Department’s antitrust division, and the Federal Trade Commission all came under the spell of dubious interpretations of history and economic theories strikingly tolerant of mergers and monopolistic practices.”). These claims don’t necessarily stand up well to careful examination. See e.g. Commissioners Noah Joshua Phillips & Christine S. Wilson, Statement Regarding the Commentary on Vertical Merger Enforcement (Dec. 22, 2020) (“[The history of vertical merger enforcement] reflects evolving antitrust jurisprudence, the steady refinement of economic analysis, and the specific facts of each case at issue. Any proposals for a new approach to vertical merger enforcement, which our colleagues have yet to articulate, would need to take into account and grapple with the law, economics, and the evidence in each case.”); Timothy J. Muris & Jonathan E. Nuechterlein, Antitrust in the Internet Era: The Legacy of United States v. A&P, Rev. of Indus. Org. (2019) (describing “the intellectual void at the heart of this populist antitrust movement”); Carl Shapiro, Antitrust in a Time of Populism, Int’l J. of Indus. Org. (2018) (“The danger to effective antitrust enforcement is that today’s populist sentiments are fueling a ‘big is bad’ mentality, leading to policies that will slow economic growth and harm consumers.”); Joshua D. Wright, et al., Requiem for a Paradox: The Dubious Rise and Inevitable Fall of Hipster Antitrust, 51 Ariz. St. L.J. 293 (2019) (“We demonstrate that, when evaluated as evidence-based policy proposals, the Hipster Antitrust agenda fails to substantiate its claims and promises.”). But that’s an issue for another article.

3 Senator Elizabeth Warren, Twitter (Apr. 22, 2019), https://twitter.com/ewarren/status/1120484639922110464.

4 See e.g. Wired, Big Music Needs to Be Broken Up to Save the Industry (Mar. 16, 2021) (“An expanded monopoly law that includes an explicit ban on self-preferencing could prevent a conglomerate like Liberty Media from using its radio and streaming properties to boost the tours and artists controlled by Live Nation.”)

5 The Wired writer mentioned above also fits into this group.

6 European Commission, Press Release, Antitrust: Commission fines Google €2.42 billion for abusing dominance as search engine by giving illegal advantage to own comparison shopping service (June 27, 2017); Case AT.39740, Google Search (Shopping) (June 27, 2017). Cleary represented Google in this matter, though Hoffman wasn’t at Cleary at the time and Shinn didn’t work on the case.

7 It is worth noting that the Commission’s case against Google was brought under Article 102 of the Treaty on the Functioning of the European Union, which prohibits abuse of a dominant position. The closest corollary in the United States, Section 2 of the Sherman Act, prohibits similar conduct only if it has the effect of creating or maintaining a monopoly. See Verizon Comms. Inc. v. Law Offices of Curtis V. Trinko, LLP, 540 U.S. 398 (2004) (rejecting “monopoly leveraging” as a theory of illegality when the standard requirements of Section 2 are not met). See also In re Google Digital Advertising Antitrust Litigation, No. 5:20-cv-003556-BLF, ECF No. 143 (N.D. Cal. May 13, 2021) (dismissing Section 2 claim against Google where plaintiff failed to plead anticompetitive conduct in addition to anticompetitive effects). We’ll return to this point later.

8 E.g. Case AT.39740, Google Search (Shopping), para. 602 (“In the first place, the Commission is not required to prove that the Conduct has the actual effect of leading certain competing comparison shopping services to cease offering their services. Rather, it is sufficient for the Commission to demonstrate that the Conduct is capable of having, or likely to have, such an effect.”)

9 European Commission, Digital Markets Act: Ensuring Fair and Open Digital Markets (Dec. 15, 2020).

10 See Complaint, Texas et al. v. Google, LLC, ECF 1, (E.D. Tex. Dec. 16, 2020).

11 See Maj. Staff of H.R. Subcomm. On Antitrust, Commercial and Administrative Law of the Comm. on the Judiciary, 116th Cong., Investigation of Competition in Digital Markets (2020).

12 See Competition and Antitrust Law Enforcement Reform Act of 2021, S. 225, 117th Cong. (2021).

13 House Committee on the Judiciary, House Judiciary Antitrust Subcommittee Announces Series of Hearings on Proposals to Curb the Dominance of Online Platforms and Modernize Antitrust Law (Feb. 18, 2021).

14 See e.g. Competition and Antitrust Law Enforcement Reform Act of 2021, S. 225, 117th Cong. (2021); Senator Elizabeth Warren, Here’s How We Can Break Up Big Tech (Mar. 8, 2019), https://medium.com/@teamwarren/heres-how-we-can-break-up-big-tech-9ad9e0da324c.

15 We’re using this commonplace example because the complexity of technology platforms (and the political venom directed at them) tends to obscure the simple economic logic underpinning self-preferencing.

16 As William of Ockham put it, “pluralitas non est ponenda sine necessitate.” A bit more plainly: the simple answer is likely the right one.

17 This may be a little different if the platform markets itself as specializing in a type of product — for example, a shoe market or jewelry market. In that case, it will probably want a good mix of sellers of the type of products it’s using for its marketing. But our overall point holds in such cases too.

18 More technically, the platform owner benefits by maximizing transactions on its platform. Thus, it wants to maximize indirect network effects operating on both sides of the platform. See e.g. Geoffrey G. Parker & Marshall W. Van Alstyne, Two-Sided Network Effects: A Theory of Information Product Design, 51 Management Science 1494 (2005). The platform’s incentive is to fuel those network effects by increasing its attractiveness to users on both sides, particularly on the side where the indirect network effects are weaker. Id. (demonstrating that under some conditions indirect market effects may lead a profit-maximizing platform to subsidize or even give away for free a product sold on its platform); Michael L. Katz & Carl Shapiro, Systems Competition and Network Effects, 8 J. Econ. Perspectives 93 (1994) (arguing that network ownership (such as occurs with platforms) allows internalization of the benefits of indirect network effects, leading to increased investment by the owner in the network and to overall welfare gains).

19 Due to indirect network effects, declines on both sides of the market could accelerate due to a negative feedback loop. E.g. Catherine Tucker, Network Effects and Market Power: What Have We Learned in the Last Decade?, Antitrust (2018) (arguing that network effects may lead to market power that is more unstable and prone to reversal than traditionally thought). This danger strengthens platform owners’ incentives to avoid degrading the platform for either side of a two-sided market; platforms don’t want to trigger this potential downward spiral. It’s important to note that this does not mean that a platform can never raise price to one side or the other — but it suggests that a price increase to one side of the market that holds without degrading the overall platform probably reflects an increase in the platform’s quality with corresponding benefits to those paying the higher price (likely in the form of increased activity on the other side of the platform).

20 Indeed, it is perhaps noteworthy in this context that a cartel member — the quintessential antitrust law-breaker — that does not go along with the anticompetitive actions of the cartel is referred to as “cheating.”

21 Phillip E. Areeda & Herbert Hovenkamp, Antitrust Law: An Analysis of Antitrust Principles and Their Application, ¶ 114 (“The overall goal is markets that maximize output, whether measured by quantity or quality.”).

22 E.g. Alaska Airlines, Inc. v. United Airlines, Inc., 948 F.2d 536, 549 (9th Cir. 1991) (“Every act exploiting monopoly power to the disadvantage of the monopoly’s customers hastens the monopoly’s end by making the potential competition more attractive.”). In our mall example, the overcharged merchants will eventually find another way to reach customers, such as opening standalone stores, creating outdoor semi-malls, going back downtown, or using the internet. Or, an enterprising firm will open a new mall.

23 Verizon Comms. Inc. v. Law Offices of Curtis V. Trinko, LLP, 540 U.S. 398, 407 (2004) (“The mere possession of monopoly power, and the concomitant charging of monopoly prices, is not only not unlawful; it is an important element of the free-market system.”).

24 United States v. Grinnell Corp., 384 U.S. 563, 570–71, (1966) (“The offense of monopoly under s 2 of the Sherman Act has two elements: (1) the possession of monopoly power in the relevant market and (2) the willful acquisition or maintenance of that power as distinguished from growth or development as a consequence of a superior product, business acumen, or historic accident.”).

25 See e.g. In re Google Digital Advertising Antitrust Litigation, No. 5:20-cv-003556-BLF, ECF No. 143 (N.D. Cal. May 13, 2021) (dismissing complaint for failure to allege anticompetitive conduct, and noting that even a monopolist generally does not have a duty to aid competitors).

26 Verizon Comms. Inc. v. Law Offices of Curtis V. Trinko, LLP, 540 U.S. 398 (2004).

27 United States v. Microsoft, 253 F.3d 34 (D.C. Cir. 2001).

28 As the Supreme Court explained in a related context, “Cutting prices in order to increase business often is the very essence of competition. Thus, mistaken inferences in cases such as this one are especially costly, because they chill the very conduct the antitrust laws are designed to protect.” Matsushita Elec. Indus. Co., Ltd. v. Zenith Radio Corp., 475 U.S. 574, 594 (1986).

29 United States v. Microsoft Corp., 253 F.3d 34, 68 (D.C. Cir. 2001) (“The rare case of price predation aside, the antitrust laws do not condemn even a monopolist for offering its product at an attractive price, and we therefore have no warrant to condemn Microsoft for offering either IE or the IEAK free of charge or even at a negative price. Likewise, as we said above, a monopolist does not violate the Sherman Act simply by developing an attractive product.”).

30 Gerry Greenstone, The History of Bloodletting, 52 BC Med. J. 12 (2010).

31 E.g. Senator Elizabeth Warren, Here’s How We Can Break Up Big Tech (Mar. 8, 2019), https://medium.com/@teamwarren/heres-how-we-can-break-up-big-tech-9ad9e0da324c; Senator Josh Hawley, Press Release, Senator Hawley Introduces the Bust Up Big Tech Act (Apr. 19, 2021).

32 E.g. U.S. Dep’t of Justice and Fed. Trade Comm’n, Vertical Merger Guidelines (“Vertical mergers combine complementary economic functions and eliminate contracting frictions, and therefore have the capacity to create a range of potentially cognizable efficiencies that benefit competition and consumers.”); Francine Lafontaine & Margaret Slade, Vertical Integration and Firm Boundaries: The Evidence, 45 J. of Econ. Lit. 629 (2007) (“[The weight of the evidence] says that, under most circumstances, profit-maximizing vertical-integration and merger decisions are efficient, not just from the firms’ but also from the consumers’ points of view.”); Joshua D. Wright, Douglas H. Ginsburg, Tad Lipsky, & John M. Yun, Connecting Vertical Merger Guidelines to Sound Economics, Truth on the Market Blog (Feb. 6, 2020), https://truthonthemarket.com/2020/02/06/wright-vmg-symposium/ (“With few exceptions, the literature does not support the view that [vertical mergers] are used for anticompetitive reasons. . . . [T]he empirical reality [is] that vertical relationships are generally procompetitive or neutral.”); D. Bruce Hoffman, Director, FTC Bureau of Competition, Vertical Merger Enforcement at the FTC, Remarks as Prepared for Delivery at Credit Suisse 2018 Washington Perspectives Conference (Jan. 10, 2018), available at https://www.ftc.gov/system/files/documents/public_statements/1304213/hoffman_vertical_merger_speech_final.pdf (“Moreover, while efficiencies are often important in horizontal mergers, they are much more intrinsic to a vertical transaction due to the cost-reducing effects of most vertical mergers, at least in the abstract. Due to the elimination of double-marginalization and the resulting downward pressure on prices, vertical mergers come with a more built-in likelihood of improving competition than horizontal mergers.”); Jon Sallet, Deputy Assistant Attorney General For Litigation, Antitrust Division, U.S. Department of Justice, The Interesting Case of the Vertical Merger, Remarks as Prepared for Delivery, ABA Fall Forum (Nov. 17, 2016) (“And here it’s worth emphasizing that vertical integration can create significant efficiencies that benefit suppliers, distributors, and consumers alike.”).

33 E.g. Fiona Scott Morton, et al., Report of the Committee for the Study of Digital Platforms Market Structure and Antitrust Subcommittee, George J. Stigler Center for the Study of the Economy and the State (July 1, 2019), at 97 (noting that while divestiture may be “the clear and simple remedy” for self-preferencing, because of the costs associated with this drastic approach, “a behavioral non-discrimination remedy might be more appropriate.”).

34 See e.g. Bust Up Big Tech Act, S. 1204, 117th Cong. (2021) (covering platforms with at least 30 million U.S. users or 300 million worldwide users and at least $1.5 billion in worldwide revenue); European Commission, Digital Markets Act: Ensuring Fair and Open Digital Markets (Dec. 15, 2020) (covering platforms with at least 6.5 billion euro EEA annual turnover or 65 billion euro market capitalization, and at least 45 million monthly EU users or 10,000 yearly EU business users); Mark MacCarthy, To Regulate Digital Platforms, Focus on Specific Business Sectors, Brookings Institution (Oct. 22, 2019) (proposing that regulators focus on social media over other types of platforms).

35 E.g. Library of Congress, Germany: New Digital Competition Act Expands Abilities of Competition Authorities to Regulate Abuse of Dominant Market Positions (Feb. 23, 2021) (describing the German Digital Competition Act, which imposes heightened regulation on digital platforms with a market share of at least 40 percent); Gene Kimmelman, The Right Way to Regulate Digital Platforms, Shorenstein Center on Media, Politics and Public Policy (Sept. 18, 2019) (proposing regulation for “dominant” platforms).

36 See e.g. U.S. Dep’t of Justice and Fed. Trade Comm’n, Commentary on the Horizontal Merger Guidelines, at 6 (describing how the FTC defined separate markets for superpremium, premium, and economy ice cream in its assessment of the Nestle/Dreyer’s merger).

37 Moreover, while too complex for this article, in vertical arrangements the potential for consumer benefit from market power tends to be isomorphic with the potential for foreclosure. Thus, generalized rules targeting market power in vertical relationships have a high potential for harming consumers and competition.