Archives For Tying

[The following is a guest post from Andrew Mercado, a research assistant at the Mercatus Center at George Mason University and an adjunct professor and research assistant at George Mason’s Antonin Scalia Law School.]

The Competition and Transparency in Digital Advertising Act (CTDAA), introduced May 19 by Sens. Mike Lee (R-Utah), Ted Cruz (R-Texas), Amy Klobuchar (D-Minn.), and Richard Blumenthal (D-Conn.), is the latest manifestation of the congressional desire to “do something” legislatively about big digital platforms. Although different in substance from the other antitrust bills introduced this Congress, it shares one key characteristic: it is fatally flawed and should not be enacted.  

Restrictions

In brief, the CTDAA imposes revenue-based restrictions on the ownership structure of firms engaged in digital advertising. The CTDAA bars a firm with more than $20 billion in annual advertising revenue (adjusted annually for inflation) from:

  1. owning a digital-advertising exchange if it owns either a sell-side ad brokerage or a buy-side ad brokerage; and
  2. owning a sell-side brokerage if it owns a buy-side brokerage, or from owning a buy-side or sell-side brokerage if it is also a buyer or seller of advertising space.

The proposal’s ownership restrictions present the clearest harm to the future of the digital-advertising market. From an efficiency perspective, vertical integration of both sides of the market can lead to enormous gains. Since, for example, Google owns and operates an ad exchange, a sell-side broker, and a buy-side broker, there are very few frictions that exist between each side of the market. All of the systems are integrated and the supply of advertising space, demand for that space, and the marketplace conducting price-discovery auctions are automatically updated in real time.

While this instantaneous updating is not unique to Google’s system, and other buy- and sell-side firms can integrate into the system, the benefit to advertisers and publishers can be found in the cost savings that come from the integration. Since Google is able to create synergies on all sides of the market, the fees on any given transaction are lower. Further, incorporating Google’s vast trove of data allows for highly relevant and targeted ads. All of this means that advertisers spend less for the same quality of ad; publishers get more for each ad they place; and consumers see higher-quality, more relevant ads.

Without the ability to own and invest in the efficiency and transaction-cost reduction of an integrated platform, there will likely be less innovation and lower quality on all sides of the market. Further, advertisers and publishers will have to shoulder the burden of using non-integrated marketplaces and would likely pay higher fees for less-efficient brokers. Since Google is a one-stop shop for all of a company’s needs—whether that be on the advertising side or the publishing side—companies can move seamlessly from one side of the market to the other, all while paying lower costs per transaction, because of the integrated nature of the platform.

In the absence of such integration, a company would have to seek out one buy-side brokerage to place ads and another, separate sell-side brokerage to receive ads. These two brokers would then have to go to an ad exchange to facilitate the deal, bringing three different brokers into the mix. Each of these middlemen would take a proportionate cut of the deal. When comparing the situation between an integrated and non-integrated market, the fees associated with serving ads in a non-integrated market are almost certainly higher.

Additionally, under this proposal, the innovative potential of each individual firm is capped. If a firm grows big enough and gains sufficient revenue through integrating different sides of the market, they will be forced to break up their efficiency-inducing operations. Marginal improvements on each side of the market may be possible, but without integrating different sides of the market, the scale required to justify those improvements would be insurmountable.

Assumptions

The CTDAA assumes that:

  1. there is a serious competitive problem in digital advertising; and
  2. the structural separation and regulation of advertising brokerages run by huge digital-advertising platforms (as specified in the CTDAA) would enhance competition and benefit digital advertising customers and consumers.

The first assumption has not been proven and is subject to debate, while the second assumption is likely to be false.

Fundamental to the bill’s assumption that the digital-advertising market lacks competition is a misunderstanding of competitive forces and the idea that revenue and profit are inversely related to competition. While it is true that high profits can be a sign of consolidation and anticompetitive outcomes, the dynamic nature of the internet economy makes this theory unlikely.

As Christopher Kaiser and I have discussed, competition in the internet economy is incredibly dynamic. Vigorous competition can be achieved with just a handful of firms,  despite claims from some quarters that four competitors is necessarily too few. Even in highly concentrated markets, there is the omnipresent threat that new entrants will emerge to usurp an incumbent’s reign. Additionally, while some studies may show unusually large profits in those markets, when adjusted for the consumer welfare created by large tech platforms, profits should actually be significantly higher than they are.

Evidence of dynamic entry in digital markets can be found in a recently announced product offering from a small (but more than $6 billion in revenue) competitor in digital advertising. Following the outcry associated with Google’s alleged abuse with Project Bernanke, the Trade Desk developed OpenPath. This allowed the Trade Desk, a buy-side broker, to handle some of the functions of a sell-side broker and eliminate harms from Google’s alleged bid-rigging to better serve its clients.

In developing the platform, the Trade Desk said it would discontinue serving any Google-based customers, effectively severing ties with the largest advertising exchange on the market. While this runs afoul of the letter of the law spelled out in CTDAA, it is well within the spirit its sponsor’s stated goal: businesses engaging in robust free-market competition. If Google’s market power was as omnipresent and suffocating as the sponsors allege, then eliminating traffic from Google would have been a death sentence for the Trade Desk.

While various theories of vertical and horizontal competitive harm have been put forward, there has not been an empirical showing that consumers and advertising customers have failed to benefit from the admittedly efficient aspects of digital-brokerage auctions administered by Google, Facebook, and a few other platforms. The rapid and dramatic growth of digital advertising and associated commerce strongly suggests that this has been an innovative and welfare-enhancing development. Moreover, the introduction of a new integrated brokerage platform by a “small” player in the advertising market indicates there is ample opportunity to increase this welfare further.  

Interfering in brokerage operations under the unproven assumption that “monopoly rents” are being charged and that customers are being “exploited” is rhetoric unmoored from hard evidence. Furthermore, if specific platform practices are shown inefficiently to exclude potential entrants, existing antitrust law can be deployed on a case-specific basis. This approach is currently being pursued by a coalition of state attorneys general against Google (the merits of which are not relevant to this commentary).   

Even assuming for the sake of argument that there are serious competition problems in the digital-advertising market, there is no reason to believe that the arbitrary provisions and definitions found in the CTDAA would enhance welfare. Indeed, it is likely that the act would have unforeseen consequences:

  • It would lead to divestitures supervised by the U.S. Justice Department (DOJ) that could destroy efficiencies derived from efficient targeting by brokerages integrated into platforms;
  • It would disincentivize improvements in advertising brokerages and likely would reduce future welfare on both the buy and sell sides of digital advertising;
  • It would require costly recordkeeping and disclosures by covered platforms that could have unforeseen consequences for privacy and potentially reduce the efficiency of bidding practices;
  • It would establish a fund for damage payments that would encourage wasteful litigation (see next two points);
  • It would spawn a great deal of wasteful private rent-seeking litigation that would discourage future platform and brokerage innovations; and
  • It would likely generate wasteful lawsuits by rent-seeking state attorneys general (and perhaps the DOJ as well).

The legislation would ultimately harm consumers who currently benefit from a highly efficient form of targeted advertising (for more on the welfare benefits of targeted advertising, see here). Since Google continually invests in creating a better search engine (to deliver ads directly to consumers) and collects more data to better target ads (to deliver ads to specific consumers), the value to advertisers of displaying ads on Google constantly increases.

Proposing a new regulatory structure that would directly affect the operations of highly efficient auction markets is the height of folly. It ignores the findings of Nobel laureate James M. Buchanan (among others) that, to justify regulation, there should first be a provable serious market failure and that, even if such a failure can be shown, the net welfare costs of government intervention should be smaller than the net welfare costs of non-intervention.

Given the likely substantial costs of government intervention and the lack of proven welfare costs from the present system (which clearly has been associated with a growth in output), the second prong of the Buchanan test clearly has not been met.

Conclusion

While there are allegations of abuses in the digital-advertising market, it is not at all clear that these abuses have had a long-term negative economic impact. As shown in a study by Erik Brynjolfsson and his student Avinash Collis—recently summarized in the Harvard Business Review (Alden Abbott offers commentary here)—the consumer surplus generated by digital platforms has far outstripped the advertising and services revenues received by the platforms. The CTDAA proposal would seek to unwind much of these gains.

If the goal is to create a multitude of small, largely inefficient advertising companies that charge high fees and provide low-quality service, this bill will deliver. The market for advertising will have a far greater number of players but it will be far less competitive, since no companies will be willing to exceed the $20 billion revenue threshold that would leave them subject to the proposal’s onerous ownership standards.

If, however, the goal is to increase consumer welfare, increase rigorous competition, and cement better outcomes for advertisers and publishers, then it is likely to fail. Ownership requirements laid out in the proposal will lead to a stagnant advertising market, higher fees for all involved, and lower-quality, less-relevant ads. Government regulatory interference in highly successful and efficient platform markets are a terrible idea.

We will learn more in the coming weeks about the fate of the proposed American Innovation and Choice Online Act (AICOA), legislation sponsored by Sens. Amy Klobuchar (D-Minn.) and Chuck Grassley (R-Iowa) that would, among other things, prohibit “self-preferencing” by large digital platforms like Google, Amazon, Facebook, Apple, and Microsoft. But while the bill has already been subject to significant scrutiny, a crucially important topic has been absent from that debate: the measure’s likely effect on startup acquisitions. 

Of course, AICOA doesn’t directly restrict startup acquisitions, but the activities it would restrict most certainly do dramatically affect the incentives that drive many startup acquisitions. If a platform is prohibited from engaging in cross-platform integration of acquired technologies, or if it can’t monetize its purchase by prioritizing its own technology, it may lose the motivation to make a purchase in the first place.

This would be a significant loss. As Dirk Auer, Sam Bowman, and I discuss in a recent article in the Missouri Law Review, acquisitions are arguably the most important component in providing vitality to the overall venture ecosystem:  

Startups generally have two methods for achieving liquidity for their shareholders: IPOs or acquisitions. According to the latest data from Orrick and Crunchbase, between 2010 and 2018 there were 21,844 acquisitions of tech startups for a total deal value of $1.193 trillion. By comparison, according to data compiled by Jay R. Ritter, a professor at the University of Florida, there were 331 tech IPOs for a total market capitalization of $649.6 billion over the same period. As venture capitalist Scott Kupor said in his testimony during the FTC’s hearings on “Competition and Consumer Protection in the 21st Century,” “these large players play a significant role as acquirers of venture-backed startup companies, which is an important part of the overall health of the venture ecosystem.”

Moreover, acquisitions by large incumbents are known to provide a crucial channel for liquidity in the venture capital and startup communities: While at one time the source of the “liquidity events” required to yield sufficient returns to fuel venture capital was evenly divided between IPOs and mergers, “[t]oday that math is closer to about 80 percent M&A and about 20 percent IPOs—[with important implications for any] potential actions that [antitrust enforcers] might be considering with respect to the large platform players in this industry.” As investor and serial entrepreneur Leonard Speiser said recently, “if the DOJ starts going after tech companies for making acquisitions, venture investors will be much less likely to invest in new startups, thereby reducing competition in a far more harmful way.” (emphasis added)

Going after self-preferencing may have exactly the same harmful effect on venture investors and competition. 

It’s unclear exactly how the legislation would be applied in any given context (indeed, this uncertainty is one of the most significant problems with the bill, as the ABA Antitrust Section has argued at length). But AICOA is designed, at least in part, to keep large online platforms in their own lanes—to keep them from “leveraging their dominance” to compete against more politically favored competitors in ancillary markets. Indeed, while covered platforms potentially could defend against application of the law by demonstrating that self-preferencing is necessary to “maintain or substantially enhance the core functionality” of the service, no such defense exists for non-core (whatever that means…) functionality, the enhancement of which through self-preferencing is strictly off limits under AICOA.

As I have written (and so have many, many, many, many others), this is terrible policy on its face. But it is also likely to have significant, adverse, indirect consequences for startup acquisitions, given the enormous number of such acquisitions that are outside the covered platforms’ “core functionality.” 

Just take a quick look at a sample of the largest acquisitions made by Apple, Microsoft, Amazon, and Alphabet, for example. (These are screenshots of the first several acquisitions by size drawn from imperfect lists collected by Wikipedia, but for purposes of casual empiricism they are well-suited to give an idea of the diversity of acquisitions at issue):

Apple:

Microsoft:

Amazon:

Alphabet (Google):

Vanishingly few of these acquisitions go to the “core functionalities” of these platforms. Alphabet’s acquisitions, for example, involve (among many other things) cybersecurity; home automation; cloud computing; wearables, smart glasses, and AR hardware; GPS navigation software; communications security; satellite technology; and social gaming. Microsoft’s acquisitions include companies specializing in video games; social networking; software versioning; drawing software; cable television; cybersecurity; employee engagement; and e-commerce. The technologies and applications involved in acquisitions by Apple and Amazon are similarly varied.

Drilling down a bit, consider the companies Alphabet acquired and put to use in the service of Google Maps:

Which, if any, of these companies would Google have purchased if it knew it would be unable to prioritize Maps in its search results? Would Google have invested more than $1 billion in these companies—and likely significantly more in internal R&D to develop Maps—if it had to speculate whether it would be required (or even be able) to prove someday in the future that prioritizing Google Maps results would enhance its core functionality?

What about Xbox? As noted, AICOA’s terms aren’t perfectly clear, so I’m not certain it would apply to Xbox (is Xbox a “website, online or mobile application, operating system, digital assistant, or online service”?). Here are Microsoft’s video-gaming-related purchases:

The vast majority of these (and all of the acquisitions for which Wikipedia has purchase-price information, totaling some $80 billion of investment) involve video games, not the development of hardware or the functionality of the Xbox platform. Would Microsoft have made these investments if it knew it would be prohibited from prioritizing its own games or exclusively using data gleaned through these games to improve its platform? No one can say for certain, but, at the margin, it is absolutely certain that these self-preferencing bills would make such acquisitions less likely.

Perhaps the most obvious—and concerning—example of the problem arises in the context of Google’s Android platform. Google famously gives Android away for free, of course, and makes its operating system significantly open for bespoke use by all comers. In exchange, Google requires that implementers of the Android OS provide some modicum of favoritism to Google’s revenue-generating products, like Search. For all its uncertainty, there is no question that AICOA’s terms would prohibit this self-preferencing. Intentionally or not, it would thus prohibit the way in which Google monetizes Android and thus hopes to recoup some of the—literally—billions of dollars it has invested in the development and maintenance of Android. 

Here are Google’s Android-related acquisitions:

Would Google have bought Android in the first place (to say nothing of subsequent acquisitions and its massive ongoing investment in Android) if it had been foreclosed from adopting its preferred business model to monetize its investment? In the absence of Google bidding for these companies, would they have earned as much from other potential bidders? Would they even have come into existence at all?

Of course, AICOA wouldn’t preclude Google charging device makers for Android and thus raising the price of mobile devices. But that mechanism may not have been sufficient to support Google’s investment in Android, and it would certainly constrain its ability to compete. Even if rules like those proposed by AICOA didn’t undermine Google’s initial purchase of and investment in Android, it is manifestly unclear how forcing Google to adopt a business model that increases consumer prices and constrains its ability to compete head-to-head with Apple’s iOS ecosystem would benefit consumers. (This excellent series of posts—1, 2, 3, 4—by Dirk Auer on the European Commission’s misguided Android decision discusses in detail the significant costs of prohibiting self-preferencing on Android.)

There are innumerable further examples, as well. In all of these cases, it seems clear not only that an AICOA-like regime would diminish competition and reduce consumer welfare across important dimensions, but also that it would impoverish the startup ecosystem more broadly. 

And that may be an even bigger problem. Even if you think, in the abstract, that it would be better for “Big Tech” not to own these startups, there is a real danger that putting that presumption into force would drive down acquisition prices, kill at least some tech-startup exits, and ultimately imperil the initial financing of tech startups. It should go without saying that this would be a troubling outcome. Yet there is no evidence to suggest that AICOA’s proponents have even considered whether the presumed benefits of the bill would be worth this immense cost.

Sens. Amy Klobuchar (D-Minn.) and Chuck Grassley (R-Iowa)—cosponsors of the American Innovation Online and Choice Act, which seeks to “rein in” tech companies like Apple, Google, Meta, and Amazon—contend that “everyone acknowledges the problems posed by dominant online platforms.”

In their framing, it is simply an acknowledged fact that U.S. antitrust law has not kept pace with developments in the digital sector, allowing a handful of Big Tech firms to exploit consumers and foreclose competitors from the market. To address the issue, the senators’ bill would bar “covered platforms” from engaging in a raft of conduct, including self-preferencing, tying, and limiting interoperability with competitors’ products.

That’s what makes the open letter to Congress published late last month by the usually staid American Bar Association’s (ABA) Antitrust Law Section so eye-opening. The letter is nothing short of a searing critique of the legislation, which the section finds to be poorly written, vague, and departing from established antitrust-law principles.

The ABA, of course, has a reputation as an independent, highly professional, and heterogenous group. The antitrust section’s membership includes not only in-house corporate counsel, but lawyers from nonprofits, consulting firms, federal and state agencies, judges, and legal academics. Given this context, the comments must be read as a high-level judgment that recent legislative and regulatory efforts to “discipline” tech fall outside the legal mainstream and would come at the cost of established antitrust principles, legal precedent, transparency, sound economic analysis, and ultimately consumer welfare.

The Antitrust Section’s Comments

As the ABA Antitrust Law Section observes:

The Section has long supported the evolution of antitrust law to keep pace with evolving circumstances, economic theory, and empirical evidence. Here, however, the Section is concerned that the Bill, as written, departs in some respects from accepted principles of competition law and in so doing risks causing unpredicted and unintended consequences.

Broadly speaking, the section’s criticisms fall into two interrelated categories. The first relates to deviations from antitrust orthodoxy and the principles that guide enforcement. The second is a critique of the AICOA’s overly broad language and ambiguous terminology.

Departing from established antitrust-law principles

Substantively, the overarching concern expressed by the ABA Antitrust Law Section is that AICOA departs from the traditional role of antitrust law, which is to protect the competitive process, rather than choosing to favor some competitors at the expense of others. Indeed, the section’s open letter observes that, out of the 10 categories of prohibited conduct spelled out in the legislation, only three require a “material harm to competition.”

Take, for instance, the prohibition on “discriminatory” conduct. As it stands, the bill’s language does not require a showing of harm to the competitive process. It instead appears to enshrine a freestanding prohibition of discrimination. The bill targets tying practices that are already prohibited by U.S. antitrust law, but while similarly eschewing the traditional required showings of market power and harm to the competitive process. The same can be said, mutatis mutandis, for “self-preferencing” and the “unfair” treatment of competitors.

The problem, the section’s letter to Congress argues, is not only that this increases the teleological chasm between AICOA and the overarching goals and principles of antitrust law, but that it can also easily lead to harmful unintended consequences. For instance, as the ABA Antitrust Law Section previously observed in comments to the Australian Competition and Consumer Commission, a prohibition of pricing discrimination can limit the extent of discounting generally. Similarly, self-preferencing conduct on a platform can be welfare-enhancing, while forced interoperability—which is also contemplated by AICOA—can increase prices for consumers and dampen incentives to innovate. Furthermore, some of these blanket prohibitions are arguably at loggerheads with established antitrust doctrine, such as in, e.g., Trinko, which established that even monopolists are generally free to decide with whom they will deal.

In response to the above, the ABA Antitrust Law Section (reasonably) urges Congress explicitly to require an effects-based showing of harm to the competitive process as a prerequisite for all 10 of the infringements contemplated in the AICOA. This also means disclaiming generalized prohibitions of “discrimination” and of “unfairness” and replacing blanket prohibitions (such as the one for self-preferencing) with measured case-by-case analysis.

Arguably, the reason why the Klobuchar-Grassley bill can so seamlessly exclude or redraw such a central element of antitrust law as competitive harm is because it deliberately chooses to ignore another, preceding one. Namely, the bill omits market power as a requirement for a finding of infringement or for the legislation’s equally crucial designation as a “covered platform.” It instead prescribes size metrics—number of users, market capitalization—to define which platforms are subject to intervention. Such definitions cast an overly wide net that can potentially capture consumer-facing conduct that doesn’t have the potential to harm competition at all.

It is precisely for this reason that existing antitrust laws are tethered to market power—i.e., because it long has been recognized that only companies with market power can harm competition. As John B. Kirkwood of Seattle University School of Law has written:

Market power’s pivotal role is clear…This concept is central to antitrust because it distinguishes firms that can harm competition and consumers from those that cannot.

In response to the above, the ABA Antitrust Law Section (reasonably) urges Congress explicitly to require an effects-based showing of harm to the competitive process as a prerequisite for all 10 of the infringements contemplated in the AICOA. This also means disclaiming generalized prohibitions of “discrimination” and of “unfairness” and replacing blanket prohibitions (such as the one for self-preferencing) with measured case-by-case analysis.

Opaque language for opaque ideas

Another underlying issue is that the Klobuchar-Grassley bill is shot through with indeterminate language and fuzzy concepts that have no clear limiting principles. For instance, in order either to establish liability or to mount a successful defense to an alleged violation, the bill relies heavily on inherently amorphous terms such as “fairness,” “preferencing,” and “materiality,” or the “intrinsic” value of a product. But as the ABA Antitrust Law Section letter rightly observes, these concepts are not defined in the bill, nor by existing antitrust case law. As such, they inject variability and indeterminacy into how the legislation would be administered.

Moreover, it is also unclear how some incommensurable concepts will be weighed against each other. For example, how would concerns about safety and security be weighed against prohibitions on self-preferencing or requirements for interoperability? What is a “core function” and when would the law determine it has been sufficiently “enhanced” or “maintained”—requirements the law sets out to exempt certain otherwise prohibited behavior? The lack of linguistic and conceptual clarity not only explodes legal certainty, but also invites judicial second-guessing into the operation of business decisions, something against which the U.S. Supreme Court has long warned.

Finally, the bill’s choice of language and recent amendments to its terminology seem to confirm the dynamic discussed in the previous section. Most notably, the latest version of AICOA replaces earlier language invoking “harm to the competitive process” with “material harm to competition.” As the ABA Antitrust Law Section observes, this “suggests a shift away from protecting the competitive process towards protecting individual competitors.” Indeed, “material harm to competition” deviates from established categories such as “undue restraint of trade” or “substantial lessening of competition,” which have a clear focus on the competitive process. As a result, it is not unreasonable to expect that the new terminology might be interpreted as meaning that the actionable standard is material harm to competitors.

In its letter, the antitrust section urges Congress not only to define more clearly the novel terminology used in the bill, but also to do so in a manner consistent with existing antitrust law. Indeed:

The Section further recommends that these definitions direct attention to analysis consistent with antitrust principles: effects-based inquiries concerned with harm to the competitive process, not merely harm to particular competitors

Conclusion

The AICOA is a poorly written, misguided, and rushed piece of regulation that contravenes both basic antitrust-law principles and mainstream economic insights in the pursuit of a pre-established populist political goal: punishing the success of tech companies. If left uncorrected by Congress, these mistakes could have potentially far-reaching consequences for innovation in digital markets and for consumer welfare. They could also set antitrust law on a regressive course back toward a policy of picking winners and losers.

The International Center for Law & Economics (ICLE) filed an amicus brief on behalf of itself and 26 distinguished law & economics scholars with the 9th U.S. Circuit Court of Appeals in the hotly anticipated and intensely important Epic Games v Apple case.

A fantastic group of attorneys from White & Case generously assisted us with the writing and filing of the brief, including George Paul, Jack Pace, Gina Chiapetta, and Nicholas McGuire. The scholars who signed the brief are listed at the end of this post. A summary of the brief’s arguments follows. For some of our previous writings on the case, see here, here, here, and here.

Introduction

In Epic Games v. Apple, Epic challenged Apple’s prohibition of third-party app stores and in-app payments (IAP) systems from operating on its proprietary iOS platform as a violation of antitrust law. The U.S. District Court for the Northern District of California ruled against Epic, finding that Epic’s real concern is its own business interests in the face of Apple’s business model—in particular, the commission Apple charges for use of its IAP system—rather than harm to consumers and to competition more broadly.

Epic appealed to the 9th Circuit on several grounds. Our brief primarily addresses two of Epic’s arguments:

  • First, Epic takes issue with the district court’s proper finding that Apple’s procompetitive justifications outweigh the anticompetitive effects of Apple’s business model. But Epic’s case fails at step one of the rule-of-reason analysis, as it didn’t demonstrate that Apple’s app distribution and IAP practices caused the significant, market-wide, anticompetitive effects that the Supreme Court, in 2018’s Ohio v. American Express (“Amex”), deemed necessary to show anticompetitive harm in cases involving two-sided transaction markets (like Apple’s App Store).
  • Second, Epic argues that the theoretical existence of less restrictive alternatives (“LRA”) to Apple’s business model is sufficient to meet its burden under the rule of reason. But the reliance on LRA in this case is misplaced. Forcing Apple to adopt the “open” platform that Epic champions would reduce interbrand competition and improperly permit antitrust plaintiffs to commandeer the judiciary to modify routine business conduct any time a plaintiff’s attorney or district court can imagine a less restrictive version of a challenged practice—irrespective of whether the practice promotes consumer welfare. This is especially true in the context of two-sided platform businesses, where such an approach would sacrifice interbrand, systems-level competition for the sake of a superficial increase in competition among a small subset of platform users.

Competitive Effects in Two-Sided Markets

Two-sided markets connect distinct sets of users whose demands for the platform are interdependent—i.e., consumers’ demand for a platform increases as more products are available, and conversely, product developers’ demand for a platform increases as additional consumers use the platform, increasing the overall potential for transactions. As a result of these complex dynamics, conduct that may appear anticompetitive when considering the effects on only one set of customers may be entirely consistent with—and actually promote—healthy competition when examining the effects on both sides.

That’s why the Supreme Court recognized in Amex that it was improper to focus on only one side of a two-sided platform. And this holding doesn’t require adherence to the Court’s contentious finding of a two-sided relevant market in Amex. Indeed, even scholars highly critical of the Amex decision recognize the importance of considering effects on both sides of a two-sided platform.

While the district court did find that Epic demonstrated some anticompetitive effects, Epic’s evidence focused only on the effects that Apple’s conduct had on certain app developers; it failed to appropriately examine whether consumers were harmed overall. As Geoffrey Manne has observed, in two-sided markets, “some harm” is not the same thing as “competitively relevant harm.” Supracompetitive prices on one side do not tell us much about the existence or exercise of (harmful) market power in two-sided markets. As the Supreme Court held in Amex:

The fact that two-sided platforms charge one side a price that is below or above cost reflects differences in the two sides’ demand elasticity, not market power or anticompetitive pricing. Price increases on one side of the platform likewise do not suggest anticompetitive effects without some evidence that they have increased the overall cost of the platform’s services.

Without further evidence of the effect of Apple’s practices on consumers, no conclusions can be drawn about the competitive effects of Apple’s conduct. 

Nor can an appropriate examination of anticompetitive effects ignore output. The ability to restrict output, after all, is what allows a monopolist to increase prices. Whereas price effects alone might appear predatory on one side of the market and supra-competitive on the other, output reflects what is happening in the market as a whole. It is therefore the most appropriate measure for antitrust law generally, and it is especially useful in two-sided markets, where asymmetrical price changes are of little use in determining anticompetitive effects.

Ultimately, the question before the court must be whether Apple’s overall pricing structure and business model reduces output, either by deterring app developers from participating in the market or by deterring users from purchasing apps (or iOS devices) as a consequence of the app-developer commission. The district court here noted that it could not ascertain whether Apple’s alleged restrictions had a “positive or negative impact on game transaction volume.”

Thus, Epic’s case fails at step one of the rule of reason analysis because it simply hasn’t demonstrated the requisite harm to competition.

Less Restrictive Alternatives and the Rule of Reason

But even if that weren’t the case, Epic’s claims also don’t make it past step three of the rule of reason analysis.

Epic’s appeal relies on theoretical “less restrictive alternatives” (LRA) to Apple’s business model, which highlights longstanding questions about the role and limits of LRA analysis under the rule of reason. 

According to Epic, because the district court identified some anticompetitive effects on one side of the market, and because alternative business models could, in theory, be implemented to achieve the same procompetitive benefits as Apple’s current business model, the court should have ruled in Epic’s favor at step three. 

There are several problems with this.

First, the existence of an LRA is irrelevant if anticompetitive harm has not been established, of course (as is the case here).

Nor does the fact that some hypothetically less restrictive alternative exists automatically render the conduct under consideration anticompetitive. As the Court held in Trinko, antitrust laws do not “give judges carte blanche to insist that a monopolist alter its way of doing business whenever some other approach might yield greater competition.” 

While, following the Supreme Court’s recent Alston decision, LRA analysis may well be appropriate in some contexts to identify anticompetitive conduct in the face of procompetitive justifications, there is no holding (in either the 9th Circuit or the Supreme Court) requiring it in the context of two-sided markets. (Amex refers to LRA analysis as constituting step three, but because that case was resolved at step one, it must be viewed as mere dictum).And for good reason. In the context of two-sided platforms, an LRA approach would inevitably require courts to second guess the particular allocation of costs, prices, and product attributes across platform users. As Tom Nachbar writes:

Platform defendants, even if they are able to establish the general procompetitive justifications for charging above and below cost prices on the two sides of their platforms, will have to defend the precise combination of prices they have chosen [under an LRA approach] . . . . The relative difficulty of defending any particular allocation of costs will present considerable risk of destabilizing platform markets.

Moreover, LRAs—like the ones proposed by Epic—that are based on maximizing competitor effectiveness by “opening” an incumbent’s platform would convert the rule of reason into a regulatory tool that may not promote competition at all. As Alan Devlin deftly puts it:

This construction of antitrust law—that dominant companies must affirmatively support their fringe rivals’ ability to compete effectively—adopts a perspective of antitrust that is regulatory in nature. . . . [I]f one adopts the increasingly prevalent view that antitrust must facilitate unfettered access to markets, thus spurring free entry and expansion by incumbent rivals, the Sherman Act goes from being a prophylactic device aimed at protecting consumers against welfare-reducing acts to being a misplaced regulatory tool that potentially sacrifices both consumer welfare and efficiency in a misguided pursuit of more of both.

Open Platforms Are not Necessarily Less Restrictive Platforms

It is also important to note that Epic’s claimed LRAs are neither viable alternatives nor actually “less restrictive.” Epic’s proposal would essentially turn Apple’s iOS into an open platform more similar to Google’s Android, its largest market competitor.

“Open” and “closed” platforms both have distinct benefits and drawbacks; one is not inherently superior to the other. Closed proprietary platforms like Apple’s iOS create incentives for companies to internalize positive indirect network effects, which can lead to higher levels of product variety, user adoption, and total social welfare. As Andrei Hagiu has written:

A proprietary platform may in fact induce more developer entry (i.e., product variety), user adoption and higher total social welfare than an open platform.

For example, by filtering which apps can access the App Store and precluding some transactions from taking place on it, a closed or semi-closed platform like Apple’s may ultimately increase the number of apps and transactions on its platform, where doing so makes the iOS ecosystem more attractive to both consumers and developers. 

Any analysis of a supposedly less restrictive alternative to Apple’s “walled garden” model thus needs to account for the tradeoffs between open and closed platforms, and not merely assume that “open” equates to “good,” and “closed” to “bad.” 

Further, such analysis also must consider tradeoffs among consumers and among developers. More vigilant users might be better served by an “open” platform because they find it easier to avoid harmful content; less vigilant ones may want more active assistance in screening for malware, spyware, or software that simply isn’t optimized for the user’s device. There are similar tradeoffs on the developer side: Apple’s model lowers the cost to join the App store, which particularly benefits smaller developers and those whose apps fall outside the popular gaming sector. In a nutshell, the IAP fee cross-subsidizes the delivery of services to the approximately 80% of apps on the App Store that are free and pay no IAP fees.

In fact, the overwhelming irony of Epic’s proposed approach is that Apple could avoid condemnation if it made its overall platform more restrictive. If, for example, Apple had not adopted an App Store model and offered a completely closed and fully integrated device, there would be no question of relative costs and benefits imposed on independent app developers; there would be no independent developers on the iOS platform at all. 

Thus, Epic’s proposed LRA approach, which amounts to converting iOS to an open platform, proves too much. It would enable any contractual or employment relationship for a complementary product or service to be challenged because it could be offered through a “less restrictive” open market mechanism—in other words, that any integrated firm should be converted into an open platform. 

At least since the Supreme Court’s seminal 1977 Sylvania ruling, U.S. antitrust law has been unequivocal in its preference for interbrand over intrabrand competition. Paradoxically, turning a closed platform into an open one (as Epic intends) would, under the guise of protecting competition, actually destroy competition where it matters most: at the interbrand, systems level.

Conclusion

Forcing Apple to adopt the “open” platform that Epic champions would reduce interbrand competition among platform providers. It would also more broadly allow antitrust plaintiffs to insist the courts modify routine business conduct any time a plaintiff’s attorney or district court can imagine a less restrictive version of a challenged practice, regardless of whether that practice nevertheless promotes consumer welfare. In the context of two-sided platform businesses, this would mean sacrificing systems-level competition for the sake of a superficial increase in competition among a small subset of platform users.

The bottom line is that an order compelling Apple to allow competing app stores would require the company to change the way in which it monetizes the App Store. This might have far-reaching distributional consequences for both groups— consumers and distributors. Courts (and, obviously, competitors) are ill-suited to act as social planners and to balance out such complex tradeoffs, especially in the absence of clear anticompetitive harm and the presence of plausible procompetitive benefits.

Amici Scholars Signing on to the Brief


(The ICLE brief presents the views of the individual signers listed below. Institutions are listed for identification purposes only.)

Alden Abbott
Senior Research Fellow, Mercatus Center, George Mason University
Former General Counsel, U.S. Federal Trade Commission
Ben Klein
Professor of Economics Emeritus, University of California Los Angeles
Thomas C. Arthur
L. Q. C. Lamar Professor of Law, Emory University School of Law
Peter Klein
Professor of Entrepreneurship and Corporate Innovation, Baylor University, Hankamer School of Business
Dirk Auer
Director of Competition Policy, International Center for Law & Economics
Adjunct Professor, University of Liège (Belgium)
Jonathan Klick
Charles A. Heimbold, Jr. Professor of Law, University of Pennsylvania Carey Law School
Jonathan M. Barnett
Torrey H. Webb Professor of Law, University of Southern California, Gould School of Law
Daniel Lyons
Professor of Law, Boston College Law School
Donald J. Boudreaux
Professor of Economics, former Economics Department Chair, George Mason University
Geoffrey A. Manne
President and Founder, International Center for Law & Economics
Distinguished Fellow, Northwestern University Center on Law, Business & Economics
Giuseppe Colangelo
Jean Monnet Chair in European Innovation Policy and Associate Professor of Competition Law and Economics, University of Basilicata and Libera Università Internazionale degli Studi Sociali
Francisco Marcos
Associate Professor of Law, IE University Law School (Spain)
Anthony Dukes
Chair and Professor of Marketing, University of Southern California, Marshall School of Business
Scott E. Masten
Professor of Business Economics and Public Policy, University of Michigan, Ross Business School
Richard A. Epstein
Laurence A. Tisch Professor of Law, New York University, School of Law James Parker Hall Distinguished Service Professor of Law Emeritus, University of Chicago Law School
Alan J. Meese
Ball Professor of Law, College of William & Mary Law School
Vivek Ghosal
Economics Department Chair and Virginia and Lloyd W. Rittenhouse Professor of Economics, Rensselaer Polytechnic Institute
Igor Nikolic
Research Fellow, Robert Schuman Centre for Advanced Studies, European University Institute (Italy)
Janice Hauge
Professor of Economics, University of North Texas
Paul H. Rubin
Samuel Candler Dobbs Professor of Economics Emeritus, Emory University
Justin (Gus) Hurwitz
Professor of Law, University of Nebraska College of Law
Vernon L. Smith
George L. Argyros Endowed Chair in Finance and Economics and Professor of Economics and Law, Chapman University Nobel Laureate in Economics (2002)
Michael S. Jacobs
Distinguished Research Professor of Law Emeritus, DePaul University College of Law
Michael Sykuta
Associate Professor of Economics, University of Missouri
Mark A. Jamison
Gerald Gunter Professor of the Public Utility Research Center, University of Florida, Warrington College of Business
Alexander “Sasha” Volokh
Associate Professor of Law, Emory University School of Law

On March 31, I and several other law and economics scholars filed an amicus brief in Epic Games v. Apple, which is on appeal to the U.S. Court of Appeals for Ninth Circuit.  In this post, I summarize the central arguments of the brief, which was joined by Alden Abbott, Henry Butler, Alan Meese, Aurelien Portuese, and John Yun and prepared with the assistance of Don Falk of Schaerr Jaffe LLP.

First, some background for readers who haven’t followed the case.

Epic, maker of the popular Fortnite video game, brought antitrust challenges against two policies Apple enforces against developers of third-party apps that run on iOS, the mobile operating system for Apple’s popular iPhones and iPads.  One policy requires that all iOS apps be distributed through Apple’s own App Store.  The other requires that any purchases of digital goods made while using an iOS app utilize Apple’s In App Purchase system (IAP).  Apple collects a share of the revenue from sales made through its App Store and using IAP, so these two policies provide a way for it to monetize its innovative app platform.   

Epic maintains that Apple’s app policies violate the federal antitrust laws.  Following a trial, the district court disagreed, though it condemned another of Apple’s policies under California state law.  Epic has appealed the antitrust rulings against it. 

My fellow amici and I submitted our brief in support of Apple to draw the Ninth Circuit’s attention to a distinction that is crucial to ensuring that antitrust promotes long-term consumer welfare: the distinction between the mere extraction of surplus through the exercise of market power and the enhancement of market power via the weakening of competitive constraints.

The central claim of our brief is that Epic’s antitrust challenges to Apple’s app store policies should fail because Epic has not shown that the policies enhance Apple’s market power in any market.  Moreover, condemnation of the practices would likely induce Apple to use its legitimately obtained market power to extract surplus in a different way that would leave consumers worse off than they are under the status quo.   

Mere Surplus Extraction vs. Market Power Extension

As the Supreme Court has observed, “Congress designed the Sherman Act as a ‘consumer welfare prescription.’”  The Act endeavors to protect consumers from harm resulting from “market power,” which is the ability of a firm lacking competitive constraints to enhance its profits by reducing its output—either quantitively or qualitatively—from the level that would persist if the firm faced vigorous competition.  A monopolist, for example, might cut back on the quantity it produces (to drive up market price) or it might skimp on quality (to enhance its per-unit profit margin).  A firm facing vigorous competition, by contrast, couldn’t raise market price simply by reducing its own production, and it would lose significant sales to rivals if it raised its own price or unilaterally cut back on product quality.  Market power thus stems from deficient competition.

As Dennis Carlton and Ken Heyer have observed, two different types of market power-related business behavior may injure consumers and are thus candidates for antitrust prohibition.  One is an exercise of market power: an action whereby a firm lacking competitive constraints increases its returns by constricting its output so as to raise price or otherwise earn higher profit margins.  When a firm engages in this sort of conduct, it extracts a greater proportion of the wealth, or “surplus,” generated by its transactions with its customers.

Every voluntary transaction between a buyer and seller creates surplus, which is the difference between the subjective value the consumer attaches to an item produced and the cost of producing and distributing it.  Price and other contract terms determine how that surplus is allocated between the buyer and the seller.  When a firm lacking competitive constraints exercises its market power by, say, raising price, it extracts for itself a greater proportion of the surplus generated by its sale.

The other sort of market power-related business behavior involves an effort by a firm to enhance its market power by weakening competitive constraints.  For example, when a firm engages in unreasonably exclusionary conduct that drives its rivals from the market or increases their costs so as to render them less formidable competitors, its market power grows.

U.S. antitrust law treats these two types of market power-related conduct differently.  It forbids behavior that enhances market power and injures consumers, but it permits actions that merely exercise legitimately obtained market power without somehow enhancing it.  For example, while charging a monopoly price creates immediate consumer harm by extracting for the monopolist a greater share of the surplus created by the transaction, the Supreme Court observed in Trinko that “[t]he mere possession of monopoly power, and the concomitant charging of monopoly prices, is not . . . unlawful.”  (See also linkLine: “Simply possessing monopoly power and charging monopoly prices does not violate [Sherman Act] § 2….”)

Courts have similarly refused to condemn mere exercises of market power in cases involving surplus-extractive arrangements more complicated than simple monopoly pricing.  For example, in its Independent Ink decision, the U.S. Supreme Court expressly declined to adopt a rule that would have effectively banned “metering” tie-ins.

In a metering tie-in, a seller with market power on some unique product that is used with a competitively supplied complement that is consumed in varying amounts—say, a highly unique printer that uses standard ink—reduces the price of its unique product (the printer), requires buyers to also purchase from it their requirements of the complement (the ink), and then charges a supracompetitive price for the latter product.  This allows the seller to charge higher effective prices to high-volume users of its unique tying product (buyers who use lots of ink) and lower prices to lower-volume users. 

Assuming buyers’ use of the unique product correlates with the value they ascribe to it, a metering tie-in allows the seller to price discriminate, charging higher prices to buyers who value its unique product more.  This allows the seller to extract more of the surplus generated by sales of its product, but it in no way extends the seller’s market power.

In refusing to adopt a rule that would have condemned most metering tie-ins, the Independent Ink Court observed that “it is generally recognized that [price discrimination] . . . occurs in fully competitive markets” and that tying arrangements involving requirements ties may be “fully consistent with a free, competitive market.” The Court thus reasoned that mere price discrimination and surplus extraction, even when accomplished through some sort of contractual arrangement like a tie-in, are not by themselves anticompetitive harms warranting antitrust’s condemnation.    

The Ninth Circuit has similarly recognized that conduct that exercises market power to extract surplus but does not somehow enhance that power does not create antitrust liability.  In Qualcomm, the court refused to condemn the chipmaker’s “no license, no chips” policy, which enabled it to enhance its profits by earning royalties on original equipment manufacturers’ sales of their high-priced products.

In reversing the district court’s judgment in favor of the FTC, the Ninth Circuit conceded that Qualcomm’s policies were novel and that they allowed it to enhance its profits by extracting greater surplus.  The court refused to condemn the policies, however, because they did not injure competition by weakening competitive constraints:

This is not to say that Qualcomm’s “no license, no chips” policy is not “unique in the industry” (it is), or that the policy is not designed to maximize Qualcomm’s profits (Qualcomm has admitted as much). But profit-seeking behavior alone is insufficient to establish antitrust liability. As the Supreme Court stated in Trinko, the opportunity to charge monopoly prices “is an important element of the free-market system” and “is what attracts ‘business acumen’ in the first place; it induces risk taking that produces innovation and economic growth.”

The Qualcomm court’s reference to Trinko highlights one reason courts should not condemn exercises of market power that merely extract surplus without enhancing market power: allowing such surplus extraction furthers dynamic efficiency—welfare gain that accrues over time from the development of new and improved products and services.

Dynamic efficiency results from innovation, which entails costs and risks.  Firms are more willing to incur those costs and risks if their potential payoff is higher, and an innovative firm’s ability to earn supracompetitive profits off its “better mousetrap” enhances its payoff. 

Allowing innovators to extract such profits also helps address the fact most of the benefits of product innovation inure to people other than the innovator.  Private actors often engage in suboptimal levels of behaviors that produce such benefit spillovers, or “positive externalities,”  because they bear all the costs of those behaviors but capture just a fraction of the benefit produced.  By enhancing the benefits innovators capture from their innovative efforts, allowing non-power-enhancing surplus extraction helps generate a closer-to-optimal level of innovative activity.

Not only do supracompetitive profits extracted through the exercise of legitimately obtained market power motivate innovation, they also enable it by helping to fund innovative efforts.  Whereas businesses that are forced by competition to charge prices near their incremental cost must secure external funding for significant research and development (R&D) efforts, firms collecting supracompetitive returns can finance R&D internally.  Indeed, of the top fifteen global spenders on R&D in 2018, eleven were either technology firms accused of possessing monopoly power (#1 Apple, #2 Alphabet/Google, #5 Intel, #6 Microsoft, #7 Apple, and #14 Facebook) or pharmaceutical companies whose patent protections insulate their products from competition and enable supracompetitive pricing (#8 Roche, #9 Johnson & Johnson, #10 Merck, #12 Novartis, and #15 Pfizer).

In addition to fostering dynamic efficiency by motivating and enabling innovative efforts, a policy acquitting non-power-enhancing exercises of market power allows courts to avoid an intractable question: which instances of mere surplus extraction should be precluded?

Precluding all instances of surplus extraction by firms with market power would conflict with precedents like Trinko and linkLine (which say that legitimate monopolists may legally charge monopoly prices) and would be impracticable given the ubiquity of above-cost pricing in niche and brand-differentiated markets.

A rule precluding surplus extraction when accomplished by a practice more complicated that simple monopoly pricing—say, some practice that allows price discrimination against buyers who highly value a product—would be both arbitrary and backward.  The rule would be arbitrary because allowing supracompetitive profits from legitimately obtained market power motivates and enables innovation regardless of the means used to extract surplus. The rule would be backward because, while simple monopoly pricing always reduces overall market output (as output-reduction is the very means by which the producer causes price to rise), more complicated methods of extracting surplus, such as metering tie-ins, often enhance market output and overall social welfare.

A third possibility would be to preclude exercising market power to extract more surplus than is necessary to motivate and enable innovation.  That position, however, would require courts to determine how much surplus extraction is required to induce innovative efforts.  Courts are poorly positioned to perform such a task, and their inevitable mistakes could significantly chill entrepreneurial activity.

Consider, for example, a firm contemplating a $5 million investment that might return up to $50 million.  Suppose the managers of the firm weighed expected costs and benefits and decided the risky gamble was just worth taking.  If the gamble paid off but a court stepped in and capped the firm’s returns at $20 million—a seemingly generous quadrupling of the firm’s investment—future firms in the same position would not make similar investments.  After all, the firm here thought this gamble was just barely worth taking, given the high risk of failure, when available returns were $50 million.

In the end, then, the best policy is to draw the line as both the U.S. Supreme Court and the Ninth Circuit have done: Whereas enhancements of market power are forbidden, merely exercising legitimately obtained market power to extract surplus is permitted.

Apple’s Policies Do Not Enhance Its Market Power

Under the legal approach described above, the two Apple policies Epic has challenged do not give rise to antitrust liability.  While the policies may boost Apple’s profits by facilitating its extraction of surplus from app transactions on its mobile devices, they do not enhance Apple’s market power in any conceivable market.

As the creator and custodian of the iOS operating system, Apple has the ability to control which applications will run on its iPhones and iPads.  Developers cannot produce operable iOS apps unless Apple grants them access to the Application Programming Interfaces (APIs) required to enable the functionality of the operating system and hardware. In addition, Apple can require developers to obtain digital certificates that will enable their iOS apps to operate.  As the district court observed, “no certificate means the code will not run.”

Because Apple controls which apps will work on the operating system it created and maintains, Apple could collect the same proportion of surplus it currently extracts from iOS app sales and in-app purchases on iOS apps even without the policies Epic is challenging.  It could simply withhold access to the APIs or digital certificates needed to run iOS apps unless developers promised to pay it 30% of their revenues from app sales and in-app purchases of digital goods.

This means that the challenged policies do not give Apple any power it doesn’t already possess in the putative markets Epic identified: the markets for “iOS app distribution” and “iOS in-app payment processing.” 

The district court rejected those market definitions on the ground that Epic had not established cognizable aftermarkets for iOS-specific services.  It defined the relevant market instead as “mobile gaming transactions.”  But no matter.  The challenged policies would not enhance Apple’s market power in that broader market either.

In “mobile gaming transactions” involving non-iOS (e.g., Android) mobile apps, Apple’s policies give it no power at all.  Apple doesn’t distribute non-iOS apps or process in-app payments on such apps.  Moreover, even if Apple were to being doing so—say, by distributing Android apps in its App Store or allowing producers of Android apps to include IAP as their in-app payment system—it is implausible that Apple’s policies would allow it to gain new market power.  There are giant, formidable competitors in non-iOS app distribution (e.g., Google’s Play Store) and in payment processing for non-iOS in-app purchases (e.g., Google Play Billing).  It is inconceivable that Apple’s policies would allow it to usurp so much scale from those rivals that Apple could gain market power over non-iOS mobile gaming transactions.

That leaves only the iOS segment of the mobile gaming transactions market.  And, as we have just seen, Apple’s policies give it no new power to extract surplus from those transactions; because it controls access to iOS, it could do so using other means.

Nor do the challenged policies enable Apple to maintain its market power in any conceivable market.  This is not a situation like Microsoft where a firm in a market adjacent to a monopolist’s could somehow pose a challenge to that monopolist, and the monopolist nips the potential competition in the bud by reducing the potential rival’s scale.  There is no evidence in the record to support the (implausible) notion that rival iOS app stores or in-app payment processing systems could ever evolve in a manner that would pose a challenge to Apple’s position in mobile devices, mobile operating systems, or any other market in which it conceivably has market power. 

Epic might retort that but for the challenged policies, rivals could challenge Apple’s market share in iOS app distribution and in-app purchase processing.  Rivals could not, however, challenge Apple’s market power in such markets, as that power stems from its control of iOS.  The challenged policies therefore do not enable Apple to shore up any existing market power.

Alternative Means of Extracting Surplus Would Likely Reduce Consumer Welfare

Because the policies Epic has challenged are not the source of Apple’s ability to extract surplus from iOS app transactions, judicial condemnation of the policies would likely induce Apple to extract surplus using different means.  Changing how it earns profits off iOS app usage, however, would likely leave consumers worse off than they are under the status quo.

Apple could simply charge third-party app developers a flat fee for access to the APIs needed to produce operable iOS apps but then allow them to distribute their apps and process in-app payments however they choose.  Such an approach would allow Apple to monetize its innovative app platform while permitting competition among providers of iOS app distribution and in-app payment processing services.  Relative to the status quo, though, such a model would likely reduce consumer welfare by:

  • Reducing the number of free and niche apps,as app developers could no longer avoid a fee to Apple by adopting a free (likely ad-supported) business model, and producers of niche apps may not generate enough revenue to justify Apple’s flat fee;
  • Raising business risks for app developers, who, if Apple cannot earn incremental revenue off sales and use of their apps, may face a greater likelihood that the functionality of those apps will be incorporated into future versions of iOS;
  • Reducing Apple’s incentive to improve iOS and its mobile devices, as eliminating Apple’s incremental revenue from app usage reduces its motivation to make costly enhancements that keep users on their iPhones and iPads;
  • Raising the price of iPhones and iPads and generating deadweight loss, as Apple could no longer charge higher effective prices to people who use apps more heavily and would thus likely hike up its device prices, driving marginal consumers from the market; and
  • Reducing user privacy and security, as jettisoning a closed app distribution model (App Store only) would impair Apple’s ability to screen iOS apps for features and bugs that create security and privacy risks.

An alternative approach—one that would avoid many of the downsides just stated by allowing Apple to continue earning incremental revenue off iOS app usage—would be for Apple to charge app developers a revenue-based fee for access to the APIs and other amenities needed to produce operable iOS apps.  That approach, however, would create other costs that would likely leave consumers worse off than they are under the status quo.

The policies Epic has challenged allow Apple to collect a share of revenues from iOS app transactions immediately at the point of sale.  Replacing those policies with a revenue-based  API license system would require Apple to incur additional costs of collecting revenues and ensuring that app developers are accurately reporting them.  In order to extract the same surplus it currently collects—and to which it is entitled given its legitimately obtained market power—Apple would have to raise its revenue-sharing percentage above its current commission rate to cover its added collection and auditing costs.

The fact that Apple has elected not to adopt this alternative means of collecting the revenues to which it is entitled suggests that the added costs of moving to the alternative approach (extra collection and auditing costs) would exceed any additional consumer benefit such a move would produce.  Because Apple can collect the same revenue percentage from app transactions two different ways, it has an incentive to select the approach that maximizes iOS app transaction revenues.  That is the approach that creates the greatest value for consumers and also for Apple. 

If Apple believed that the benefits to app users of competition in app distribution and in-app payment processing would exceed the extra costs of collection and auditing, it would have every incentive to switch to a revenue-based licensing regime and increase its revenue share enough to cover its added collection and auditing costs.  As such an approach would enhance the net value consumers receive when buying apps and making in-app purchases, it would raise overall app revenues, boosting Apple’s bottom line.  The fact that Apple has not gone in this direction, then, suggests that it does not believe consumers would receive greater benefit under the alternative system.  Apple might be wrong, of course.  But it has a strong motivation to make the consumer welfare-enhancing decision here, as doing so maximizes its own profits.

The policies Epic has challenged do not enhance or shore up Apple’s market power, a salutary pre-requisite to antitrust liability.  Furthermore, condemning the policies would likely lead Apple to monetize its innovative app platform in a manner that would reduce consumer welfare relative to the status quo.  The Ninth Circuit should therefore affirm the district court’s rejection of Epic’s antitrust claims.  

The Biden administration finally has taken a public position on parallel House (H.R. 3816) and Senate (S. 2992) bills that would impose new welfare-reducing regulatory constraints on the ability of large digital platforms to engage in innovative business practices that benefit consumers and the economy.

The administration’s articulation of its position—set forth in a March 28 U.S. Justice Department (DOJ letter to House and Senate Judiciary Committee leadership—is a fine example of draftsmanship. With just a few very minor redline edits, which I suggest below, the letter would advance sound and enlightened procompetitive policy.

I hope the DOJ will accept my modest redlines and incorporate them into a new letter to Congress, superseding the March 28 draft. My edited redline and clean revisions of the current draft follow (redline draft is in italics, clean draft in bold italics):

Redline Version

Dear Chairman Nadler, Chairman Cicilline, Representative Jordan, and Representative Buck:

The Department of Justice (Department) appreciates the considerable attention and resources devoted by the House and Senate Committees on the Judiciary over the past several years to ensuring the competitiveness of our digital economy, and writes today to express support for oppose the American Innovation and Choice Online Act, Senate bill S. 2992, and the American Innovation and Choice Online Act, House bill H.R. 3816, which contain similar prohibitions on discriminatory conduct by dominant platforms (the “bills”). Unfortunately, the legislative efforts expended on these bills have been a waste of time.

The Department views the rise of major digitaldominant platforms as presenting a great boon tothreat to open markets and competition, with bestowing benefits onrisks for consumers, businesses, innovation, resiliency, global competitiveness, and our democracy. By enhancing value controllingtransmitted through key arteries of the nation’s commerce and communications, such platforms have promoted a more vibrant and innovative can exercise outsized market power in our modern economy. Vesting in government the power to pick winners and losers across markets through legislative regulation as found in the bills in a small number of corporations contravenes the foundations of our capitalist system, and given the increasing importance of these markets, the economic benefits flowing frompower of such platforms activity areis likely to be curtailed if the bills are passed continue to grow unless checked. Enactment of the bills wouldThis puts at risk the nation’s economic progress and prosperity, ultimately threatening the economic liberty that undergirds our democracy.

The legislation, if enacted, would emphasize causes of action prohibiting the largest digital platforms from discriminating in favor of their own products or services, or among third parties. In so doing, it would eliminate and disincentivize many provide important clarification from Congress on types of discriminatory conduct efficient business arrangements that can materially enhanceharm competition. This would thereby undermineimprove upon the system of ex ante enforcement through which the United States maintains competitive markets with legal prohibitions on competitively harmful corporate conduct. By mistakenly characterizing confirming the illegality of as anticompetitive platform behaviors that in reality enhancereduce incentives for vigorous innovation and dynamic competition, smaller or newer firms to innovate and compete, the legislation would underminesupplement the existing antitrust laws. Specifically, the legislation would  in preventing the largest digital companies from managing their business transactions in an efficient welfare-enhancing manner, abusing and exploiting their dominant positions to the detriment of competition and the competitive process. The Department is strongly concerned aboutsupportive of these harmful effects.objectives and As such, it encourages both the Committees and Congress to work to abandon all efforts to finalize this legislationfinalize this legislation[1] and pass it into law.

The Department views the legislation’s new prohibitions on discrimination as a harmful detrimenthelpful complement to, and interference withclarification of, existing antitrust authority. In our view, the most significant harmbenefits would arise where the legislation seeks to elucidates Congress’ views of anticompetitive conduct—particularly with respect to harmful types of discrimination and self-preferencing by dominant platforms. Enumerating specific  discriminatory and self-preferencing conduct that Congress views as anticompetitive and therefore illegal would undermine the economically informed, fact-specific evaluation of business conduct that lies at the heart of modern antitrust analysis, centered on consumer welfare. Modern economic analysis demonstrates that a great deal of superficially “discriminatory” and “self-preferencing” conduct often represents consumer welfare-enhancing behavior that adds to economic surplus. Deciding whether such conduct is procompetitive (welfare-enhancing) or anticompetitive in a particular instance is the role of existing case-by-case antitrust enforcement. This approach vindicates competition while avoiding the wrongful condemnation of economically beneficial behavior. In contrast, by creating a new class of antitrust “wrongs,” the bills would lead to the incorrect condemnation of many business practices that enhance market efficiency and strengthen the economy.clarify the antitrust laws and supplement the available causes of action and legal frameworks to pursue that conduct. Doing so would enhance the ability of the DOJ and FTC to challenge that conduct efficiently and effectively and better enable them to promote competition in digital markets. The legislation also has the potential to effectively harmonize broad prohibitions with the particularized needs and business practices of individual platforms over time.

If enacted, we believe that this legislation has the potential to have a major negativepositive effect on dynamism in digital markets going forward. Our future global competitiveness depends on innovators and entrepreneurs having the ability to access markets, free from counterproductive inflexible government market regulationdominant incumbents that impede innovation, competition, resiliency, and widespread prosperity. Discriminatory conduct by majordominant platforms, properly understood, often benefits can sap the rewards from other innovators and entrepreneurs, increasingreducing the incentives for entrepreneurship and innovation. Even more importantly, the legislation may undercutsupport the creation and growth of new tech businesses adjacent to the platforms., Such an unfortunate result would reduce the welfare-enhancing initiatives of new businesses that are complementary to the economically beneficial activity (to consumers and producers) generated by the platforms. which may ultimately pose a critically needed competitive check to the covered platforms themselves. We view reduction of these new business initiatives benefits  as a significant harm that would stem from passage of the bills. For these reasons, the Department strongly supports the principles and goals animating opposes the legislation and looks forward to working with Congress to further explain why this undoubtedly well-meaning legislative initiative is detrimental to vigorous competition and a strong American economy.ensure that the final legislation enacted meets these goals.

Thank you for the opportunity to present our views. We hope this information is helpful. Please do not hesitate to contact this office if we may be of additional assistance to you.


[1] In other words,As , the Department respectfully recommends that members of Congress stop wasting time seeking to revise and enact this legislation.members continue to revise the legislation, the Department will provide under separate cover additional assistance to ensure that the bills achieve their goals.

Clean Version (incorporating all redline edits)

Dear Chairman Nadler, Chairman Cicilline, Representative Jordan, and Representative Buck:

The Department of Justice (Department) appreciates the considerable attention and resources devoted by the House and Senate Committees on the Judiciary over the past several years to ensuring the competitiveness of our digital economy, and writes today to oppose the American Innovation and Choice Online Act, Senate bill S. 2992, and the American Innovation and Choice Online Act, House bill H.R. 3816, which contain similar prohibitions on discriminatory conduct by dominant platforms (the “bills”). Unfortunately, the legislative efforts expended on these bills have been a waste of time.

The Department views the rise of major digital platforms as presenting a great boon to open markets and competition, bestowing benefits on consumers, businesses, innovation, resiliency, global competitiveness, and our democracy. By enhancing value transmitted through key arteries of the nation’s commerce and communications, such platforms have promoted a more vibrant and innovative modern economy. Vesting in government the power to pick winners and losers across markets through legislative regulation as found in the bills contravenes the foundations of our capitalist system, and given the increasing importance of these markets, the economic benefits flowing from platform activity are likely to be curtailed if the bills are passed. Enactment of the bills would put at risk the nation’s economic progress and prosperity, ultimately threatening the economic liberty that undergirds our democracy.

The legislation, if enacted, would emphasize causes of action prohibiting the largest digital platforms from “discriminating” in favor of their own products or services, or among third parties. In so doing, it would eliminate and disincentivize many efficient business arrangements that can materially enhance competition. This would thereby undermine the system of ex ante enforcement through which the United States maintains competitive markets with legal prohibitions on competitively harmful corporate conduct. By mistakenly characterizing as anticompetitive platform behaviors that in reality enhance incentives for vigorous innovation and dynamic competition, the legislation would undermine the existing antitrust laws. Specifically, the legislation would prevent the largest digital companies from managing their business transactions in an efficient welfare-enhancing manner, to the detriment of competition and the competitive process. The Department is strongly concerned about these harmful effects. As such, it encourages both the Committees and Congress to abandon all efforts to finalize this legislation and pass it into law.[1]

The Department views the legislation’s new prohibitions on discrimination as a harmful detriment to, and interference with, existing antitrust authority. In our view, the most significant harm would arise where the legislation seeks to elucidate Congress’ views of anticompetitive conduct—particularly with respect to harmful types of discrimination and self-preferencing by dominant platforms. Enumerating specific “discriminatory” and “self-preferencing” conduct that Congress views as anticompetitive and therefore illegal would undermine the economically informed, fact-specific evaluation of business conduct that lies at the heart of modern antitrust analysis, centered on consumer welfare. Modern economic analysis demonstrates that a great deal of superficially “discriminatory” and “self-preferencing” conduct often represents consumer welfare-enhancing behavior that adds to economic surplus. Deciding whether such conduct is procompetitive (welfare-enhancing) or anticompetitive in a particular instance is the role of existing case-by-case antitrust enforcement. This approach vindicates competition while avoiding the wrongful condemnation of economically beneficial behavior. In contrast, by creating a new class of antitrust “wrongs,” the bills would lead to the incorrect condemnation of many business practices that enhance market efficiency and strengthen the economy.

If enacted, we believe that this legislation has the potential to have a major negative effect on dynamism in digital markets going forward. Our future global competitiveness depends on innovators and entrepreneurs having the ability to access markets, free from counterproductive inflexible government market regulation that impede innovation, competition, resiliency, and widespread prosperity. “Discriminatory” conduct by major platforms, properly understood, often benefits innovators and entrepreneurs, increasing the incentives for entrepreneurship and innovation. Even more importantly, the legislation may undercut the creation and growth of new tech businesses adjacent to the platforms. Such an unfortunate result would reduce the welfare-enhancing initiatives of new businesses that are complementary to the economically beneficial activity (to consumers and producers) generated by the platforms. We view reduction of these new business initiatives as a significant harm that would stem from passage of the bills. For these reasons, the Department strongly opposes the legislation and looks forward to working with Congress to further explain why this undoubtedly well-meaning legislative initiative is detrimental to vigorous competition and a strong American economy.

Thank you for the opportunity to present our views. We hope this information is helpful. Please do not hesitate to contact this office if we may be of additional assistance to you.


[1] In other words, the Department respectfully recommends that members of Congress stop wasting time seeking to revise and enact this legislation.

After years of debate and negotiations, European Lawmakers have agreed upon what will most likely be the final iteration of the Digital Markets Act (“DMA”), following the March 24 final round of “trilogue” talks. 

For the uninitiated, the DMA is one in a string of legislative proposals around the globe intended to “rein in” tech companies like Google, Amazon, Facebook, and Apple through mandated interoperability requirements and other regulatory tools, such as bans on self-preferencing. Other important bills from across the pond include the American Innovation and Choice Online Act, the ACCESS Act, and the Open App Markets Act

In many ways, the final version of the DMA represents the worst possible outcome, given the items that were still up for debate. The Commission caved to some of the Parliament’s more excessive demands—such as sweeping interoperability provisions that would extend not only to “ancillary” services, such as payments, but also to messaging services’ basic functionalities. Other important developments include the addition of voice assistants and web browsers to the list of Core Platform Services (“CPS”), and symbolically higher “designation” thresholds that further ensure the act will apply overwhelmingly to just U.S. companies. On a brighter note, lawmakers agreed that companies could rebut their designation as “gatekeepers,” though it remains to be seen how feasible that will be in practice. 

We offer here an overview of the key provisions included in the final version of the DMA and a reminder of the shaky foundations it rests on.

Interoperability

Among the most important of the DMA’s new rules concerns mandatory interoperability among online platforms. In a nutshell, digital platforms that are designated as “gatekeepers” will be forced to make their services “interoperable” (i.e., compatible) with those of rivals. It is argued that this will make online markets more contestable and thus boost consumer choice. But as ICLE scholars have been explaining for some time, this is unlikely to be the case (here, here, and here). Interoperability is not the panacea EU legislators claim it to be. As former ICLE Director of Competition Policy Sam Bowman has written, there are many things that could be interoperable, but aren’t. The reason is that interoperability comes with costs as well as benefits. For instance, it may be worth letting different earbuds have different designs because, while it means we sacrifice easy interoperability, we gain the ability for better designs to be brought to the market and for consumers to be able to choose among them. Economists Michael L. Katz and Carl Shapiro concur:

Although compatibility has obvious benefits, obtaining and maintaining compatibility often involves a sacrifice in terms of product variety or restraints on innovation.

There are other potential downsides to interoperability.  For instance, a given set of interoperable standards might be too costly to implement and/or maintain; it might preclude certain pricing models that increase output; or it might compromise some element of a product or service that offers benefits specifically because it is not interoperable (such as, e.g., security features). Consumers may also genuinely prefer closed (i.e., non-interoperable) platforms. Indeed: “open” and “closed” are not synonyms for “good” and “bad.” Instead, as Boston University’s Andrei Hagiu has shown, there are fundamental welfare tradeoffs at play that belie simplistic characterizations of one being inherently superior to the other. 

Further, as Sam Bowman observed, narrowing choice through a more curated experience can also be valuable for users, as it frees them from having to research every possible option every time they buy or use some product (if you’re unconvinced, try turning off your spam filter for a couple of days). Instead, the relevant choice consumers exercise might be in choosing among brands. In sum, where interoperability is a desirable feature, consumer preferences will tend to push for more of it. However, it is fundamentally misguided to treat mandatory interoperability as a cure-all elixir or a “super tool” of “digital platform governance.” In a free-market economy, it is not—or, it should not—be up to courts and legislators to substitute for businesses’ product-design decisions and consumers’ revealed preferences with their own, based on diffuse notions of “fairness.” After all, if we could entrust such decisions to regulators, we wouldn’t need markets or competition in the first place.

Of course, it was always clear that the DMA would contemplate some degree of mandatory interoperability – indeed, this was arguably the new law’s biggest selling point. What was up in the air until now was the scope of such obligations. The Commission had initially pushed for a comparatively restrained approach, requiring interoperability “only” in ancillary services, such as payment systems (“vertical interoperability”). By contrast, the European Parliament called for more expansive requirements that would also encompass social-media platforms and other messaging services (“horizontal interoperability”). 

The problem with such far-reaching interoperability requirements is that they are fundamentally out of pace with current privacy and security capabilities. As ICLE Senior Scholar Mikolaj Barczentewicz has repeatedly argued, the Parliament’s insistence on going significantly beyond the original DMA’s proposal and mandating interoperability of messaging services is overly broad and irresponsible. Indeed, as Mikolaj notes, the “likely result is less security and privacy, more expenses, and less innovation.”The DMA’s defensers would retort that the law allows gatekeepers to do what is “strictly necessary” (Council) or “indispensable” (Parliament) to protect safety and privacy (it is not yet clear which wording the final version has adopted). Either way, however, the standard may be too high and companies may very well offer lower security to avoid liability for adopting measures that would be judged by the Commission and the courts as going beyond what is “strictly necessary” or “indispensable.” These safeguards will inevitably be all the more indeterminate (and thus ineffectual) if weighed against other vague concepts at the heart of the DMA, such as “fairness.”

Gatekeeper Thresholds and the Designation Process

Another important issue in the DMA’s construction concerns the designation of what the law deems “gatekeepers.” Indeed, the DMA will only apply to such market gatekeepers—so-designated because they meet certain requirements and thresholds. Unfortunately, the factors that the European Commission will consider in conducting this designation process—revenues, market capitalization, and user base—are poor proxies for firms’ actual competitive position. This is not surprising, however, as the procedure is mainly designed to ensure certain high-profile (and overwhelmingly American) platforms are caught by the DMA.

From this perspective, the last-minute increase in revenue and market-capitalization thresholds—from 6.5 billion euros to 7.5 billion euros, and from 65 billion euros to 75 billion euros, respectively—won’t change the scope of the companies covered by the DMA very much. But it will serve to confirm what we already suspected: that the DMA’s thresholds are mostly tailored to catch certain U.S. companies, deliberately leaving out EU and possibly Chinese competitors (see here and here). Indeed, what would have made a difference here would have been lowering the thresholds, but this was never really on the table. Ultimately, tilting the European Union’s playing field against its top trading partner, in terms of exports and trade balance, is economically, politically, and strategically unwise.

As a consolation of sorts, it seems that the Commission managed to squeeze in a rebuttal mechanism for designated gatekeepers. Imposing far-reaching obligations on companies with no  (or very limited) recourse to escape the onerous requirements of the DMA would be contrary to the basic principles of procedural fairness. Still, it remains to be seen how this mechanism will be articulated and whether it will actually be viable in practice.

Double (and Triple?) Jeopardy

Two recent judgments from the European Court of Justice (ECJ)—Nordzucker and bpost—are likely to underscore the unintended effects of cumulative application of both the DMA and EU and/or national competition laws. The bpost decision is particularly relevant, because it lays down the conditions under which cases that evaluate the same persons and the same facts in two separate fields of law (sectoral regulation and competition law) do not violate the principle of ne bis in idem, also known as “double jeopardy.” As paragraph 51 of the judgment establishes:

  1. There must be precise rules to determine which acts or omissions are liable to be subject to duplicate proceedings;
  2. The two sets of proceedings must have been conducted in a sufficiently coordinated manner and within a similar timeframe; and
  3. The overall penalties must match the seriousness of the offense. 

It is doubtful whether the DMA fulfills these conditions. This is especially unfortunate considering the overlapping rules, features, and goals among the DMA and national-level competition laws, which are bound to lead to parallel procedures. In a word: expect double and triple jeopardy to be hotly litigated in the aftermath of the DMA.

Of course, other relevant questions have been settled which, for reasons of scope, we will have to leave for another time. These include the level of fines (up to 10% worldwide revenue, or 20% in the case of repeat offenses); the definition and consequences of systemic noncompliance (it seems that the Parliament’s draconian push for a general ban on acquisitions in case of systemic noncompliance has been dropped); and the addition of more core platform services (web browsers and voice assistants).

The DMA’s Dubious Underlying Assumptions

The fuss and exhilaration surrounding the impending adoption of the EU’s most ambitious competition-related proposal in decades should not obscure some of the more dubious assumptions which underpin it, such as that:

  1. It is still unclear that intervention in digital markets is necessary, let alone urgent.
  2. Even if it were clear, there is scant evidence to suggest that tried and tested ex post instruments, such as those envisioned in EU competition law, are not up to the task.
  3. Even if the prior two points had been established beyond any reasonable doubt (which they haven’t), it is still far from clear that DMA-style ex ante regulation is the right tool to address potential harms to competition and to consumers that arise in digital markets.

It is unclear that intervention is necessary

Despite a mounting moral panic around and zealous political crusading against Big Tech (an epithet meant to conjure antipathy and distrust), it is still unclear that intervention in digital markets is necessary. Much of the behavior the DMA assumes to be anti-competitive has plausible pro-competitive justifications. Self-preferencing, for instance, is a normal part of how platforms operate, both to improve the value of their core products and to earn returns to reinvest in their development. As ICLE’s Dirk Auer points out, since platforms’ incentives are to maximize the value of their entire product ecosystem, those that preference their own products frequently end up increasing the total market’s value by growing the share of users of a particular product (the example of Facebook’s integration of Instagram is a case in point). Thus, while self-preferencing may, in some cases, be harmful, a blanket presumption of harm is thoroughly unwarranted

Similarly, the argument that switching costs and data-related increasing returns to scale (in fact, data generally entails diminishing returns) have led to consumer lock-in and thereby raised entry barriers has also been exaggerated to epic proportions (pun intended). As we have discussed previously, there are plenty of counterexamples where firms have easily overcome seemingly “insurmountable” barriers to entry, switching costs, and network effects to disrupt incumbents. 

To pick a recent case: how many of us had heard of Zoom before the pandemic? Where was TikTok three years ago? (see here for a multitude of other classic examples, including Yahoo and Myspace).

Can you really say, with a straight face, that switching costs between messaging apps are prohibitive? I’m not even that active and I use at least six such apps on a daily basis: Facebook Messenger, Whatsapp, Instagram, Twitter, Viber, Telegram, and Slack (it took me all of three minutes to download and start using Slack—my newest addition). In fact, chances are that, like me, you have always multihomed nonchalantly and had never even considered that switching costs were impossibly high (or that they were a thing) until the idea that you were “locked-in” by Big Tech was drilled into your head by politicians and other busybodies looking for trophies to adorn their walls.

What about the “unprecedented,” quasi-fascistic levels of economic concentration? First, measures of market concentration are sometimes anchored in flawed methodology and market definitions  (see, e.g., Epic’s insistence that Apple is a monopolist in the market for operating systems, conveniently ignoring that competition occurs at the smartphone level, where Apple has a worldwide market share of 15%—see pages 45-46 here). But even if such measurements were accurate, high levels of concentration don’t necessarily mean that firms do not face strong competition. In fact, as Nicolas Petit has shown, tech companies compete vigorously against each other across markets.

But perhaps the DMA’s raison d’etre rests less on market failure, but rather on a legal or enforcement failure? This, too, is misguided.

EU competition law is already up to the task

As Giuseppe Colangelo has argued persuasively (here and here), it is not at all clear that ex post competition regulation is insufficient to tackle anti-competitive behavior in the digital sector:

Ongoing antitrust investigations demonstrate that standard competition law still provides a flexible framework to scrutinize several practices described as new and peculiar to app stores. 

The recent Google Shopping decision, in which the Commission found that Google had abused its dominant position by preferencing its own online-shopping service in Google Search results, is a case in point (the decision was confirmed by the General Court and is now pending review before the European Court of Justice). The “self-preferencing” category has since been applied by other EU competition authorities. The Italian competition authority, for instance, fined Amazon 1 billion euros for preferencing its own distribution service, Fulfilled by Amazon, on the Amazon marketplace (i.e., Amazon.it). Thus, Article 102, which includes prohibitions on “applying dissimilar conditions to similar transactions,” appears sufficiently flexible to cover self-preferencing, as well as other potentially anti-competitive offenses relevant to digital markets (e.g., essential facilities).

For better or for worse, EU competition law has historically been sufficiently pliable to serve a range of goals and values. It has also allowed for experimentation and incorporated novel theories of harm and economic insights. Here, the advantage of competition law is that it allows for a more refined, individualized approach that can avoid some of the pitfalls of applying a one-size fits all model across all digital platforms. Those pitfalls include: harming consumers, jeopardizing the business models of some of the most successful and pro-consumer companies in existence, and ignoring the differences among platforms, such as between Google and Apple’s app stores. I turn to these issues next.

Ex ante regulation probably isn’t the right tool

Even if it were clear that intervention is necessary and that existing competition law was insufficient, it is not clear that the DMA is the right regulatory tool to address any potential harms to competition and consumers that may arise in the digital markets. Here, legislators need to be wary of unintended consequences, trade-offs, and regulatory fallibility. For one, It is possible that the DMA will essentially consolidate the power of tech platforms, turning them into de facto public utilities. This will not foster competition, but rather will make smaller competitors systematically dependent on so-called gatekeepers. Indeed, why become the next Google if you can just free ride off of the current Google? Why download an emerging messaging app if you can already interact with its users through your current one? In a way, then, the DMA may become a self-fulfilling prophecy. 

Moreover, turning closed or semi-closed platforms such as the iOS into open platforms more akin to Android blurs the distinctions among products and dampens interbrand competition. It is a supreme paradox that interoperability and sideloading requirements purportedly give users more choice by taking away the option of choosing a “walled garden” model. As discussed above, overriding the revealed preferences of millions of users is neither pro-competitive nor pro-consumer (but it probably favors some competitors at the expense of those two things). 

Nor are many of the other obligations contemplated in the DMA necessarily beneficial to consumers. Do users really not want to have default apps come preloaded on their devices and instead have to download and install them manually? Ditto for operating systems. What is the point of an operating system if it doesn’t come with certain functionalities, such as a web browser? What else should we unbundle—keyboard on iOS? Flashlight? Do consumers really want to choose from dozens of app stores when turning on their new phone for the first time? Do they really want to have their devices cluttered with pointless split-screens? Do users really want to find all their contacts (and be found by all their contacts) across all messaging services? (I switched to Viber because I emphatically didn’t.) Do they really want to have their privacy and security compromised because of interoperability requirements?Then there is the question of regulatory fallibility. As Alden Abott has written on the DMA and other ex ante regulatory proposals aimed at “reining in” tech companies:

Sorely missing from these regulatory proposals is any sense of the fallibility of regulation. Indeed, proponents of new regulatory proposals seem to implicitly assume that government regulation of platforms will enhance welfare, ignoring real-life regulatory costs and regulatory failures (see here, for example). 

This brings us back to the second point: without evidence that antitrust law is “not up to the task,” far-reaching and untested regulatory initiatives with potentially high error costs are put forth as superior to long-established, consumer-based antitrust enforcement. Yes, antitrust may have downsides (e.g., relative indeterminacy and slowness), but these pale in comparison to the DMA’s (e.g., large error costs resulting from high information requirements, rent-seeking, agency capture).

Conclusion

The DMA is an ambitious piece of regulation purportedly aimed at ensuring “fair and open digital markets.” This implies that markets are not fair and open; or that they risk becoming unfair and closed absent far-reaching regulatory intervention at EU level. However, it is unclear to what extent such assumptions are borne out by the reality of markets. Are digital markets really closed? Are they really unfair? If so, is it really certain that regulation is necessary? Has antitrust truly proven insufficient? It also implies that DMA-style ex ante regulation is necessary to tackle it, and that the costs won’t outweigh the benefits. These are heroic assumptions that have never truly been seriously put to the test. 

Considering such brittle empirical foundations, the DMA was always going to be a contentious piece of legislation. However, there was always the hope that EU legislators would show restraint in the face of little empirical evidence and high error costs. Today, these hopes have been dashed. With the adoption of the DMA, the Commission, Council, and the Parliament have arguably taken a bad piece of legislation and made it worse. The interoperability requirements in messaging services, which are bound to be a bane for user privacy and security, are a case in point.

After years trying to anticipate the whims of EU legislators, we finally know where we’re going, but it’s still not entirely sure why we’re going there.

There has been a wave of legislative proposals on both sides of the Atlantic that purport to improve consumer choice and the competitiveness of digital markets. In a new working paper published by the Stanford-Vienna Transatlantic Technology Law Forum, I analyzed five such bills: the EU Digital Services Act, the EU Digital Markets Act, and U.S. bills sponsored by Rep. David Cicilline (D-R.I.), Rep. Mary Gay Scanlon (D-Pa.), Sen. Amy Klobuchar (D-Minn.) and Sen. Richard Blumenthal (D-Conn.). I concluded that all those bills would have negative and unaddressed consequences in terms of information privacy and security.

In this post, I present the main points from the working paper regarding two regulatory solutions: (1) mandating interoperability and (2) mandating device neutrality (which leads to a possibility of sideloading applications, a special case of interoperability.) The full working paper  also covers the risks of compulsory data access (by vetted researchers or by authorities).

Interoperability

Interoperability is increasingly presented as a potential solution to some of the alleged problems associated with digital services and with large online platforms, in particular (see, e.g., here and here). For example, interoperability might allow third-party developers to offer different “flavors” of social-media newsfeeds, with varying approaches to content ranking and moderation. This way, it might matter less than it does now what content moderation decisions Facebook or other platforms make. Facebook users could choose alternative content moderators, delivering the kind of news feed that those users expect.

The concept of interoperability is popular not only among thought leaders, but also among legislators. The DMA, as well as the U.S. bills by Rep. Scanlon, Rep. Cicilline, and Sen. Klobuchar, all include interoperability mandates.

At the most basic level, interoperability means a capacity to exchange information between computer systems. Email is an example of an interoperable standard that most of us use today. It is telling that supporters of interoperability mandates use services like email as their model examples. Email (more precisely, the SMTP protocol) originally was designed in a notoriously insecure way. It is a perfect example of the opposite of privacy by design. A good analogy for the levels of privacy and security provided by email, as originally conceived, is that of a postcard message sent without an envelope that passes through many hands before reaching the addressee. Even today, email continues to be a source of security concerns, due to its prioritization of interoperability (see, e.g., here).

Using currently available technology to provide alternative interfaces or moderation services for social-media platforms, third-party developers would have to be able to access much of the platform content that is potentially available to a user. This would include not just content produced by users who explicitly agree to share their data with third parties, but also content—e.g., posts, comments, likes—created by others who may have strong objections to such sharing. It does not require much imagination to see how, without adequate safeguards, mandating this kind of information exchange would inevitably result in something akin to the 2018 Cambridge Analytica data scandal.

There are several constraints for interoperability frameworks that must be in place to safeguard privacy and security effectively.

First, solutions should be targeted toward real users of digital services, without assuming away some common but inconvenient characteristics. In particular, solutions should not assume unrealistic levels of user interest and technical acumen.

Second, solutions must address the issue of effective enforcement. Even the best information privacy and security laws do not, in and of themselves, solve any problems. Such rules must be followed, which requires addressing the problems of procedure and enforcement. In both the EU and the United States, the current framework and practice of privacy law enforcement offers little confidence that misuses of broadly construed interoperability would be detected and prosecuted, much less that they would be prevented. This is especially true for smaller and “judgment-proof” rulebreakers, including those from foreign jurisdictions.

If the service providers are placed under a broad interoperability mandate with non-discrimination provisions (preventing effective vetting of third parties, unilateral denials of access, and so on), then the burden placed on law enforcement will be mammoth. Just one bad actor, perhaps working from Russia or North Korea, could cause immense damage by taking advantage of interoperability mandates to exfiltrate user data or to execute a hacking (e.g., phishing) campaign. Of course, such foreign bad actors would be in violation of the EU GDPR, but that is unlikely to have any practical significance.

It would not be sufficient to allow (or require) service providers to enforce merely technical filters, such as a requirement to check whether the interoperating third parties’ IP address comes from a jurisdiction with sufficient privacy protections. Working around such technical limitations does not pose a significant difficulty to motivated bad actors.

Art 6(1) of the original DMA proposal included some general interoperability provisions applicable to “gatekeepers”—i.e., the largest online platforms. Those interoperability mandates were somewhat limited – applying only to “ancillary services” (e.g., payment or identification services) or requiring only one-way data portability. However, even here, there may be some risks. For example, users may choose poorly secured identification services and thus become victims of attacks. Therefore, it is important that gatekeepers not be prevented from protecting their users adequately.

The drafts of the DMA adopted by the European Council and by the European Parliament attempt to address that, but they only allow gatekeepers to do what is “strictly necessary” (Council) or “indispensable” (Parliament). This standard may be too high and could push gatekeepers to offer lower security to avoid liability for adopting measures that would be judged by EU institutions and the courts as going beyond what is strictly necessary or indispensable.

The more recent DMA proposal from the European Parliament goes significantly beyond the original proposal, mandating full interoperability of a number of “independent interpersonal communication services” and of social-networking services. The Parliament’s proposals are good examples of overly broad and irresponsible interoperability mandates. They would cover “any providers” wanting to interconnect with gatekeepers, without adequate vetting. The safeguard proviso mentioning “high level of security and personal data protection” does not come close to addressing the seriousness of the risks created by the mandate. Instead of facing up to the risks and ensuring that the mandate itself be limited in ways that minimize them, the proposal seems just to expect that the gatekeepers can solve the problems if they only “nerd harder.”

All U.S. bills considered here introduce some interoperability mandates and none of them do so in a way that would effectively safeguard information privacy and security. For example, Rep. Cicilline’s American Choice and Innovation Online Act (ACIOA) would make it unlawful (in Section 2(b)(1)) to:

All U.S. bills considered here introduce some interoperability mandates and none of them do so in a way that would effectively safeguard information privacy and security. For example, Rep. Cicilline’s American Choice and Innovation Online Act (ACIOA) would make it unlawful (in Section 2(b)(1)) to:

restrict or impede the capacity of a business user to access or interoperate with the same platform, operating system, hardware and software features that are available to the covered platform operator’s own products, services, or lines of business.

The language of the prohibition in Sen. Klobuchar’s American Innovation and Choice Online Act (AICOA) is similar (also in Section 2(b)(1)). Both ACIOA and AICOA allow for affirmative defenses that a service provider could use if sued under the statute. While those defenses mention privacy and security, they are narrow (“narrowly tailored, could not be achieved through a less discriminatory means, was nonpretextual, and was necessary”) and would not prevent service providers from incurring significant litigation costs. Hence, just like the provisions of the DMA, they would heavily incentivize covered service providers not to adopt the most effective protections of privacy and security.

Device Neutrality (Sideloading)

Article 6(1)(c) of the DMA contains specific provisions about “sideloading”—i.e., allowing installation of third-party software through alternative app stores other than the one provided by the manufacturer (e.g., Apple’s App Store for iOS devices). A similar express provision for sideloading is included in Sen. Blumenthal’s Open App Markets Act (Section 3(d)(2)). Moreover, the broad interoperability provisions in the other U.S. bills discussed above may also be interpreted to require permitting sideloading.

A sideloading mandate aims to give users more choice. It can only achieve this, however, by taking away the option of choosing a device with a “walled garden” approach to privacy and security (such as is taken by Apple with iOS). By taking away the choice of a walled garden environment, a sideloading mandate will effectively force users to use whatever alternative app stores are preferred by particular app developers. App developers would have strong incentive to set up their own app stores or to move their apps to app stores with the least friction (for developers, not users), which would also mean the least privacy and security scrutiny.

This is not to say that Apple’s app scrutiny is perfect, but it is reasonable for an ordinary user to prefer Apple’s approach because it provides greater security (see, e.g., here and here). Thus, a legislative choice to override the revealed preference of millions of users for a “walled garden” approach should not be made lightly. 

Privacy and security safeguards in the DMA’s sideloading provisions, as amended by the European Council and by the European Parliament, as well as in Sen. Blumenthal’s Open App Markets Act, share the same problem of narrowness as the safeguards discussed above.

There is a more general privacy and security issue here, however, that those safeguards cannot address. The proposed sideloading mandate would prohibit outright a privacy and security-protection model that many users rationally choose today. Even with broader exemptions, this loss will be genuine. It is unclear whether taking away this choice from users is justified.

Conclusion

All the U.S. and EU legislative proposals considered here betray a policy preference of privileging uncertain and speculative competition gains at the expense of introducing a new and clear danger to information privacy and security. The proponents of these (or even stronger) legislative interventions seem much more concerned, for example, that privacy safeguards are “not abused by Apple and Google to protect their respective app store monopoly in the guise of user security” (source).

Given the problems with ensuring effective enforcement of privacy protections (especially with respect to actors coming from outside the EU, the United States, and other broadly privacy-respecting jurisdictions), the lip service paid by the legislative proposals to privacy and security is not much more than that. Policymakers should be expected to offer a much more detailed vision of concrete safeguards and mechanisms of enforcement when proposing rules that come with significant and entirely predictable privacy and security risks. Such vision is lacking on both sides of the Atlantic.

I do not want to suggest that interoperability is undesirable. The argument of this paper was focused on legally mandated interoperability. Firms experiment with interoperability all the time—the prevalence of open APIs on the Internet is testament to this. My aim, however, is to highlight that interoperability is complex and exposes firms and their users to potentially large-scale cyber vulnerabilities.

Generalized obligations on firms to open their data, or to create service interoperability, can short-circuit the private ordering processes that seek out those forms of interoperability and sharing that pass a cost-benefit test. The result will likely be both overinclusive and underinclusive. It would be overinclusive to require all firms in the regulated class to broadly open their services and data to all interested parties, even where it wouldn’t make sense for privacy, security, or other efficiency reasons. It is underinclusive in that the broad mandate will necessarily sap regulated firms’ resources and deter them from looking for new innovative uses that might make sense, but that are outside of the broad mandate. Thus, the likely result is less security and privacy, more expense, and less innovation.

In a new paper, Giuseppe Colangelo and Oscar Borgogno investigate whether antitrust policy is sufficiently flexible to keep up with the dynamics of digital app stores, and whether regulatory interventions are required in order to address their unique features. The authors summarize their findings in this blog post.

App stores are at the forefront of policy debates surrounding digital markets. The gatekeeping position of Apple and Google in the App Store and Google Play Store, respectively, and related concerns about the companies’ rule-setting and dual role, have been the subject of market studies launched by the Australian Competition and Consumer Commission (ACCC), the Netherlands Authority for Consumers & Markets (ACM), the U.K. Competition and Markets Authority (CMA), the Japan Federal Trade Commission (JFTC), and the U.S. House of Representatives.

Likewise, the terms and conditions for accessing app stores—such as in-app purchasing rules, restrictions on freedom of choice for smartphone payment apps, and near field communication (NFC) limitations—face scrutiny from courts and antitrust authorities around the world.

Finally, legislative initiatives envisage obligations explicitly addressed to app stores. Notably, the European Digital Markets Act (DMA) and some U.S. bills (e.g., the American Innovation and Choice Online Act and the Open App Markets Act, both of which are scheduled to be marked up Jan. 20 by the Senate Judiciary Committee) prohibit designated platforms from, for example: discriminating among users by engaging in self-preferencing and applying unfair access conditions; preventing users from sideloading and uninstalling pre-installed apps; impeding data portability and interoperability; or imposing anti-steering provisions. Likewise, South Korea has recently prohibited app-store operators in dominant market positions from forcing payment systems upon content providers and inappropriately delaying the review of, or deleting, mobile content from app markets.

Despite their differences, these international legislative initiatives do share the same aims and concerns. By and large, they question the role of competition law in the digital economy. In the case of app stores, these regulatory interventions attempt to introduce a neutrality regime, with the aim of increasing contestability, facilitating the possibility of switching by users, tackling conflicts of interests, and addressing imbalances in the commercial relationship. Ultimately, these proposals would treat online platforms as akin to common carriers or public utilities.

All of these initiatives assume antitrust is currently falling, because competition rules apply ex post and require an extensive investigation on a case-by-case basis. But is that really the case?

Platform and Device Neutrality Regime

Focusing on the content of the European, German, and U.S. legislative initiatives, the neutrality regime envisaged for app stores would introduce obligations in terms of both device and platform neutrality. The former includes provisions on app uninstalling, sideloading, app switching, access to technical functionality, and the possibility of changing default settings.  The latter entail data portability and interoperability obligations, and the ban on self-preferencing, Sherlocking, and unfair access conditions.

App Store Obligations: Comparison of EU, German, and U.S. Initiatives

Antitrust v. Regulation

Despite the growing consensus regarding the need to rely on ex ante regulation to govern digital markets and tackle the practices of large online platforms, recent and ongoing antitrust investigations demonstrate that standard competition law still provides a flexible framework to scrutinize several practices sometimes described as new and peculiar to app stores.

This is particularly true in Europe, where the antitrust framework grants significant leeway to antitrust enforcers relative to the U.S. scenario, as illustrated by the recent Google Shopping decision.

Indeed, considering legislative proposals to modernize antitrust law and to strengthen its enforcement, the U.S. House Judiciary Antitrust Subcommittee, along with some authoritative scholars, have suggested emulating the European model—imposing particular responsibility on dominant firms through the notion of abuse of dominant position and overriding several Supreme Court decisions in order to clarify the prohibitions on monopoly leveraging, predatory pricing, denial of essential facilities, refusals to deal, and tying.

By contrast, regulation appears better suited to support interventions intended to implement industrial-policy objectives. This applies, in particular, to provisions prohibiting app stores from impeding or restricting sideloading, app uninstalling, the possibility of choosing third-party apps and app stores as defaults, as well as provisions that would mandate data portability and interoperability.

However, such regulatory proposals may ultimately harm consumers. Indeed, by questioning the core of digital platform business models and affecting their governance design, these interventions entrust public authorities with mammoth tasks that could ultimately jeopardize the profitability of app-store ecosystems. They also overlook the differences that may exist between the business models of different platforms, such as Google and Apple’s app stores.

To make matters worse, the  difficulties encountered by regulators that have imposed product-design remedies on firms suggest that regulators may struggle to craft feasible and effective solutions. For instance, when the European General Court found that Google favored its own services in the Google Shopping case, it noted that this finding rested on the differential positioning and display of Shopping Units when compared to generic results. As a consequence, it could be argued that Google’s proposed auction remedy (whereby Google would compete with rivals for Shopping box placement) is compliant with the Court’s ruling because there is no dicrimination, regardless of the fact that Google might ultimately outbid its rivals (see here).

Finally, the neutrality principle cannot be transposed perfectly to all online platforms. Indeed, the workings of the app-discovery and distribution markets differ from broadband networks, as rankings and mobile services by definition involve some form of continuous selection and differentiated treatment to optimize the mobile-customer experience.

For all these reasons, our analysis suggests that antitrust law provides a less intrusive and more individualized approach, which would eventually benefit consumers by safeguarding quality and innovation.

Intermediaries may not be the consumer welfare hero we want, but more often than not, they are one that we need.

In policy discussions about the digital economy, a background assumption that frequently underlies the discourse is that intermediaries and centralization always and only serve as a cost to consumers, and to society more generally. Thus, one commonly sees arguments that consumers would be better off if they could freely combine products from different trading partners. According to this logic, bundled goods, walled gardens, and other intermediaries are always to be regarded with suspicion, while interoperability, open source, and decentralization are laudable features of any market.

However, as with all economic goods, intermediation offers both costs and benefits. The challenge for market players is to assess these tradeoffs and, ultimately, to produce the optimal level of intermediation.

As one example, some observers assume that purchasing food directly from a producer benefits consumers because intermediaries no longer take a cut of the final purchase price. But this overlooks the tremendous efficiencies supermarkets can achieve in terms of cost savings, reduced carbon emissions (because consumers make fewer store trips), and other benefits that often outweigh the costs of intermediation.

The same anti-intermediary fallacy is plain to see in countless other markets. For instance, critics readily assume that insurance, mortgage, and travel brokers are just costly middlemen.

This unduly negative perception is perhaps even more salient in the digital world. Policymakers are quick to conclude that consumers are always better off when provided with “more choice.” Draft regulations of digital platforms have been introduced on both sides of the Atlantic that repeat this faulty argument ad nauseam, as do some antitrust decisions.

Even the venerable Tyler Cowen recently appeared to sing the praises of decentralization, when discussing the future of Web 3.0:

One person may think “I like the DeFi options at Uniswap,” while another may say, “I am going to use the prediction markets over at Hedgehog.” In this scenario there is relatively little intermediation and heavy competition for consumer attention. Thus most of the gains from competition accrue to the users. …

… I don’t know if people are up to all this work (or is it fun?). But in my view this is the best-case scenario — and the most technologically ambitious. Interestingly, crypto’s radical ability to disintermediate, if extended to its logical conclusion, could bring about a radical equalization of power that would lower the prices and values of the currently well-established crypto assets, companies and platforms.

While disintermediation certainly has its benefits, critics often gloss over its costs. For example, scams are practically nonexistent on Apple’s “centralized” App Store but are far more prevalent with Web3 services. Apple’s “power” to weed out nefarious actors certainly contributes to this difference. Similarly, there is a reason that “middlemen” like supermarkets and travel agents exist in the first place. They notably perform several complex tasks (e.g., searching for products, negotiating prices, and controlling quality) that leave consumers with a manageable selection of goods.

Returning to the crypto example, besides being a renowned scholar, Tyler Cowen is also an extremely savvy investor. What he sees as fun investment choices may be nightmarish (and potentially dangerous) decisions for less sophisticated consumers. The upshot is that intermediaries are far more valuable than they are usually given credit for.

Bringing People Together

The reason intermediaries (including online platforms) exist is to reduce transaction costs that suppliers and customers would face if they tried to do business directly. As Daniel F. Spulber argues convincingly:

Markets have two main modes of organization: decentralized and centralized. In a decentralized market, buyers and sellers match with each other and determine transaction prices. In a centralized market, firms act as intermediaries between buyers and sellers.

[W]hen there are many buyers and sellers, there can be substantial transaction costs associated with communication, search, bargaining, and contracting. Such transaction costs can make it more difficult to achieve cross-market coordination through direct communication. Intermediary firms have various means of reducing transaction costs of decentralized coordination when there are many buyers and sellers.

This echoes the findings of Nobel laureate Ronald Coase, who observed that firms emerge when they offer a cheaper alternative to multiple bilateral transactions:

The main reason why it is profitable to establish a firm would seem to be that there is a cost of using the price mechanism. The most obvious cost of “organising ” production through the price mechanism is that of discovering what the relevant prices are. […] The costs of negotiating and concluding a separate contract for each exchange transaction which takes place on a market must also be taken into account.

Economists generally agree that online platforms also serve this cost-reduction function. For instance, David Evans and Richard Schmalensee observe that:

Multi-sided platforms create value by bringing two or more different types of economic agents together and facilitating interactions between them that make all agents better off.

It’s easy to see the implications for today’s competition-policy debates, and for the online intermediaries that many critics would like to see decentralized. Particularly salient examples include app store platforms (such as the Apple App Store and the Google Play Store); online retail platforms (such as Amazon Marketplace); and online travel agents (like Booking.com and Expedia). Competition policymakers have embarked on countless ventures to “open up” these platforms to competition, essentially moving them further toward disintermediation. In most of these cases, however, policymakers appear to be fighting these businesses’ very raison d’être.

For example, the purpose of an app store is to curate the software that users can install and to offer payment solutions; in exchange, the store receives a cut of the proceeds. If performing these tasks created no value, then to a first approximation, these services would not exist. Users would simply download apps via their web browsers, and the most successful smartphones would be those that allowed users to directly install apps (“sideloading,” to use the more technical terms). Forcing these platforms to “open up” and become neutral is antithetical to the value proposition they offer.

Calls for retail and travel platforms to stop offering house brands or displaying certain products more favorably are equally paradoxical. Consumers turn to these platforms because they want a selection of goods. If that was not the case, users could simply bypass the platforms and purchase directly from independent retailers or hotels.Critics sometimes retort that some commercial arrangements, such as “most favored nation” clauses, discourage consumers from doing exactly this. But that claim only reinforces the point that online platforms must create significant value, or they would not be able to obtain such arrangements in the first place.

All of this explains why characterizing these firms as imposing a “tax” on their respective ecosystems is so deeply misleading. The implication is that platforms are merely passive rent extractors that create no value. Yet, barring the existence of market failures, both their existence and success is proof to the contrary. To argue otherwise places no faith in the ability of firms and consumers to act in their own self-interest.

A Little Evolution

This last point is even more salient when seen from an evolutionary standpoint. Today’s most successful intermediaries—be they online platforms or more traditional brick-and-mortar firms like supermarkets—mostly had to outcompete the alternative represented by disintermediated bilateral contracts.

Critics of intermediaries rarely contemplate why the app-store model outpaced the more heavily disintermediated software distribution of the desktop era. Or why hotel-booking sites exist, despite consumers’ ability to use search engines, hotel websites, and other product-search methods that offer unadulterated product selections. Or why mortgage brokers are so common when borrowers can call local banks directly. The list is endless.

Indeed, as I have argued previously:

Digital markets could have taken a vast number of shapes, so why have they systematically gravitated towards those very characteristics that authorities condemn? For instance, if market tipping and consumer lock-in are so problematic, why is it that new corners of the digital economy continue to emerge via closed platforms, as opposed to collaborative ones? Indeed, if recent commentary is to be believed, it is the latter that should succeed because they purportedly produce greater gains from trade. And if consumers and platforms cannot realize these gains by themselves, then we should see [other] intermediaries step into the breach – i.e. arbitrage. This does not seem to be happening in the digital economy. The naïve answer is to say that this is precisely the problem, the harder one is to actually understand why.

Fiat Versus Emergent Disintermediation

All of this is not to say that intermediaries are perfect, or that centralization always beats decentralization. Instead, the critical point is about the competitive process. There are vast differences between centralization that stems from government fiat and that which emerges organically.

(Dis)intermediation is an economic good. Markets thus play a critical role in deciding how much or little of it is provided. Intermediaries must charge fees that cover their costs, while bilateral contracts entail transaction costs. In typically Hayekian fashion, suppliers and buyers will weigh the costs and benefits of these options.

Intermediaries are most likely to emerge in markets prone to excessive transaction costs and competitive processes ensure that only valuable intermediaries survive. Accordingly, there is no guarantee that government-mandated disintermediation would generate net benefits in any given case.

Of course, the market does not always work perfectly. Sometimes, market failures give rise to excessive (or insufficient) centralization. And policymakers should certainly be attentive to these potential problems and address them on a case-by-case basis. But there is little reason to believe that today’s most successful intermediaries are the result of market failures, and it is thus critical that policymakers do not undermine the valuable role they perform.

For example, few believe that supermarkets exist merely because government failures (such as excessive regulation) or market failures (such as monopolization) prevent the emergence of smaller rivals. Likewise, the app-store model is widely perceived as an improvement over previous software platforms; few consumers appear favorably disposed toward its replacement with sideloading of apps (for example, few Android users choose to sideload apps rather than purchase them via the Google Play Store). In fact, markets appear to be moving in the opposite direction: even traditional software platforms such as Windows OS increasingly rely on closed stores to distribute software on their platforms.

More broadly, this same reasoning can (and has) been applied to other social institutions, such as the modern family. For example, the late Steven Horwitz observed that family structures have evolved in order to adapt to changing economic circumstances. Crucially, this process is driven by the same cost-benefit tradeoff that we see in markets. In both cases, agents effectively decide which functions are better performed within a given social structure, and which ones are more efficiently completed outside of it.

Returning to Tyler Cowen’s point about the future of Web3, the case can be made that whatever level of centralization ultimately emerges is most likely the best case scenario. Sure, there may be some market failures and suboptimal outcomes along the way, but they ultimately pale in comparison to the most pervasive force: namely, economic agents’ ability to act in what they perceive to be their best interest. To put it differently, if Web3 spontaneously becomes as centralized as Web 2.0 has been, that would be testament to the tremendous role that intermediaries play throughout the economy.

Early last month, the Italian competition authority issued a record 1.128 billion euro fine against Amazon for abuse of dominance under Article 102 of the Treaty on the Functioning of the European Union (TFEU). In its order, the Agenzia Garante della Concorrenza e del Mercato (AGCM) essentially argues that Amazon has combined its Amazon.it marketplace and Fulfillment by Amazon (FBA) services to exclude logistics rivals such as FedEx, DHL, UPS, and Poste Italiane. 

The sanctions came exactly one month after the European General Court seconded the European Commission’s “discovery” in the Google Shopping case of a new antitrust infringement known as “self-preferencing,” which also cited Article 102 TFEU. Perhaps not entirely coincidentally, legislation was introduced in the United States earlier this year to prohibit the practice. Meanwhile, the EU’s legislative bodies have been busy taking steps to approve the Digital Markets Act (DMA), which would regulate so-called digital “gatekeepers.”

Italy thus joins a wave of policymakers that have either imposed heavy-handed decisions to “rein in” online platforms, or are seeking to implement ex ante regulations toward that end. Ultimately, the decision is reminiscent of the self-preferencing prohibition contained in Article 6a of the current draft of the DMA and reflects much of what is wrong with the current approach to regulating tech. It presages some of the potential problems with punishing efficient behavior for the sake of protecting competitors through “common carrier antitrust.” However, if this decision is anything to go by, these efforts will end up hurting the very consumers authorities purport to protect and lending color to more general fears over the DMA. 

In this post, we discuss how the AGCM’s reasoning departs from sound legal and economic thinking to reach a conclusion at odds with the traditional goal of competition law—i.e., the protection of consumer welfare. Neo-Brandeisians and other competition scholars who dispute the centrality of the consumer welfare standard and would use antitrust to curb “bigness” may find this result acceptable, in principle. But even they must admit that the AGCM decision ultimately serves to benefit large (if less successful) competitors, and not the “small dealers and worthy men” of progressive lore.

Relevant Market Definition

Market definition constitutes a preliminary step in any finding of abuse under Article 102 TFEU. An excessively narrow market definition can result in false positives by treating neutral or efficient conduct as anticompetitive, while an overly broad market definition might allow anticompetitive conduct to slip through the cracks, leading to false negatives. 

Amazon Italy may be an example of the former. Here, the AGCM identified two relevant markets: the leveraging market, which it identified as the Italian market for online marketplace intermediation, and the leveraged market, which it identified as the market for e-commerce logistics. The AGCM charges that Amazon is dominant in the former and that it gained an illegal advantage in the latter. It found, in this sense, that online marketplaces constitute a uniquely relevant market that is not substitutable for other offline or online sales channels, such as brick-and-mortar shops, price-comparison websites (e.g., Google Shopping), or dedicated sales websites (e.g., Nike.com/it). Similarly, it concluded that e-commerce logistics are sufficiently different from other forms of logistics as to comprise a separate market.

The AGCM’s findings combine qualitative and quantitative evidence, including retailer surveys and “small but significant and non-transitory increase in price” (SSNIP) tests. They also include a large dose of speculative reasoning.

For instance, the AGCM asserts that online marketplaces are fundamentally different from price-comparison sites because, in the latter case, purchase transactions do not take place on the platform. It asserts that e-commerce logistics are different from traditional logistics because the former require a higher degree of automation for transportation and storage. And in what can only be seen as a normative claim, rather than an objective assessment of substitutability, the Italian watchdog found that marketplaces are simply better than dedicated websites because, e.g., they offer greater visibility and allow retailers to save on marketing costs. While it is unclear what weights the AGCM assigned to each of these considerations when defining the relevant markets, it is reasonable to assume they played some part in defining the nature and scope of Amazon’s market presence in Italy.

In all of these instances, however, while the AGCM carefully delineated superficial distinctions between these markets, it did not actually establish that those differences are relevant to competition. Fetishizing granular but ultimately irrelevant differences between products and services—such as between marketplaces and shopping comparison sites—is a sure way to incur false positives, a misstep tantamount to punishing innocuous or efficient business conduct.

Dominance

The AGCM found that Amazon was “hyper-dominant” in the online marketplace intermediation market. Dominance was established by looking at revenue from marketplace sales, where Amazon’s share had risen from about 65% in 2016 to 75% in 2019. Taken in isolation, this figure might suggest that Amazon’s competitors cannot thrive in the market. A broader look at the data, however, paints a picture of more generalized growth, with some segments greatly benefiting newcomers and small, innovative marketplaces. 

For instance, virtually all companies active in the online marketplace intermediation market have experienced significant growth in terms of monthly visitors. It is true that Amazon’s visitors grew significantly, up 150%, but established competitors like Aliexpress and eBay also saw growth rates of 90% and 25%, respectively. Meanwhile, Wish grew a massive 10,000% from 2016 to 2019; while ManoMano and Zalando grew 450% and 100%, respectively.

In terms of active users (i.e., visits that result in a purchase), relative numbers seem to have stayed roughly the same, although the AGCM claims that eBay saw a 20-30% drop. The number of third-party products Amazon offered through Marketplace grew from between 100 and 500 million to between 500 million and 1 billion, while other marketplaces appear to have remained fairly constant, with some expanding and others contracting.

In sum, while Amazon has undeniably improved its position in practically all of the parameters considered by the AGCM, indicators show that the market as a whole has experienced and is experiencing growth. The improvement in Amazon’s position relative to some competitors—notably eBay, which AGCM asserts is Amazon’s biggest competitor—should therefore not obscure the fact that there is entry and expansion both at the fringes (ManoMano, Wish), and in the center of the market for online marketplace intermediation (Aliexpress).

Amazon’s Allegedly Abusive Conduct

According to the AGCM, Amazon has taken advantage of vertical integration to engage in self-preferencing. Specifically, the charge is that the company offers exclusive and purportedly crucial advantages on the Amazon.it marketplace to sellers who use Amazon’s own e-commerce logistics service, FBA. The purported advantages of this arrangement include, to name a few, the coveted Prime badge, the elimination of negative user feedback on sale or delivery, preferential algorithmic treatment, and exclusive participation in Amazon’s sales promotions (e.g., Black Friday, Cyber Monday). As a result, according to the AGCM, products sold through FBA enjoy more visibility and a better chance to win the “Buy Box.”

The AGCM claims this puts competing logistics operators like FedEx, Poste Italiane, and DHL at a disadvantage, because non-FBA products have less chance to be sold than FBA products, regardless of any efficiency or quality criteria. In the AGCM’s words, “Amazon has stolen demand for other e-commerce logistics operators.” 

Indirectly, Amazon’s “self-preferencing” purportedly also harms competing marketplaces like eBay by creating incentives for sellers to single-home—i.e., to sell only through Amazon Marketplace. The argument here is that retailers will not multi-home to avoid duplicative costs associated with FBA, e.g., storing goods in several warehouses. 

Although it is not necessary to demonstrate anticompetitive effects under Article 102 TFEU, the AGCM claims that Amazon’s behavior has caused drastic worsening in other marketplaces’ competitive position by constraining their ability to reach the minimum scale needed to enjoy direct and indirect network effects. The Italian authorities summarily assert that this results in consumer harm, although the gargantuan 250-page decision spends scarcely one paragraph on this point. 

Intuitively, however, Amazon’s behavior should, in principle, benefit consumers by offering something that most find tremendously valuable: a guarantee of quick delivery for a wide range of goods. Indeed, this is precisely why it is so misguided to condemn self-preferencing by online platforms.

As some have already argued, we cannot assume that something is bad for competition just because it is bad for certain competitors. For instance, a lot of unambiguously procompetitive behavior, like cutting prices, puts competitors at a disadvantage. The same might be true for a digital platform that preferences its own service because it is generally better than the alternatives provided by third-party sellers. In the case at hand, for example, Amazon’s granting marketplace privileges to FBA products may help users to select the products that Amazon can guarantee will best satisfy their needs. This is perfectly plausible, as customers have repeatedly shown that they often prefer less open, less neutral options.

The key question, therefore, should be whether the behavior in question excludes equally efficient rivals in such a way as to harm consumer welfare. Otherwise, we would essentially be asking companies to refrain from offering services that benefit their users in order to make competing products comparatively more attractive. This is antithetical to the nature of competition, which is based on the principle that what is good for consumers is frequently bad for competitors.

AGCM’s Theory of Harm Rests on Four Weak Pillars

Building on the logic that Amazon enjoys “hyper-dominance” in marketplace intermediation; that most online sales are marketplace sales; and that most marketplace sales are, in turn, Amazon.it sales, the AGCM decision finds that succeeding on Amazon.it is indispensable for any online retailer in Italy. This argument hinges largely on whether online and offline retailers are thought of as distinct relevant markets—i.e., whether, from the perspective of the retailer, online and offline sales channels are substitutable (see also the relevant market definition section above). 

Ultimately, the AGCM finds that they are not, as online sales enjoy such advantages as lower fixed costs, increased sale flexibility, and better geographical reach. To an outsider, the distinction between the two markets may seem artificial—and it largely is—but such theoretical market segmentation is the bread-and-butter of antitrust analysis. Still, even by EU competition law standards, the relevant market definitions on which the AGCM relies to conclude that selling on Amazon is indispensable appear excessively narrow. 

This market distinction also serves to set up the AGCM’s second, more controversial argument: that the benefits extended to products sold through the FBA channel are also indispensable for retailers’ success on the Amazon.it marketplace. Here, the AGCM seeks a middle ground between competitive advantage and indispensability, finally settling on the notion that a sufficiently large competitive advantage itself translates into indispensability.

But how big is too big? The facts that 40-45% of Amazon’s third-party retailers do not use FBA (p. 57 of the decision) and that roughly 40 of the top 100 products sold on Amazon.it are not fulfilled through Amazon’s logistics service (p. 58) would appear to suggest that FBA is more of a convenience than an obligation. At the least, it does not appear that the advantage conferred is so big as to amount to indispensability. This may be because sellers that choose not to use Amazon’s logistics service (including offline, of course) can and do cut prices to compete with FBA-sold products. If anything, this should be counted as a good thing from the perspective of consumer welfare.

Instead, and signaling the decision’s overarching preoccupation with protecting some businesses at the expense of others (and, ultimately, at the expense of consumers), the AGCM has expanded the already bloated notion of a self-preferencing offense to conclude that expecting sellers to compete on pricing parameters would unfairly slash profit margins for non-FBA sellers.

The third pillar of the AGCM’s theory of harm is the claim that the benefits conferred on products sold through FBA are not awarded based on any objective quality criteria, but purely on whether the seller has chosen FBA or third-party logistics. Thus, even if a logistics operator were, in principle, capable of offering a service as efficient as FBA’s, it would not qualify for the same benefits. 

But this is a disingenuous line of reasoning. One legitimate reason why Amazon could choose to confer exclusive advantages on products fulfilled by its own logistics operation is because no other service is, in fact, routinely as reliable. This does not necessarily mean that FBA is always superior to the alternatives, but rather that it makes sense for Amazon to adopt this presumption a general rule based on past experience, without spending the resources to constantly evaluate it. In other words, granting exclusive benefits is based on quality criteria, just on a prior measurement of quality rather than an ongoing assessment. This is presumably what a customer-obsessed business that does not want to take chances with consumer satisfaction would do. 

Fourth, the AGCM posits that Prime and FBA constitute two separate products that have been artificially tied by Amazon, thereby unfairly excluding third-party logistics operators. Co-opting Amazon’s own terminology, the AGCM claims that the company has created a flywheel of artificial interdependence, wherein Prime benefits increase the number of Prime users, which drives demand for Prime products, which creates demand for Prime-eligible FBA products, and so on. 

To support its case, the AGCM repeatedly adduces a 2015 letter in which Jeff Bezos told shareholders that Amazon Marketplace and Prime are “happily and deeply intertwined,” and that FBA is the “glue” that links them together. Instead of taking this for what it likely is—i.e., a case of legitimate, efficiency-enhancing vertical integration—the AGCM has preferred to read into it a case of illicit tying, an established offense under Article 102 TFEU whereby a dominant firm makes the purchase of one product conditional on the purchase of another, unrelated one. 

The problem with this narrative is that it is perfectly plausible that Prime and FBA are, in fact, meant to be one product that is more than the sum of its parts. For one, the inventory of sellers who use FBA is stowed in fulfillment centers, meaning that Amazon takes care of all logistics, customer service, and product returns. As Bezos put it in the same 2015 letter, this is a huge efficiency gain. It thus makes sense to nudge consumers towards products that use FBA.

In sum, the AGCM’s case rests on a series of questionable assumptions that build on each other: a narrow relevant market definition; a finding of “hyper-dominance” that downplays competitors’ growth and expansion, as well as competition from outside the narrowly defined market; a contrived notion of indispensability at two levels (Marketplace and FBA); and a refusal to contemplate the possibility that Amazon integrates its marketplace and logistics services in orders to enhance efficiency, rather than to exclude competitors.

Remedies

The AGCM sees “only one way to restore a level-playing field in e-commerce logistics”: Amazon must redesign its existing Self-Fulfilled Prime (SFP) program in such a way as to grant all logistics operators—FBA or non-FBA—equal treatment on Amazon.it, based on a set of objective, transparent, standard, uniform, and non-discriminatory criteria. Any logistics operator that demonstrates the ability to fulfill such criteria must be awarded SFP status and the accompanying Prime badge, along with all the perks associated with it. Further, SFP- and FBA-sold products must be subject to the same monitoring mechanism with regard to the observance of Prime standards, as well as to the same evaluation standards. 

In sum, Amazon Italy now has a duty to treat Marketplace sales fulfilled by third-party operators the same as those fulfilled by its own logistics service. This is a significant step toward “common carrier antitrust.” in which vertically integrated firms are expected to comply with perfect neutrality obligations with respect to customers, suppliers, and competitors

Beyond the philosophical question of whether successful private companies should be obliged by law to treat competitors analogously to its affiliates (they shouldn’t), the pitfalls of this approach are plain to see. Nearly all consumer-facing services use choice architectures as a means to highlight products that rank favorably in terms of price and quality, and ensuring consumers enjoy a seamless user experience: Supermarkets offer house brands that signal a product has certain desirable features; operating system developers pre-install certain applications to streamline users’ “out of the box “experience; app stores curate the apps that users will view; search engines use specialized boxes that anticipate the motives underlying users’ search queries, etc. Suppressing these practices through heavy-handed neutrality mandates is liable to harm consumers. 

Second, monitoring third-party logistics operators’ compliance with the requisite standards is going to come at a cost for Amazon (and, presumably, its customers)—a cost likely much higher than that of monitoring its own operations—while awarding the Prime badge liberally may deteriorate the consumer experience on Amazon Marketplace.

Thus, one way for Amazon to comply with AGCM’s remedies while also minimizing monitoring costs is simply to dilute or even remove the criteria for Prime, thereby allowing sellers using any logistics provider to be eligible for Prime. While this would presumably insulate Amazon from any future claims against exclusionary self-preferencing, it would almost certainly also harm consumer welfare. 

A final point worth noting is that vertical integration may well be subsidizing Amazon’s own first-party products. In other words, even if FBA is not fully better than other logistics operators, the revenue that it derives from FBA enables Amazon to offer low prices, as well as a range of other benefits from Prime, such as, e.g., free video. Take that source of revenue away, and those subsidized prices go up and the benefits disappear. This is another reason why it may be legitimate to consider FBA and Prime as a single product.

Of course, this argument is moot if all one cares about is how Amazon’s vertical integration affects competitors, not consumers. But consumers care about the whole package. The rationale at play in the AGCM decision ultimately ends up imposing a narrow, boring business model on all sellers, precluding them from offering interesting consumer benefits to bolster their overall product.

Conclusion

Some have openly applauded AGCM’s use of EU competition law to protect traditional logistics operators like FedEx, Poste Italiane, DHL, and UPS. Others lament the competition authority’s apparent abandonment of the consumer welfare standard in favor of a renewed interest in punishing efficiency to favor laggard competitors under the guise of safekeeping “competition.” Both sides ultimately agree on one thing, however: Amazon Italy is about favoring Amazon’s competitors. If competition authorities insist on continuing down this populist rabbit hole,  the best they can hope for is a series of Pyrrhic victories against the businesses that are most bent on customer satisfaction, i.e., the successful ones.

Some may intuitively think that this is fair; that Amazon is just too big and that it strangles small competitors. But Amazon’s “small” competitors are hardly the “worthy men” of Brandeisian mythology. They are FedEx, DHL, UPS, and the state-backed goliath Poste Italiane; they are undeniably successful companies like eBay, Alibaba – or Walmart in the United States. It is, conversely, the smallest retailers and consumers who benefit the most from Amazon’s integrated logistics and marketplace services, as the company’s meteoric rise in popularity in Italy since 2016 attests. But it seems that, in the brave new world of antitrust, such stakeholders are now too small to matter.

A lawsuit filed by the State of Texas and nine other states in December 2020 alleges, among other things, that Google has engaged in anticompetitive conduct related to its online display-advertising business.

Broadly, the Texas complaint (previously discussed in this TOTM symposium) alleges that Google possesses market power in ad-buying tools and in search, illustrated in the figure below.

The complaint also alleges anticompetitive conduct by Google with respect to YouTube in a separate “inline video-advertising market.” According to the complaint, this market power is leveraged to force transactions through Google’s exchange, AdX, and its network, Google Display Network. The leverage is further exercised by forcing publishers to license Google’s ad server, Google Ad Manager.

Although the Texas complaint raises many specific allegations, the key ones constitute four broad claims: 

  1. Google forces publishers to license Google’s ad server and trade in Google’s ad exchange;
  2. Google uses its control over publishers’ inventory to block exchange competition;
  3. Google has disadvantaged technology known as “header bidding” in order to prevent publishers from accessing its competitors; and
  4. Google prevents rival ad-placement services from competing by not allowing them to buy YouTube ad space.

Alleged harms

The Texas complaint alleges Google’s conduct has caused harm to competing networks, exchanges, and ad servers. The complaint also claims that the plaintiff states’ economies have been harmed “by depriving the Plaintiff States and the persons within each Plaintiff State of the benefits of competition.”

In a nod to the widely accepted Consumer Welfare Standard, the Texas complaint alleges harm to three categories of consumers:

  1. Advertisers who pay for their ads to be displayed, but should be paying less;
  2. Publishers who are paid to provide space on their sites to display ads, but should be paid more; and
  3. Users who visit the sites, view the ads, and purchase or use the advertisers’ and publishers’ products and services.

The complaint claims users are harmed by above-competitive prices paid by advertisers, in that these higher costs are passed on in the form of higher prices and lower quality for the products and services they purchase from those advertisers. The complaint simultaneously claims that users are harmed by the below-market prices received by publishers in the form of “less content (lower output of content), lower-quality content, less innovation in content delivery, more paywalls, and higher subscription fees.”

Without saying so explicitly, the complaint insinuates that if intermediaries (e.g., Google and competing services) charged lower fees for their services, advertisers would pay less, publishers would be paid more, and consumers would be better off in the form of lower prices and better products from advertisers, as well as improved content and lower fees on publishers’ sites.

Effective competition is not an antitrust offense

A flawed premise underlies much of the Texas complaint. It asserts that conduct by a dominant incumbent firm that makes competition more difficult for competitors is inherently anticompetitive, even if that conduct confers benefits on users.

This amounts to a claim that Google is acting anti-competitively by innovating and developing products and services to benefit one or more display-advertising constituents (e.g., advertisers, publishers, or consumers) or by doing things that benefit the advertising ecosystem more generally. These include creating new and innovative products, lowering prices, reducing costs through vertical integration, or enhancing interoperability.

The argument, which is made explicitly elsewhere, is that Google must show that it has engineered and implemented its products to minimize obstacles its rivals face, and that any efficiencies created by its products must be shown to outweigh the costs imposed by those improvements on the company’s competitors.

Similarly, claims that Google has acted in an anticompetitive fashion rest on the unsupportable notion that the company acts unfairly when it designs products to benefit itself without considering how those designs would affect competitors. Google could, it is argued, choose alternate arrangements and practices that would possibly confer greater revenue on publishers or lower prices on advertisers without imposing burdens on competitors.

For example, a report published by the Omidyar Network sketching a “roadmap” for a case against Google claims that, if Google’s practices could possibly be reimagined to achieve the same benefits in ways that foster competition from rivals, then the practices should be condemned as anticompetitive:

It is clear even to us as lay people that there are less anticompetitive ways of delivering effective digital advertising—and thereby preserving the substantial benefits from this technology—than those employed by Google.

– Fiona M. Scott Morton & David C. Dinielli, “Roadmap for a Digital Advertising Monopolization Case Against Google”

But that’s not how the law—or the economics—works. This approach converts beneficial aspects of Google’s ad-tech business into anticompetitive defects, essentially arguing that successful competition and innovation create barriers to entry that merit correction through antitrust enforcement.

This approach turns U.S. antitrust law (and basic economics) on its head. As some of the most well-known words of U.S. antitrust jurisprudence have it:

A single producer may be the survivor out of a group of active competitors, merely by virtue of his superior skill, foresight and industry. In such cases a strong argument can be made that, although, the result may expose the public to the evils of monopoly, the Act does not mean to condemn the resultant of those very forces which it is its prime object to foster: finis opus coronat. The successful competitor, having been urged to compete, must not be turned upon when he wins.

– United States v. Aluminum Co. of America, 148 F.2d 416 (2d Cir. 1945)

U.S. antitrust law is intended to foster innovation that creates benefits for consumers, including innovation by incumbents. The law does not proscribe efficiency-enhancing unilateral conduct on the grounds that it might also inconvenience competitors, or that there is some other arrangement that could be “even more” competitive. Under U.S. antitrust law, firms are “under no duty to help [competitors] survive or expand.”  

To be sure, the allegations against Google are couched in terms of anticompetitive effect, rather than being described merely as commercial disagreements over the distribution of profits. But these effects are simply inferred, based on assumptions that Google’s vertically integrated business model entails an inherent ability and incentive to harm rivals.

The Texas complaint claims Google can surreptitiously derive benefits from display advertisers by leveraging its search-advertising capabilities, or by “withholding YouTube inventory,” rather than altruistically opening Google Search and YouTube up to rival ad networks. The complaint alleges Google uses its access to advertiser, publisher, and user data to improve its products without sharing this data with competitors.

All these charges may be true, but they do not describe inherently anticompetitive conduct. Under U.S. law, companies are not obliged to deal with rivals and certainly are not obliged to do so on those rivals’ preferred terms

As long ago as 1919, the U.S. Supreme Court held that:

In the absence of any purpose to create or maintain a monopoly, the [Sherman Act] does not restrict the long recognized right of [a] trader or manufacturer engaged in an entirely private business, freely to exercise his own independent discretion as to parties with whom he will deal.

– United States v. Colgate & Co.

U.S. antitrust law does not condemn conduct on the basis that an enforcer (or a court) is able to identify or hypothesize alternative conduct that might plausibly provide similar benefits at lower cost. In alleging that there are ostensibly “better” ways that Google could have pursued its product design, pricing, and terms of dealing, both the Texas complaint and Omidyar “roadmap” assert that, had the firm only selected a different path, an alternative could have produced even more benefits or an even more competitive structure.

The purported cure of tinkering with benefit-producing unilateral conduct by applying an “even more competition” benchmark is worse than the supposed disease. The adjudicator is likely to misapply such a benchmark, deterring the very conduct the law seeks to promote.

For example, Texas complaint alleges: “Google’s ad server passed inside information to Google’s exchange and permitted Google’s exchange to purchase valuable impressions at artificially depressed prices.” The Omidyar Network’s “roadmap” claims that “after purchasing DoubleClick, which became its publisher ad server, Google apparently lowered its prices to publishers by a factor of ten, at least according to one publisher’s account related to the CMA. Low prices for this service can force rivals to depart, thereby directly reducing competition.”

In contrast, as current U.S. Supreme Court Associate Justice Stephen Breyer once explained, in the context of above-cost low pricing, “the consequence of a mistake here is not simply to force a firm to forego legitimate business activity it wishes to pursue; rather, it is to penalize a procompetitive price cut, perhaps the most desirable activity (from an antitrust perspective) that can take place in a concentrated industry where prices typically exceed costs.”  That commentators or enforcers may be able to imagine alternative or theoretically more desirable conduct is beside the point.

It has been reported that the U.S. Justice Department (DOJ) may join the Texas suit or bring its own similar action against Google in the coming months. If it does, it should learn from the many misconceptions and errors in the Texas complaint that leave it on dubious legal and economic grounds.