Spotlight

November 2023

HIGHLIGHTS

ICLE Comments on Artificial Intelligence and Copyright

Introduction We thank you for the opportunity to comment on this important notice of inquiry (NOI)[1] on artificial intelligence (AI) and copyright. We appreciate the . . .

Introduction

We thank you for the opportunity to comment on this important notice of inquiry (NOI)[1] on artificial intelligence (AI) and copyright. We appreciate the U.S. Copyright Office undertaking a comprehensive review of the policy and copyright-law issues raised by recent advances in generative AI systems. This NOI covers key areas that require attention, from legal questions regarding infringement and fair use, to questions about how policy choices could shape opportunities for creators and AI producers to engage in licensing.

At this early date, AI systems have already generated some incredible visual art and impressive written texts, as well as a good deal of controversy. Some artists have banded together as part of an anti-AI campaign;[2] lawsuits have been filed;[3] and policy experts have attempted to think through the various legal questions raised by these machine-learning systems.

The debates over the role of AI in creative industries have particular salience for intellectual-property rights. Copyright is notoriously difficult to protect online, and the emergence of AI may exacerbate that difficulty. AI systems also potentially pose an additional wrinkle: it is at least arguable that the outputs they produce can themselves be considered unique creations. There are, of course, other open questions whose answers are relevant here, not the least being whether it is fair to assert that only a human can be “creative” (at least, so far).[4]

But leaving these questions aside, we can say that at least some AI systems produce unique outputs and are not merely routinely duplicating other pieces of work in a digital equivalent of collage. That is, at some level, the machines are engaged in a rudimentary sort of “learning” about how humans arrange creative inputs when generating images, music, or written works. The machines appear to be able to reconstruct this process and produce new sets of words, sounds, or lines and colors that conform to the patterns found in human art, in at least a simulacrum of “creativity.”

But that conclusion isn’t the end of the story. Even if some of these AI outputs are unique and noninfringing, the way that AI systems learn—by ingesting massive quantities of existing creative work—raises a number of thorny copyright-law issues. Indeed, some argue that these systems inherently infringe copyright during the learning phase and that, as discussed below, such processes may not survive a “fair use” analysis.

But nor is that assertion the end of the analysis. Rather, it raises the question of whether applying existing doctrine in this novel technological context yields the best results for society. Moreover, it heightens the need for a comprehensive analytical framework to help parse these questions.

A.            The Law & Economics of Copyright and AI

Nearly all would agree that it is crucial that law and public policy strike the appropriate balance between protecting creators’ existing rights and enabling society to enjoy the potentially significant benefits that could arise from the development of AI systems. Indeed, the subject is often cast as a dramatic conflict between creative professionals struggling to make ends meet and innovative firms working to provide cutting-edge AI technology. For the moment, however, it is likely more important to determine the right questions to ask and the proper analytical framework to employ than it is to identify any precise balancing point.

What is important to remember is that copyright policy is foremost economic in nature and “can be explained as a means for promoting efficient allocation of resources.”[5] That is to say, the reason that property rights in creative expression exist is to guarantee the continued production of such works.[6] The fundamental tradeoff in copyright policy is between the costs of limiting access to creative works, and the value obtained by encouraging production of such works.[7] The same applies in the context of AI: identifying the key tradeoffs and weighing the costs and benefits of restricting access to protected works by the producers (and users) of AI systems.[8]

This entails examining the costs and benefits of relatively stronger or weaker forms copyright protection in terms of their effects on both incentives and access, and as they relate to both copyright holders and AI-system developers. It also requires considering where the transaction costs should be allocated for negotiating access to both copyright and, as discussed infra,[9] the use of name/image/likeness, as well as how those allocations are likely to shape outcomes.

At root, these questions center on how to think about the property rights that limit access to protected works and, possibly even more importantly, how to assign new property rights governing the ability to control the use of a name/image/likeness. As we know from the work of the late Nobel laureate Ronald Coase, the actual demarcation of rights affects parties’ abilities to negotiate superior solutions.[10] The development of nuisance law provides a good example of the problem at hand. When a legal regime provides either strict liability or no-liability rules around pollution, parties have little incentive to minimize harmful conduct:

The factory that has the absolute right to pollute will, if transaction costs are prohibitive, have no incentives to stop (or reduce) pollution even if the cost of stopping would be much less than the cost of pollution to the homeowners. Conversely, homeowners who have an absolute right to be free from pollution will, if transaction costs are prohibitive, have no incentive to take steps of their own to reduce the effects of pollution even if the cost to them of doing so (perhaps by moving away) is less than the cost to the factory of not polluting or of polluting less.[11]

As Coase observed, this class of problem is best regarded as reciprocal in nature, and the allocation of rights matters in obtaining an efficient outcome. This is necessarily so because, when fully considered, B’s ability to restrain A from the pollution-generating activity can itself be conceived of as another kind of harm that B can impose on A. Therefore, the problem requires a balancing of the relative harms generated by both A and B in exercising conflicting claims in a particular context.

When thinking about how to minimize harms—whether from pollution or other activity that generates social costs (which is to say, nearly every activity)—the aim is to decide whether “the gain from preventing the harm is greater than the loss which would be suffered elsewhere as a result of stopping the action which produces the harm.”[12] Theoretically, in a world without transaction costs, even assignments of no-liability or strict-liability rules could be bargained around. But we do not live in such a world.[13] Thus, “[i]n a world in which there are costs of rearranging the rights established by the legal system [common law and statutory assignments of liability] are, in effect, making a decision on the economic problem and determining how resources are to be employed.”[14]

While pollution rules, unlicensed uses of intellectual property, and a host of other activities subject to legal sanction are not typically framed as resource-allocation decisions, it is undeniable that they do have this character. This is true even where legislation attempts to correct deficiencies in the system. We experience a form of blindness when we focus on correcting what may be rightly perceived as problems in a liability regime. Such analysis tends to concentrate attention on particular deficiencies of the system and to nourish the belief that any measure that removes the deficiency is necessarily desirable. It diverts attention from other changes inevitably associated with the corrective measure—changes that may well produce more harm than the original deficiency.[15]

All of this is to say that one solution to the costs generated by the need for AI systems to process a massive corpus of expensive, copyright-protected material is neither to undermine property rights, nor to make AI impossible, but to think about how new property rights could make the system work. It may be that some entirely different form or allocation of property right would facilitate bargaining between rightsholders and AI creators, optimizing resource allocation in a way the existing doctrinal regime may not be able to.

A number of other questions flow from this insight into the allocative nature of copyright. How would the incentives for human creators change under different copyright rules for AI systems, or in the face of additional rights? And how would access to copyrighted works for AI training change with different rules, and what effects would that access have on AI innovation?

Above all, our goal today should be to properly frame the AI and copyright debate by identifying tradeoffs, quantifying effects (where possible), and asking what rules best serve the overall objectives of the copyright system and the social goal of encouraging AI innovation. The best chance of striking the right balance will come from a rigorous framing of the questions and from the use of economic analysis to try to answer them.

B.            Copyright Law and AI: Moving Forward

As the Copyright Office undertakes this inquiry, it is important to recognize that, regardless of how the immediate legal questions around AI and copyright are resolved, the growing capabilities and adoption of generative AI systems will likely necessitate some changes in the long term.

The complex questions surrounding the intersection of AI and copyright law admit reasonable arguments on both sides. But AI is here to stay, regardless, and if copyright law is applied in an unduly restrictive manner that substantially hinders socially beneficial AI innovation, it could provoke a broader public-policy backlash that does more to harm copyright’s ability to protect creative works than it does to stanch AI’s ability to undermine it. Copyright law risks being perceived as an obstruction to technological progress if it is used preemptively to kill AI in the cradle. Such an outcome could galvanize calls for recalibrating copyright’s scope and protections in the name of the public interest.

This illustrates the precarious balancing act that copyright law faces in the wake of rapidly evolving technologies like AI. Aggressive copyright restrictions that curtail AI development could instigate a public-policy counter-reaction before Congress and the courts that ultimately undermines copyright’s objectives. The judicious course is to adapt copyright law cautiously to enable AI’s responsible evolution, while resolutely preserving the incentives for human creativity.

In the remainder of this analysis, we offer our perspective on the likely outcomes of the AI-copyright issues raised in this NOI, given the current state of the law. These assessments reflect our perspective formed through the rigorous application of established copyright principles and precedent to the novel technological context of generative AI systems. Reasonable arguments rooted in existing doctrine could be made to support different conclusions. We submit these comments not as definitive predictions or normative preferences, but rather as informed appraisals of how courts may analyze AI under present copyright law, absent legislative intervention.

We appreciate the Copyright Office starting this process to modernize copyright law for the AI age. This inquiry is an important first step, but openness to further evolution will be key to promoting progress in both AI and the arts. We believe an open, evidence-based discussion of these issues will lead to balanced solutions that uphold copyright’s constitutionally mandated purpose, while allowing responsible AI innovation for the public benefit.

II.            The Training of AI Systems and the Applicability of Fair Use

In the NOI, the Copyright Offices asks: “[u]nder what circumstances would the unauthorized use of copyrighted works to train AI models constitute fair use?”[16]

To answer this question, it would be useful to first briefly walk through a high-level example of how AI systems work, in order to address the most relevant points of contact between AI systems and copyright law.

A.            A Brief Technical Description of AI Training

AI-generated content is not a single “thing,” but a collection of differing processes, each with different implications for the law. For the purposes of this discussion, we will discuss image generation using “generated adversarial networks” (GANs) and diffusion models. Although different systems and different types of content generation will vary, the basic concepts discussed below are nonetheless useful at a general level.[17]

A GAN is a type of machine-learning model that consists of two parts: a generator and a discriminator.[18] The generator is trained to create new images that look like they come from a particular dataset, while the discriminator is trained to distinguish the generated images from real images in its original dataset.[19] The two parts are trained together in an adversarial manner, with the generator trying to produce images that can fool the discriminator and the discriminator trying to correctly identify the generated images.[20]

A diffusion model, by contrast, analyzes the distribution of information in an image, as noise is progressively added to it.[21] This kind of algorithm analyzes characteristics of sample images, like the distribution of colors or lines, in order to understand what counts as an accurate representation of a subject (i.e., what makes a picture of a cat look like a cat, and not like a dog).[22]

For example, in the generation phase, diffusion-based systems start with randomly generated noise, and work backward in “denoising” steps to essentially “see” shapes:

The sampled noise is predicted so that if we subtract it from the image, we get an image that’s closer to the images the model was trained on (not the exact images themselves, but the distribution – the world of pixel arrangements where the sky is usually blue and above the ground, people have two eyes, cats look a certain way – pointy ears and clearly unimpressed).[23]

While it is possible that some implementations might be designed in a way that saves copies of the training images,[24] for at least some systems, once the network is trained using these techniques, it will not need to rely on saved copies of input work in order to produce outputs. The models that are produced during training are, in essence, instructions to a different piece of software about how to start with a prompt from a user, a palette of pure noise, and progressively “discover” signal in that image until some new image emerges.

B.            Fair Use

The creator of some of the most popular AI tools, OpenAI, is not shy about their use of protected works in the training phase of the algorithms. In comments to the U.S. Patent and Trademark Office (PTO), OpenAI noted that:

Modern AI systems require large amounts of data. For certain tasks, that data is derived from existing publicly accessible “corpora”… of data that include copyrighted works. By analyzing large corpora (which necessarily involves first making copies of the data to be analyzed), AI systems can learn patterns inherent in human-generated data and then use those patterns to synthesize similar data which yield increasingly compelling novel media in modalities as diverse as text, image, and audio. (emphasis added).[25]

Thus, at the training stage, the most popular forms of AI systems require making copies of existing works. And where that material is either not in the public domain or is not licensed, an infringement can occur. Thus, the copy must not be infringing (say, because it is transient), or some affirmative defense is needed to excuse the infringement. Toward this end, OpenAI believes that this use should qualify as fair use,[26] as do most or all the other major producers of generative AI systems.[27]

But as OpenAI has framed the fair-use analysis, it is not clear that these uses should qualify. There are two major questions in this respect: will the data used to train these systems count as “copies” under the Copyright Act, and, if so, is the use of these “copies” sufficiently “transformative” to qualify for the fair-use defense?

1.              Are AI systems being trained with ‘copies’ of protected works?

Section 106 of the Copyright Act grants the owner of a copyright the exclusive right “to reproduce… copyrighted work in copies” and to authorize others to do so.[28] If an AI system makes a copy of a file to a computer during training, this would likely constitute a prima facie violation of the copyright owner’s exclusive right of reproduction under Section 106. This is fairly straightforward.

But what if the “copy” is “transient” and/or only partial pieces of content are used in the training? For example, what if a training program merely streamed small bits of a protected work into temporary memory as part of its training, and retained no permanent copy?

As the Copyright Office has previously observed, even temporary reproductions of a work in a computer’s memory can constitute “copies” under the Copyright Act.[29] Critically, this includes even temporary reproductions made as part of a packet-switching network transmission, where a particular file is broken into individual packets, because the packets can be reassembled into substantial portions or even entire works.[30] On the topic of network-based transmission, the Copyright Office further observed that:

Digital networks permit a single disk copy of a work to meet the demands of many users by creating multiple RAM copies. These copies need exist only long enough to be perceived (e.g., displayed on the screen or played through speakers), reproduced or otherwise communicated (e.g., to a computer’s processing unit) in order for their economic value to be realized. If the network is sufficiently reliable, users have no need to retain copies of the material. Commercial exploitation in a network environment can be said to be based on selling a right to perceive temporary reproductions of works.[31]

This is a critical insight that translates well to the context of AI training. The “transience” of the copy matters with respect to the receiver’s ability to perceive the work in a way that yields commercial value. Under this reasoning, the relevant locus of analysis is on the AI system’s ability to “perceive” a work for the purposes of being trained to “understand” the work. In this sense, you could theoretically find the existence of even more temporary copies than that necessary for human perception to implicate the reproduction right.

Even where courts have been skeptical of extending the definition of “copy” to “fleeting” copies in computer memory, this underlying logic is revealed. In Cartoon Network LP, LLLP v. CSC Holdings, Inc., 536 F.3d 121 (2008), the 2nd U.S. Circuit Court of Appeals had to determine whether buffered media sent to a DVR device was too “transient” to count as a “copy”:

No bit of data remains in any buffer for more than a fleeting 1.2 seconds. And unlike the data in cases like MAI Systems, which remained embodied in the computer’s RAM memory until the user turned the computer off, each bit of data here is rapidly and automatically overwritten as soon as it is processed. While our inquiry is necessarily fact-specific, and other factors not present here may alter the duration analysis significantly, these facts strongly suggest that the works in this case are embodied in the buffer for only a “transitory” period, thus failing the duration requirement.[32]

In Cartoon Network, the court acknowledged both that the duration analysis was fact-bound, and also that the “fleeting” nature of the reproduction was important. “Fleeting” is a relative term, based on the receiver’s capacities. A ball flying through the air may look “fleeting” to a human observer, but may appear to go much more cognizable to a creature with faster reaction time, such as a house fly. So, too, with copies of a work in a computer’s memory and the ability to “perceive” what is fixed in a buffer: what may be much too quick for a human to perceive may very well be within an AI system’s perceptual capabilities.

Therefore, however the training copies are held, there is a strong possibility that a court will find them to be “copies” for the purposes of the reproduction right—even with respect to partial copies that exist for very small amounts of time.

2.              The purpose and character of using protected works to train AI systems

Fair use provides for an affirmative defense against infringement when the use is, among other things, “for purposes such as criticism, comment, news reporting, teaching…, scholarship, or research.”[33] When deciding whether a fair-use defense is applicable, a court must balance a number of factors:

  1. the purpose and character of the use, including whether such use is of a commercial nature or is for nonprofit educational purposes;
  2. the nature of the copyrighted work;
  3. the amount and substantiality of the portion used in relation to the copyrighted work as a whole; and
  4. the effect of the use upon the potential market for or value of the copyrighted work.[34]

The fair-use defense that AI creators have advanced is rooted in the first factor: the nature and character of the use. Although a full analysis of all the factors is ultimately necessary, analysis of the first factor is sufficiently complicated to warrant full attention here. In particular, the complex issue at hand is whether uses of protected works to train AI systems are sufficiently “transformative” or not.[35]

Whether the use of a copyrighted work to train an AI is “transformative” is certainly a novel question, but it is one that will likely be answered in light of an observation the U.S. Supreme Court made in Campbell v. Acuff Rose Music:

[W]hen a commercial use amounts to mere duplication of the entirety of an original, it clearly “supersede[s] the objects,”… of the original and serves as a market replacement for it, making it likely that cognizable market harm to the original will occur… But when, on the contrary, the second use is transformative, market substitution is at least less certain, and market harm may not be so readily inferred.[36]

Moreover, “[t]he word ‘transformative’ cannot be taken too literally as a sufficient key to understanding the elements of fair use. It is rather a suggestive symbol for a complex thought, and does not mean that any and all changes made to an author’s original text will necessarily support a finding of fair use.”[37] A key question, then, is whether training AI systems on copyrighted works amounts to a mere “duplication of the entirety of an original” or is sufficiently “transformative” to support a fair-use defense. As noted above, OpenAI believes that its use is transformative. According to its comments:

Training of AI systems is clearly highly transformative. Works in training corpora were meant primarily for human consumption for their standalone entertainment value. The “object of the original creation,” in other words, is direct human consumption of the author’s ?expression.? Intermediate copying of works in training AI systems is, by contrast, “non-expressive” the copying helps computer programs learn the patterns inherent in human-generated media. The aim of this process—creation of a useful generative AI system—is quite different than the original object of human consumption. The output is different too: nobody looking to read a specific webpage contained in the corpus used to train an AI system can do so by studying the AI system or its outputs. The new purpose and expression are thus both highly transformative.[38]

This framing, however, works against OpenAI’s interests. As noted above, and reinforced in the immediately preceding quote, generative AI systems are made of at least two distinct pieces. The first is a piece of software that ingests existing works and creates a file that can serve as instructions to the second piece of software. The second piece of software takes the output of the first and can produce independent results. Thus, there is a clear discontinuity in the process whereby the ultimate work created by the system is disconnected from the creative inputs used to train the software.

Therefore, the protected works are arguably ingested into the first part of the system “for their standalone entertainment value.” That is to say, the goal of copying and showing a protected work to an AI system is for the analog of “direct human consumption of the author’s expression” in order for the system to learn about that expression.

The software is learning what counts as “standalone entertainment value” and therefore the works must be used in those terms. Surely, a computer is not sitting on a couch and surfing for its pleasure. But it is solely for the very “standalone entertainment value” that the first piece of software is being shown copyrighted material. By contrast, parody or “remixing” uses incorporate a work into some secondary expression that directly transforms the input. The way these systems work is to learn what makes a piece entertaining and then to discard that piece altogether. Moreover, this use for the art qua art most certainly interferes with the existing market, insofar as this use is in lieu of reaching a licensing agreement with rightsholders.

A good analogy is art students and art textbooks. Art students view protected works in an art textbook in order to learn how to reproduce the styles contained therein. The students would not be forgiven for pirating the textbooks merely because they intend to go on to make new paintings. They would still be liable for copyright infringement if they used unlicensed protected works as part of their education.

The 2nd U.S. Circuit Court of Appeals dealt with a case that operates similarly to this dynamic. In American Geophysical Union v. Texaco, 60 F.3d 913 (2d Cir. 1994), the 2nd Circuit considered whether Texaco’s photocopying of scientific articles produced by the plaintiffs qualified for a fair-use defense. Texaco employed between 400 and 500 research scientists and, as part of supporting their work, maintained subscriptions to a number of scientific journals.[39]

It was common practice for Texaco’s scientists to photocopy entire articles and save them in a file.[40] The plaintiffs sued for copyright infringement.[41] Texaco asserted that photocopying by its scientists for the purposes of furthering scientific research—that is to train the scientists on the content of the journal articles—should count as a fair use. The argument was, at least in part, that this was sufficiently “transformative,” because the scientists were using that knowledge to invent new products.[42] The 2nd Circuit disagreed:

The “transformative use” concept is pertinent to a court’s investigation under the first factor because it assesses the value generated by the secondary use and the means by which such value is generated. To the extent that the secondary use involves merely an untransformed duplication, the value generated by the secondary use is little or nothing more than the value that inheres in the original. Rather than making some contribution of new intellectual value and thereby fostering the advancement of the arts and sciences, an untransformed copy is likely to be used simply for the same intrinsic purpose as the original, thereby providing limited justification for a finding of fair use….[43]

The 2nd Circuit thus observed that copies of the scientific articles were made solely to consume the material itself. AI developers often make an argument analogous to that made by Texaco: that training AI systems surely advances scientific research, and therefore fosters the “advancement of the arts and sciences.” But in American Geophysical Union, the initial copying of copyrighted content, even where it was ultimately used for the “advancement of the arts and sciences,” was not held to be sufficiently “transformative.”[44] The case thus stands for the proposition that one cannot merely identify a social goal down that would be advanced at some future date in order to permit an exception to copyright protection. As the court put it:

[T]he dominant purpose of the use is a systematic institutional policy of multiplying the available number of copies of pertinent copyrighted articles by circulating the journals among employed scientists for them to make copies, thereby serving the same purpose for which additional subscriptions are normally sold, or… for which photocopying licenses may be obtained.[45]

The use itself must be transformative and different, and copying is not transformative merely because it may be used as an input into a later transformative use. By the same token, therefore, it seems likely that where an AI system ingests (copies) copyrighted works, that use is similarly not transformative, despite its ultimate use as an input in the creation of other original works.

Comparing the American Geophysical Union analysis with the search-engine “snippets” and “thumbnails” cases provides a useful comparison relevant to the AI analysis. In Kelly v. Arriba Soft Corp., 336 F.3d 811 (9th Cir. 2002), the 9th U.S. Circuit Court of Appeals ruled that a search engine’s creation of thumbnail images from original copies was a transformative fair use.[46] Arriba’s search-engine crawler made full-sized copies of Kelly’s images and stored them temporarily on Arriba’s server to generate thumbnail versions. After the thumbnails were created, the full-sized originals were deleted. The thumbnails were used to facilitate Arriba’s image-based search engine. In reaching its fair-use conclusion, the 9th Circuit opined that:

Arriba’s use of Kelly’s images promotes the goals of the Copyright Act and the fair use exception. The thumbnails do not stifle artistic creativity because they are not used for illustrative or artistic purposes and therefore do not supplant the need for the originals.[47]

Further, although “Arriba made exact replications of Kelly’s images, the thumbnails were much smaller, lower-resolution images that served an entirely different function than Kelly’s original images.”[48]

The court found it important that the search engine did not use the protected works for their intended “aesthetic experience,” but rather for the purpose of constructing a search index.[49] Indeed, the entire point of a search engine is not to “supersede” the original, but in many or most cases to provider users an efficient means to find that original online.[50]

The court discussed, but only briefly, the benefit to the public of Arriba’s transformative use,[51] noting that “[Arriba’s thumbnails] benefit the public by enhancing information-gathering techniques on the internet.”[52] Five years later, in Perfect 10 Inc. v. Amazon.com Inc., 487 F.3d 701 (2007), the 9th Circuit expanded on this question somewhat.[53] There, in holding that the novelty of the use was of crucial importance to the analysis,[54] the court also stressed that the value of that use was a function of its newness:

[A] search engine provides social benefit by incorporating an original work into a new work, namely, an electronic reference tool. Indeed, a search engine may be more transformative than a parody [the use at issue in Campbell] because a search engine provides an entirely new use for the original work, while a parody typically has the same entertainment purpose as the original work.[55]

Indeed, even in light of the commercial nature of Google’s use of copyrighted content in its search engine, its significant public benefit carried the day: “We conclude that the significantly transformative nature of Google’s search engine, particularly in light of its public benefit, outweighs Google’s superseding and commercial uses of the thumbnails in this case.”[56] And, of particular relevance to these questions in the context of AI, the court in Perfect 10 went on to “note the importance of analyzing fair use flexibly in light of new circumstances.”[57]

Ultimately, the Perfect 10 decision tracked Kelly fairly closely on the rest of the “transformativeness” analysis in finding fair use, because “[a]lthough an image may have been created originally to serve an entertainment, aesthetic, or informative function, a search engine transforms the image into a pointer directing a user to a source of information.”[58]

The core throughline in this line of cases is the question of whether a piece of content is being used for its expressive content, weighed against the backdrop of whether the use is for some new (and, thus, presumptively valuable) purpose. In Perfect 10 and Kelly, the transformative use was the creation of a search index.

“Snippets” fair-use cases track a similar line of reasoning. For example, in Authors Guild v. Google Inc., 804 F.3d 202 (2d Cir. 2015), the 2nd Circuit ruled that Google’s use of “snippets” of copyrighted books in its Library Project and Google Books website was a “transformative” fair use.[59] Holding that the “snippet view” of books digitized as part of the Google Books project did not constitute an effectively competing substitute to the original works, the circuit court noted that copying for the purpose of “criticism” or—as in that case—copying for the purpose of “provision of information about” the protected work, “tends most clearly to satisfy Campbell’s notion of the ‘transformative’ purpose.”[60]

Importantly, the court emphasized the importance of the public-benefit aspect of transformative uses: “[T]ransformative uses tend to favor a fair use finding because a transformative use is one that communicates something new and different from the original or expands its utility, thus serving copyright’s overall objective of contributing to public knowledge.”[61]

Underscoring the idea that the “transformativeness” analysis weighs whether a use is merely for expressive content against the novelty/utility of the intended use, the court observed:

Google’s division of the page into tiny snippets is designed to show the searcher just enough context surrounding the searched term to help her evaluate whether the book falls within the scope of her interest (without revealing so much as to threaten the author’s copyright interests). Snippet view thus adds importantly to the highly transformative purpose of identifying books of interest to the searcher.[62]

Thus, the absence of use of the work’s expressive content, coupled with a fairly circumscribed (but highly novel) use was critical to the outcome.

The entwined questions of transformative use and the public benefit it confers are significantly more complicated in the AI context, however. Unlike the incidental copying involved in search-engine indexing or thumbnails, training generative AI systems directly leverages copyrighted works for their expressive value. In the Google Books and Kelly cases, the defendant systems extracted limited portions of works or down-sampled images solely to identify and catalog their location for search purposes. The copies enabled indexing and access, and they expanded public knowledge through a means unrelated to the works’ protected aesthetics.

But in training AI models on copyrighted data, the systems necessarily parse the intrinsic creative expression of those works. The AI engages with the protected aesthetic elements themselves, not just superficial markers (like title, length, location on the internet, etc.), in order to internalize stylistic and compositional principles. This appropriates the heart of the works’ copyright protection for expressive ends, unlike the more tenuous connections in search systems.

The AI is thus “learning” directly from the protected expression in a manner akin to a human student studying an art textbook, or like the scientists learning from the journals in American Geophysical Union. The subsequent AI generations are built from mastery of the copyrighted training materials’ creative expression. Thus, while search-engine copies only incidentally interact with protected expression to enable unrelated innovation, AI training is predicated on excavating the protected expression itself to fuel iterative creation. These meaningfully different purposes have significant fair-use implications.

This functional difference is, as noted, central to the analysis of a use’s “purpose and character.” Indeed, “even making an exact copy of a work may be transformative so long as the copy serves a different function than the original work.”[63] But the benefit to the public from the new use is important, as well, particularly with respect to the possible legislative response that a restrictive interpretation of existing doctrine may engender.

If existing fair-use principles prohibit the copying required for AI, absent costly item-by-item negotiation and licensing, the transaction costs could become prohibitive, thwarting the development of technologies that promise great public value.[64] Copyright law has faced similar dilemmas before, where the transaction costs of obtaining permission for socially beneficial uses could frustrate those uses entirely.[65] In such cases, we have developed mechanisms like compulsory licensing to facilitate the necessary copying, while still attempting to compensate rightsholders. An unduly narrow fair-use finding for AI training could spur calls for similar interventions in service of enabling AI progress.

In other words, regardless of the veracity of the above conclusion that AI’s use of copyrighted works may not, in fact, serve a different function than the original, courts and legislators may be reluctant to allow copyright doctrine to serve as an absolute bar against self-evidently valuable activity like AI development. Our aim should be to interpret or recalibrate copyright law to permit such progress while upholding critical incentives for creators.

C.            Opt-In vs. Opt-Out Use of Protected Works

The question at the heart of the prior discussion—and, indeed, at the heart of the economic analysis of copyright—is whether the transaction costs that accompany requiring express ex ante permission for the use of protected works are so high that they impedes socially beneficial conduct whose value would outweigh the social cost of allowing permissionless and/or uncompensated use.[66] The NOI alludes to this question when it asks: “Should copyright owners have to affirmatively consent (opt in) to the use of their works for training materials, or should they be provided with the means to object (opt out)?”[67]

This is a complex problem. Given the foregoing thoughts on fair use, it seems quite possible that, at present, the law requires creators of AI systems to seek licenses for protected content, or else must resort to public-domain works for training. Given the volume of copyrighted works that AI developers currently use to train these systems, such requirements may be broadly infeasible.

On one hand, requiring affirmative opt-in consent from copyright holders imposes significant transaction costs on AI-system developers to identify and negotiate licenses for the vast amounts of training data required. This could hamper innovation in socially beneficial AI systems. On the other hand, an opt-out approach shifts more of the transaction-cost burden to copyright holders, who must monitor and object to unwanted uses of their works. This raises concerns about uncompensated use.

Ultimately, the question is where the burden should lie: with AI-system developers to obtain express consent, or with copyright holders to monitor and object to uses? Requiring some form of consent may be necessary to respect copyright interests. Yet an opt-out approach may strike the right balance, by shifting some of the burden back to AI developers while avoiding the infeasibly high transaction costs of mandatory opt-in consent. The optimal approach likely involves nuanced policymaking to balance these competing considerations. Moreover, as we discuss infra, the realistic outcome is most likely going to require rethinking the allocation of property rights in ways that provide for large-scale licensing. Ideally, this could be done through collective negotiation, but perhaps at a de minimis rate, while allowing creators to bargain for remuneration on the basis of other rights, like a right of publicity or other rights attached to the output of AI systems, rather than the inputs.[68]

1.              Creator consent

Relatedly, the Copyright Office asks: “If copyright owners’ consent is required to train generative AI models, how can or should licenses be obtained?”[69]

Licensing markets exist, and it is entirely possible that major AI developers and large groups of rightsholders can come to mutually beneficial terms that permit a sufficiently large body of protected works to be made available as training data. Something like a licensing agency for creators who choose to make their works available could arise, similar to the services that exist to provide licensed music and footage for video creators.[70] It is also possible for some to form collective-licensing organizations to negotiate blanket permissions covering many works.

It’s important to remember that our current thinking is constrained by our past experience. All we know today are AI models trained on vast amounts of unlicensed works. It is entirely possible that, if firms were required to seek licenses, unexpected business models would emerge to satisfy both sides of the equation.

For example, an AI firm could develop its own version of YouTube’s ContentID, which would allow creators to control when their work is used in AI training. For some well-known artists, this could be negotiated with an upfront licensing fee. On the user side, any artist who has opted in could then be selected as a “style” for the AI to emulate—triggering a royalty payment to the artist when a user generates an image or song in that style. Creators could also have the option of removing their influence from the system if they so desire.

Undoubtedly, there are other ways to structure the relationship between creators and AI systems  that would facilitate creators’ monetization of the use of their work in AI systems, including legal and commercial structures that create opportunities for both creators and AI firms to succeed.

III.          Generative AI Outputs: Protection of Outputs and Outputs that Infringe

The Copyright Office asks: “Under copyright law, are there circumstances when a human using a generative AI system should be considered the ‘author’ of material produced by the system?”[71]

Generally speaking, we see no reason why copyright law should be altered to afford protection to purely automatic creations generated by AI systems. That said, when a human makes a nontrivial contribution to generative AI output—such as editing, reframing, or embedding the AI-generated component within a larger work—the resulting work should qualify for copyright protection.

Copyright law centers on the concept of original human authorship.[72] The U.S. Constitution expressly limits copyright to “authors.”[73] As of this writing, however, generative AI’s capacities do not rise to the level of true independent authorship. AI systems remain tools that require human direction and judgment.[74] As such, when a person provides the initial prompt or framing, makes choices regarding the iterative development of the AI output, and decides that the result is satisfactory for inclusion in a final work, they are fundamentally engaging in creative decision making that constitutes authorship under copyright law.

As Joshua Gans has observed of recent Copyright Review Board decisions:

Trying to draw some line between AI and humans with the current technology opens up a massive can of worms. There is literally no piece of digital work these days that does not have some AI element to it, and some of these mix and blur the lines in terms of what is creative and what is not. Here are some examples:

A music artist uses AI to denoise a track or to add an instrument or beat to a track or to just get a composition started.

A photographer uses Photoshop or takes pictures with an iPhone that already uses AI to focus the image and to sort a burst of images into one that is appropriate.

A writer uses AI to prompt for some dialogue when stuck at some point or to suggest a frame for writing a story.[75]

Attempting to separate out an “AI portion” from the final work, as the Copyright Review Board proposed, fundamentally misunderstands the integrated nature of the human-AI collaborative process. The AI system cannot function without human input, and its output remains raw material requiring human creativity to incorporate meaningfully into a finished product.

Therefore, when a generative AI system is used as part of a process guided by human creative choices, the final work should be protected by copyright, just as a work created using any other artistic tool or collaborator would be. Attenuating copyrightability due to the use of AI would undermine basic copyright principles and fail to recognize the essentially human nature of the creative process.

A.            AI Outputs and Infringement

The NOI asks: “Is the substantial similarity test adequate to address claims of infringement based on outputs from a generative AI system, or is some other standard appropriate or necessary?” (Question 23)

The outputs of AI systems may or may not violate IP laws, but there is nothing inherent in the processes described above that dictates that they must. As noted, the most common AI systems do not save copies of existing works, but merely “instructions” (more or less) on how to create new work that conforms to patterns found by examining existing work. If we assume that a system isn’t violating copyright at the input stage, it’s entirely possible that it can produce completely new pieces of art that have never before existed and do not violate copyright.

They can, however, be made to violate copyrights. For example, these systems can be instructed to generate art, not just in the style of a particular artist, but art that very closely resembles existing pieces. In this sense, it would be making a copy that theoretically infringes. The fact of an AI’s involvement would not change the analysis: just as with a human-created work, if it is substantially similar to a copyrighted work, it may be found infringing.

There is, however, a common bug in AI systems that leads to outputs that are more likely to violate copyright in this way. Known as “overfitting,” the training leg of these AI systems can be presented with samples that contain too many instances of a particular image.[76] This leads to a dataset that contains too much information about the specific image, such that—when the AI generates a new image—it is constrained to producing something very close to the original. Similarly, there is evidence that some AI systems are “memorizing” parts of protected books.[77] This could lead to AI systems repeating copyright-protected written works.

1.              The substantial-similarity test

The substantial-similarity test remains functionally the same when evaluating works generated using AI. To find “substantial similarity,” courts require evidence of copying, as well as an expression that is substantially similar to a protected work.[78] “It is now an axiom of copyright law that actionable copying can be inferred from the defendant’s access to the copyrighted work and substantial similarity between the copyrighted work and the alleged infringement.”[79] In many or most cases, it will arguably be the case that AI systems have access to quite a wide array of protected works that are posted online. Thus, there may not be a particularly high hurdle to determine that an AI system actually copied a protected work.

There is, however, one potential problem for the first prong of this analysis. Models produced during a system’s training process do not (usually) contain the original work, but are the “ideas” that the AI systems generated during training. Thus, where the provenance of works contained in a training corpus is difficult to source, it may not be so straightforward to make inferences about whether a model “saw” a particular work. This is because the “ideas” that the AI “learns” from its training corpus are unprotected under U.S. copyright law, as it is permissible to mimic unprotected elements of a copyrighted work (such as ideas).[80]

Imagine a generative AI system trained on horror fiction. It would be possible for this system to produce a new short story that is similar to one written by Stephen King, but the latent data in the model almost certainly would not violate any copyrights that King holds in his work. The model would contain “ideas” about horror stories, including those learned from an array of authors who were themselves influences on Stephen King, and potentially some of King’s own stories. What the AI system “learns” in this case is the relationship between words and other linguistic particularities that are commonly contained in horror fiction. That is, it has “ideas” about what goes into a horror story, not (theoretically) the text of the horror story itself.

Thus, when demonstrating indirect proof of copying in the case of a Stephen King story, it may pose a difficulty that an AI system has ingested all of H.P. Lovecraft’s work—an author who had a major influence on King. The “ideas” in the model and the output it subsequently produces may, in fact, produce something similar to a Stephen King work, but it may have been constructed largely or entirely on material from Lovecraft and other public-domain horror writers. The problem becomes only more complicated when you realize that this system could also have been trained on public-domain fan fiction written in the style of Stephen King. Thus, for the purposes of the first prong of this analysis, courts may place greater burden on plaintiffs in copyright actions against model producers to demonstrate more than merely that a work was merely available online.

Assuming that plaintiffs are able to satisfy the first prong, once an AI system “expresses” those ideas, that expression could violate copyright law under the second prong of the substantial-similarity test. The second prong inquires whether the final work appropriated the protected original expression.[81] Any similarities in unprotectable ideas, facts, or common tropes are disregarded.[82] So, in both traditional and AI contexts, the substantial-similarity test ultimately focuses on the protected components of creative expression, not surface similarity.

The key determination is whether the original work’s protected expression itself has been impermissibly copied, no matter the process that generated the copy. AI is properly viewed as simply another potential tool that could be used in certain acts of copying. It does not require revisiting settled principles of copyright law.

B.            Direct and Secondary Liability

The NOI asks: “If AI-generated material is found to infringe a copyrighted work, who should be directly or secondarily liable—the developer of a generative AI model, the developer of the system incorporating that model, end users of the system, or other parties?”[83]

Applying traditional copyright-infringement frameworks to AI-generated works poses unique challenges in determining direct versus secondary liability. In some cases, the AI system itself may create infringing content without any direct human causation.

1.              Direct liability

If the end user prompts an AI system in a way that intentionally targets copyrighted source material, they may meet the threshold for direct infringement by causing the AI to reproduce protected expression.[84] Though many AI prompts contain only unprotected ideas, users may sometimes input copyrightable material as the basis for the AI output. For example, a user could upload a copyrighted image and request the AI to make a new drawing based on the sample. In such cases, the user is intentionally targeting copyrighted works and directly “causing” the AI system to reproduce output that is similar. If sufficiently similar, that output could infringe on the protected input. This would be a question of first impression, but it is a plausible reading of available cases.

For example, in CoStar Grp. Inc. v. LoopNet Inc., 373 F.3d 544 (4th Cir. 2004), the 4th U.S. Circuit Court of Appeals had to consider whether an internet service provider (ISP) could be directly liable when third parties reposted copyrighted material owned by the plaintiff. In determining that merely owning the “machine” through which copies were made or transmitted was not enough to “cause” a direct infringement, the court held that:

[T]o establish direct liability under §§ 501 and 106 of the Act, something more must be shown than mere ownership of a machine used by others to make illegal copies. There must be actual infringing conduct with a nexus sufficiently close and causal to the illegal copying that one could conclude that the machine owner himself trespassed on the exclusive domain of the copyright owner. The Netcom court described this nexus as requiring some aspect of volition or causation… Indeed, counsel for both parties agreed at oral argument that a copy machine owner who makes the machine available to the public to use for copying is not, without more, strictly liable under § 106 for illegal copying by a customer. The ISP in this case is an analogue to the owner of a traditional copying machine whose customers pay a fixed amount per copy and operate the machine themselves to make copies. When a customer duplicates an infringing work, the owner of the copy machine is not considered a direct infringer. Similarly, an ISP who owns an electronic facility that responds automatically to users’ input is not a direct infringer.[85]

Implied in the 4th Circuit’s analogy is that, while the owner of a copying machine might not be a direct infringer, a user employing such a machine could be a direct infringer. It’s an imperfect analogy, but a user of an AI system prompting it to create a “substantially similar” reproduction of a protected work could very well be a direct infringer under this framing. Nevertheless, the analogy is inexact, because the user feeds an original into a copying machine in order to make a more-or-less perfect copy of the original, whereas an AI system generates something new but similar. The basic mechanism of using a machine to try to reproduce a protected work, however, remains essentially the same. Whether there is an infringement would be a question of “substantial similarity.”

2.              Secondary liability

As in the case of direct liability, the nature of generative AI makes the secondary-liability determination slightly more complicated, as well. That is, paradoxically, the basis for secondary liability could theoretically arise even where there was no direct infringement.[86]

The first piece of this analysis is relatively easier. If a user is directly liable for infringing a protected work, as noted above, the developer and provider of a generative AI system may face secondary copyright liability. If the AI developer or distributor knows the system can produce infringing outputs, and provides tools or material support that allows users to infringe, it may be liable for contributory infringement.[87] Critically, merely designing a system that is capable of infringing is not enough to find contributory liability.[88]

An AI producer or distributor may also have vicarious liability, insofar as it has the right and ability to supervise users’ activity and a direct financial interest in that activity.[89] AI producers have already demonstrated their ability to control users’ behavior to thwart unwanted uses of the service.[90] Thus, if there is a direct infringement by a user, a plausible claim for vicarious liability could be made so long as there is sufficient connection between the user’s behavior and the producer’s financial interests.

The question becomes more complicated when a user did not direct the AI system to infringe. When the AI generates infringing content without user direction, it’s not immediately clear who would be liable for the infringement.[91] Consider the case where, unprompted by either the user or the AI producer, an AI system creates an output that would infringe under the substantial-similarity test. Assuming that the model has not been directed by the producer to “memorize” the works it ingests, the model itself consists of statistical information about the relationship between different kinds of data. The infringer, in a literal sense, is the AI system itself, as it is the creator of the offending output. Technically, this may be a case of vicarious liability, even without an independent human agent causing the direct infringement.

We know that copyright protection can only be granted to humans. As the Copyright Review Board recently found in a case deciding whether AI-generated outputs can be copyrighted:

The Copyright Act protects, and the Office registers, “original works of authorship fixed in any tangible medium of expression.” 17 U.S.C. § 102(a). Courts have interpreted the statutory phrase “works of authorship” to require human creation of the work.[92]

But can an AI system directly violate copyright? In his Aereo dissent, Justice Clarence Thomas asserted that it was a longstanding feature of copyright law that violation of the performance right required volitional behavior.[93] But the majority disagreed with him, holding that, by running a fully automated system of antennas intended to allow users to view video at home, the system gave rise to direct copyright liability.[94] Thus, implied in the majority’s opinion is the idea that direct copyright infringement does not require “volitional” conduct.

It is therefore plausible that a non-sentient, fully automated AI system could infringe copyright, even if, ultimately, there is no way to recover against the nonhuman agent. That does, however, provide an opportunity for claims of vicarious liability against the AI producer or distributor— at least, where the producer has the power to control the AI system’s behavior and that behavior appears to align with the producer’s financial interests.

3.              Protecting the ‘style’ of human creators

The NOI asks: “Are there or should there be protections against an AI system generating outputs that imitate the artistic style of a human creator (such as an AI system producing visual works ‘in the style of’ a specific artist)?”[95]

At the federal level, one candidate for protection against AI imitating some aspects of a creator’s works can currently be found in trademark law. Trademark law, governed by the Lanham Act, protects names, symbols, and other source identifiers that distinguish goods and services in commerce.[96] Unfortunately, a photograph or likeness, on its own, typically does not qualify for trademark protection, unless it is consistently used on specific goods.[97] Even where there is a likeness (or similar “mark”) used consistently as part of branding a distinct product, many trademark-infringement claims would be difficult to establish in this context, because trademark law does little to protect many aspects of a creator’s work.

Moreover, the Supreme Court has been wary about creating a sort of “mutant copyright” in cases that invoke the Lanham Act as a means to enforce a sort of “right of attribution,” which would potentially give creators the ability to control the use of their name in broader contexts.[98] In this context, the Court has held that the relevant parts of the Lanham Act were not designed to “protect originality or creativity,”[99] but are focused solely on “actions like trademark infringement that deceive consumers and impair a producer’s goodwill.”[100]

In many ways, there is a parallel here to the trademark cases involving keyword bidding in online ads. At a high level, search engines and other digital-advertising services do not generally infringe trademark when they allow businesses to purchase ads triggered by a user’s search for competitor trademarks (i.e., rivals’ business names).[101] But in some contexts, this can be infringing—e.g., where the use of trademarked terms in combination with advertising text can mislead consumers about the origin of a good or service.[102]

Thus, the harm, when it arises, would not be in a user asking an AI system to generate something “in the style of” a known creator, but when that user subsequently seeks to release a new AI-generated work and falsely claims it originated from the creator, or leaves the matter ambiguous and misleading to consumers.

Alternative remedies for creators could be found in the “right of publicity” laws in various states. A state-level right of publicity “is not merely a legal right of the ‘celebrity,’ but is a right inherent to everyone to control the commercial use of identity and persona and recover in court damages and the commercial value of an unpermitted taking.”[103] Such rights are recognized under state common law and statutes, which vary considerably in scope across jurisdictions—frequently as part of other privacy statutes.[104] For example, some states only protect an individual’s name, likeness, or voice, while others also cover distinctive appearances, gestures, and mannerisms.[105] The protections afforded for right-of-publicity claims vary significantly based on the state where the unauthorized use occurs or the individual is domiciled.[106] This creates challenges for the application of uniform nationwide protection of creators’ interests in the various aspects that such laws protect.

In recent hearings before the U.S. Senate Judiciary Subcommittee on Intellectual Property, several witnesses advocated creating a federal version of the right of publicity.[107] The Copyright Office has also previously opined that it may be desirable for Congress to enact some form of a “right of publicity” law.[108] If Congress chose to enact a federal “right of privacy” statute, several key issues would need to be addressed regarding the scope of protection, effect on state laws, constitutional authority, and First Amendment limitations.

Congress would have to delineate the contours of the federal right of publicity, including the aspects of identity covered and the types of uses prohibited. A broad right of privacy could protect names, images, likenesses, voices, gestures, distinctive appearances, and biographical information from any unauthorized commercial use. Or Congress could take a narrower approach focused only on particular identity attributes, like name and likeness. Congress would also need to determine whether a federal right-of-publicity statute preempts state right-of-publicity laws or sets a floor that would allow state protections to exceed the federal standards.

4.              Bargaining for the use of likenesses

A federal right of publicity could present an interesting way out of the current dispute between rightsholders and AI producers. Most of the foregoing comment attempts to pull apart different pieces of potential infringement actions, but such actions are only necessary, obviously, if a mutually beneficial agreement cannot be struck between creators and AI producers. The main issue at hand is that, given the vast amount of content necessary to train an AI system, it could be financially impractical for even the largest AI firms to license all the necessary content. Even if the comments above are correct, and fair use is not available, it could very well be the case that AI producers will not license very much content, possibly relying on public-domain material, and choosing to license only a very small selection.

Something like a “right of publicity,” or an equivalent agreement between creators and AI producers, could provide alternative licensing and monetization strategies that encourage cooperation between the parties. If creators had the opportunity to opt into the use of their likeness (or the relevant equivalent for the sort of AI system in question), the creators could generate revenue when the AI system actually uses the results of processing their content. Thus, the producers would not need to license content that contributes an unknown and possibly de minimis value to their systems, and would only need to pay for individual instances of use.

Indeed, in this respect, we are already beginning to see some experimentation with business models. The licensing of celebrity likenesses for Meta’s new AI chatbots highlights an emerging opportunity for creators to monetize their brand through contractual agreements that grant usage rights to tech companies that commercialize conversational AI.[109] As this technology matures, there will be more opportunities for collaborations between AI producers—who are eager to leverage reputable and recognizable personalities—and celebrities or influencers seeking new income streams.

As noted, much of the opportunity for creators and AI producers to reach these agreements will depend on how rights are assigned.[110] It may be the case that a “right of publicity” is not necessary to make this sort of bargaining happen, as creators could—at least theoretically—pursue litigation on a state-by-state basis. This disparate-litigation strategy could deter many creators, however, and it could also be the case that a single federal standard outlining a minimal property right in “publicity” could help to facilitate bargaining.

Conclusion

The advent of generative AI systems presents complex new public-policy challenges centered on the intersection of technology and copyright law. As the Copyright Office’s inquiry recognizes, there are open questions around the legal status of AI-training data, the attribution of AI outputs, and infringement liability, which all require thoughtful analysis.

Ultimately, maintaining incentives for human creativity, while also allowing AI systems to flourish, will require compromise and cooperation between stakeholders. Rather than an outright ban on the unauthorized use of copyrighted works for training data, a licensing market that enables access to a large corpora could emerge. Rightsholders may need to accept changes to how they typically license content. In exchange, AI producers will have to consider how they can share the benefit of their use of protected works with creators.

Copyright law retains flexibility to adapt to new technologies, as past reforms reacting to photography, sound recordings, software, and the internet all demonstrate. With careful balancing of interests, appropriate limitations, and respect for constitutional bounds, copyright can continue to promote the progress of science and the useful arts even in the age of artificial intelligence. This inquiry marks a constructive starting point, although ongoing reassessment will likely be needed as generative AI capabilities continue to advance rapidly.

[1] Artificial Intelligence and Copyright, Notice of Inquiry and Request for Comments, U.S. Copyright Office, Library of Congress (Aug. 30, 2023) [hereinafter “NOI”].

[2] Tim Sweeney (@TimSweeneyEpic), Twitter (Jan. 15, 2023, 3:35 AM), https://twitter.com/timsweeneyepic/status/1614541807064608768?s=46&t=0MH_nl5w4PJJl46J2ZT0Dw.

[3] Pulitzer Prize Winner and Other Authors Accuse OpenAI of Misusing Their Writing, Competition Policy International (Sep. 11, 2023), https://www.pymnts.com/cpi_posts/pulitzer-prize-winner-and-other-authors-accuse-openai-of-misusing-their-writing; Getty Images Statement, Getty Images (Jan. 17, 2023), https://newsroom.gettyimages.com/en/getty-images/getty-images-statement.

[4] See, e.g., Anton Oleinik, What Are Neural Networks Not Good At? On Artificial Creativity, 6 Big Data & Society (2019), available at https://journals.sagepub.com/doi/full/10.1177/2053951719839433#bibr75-2053951719839433.

[5] William M. Landes & Richard A. Posner, An Economic Analysis of Copyright Law, 18 J. Legal Stud. 325 (1989).

[6] Id. at 332.

[7] Id. at 326.

[8] Id.

[9] See infra, notes 102-103 and accompanying text.

[10] See generally R.H. Coase, The Problem of Social Cost, 3 J. L. & Econ. 1, 2 (1960).

[11] Richard Posner, Economic Analysis of Law (Aspen 5th ed 1998) 65, 79.

[12] Coase, supra note 9, at 27.

[13] Id.

[14] Id. at 27.

[15] Id. at 42-43.

[16] U.S. Copyright Office, Library of Congress, supra note 1, at 14.

[17] For more detailed discussion of GANs and Stable Diffusion see Ian Spektor, From DALL E to Stable Diffusion: How Do Text-to-image Generation Models Work?, Tryo Labs Blog (Aug. 31, 2022), https://tryolabs.com/blog/2022/08/31/from-dalle-to-stable-diffusion.

[18] Id.

[19] Id.

[20] Id.

[21] Id.

[22] Id.

[23] Jay Alammar, The Illustrated Stable Diffusion, Blog (Oct. 4, 2022), https://jalammar.github.io/illustrated-stable-diffusion.

[24] Indeed, there is evidence that some models may be trained in a way that they “memorize” their training set, to at least some extent. See, e.g., Kent K. Chang, Mackenzie Cramer, Sandeep Soni, & David Bamman, Speak, Memory: An Archaeology of Books Known to ChatGPT/GPT-4, arXiv Preprint (Oct. 20, 2023), https://arxiv.org/abs/2305.00118; OpenAI LP, Comment Regarding Request for Comments on Intellectual Property Protection for Artificial Intelligence Innovation, Before the USPTO, Dep’t of Com. (2019), available at https://www.uspto.gov/sites/default/files/documents/OpenAI_RFC-84-FR-58141.pdf.

[25] OpenAI, LP, Comment Regarding Request for Comments on Intellectual Property Protection for Artificial Intelligence, id. (emphasis added).

[26] 17 U.S.C. § 107.

[27] See, e.g., Blake Brittain, Meta Tells Court AI Software Does Not Violate Author Copyrights, Reuters (Sep. 19, 2023), https://www.reuters.com/legal/litigation/meta-tells-court-ai-software-does-not-violate-author-copyrights-2023-09-19; Avram Piltch, Google Wants AI Scraping to be ‘Fair Use.’ Will That Fly in Court?, Tom’s Hardware (Aug. 11, 2023), https://www.tomshardware.com/news/google-ai-scraping-as-fair-use.

[28] 17 U.S.C. § 106.

[29] Register of Copyrights, DMCA Section 104 Report (U.S. Copyright Office, Aug. 2001), at 108-22, available at https://www.copyright.gov/reports/studies/dmca/sec-104-report-vol-1.pdf.

[30] Id. at 122-23.

[31] Id. at 112 (emphasis added).

[32] Id. at 129–30.

[33] 17 U.S.C. § 107.

[34] Id.; see also Campbell v. Acuff-Rose Music Inc., 510 U.S. 569 (1994).

[35] Critically, a fair use analysis is a multi-factor test, and even within the first factor, it’s not a mandatory requirement that a use be “transformative.” It is entirely possible that a court balancing all of the factors could indeed find that training AI systems is fair use, even if it does not hold that such uses are “transformative.”

[36] Campbell, supra note 22, at 591.

[37] Authors Guild v. Google, Inc., 804 F.3d 202, 214 (2d Cir. 2015).

[38] OpenAI submission, supra note 13, at 5.

[39] Id. at 915.

[40] Id.

[41] Id.

[42] Id. at 933-34.

[43] Id. at 923. (emphasis added)

[44] Id.

[45] Id. at 924.

[46] Kelly v. Arriba Soft Corp., 336 F.3d 811 (9th Cir. 2002).

[47] Id.

[48] Id. at 818.

[49] Id.

[50] Id. at 819 (“Arriba’s use of the images serves a different function than Kelly’s use—improving access to information on the internet versus artistic expression.”).

[51] The “public benefit” aspect of copyright law is reflected in the fair-use provision, 17 U.S.C. § 107. In Campbell v. Acuff-Rose Music, Inc., 510 U.S. 569, 579 (1994), the Supreme Court highlighted the “social benefit” that a use may provide, depending on the first of the statute’s four fair-use factors, the “the purpose and character of the use.”

[52] Supra note 46, at 820.

[53] Perfect 10 Inc. v. Amazon.com Inc., 487 F.3d 701 (9th Cir., 2007)

[54] Id. at 721 (“Although an image may have been created originally to serve an entertainment, aesthetic, or informative function, a search engine transforms the image into a pointer directing a user to a source of information.”).

[55] Id. at 721.

[56] Id. at 723 (emphasis added).

[57] Id. (emphasis added).

[58] Id.

[59] Supra note 37, at 218.

[60] Id. at 215-16.

[61] Id. at 214. See also id. (“The more the appropriator is using the copied material for new, transformative purposes, the more it serves copyright’s goal of enriching public knowledge and the less likely it is that the appropriation will serve as a substitute for the original or its plausible derivatives, shrinking the protected market opportunities of the copyrighted work.”).

[62] Id. at 218.

[63] Perfect 10, 487 F.3d at 721-22 (citing Kelly, 336 F.3d at 818-19). See also Campbell, 510 U.S. at 579 (“The central purpose of this investigation is to see, in Justice Story’s words, whether the new work merely ‘supersede[s] the objects’ of the original creation, or instead adds something new, with a further purpose or different character….”) (citations omitted).

[64] See supra, notes 9-14 and accompanying text.

[65] See, e.g., the development of the compulsory “mechanical royalty,” now embodied in 17 U.S.C. § 115, that was adopted in the early 20th century as a way to make it possible for the manufacturers of player pianos to distribute sheet music playable by their instruments.

[66] See supra notes 9-14 and accompanying text.

[67] U.S. Copyright Office, Library of Congress, supra note 1, at 15.

[68] See infra, notes at 102-103 and accompanying text.

[69] U.S. Copyright Office, Library of Congress, supra note 1, at 15.

[70] See, e.g., Copyright Free Music, Premium Beat By Shutterstock, https://www.premiumbeat.com/royalty-free/licensed-music; Royalty-free stock footage at your fingertips, Adobe Stock, https://stock.adobe.com/video.

[71] U.S. Copyright Office, Library of Congress, supra note 1, at 19.

[72] Id.

[73] U.S. Const. art. I, § 8, cl. 8.

[74] See Ajay Agrawal, Joshua S. Gans, & Avi Goldfarb, Exploring the Impact of Artificial Intelligence: Prediction Versus Judgment, 47 Info. Econ. & Pol’y 1, 1 (2019) (“We term this process of understanding payoffs, ‘judgment’. At the moment, it is uniquely human as no machine can form those payoffs.”).

[75] Joshua Gans, Can AI works get copyright protection? (Redux), Joshua Gans’ Newsletter (Sept. 7, 2023), https://joshuagans.substack.com/p/can-ai-works-get-copyright-protection.

[76] See Nicholas Carlini, et al., Extracting Training Data from Diffusion Models, Cornell Univ. (Jan. 30, 2023), available at https://arxiv.org/abs/2301.13188.

[77] See Chang, Cramer, Soni, & Bamman, supra note 24; see also Matthew Sag, Copyright Safety for Generative AI, Working Paper (May 4, 2023), available at https://ssrn.com/abstract=4438593.; Andrés Guadamuz, A Scanner Darkly: Copyright Liability and Exceptions in Artificial Intelligence Inputs and Outputs, 25-27 (Mar. 1, 2023), available at https://papers.ssrn.com/sol3/papers.cfm?abstract_id=4371204.

[78] Laureyssens v. Idea Grp. Inc., 964 F.2d 131, 140 (2d Cir. 1992), as amended (June 24, 1992).

[79] Id. at 139.

[80] Harney v. Sony Pictures Television Inc., 704 F.3d 173, 178 (1st Cir. 2013). This assumes, for argument’s sake, that a given model is not “memorizing,” as noted above.

[81] Id. at 178-79.

[82] Id.

[83] U.S. Copyright Office, Library of Congress, supra note 1, at 25.

[84] Notably, the state of mind of the user would be irrelevant from the point of view of whether an infringement occurs. All that is required is that a plaintiff owns a valid copyright, and that the defendant infringed it. 17 U.S.C. 106. There are cases where the state of mind of the defendant will matter, however. For one, willful or recklessly indifferent infringement by a plaintiff will open the door for higher statutory damages. See, e.g., Island Software & Computer Serv., Inc. v. Microsoft Corp., 413 F.3d 257, 263 (2d Cir. 2005). For another, a case of criminal copyright infringement will require that a defendant have acted “willfully.” 17 U.S.C. § 506(a)(1) (2023), 18 U.S.C. § 2319 (2023).

[85] Id. at 550.

[86] Legally speaking, it would be incoherent to suggest that there can be secondary liability without primary liability. The way that AI systems work, however, could prompt Congress to modify the law in order to account for the identified situation.

[87] See, e.g., Metro-Goldwyn-Mayer Studios Inc. v. Grokster Ltd., 380 F.3d 1154, 1160 (9th Cir. 2004), vacated and remanded, 545 U.S. 913, 125 S. Ct. 2764, 162 L. Ed. 2d 781 (2005).

[88] See BMG Rts. Mgmt. (US) LLC v. Cox Commc’ns Inc., 881 F.3d 293, 306 (4th Cir. 2018); Sony Corp. of Am. v. Universal City Studios Inc., 464 U.S. 417, 442 (1984).

[89] A&M Recs. Inc. v. Napster Inc., 239 F.3d 1004, 1022 (9th Cir. 2001), as amended (Apr. 3, 2001), aff’d sub nom. A&M Recs. Inc. v. Napster Inc., 284 F.3d 1091 (9th Cir. 2002), and aff’d sub nom. A&M Recs. Inc. v. Napster Inc., 284 F.3d 1091 (9th Cir. 2002).

[90] See, e.g., Content Filtering, Microsoft Ignite, available at https://learn.microsoft.com/en-us/azure/ai-services/openai/concepts/content-filter (last visited Oct. 27, 2023).

[91] Note that, if an AI producer can demonstrate that they used no protected works in the training phase, there may in fact be no liability for infringement at all. If a protected work is never made available to the AI system, even an output very similar to that protected work might not be “substantially similar” in a legal sense.

[92] Copyright Review Board, Second Request for Reconsideration for Refusal to Register Théâtre D’opéra Spatial (SR # 1-11743923581; Correspondence ID: 1-5T5320R), U.S. Copyright Office (Sep. 5, 2023), available at https://fingfx.thomsonreuters.com/gfx/legaldocs/byprrqkqxpe/AI%20COPYRIGHT%20REGISTRATION%20decision.pdf.

[93] Am. Broad. Companies Inc. v. Aereo Inc., 573 U.S. 431, 453 (2014). (Thomas J, dissenting).

[94] Id. at 451.

[95] U.S. Copyright Office, Library of Congress, supra note 1, at 21.

[96] See 5 U.S.C. § 1051 et seq. at § 1127.

[97] See, e.g., ETW Corp. v. Jireh Pub. Inc., 332 F.3d 915, 923 (6th Cir. 2003).

[98] Dastar Corp. v. Twentieth Century Fox Film Corp., 539 U.S. 23, 34 (2003).

[99] Id. at 37.

[100] Id. at 32.

[101] See, e.g., Multi Time Mach. Inc. v. Amazon.com Inc., 804 F.3d 930, 938 (9th Cir. 2015); EarthCam Inc. v. OxBlue Corp., 49 F. Supp. 3d 1210, 1241 (N.D. Ga. 2014); Coll. Network Inc. v. Moore Educ. Publishers Inc., 378 F. App’x 403, 414 (5th Cir. 2010).

[102] Digby Adler Grp. LLC v. Image Rent a Car Inc., 79 F. Supp. 3d 1095, 1102 (N.D. Cal. 2015).

[103] J. Thomas McCarthy, The Rights of Publicity and Privacy § 1:3. Introduction—Definition and History of the Right of Publicity—Simple Definition of the Right of Publicity, 1 Rights of Publicity and Privacy § 1:3 (2d ed).

[104] See id. at § 6:3.

[105] Compare Ind. Code § 32-36-1-7 (covering name, voice, signature, photograph, image, likeness, distinctive appearance, gesture, or mannerism), with Ky. Rev. Stat. Ann. § 391.170 (limited to name and likeness for “public figures”).

[106] See Restatement (Third) of Unfair Competition § 46 (1995).

[107] See, e.g., Jeff Harleston, Artificial Intelligence and Intellectual Property – Part II: Copyright, U.S. Senate Comm. on the Judiciary Subcomm. on Intellectual Property (Jul.12, 2023), available at https://www.judiciary.senate.gov/imo/media/doc/2023-07-12_pm_-_testimony_-_harleston1.pdf; Karla Ortiz, “AI and Copyright”, U.S. Senate Comm. on the Judiciary Subcomm. on Intellectual Property (Jul. 7, 2023), available at https://www.judiciary.senate.gov/imo/media/doc/2023-07-12_pm_-_testimony_-_ortiz.pdf; Matthew Sag, “Artificial Intelligence and Intellectual Property – Part II: Copyright and Artificial Intelligence”, U.S. Senate Comm. on the Judiciary Subcomm. on Intellectual Property (Jul. 12, 2023), available at https://www.judiciary.senate.gov/imo/media/doc/2023-07-12_pm_-_testimony_-_sag.pdf.

[108] Authors, Attribution, and Integrity: Examining Moral Rights in the United States, U.S. Copyright Office (Apr. 2019) at 117-119, https://www.copyright.gov/policy/moralrights/full-report.pdf.

[109] Benj Edwards, Meta Launches Consumer AI Chatbots with Celebrity Avatars in its Social Apps, ArsTechnica (Sep. 28, 2023), https://arstechnica.com/information-technology/2023/09/meta-launches-consumer-ai-chatbots-with-celebrity-avatars-in-its-social-apps; Max Chafkin, Meta’s New AI Buddies Aren’t Great Conversationalists, Bloomberg (Oct. 17, 2023), https://www.bloomberg.com/news/newsletters/2023-10-17/meta-s-celebrity-ai-chatbots-on-facebook-instagram-are-surreal.

[110] See supra, notes 8-14 and accompanying text.

ICLE Amicus to US Supreme Court in Apple v Epic

Amicus respectfully submits this brief in support of Petitioner Apple Inc.[1] INTEREST OF AMICUS CURIAE The International Center for Law & Economics (“ICLE”) is a . . .

Amicus respectfully submits this brief in support of Petitioner Apple Inc.[1]

INTEREST OF AMICUS CURIAE

The International Center for Law & Economics (“ICLE”) is a nonprofit, non-partisan global research and policy center aimed at building the intellectual foundations for sensible, economically grounded policy. ICLE promotes the use of law and economics methodologies and economic learning to inform policy debates and has longstanding expertise evaluating antitrust law and policy.

ICLE has an interest in ensuring that antitrust law promotes the public interest by remaining grounded in sensible rules informed by sound economic analysis. That includes fostering consistency between antitrust law and other laws that proscribe unfair methods of competition, such as California’s Unfair Competition Law, and advising against far-reaching injunctions that could deteriorate the quality of mobile ecosystems, thereby harming the interests of consumers and app developers.

INTRODUCTION AND SUMMARY OF ARGUMENT

This Court has admonished that “injunctive relief should be no more burdensome to the defendant than necessary to provide complete relief to the plaintiffs.” Califano v. Yamasaki, 442 U.S. 682, 702 (1979); see also Application for a Stay at 5, Murthy v. Missouri, No. 23-411 (Sept. 14, 2023) (review granted on government’s stay motion arguing “the injunction sweeps far beyond what is necessary to address any cognizable harm to respondents”). The nationwide injunction issued in this case, which applies to millions of non-party app developers, cannot be reconciled with that principle.

The lower court’s use of a nationwide injunction to address narrow alleged injuries has severe consequences that are best understood through the lens of law and economics principles. The district court recognized that Apple’s walled-garden ecosystem yields procompetitive consumer benefits, including greater privacy and data security, and that such benefits are cognizable under federal antitrust law. Pet. App. 261a-270a. Yet the district court’s nationwide injunction undercuts precisely those benefits. Apple’s practice of vetting unsafe payment systems and malware on its App Store depends on its ability to prevent third parties from “steering” consumers towards purchase mechanisms other than Apple’s secure in-app purchasing (“IAP”) system. In addition, the anti-steering policy prevents free-riding and protects Apple’s incentive to invest in its platform to improve the curation of apps, privacy, safety, and security.

These harms to Apple’s platform are not offset by benefits to consumers, or even to developers taken as a whole. All the injunction does is alter the allocation of app store fees between developers, because even if Apple’s ability to collect a commission through its IAP is limited, Apple would still have the right to collect a commission in other ways for the use of its proprietary software and technology. It could do so by readjusting whom it charges for access to the App Store, and how much it charges.

For instance, rather than charge a commission to developers on paid downloads of apps and on in-app purchases of digital goods and services, as it does now, Apple could instead charge all developers a fee for accessing the App Store. While this might ostensibly benefit big developers who rely heavily on in-app purchases and paid downloads to monetize their apps, it is not at all clear that the net effects would be positive. One thing does seem clear, however: The current model, in which small, free apps pay few fees, would likely cease to be tenable under a nationwide federal injunction.

Put differently, despite not violating federal antitrust law, the district court’s sweeping remedy risks harming the vast majority of app developers, who have not requested the injunction and are now operating on the iOS for free. And it may ultimately harm tens of millions of consumers using Apple’s App Store and iOS.

ARGUMENT

I.              The Injunction Is Unnecessarily Broad and Would Affect Millions of Developers, Not Just Epic

The district court imposed an injunction that affects Apple’s anti-steering provisions across the board, and thus redefines Apple’s relationship with many developers—not just Epic. As it stands, the injunction is overly broad and at odds with established jurisprudence. Gill v. Witford, 138 S. Ct. 1916, 1933-34 (2018); Califano, 442 U.S. at 702. And it reduces consumer welfare by precluding more beneficial conduct than the harmful behavior it deters.

There are about thirty million registered app developers of native iOS apps. Pet. App. 10a. There are about two million apps  available in the United States storefront for the App Store, and most of them were created by third-party developers. See Apple Inc. v. Pepper, 139 S. Ct. 1514, 1519 (2019). All the developers have signed Apple’s guidelines regarding the exclusive use of Apple’s IAP and the related anti-steering provisions. By contrast, the trial evidence established that a little over 100 developers use Epic’s Epic Store. See Pet. App. 115a (citing Trial Tr. 1220:18-20). Yet, the anti-steering injunction would affect all App Store developers. The plaintiff is not even among these developers, because Epic was jettisoned from the App Store in 2020 for introducing an in-app payment system that bypassed Apple’s IAP. Epic has only one subsidiary that is active on the App Store. See Pet. App. 12a; D.Ct. ECF No. 825-8.

It is thus unclear why the district court found it necessary to issue an injunction covering all developers who are licensed to make iOS apps for the App Store’s U.S. storefront, not just Epic’s subsidiary and the approximately 100 developers who use the Epic Store.

Two considerations are especially pertinent. First, Califano precludes the Ninth Circuit’s erroneous assertion that an injunction need only be “tied to Epic’s injuries.” Pet. App. 82a; Califano, 442 U.S. at 702. Indeed, as the government argued in a recently granted petition that raises similar issues, an overbroad injunction cannot be justified on the theory that the non-parties are simply incidental beneficiaries of the injunction for the prevailing parties. Application for a Stay, supra, at 34-36; see Order Granting Review & Order Granting Stay, Murthy v. Missouri, No. 23-411 (Oct. 20, 2023). Instead, “[i]njunctive relief may ‘be no more burdensome to the defendant than necessary to provide complete relief to the plaintiffs.’” Id. at 34-35 (quoting Califano, 442 U.S. at 702).

Second, Apple already settled a class-action lawsuit with developers regarding developer-consumer communications. As part of the Cameron v. Apple Inc. settlement, Apple deleted a prohibition on targeted communication between developers and consumers outside of the app, meaning that developers are now free to communicate outside the apps about external purchasing options (or anything else). See Order: Granting Motion for Final Approval of Class Action Settlement; Granting in Part and Denying in Part Mot. for Attorney’s Fees, Costs, and Service Award; and Judgment at 13, Cameron v. Apple Inc., No. 19-cv-03074 (N.D. Cal. June 10, 2022), ECF No. 491. That settlement, spurred by a properly certified Rule 23 class action representing around 6,700 app developers, did not, however, require Apple to modify or remove the anti-steering provision at issue here (links and buttons within apps). See Declaration of Steve W. Berman in Support of Developer Plaintiffs’ Motion for Preliminary Approval of Settlement with Defendant Apple Inc. at 7-41, Cameron v. Apple Inc., No. 19-cv-3074 (Aug. 26, 2021), ECF No. 396-1.

It is jarring that the courts would now issue a much broader injunction in a case involving a single plaintiff. This could cause serious harm to nonparties who had no opportunity to argue for more limited relief. Zayn Siddique, Nationwide Injunctions, 117 Colum. L. Rev. 2095, 2125 (2017). And it also raises the question whether such a blanket remedy is even necessary given that Cameron v. Apple strikes a balance between Apple’s ability to safeguard its investments and maintain the safety and security of its ecosystem, and app developers’ ability to steer users to alternative payment systems. That agreement was found acceptable by Apple and some 6,700 app developers. Why should it be overridden by an injunction in a case involving a single plaintiff, when app developers have already had the opportunity to join a properly certified class action before, and have either chosen not to do so or have agreed to a different settlement? Further, if a single plaintiff’s allegations of harm can undercut a court-approved, negotiated settlement involving a much larger number of plaintiffs, that diminishes the incentives of parties to fashion and negotiate reasonable settlements in the first instance.

A broad injunction may well be warranted when it is difficult to separate the parties affected by the enjoined conduct from those that are not. But this is not the case here. The identity of the parties that have supposedly been harmed is clear—they are, at most, Epic’s subsidiary and the approximately 100 developers that use the Epic Store. Even if the district court’s conclusions regarding harm to Epic’s subsidiary and other developers with apps on the Epic Store were correct, it would be easy—and necessary—to carve a much narrower remedy than the one the district court imposed. See Barr v. Am. Ass’n of Pol. Consultants, Inc., 140 S. Ct. 2335, 2354-55 (2020).

Overly broad injunctions represent a Constitutional threat, as several members of this Court have warned. See, e.g., United States v. Texas, 143 S. Ct. 1964, 1980 (2023) (Gorusch, J., concurring); Trump v. Hawaii, 138 S. Ct. 2392, 2425 (2018) (Thomas, J., concurring); see also Lewis v. Casey, 518 U.S. 343, 360 (1996). “[G]ranting a remedy beyond what [is] necessary to provide relief to [the plaintiff is] improper.” Lewis, 518 U.S. at 360. In addition to such constitutional implications, overly broad injunctions also raise problems from a law and economics perspective such as hindering and even destroying beneficial conduct. If an injunction is not properly tailored, the beneficial conduct which it precludes may be greater than the harmful conduct which it prevents, resulting in a loss to both total social welfare and consumer welfare.

II.           Platforms Have Legitimate Business Reasons for Anti-Steering Provisions

By casting an overly wide net, the district court’s injunction throws the proverbial baby out with the bathwater. Anti-steering provisions are commonly used by digital platforms and other businesses because they serve a series of legitimate aims, such as allowing for the recoupment of investments. They also result in tangible procompetitive benefits, such as increased privacy, security, and market-wide output. These rules can be procompetitive, as this Court has recognized. Ohio v. Am. Express Co., 138 S. Ct. 2274, 2289 (2018) [hereinafter Amex].

Absent intervention by this Court, Apple will have to comply with a nationwide injunction that risks diminishing these benefits. If the decision is not corrected, the precedent could have a harmful ripple effect, subjecting other platforms to overly broad injunctions against anti-steering provisions, even though those anti-steering provisions may help sustain and improve the overall quality of those platforms.

A.            The framework for assessing competitive effects in a two-sided market requires a broad examination of the market as a whole

The district court properly found that Apple’s procompetitive justifications for the anti-steering provisions in its IAP system outweighed any anticompetitive effects of those provisions. In fact, Epic failed to make even a prima facie case under the requisite rule-of-reason analysis, as Epic failed to show that Apple’s app distribution and IAP system caused the significant, market-wide competitive harm that the Supreme Court deemed necessary to a showing of anticompetitive harm in Amex.

In Amex, the Court recognized the importance of platform economics and network effects to understanding the market and competitive effects at issue. Two-sided platforms intermediate between two groups, offering a different product or service to each. 138 S. Ct. at 2280 (citing e.g., David Evans & Richard Schmalensee, Markets with Two-Sided Platforms, 1 Issues in Competition L. & Pol’y 667 (2008); David Evans & Michael Noel, Defining Antitrust Markets When Firms Operate Two-Sided Platforms, 2005 Colum. Bus. L. Rev. 667 (2005)).

The Court noted that two-sided platforms are characterized by indirect network effects, where the value of the platform to each group depends on the scale of, or number of members in, the other. Id. at 2280-81. Specifically, the Court observed that “two-sided transaction platforms exhibit more pronounced indirect network effects and interconnected pricing and demand.” Id. at 2286 (emphasis added) (citing Benjamin Klein et al., Competition in Two-Sided Markets: The Antitrust Economics of Payment Card Interchange Fees, 73 Antitrust L.J. 571, 583 (2006)). Hence, “[e]valuating both sides of a two-sided transaction platform is . . . necessary to accurately assess competition.” Id. at 2287.

B.            Anti-steering provisions can be procompetitive

At issue in Amex were various anti-steering provisions American Express had placed in its contracts with merchants. The plaintiffs had alleged that the anti-steering provisions violated Section 1 of the Sherman Act. 138 S. Ct. at 2283. But in Amex, the Court recognized that “there is nothing inherently anticompetitive about . . . antisteering provisions.” Id. at 2289. Those vertical provisions can, among other things, prevent merchants from free-riding, thereby increasing the availability of “‘tangible or intangible services or promotional efforts’ that enhance competition and consumer welfare.” Id. at 2290 (quoting Leegin Creative Leather Prods., Inc. v. PSKS, Inc., 551 U.S. 877, 890-91 (2007)).

As in Amex, understanding the competitive effects of conduct between the platform and parties on either side of the platform—for example, vertical agreements between the IAP and app developers—requires examining effects on the system as a whole. And just as in Amex, there are legitimate, procompetitive reasons for anti-steering provisions.

First, as discussed above, anti-steering provisions help prevent free-riding. Simply put, “free-riding” occurs when someone uses a valuable resource without paying for it. Free-riding—even the potential for free-riding—tends to undermine incentives to provide the resource in the first place, as well as incentives to improve it in later development. It presents an especially serious challenge to the provision of goods or services where it is difficult to exclude those who have not paid, as with city parks or policing. Everyone, even those who would be willing to pay if they had to, has an incentive to avoid fees. Thus, where free-riding is possible, desirable goods and services tend to be underfunded, reducing their provision (or, in antitrust terms, output), or, in the alternative, are provided dependent on government subsidy. The most common solution to free-rider problems is to create ways to exclude those who are unwilling to pay.

In this case, Apple owns a valuable resource that it has created and steadily improved—the iPhone and iOS ecosystem, including the App Store. Apple currently charges commissions between 15% and 30% for digital goods sold through the App Store, including for certain in-app purchases. Epic would like to access that ecosystem without paying. But while Epic may benefit from its long-term strategy to reduce the fees it pays to Apple, consumers might not. If reductions in revenue from the iOS ecosystem mean that Apple has less incentive to invest in it, Epic’s gain may come at the consumer’s expense.

The district court correctly rejected Epic’s main claim, as Epic failed to establish cognizable harm under the antitrust laws. That foreclosed Epic’s ability to directly circumvent the App Store and pay a lower commission, or none at all. In granting a nationwide injunction against Apple’s anti-steering provisions, the district court facilitated precisely the type of free-riding that failed to gain traction under federal antitrust law. Doing so will greatly exacerbate any free-rider problem Epic itself might have caused Apple, to the likely detriment of many developers and most consumers.

The situation is further complicated by the fact that the district court’s injunction is vaguely written and is thus likely to be interpreted quite differently by different parties. Ultimately, if it allows app developers to link users outside of the in-app payments flow, and bypass Apple’s IAP fees, it will further enable free-riding and undermine Apple’s incentives to invest in iOS, iPhones, and iPads. And the injunction could undermine the incentive for Apple’s competitors to develop whatever products might someday displace the current ones through competition.

Second, as a two-sided market, the App Store is valuable only because it is used by both consumers and developers, and Apple has to balance both sides of the market. The risk of developers leaving the iOS ecosystem creates a built-in ceiling on the prices Apple can charge, as users will be less inclined to pay for Apple products if valuable apps are not there. The commission-fee business model gives Apple and other platforms significant incentives to develop new distribution mediums (like smart TVs, for example) and to improve existing ones. Such development expands the audience that software can reach.

Apple’s “closed” distribution model also allows the company to curate the App Store’s apps and payment options. For example, Apple’s guidelines exclude apps that pose data security threats, threaten to impose physical harm on users, or undermine child-safety filters. These rules increase trust between users and previously unknown developers, because users do not have to fear their apps contain malware. They also reduce user fears about payment fraud. Rivals could free-ride on Apple’s curation by mimicking its decisions and undercutting it on price. Doing so does not enhance competition on the merits: It eviscerates it by eroding Apple’s incentives to enforce such rules.

Apple’s closed business model also enables it to maintain a high standard of performance on iOS devices by excluding apps and payment systems that might slow devices or crash frequently. Users may not know when device performance is affected by a given app or purchase mechanism, so an open system would mean the potential for apps that crash the entire device. Apple’s closed model ensures that unscrupulous developers cannot impose negative externalities on the entire ecosystem.

By increasing the total value of the platform, these benefits also increase the number of market-wide transactions. In a two-sided market, it is output—not prices—that tells us what is happening on the market as a whole, and it is therefore output that should be used as the relevant parameter to determine whether conduct is procompetitive or anticompetitive. “What is material is whether Apple’s overall pricing structure reduces output by deterring app developers from participating in the market or users from purchasing apps (or iOS devices at all) because of the amount of the app developer commission.” Geoffrey A. Manne & Kristian Stout, The Evolution of Antitrust Doctrine After Ohio v. Amex and the Apple v. Pepper Decision That Should Have Been, 98 Neb. L. Rev. 425, 457 (2019). Notably, the district court found that it could not ascertain whether Apple’s alleged restrictions had “a negative or a positive impact on game transaction volume.” Pet. App. 253a; see also id. (“no evidence that a substantial number of developers actually forego making games because of Apple’s commission.”); id. at 319a (finding Epic failed to show reduction in output and that “[t]he record contains substantial evidence that output has increased.”).

Ultimately, however, the benefits of anti-steering provisions are obvious only if one adopts the correct, holistic vision of app stores as a two-sided market; conversely, they appear less relevant if one applies “one-sided logic in two-sided markets.” Julian Wright, One-sided Logic in Two-sided Markets, 3 Review of Network Econ. 44, 45-51 (2004). In this sense, in a two-sided market, anti-steering provisions can reduce transaction friction and bolster security and safety, thereby improving the platform’s overall quality and, ultimately, attracting more users. See Amex, 138 S. Ct. at 2889 (sustaining similar anti-circumvention rules as procompetitive for these reasons). Developers may get a smaller share of revenues, but it is a smaller slice of a much larger pie. Thus, while the ability to circumvent Apple’s commission fee can, on the surface, appear to benefit some developers, in the longer term most developers and consumers will be worse off.

Apple’s anti-steering provisions increase safety and curation, and an injunction against them can reduce the overall value of Apple’s platform. That would in turn discourage developers and users from using the iOS ecosystem, and would prompt a downward spiral in quality and choice for both sides of the market—which would depreciate the value of the platform even further.

C.            Open and closed platforms are not inherently good or bad: They represent alternative business models with potential advantages and disadvantages

Any comparison between “open” and “closed” platforms should account for the fact that there are tradeoffs between the two; it should not simply assume that “open” equals “good” while “closed” equals “bad.” Such analysis also must consider tradeoffs among consumers, and among developers, in addition to tradeoffs between developers and consumers. 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 especially benefits smaller developers and those whose apps fall outside the popular gaming sector. In short, the IAP fee cross-subsidizes the delivery of services to the approximately 80% of apps on the App Store that are free to consumers and pay no IAP fees.

Centralized app distribution and Apple’s “walled garden” model (including IAP) increase interbrand competition because they are at the core of what differentiates Apple from Android, the other major competing platform. 1-ER-148–49. They play into Apple’s historical business model, which focuses on being user-friendly, reliable, safe, private, and secure. 1?ER-86; see also 1-ER-107 (recognizing that the safety and security of Apple’s closed system is a “competitive differentiator for its devices and operating system”). Even Epic recognized that Apple would lose its competitive advantage if it were to compromise its safety and security features. 1-ER-48 n.250 (noting Epic’s expert, Susan Athey, testified that “privacy and security are competitive differentiators for Apple”).

For Apple and its users, the touchstone of a good platform is not “openness,” but carefully curated selection and security, understood broadly as encompassing the removal of objectionable content, protection of privacy, and protection from “social engineering,” and the like. 1-ER-148–49. By contrast, Android’s bet is on the open platform model, which sacrifices some degree of security for the greater variety and customization associated with more open distribution. These are legitimate differences in product design and business philosophy. See Andrei Hagiu, Proprietary vs. Open Two-Sided Platforms and Social Efficiency 2-3 (AEI-Brookings Joint Ctr. for Regul. Stud., Working Paper No. 06-12, 2006) [hereinafter Proprietary vs. Open Platforms] (explaining that there is a “fundamental welfare tradeoff between two-sided proprietary . . . platforms and two-sided open platforms, which allow ‘free entry’ on both sides of the market” and thus “it is by no means obvious which type of platform will create higher product variety, consumer adoption and total social welfare”) (emphasis omitted); Jonathan M. Barnett, The Host’s Dilemma: Strategic Forfeiture in Platform Mkts. for Informational Goods, 124 Harv. L. Rev. 1861, 1927 (2011) (“Open systems may yield no net social gain over closed systems, can impose a net social loss under certain circumstances, and . . . can impose a net social gain under yet other circumstances.”).

Because consumers and developers could reasonably prefer either ecosystem, it is not clear that loosening Apple’s control over the App Store would necessarily lead to more app transactions market wide. Indeed, in a two-sided market context, a proprietary platform like Apple’s “may in fact induce more developer entry (i.e. product variety), user adoption and higher total social welfare than an open platform.” Proprietary vs. Open Platforms, at 15-16. In other words, preventing certain apps from accessing the App Store, and preventing certain transactions from taking place on it, may ultimately have increased the number of apps and transactions on Apple’s platform, because doing so made it attractive to a wider set of consumers and developers.

Yet the injunction brings Apple’s iOS closer to an “open” system, effectively rendering Apple’s platform more similar to Android’s. The district court found that Apple did not have a monopoly, yet under the guise of fostering competition on Apple’s platform the injunction eliminates competition where it matters most—at the interbrand, systems level between Apple and Android. See Michael L. Katz & Carl Shapiro, Systems Competition and Network Effects, 8 J. Econ. Persps. 93, 110 (1994), (“[T]he primary cost of standardization is a loss of variety: consumers have fewer differentiated products to pick from, especially if standardization prevents the development of promising but unique and incompatible new systems”). By limiting intrabrand competition, in other words, Apple ultimately promotes interbrand competition. 1-ER-148–49. Again, Amex provides useful insight here, because the Court noted that the business model had “spurred robust interbrand competition,” while increasing both the quality and quantity of transactions. Amex, 138 S. Ct. at 2290.

D.            Anti-steering provisions are a legitimate way of recouping a platform’s investments

Anti-steering provisions are a legitimate way for a platform to recoup its investments. Epic has argued that Apple could simply lift restrictions on the use of third-party IAP processors (e.g., Visa and MasterCard), but still be appropriately compensated for the use of its intellectual property, ensure that iPhone users’ IAP are sufficiently secure, and guarantee quality. 1-ER-153; Epic 9th Cir. Br. 44-47. But exactly how Apple could achieve these ends without increasing its costs is a question Epic has not even tried to answer. See, e.g., 1-ER-151–52 (noting that Epic’s requests for relief “leave unclear whether Apple can collect licensing royalties and, if so, how it would do so”); 1-ER-153 & n.617 (noting it would “be more difficult” and more costly for Apple to collect commission without the IAP system). Nor did Epic, the Epic amici, or the district court properly address the effect of the proposed less restrictive alternatives on consumers rather than competing developers. See 1-ER-148 n.605 (noting it is “unclear the extent or degree to which developers would pass on any savings to consumers”).

Consistent with Epic’s proposed approach, Apple could allow independent payment processors to compete, and charge an all-in fee of 30% when Apple’s IAP is chosen. To recoup the costs of developing and running its App Store, Apple could then charge app developers a reduced, mandatory per-transaction fee (on top of developers’ “competitive” payment to a third-party IAP provider) when Apple’s IAP is not used. Indeed, where a similar remedy has been imposed already, Apple has taken similar steps. In the Netherlands, for example, where Apple is required by the Authority for Consumers and Markets to uncouple distribution and payments for dating apps, Apple has adopted a policy under which any apps that want to use a non-Apple payment provider must still “pay Apple a commission on transactions” that is 3% less than normal (so 27% for most transactions), a slightly “reduced rate that excludes value related to payment processing and related activities.” Apple, Distributing Dating Apps in the Netherlands, (last visited Oct. 26, 2023).

III.         A State Law Should Not Undermine the Fundamental Goals of Federal Antitrust Policy

When assessing the effects of Apple’s anti-steering provisions, the courts should not ignore Federal antitrust law and, especially, the effects on competition and consumers. In other words, the fact that anti-steering provisions are procompetitive should be a relevant factor in whether a federal court grants nationwide injunctive relief. To interpret California’s Unfair Competition Law (“UCL”) as the district court has done—in a way that is at loggerheads with federal antitrust law but yet permits a nationwide injunction—is to undermine the fundamental goal of antitrust policy, and to do so on a national level. As the Court has observed, “The heart of our national economic policy long has been faith in the value of competition.” Nat’l Soc’y of Prof. Eng’rs v. United States, 435 U.S. 679, 695 (1978) (quoting Standard Oil Co. v. FTC, 340 U.S. 231, 248 (1951)).

The district court recognized Apple’s security arguments as a key procompetitive factor that determines Apple’s success and increases output across the platform, ultimately benefitting both consumers and developers. Yet the court issued an unnecessarily broad injunction against Apple’s anti-steering provisions that risks chilling procompetitive conduct by deterring investment in efficiency-enhancing business practices, such as Apple’s “walled-garden” iOS (see sections II.B and II.D on the procompetitive benefits of anti-steering provisions). See also Verizon Commc’ns, Inc. v. Law Offices of Curtis V. Trinko, LLP, 540 U.S. 398, 414 (2004) (“[F]alse condemnations ‘are especially costly, because they chill the very conduct the antitrust laws are designed to protect.’”) (quoting Matsushita Elec. Indus. Co. v. Zenith Radio Corp., 475 U.S. 574, 594 (1986)).

Even more egregiously, perhaps: It risks undermining federal antitrust law by enjoining conduct under state unfair competition law that is recognized as benign—and even beneficial—under federal antitrust law. If the district court’s remedy is left to stand, state laws will be stretched beyond their territorial remit and used to contradict federal antitrust laws nationally, thus eviscerating federal antitrust policy from the bottom-up. This is not a hypothetical threat, either: California has already expressed its intent to use the UCL to “seek nationwide injunctions” on the same theory as the ruling below. Michael Acton, Epic Games-Apple US Appeals Court Ruling Shows Power of California’s Competition Law, Blizzard Says, MLex (May 10, 2023).

The district court erred in finding Apple’s anti-steering provision “unfair” despite a concurrent finding that there is no incipient antitrust violation. And a nationwide injunction based on that finding lifts what could have been a relatively contained mistake to the national level, and thereby magnifies it.

This is misguided from an antitrust perspective because it undermines some of the procompetitive benefits that anti-steering provisions in closed two-sided platforms can give to consumers and app developers. A national injunction that subverts Apple’s ability to charge a commission for the use of its software and technology through paid apps and in-app payments might also alter the current balance between the two sides of the App Store, to the detriment of smaller developers of free apps. In this zero-sum game, the gain of a handful of developers who rely on paid downloads of apps and frequent in-app purchases by users will come at the expense of the majority who do not.

CONCLUSION

For the foregoing reasons, this Court should grant Apple’s petition for a writ of certiorari.

[1] Amicus notified counsel for the parties of its intent to file this brief more than ten days before the deadline. No counsel in this matter for any party authored this brief in whole or in part, and no person other than amicus or its counsel have made any monetary contribution intended to fund the preparation or submission of this brief.

Food-Retail Competition, Antitrust Law, and the Kroger/Albertsons Merger

Executive Summary In October 2022, the Kroger Co. and Albertsons Cos. Inc. announced their intent to merge in a deal valued at $24.6 billion.[1] Given . . .

Executive Summary

In October 2022, the Kroger Co. and Albertsons Cos. Inc. announced their intent to merge in a deal valued at $24.6 billion.[1] Given the Federal Trade Commission’s (FTC) increasingly aggressive enforcement stance against mergers and acquisitions, as well as Chair Lina Khan’s previous writings on food retail specifically,[2] the agency appears poised to try to block the transaction—even with divestitures.[3] The FTC and U.S. Justice Department’s (DOJ) recently unveiled draft revisions to the agencies’ merger guidelines further suggest that they plan to challenge more mergers—and to do so more aggressively than under past administrations.[4]

But attempting to block this transaction would go against the analytical framework the FTC has historically used to evaluate similar transactions, as well as the agency’s historical precedent of accepting divestures as a remedy to address localized problems where they arise. Such breaks with the past sometimes happen; our understanding of the law and economics evolves. But in the case at hand, these breaks from tradition would reflect a failure to consider relevant and significant changes in how consumers shop for food and groceries in today’s world.

The FTC has a long history of assessing retail mergers in a manner significantly at odds with the aggressive approach it is currently signaling. Only one supermarket merger has been challenged in court since American Store’s acquisition of Lucky Stores in 1988: the Whole Foods/Wild Oats merger in 2007.[5] Over the last 35 years, the FTC has allowed every other supermarket merger and most retail-store transactions to proceed with divestitures. Within the last two years alone, these have included Tractor Supply/Orschlein and 7-Eleven/Speedway.[6] The FTC’s historic approach recognizes the reality that competitive concerns regarding supermarket mergers can be readily and adequately remedied by divestitures in geographic markets of concern; indeed, even Whole Foods/Wild Oats was ultimately resolved with a divestiture agreement following a fractured circuit court decision.[7]

The retail food and grocery landscape has changed dramatically since American/Lucky and Whole Foods/Wild Oats. But the ways in which the market has changed point toward its becoming more competitive, further undermining a possible FTC case. With the growth of wholesale clubs, delivery services, e-commerce, and other retail formats, the industry is no longer dominated by traditional supermarkets. In addition, these changing dynamics have made geographic distance, traffic patterns, and population density decreasingly relevant in a consumer’s choice of where they purchase food and groceries. Today, Kroger is only the fourth-largest food and grocery retailer in the United States, behind Walmart, Amazon, and Costco. If the merger goes through, the combined firm will move into third place in market share but would still account for just 9% of nationwide sales.[8]

The upshot is that the food and grocery industry is arguably as competitive as it has ever been. Unfortunately, recent developments suggest the FTC may well ignore or dismiss the economic realities of this rapid transformation of the food and grocery industry, substituting instead the outdated approach to market definition and industry concentration signaled by the draft guidelines.[9]

In light of these developments in the food-retail market and the FTC’s likely break from precedent, this paper highlights important areas where both the commission and commentators’ stances appear to run headlong into legal precedent that mandates an evidence-based approach to merger review, even as the best available evidence points to a dynamic and competitive grocery industry. The correct understanding of the law and the industry appears entirely at odds with a challenge to the proposed merger.

A.     The FTC’s Merger-Enforcement Policy Is on a Collision Course with the Law

The Kroger/Albertsons merger proceeds against a backdrop of tough merger-enforcement rhetoric and actions from the FTC. Recent developments include the publication of aggressive revised merger guidelines, a string of cases brought to block seemingly benign mergers, process revisions that burden even unproblematic mergers, and FTC leadership’s contentious and expansive interpretation of the merger laws. The FTC’s ambition to remake U.S. merger law is likely to falter before the courts, but not before imposing a substantial tax on all corporate transactions—and, ultimately, on consumers.

The retail food and grocery market has changed substantially since the last time a supermarket merger was challenged in court. If the merger goes to trial, the court will need to address issues that have not been litigated in decades, if ever. Depending how the court rules, the market definition for future supermarket mergers may be substantially revised. Moreover, if the FTC attempts to litigate allegations of labor-market or input-market monopsony, the agency runs the risk of a humiliating loss that could stymie future attempts to expand the role of monopsony in competition enforcement and policy. The FTC thus would do well to even-handedly assess the Kroger Albertsons merger, remaining open to new evidence and sensible remedies. This is especially true given the agency’s losing streak in court—culminating with its unsuccessful challenges of the Meta/Within and Microsoft/Activision Blizzard deals.[10]

B.     The Product Market Is Broader Than Local Supermarkets

Because of recent changes in market dynamics, it no longer makes sense to limit the relevant market to supermarkets alone. Rather, consumer behavior in the face of omnipresent wholesale clubs, e-commerce, and local delivery platforms significantly constrains supermarkets’ pricing decisions.

Recent FTC consent orders involving supermarket mergers have limited the relevant product market to local brick-and-mortar supermarkets and food and grocery sales at nearby hypermarkets (e.g., Walmart supercenters), while excluding wholesale-club stores (e.g., Costco), e-commerce (e.g., Amazon), and further-flung stores accessible through online-delivery platforms (e.g., Instacart). This is based on an assertion that the relevant market includes only those retail formats in which a consumer can purchase nearly all of a household’s weekly food and grocery needs from a single stop, at a single retailer, in the shopper’s neighborhood. This is, however, no longer how most of today’s consumers shop. Instead, shoppers purchase different bundles of groceries from multiple sources, often simultaneously.[11] This pattern has substantial implications for supermarkets’ competitive environment, and underscores why the FTC should not rely on outdated market definitions.

Past FTC consent orders have defined the relevant geographic markets to be areas that range from a two- to ten-mile radius around each of the merging parties’ supermarkets.[12] The radius depends on such factors as population density, traffic patterns, and the unique characteristics of each market. It would, however, be reasonable to expand the relevant geographic market when club stores are present, as these have a larger catchment area than supermarkets. Finally, the rapid growth of e-commerce and delivery services make distance, traffic patterns, and population density decreasingly relevant in a consumer’s choice of where they purchase food and groceries. As with product-market definitions, this is a crucial empirical issue that should be evaluated in the FTC’s merger review and any litigation.

C.     Labor Monopsony Concerns Are Unlikely to Hold Up in Court

More than in any previous retail merger, opponents of the Kroger/Albertsons deal have raised the specter of potential monopsony power in labor markets. But these concerns reflect a manifestly unrealistic conception of labor-market competition. Fundamentally, the market for labor in the retail sector is extremely competitive, and workers have a wide range of alternative employment options—both in and out of the retail sector. At the same time, both Kroger and Albertsons are highly unionized, providing a counterbalance to any potential exercise of monopsony power by the merged firm.

D.    The Alleged ‘Waterbed Effect’ Is Not Borne Out by Evidence

Some critics of the merger have speculated that the merged company would be able to exercise monopsony power against its food and grocery suppliers (i.e., wholesalers and small manufacturers), often invoking an economic concept called the “waterbed effect.” The intuition is that the largest buyers may use their monopsony power to negotiate lower input prices from suppliers, leading the suppliers to make up the lost revenue by raising prices for their smaller, weaker buyers.

But these arguments are far from compelling. It is very difficult, for example, to hypothesize any relevant market for purchasing a good where the merged firm would have market power. Critics also often fail to consider the ability of many producers—both small and large—to sell directly to consumers, as demonstrated by the rise of online shopping, with its low entry barriers entry and low-cost structure.

Much of the discussion of the waterbed effect focuses on harm to competing retailers, rather than consumers. But this is not the harm that U.S. antitrust law seeks to prevent. It is thus not surprising that at least one U.S. court has rejected waterbed-effect claims on grounds that there was no harm to consumers.

E.     Divestitures Historically Have Proven an Appropriate and Adequate Remedy

Historically, the FTC has allowed most grocery-store transactions to proceed with divestitures, such as Ahold/Delhaize (81 stores divested), Albertsons/Safeway (168 stores), and Price Chopper/Tops (12 stores). The extent of the remedies sought depends on the extent of post-merger competition in the relevant local markets, as well as the likelihood of significant entry by additional competitors into the relevant markets. The benefit of selling off stores is that you can allow the vast majority of stores—where there is no worry about anticompetitive effect—to merge, while targeting the areas that have the highest concern.

Despite a long history of divestitures serving as an appropriate and adequate remedy in supermarket mergers, some point to the Albertsons/Safeway merger divestitures to Haggen as evidence that divestitures are no longer an appropriate remedy. But several factors idiosyncratic to Haggen and its acquisition strategy led to the failure of that divestiture, and it does not properly stand for the claim that all supermarket divestitures are doomed.

In September 2023, Kroger and Albertsons announced a $1.9 billion divestiture proposal to sell 413 stores, eight distribution centers, two offices, and five private-label brands to C&S Wholesale Grocers LLC.[13] If consummated, the deal would cover operations spanning 17 states and the District of Columbia, and C&S has committed to maintain collective-bargaining agreements with labor.[14] As antitrust enforcers review whether these proposed divestitures are adequate, they should learn from the Haggen experience, rather than view it as justification to reject reasonable divestiture options that have worked for other mergers.

I.  Introduction

In October 2022, the Kroger Co. and Albertsons Cos. announced their intent to merge the two companies in a deal valued at $24.6 billion.[15] Kroger is the fourth-largest food and grocery retailer in the United States—behind Walmart, Amazon, and Costco—while Albertsons is fifth.[16] Both chains trail market-leading Walmart  by a considerable margin. Kroger operates 2,726 stores under the Kroger, Harris Teeter, and Smith’s banners, while Albertsons operates 2,278 stores under the Safeway, Albertsons, and Von’s grocery banners.[17] By contrast, Walmart and Sam’s Club combined store count is greater than 5,300, and the company’s grocery revenue is more than twice that of Kroger and Albertsons combined.[18]

While the proposed Kroger/Albertsons merger is a large transaction in terms of dollar valuation and the combined firm would move into third place in market share, it would still account for just 9% of nationwide sales.[19] In some localities, the market share would be much larger, however, raising questions regarding whether the merger would increase Kroger/Albertson’s monopoly power in those retail markets and convey monopsony power in wholesale markets and local labor markets. To address such questions, Kroger and Albertsons have announced a $1.9 billion divestiture proposal that would include the sale of 413 stores and eight distribution centers across 17 states to C&S Wholesale Grocers LLC.[20]

But given the Federal Trade Commission’s (FTC) recent disposition toward proposed mergers and Chair Lina Khan’s previous writings about food retail, it is widely expected that the FTC will not be satisfied with any remedy offers from the companies—including offers to divest stores—and will instead attempt to block the merger.[21] California Attorney General Rob Bonta has also signaled that he is likely to challenge the deal.[22] In anticipation, Kroger’s chief executive officer announced that both companies are “committed to litigate” if the enforcers act to block the merger.[23]

What can we expect of such a court battle? The precedent is slightly complicated. While several retail mergers have been challenged in court, including Staples/Office Depot in 2015[24] and Whole Foods/Wild Oats in 2007,[25] no supermarket mergers have been litigated since the State of California’s 1988 challenge to American Store’s acquisition of Lucky Stores.[26] Both the supermarket business and antitrust analysis have changed dramatically over the intervening 35 years. In this paper, we describe some of the changes that have occurred within food retail over the past 35 years, and how they should be addressed by the merging parties, the FTC, and—if litigated—by the courts.

As with most merger analysis, many of the most important questions hinge on a proper definition of the relevant market. The most obvious changes we have observed in food retail in recent decades, including the rise of wholesale club stores and e-commerce, are directly relevant to the question of market definition. The most recent FTC consent orders involving supermarket mergers defined the relevant product market to include “traditional” brick-and-mortar supermarkets, as well as food and grocery sales at hypermarkets (e.g., Walmart supercenters), while excluding wholesale club stores (e.g., Costco) and e-commerce (e.g., Amazon and home-delivery services).[27] This is based on the longstanding assertion that the relevant market includes only those retail formats in which a consumer can purchase all or nearly all of their household’s weekly food and grocery needs during a single stop at a single retailer.

Any attempt by the FTC to maintain this outdated market definition will likely be challenged in court. Research has shown that consumer behavior has changed over time, in that the typical consumer no longer makes once-a-week shopping trips to a single food and grocery retailer.[28] Instead, the typical consumer makes multiple weekly trips and multi-homes across several different retailers and retail formats. This change alone blurs the line between traditional supermarkets and other retail formats. The extent to which wholesale clubs, e-commerce, and delivery services should be included in the relevant market is a key empirical issue that surely will—and should—be evaluated in the FTC’s merger review and any ensuing litigation.

Geographic market is also an issue. The most recent FTC consent orders have defined the relevant geographic markets to be areas that range from a two- to ten-mile radius surrounding each of the merging parties’ supermarkets.[29] The radius depends on such factors as population density, traffic patterns, and the unique characteristics of each market. That, too, needs revision. Based on academic research, industry surveys, and reports from companies that find club stores compete with supermarkets and have larger catchment areas, we conclude that it would be reasonable to expand the relevant geographic market when club stores are present. In addition, the rapid growth of e-commerce and delivery services make distance, traffic patterns, and population density decreasingly relevant in a consumer’s choice of where they purchase food and groceries. As with product-market definition, this is a crucial empirical issue that should be evaluated in the FTC’s merger review and any ensuing litigation.

More than in any previous retail merger, opponents of the Kroger/Albertsons merger have raised the specter of potential monopsony power in labor markets. We argue that these concerns are likely overblown and will be nearly impossible to demonstrate if the merger were to be litigated. Fundamentally, the market for labor in the retail sector is highly competitive, with workers having a wide range of alternative employment if a particular employer attempted to exploit any claimed monopsony power. In addition, both Kroger and Albertsons are highly unionized. Through their collective-bargaining agreements, unions exercise monopoly power in labor negotiations that likely counterbalances any attempted exercise of monopsony power by the merged firm.

Lastly, some critics of the merger have speculated that the merged company may have and exercise monopsony power among its food and grocery suppliers (e.g., wholesalers and small manufacturers). In particular, critics invoke a concept colloquially called the “waterbed effect,” where pushing input prices down for some retailers ends up raising the price for other retailers. Why prices are “pushed” down is not always clear in popular discussions, nor is it clear that it qualifies as an antitrust harm in any way. Being the easiest trading partner would also result in lower prices.

We conclude this may be the weakest argument raised in opposition to the merger. The United Kingdom has evaluated “waterbed effect” allegations in at least two supermarket mergers and found no evidence indicating any anticipated effects from the mergers on input prices that would harm consumers.[30] More importantly, much of the discussion of waterbed effects focuses on harm to competing retailers, rather than to consumers. At least one U.S. court has rejected waterbed-effect claims on the grounds that (1) the plaintiffs did not demonstrate any harm to consumers, and (2) firms can substitute to other suppliers, thereby mitigating any anticipated waterbed effect.[31]

Given the size of a merged Kroger and Albertsons, it would be easy, but naïve, to conclude that the merger should be blocked.[32] The retail food and grocery market has changed substantially since the last time a supermarket merger was challenged in court. If the merger goes to trial, the court will address issues that have not been litigated in decades, if ever. Depending how the court rules, the market definition for future supermarket mergers may be substantially revised. Moreover, if the FTC attempts to litigate allegations of labor-market or input-market monopsony, the agency runs the risk of a humiliating loss that could stymie future attempts to expand the role of monopsony in competition enforcement and policy.

II.     The Agencies Are Trying to Rewrite Merger-Review Standards

The recently published FTC-DOJ draft merger guidelines are a particularly notable backdrop for the Kroger/Albertsons merger, leading many commentators to expect the FTC to take a hardline stance on the deal.[33] Merger case law, however, has not changed much in recent years. Given the merging parties’ apparent willingness to litigate the case, if necessary, the likelihood of a protracted legal battle appears to be high. As we explain below, at least at first sight, any case against the merger would be largely built on sand, and the commission’s chances of succeeding in court appear slim.

The Clayton Act of 1914 grants the U.S. government authority to review and challenge mergers that may substantially lessen competition. The FTC and DOJ are the two antitrust agencies that share responsibility to enforce this law. Traditionally, the FTC investigates retail mergers, while the DOJ oversees other sectors, such as telecommunications, banking, and transportation.

Before the FTC and DOJ officials appointed by the current administration came into office, the settled practice was for the antitrust agencies to follow the 2010 Horizontal Merger Guidelines, which outline the analytical framework and evidence they use to evaluate mergers. The 2010 guidelines describe four major steps of merger analysis:

  1. The first step is to define the relevant product and geographic markets affected by the merger. The goal is to identify the set of products and regions that are close substitutes to the merging parties’ products and regions.
  2. The second step is to assess the merger’s competitive effects, or how the merger may harm competition in the defined markets.
  3. The third step is to examine the role of market entry as a potential counterbalance to the merger’s competitive effects. For entry to be sufficient to deter or undo the anticompetitive effects of a merger, it must be timely, likely, and sufficient in scale and scope.
  4. The fourth and final step is to evaluate the efficiencies the merger would generate, or how the merger may benefit consumers by reducing costs and improving quality.

The antitrust agencies weigh all these factors to determine whether a merger is likely to harm competition and consumers. If they find that a merger raises significant competitive concerns, they may seek to block it or require remedies such as divestitures or behavioral commitments from the merging parties.

Several factors, however, suggest that authorities are unlikely to follow this measured approach when reviewing the Kroger/Albertsons merger. Primarily, the FTC and DOJ have recently issued draft revised merger guidelines. The 2023 guidelines have not yet been adopted, although the public comment period is closed. Compared to the previous iteration, which guided recent consent decrees, the new guidelines contain more stringent structural presumptions—that is, a presumption that a merger that merely increases concentration (as all horizontal mergers do) by a certain amount violates the law, rather than deferring to more nuanced economic analysis connecting specific market attributes to a likelihood of actual consumer harm.[34] These new, more stringent structural presumptions are not justified by new economic learnings about the economic effects of concentration. As now-FTC Bureau of Economics Director Aviv Nevo and colleagues wrote in 2022 (just before he joined the commission):

If the agencies were to substantially change the presumption thresholds, they would also need to persuade courts that the new thresholds were at the right level. Is the evidence there to do so? The existing body of research on this question is, today, thin and mostly based on individual case studies in a handful of industries. Our reading of the literature is that it is not clear and persuasive enough, at this point in time, to support a substantially different threshold that will be applied across the board to all industries and market conditions.[35]

Although elements of Nevo and coauthors’ proposed framework are present in the new proposed guidelines, the guidelines also incorporate new language that reflects a persistent thumb on the scale, systematically undermining merging parties’ ability to justify their merger. For example, while a presumption of harm is triggered at a certain level of concentration (an HHI of 1800), in markets where there has previously been consolidation (over an unspecified timeframe), an impermissible “trend [toward concentration] can be established by… a steadily increasing HHI [that] exceeds 1,000 and rises toward 1,800.”[36] Traditionally, an HHI under 1500 would be considered “unconcentrated” and presumed to raise no competitive concerns.[37]

While the FTC will likely point to the renewed focus on concentration measures as capturing the Clayton Act’s focus on lessening competition and the tendency to create a monopoly, the draft guidelines make clear that commission now views concentration as problematic in itself, regardless of whether it lessens competition. For example, the draft guidelines state “efficiencies are not cognizable if they will accelerate a trend toward concentration.”[38] Such a statement effectively negates any efficiency defense available to all but the very smallest firms. Efficiencies will almost always increase concentration—especially if those efficiencies come from economies of scale. If a merger creates efficiencies, the merged firm can lower costs, cut prices, and attract more customers. Attracting more customers with better products and prices will likely increase competition.

The economic evidence is quite strong that efficiency increases concentration.[39] If no efficiency defense is possible, any horizontal merger could accelerate a trend toward concentration (if it had been previously becoming more concentrated). Spinning in these circles is why the notion of a “trend” toward concentration raising particular concern hasn’t been reflected in guidelines since 1968,[40] and reached its apotheosis in Von’s Grocery in 1966[41]—one of the most thoroughly reviled merger cases in U.S. history.[42]

Before updating the merger guidelines, the FTC had already started to tighten its merger-enforcement policy. Among other actions, the agency brought high-profile cases against the Illumina/GRAIL, Meta/Within, and Microsoft/Activision Blizzard deals.[43] So far, all three challenges have resulted in defeat for the FTC in adjudication. Taken together, these cases suggest the agency is willing to push the law beyond its limits in an attempt to limit corporate consolidation, whatever the actual competitive effect. The courts have thus far shown themselves unwilling to buy the agencies more speculative claims of harm. In contrast, the DOJ recently was able to block Penguin-Random House from merging with Simon & Schuster.[44] On the surface, this may seem like a novel case similar to those the FTC has been pursuing; it involved monopsony power against authors. But in this case, the parties agreed that, if there was harm to the authors, there would be fewer books, thereby harming consumers.[45] Fighting over harms to consumers (not concentration) is textbook antitrust litigation.[46]

Finally, the FTC’s leadership has been particularly bearish about the potential consumer benefits of corporate mergers and acquisitions. This inclination is reflected in Chair Khan’s assertion that the Clayton Act embodies a “broad mandate aimed at prohibiting mergers even when they do not constitute monopolization and even when their tendency to lessen competition is not certain.”[47] One way to prohibit mergers is to make them more costly without even going to court. This is what will happen under the agencies’ proposed changes to the premerger notification rules (“Hart-Scott-Rodino Act”).[48] Despite no evidence presented of anticompetitive mergers slipping through the cracks due to the current reporting being too lax, the revised guidelines would greatly increase the cost of merging, thereby reducing the number of mergers.

Even the FTC estimates a massive increase in compliance costs of approximately $350 million, to more than $470 million per year. But that estimate is likely a serious underestimate, as it is based on, among other things, an unscientific “estimate” of the time involved and a dated assumption about the average hourly costs imposed on filers’ senior executives and firms’ counsel.

A survey that the U.S. Chamber of Commerce conducted of 70 antitrust practitioners about the proposed HSR revisions found that the new rules would increase compliance costs by $1.66 billion, almost five times the FTC’s $350 million estimate.”[49] While this general approach to blocking more mergers will not be directly applicable in any particular case, it highlights the FTC’s willingness to not follow “the old rules.”

All of these factors—in concert with the merging parties’ claim that they are prepared to go to court if the FTC decides to block the transaction outright[50]—suggest that there is a particularly high likelihood that the Kroger/Albertsons merger will be challenged and litigated, rather than approved, or challenged and settled.

For the reasons outlined in the following sections, however, the FTC is unlikely to prevail in court. The market overlaps between the merging parties are few and can be resolved by relatively straightforward divestiture remedies, as already proposed by the parties—which, even if disfavored by the agency, are routinely accepted by courts. Likewise, the FTC’s likely market definition and potential theories of harm pertaining to labor monopsony and purchasing power more generally appear speculative at best. The upshot is that the FTC’s desire to bring tougher merger enforcement appears to be on a collision course with the law as it is currently enforced by U.S. courts.

III.   The Relevant Market Is Broader Than Hyper-Local Supermarkets

The retail food and grocery landscape has changed dramatically since the last litigated supermarket merger. Consumer-shopping behavior has shifted toward more frequent shopping trips across a wide variety of formats, which include warehouse clubs (e.g., Costco); e-commerce (e.g., Amazon); online-delivery platforms (e.g., Instacart); limited-assortment stores (e.g., Trader Joe’s and Aldi); natural and organic markets (e.g., Whole Foods); and ethnic-specialty stores (e.g., H Mart); in addition to traditional supermarkets. Because of these enormous changes, the market definition assumed in previous FTC consent orders likely will be—and should be—challenged, given the empirical evidence.

A.     Recent Trends in Retailing Have Upended the ‘Traditional’ Grocery Market Definition

The FTC is likely to find the relevant product market to be supermarkets, which the agency has previously defined as retail stores that enable consumers to purchase all of their weekly food and grocery requirements during a single shopping visit. This product-market definition has remained mostly unchanged for at least a quarter of a century. In both the Albertsons/Safeway merger and the Ahold/Delhaize merger, consent orders between the FTC and the merging parties defined the relevant market to be “the retail sale of food and other grocery products in supermarkets.”[51] The orders defined supermarkets as “full-line grocery stores” that provide “one-stop shopping” that enables consumers “to shop in a single store for all of their food and grocery needs.”[52]

On the one hand, the consent orders’ product-market definitions included supermarkets located within hypermarkets, such as Walmart supercenters. Hypermarkets sell both products that are not typically sold in traditional supermarkets, as well as a sufficient range of food and grocery products such that consumers can “purchase all of their weekly grocery requirements in a single shopping visit.”[53]

On the other hand, the consent orders excluded club stores—such as Costco, Sam’s Club, and BJ’s Wholesale Club—as well as “hard discounters, limited assortment stores, natural and organic markets, [and] ethnic specialty stores.”[54] The orders claim that these stores are excluded from the relevant product market because “they offer a more limited range of products and services than supermarkets and because they appeal to a distinct customer type.”[55] In addition, the orders indicate that “supermarkets do not view them as providing as significant or close competition as traditional supermarkets.”[56] Prior consent orders omitted any discussion of whether online retailers or delivery services should be included or excluded from the relevant market. This product-market definition has remained mostly unchanged—and mostly unchallenged—since the Ahold/Giant merger a quarter-century ago.[57]

Figure 1 shows that retail sales by supermarkets, warehouse clubs, supercenters, and other grocery stores have been relatively stable at 5-6% of U.S. gross domestic product (GDP).[58] Figure 2 shows that supermarkets’ share of retail sales dropped sharply from the early 1990s through the mid-2000s, with that share shifting to warehouse clubs and supercenters. These figures are consistent with the conclusion that warehouse clubs and supercenters successfully compete against traditional grocery stores. Indeed, it would be reasonable to conclude that the rise of warehouse clubs and supercenters at the expense of traditional supermarkets is one of the most significant long-run trends in retail.

The retail food and grocery industry has changed dramatically. In particular, a great deal of attention has been paid to consolidation in the industry. Lina Khan & Sandeep Vaheesan, for example, note that “[t]he share of groceries sold by the four biggest food retailers more than doubled between 1997 and 2009, from seventeen percent in 1994 to twenty-eight percent in 1999 and thirty-four percent in 2004.”[59]

Below, we note that the average consumer shops for food and groceries more than once a week and shops at more than one retail format in a given week. Competition in groceries is not just between supermarkets. While Khan & Vaheesan recognize that supermarkets started to compete with warehouse clubs and supercenters, they fail to update their market definition to reflect that competition.[60] Over the period of increasing concentration within groceries, warehouse clubs and supercenters were steadily eroding supermarkets’ share of retail sales.

Figure 2 shows that, in 1994, supermarkets accounted for 81% of retail sales, which fell to 61% by 2004.[61] Over the same period warehouse clubs and supercenters grew from 14% of retail sales to 35%. In 2021, supermarkets accounted for 56% of retail sales and warehouse clubs and supercenters accounted for 42%. Since the Ahold/Giant merger in 1998, warehouse clubs and supercenters have doubled their share of retail sales, while supermarkets’ share has dropped by more than 25%.[62] Put simply, the four largest supermarket retailers were occupying a larger share of a shrinking segment.

Based on these observations, the product-market definition that the FTC has employed in its consent orders over the past more than two decades is likely to be—and should be—challenged to include warehouse clubs, in addition to accounting for online retail and delivery.

B.     The Once-a-Week Shopper Is No Longer the Norm

In the past, the FTC has specified that, for a firm to be in the relevant market of “supermarkets,” it must be able to “enable[e] consumers to purchase substantially all of their weekly food and grocery shopping requirements in a single shopping visit.”[63] This definition suffers from several deficiencies.

The first deficiency is that this hypothetical consumer behavior is at odds with how many or most consumers behave today.

  • Surveys conducted by the Food Marketing Institute and The Hartman Group report the average shopper makes 1.6 weekly trips to buy groceries.[64]
  • Earlier research by FMI and Hartman find, in addition, other household members take another 0.6 trips, implying that the total number of trips per household each week is about 2.2, or approximately one trip every three days.[65]
  • Surveys conducted by Drive Research show the average household makes an average of 8.1 grocery shopping trips a month, or around two trips a week F(Table 1).[66]

There is no evidence that consumers view retailers that provide one-stop shopping for an entire week’s food and grocery needs as distinct from other retailers who provide food and groceries. In fact, evidence suggests that many consumers “multi-home” across several different retail categories.

  • Survey data published by Drive Research indicate that many households spread their shopping across grocery stores, mass merchants, warehouse clubs, independent grocery stores, natural and specialty grocery stores, dollar stores, and online retailers (Table 1).[67]
  • Acosta, a sales and marketing consulting firm, reports that 76% of consumers shop at more than one retailer a week and about one third “retail hop” among three or more retailers a week for groceries and staples.[68]
  • Research from the University of Florida found that, in 2017, an average consumer visited 3.2 to 4.3 different formats of food outlets a month, depending on income level.[69]
  • Survey results from PYMNTS, a data-analysis firm, demonstrate a significant shift in consumer spending away from traditional supermarkets. For example, in 2020, approximately 98% of consumers who bought at least one common household product each week—“center aisle” products, such as paper towels, cleaning supplies, and canned goods—made the purchase at a grocery store. In 2023, more than a third of consumers (37%) say they purchase none of these products from a traditional grocery store, with online purchases accounting for much of the shift.[70]

Thus, while one-stop weekly food and grocery shopping at single retailers was once typical, the evidence indicates such a phenomenon is much less common today.

C.     Supermarkets Compete with Warehouse Clubs

The FTC’s earlier consent orders provide four reasons to exclude warehouse clubs from the relevant market that includes supermarkets:

  1. They offer a “more limited range of products and services” than supermarkets;
  2. They “appeal to a distinct customer type” from supermarket customers;
  3. Shoppers do not view warehouse clubs as “adequate substitutes for supermarkets;” and
  4. Supermarkets do not view warehouse clubs as “significant or close competition,” relative to other supermarkets.[71]

In contrast to these conclusions, there is widespread recognition today that warehouse clubs impose significant competitive pressure on supermarkets.[72]

  • The National Academies of Sciences concludes that, over time, the entry and growth of warehouse clubs, superstores, and online retail has “blurred” the distinctions among retail formats.[73] More importantly for merger-review analysis, the National Academies concludes that the retail sector is “highly competitive,” in part because of the entry and growth of warehouse clubs, superstores, and online retail.[74]
  • Based on their empirical analysis, Paul Ellickson and coauthors conclude that warehouse clubs are “relevant substitutes” for supermarkets, even when the club stores are outside the geographic area typically used by the FTC in merger reviews.3[75]
  • Prior to her appointment as FTC chair, Lina Khan and her coauthor concluded that competition from warehouse clubs “fueled” grocery mergers in the late 1990s.[76]

The FTC’s consent orders note that warehouse clubs offer a “more limited range of products and services” than supermarkets. The orders identify products sold at supermarket as “including, but not limited to, fresh meat, dairy products, frozen foods, beverages, bakery goods, dry groceries, detergents, and health and beauty products.”[77]

The annual reports for Costco, Walmart (which include Sam’s Club), and BJ’s, however, indicate that each company offers the same range of products the FTC consent orders identify as being offered by supermarkets.[78] Indeed, the reports indicate that supermarkets do consider wholesale clubs to be competitors and vice versa. BJ’s annual report goes to great lengths to explain its efforts to compete with supermarkets by offering similar manufacturer-branded products at lower prices.[79] The role of warehouse clubs cannot be assumed or asserted away. Indeed, any identification of the relevant market for retail food and grocery sales should begin with a presumption that warehouse clubs provide competitive pressure, since that is what the economics research finds.

It is true that warehouse stores typically carry fewer stock-keeping units (“SKUs”) than supermarkets. Costco’s annual report indicates the company carries fewer than 4,000 SKUs in its warehouse stores and offers 10,000 to 11,000 SKUs for online purchases.[80] BJ’s annual report says that the company carries “approximately 7,000 core active stock keeping units.”[81] In contrast, the BJ’s report notes, “supermarkets normally carry an average of 40,000 SKUs, and supercenters may stock 100,000 SKUs or more.” But these differences in SKU counts do not in themselves demonstrate fundamental differences in product offerings between the two store formats.[82]

The main difference between the two formats is that warehouse clubs tend to offer a smaller range of sizes and packaging and a smaller variety of brands. For example, Costco notes that many of the products it stocks “are offered for sale in case, carton, or multiple-pack quantities only.”[83] Another difference can be observed in the variety of brands offered. For example, Costco has invested in developing its Costco Wholesale and Kirkland Signature private labels.[84] In contrast, CFRA Research analyst Arun Sundaram observes that Sam’s Club “typically offers a bigger national-brand product selection than its club cohorts.”[85] Such differences are insufficient to exclude wholesale clubs from the relevant market. If warehouse clubs offer similar products, see themselves as competing with supermarkets, and customers view them as substitutes, warehouse clubs must be in the same market.

D.    E-Commerce Has Changed the Food Landscape

In the years since the Ahold/Giant merger in 1998, online shopping and home delivery have grown from niche services serving only 10,000 households nationwide to a landscape where approximately one-in-eight consumers purchase groceries “exclusively” or “mostly” online.[86] This shift has increased competition and innovation in the supermarket industry, as traditional supermarkets have adapted to changing consumer preferences and behaviors by offering more delivery and pickup options, expanding their online assortments, and enhancing their digital capabilities.[87] Some have invested in their own e-commerce platforms and many have partnered with such third-party providers as Instacart, Shipt, and Peapod.[88]

One might surmise that e-commerce simply replaced in-person shopping, but with the same stores competing. This is not what has been observed. E-commerce has also increased competition by bringing in new companies with which traditional stores need to compete (e.g., Amazon) and by increasing the options available to consumers through services like Instacart, which allow for direct price and product comparisons among many stores. Each of these innovations has blurred the lines between brick-and-mortar food and grocery retailers and e-commerce, as well as the lines between supermarkets and other retail formats.

In 2023, more than a third of consumers (37%) say they purchase no center-aisle products (products such as paper towels, cleaning supplies, and canned goods) from a traditional grocery store, with online purchases accounting for much of the shift.[89] The fluidity between supermarkets, grocery stores, and online purchases makes the distinction nearly meaningless.

If a consumer uses Instacart to purchase and deliver groceries from Safeway, is that a supermarket purchase or e-commerce? What if the same consumer uses Instacart to purchase and deliver the same goods from Costco or Target? Does the consumer care which retailer the food and groceries came from when competition is just a click away? Indeed, the National Academies of Sciences concludes it is “often impossible” to distinguish between brick-and-mortar retail sales and e-commerce.[90]

E-commerce and club stores also matter for defining the geographic market. Past FTC consent orders have defined the relevant geographic markets to be areas that range from a two- to ten-mile radius around each of the merging parties’ supermarkets. We conclude that, because club stores have a larger catchment area than supermarkets, it would be reasonable to expand the relevant geographic market in localities where club stores are present.

Combined, the rapid growth of e-commerce and delivery services make distance, traffic patterns, and population density decreasingly relevant in a consumer’s choice of where they purchase food and groceries. Dimitropoulos and coauthors note (1) the presence of warehouse clubs expands the relevant geographic market, (2) online-delivery options expand the geographic market “far away,” and (3) online food and grocery purchases can be delivered from fulfillment centers, as well as from traditional stores.[91]

Because of these observations, the product market-definition that has been employed in the FTC’s consent orders over the past more than two decades is likely to be—and should be—challenged and should be revised to include warehouse clubs and to account for online retail and delivery.

IV.   The Merger Is Unlikely to Increase Labor Monopsony Power

In recent years, there has been an increasing emphasis in antitrust discussions on labor markets and potential harms to workers. The recent draft merger guidelines added an explicit section on mergers that “May Substantially Lessen Competition for Workers or Other Sellers.”[92] Even before the guidelines’ publication, some observers predicted that the FTC was set to push a case on labor competition.[93] While, in theory, antitrust harms can occur in labor markets, just as in product markets, proving that harm is more difficult.

An important fact about the proposed Kroger/Albertsons merger is that both companies have many unionized workers. Around two-thirds of Kroger employees[94] and a majority of Albertsons employees[95] are part of the United Food and Commercial Workers International Union (UFCW), which represents 1.3 million members. Even if the merger would increase labor monopsony power in the absence of unions, the FTC will have to acknowledge the reality of the unions’ own bargaining power.

Delegates of the UFCW unanimously voted to oppose the merger[96] Rather than monopsony power or lower wages, however, the union’s stated reason for their opposition was lack of transparency.[97] While lack of transparency may be problematic for the UFCW members, it does not constitute an antitrust harm. Kroger, for its part, has contended that the merger will benefit employees, citing a commitment to invest an additional $1 billion toward increased wages and expanded benefits, starting from the day the deal closes.[98] Albertsons claims that no store closures or frontline associate layoffs will result from the transaction and that the merger will “secure the long-term future of union jobs by establishing a more competitive alternative to large, non-union retailers.”[99]?As with most announced goals, however, there is no enforcement mechanism for this commitment at present, although one could be litigated.

Rather than relying on proclamations from any of the parties, we need economic analysis of the relevant labor markets, asking the types of questions raised above regarding the output markets. A policy report from Economic Policy Institute estimates that “workers stand to lose over $300 million annually” from the merger,[100] but the report arrives at that estimate by using an estimate of the correlation between concentration (HHI) in labor markets and wages. While academic research may benefit from such an estimate, it is unhelpful in merger analysis. As a long list of prominent antitrust economists recently wrote, “regressions of price on HHI should not be used in merger review… [A] regression of price on the HHI does not recover a causal effect that could inform the likely competitive effects of a merger.”[101]

While HHI regressions are of little practical help in this context, according to standard economic theory, it is possible that the average worker would be harmed for traditional labor-monopsony reasons. It is, however, more difficult to identify anticompetitive labor-market harms than to identify analogous product-market harms. For the product market, if the merger simply enhanced monopoly power without producing efficiency gains, the quantity sold would decrease, either because the merging parties raise prices or because quality declines. A merger that creates monopsony power will necessarily reduce the prices and quantity purchased of inputs like labor and materials. A strong union could counteract a firm’s monopsony power to some extent, by collectively advocating for higher wages, fewer layoffs, and other worker benefits. Indeed, obtaining and exerting labor market power is a union’s raison d’être.

Of course, the presence of a unionized workforce does not render monopsony impossible; unions’ ability to offset the effects of a monopsony or monopoly may also be limited, and monopsony power could increase under a merger, even with unions. Still, the existence of union bargaining power makes any monopsony case more difficult and is an important factor to consider in evaluating a merger’s likely labor-market effects—particularly in this case, given the high rates of union membership at both companies.

Moreover, the FTC needs to be careful with any labor case. For labor markets, a decline in the number of workers employed (which harms the workers) may not be anticompetitive. The reduction in input purchases may be because of efficiency gains.[102] For example, if two merging hospitals integrate their information-technology resources, therefore requiring fewer overlapping workers, the merged firm will employ fewer IT workers. This may even reduce the wages of specialized IT workers, even if the newly merged hospital does not exercise any market power to suppress wages.

The same applies for any inputs from an efficiency-enhancing merger: inputs may decrease. But using fewer inputs is not an antitrust harm. The key point is that monopsony needs to be treated differently than monopoly. The antitrust agencies cannot simply look at the quantity of inputs purchased in the monopsony case as the flip side of the quantity sold in the monopoly case, because the efficiency-enhancing merger can look like the monopsony merger in terms of the level of inputs purchased.

Another difficulty with pursuing a labor monopsony case is that the usual antitrust tools, such as merger simulation, cannot be easily applied to the labor market. Unlike the DOJ’s recent success in blocking Penguin-Random House from merging with Simon & Schuster on grounds that the merger would hurt authors with advances above $250,000,[103] the labor market for most employees of Kroger and Albertsons is much larger than those two companies, or even the largest definition of grocery stores. Indeed, it cannot be narrowed down to a handful of companies.

Any labor case would require showing that the merger would harm workers by reducing their bargaining power. For most workers involved, there are still many potential employers competing. One relevant piece of evidence for this is that press releases Kroger issued during the COVID-19 pandemic highlighted that the company was hiring workers from a wide variety of firms and industries—from hospitality (Marriott International) to restaurants (Waffle House) to food distribution (Sysco).[104] While we are not aware of publicly available data that would more comprehensively illustrated worker flows among different companies, such flows of retail workers into and out of roughly adjacent labor markets makes intuitive sense. As economist Kevin Murphy has explained:

If you look at where people go when they leave a firm or where people come from when they go to the firm, often very diffuse. People go many, many different places. If you look at employer data and you ask where do people go when they leave, often you’ll find no more than 5 percent of them go to any one firm, that they go all over the place. And some go in the same industry. Some go in other industries. Some change occupations. Some don’t. You look at plant closings, where people go. Again, not so often a big concentration of where they go to. If you look at data on where people are hired from, you see much the same patterns. That’s kind of a much more diffuse nature.[105]

If, as is likely, an overwhelming majority of Kroger’s workers’ next best option (what they would do if a store closed) was not an Albertsons store, but something completely outside of the market for grocery-store labor (or even outside the retail-food industry more broadly), the merger would not take away those workers’ next best option. If true, the merger cannot be said to increase labor monopsony power to the extent necessary to justify blocking a merger.

V.     ‘Waterbed Effects’ Are Highly Speculative

One potential antitrust harm that has been discussed frequently in recent years is the so-called “waterbed effect,” in which “a large and powerful firm improves its own terms of supply by exercising its bargaining power, [but] the terms of its competitors can deteriorate sufficiently so as ultimately to increase average retail prices and, thereby, reduce total consumer surplus.”[106] The waterbed effect is not unique to mergers, but can apply any time there is differential buyer-market power. The firm with more market power gets a better deal from suppliers and its competitors are ostensibly harmed because they cannot get the same deal. Long before the proposed merger, but still in the context of retail, people were speculating about a waterbed effect regarding Walmart.[107]

In the context of the Kroger/Albertsons merger, critics have again raised the possibility of a waterbed effect. Michael Needler Jr.—the president and chief executive of Fresh Encounter, a chain of 98 grocery stores based in Findlay, Ohio—raised the possibility in a U.S. Senate hearing on the merger.[108] He was also quoted by The New York Times as saying:

When the large power buyers demand full orders, on time and at the lowest cost, it effectively causes the water-bed effect. They push down, and the consumer packaged goods companies have no option but to supply them at their demands, leaving rural stores with higher costs and less availability to products.[109]

In a letter to the FTC, the American Antitrust Institute raised several concerns about the merger, arguing that:

The waterbed effect is likely to worsen with Kroger-Albertsons enhanced buyer power post-merger, with adverse effects on the ability of independent grocers to compete in a tighter oligopoly of large grocery chains.[110]

The implied argument in this version of the waterbed effect goes as follows: A merged Kroger and Albertsons would have additional market power over some of its suppliers. It could then exercise that market power to extract discounts from those suppliers, which would be unavailable to its competitors. The merged firm could then pass those cost savings on to consumers in the form of lower retail prices, thereby increasing Kroger/Albertsons’ sales. Some of these sales would come at the expense of smaller competitors, who could no longer compete on price. And because of these reduced sales, they would purchase less from their suppliers, further eroding their bargaining power with suppliers. Ultimately, consumers of the smaller retailers may face higher retail prices. Thus, under this theory, consumers of the merged firm would pay lower retail prices, while consumers of the smaller retailers would pay more.

Even if all of that were true, however, what remains unknown (and unaddressed under this argument) is whether consumers as a whole would be better or worse off. That, of course, is precisely the result that would be required to establish harm under antitrust law.

Roman Inderst & Tommaso Valletti are credited with the first formal theoretical model of a waterbed effect and how it could potentially harm consumers as a whole (as opposed to merely certain competitors or the subset of consumers who continue to buy from them).[111] In order to establish consumer harm under their model, three key assumptions must be met:

  1. Retailers buy their inputs from a supplier who can price discriminate among retailers;
  2. Retailers can access an alternative source of supply, but must incur a fixed switching cost, which is the same across all retailers; and
  3. Retailers compete on price.

Because a larger retailer can spread the fixed switching costs across more units, its per-unit costs will be lower. This provides the larger retailer with a better bargaining position with its suppliers to extract lower input prices. If the large retailer reduces its prices to consumers, the reduced sales to smaller competing retailers results in those competitors having higher per-unit switching costs, thus reducing their ability to change suppliers, reducing their bargaining power with the initial supplier, and increasing the price they pay to the supplier for inputs.

While the model shows how the effect is possible and that it could harm consumers, it does not imply that the waterbed effect necessarily harms consumers. In fact, the same waterbed effect would also occur if a merger generated efficiency gains (as the authors point out), but with considerably different welfare and antitrust implications. Setting aside mergers, in Inderst & Valletti’s model, if one firm discovers a cheaper importer, for example, it would give that firm more buyer power, because it presents a more credible threat of leaving for a competitor. Recognizing that the firm has a better “outside option,” the wholesaler offers better terms. This, too, generates a waterbed effect, but it is clearly pro-competitive, as it would help consumers. Unless we are willing to declare finding another source of supply to be anticompetitive, we should be hesitant about jumping to the conclusion that the waterbed effect is anticompetitive.

Inderst & Valletti’s model also demonstrates that, with relatively low supplier-switching costs, the supplier has little scope to price discriminate among retailers. As a result, “any further growth of the large buyer… will reduce all retail prices.”[112] This is true even in the presence of a waterbed effect. Thus, for a waterbed effect to result in higher average retail prices for consumers, the large retailer’s buying advantage must be “sufficiently larger in size,” and smaller retailers must face much high switching costs, with those switching costs serving as the reason why the supplier can effectively price discriminate across the retailers.[113] For many products, this simply won’t be the case.

A competition authority that pursued a waterbed theory to block a merger must first demonstrate that a waterbed effect exists. Because each product sold in a food and grocery retailer may have its own idiosyncratic manufacturing, wholesale, and distribution characteristics, this evaluation likely must be conducted on a product-by-product basis. Then, for each market, the authority must evaluate the suppliers’ abilities to price discriminate (which could raise additional antitrust issues). Last, the authority must evaluate competing firms’ anticipated price response to any identified waterbed effect. While Inderst & Valletti provide a seemingly straightforward theoretical approach to evaluating allegations of a waterbed effect, applying their model to the real world of food and grocery mergers would likely amount to a costly and time-consuming wild goose chase.

That is likely why finding empirical demonstration of a waterbed effect has been so elusive. Indeed, we are not aware of any empirical literature indicating the existence of a waterbed effect in retail markets, let alone any evidence of consumer harms from such an effect.[114] Indeed, UK competition authorities have been unconvinced of any waterbed effects in the food and grocery mergers in which the issue has been raised. In 2006, the UK Office of Fair Trading concluded:

[T]here are theoretical questions that would need to be resolved before concluding that the price differentials observed are evidence of a waterbed effect. For example, it is not clear how suppliers would be able to charge significantly above cost to smaller retailers without rivals undercutting them in the market; similarly, it is not clear why suppliers would price persistently below cost to the large supermarkets.[115]

In the United States, only one district court has issued an opinion on the waterbed effect. In DeHoog, consumers sued to block AB InBev’s acquisition of SABMiller.[116] The consumers alleged that the merged firm would be a “powerful buyer” that “demands lower prices or other concessions from its suppliers, causing the supplier to, in turn, increase prices to smaller buyers.” The district court rejected the consumers’ waterbed claim because (1) the alleged harm was to competing brewers, not to consumers, and (2) competing brewers could switch to different hops, thereby avoiding any waterbed effect.

DeHoog highlights the high hurdles an antitrust authority or private plaintiff would need to clear in order to successfully allege a waterbed effect. Challengers must demonstrate that switching costs are insurmountably high and that a waterbed effect exists. They must demonstrate then that the waterbed effect harms consumers, rather than competitors.

Demonstrating a waterbed effect in the Kroger/Albertsons merger may be especially challenging. Although the notion has been invoked by several critics of the merger, we are not aware of any specific product or product category in which a potential waterbed effect has been alleged.[117] If the FTC chooses to pursue the waterbed-effect theory, it must identify the relevant products that would be subject to the effect and demonstrate the anticipated consumer harm associated with it. If the agency relies on the waterbed effect in an effort to block the Kroger/Albertsons merger, then it would be reasonable to conclude that its “traditional” claims of horizontal market power are especially weak.

VI.   Remedies Can Resolve Any Remaining Competitive Concerns

While the above sections argue that the FTC will (and should) have a hard time making a case that the Kroger/Albertsons merger is overall anticompetitive, there may be some specific geographic markets where concerns remain. In the face of such concerns, the FTC historically has allowed most supermarket transactions to proceed with divestitures, such as Ahold/Delhaize (81 stores divested), Albertsons/Safeway (168 stores), and Price Chopper/Tops (12 stores).[118] The extent of the remedies sought depends on the extent of post-merger competition in the relevant markets, as well as the likelihood of entry by additional competitors.[119] Dimitropoulos and coauthors have noted that most divestitures required by consent orders in recent supermarket mergers have occurred in geographic markets with fewer than five remaining competitors.[120]

There is good reason (and a long history of examples in previous grocery-merger settlements) to think that targeted divestitures in certain markets—as have been proposed from the start of this process by the merging parties[121]—should be sufficient to address any geographic-market-specific concerns that may arise.

One reason that divestiture—instead of outright blocking—should be appropriate in this case is that the vast majority of Kroger and Albertsons stores are in geographic markets where the other is not located (Figure 3). As such, there is no antitrust concerns from a product-market perspective. The merger does not affect competition in the South (where Kroger is focused) or in the Northeast (where Albertsons is focused). In these regions, the merger generates all of the efficiencies without the possible downside of a loss to competition.

In some other geographic locations, however, the companies do currently compete, and antitrust concerns could therefore arise. This is where divestiture comes in. By most measures, there appear to be some 1,400 overlapping stores; resolving this overlap entails divestiture of no more than 700, or 14% of the two companies’ more than 5,000 stores.[122] By the same token, only a limited number of geographic markets have Kroger and Albertsons stores in close proximity, suggesting that targeted divestitures could address those concerns, while allowing the merger to proceed unimpeded in the great majority of markets.[123]

Previous remedies sought by the FTC in merger cases have generally been successful in achieving their goals. The FTC’s most recent merger-remedies study, covering 89 orders from 2006-2012, provides additional support for the feasibility of divestitures as an effective remedy.[124] The study found that the vast majority of divestitures succeeded in maintaining competition in the affected markets. All remedies involving divestiture of an ongoing business were successful. Divestitures of more limited “selected assets” also largely succeeded, although at a lower rate. Overall, the FTC concluded that more than 80% of the orders examined achieved the goal of maintaining or restoring competition post-merger.[125]

Nonetheless, despite a long history of divestitures serving as an appropriate and adequate remedy in food-retail mergers, some advocates for stronger antitrust are extremely skeptical of divestiture remedies. As authors from the American Economic Liberties Project and the Open Markets Institute put it in one recent article:

It should not fall on our overburdened antitrust enforcers to pore over the individual assets changing hands in service of coming up with a carve-out that somehow brings a merger into technical compliance with an arbitrary Reaganite standard devised by bad-faith ideologues.[126]

Such concerns are leveled at the grocery industry, in particular, with critics consistently pointing to the Albertsons/Safeway merger divestitures to Haggen as evidence that, in this industry (if not elsewhere), divestiture is no longer an appropriate merger remedy.[127] But these arguments ring hollow. Several factors idiosyncratic to Haggen and its acquisition strategy led to that divestiture’s failure.

A.     Distinguishing the 2014 Haggen Divestiture

In 2014, the parent company of Albertsons announced plans to purchase rival food and grocery chain Safeway for $9.4 billion.[128] Prior to the merger, Albertsons was the fifth-largest grocer in the United States and operated approximately 1,075 supermarkets in 29 U.S. states. At the time, Safeway was the second-largest and operated more than 1,300 stores nationwide.[129] During its merger review, the FTC identified 130 local markets in Western and Midwestern states where it alleged the merger would be anticompetitive.[130] In response, Albertsons and Safeway agreed to divest 168 supermarkets in those geographic markets.[131] Haggen Holdings LLC was the largest buyer of the divested stores, acquiring 146 Albertsons and Safeway stores in Arizona, California, Nevada, Oregon, and Washington.

Following the acquisition, Haggen almost immediately encountered numerous problems at the converted stores. Consumers complained of high prices, and sales plummeted at some stores. The company struggled and began selling some of its stores. Less than a year after the FTC announced the divestiture agreement, Haggen filed for bankruptcy. Following the bankruptcy, Albertsons bought back 33 of the stores it had divested in its merger with Safeway.

In retrospect, Haggen may not have been an appropriate buyer for the divested stores. Before acquiring the divested stores, Haggen was a small regional chain with only 18 stores, mostly in Washington State. The acquisition represented a tenfold increase in the number of stores the company would operate. While Haggen was once an independent, family-owned firm, at the time of the acquisition, the company was owned by a private investment firm that used a sale-leaseback scheme to finance the purchase. Christopher Wetzel notes that Haggen failed to invest sufficiently in the marketing necessary to create brand awareness in regions where Haggen had not previously operated.[132] Such issues would need to be avoided in any future divestitures, and experience shows they can be.

Around the same time that it filed for bankruptcy, Haggen also filed a lawsuit against Albertsons in federal district court, arguing that Albertsons engaged in “coordinated and systematic efforts to eliminate competition and Haggen as a viable competitor.”[133] Haggen claimed that Albertsons made false representations to both Haggen and the FTC about its commitment to providing a smooth transition that would allow Haggen to be a viable competitor. Among other allegations, Haggen claimed that Albertsons overstocked the divested stores with perishable meat and produce, provided inaccurate and misleading price information that led to inflated prices, and failed to perform maintenance on stores and equipment.

None of these claims were demonstrated, as the matter settled months after the complaint was filed. Even so, FTC consent orders typically provide asset-maintenance agreements to address the kinds of issues raised by Haggen. David Balto reports that, after the 1995 Schnucks/National merger, the FTC sued Schnucks, alleging that it had violated a provision of the asset-maintenance agreement in the consent order.[134] The suit resulted in a settlement in which Schnucks paid a $3 million civil penalty and was required to divest two additional properties. These two properties were stores that had been closed by Schnucks, but that presumably could be reopened by a new buyer.

The problems with the Haggen divestiture need not be repeated. In particular, there are many companies of various sizes that have the capabilities and desire to expand. In recent merger-consent orders, divested stores have been acquired by both retail supermarkets and wholesalers with retail outlets, including Publix, Supervalu, Big Y, Weis, Associated Wholesale Grocers, Associated Food Stores, and C&S Wholesale Grocers.[135] Several of these companies have successfully expanded—in some cases outside of their “home” territories. For example, Publix is a Florida-based chain that operates nearly 1,350 stores in seven southeastern states.[136] Publix expanded to North Carolina in 2014, Virginia in 2017, and has announced plans to expand into Kentucky this year.[137] Weis Markets is a Pennsylvania-based chain that operates more than 200 stores in seven northeastern states.[138] Last year, the company announced plans to spend more than $150 million on projects, including new retail locations and upgrades to existing facilities.[139] And Rochester, New York-based Wegmans has successfully entered Delaware, Virginia, and the District of Columbia in recent years.[140]

While the relevant product and geographic markets for supermarket mergers has shifted enormously over the past few decades, divestitures remain an appropriate and adequate remedy for any competitive concerns. The FTC has knowledge and experience with divestiture remedies and should have a good understanding of what works. In particular—and, perhaps, unlike in the Haggen example—firms acquiring divested assets should have an adequate cushion of capital, experience with the market conditions in which the stores are located, and the operational and marketing expertise to transition customers through the change.

B.     Proposed C&S Divestiture

As noted, Kroger and Albertsons have contemplated divestitures from the beginning, even including a provision in their merger agreement preemptively agreeing to divest up to 650 stores.[141] More recently, however, the companies have made their divestiture plans more concrete. In September 2023, the companies presented a proposal (both publicly and to the FTC) proposing to divest 413 stores, eight distribution centers, and three store brands to C&S Wholesale Grocers for $1.9 billion.[142] The agreement also allows C&S to purchase up to 237 additional stores if needed to resolve antitrust concerns. C&S also has committed to maintain any existing collective-bargaining agreements with labor unions.[143]

The specific characteristics of the proposed buyer of the divested stores suggests that it is unlikely to fall prey to the limitations that scuttled the Haggen divestiture. Unlike Haggen, the purchasing party here has experience operating more than 160 supermarkets under brands like Grand Union. This existing operation of stores makes C&S better positioned as a buyer than Haggen was when it attempted to rapidly expand from 18 to 168 stores.[144]

While C&S is primarily a wholesaler, its Grand Union retail operations and the transition support offered under the divestiture agreement should position it to successfully operate the divested stores. In that way, the divestiture does not just spin off or increase the size of a horizontal competitor. Rather, the plan jumpstarts greater vertical integration by C&S, whose wholesale operations include the production of private-label products.

Indeed, by enabling C&S to take better advantage of the benefits of vertical integration, the divestiture appears to ensure that C&S will emerge with a structure more in line with the rest of the food-retail market. Over the past decade, many retailers (including Kroger and Albertsons) have shifted toward “bringing private label production in-house.”[145] This move by firms (even without any market power) likely reflects competitive advantages gained from vertical integration.

The targeted nature of the divestiture would allow the merger to proceed in the majority of geographic markets where there are no competitive concerns between Kroger and Albertsons. Divesting stores only where localized overlaps in specific regions exist enables the realization of efficiencies and benefits in the many remaining markets. The FTC will still need to closely scrutinize the buyer and the proposed divestiture package. But the announced plan demonstrates that the merging parties are taking seriously the need for divestitures.

Of course, as with any complex business transaction, there is always some possibility that aspects of a divestiture may not fully go according to plan. The recent piece by Maureen Tkacik & Claire Kelloway throws out many of these possibilities.[146] It’s possible that C&S turns out to not want to run grocery stores but only wants to resell the properties. It’s possible that C&S will be unable to afford the leases. Regulators and merging parties alike operate under inherent uncertainty when predicting competitive outcomes. Antitrust analysis does not deal with certainties, but rather with probabilistic assessments of likely competitive effects.

The relevant question is whether the divestiture is likely to effectively maintain competition in the markets of concern, not whether it can be guaranteed to perfectly do so in all scenarios. When we take more episodes than Haggen’s into account, despite the uncertainty, the FTC’s experience shows that targeted divestitures with an experienced buyer are likely to adequately protect consumers post-merger. The possibility that some unforeseen complication may arise does not negate the high probability that competition will be preserved. Antitrust regulation requires reasonable predictive judgments, acknowledging that business transactions inherently carry risks.

With the FTC’s knowledge of the industry and of its own past successes and failures, divestitures remain an appropriate and adequate remedy for this merger. The parties appear committed to working cooperatively with regulators to craft divestitures that fully resolve competitive concerns. Rather than blocking the deal outright, the FTC can allow the merger to proceed, conditioned on acceptable divestitures that protect consumers, while permitting efficiency gains across the majority of stores.

 

[1] Press Release, Kroger and Albertsons Companies Announce Definitive Merger Agreement, Kroger (Oct. 14, 2022), https://ir.kroger.com/CorporateProfile/press-releases/press-release/2022/Kroger-and-Albertsons-Companies-Announce-Definitive-Merger-Agreement/default.aspx.

[2] In an article written with Sandeep Vaheesan before she became chair of the FTC, Lina Khan expressed disdain for grocery-industry consolidation and deep skepticism of even the best divestiture packages. See Lina Khan & Sandeep Vaheesan, Market Power and Inequality: The Antitrust Counterrevolution and Its Discontents, 11 Harv. L. & Pol’y Rev. 235, 254 (2017) (“Retail consolidation has enabled firms to squeeze their suppliers… and led to worse outcomes for consumers.”) & 289 (“Even if divestitures could be perfectly tailored and if they preserved competition in narrow markets in every instance, they would fail to advance the citizen interest standard.”).

[3] See, e.g., David Dayen, Proposed Kroger-Albertsons Merger Would Create a Grocery Giant, The American Prospect (Oct. 17, 2022), https://prospect.org/power/proposed-kroger-albertsons-merger-would-create-grocery-giant; Richard Smoley, Kroger, Albertsons, and Lina Khan, Blue Book Services (May 2, 2023), https://www.producebluebook.com/2023/05/02/kroger-albertsons-and-lina-khan.

[4] U.S. Dep’t of Justice & Fed. Trade Comm’n, Draft Merger Guidelines (Jul. 19, 2023), available at https://www.justice.gov/d9/2023-07/2023-draft-merger-guidelines_0.pdf. See also Gus Hurwitz & Geoffrey Manne, Antitrust Regulation by Intimidation, Wall St. J. (Jul. 24, 2023), https://www.wsj.com/articles/antitrust-regulation-by-intimidation-khan-kanter-case-law-courts-merger-27f610d9.

[5] Prior to Whole Foods/Wild Oats, the last litigated supermarket merger was the State of California’s 1988 challenge to American Store’s acquisition of Lucky Stores. Several retail mergers have been challenged in court, however, such as Staples/Office Depot in 2015. See Press Release, FTC Challenges Proposed Merger of Staples, Inc. and Office Depot, Inc., Federal Trade Commission (Dec. 7, 2015), https://www.ftc.gov/news-events/news/press-releases/2015/12/ftc-challenges-proposed-merger-staples-inc-office-depot-inc.

[6] This includes approving Albertsons/Safeway (2015), Ahold/Delhaize (2016), and Price Chopper/Tops (2022). See Analysis of Agreement Containing Consent Order to Aid Public Comment, In the Matter of Cerberus Institutional Partners V, L.P., AB Acquisition, LLC, and Safeway Inc. (File No. 141 0108) (Jan. 27, 2015) available at https://www.ftc.gov/system/files/documents/cases/150127cereberusfrn.pdf; Analysis of Agreement Containing Consent Order to Aid Public Comment, In the Matter of Koninklijke Ahold N.V. and Delhaize Group NV/SA (File No. 151-0175) (Jul. 22, 2016), available at https://www.ftc.gov/system/files/documents/cases/160722koninklijkeanalysis.pdf; Analysis of Agreement Containing Consent Order to Aid Public Comment, In the Matter of The Golub Corporation and Tops Markets Corporation (File No. 211-0002, Docket No. C-4753) (Nov. 8, 2021), available at https://www.ftc.gov/system/files/documents/cases/2110002pricechoppertopsaapc.pdf.

[7] Decision and Order, In the Matter of Whole Foods Market, Inc., (Docket No. 9324) (May 28, 2009), available at https://www.ftc.gov/sites/default/files/documents/cases/2009/05/090529wfdo.pdf; FTC v. Whole Foods Market, 548 F.3d 1028 (D.C. Cir. 2008).

[8] Number based on authors’ calculations, using data from 90th Annual Report, Progressive Grocer (May 2023), https://progressivegrocer.com/crossroads-progressive-grocers-90th-annual-report.

[9] See Draft Merger Guidelines, supra note 4.

[10] FTC v. Meta Platforms Inc., U.S. Dist. LEXIS 29832 (2023); FTC v. Microsoft Corporation et al., No. 23-cv-02880-JSC (N.D. Cal. Jul. 10, 2023), available at https://s3.documentcloud.org/documents/23870711/ftc-v-microsoft-preliminary-injunction-opinion.pdf.

[11] See, e.g., George Kuhn, Grocery Shopping Consumer Segmentation, Drive Research (2002), available at https://www.driveresearch.com/media/4725/final-2022-grocery-segmentation-report.pdf.

[12] See, e.g., In the Matter of Cerberus Institutional Partners, supra note 6, at 3.

[13] Press Release, Kroger and Albertsons Companies Announce Comprehensive Divestiture Plan with C&S Wholesale Grocers, LLC in Connection with Proposed Merger, The Kroger Co. (Sep. 8, 2023), https://www.prnewswire.com/news-releases/kroger-and-albertsons-companies-announce-comprehensive-divestiture-plan-with-cs-wholesale-grocers-llc-in-connection-with-proposed-merger-301921933.html.

[14] Id.

[15] See Press Release, supra note 1.

[16] Progressive Grocer, supra note 8.

[17] Who Are the Top 10 Grocers in the United States?, Foodindustry.com (last visited Oct. 10, 2023), https://www.foodindustry.com/articles/top-10-grocers-in-the-united-states-2019.

[18] Id.

[19] Number based on authors’ calculations, using data from Progressive Grocer Staff, 90th Annual Report, Progressive Grocer (May 2023), https://progressivegrocer.com/crossroads-progressive-grocers-90th-annual-report.

[20] Kroger, supra note 13.

[21] See Khan & Vaheesan, supra note 2.

[22] See Leah Nylen & Jeannette Neumann, California Preparing Lawsuit to Block Merger of Kroger, Jewel Parent, Bloomberg (Oct. 12, 2023), https://www.chicagobusiness.com/retail/california-preparing-lawsuit-block-kroger-albertsons-deal.

[23] Alexander Coolidge, Report: Kroger CEO Is “Committed to Litigate” If FTC Regulators Fight Albertsons Merger, Cincinnati Enquirer (May 11, 2023), https://www.cincinnati.com/story/money/2023/05/11/kroger-commited-to-litigate-if-ftc-blocks-albertsons-deal/70206692007.

[24] Press Release, FTC Challenges Proposed Merger of Staples, Inc. and Office Depot, Inc., Federal Trade Commission (Dec. 7, 2015), https://www.ftc.gov/news-events/news/press-releases/2015/12/ftc-challenges-proposed-merger-staples-inc-office-depot-inc.

[25] Jesse Greenspan, FTC To Challenge Whole Foods, Wild Oats Merger, Law360 (Jun. 5, 2007), https://www.law360.com/texas/articles/26191/ftc-to-challenge-whole-foods-wild-oats-merger.

[26] State of Cal. v. American Stores Co., 872 F.2d 837 (9th Cir. 1989) (granting preliminary injunction).

[27] See, e.g., In the Matter of Cerberus Institutional Partners, supra note 6, at 2-3.

[28] See, e.g., Food Marketing Institute & The Hartman Group, Consumers’ Weekly Grocery Shopping Trips in the United States from 2006 to 2022 (Average Weekly Trips per Household), Statista (May 2022), available at https://www.statista.com/statistics/251728/weekly-number-of-us-grocery-shopping-trips-per-household.

[29] See, e.g., In the Matter of Cerberus Institutional Partners, supra note 6, at 3.

[30] Safeway Merger Report, UK Competition Commission (2003), available at https://webarchive.nationalarchives.gov.uk/ukgwa/20120119163858/http:/www.competition-commission.org.uk/inquiries/completed/2003/safeway/index.htm (“Overall, therefore, there is little evidence of an immediate or short-term ‘waterbed’ effect. … [O]ur surveys produced insufficient evidence on this point for us to conclude that any waterbed effect would be exacerbated by any of the mergers.”); Anticipated Merger between J Sainsbury PLC and ASDA Group Ltd: Summary of Final Report, UK Competition & Markets Authority (Apr. 25, 2019), available at https://assets.publishing.service.gov.uk/media/5cc1434ee5274a467a8dd482/Executive_summary.pdf (“Overall, it seems unlikely that many retailers will raise their prices in response to the Merger; and even if some individual retailers do, the overall effect on UK households is unlikely to be negative. On that basis, our finding is that the Merger is unlikely to lead to customer harm through a waterbed effect.”).

[31] DeHoog v. Anheuser-Busch InBev, SA/NV, No. 1:15-CV-02250-CL, 2016 U.S. Dist. LEXIS 137759, at *13-16 (D. Or. July 22, 2016).

[32] Leading to truculent statements like that of California Attorney General Rob Bonta that “[r]ight now, there’s not a lot of reason not to sue [to block the merger].” See Nylen & Neumann, supra note 19.

[33] See, e.g., Dayen, supra note 3; Smoley, id.

[34] For instance, the Herfindahl–Hirschman Index (HHI) at which mergers are deemed problematic has been lowered from 2500 (and a post-merger increase of 200) to 1800 (and a post-merger increase of 100). Likewise, combined market shares of more than 30% are generally deemed problematic under the new guidelines (if a merger also increase the market’s HHI by 100 or more). The revised guidelines also focus more heavily on monopsony and labor-market issues. See Draft Merger Guidelines, supra note 4, at 6-7.

[35] John Asker et al, Comments on the January 2022 DOJ and FTC RFI on Merger Enforcement, available at https://www.regulations.gov/comment/FTC-2022-0003-1847 at 15-16 (emphasis added).

[36] Draft Merger Guidelines, supra note 4, at 21.

[37] See U.S. Dep’t of Justice & Fed. Trade Comm’n, 2010 Horizontal Merger Guidelines (Aug. 19, 2010) at §5.3, available at https://www.justice.gov/atr/horizontal-merger-guidelines-08192010#5c.

[38] Draft Merger Guidelines, supra note 4, at 26.

[39] Chad Syverson, Macroeconomics and Market Power: Context, Implications, and Open Questions 33 J. Econ. Persp. 23, 27 (2019).

[40] See U.S. Dep’t of Justice, Merger Guidelines (1968) at 6-7, available at https://www.justice.gov/archives/atr/1968-merger-guidelines.

[41] United States v. Von’s Grocery Co., 384 U.S. 270 (1966).

[42] See, e.g., Robert H. Bork, The Goals of Antitrust Policy, 57 Am. Econ. Rev. Papers & Proceedings 242 (1967) (“In the Von’s Grocery case a majority of the Supreme Court was willing to outlaw a merger which did not conceivably threaten consumers in order to help preserve small groceries in the Los Angeles area against the superior efficiency of the chains.”).

[43] Supra note 10; FTC v. Illumina, Inc., U.S. Dist. LEXIS 75172 (2021).

[44] United States v. Bertelsmann SE & Co. KGaA, No. CV 21-2886-FYP, 2022 WL 16949715 (D.D.C. Nov. 15, 2022).

[45] Id. (“The defendants do not dispute that if advances are significantly decreased, some authors will not be able to write, resulting in fewer books being published, less variety in the marketplace of ideas, and an inevitable loss of intellectual and creative output.”)

[46] Brian Albrecht, Business as Usual for Antitrust, City Journal (Nov 22, 2022), available at https://www.city-journal.org/article/business-as-usual-for-antitrust.

[47] Lina M. Khan, Rohit Chopra, & Kelly Slaughter, Comm’rs, Fed. Trade Comm’n, Statement on the Withdrawal of the Vertical Merger Guidelines (Sep. 15, 2021) at 3, available at https://www.ftc.gov/system/files/documents/public_statements/1596396/statement_of_chair_lina_m_khan_commissioner_rohit_chopra_and_commissioner_rebecca_kelly_slaughter_on.pdf.

[48] Premerger Notification Rules, 88 Fed. Reg. 42178 (RIN 3084-AB46), proposed Jun. 29, 2023 (to be codified at 16 C.F.R. Parts 801 and 803).

[49] Antitrust Experts Reject FTC/DOJ Changes to Merger Process, U.S. Chamber of Commerce (Sep. 19, 2023), https://www.uschamber.com/finance/antitrust/antitrust-experts-reject-ftc-doj-changes-to-merger-process. The surveyed group was made up seasoned antitrust veterans from across a variety of backgrounds: 80% had been involved in more than 50 mergers and 59% in more than 100.

[50] Supra note 23.

[51] Analysis of Agreement Containing Consent Order to Aid Public Comment, In the Matter of Cerberus Institutional Partners V, L.P., AB Acquisition, LLC, and Safeway Inc. (File No. 141 0108) (Jan. 27, 2015) available at https://www.ftc.gov/system/files/documents/cases/150127cereberusfrn.pdf. Analysis of Agreement Containing Consent Order to Aid Public Comment, In the Matter of Koninklijke Ahold N.V. and Delhaize Group NV/SA (Jul. 22, 2016) (File No. 151-0175) available at https://www.ftc.gov/system/files/documents/cases/160722koninklijkeanalysis.pdf.

[52] In the Matter of Cerberus Institutional Partners, supra note 6, at 2.

[53] In this paper, the terms “hypermarket” and “supercenter” are used synonymously. See Richard Volpe, Annemarie Kuhns, & Ted Jaenicke, Store Formats and Patterns in Household Grocery Purchases, Economic Research Service Economic Information Bulletin No. 167 (Mar. 2017), https://www.ers.usda.gov/webdocs/publications/82929/eib-167.pdf?v=0 (supercenters are also known as hypermarkets or superstores).

[54] In the Matter of Cerberus Institutional Partners, supra note 6, at 3.

[55] Id.

[56] Id.

[57] See In the Matter of Koninklijke Ahold, supra note 6.

[58] Data obtained from: U.S. Census Bureau, Report on Retail Sales and Trends: Annual Retail Trade Survey: 2021, https://www.census.gov/data/tables/2021/econ/arts/annual-report.html.

[59] Khan & Vaheesan, supra note 2, at 255.

[60] Id. (“Grocers sought to bulk up in order to compete with the scale of warehouse clubs and large discount stores, fueling further mergers and leading many local grocers to close….”).

[61] U.S. Census Bureau, supra note 58.

[62] Id.

[63] In the Matter of Cerberus Institutional Partners, supra note 6, at 2.

[64] Food Marketing Institute & The Hartman Group, Consumers’ Weekly Grocery Shopping Trips in the United States from 2006 to 2022 (Average Weekly Trips per Household), Statista (May 2022), available at https://www.statista.com/statistics/251728/weekly-number-of-us-grocery-shopping-trips-per-household.

[65] Michael Browne, Grocery Shopping Has a Hold on Consumers, Study Finds, Supermarket News (Jun. 27, 2018), https://www.supermarketnews.com/issues-trends/grocery-shopping-has-hold-consumers-study-finds.

[66] Kuhn, supra note 11.

[67] Id.

[68] Trip Drivers: Top Influencers Driving Shopper Traffic, Acosta (2017), available at https://acostastorage.blob.core.windows.net/uploads/prod/newsroom/publication_phetw_0rzq.pdf.

[69] Lijun Angelia Chen & Lisa House, US Food Shopper Trends in 2017, Univ. of Fla, IFAS Extension Pub. No. FE1126 (Dec. 7, 2022), https://edis.ifas.ufl.edu/publication/FE1126.

[70] Karen Webster, Consumer Shopping Data Shows Troubling Signs for Grocery Stores’ Future, PYMNTS (Feb. 6, 2023), https://www.pymnts.com/news/retail/2023/consumer-shopping-data-shows-troubling-signs-for-grocery-stores-future.

[71] See, e.g., In the Matter of Cerberus Institutional Partners, supra note 6, at 3.

[72] See Paul B. Ellickson, Paul L.E. Grieco, & Oleksii Khvastunov, Measuring Competition in Spatial Retail, 51 RAND J. Econ. 189 (2020) (“[C]lub stores are able to draw revenue from a significantly larger geographic area than traditional grocers. Hence, club stores are relevant substitutes for grocery stores, even if they are located even several miles away, a fact that could easily be overlooked in an analysis in which stores are simply clustered by geographic market.”).

[73] National Academies of Sciences, Engineering, and Medicine, A Satellite Account to Measure the Retail Transformation: Organizational, Conceptual, and Data Foundations (2021), available at https://www.bls.gov/evaluation/a-satellite-account-to-measure-the-retail-transformation.pdf (“[T]he restructuring that started first with the warehouse clubs and superstores and then moved on to e-commerce has begun to blur the lines between the retail industry and several other sectors….”).

[74] Id. at 25 (“[C]hanges experienced by retail over the past few decades suggest that the sector is highly competitive and is undergoing substantial change and reorganization. As discussed earlier, the changes described involve warehouse clubs and superstores … e-commerce … digital goods, imports, and large firms….”).

[75] Ellickson et al., supra note 72, (“Due to their size and attractiveness for larger purchases, club stores represent strong competitors to grocery stores even, when they are a significant distance away.”).

[76] Khan & Vaheesan, supra note 2, at 255 (“Grocers sought to bulk up in order to compete with the scale of warehouse clubs and large discount stores, fueling further mergers and leading many local grocers to close….”).

[77] See, e.g., In the Matter of Cerberus Institutional Partners, supra note 6, at 2.

[78] Costco Wholesale Corporation, Annual Report (Form 10-K) (Aug. 28, 2022), https://www.sec.gov/ix?doc=/Archives/edgar/data/0000909832/000090983222000021/cost-20220828.htm; BJ’s Wholesale Club Holdings, Inc., Annual Report (Form 10-K) (Mar. 16, 2023), https://www.sec.gov/ix?doc=/Archives/edgar/data/1531152/000153115223000026/bj-20230128.htm; Walmart Inc., Annual Report (Form 10-K) (Mar. 27, 2023), https://www.sec.gov/ix?doc=/Archives/edgar/data/104169/000010416923000020/wmt-20230131.htm.

[79] BJ’s Wholesale Club Holdings, Inc, id.

[80] Costco Wholesale Corporation, id.

[81] BJ’s Wholesale Club Holdings, Inc, id.

[82] Id.

[83] Costco Wholesale Corporation, id.

[84] Id.

[85] Russell Redman, Report: Club Stores Absorbing Grocery Market Share from Supermarkets, Winsight Grocery Business (Apr. 20, 2023), https://www.winsightgrocerybusiness.com/retailers/report-club-stores-absorbing-grocery-market-share-supermarkets.

[86] Hean Tat Keh & Elain Shieh, Online Grocery Retailing: Success Factors and Potential Pitfalls, 44 Bus. Horizons 73 (Jul.-Aug., 2001); Appinio & Spryker, Share of Consumers Purchasing Groceries Online in the United States in 2022, by Channel, Statista (Sep. 2002).

[87] Navigating the Market Headwinds: The State of Grocery Retail 2022, McKinsey & Co. (May 2022), available at https://www.mckinsey.com/~/media/mckinsey/industries/retail/how%20we%20help%20clients/the%20state%20of%20grocery%20retail%202022%20north%20america/mck_state%20of%20grocery%20na_fullreport_v9.pdf.

[88] Id.; Dimitri Dimitropoulos, Renée M. Duplantis, & Loren K. Smith, Trends in Consumer Shopping Behavior and Their Implications for Retail Grocery Merger Reviews, CPI Antitrust Chron. (Dec. 2021), available at https://www.brattle.com/wp-content/uploads/2022/01/Trends-in-Consumer-Shopping-Behavior-and-their-Implications-for-Retail-Grocery-Merger-Review.pdf.

[89] See Webster, supra note 70.

[90] National Academies of Sciences, Engineering, and Medicine, supra note 73 (“As e-commerce has grown in recent years, it has become increasingly difficult to separate out the e-commerce portion of the industry. Most e-commerce could be identified within the nonstore retailer category as of 2013, but e-commerce is becoming so pervasive that it is now not only difficult to clearly identify individual firms as predominantly e-commerce firms, but also often impossible to clearly classify individual retail sales as either e-commerce or not.” citations omitted).

[91] Dimitropoulos, et al., supra note 88 (“Of course, adjustments to geographic market definition likely would need to be factored into the analysis, as club stores tend to have larger catchment areas than traditional grocery stores, and online delivery can reach as far away as can be travelled by truck from a central fulfilment center.”)

[92] Draft Merger Guidelines, supra note 4, at 25-7.

[93] Maeve Sheehey & Dan Papscun, Kroger-Albertsons Merger Tests FTC’s Focus on Labor Competition, Bloomberg Law (Dec. 2, 2022) https://news.bloomberglaw.com/antitrust/kroger-albertsons-merger-tests-ftcs-focus-on-labor-competition.

[94] Kroger Union, UFCW (last accessed Jul. 26, 2023), https://www.ufcw.org/actions/campaign/kroger-union.

[95] Albertsons and Safeway Union, UFCW (last accessed Jul. 26, 2023), https://www.ufcw.org/actions/campaign/albertsons-and-safeway-union.

[96] Press Release, America’s Largest Union of Essential Grocery Workers Announces Opposition to Kroger and Albertsons Merger, UFCW (May 5, 2023), https://www.ufcw.org/press-releases/americas-largest-union-of-essential-grocery-workers-announces-opposition-to-kroger-and-albertsons-merger.

[97] Id. (“Given the lack of transparency and the impact a merger between two of the largest supermarket companies could have on essential workers – and the communities and customers they serve – the UFCW stands united in its opposition to the proposed Kroger and Albertsons merger”).

[98] Press Release, Kroger and Albertsons Companies Announce Definitive Merger Agreement, Kroger (Oct. 14, 2022), https://ir.kroger.com/CorporateProfile/press-releases/press-release/2022/Kroger-and-Albertsons-Companies-Announce-Definitive-Merger-Agreement/default.aspx.

[99] Bill Wilson, Teamsters Union Says ‘No’ to Kroger, Albertsons Merger, Supermarket News (Jun. 13, 2023),  https://www.supermarketnews.com/retail-financial/teamsters-union-says-no-kroger-albertsons-merger.

[100] Ben Zipperer, Kroger-Albertsons Merger Will Harm Grocery Store Worker Wages, Economic Policy Institute (May 1, 2023), https://www.epi.org/publication/kroger-albertsons-merger.

[101] Nathan Miller et al., On the Misuse of Regressions of Price on the HHI in Merger Review, 10 J. Antitrust Enforcement 248 (2022).

[102] See Geoffrey A. Manne, Dirk Auer, Brian C. Albrecht, Eric Fruits & Lazar Radi?, Comments of the International Center for Law and Economics on the DOJ-FTC Request for Information on Merger Enforcement (2022), at 29, available at https://papers.ssrn.com/sol3/papers.cfm?abstract_id=4090844.

[103] See United States v. Bertelsmann SE & Co. KGaA, 1:21-CV-02886, 2022 WL 16949715 (D.D.C. Nov. 15, 2022).

[104] Press Release, The Kroger Family of Companies Provides New Career Opportunities to 100,000 Workers, Kroger (May 14, 2020), https://ir.kroger.com/CorporateProfile/press-releases/press-release/2020/The-Kroger-Family-of-Companies-Provides-New-Career-Opportunities-to-100000-Workers/default.aspx. While the exact job-to-job switches are unknown, at least during the pandemic we know that some workers at non-grocery employers viewed at least some grocery-industry jobs as substitutes.

[105] Transcript of Proceedings at the Public Workshop Held by the Antitrust Division of the United States Department of Justice, U.S. Justice Department (Sep. 23, 2019), available at https://www.justice.gov/atr/page/file/1209071/download.

[106] Roman Inderst & Tommaso M. Valletti, Buyer Power and the ‘Waterbed Effect’ 59 J. Ind. Econ. 1, 2 (2011).

[107] Albert Foer, Mr. Magoo Visits Wal-Mart: Finding the Right Lens for Antitrust, American Antitrust Institute Working Paper No. 06-07, https://papers.ssrn.com/sol3/papers.cfm?abstract_id=1103609.

[108] Michael Needler Jr., Senate Hearing on Kroger and Albertsons Grocery Store Chains, at 1:43:00, available at https://www.c-span.org/video/?524439-1/senate-hearing-kroger-albertsons-grocery-store-chains.

[109] Julie Creswell, Kroger-Albertsons Merger Faces Long Road Before Approval, New York Times (Jan. 23, 2023), https://www.nytimes.com/2023/01/23/business/kroger-albertsons-merger.html.

[110] Diana Moss, The American Antitrust Institute to the Honorable Lina M. Khan, American Antitrust Institute (Feb. 7, 2023), available at https://www.antitrustinstitute.org/wp-content/uploads/2023/02/Kroger-Albertsons_Ltr-to-FTC_2.7.23.pdf.

[111] Inderst & Valletti, supra note 106. For a short history of the development of the waterbed model, see Eric Fruits, Sloshing Around with the “Waterbed Effect,” Truth on the Market (Sep. 5, 2023), https://truthonthemarket.com/2023/09/05/sloshing-around-with-the-waterbed-effect.

[112] Inderst & Valletti, supra note 106, at 9.

[113] Id. at 10.

[114] There has been some investigation of the waterbed effect in two-sided markets in telecommunications, but these markets are very different from retail food and grocery. See Christos Genakos & Tommaso Valletti, Testing the “Waterbed” Effect in Mobile Telephony, 9 J. Eur. Econ. Assoc. 1114 (Dec. 2011) (evaluating the effect of cutting mobile-termination fees on mobile-subscription prices).

[115] UK Competition & Markets Authority, supra note 30.

[116] DeHoog v. Anheuser-Busch InBev, supra note 31.

[117] See, e.g., Albert Foer, Mr. Magoo Visits Wal-Mart: Finding the Right Lens for Antitrust, American Antitrust Institute Working Paper No. 06-07 (Nov. 30, 2006), available at https://ssrn.com/abstract=1103609 (alleging a Walmart waterbed effect without identifying any product or product category with the relevant characteristics that would make it subject to the effect).

[118] See In the Matter of Cerberus Institutional Partners, supra note 6; In the Matter of Koninklijke Ahold, id., In the Matter of the Golub Corporation, id.

[119] See Dimitropoulos, et al., supra note 88.

[120] See id.

[121] See, e.g., Abigail Summerville and Anirban Sen, Analysis: Kroger, Albertsons Spin-Off Is Extra Ammunition in Regulatory Battle, Reuters (Oct. 17, 2022), https://www.reuters.com/business/retail-consumer/kroger-albertsons-spin-off-is-extra-ammunition-regulatory-battle-2022-10-17 (“Kroger Co and Albertsons Cos Inc are willing to divest up to 650 supermarket stores to secure regulatory clearance for their $24.6 billion deal….”); Dan Papscun, Kroger-Albertsons Divestiture Bid Aims to Head Off Challenge, Bloomberg Law (Oct. 14, 2022) https://news.bloomberglaw.com/antitrust/kroger-albertsons-divestiture-plan-is-bid-to-deflect-regulators (“The FTC must factor the divestiture proposal in its deal analysis, now that the companies themselves have built it into their own proposal, said Steven Cernak, a Bona Law partner. The companies’ divestiture proposal makes the tie-up ‘a tougher deal for the FTC to challenge,’ Cernak said.”).

[122] See Kroger/Albertsons: Companies Have Overlap of More Than 1,400 Stores; Khan Highly Critical of Failed Supermarket Divestitures, The Capitol Forum (Nov. 2, 2022), https://thecapitolforum.com/kroger-albertsons-companies-have-overlap-of-more-than-1400-stores-khan-highly-critical-of-failed-supermarket-divestitures.

[123] Id.

[124] The FTC’s Merger Remedies 2006-2012: A Report of the Bureaus of Competition and Economics, Fed. Trade Comm’n (Jan. 2017), available at https://www.ftc.gov/system/files/documents/reports/ftcs-merger-remedies-2006-2012-report-bureaus-competition-economics/p143100_ftc_merger_remedies_2006-2012.pdf.

[125] Id. at 2.

[126] Maureen Tkacik & Claire Kelloway, The No Spin-Off Zone, The American Prospect (Oct. 11, 2023), https://prospect.org/power/2023-10-11-no-spin-off-zone-kroger-albertsons-merger/.

[127] See, e.g., Dayen, supra note 3 (“As the Haggen affair makes clear, the whole idea of using conditions to allow high-level mergers and competition simultaneously has been a failure.”). See also Tkacik & Kelloway, id. (“The Kroger-Albertsons merger shows us why regulators need to permanently divest the concept of, well, divesting.”).

[128] Scott Neuman, Grocery Chains Safeway and Albertson’s Announce Merger Deal, The Two Way (Mar. 6, 2014), https://www.npr.org/sections/thetwo-way/2014/03/06/286935900/grocery-chains-safeway-and-albertsons-announce-merger-deal.

[129] See In the Matter of Cerberus Institutional Partners, supra note 6, at 2.

[130] See id., at 3-5.

[131] See id., at 5.

[132] Christopher A. Wetzel, Strict(er) Scrutiny: The Impact of Failed Divestitures on U.S. Merger Remedies, 64 Antitrust Bull. 341 (2019).

[133] Jon Talton, Haggen: What Went Wrong?, Seattle Times (Mar. 15, 2016), https://www.seattletimes.com/business/economy/haggen-what-went-wrong.

[134] David A. Balto, Supermarket Merger Enforcement, 20 J. Pub. Pol’y & Marketing 38 (Spr. 2001).

[135] See In the Matter of Cerberus Institutional Partners, supra note 6; In the Matter of Koninklijke Ahold, id., In the Matter of the Golub Corporation, id.

[136] See Facts and Figures, Publix (last visited Oct. 10, 2023), https://corporate.publix.com/about-publix/company-overview/facts-figures.

[137] See Caroline A., The History of Publix: Entering New States, The Publix Checkout (Jan. 4, 2018), https://blog.publix.com/publix/the-history-of-publix-entering-new-states; Press Release, Publix Breaks Ground on First Kentucky Store and Announces Third Location, Publix (Jun. 23, 2022), https://corporate.publix.com/newsroom/news-stories/publix-breaks-ground-on-first-kentucky-store-and-announces-third-location.

[138] Weis Markets, LinkedIn https://www.linkedin.com/company/weis-markets/about, (last accessed Jul. 26, 2023).

[139] Sam Silverstein, Weis Markets Unveils $150M Expansion and Upgrade Plan, Grocery Dive (May 2, 2022), https://www.grocerydive.com/news/weis-markets-unveils-150m-expansion-and-upgrade-plan/623015.

[140] Russell Redman, Wegmans lines up its next new store locations, Winsight Grocery Business (Dec. 1, 2022) https://www.winsightgrocerybusiness.com/retailers/wegmans-lines-its-next-new-store-locations.

[141] See, e.g., Summerville & Sen, supra note 121.

[142] See Catherine Douglas Moran & Petyon Giora, Mapping Kroger and Albertson’s Store Divestiture Deal with C&S, Grocery Dive (Sept. 12, 2023), https://www.grocerydive.com/news/mapping-kroger-and-albertsons-store-divestiture-deal-with-cs/693186.

[143] Kroger, supra note 13.

[144] Grocery Dive Staff, The Friday Checkout: C&S Would Catapult to Major Retailer Status with Kroger-Albertsons Deal, Grocery Dive (Sept. 8, 2023), available at https://www.grocerydive.com/news/cs-wholesale-grocers-major-grocer-kroger-albertsons-merger-deal/693127.

[145] Catherine Douglas Moran, Why More Grocers Are Bringing Private Label Production In-House, Grocery Dive (June 13, 2023), available at https://www.grocerydive.com/news/grocers-private-label-production-in-house-sales-manufacturing/651986.

[146] Tkacik & Kelloway, supra note 119

IN THE MEDIA

ICLE on Kroger’s Merger with Albertsons

Cincinnati Business Courier – An ICLE white paper about the proposed merger of Kroger and Albertsons was the subject of a story in the Cincinnati Business . . .

Cincinnati Business Courier – An ICLE white paper about the proposed merger of Kroger and Albertsons was the subject of a story in the Cincinnati Business Courier. You can read full piece here.

But analysts at the International Center for Law & Economics, a Portland, Ore.-based nonpartisan research center, say changes in the competitive landscape and the rise of e-commerce mean there’s no reason for regulators to reject the deal.

“Attempting to block this transaction would go against the analytical framework the FTC (Federal Trade Commission, a key antitrust regulator) has historically used to evaluate similar transactions, as well as the agency’s historical precedent of accepting divestures as a remedy to address localized problems where they arise,” the report said.

The report acknowledges breaks from precedent happen. But they don’t apply in the Kroger-Albertsons deal, the center’s experts said.

“In the case at hand, these breaks from tradition would reflect a failure to consider relevant and significant changes in how consumers shop for food and groceries in today’s world.”

“The agency appears poised to try to block the transaction – even with divestitures,” the report said.

The FTC has historically ruled that divestitures in supermarket mergers adequately address competition concerns, the report said. The divestitures planned by Kroger and Albertsons should resolve any antitrust concerns, it added.

“The parties appear committed to working cooperatively with regulators to craft divestitures that fully resolve competitive concerns,” it said. “Rather than blocking the deal outright, the FTC can allow the merger to proceed, conditioned on acceptable divestitures that protect consumers while permitting efficiency gains across the majority of stores.”

Dan Gilman on Epic v Apple

Law360 – ICLE Senior Scholar Daniel J. Gilman was quoted by Law360 in a story about ICLE’s amicus brief in the Epic v Apple case. You . . .

Law360 – ICLE Senior Scholar Daniel J. Gilman was quoted by Law360 in a story about ICLE’s amicus brief in the Epic v Apple case. You can read full piece here.

A think tank has urged the U.S. Supreme Court to take up Apple’s bid to nix an injunction blocking anti-steering rules in its App Store, saying the nationwide ban would affect a host of developers, not just Epic Games, who brought the case.

The International Center for Law & Economics, or ICL&E, filed an amicus brief Friday supporting Apple’s petition targeting the Ninth Circuit’s findings that its rules barring app developers from steering users to outside payment options violate California’s Unfair Competition Law.

…The brief on Friday argued that the lower court recognized the benefits of Apple’s “walled-garden” ecosystem, where it uses strict control to increase privacy and data security, but said the order blocking the anti-steering rules “undercuts precisely those benefits.”

“The district court’s sweeping remedy risks harming the vast majority of app developers, who have not requested the injunction and are now operating on the iOS for free,” the brief said. “And it may ultimately harm tens of millions of consumers using Apple’s App Store and iOS.”Daniel Gilman, a senior scholar of competition policy for ICL&E, told Law360 on Monday that the key issue from a legal and economic standpoint is that the injunction impacts “millions of developers,” not just Epic. Gilman also noted that a group of smaller developers reached a settlement with Apple that included changes that did not go as far as those imposed by the court.

“Many developers — especially smaller ones and those distributing free apps — would actually be harmed by this injunction, and they’ve had no voice in Epic’s case,” Gilman said. “Apple would be able to charge them directly, and it would have every incentive to do so.”

…Gilman also told Law360 on Monday there’s no reason for the court to issue a nationwide injunction under California law anyway, since it would be easy to identify the developers that have allegedly been harmed, saying it would include at most Epic and the 100 or so developers that use Epic’s app store.

Gus Hurwitz on Regulatory Capture and Big Tech

Gearrice – ICLE Director of Law & Economics Programs Gus Hurwitz was quoted by Gearrice, citing comments he’d earlier given to the New York Times, in a story . . .

Gearrice – ICLE Director of Law & Economics Programs Gus Hurwitz was quoted by Gearrice, citing comments he’d earlier given to the New York Times, in a story about jockeying by Big Tech firms to weaponize EU regulations against one another. You can read full piece here.

In practice, two of the largest companies in the world would be fighting back, using legislation aimed at helping small businesses. In this regard, Gus Hurwitz, senior researcher at the Carey Law School of the University of Pennsylvania, specializing in technology and competition, explained

“Regulations intended to help smaller companies enter the market can also very often be used by incumbents to gain an advantage over their rivals”

Gus Hurwitz on Big Tech and Regulatory Capture

New York Times – ICLE Director of Law & Economics Programs Gus Hurwitz was quoted by The New York Times in a story about jockeying by . . .

New York Times – ICLE Director of Law & Economics Programs Gus Hurwitz was quoted by The New York Times in a story about jockeying by Big Tech firms to weaponize EU regulations against one another. You can read full piece here.

Regulations intended to help smaller companies enter the marketplace “very frequently can also be used by incumbents to gain advantage over their rivals,” Gus Hurwitz, a senior fellow at the University of Pennsylvania Carey Law School who focuses on technology and competition, said in an interview.

…“I prefer companies to compete on the merits for consumers to want to use their products by offering higher-quality products,” Mr. Hurwitz said. “Not by paying lawyers to go to the European Union and get rules in place in order to obtain access to their competitors’ platforms.”

Gus Hurwitz on Weaponizing Regulation

9 to 5 Mac – ICLE Director of Law & Economics Programs Gus Hurwitz was quoted by 9 to 5 Mac, citing comments he’d earlier given . . .

9 to 5 Mac – ICLE Director of Law & Economics Programs Gus Hurwitz was quoted by 9 to 5 Mac, citing comments he’d earlier given to the New York Times, in a story about jockeying by Big Tech firms to weaponize EU regulations against one another. You can read full piece here.

Regulations intended to help smaller companies enter the marketplace “very frequently can also be used by incumbents to gain advantage over their rivals,” Gus Hurwitz, a senior fellow at the University of Pennsylvania Carey Law School who focuses on technology and competition, said.

Ben Sperry on the States’ Case Against Meta

Sacramento Bee – ICLE Senior Scholar Ben Sperry was quoted by The Sacramento Bee in a story about litigation filed by California Attorney General Rob Bonta . . .

Sacramento Bee – ICLE Senior Scholar Ben Sperry was quoted by The Sacramento Bee in a story about litigation filed by California Attorney General Rob Bonta and other state attorneys general against Meta. You can read full piece here.

Meanwhile, Ben Sperry of the International Center for Law and Economics, a nonpartisan research center, in a statement said that this lawsuit is “another effort by enterprising state AGs to make a name for themselves while states and the federal government grapple with how to properly regulate social media in order to deter harms to children.”

Sperry notes that there are several pending lawsuits already challenging Meta’s business practices, “and this suit is in some ways duplicative of those efforts.”

Sperry also notes that federal law granting social media companies a measure of immunity from state consumer protection laws has not been tested by the courts, and that some laws, including the California Age Appropriate Design Code Act — which aims to protect children’s privacy online — have already been enjoined by federal courts for likely violating the First Amendment.

Gus Hurwitz on HSR Premerger Notifications

National Review Capital Matters – ICLE Director of Law & Economics Programs Gus Hurwitz was cited by National Review Capital Matters in a story about proposed . . .

National Review Capital Matters – ICLE Director of Law & Economics Programs Gus Hurwitz was cited by National Review Capital Matters in a story about proposed changed to the premerger notification rules called for under the Hart-Scott-Rodino (HSR) Act of 1976. You can read full piece here.

The additional regulatory burden alone would cost more than the FTC and Justice Department’s combined 2023 antitrust budgets, according to former law professor Gus Hurwitz. And as the FTC increases the amount of paperwork to be submitted, the time it takes for the agencies to review initial HSR filings will increase as well. All this extra attention to pre-merger filings would take away resources from other agency priorities, such as litigation and consumer protection. The costs to the private sector would be even larger.

Kristian Stout on the FCC’s Title II Vote

Communications Daily – ICLE Director of Innovation Policy Kristian Stout was quoted by Communications Daily in a story about the Federal Communications Commission’s vote to reinstate . . .

Communications Daily – ICLE Director of Innovation Policy Kristian Stout was quoted by Communications Daily in a story about the Federal Communications Commission’s vote to reinstate Title II regulations on broadband providers. You can read full piece here.

The reintroduction of Title II regulation “stands on even shakier ground now than it did in 2015,” said Kristian Stout, International Center for Law & Economics director-innovation policy: “The wisdom behind the commission’s 2018 Restoring Internet Freedom Order is more evident with each passing year, as the broadband market continues to demonstrate no need for utility-style regulations.” He said the rulemaking presents a major questions doctrine issue “that will leave the FCC in an uphill battle before the U.S. Supreme Court in the face of inevitable legal challenges.”

Kristian Stout on the FCC’s Net Neutrality Vote

Law360 – ICLE Director of Innovation Policy Kristian Stout was quoted by Law360 in a story about the Federal Communications Commission’s vote to reinstate net neutrality . . .

Law360 – ICLE Director of Innovation Policy Kristian Stout was quoted by Law360 in a story about the Federal Communications Commission’s vote to reinstate net neutrality rules. You can read full piece here.

Some think tanks, including the International Center for Law & Economics, continued to oppose the FCC’s initiative on both policy and legal grounds.

“The justifications the FCC has put forward for imposing Title II obligations on broadband providers not only lack merit, but emphasize the likely legal challenges this order will face,” said Kristian Stout, ICLE director of innovation policy, in a statement.

Stout said the commission’s rulemaking presents a major-questions issue “that will leave the FCC in an uphill battle before the U.S. Supreme Court in the face of inevitable legal challenges.” The major-questions doctrine says large-scale regulatory initiatives that have broad impacts can’t be grounded in vague, minor and obscure provisions of law.

ICLE on the Kroger-Albertsons Merger

Blue Book Services – An ICLE white paper about the proposed merger of Kroger and Albertsons was the subject of a story in Blue Book Services. . . .

Blue Book Services – An ICLE white paper about the proposed merger of Kroger and Albertsons was the subject of a story in Blue Book Services. You can read full piece here.

The proposed merger of the Kroger and Albertsons retail grocery chains is the subject of a recent white paper issued by the International Center for Law and Economics in Portland, OR.

The paper argues on behalf of the merger and against Federal Trade Commission (FTC) policies that would block it.

“Attempting to block this transaction would go against the analytical framework the FTC has historically used to evaluate similar transactions, as well as the agency’s historical precedent of accepting divestures as a remedy to address localized problems where they arise,” says the paper.

ICLE on Kroger/Albertsons

Winsight Grocery Business – An ICLE white paper about the proposed merger of Kroger and Albertsons was the subject of a story in Winsight Grocery Business. . . .

Winsight Grocery Business – An ICLE white paper about the proposed merger of Kroger and Albertsons was the subject of a story in Winsight Grocery Business. You can read full piece here.

The Kroger Co. and Albertsons Cos.’ $1.9 billion deal to divest over 400 stores and supporting assets to C&S Wholesale Grocers offers a viable antitrust remedy to the supermarket giants’ planned mega-merger, the International Center for Law & Economics (ICLE) reported.

A nonprofit, nonpartisan research group, ICLE on Tuesday released a white paper titled “Food Retail Competition, Antitrust Law and the Kroger-Albertsons Merger,” in which the think tank evaluated $24.6 billion Kroger-Albertsons transaction, the divestiture pact with C&S and the Federal Trade Commission’s recent grocery merger experiences, including the failed Safeway-Albertsons divestiture to Haggen.

“With the FTC’s knowledge of the industry and of its own past successes and failures, divestitures remain an appropriate and adequate remedy for this merger,” Portland, Oregon-based ICLE wrote about the Kroger-Albertsons agreement. “The parties appear committed to working cooperatively with regulators to craft divestitures that fully resolve competitive concerns. Rather than blocking the deal outright, the FTC can allow the merger to proceed, conditioned on acceptable divestitures that protect consumers while permitting efficiency gains across the majority of stores.”

ICLE on the Merger of Kroger and Albertsons

Supermarket News – An ICLE white paper about the proposed merger of Kroger and Albertsons was the subject of a story in Supermarket News. You can . . .

Supermarket News – An ICLE white paper about the proposed merger of Kroger and Albertsons was the subject of a story in Supermarket News. You can read full piece here.

History says the $24.6 billion Kroger, Albertsons merger should survive any lawsuit, ultimately getting the stamp of approval from the Federal Trade Commission (FTC), according to a new report.

A white paper recently released by the International Center for Law & Economics, titled “Food-Retail Competition, Antitrust Law, and the Kroger/Albertsons merger,” says the only supermarket merger that has been challenged in court since American Store’s acquisition of Lucky Stores in 1988 is the Whole Foods/Wild Oats merger in 2007.

ICLE on Digital Discrimination

Communications Daily – ICLE comments to the Federal Communications Commission regarding digital discrimination in broadband were cited by Communications Daily in a story about the National . . .

Communications Daily – ICLE comments to the Federal Communications Commission regarding digital discrimination in broadband were cited by Communications Daily in a story about the National Telecommunications and Information Administration’s comments on the subject. You can read full piece here.

The International Center for Law & Economics agreed, adding the FCC “has for years been explicit about its apprehension to impose direct rate regulation.” It also asked the agency to reject NTIA’s proposal to adopt a disparate impact standard in the definition of digital discrimination. “Mere statistical correlation between outcomes and protected characteristics is insufficient to demonstrate discrimination,” the group said.

Brian Albrecht on Claudia Goldin’s Research

MarketWatch – ICLE Chief Economist Brian Albrecht was cited by The Guardian in a story about Claudia Goldin winning the 2023 Nobel prize in economics. You . . .

MarketWatch – ICLE Chief Economist Brian Albrecht was cited by The Guardian in a story about Claudia Goldin winning the 2023 Nobel prize in economics. You can read full piece here.

“We see ripples of her impact through lots of different fields across labor, economics, across education nationally,” said Brian Albrecht, chief economist at the International Center of Law and Economics, a nonprofit and nonpartisan global research and policy think tank based in Portland, Ore.

…Over the course of the pandemic, many workplaces have transitioned to a hybrid setup that offers remote work and flexible working hours. The biggest beneficiary has been working women: Labor-force participation for women ages 25 to 54 hit an all-time high in July, at 77.8%, according to the Bureau of Labor Statistics. A significant number of working parents, including Albrecht, were able to work from home and take care of their family life on the side.

More employers also now offer paid leave for parents after a child is born or adopted — an indirect impact of Goldin and her work, Albrecht said. That benefit helped employers and workers, especially working women, to redraw the line between work and life, he said.

“I don’t see how you get men getting paternity leave without maternity leave coming first,” he added.

 

Giuseppe Colangelo on Competition Authorities’ Use of the GDPR

EurActiv – A recent ICLE paper by Academic Affiliate Giuseppe Colangelo was noted as a newsletter item by EurActiv. You can read full piece here. GDPR . . .

EurActiv – A recent ICLE paper by Academic Affiliate Giuseppe Colangelo was noted as a newsletter item by EurActiv. You can read full piece here.

GDPR white paper. A white paper, published by the International Center for Law & Economics’s Academic Affiliate Giuseppe Colangelo of the University of Basilicata on Thursday, examines how EU policymakers apply the General Data Protection Regulation when it comes to antitrust inquiries.

Brian Albrecht on the FTC’s Resource Constraints

Axios – ICLE Chief Economist Brian Albrecht was quoted by Axios in a story about companies’ evaluation of government-intervention risk as agencies like the Federal Trade . . .

Axios – ICLE Chief Economist Brian Albrecht was quoted by Axios in a story about companies’ evaluation of government-intervention risk as agencies like the Federal Trade Commission amp up their antitrust enforcement. You can read full piece here.

Meanwhile, Brian Albrecht, chief economist at the International Center for Law & Economics, questioned how the FTC will compete with a constrained budget during a time that is drawing on more resources.
  • “I don’t see how you can push forward without a much larger budget.”

George Mocsary on Crypto Regulation

The Crypto Basic – ICLE Academic Affiliate George Mocsary was mentioned by The Crypto Basic in a story about the the fifth annual university blockchain research . . .

The Crypto Basic – ICLE Academic Affiliate George Mocsary was mentioned by The Crypto Basic in a story about the the fifth annual university blockchain research initiative. You can read full piece here.

He wrote in the post, “It is incredible to see global blockchain & engineering academic talent coming together via this program.” Meanwhile, Alderoty also gave a shoutout to Prof. George Mocsary of the Wyoming School of Law, who hosted the panel on the role of regulations in the crypto space.

Kristian Stout on FCC Spectrum-Screen Rulemaking

Communications Daily – ICLE Director of Innovation Policy Kristian Stout was quoted by Communications Daily in a story about  possible spectrum-screen rulemaking by the Federal Communications . . .

Communications Daily – ICLE Director of Innovation Policy Kristian Stout was quoted by Communications Daily in a story about  possible spectrum-screen rulemaking by the Federal Communications Commission. You can read full piece here.

Ben Sperry on Misinformation

Mark G. Goldberg & Assoc. – ICLE Senior Scholar Ben Sperry was cited by telecom attorney Mark Goldberg in a blog post about attempts to regulate . . .

Mark G. Goldberg & Assoc. – ICLE Senior Scholar Ben Sperry was cited by telecom attorney Mark Goldberg in a blog post about attempts to regulate misinformation. You can read full piece here.

Ben Sperry provides another perspective in a paper soon to be published in the Gonzaga Law Review. “The thesis of this paper is that the First Amendment forecloses government agents’ ability to regulate misinformation online, but it protects the ability of private actors — ie. the social-media companies themselves — to regulate misinformation on their platforms as they see fit.”

The Sperry paper concludes that in the US, regulating misinformation cannot be government mandated. Government could “invest in telling their own version of the facts”, but it has “no authority to mandate or pressure social-media companies into regulating misinformation.”

So, if government can’t mandate how misinformation is handled, by what rights can social media companies edit or block content? The author discusses why the “state-action doctrine” protects private intermediaries. According to Sperry, the social media platforms are positioned best to make decisions about the benefits and harms of speech through their moderation policies.

Geoff Manne on the FTC’s Amazon Lawsuit

American Institute for Economic Research – ICLE President Geoffrey Manne was cited by the American Institute for Economic Research in a blog post about the FTC’s antitrust . . .

American Institute for Economic Research – ICLE President Geoffrey Manne was cited by the American Institute for Economic Research in a blog post about the FTC’s antitrust case against Amazon. You can read full piece here.

In the FTC’s reductively defined market of “online superstores,” Amazon could control more than 60 percent. “This alleged market is so narrowly drawn that it appears to include just Amazon, eBay, and the online stores offered by Walmart and Target,” writes the International Center for Law and Economics (ICLE) President Geoffrey Manne. However, Amazon likely accounts for less than 30 percent of domestic online retail and just 5 percent of domestic retail generally. Roughly 80 percent of global retail still occurs in physical stores. The FTC omits key retail competitors such as brick-and-mortar stores, online specialty retailers, product brands’ own websites, and the like.

Brian Albrecht on Claudia Goldin

The Guardian – ICLE Chief Economist Brian Albrecht was cited by The Guardian in a story about Claudia Goldin winning the 2023 Nobel prize in economics. . . .

The Guardian – ICLE Chief Economist Brian Albrecht was cited by The Guardian in a story about Claudia Goldin winning the 2023 Nobel prize in economics. You can read full piece here.

Brian Albrecht, economist, has written a handy thread on Claudia Goldin’s academic work

Kristian Stout on vMVPDs

Law360 – ICLE Director of Innovation Policy Kristian Stout was cited extensively by Law360 in a story about his recent Truth on the Market post about . . .

Law360 – ICLE Director of Innovation Policy Kristian Stout was cited extensively by Law360 in a story about his recent Truth on the Market post about proposals for the Federal Communications Commission to revive rulemaking to impose cable television rules on some internet video services. You can read full piece here.

The International Center for Law & Economics is wading into the public debate over whether certain streamers should be regulated like traditional video service providers, warning the Federal Communications Commission that heaping regulations meant for other businesses upon streaming services is a bad idea.

A scholar from the free market-promoting think tank — which has called the current federal government “hostile” to mergers — penned a thinkpiece published Tuesday urging the agency not to reopen a 2014 proceeding that would have made streamers known as virtual multichannel video programming distributors subject to regulations made for traditional MVPDs, which include cable and satellite TV providers.

“[G]iven that the FCC will already have a plate full of difficult docket items, it should continue to avoid a further landmine that some advocates have been pressing to take up this year,” Kristian Stout of the ICLE said of potentially reopening the 2014 proceeding.

Kristian Stout on BEAD Implementation

Broadband Breakfast – ICLE Director of Innovation Policy Kristian Stout was quoted by Broadband Breakfast in a story about difficulties in implementing the Broadband Equity, Access . . .

Broadband Breakfast – ICLE Director of Innovation Policy Kristian Stout was quoted by Broadband Breakfast in a story about difficulties in implementing the Broadband Equity, Access and Deployment program. You can read full piece here.

Kristian Stout, director of innovation policy at the International Center for Law and Economics, said in comments to Broadband Breakfast that the biggest problem concerning pole attachments is that it is not a single problem.

“Because of how Section 224 operates, there are many areas of the country not subject to FCC oversight on access to poles,” Stout explained. “So on the one hand, there are issues before the FCC with access to some of the poles under its jurisdiction. In that case, the FCC needs to look at expediting disputes and developing a rate that doesn’t favor one party or the other.”

The problem is then multiplied across the states not subject to FCC oversight and the municipal and co-op pole owners, he said. “This creates a very complicated puzzle for any deployer to solve, and the incremental costs from each pole can quickly add up.”

…“The concern in because access to poles is critical for deployment,” stated Stout. “Unnecessary time delays and costs function as a hidden tax on broadband because they appear relatively small for any given pole but quickly explode when you multiply them by the hundreds or thousands of poles a particular deployment needs to cross. So even when a relatively small amount of pole owners play games, those costs are very significant.”

Stout added that he would be shocked if there were not delays and cost overruns from pole attachments issues in projects funded by the BEAD program. “It’s what we’ve seen in the past and nothing much has changed, so I expect to see it again in the future.”

…Stout suggested that FCC sets its rates correctly and ensures that disputes under its jurisdiction are expedited. He added that Congress needs to amend Section 224 to rationalize the landscape, saying that there is no reason why municipalities and co-ops should not be subject to either FCC or state regulation.

…However, Stout added that by and large, most problems can work themselves out because they are simply business negotiations in which each side wants to resolve the matter and move on. “It’s really when some pole owners create hold out problems on negotiations — and where some owners like municipalities and co-ops have no incentive to bargain at all — that very expensive problems can arise.”

Gus Hurwitz on Net Neutrality

DC Journal – ICLE Director of Law & Economics Programs Gus Hurwitz was cited in DC Journal in a story about the Federal Communications Commission’s play . . .

DC Journal – ICLE Director of Law & Economics Programs Gus Hurwitz was cited in DC Journal in a story about the Federal Communications Commission’s play to reimpose net neutrality. You can read full piece here.

Beyond the army of digital activists calling for net neutrality, Silicon Valley giants like Google, Netflix and Apple led the charge on net neutrality. University of Nebraska Law Scholar Gus Hurwitz wrote in The Washington Post in 2017, “The net neutrality rules are pure rent seeking by a content/edge industry as a way to hinder potential industrial competitors.”

Gus Hurwitz and Brian Albrecht on the DOJ’s Google Case

The Drum – ICLE Director of Law & Economics Programs Gus Hurwitz and Chief Economist Brian Albrecht were quoted by The Drum in a story about . . .

The Drum – ICLE Director of Law & Economics Programs Gus Hurwitz and Chief Economist Brian Albrecht were quoted by The Drum in a story about the antitrust case brought by the U.S. Justice Department against Google. You can read full piece here.

“Anyone who has been in sales knows that there can be intense pressure to ‘make your numbers.’ Sometimes that means making short-term decisions that generate quick revenue but are bad for the business longer term,” says Gus Hurwitz, senior fellow and academic director at University of Pennsylvania’s Center for Technology, Innovation, and Competition.

“As with having market power, this isn’t necessarily illegal,” Hurwitz says. “But it is dangerous: raising your prices is one of the fastest ways to lose customers. And, unlike in the search market, it is far less clear whether Google has market power in ads – Meta and Amazon have been growing and newer entrants like TikTok have been gaining a substantial foothold.”

… Antitrust law expert Hurwitz suggests that “there’s always a tradeoff in trials over how much should be open to the public.” As he put it to The Drum: “These are important public events, but they also significantly intrude upon private rights. The Department of Justice pushed the envelope a bit early in the trial with how aggressively they were sharing information.”

…Some competition experts, however, say that Google’s dominance over Microsoft in the search market isn’t necessarily unfair if Google’s product is simply superior. “Google definitely has a competitive advantage. That is why they are so popular. But they also have a competitive advantage because they are so popular,” says As Brian Albrecht, chief economist at the International Center for Law & Economics, a nonprofit research organization. “The question is why. Competitors like Microsoft have access to tons of data and high quality talent, as well as the money to spend, but they struggle to replicate Google’s quality. That’s not necessarily unfair, and is more often than not is the natural outcome of the competitive process.”

ICLE on Insurance Regulation in California

Insurance Portal – An ICLE white paper about California’s Proposition 103 insurance-regulatory system  was cited the subject of a story in Insurance Portal. You can read . . .

Insurance Portal – An ICLE white paper about California’s Proposition 103 insurance-regulatory system  was cited the subject of a story in Insurance Portal. You can read full piece here.

For those curious about why a company would exit certain markets altogether, as has been the case with State Farm exiting California’s homeowners’ insurance market in 2023, followed shortly by the exit of Allstate – a new whitepaper from the International Center for Law & Economics takes an in-depth look at California’s Proposition 103, which they say makes the ratemaking process there unpredictable at best.

 

PRESENTATIONS & INTERVIEWS

Brian Albrecht on Claudia Goldin

ICLE Chief Economist Brian Albrecht joined the Human Action Podcast to discuss the work of Nobel Prize winner Claudia Goldin, with an emphasis on the . . .

ICLE Chief Economist Brian Albrecht joined the Human Action Podcast to discuss the work of Nobel Prize winner Claudia Goldin, with an emphasis on the male-female wage gap. Video of the full episode is embedded below.

Kristian Stout on Title II Net Neutrality

ICLE Director of Innovation Policy Kristian Stout appeared as a guest on Minnesota Public Radio’s Marketplace in a segment on the Federal Communications Commission’s decision to . . .

ICLE Director of Innovation Policy Kristian Stout appeared as a guest on Minnesota Public Radio’s Marketplace in a segment on the Federal Communications Commission’s decision to reinstate so-called “net neutrality” for broadband providers.

But Kristian Stout, director of innovation policy at the International Center for Law and Economics, argues that we don’t need net neutrality as much as we once did because most of us are already online now. So how do we ensure access for every last American?

“You don’t do that by upending or frustrating the investment incentives that have made this work really well for 90 to 95% of the country. What you do is try to figure out targeted solutions,” Stout said.

Audio of the full segment is embedded below.

Ben Sperry on State Action Doctrine

ICLE Senior Scholar Ben Sperry joined The Dynamist podcast to discuss whether the Biden administration turned Big Tech companies into “state actors” in trying to . . .

ICLE Senior Scholar Ben Sperry joined The Dynamist podcast to discuss whether the Biden administration turned Big Tech companies into “state actors” in trying to regulate social-media content. The full episode is embedded below.

SHORT FORM WRITTEN OUTPUT

Market Power as a Limiting Principle in Merger Enforcement

One of the most important changes in the Federal Trade Commission (FTC) and U.S. Justice Department’s (DOJ) draft merger guidelines is the abandonment of market power . . .

One of the most important changes in the Federal Trade Commission (FTC) and U.S. Justice Department’s (DOJ) draft merger guidelines is the abandonment of market power as the central element of merger enforcement. The “unifying theme” of the 2010 horizontal merger guidelines was that “mergers should not be permitted to create, enhance, or entrench market power or to facilitate its exercise.” The draft guidelines have dropped the unifying theme language.

The guidelines’ abandonment of enhancement of market power as the central element of merger enforcement will have profound consequences for antitrust. One consequence is that merger enforcement will no longer prioritize consumers over competitors of the merging firms. Another important consequence, however, is the loss of a limiting principle in merger enforcement. Courts recognize that enhancement of market power is a necessary element of a merger challenge under antitrust law. The U.S. Circuit Court of Appeals for the D.C. Circuit made this point clear in its 2001 FTC v. H.J. Heinz opinion when it held that “[m]erger enforcement, like other areas of antitrust, is directed at market power.” The draft guidelines have removed enhancement of market power as a necessary element of a merger case.

Read the full piece here.

Everyone Discriminates Under the FCC’s Proposed New Rules

The Federal Communications Commission’s (FCC) proposed digital-discrimination rules hit the streets earlier this month and, as we say at Hootenanny Central, they’re a real humdinger. It looks . . .

The Federal Communications Commission’s (FCC) proposed digital-discrimination rules hit the streets earlier this month and, as we say at Hootenanny Central, they’re a real humdinger.

It looks like the National Telecommunications and Information Agency (NTIA) got most of their wishlist incorporated into the proposed rules. We’ve got disparate impact and a wide-open door for future rate regulation.

Here’s the tl;dr version of the new rules. More details at the end of this post.

Read the full piece here.

Why Armen Alchian Is the GOAT

Tyler Cowen has a new online book out titled “GOAT: Who is the Greatest Economist of All Time, and Why Does it Matter?” While there . . .

Tyler Cowen has a new online book out titled “GOAT: Who is the Greatest Economist of All Time, and Why Does it Matter?” While there are potential problems in treating ideas like basketball, the project is a fun, fast read overall. As the author of a newsletter with frequent gifs, I’m all for encouraging light-hearted discussions of economics and economists (in addition to the super serious work that needs to be done).

What does it mean to be the GOAT, according to Tyler?

To qualify as “GOAT the greatest economist of all time,” I expect the following from a candidate. The economist must be original, of great historical import, serve as a creator and carrier of important ideas, have a hand in both theory and empirics, have a hand in both macro and micro, and be “not too wrong” on the substance of issues. Furthermore, the person also must be a pretty good economist! That is, if you sat down with the person and discussed economic issues, you would be in some way impressed.

I won’t spoil Tyler’s answer, but we see from the table of contents that the main contenders are Milton Friedman, F.A. Hayek, J.M. Keynes, John Stuart Mill, Thomas Malthus, and Adam Smith.

Readers of our Economic Forces (heck, readers of the title of this post) newsletter will notice that Armen Alchian, our newsletter’s avatar, is not on the list. Alchian isn’t even on Tyler’s list of “names who deserve greater consideration,” a list that includes Homer. Yes, the poet. I know Homer reads The Economist, but c’mon! No mention of Alchian is a travesty! (It wouldn’t be a GOAT discussion without some hyperbole.)

So here is my case for why Armen Alchian is the GOAT.

Read the full piece here.

Four Misconceptions About the Consumer Welfare Standard

The consumer welfare standard has been the subject of a very effective contestation in modern antitrust law and policy literature. This contestation targets mostly United States law but, . . .

The consumer welfare standard has been the subject of a very effective contestation in modern antitrust law and policy literature. This contestation targets mostly United States law but, as we know, ideas travel fast. In spite of differences in law, policy, and institutions, contestations of consumer welfare frameworks have also emerged in slightly different terms in the European Union .

In this article, we lay bare the fundamental flaws of the modern critique of the consumer welfare standard. We show that critics misrepresent the meaning of the consumer welfare standard, distort the U.S. case law, and ignore important facts that do not align with their normative preferences. We conclude with the assertion that many criticisms of the consumer welfare standard among U.S. antitrust scholars reflect a critique of the U.S. judiciary’s attitude toward uncertainty and hard evidence rather than a critique of the consumer welfare standard itself.

Read the full piece here.

The FCC’s Digital-Discrimination Rules

tl;dr Background: Section 60506 of 2021’s Infrastructure Investment and Jobs Act (IIJA) mandated that the Federal Communications Commission (FCC) adopt rules to prevent discrimination in . . .

tl;dr

Background: Section 60506 of 2021’s Infrastructure Investment and Jobs Act (IIJA) mandated that the Federal Communications Commission (FCC) adopt rules to prevent discrimination in the deployment of broadband internet access “based on income level, race, ethnicity, color, religion, or national origin.” FCC Chair Jessica Rosenworcel recently outlined that the rules the commission intends to promulgate would define such digital discrimination “to include both disparate treatment and disparate impact.”

But… This approach conflicts with U.S. Supreme Court precedent on when a statute calls for disparate-impact analysis. The commission’s rulemaking will therefore likely invite lawsuits that challenge the agency’s authority to adopt these rules under the statute. 

This is particularly true under the Supreme Court’s emerging “major questions” doctrine, which requires that Congress speak clearly if it wants to delegate authority over questions of major economic or political significance to executive agencies.

The FCC’s broad interpretation of its mandate to promulgate digital-discrimination rules under the IIJA faces significant risk of being vacated by the courts, particularly if a challenge were to reach the Supreme Court.

KEY TAKEAWAYS

DISPARATE TREATMENT, DISPARATE IMPACT 

In discrimination law, disparate treatment refers to conduct intended to discriminate against one or more protected groups. In contrast, disparate impact is a finding that one or more protected groups is observed to experience different outcomes.

For example, disparate-impact analysis might find that low-income households have lower rates of internet adoption, and infer this was due to discrimination. Disparate-treatment analysis would evaluate whether the lower rate of adoption was due to provider policies or practices that were intended to stifle adoption by low-income households. 

In general, the bar to demonstrate a claim of discrimination is much lower under disparate impact than disparate treatment. But the FCC decided to incorporate both standards. In other words, a plaintiff would need to show disparate impact or disparate treatment in order to prove discrimination. 

But Section 60506’s language mandating the FCC prevent digital discrimination “based on” protected characteristics arguably indicates that Congress intended the FCC adopt a disparate-treatment approach. The Supreme Court has found that a statute must include “results-oriented language” to justify a disparate-impact approach to discrimination, which Section 60506 lacks.

MAJOR QUESTIONS AND CHEVRON

The so-called “major questions” doctrine affects how courts interpret congressional delegations of authority to federal agencies. The courts could find, for example, that if Congress intended the FCC to use a disparate-impact standard, it needed to say so clearly. The terse wording of Section 60506 does not appear to meet this level of clarity.

Even under longstanding Chevron analysis, an executive agency’s interpretation of a statute does not receive deference unless there is ambiguity in the enabling statute. Given the precedent, Section 60506 does not appear ambiguous in calling for a disparate-treatment standard.

TECHNICAL AND ECONOMIC FEASIBILITY

The IIJA requires that the FCC “tak[e] into account the issues of technical and economic feasibility” in crafting its digital-discrimination rules. Among the universe of potentially profitable broadband projects, firms will give priority to those that promise greater returns on investment. Such returns depend on factors like population density, terrain, regulations, and taxes, as well as a given consumer population’s willingness to adopt and pay for broadband. Many of these factors are, in turn, correlated with protected characteristics under the IIJA. A disparate-impact standard could thus incorrectly deem it to be improper discrimination when a firm responds to purely economic factors in its deployment decisions.

THE INCOME CONUNDRUM 

Congress’ inclusion of income level as a protected class in the IIJA made the FCC’s job much more difficult. Because income level is highly correlated with various protected (e.g., race and national origin) and unprotected (e.g., education level and home-computer ownership) characteristics, evaluations of income-based discrimination claims face a high likelihood of false positives, especially under a disparate-impact standard. Adoption of digital-discrimination rules that fail to recognize this “income conundrum” will invite costly and time-consuming litigation, both where no such discrimination exists and where it should be excused by considerations of economic feasibility. 

SLOUCHING TOWARD RATE REGULATION

Though the FCC has for years explicitly denied that it intends to impose direct rate regulation on broadband-internet providers, the National Telecommunications and Information Administration (NTIA) recently advised the FCC that: “Without addressing pricing as a possible source of discrimination, the Commission will be hard pressed to meet its statutory mandate to prevent digital discrimination of access.”

Any attempt to impose rate regulation under the language of Section 60506 would similarly face legal challenges under the major questions doctrine and Chevron.

For more on this issue, see “ICLE Ex Parte on Digital Discrimination.”

All Aboard! The Title II Express Is Leaving the Station

At lunch last week, I handed out the first of my new business cards with the title “Director, Hootenanny Division.” My lunchmate looked down and . . .

At lunch last week, I handed out the first of my new business cards with the title “Director, Hootenanny Division.” My lunchmate looked down and said, “Sounds fun, what do you do?”

Then, I had to explain that part of the job involves watching open meetings of the Federal Communications Commission (FCC) and reporting on what our federal government has in store for us next. It’s a bit like being a passenger on a steam train. No matter how much fuss you make in the coach cars, the engineer can’t hear you, and wouldn’t care if he could. The engineer’s got places to go, and nothing is going hold him back.

That’s like the FCC and its latest efforts to impose Title II regulation on much of the internet—nothing’s going to hold ‘em back.

Read the full piece here.

IBM, Microsoft and Big Tech Antitrust Folly

The continuing trial of Google, along with lawsuits against Amazon and Meta, have brought antitrust back into the public eye. These suits recall the 1969 . . .

The continuing trial of Google, along with lawsuits against Amazon and Meta, have brought antitrust back into the public eye. These suits recall the 1969 case against IBM and the 1998 case against Microsoft, the great antitrust battles of the latter half of the 20th century.

Supporters of aggressive antitrust enforcement think that only antitrust suits prevented IBM from commandeering the personal-computer market and Microsoft from taking over the internet. But that’s an urban legend.

Read the full piece here.

The Digital Markets, Competition and Consumers Bill Could Strike a Hammer Blow to UK Digital Investment

A far-reaching bill currently before Parliament would turn the UK Competition and Markets Authority (CMA) into one of the world’s most powerful tech regulators. Unfortunately, taking the . . .

A far-reaching bill currently before Parliament would turn the UK Competition and Markets Authority (CMA) into one of the world’s most powerful tech regulators. Unfortunately, taking the lead on regulation would almost certainly threaten Britain’s status as a leader in tech innovation.

Read the full piece here.

The Draft Merger Guidelines Risk Reducing Innovation

The draft Merger Guidelines released by the United States Department of Justice and the Federal Trade Commission (the Agencies) on July 19 feature many significant changes from earlier Guidelines. Of . . .

The draft Merger Guidelines released by the United States Department of Justice and the Federal Trade Commission (the Agencies) on July 19 feature many significant changes from earlier Guidelines. Of the 13 guidelines highlighted in the draft, two are particularly new and important for tech acquisitions. One is Guideline #4, which states that “mergers should not eliminate a potential entrant in a concentrated market.” The other is Guideline #9, which says that “when a merger is part of a series of multiple acquisitions, the agencies may examine the whole series” (emphases added).

Read the full piece here.

The View From Brazil: A TOTM Q&A with Mariana Tavares de Araujo

How did you come to be interested in the regulation of digital markets? Prior to joining Levy & Salomão Advogados, I worked with the Brazilian . . .

How did you come to be interested in the regulation of digital markets?

Prior to joining Levy & Salomão Advogados, I worked with the Brazilian government for nine years, four of which I served as head of the government agency in charge of antitrust enforcement and consumer protection policy. During this time, I was very lucky to participate in the early beginnings of the policy discussions on the need for enforcement in digital markets. Also, for a long time, this has been a very popular dinner conversation topic at home: my husband is in the software business and my stepdaughter is a computer engineer.

Read the full piece here.

Merger Divestitures: A Valuable Remedy for Competition Concerns

tl;dr Background: Divestitures are a common remedy sought by antitrust enforcers like the Federal Trade Commission (FTC) and U.S. Justice Department (DOJ). They have historically . . .

tl;dr

Background: Divestitures are a common remedy sought by antitrust enforcers like the Federal Trade Commission (FTC) and U.S. Justice Department (DOJ). They have historically been viewed as an appropriate tool to allow mergers to proceed while maintaining competition in particular markets where antitrust concerns might be most acute. The merging parties sell off select assets, often to a competitor. They thereby enable efficiencies in non-overlapping markets, while protecting consumers from anticompetitive effects in localized areas of concern.

But… Some critics argue that divestitures inevitably fail to preserve competition. They point to examples like Washington State-based Haggen Food and Pharmacy’s  failed acquisition of 146 stores divested from the 2014 merger of U.S. supermarkets Albertsons and Safeway. By 2016, Haggen sold the remaining 29 stores it was still operating back to Albertsons. Concerns about the inadequacy of divestitures have also been raised more recently in the context of the proposed $24.6 billion merger of Kroger Co. and Albertsons Cos. The firms recently announced a $1.9 billion agreement to sell 413 stores and eight distribution centers to C&S Wholesale Grocers to alleviate local concerns about the merger. Critics nonetheless claim that allowing mergers to proceed contingent on divestitures is inadequate to protect consumers.

KEY TAKEAWAYS

DIVESTITURES HAVE SUCCEEDED ACROSS INDUSTRIES WHEN PROPERLY STRUCTURED

Despite criticisms, empirical evidence demonstrates that divestitures have effectively maintained competition after mergers across industries ranging from retail to manufacturing to technology. 

The FTC’s own retrospective studies of past remedies find that most divestitures (around 80%) succeeded at achieving the high bar of “maintaining or restoring competition in the relevant market.” Divestitures succeed by equipping buyers with the assets, capabilities, and transitional support needed to replace lost competition. While an imperfect tool, with experienced buyers and FTC oversight, divested assets usually continue serving customers and thereby preserve pre-merger market dynamics. Ultimately, tailored divestitures can help competitors to navigate the complexities of business transfers to sufficiently replicate pre-merger discipline on the merged firm.

In retail, grocery divestitures have succeeded for combinations like Ahold/Delhaize (81 stores divested) and Albertsons/Safeway (168 stores). Divested assets must go to capable buyers with the expertise and capital to operate and expand them.

In the context of antitrust investigations, the FTC or DOJ oversees the buyer-selection process and integrates transition-assistance requirements into the consent orders it signs with the merging parties. While asset sales are the key to accomplish the goals of divestiture, ongoing agency monitoring is also possible, but requires resources.

TAILORED DIVESTITURE PACKAGES CREATE APPROPRIATE  COST-BENEFIT TRADEOFFS

No remedy perfectly maintains pre-merger conditions, and nor should it. Markets are always changing But if structured properly, asset divestitures can effectively preserve competition while allowing merger efficiencies.

Rather than blocking mergers outright, enforcers can approve deals conditioned on acceptable divestitures that target specific competitive concerns. This approach balances consumer protection with the economic benefits of consolidation.

ANTITRUST LAW DEALS IN LIKELIHOODS, NOT CERTAINTIES

In evaluating divestitures, enforcers and courts weigh the probabilities involved in maintaining competition, not hypothetical perfection. No structured resolution can guarantee that divested assets will perform identically to pre-merger conditions in all scenarios.

But antitrust law permits transactions where economic evidence indicates divestitures are reasonably capable of preserving competition post-merger. Enforcers make predictive judgments, while acknowledging that all business deals inherently carry uncertainty.

With proper buyer vetting, transitional support, and ongoing monitoring, empirical data shows that divestitures often succeed. The remote possibility of unforeseen complications does not negate high probabilities of sustained competitiveness.

Even critics agree that divestitures usually maintain some competition. So the key question is one of degree. If well-structured, data indicates divested assets can sufficiently maintain competitive pressures on the merged firm.

THE AGENCIES CAN LEARN FROM PAST FAILURES

Divestitures fail when buyers lack operational capabilities, face high barriers to expansion, take on too much debt, or merging parties act opportunistically to undermine divested assets.

In acquiring the divested Albertsons/Safeway locations, Haggen rapidly expanded from 18 to 168 stores, but quickly failed and entered bankruptcy. Its small size and debt load were vulnerabilities that proved fatal, which partially undid the divestiture plan’s pro-competitive effects. 

The FTC and DOJ must learn from past divestiture breakdowns to improve future remedies. Merger approval should require experienced buyers with resources to operate assets and transitional support agreements to ensure smooth transfers.

For more on this issue, see the ICLE issue brief “Five Problems with a Potential FTC Challenge to the Kroger/Albertsons Merger.” See also “FTC Should Allow Kroger-Albertsons Merger to Go Through” by Eric Fruits and Geoffrey Manne.

Is Amazon’s Scale a Harm?

Under the leadership of its professional anti-Amazoner Chair Lina Khan, the Federal Trade Commission (FTC) has finally filed its antitrust complaint against Amazon. No, not . . .

Under the leadership of its professional anti-Amazoner Chair Lina Khan, the Federal Trade Commission (FTC) has finally filed its antitrust complaint against Amazon. No, not the complaint about how it’s unfair to take six clicks to cancel your Prime membership. This is the big one. It mostly revolves around sellers needing to use Amazon’s fulfillment services to be part of Amazon Prime and lowering reach rankings if products are priced lower on other sites.

Instead of covering the arguments in the complaint, I want to use the complaint as an example of how I use the basics of supply and demand to sort through one of the arguments made by the FTC. Nothing about the use of price theory implies certain policy conclusions about the case. I’m just trying to be transparent, as I’ve done in the past, about how I use economics to reason about these important questions. Besides self-indulgence, the hope is that the examples help readers do the same.

Read the full piece here.

Net Neutrality Is a Solution in Search of a Hypothetical Problem

The Federal Communications Commission (FCC) has issued a new proposed rule that would impose far-reaching regulations on large internet service providers like AT&T, Comcast, and . . .

The Federal Communications Commission (FCC) has issued a new proposed rule that would impose far-reaching regulations on large internet service providers like AT&T, Comcast, and Verizon. These “net neutrality” rules would reclassify broadband internet access as a telecommunications service subject to common-carrier regulations under Title II of the Communications Act.

Read the full piece here.

A Response to Gus on Our New FTC Overlords

Gus Hurwitz’s closing post in this symposium was a very cogent and persuasive (albeit overly optimistic) take on the current state of antitrust enforcement. I hesitate . . .

Gus Hurwitz’s closing post in this symposium was a very cogent and persuasive (albeit overly optimistic) take on the current state of antitrust enforcement. I hesitate to quibble with my intellectual superior, but on some points I have a slightly different take.

Read the full piece here.

NTIA Tries to Call the Tune on Digital Discrimination

We’ve all been there. You’re enjoying a hootenanny, and someone tries to change the tune. Kind of like yelling “Free Bird” at a Taylor Swift . . .

We’ve all been there. You’re enjoying a hootenanny, and someone tries to change the tune. Kind of like yelling “Free Bird” at a Taylor Swift concert or asking a wedding DJ to play the chicken dance.

Well, the National Telecommunications and Information Administration (NTIA) has jumped into the digital-discrimination hootenanny and put in some requests.

Read the full piece here.

I, For One, Welcome Our New FTC Overlords

In this post—the last planned post for this symposium on The FTC’s New Normal (though we will continue to accept unsolicited submissions of responses)—I will offer some . . .

In this post—the last planned post for this symposium on The FTC’s New Normal (though we will continue to accept unsolicited submissions of responses)—I will offer some summary of the ideas that have been shared here over the past month, before turning to some of my own thoughts. To keep your attention rapt, I will preview that my thoughts will live up to the title of this post: I will sing some sincere praise for the Federal Trade Commission (FTC) and U.S. Justice Department’s (DOJ) honesty and newfound litigiousness.

Read the full piece here.

Competition in the Low-Earth-Orbit Satellite Industry

Amazon on Friday launched its first two prototype satellites for its planned Project Kuiper internet-satellite network. It was the latest milestone in the rapid evolution of the . . .

Amazon on Friday launched its first two prototype satellites for its planned Project Kuiper internet-satellite network. It was the latest milestone in the rapid evolution of the low-Earth-orbit (LEO) satellite industry, with companies like SpaceX and OneWeb joining Project Kuiper in launching thousands of satellites to provide broadband internet access globally.

As this nascent industry takes shape, it is important that U.S. policymakers understand its competitive dynamics. With the number of LEO satellites set to increase in the coming years, establishing a regulatory framework that spurs innovation and investment while fostering a competitive marketplace will be essential to ensure the industry’s growth benefits consumers. In this post, we will examine some of the most urgent public-policy issues that directly impact competitiveness in the LEO industry.

Read the full piece here.

The Changing Role of Structural Presumption at the Federal Trade Commission

The draft merger guidelines that were released July 19 by the Federal Trade Commission (FTC) and U.S. Justice Department (DOJ) indicate a shift by the . . .

The draft merger guidelines that were released July 19 by the Federal Trade Commission (FTC) and U.S. Justice Department (DOJ) indicate a shift by the agencies toward an overreliance on structural market factors to trigger merger scrutiny. 

For example, Draft Guideline 1—titled “Mergers Should Not Significantly Increase Concentration in Highly Concentrated Markets”—would lower the bar for what constitutes a “highly concentrated” market on the Herfindahl-Hirschman Index (HHI). Evaluating mergers through this structural lens assumes both that concentration is present or escalating in the economy at large and that concentration always and only leads to harmful anticompetitive effects. Most detrimentally, it likely precludes the more nuanced assessment of a given merger’s impact on competition that the FTC had used over the preceding four decades. This shift will increase Type I errors in antitrust enforcement and rob or delay the benefits of competition-enhancing mergers to consumers. 

Read the full piece here.

Draft Merger Guidelines Do Not ‘Return Antitrust to a Sound Economic and Legal Foundation’ – A Response to Professor Kwoka

In a recently published article in ProMarket, John Kwoka of Northeastern University (who “worked on the draft Merger Guidelines while serving at the Federal Trade Commission . . .

In a recently published article in ProMarket, John Kwoka of Northeastern University (who “worked on the draft Merger Guidelines while serving at the Federal Trade Commission as chief economist to the chair in 2022”) asserts that the U.S. Justice Department (DOJ) and Federal Trade Commission’s (FTC) draft merger guidelines aim to improve “deficient merger enforcement” by focusing on “lessening of competition,” rather than on “consumer surplus.”

With due respect to Kwoka, the draft guidelines, if adopted, would not “return antitrust to a sound economic and legal foundation.” Rather, they would generate costly uncertainty by returning federal merger enforcement to the pre-1980s structuralist era, when “efficiencies” was a dirty word. By discouraging welfare-enhancing merger proposals, they would harm the American economy.

Read the full piece here.

How ETNO’s ‘Fair Share’ Proposal Threatens Europe’s Digital Future:

The digital transformation of Europe—and, indeed, the world—has been a defining theme of the 21st century. As with all significant shifts, it has also come . . .

The digital transformation of Europe—and, indeed, the world—has been a defining theme of the 21st century. As with all significant shifts, it has also come with its share of challenges, opportunities, and controversies. 

One such controversy that has recently reemerged is the so-called “fair share” proposal for network traffic—championed most recently in a statement from the European Telecommunications Network Operators’ Association (ETNO)—under which the major tech platforms would be made to finance improvements to EU telecom networks. While ostensibly a call for regulatory change, the deeper one delves, the more evident it becomes that ETNO’s proposal is less about fairness and more a strategic play for legacy telecoms to tap into the vast revenues of major content producers.

Read the full piece here.

Net Neutrality Is Dead, Long Live National Security

Federal Communications Commission (FCC) Chair Jessica Rosenworcel announced plans last week for the commission to vote Oct. 19 on whether to take the first steps toward . . .

Federal Communications Commission (FCC) Chair Jessica Rosenworcel announced plans last week for the commission to vote Oct. 19 on whether to take the first steps toward reinstating Title II regulations on broadband providers. Two days later, the FCC issued a notice of proposed rulemaking (NPRM) for the Safeguarding and Securing the Open Internet (SSOI) order.

If adopted, the new rules would revive much of the Open Internet Order (OIO) the commission passed in 2015 under former FCC Chairman Tom Wheeler. That order classified broadband-internet service as a Title II telecommunications service under the Communications Act, treating many broadband services as public utilities. This allowed the FCC to impose common-carrier obligations on internet service providers (ISPs), including bans on blocking or throttling lawful content, paid prioritization of content, and other practices seen as contrary to so-called “net neutrality” principles.

Read the full piece here.

Abby Normal, a Flood of Ill-Considered Withdrawals, and the FTC’s Theatre of Listening

Lina M. Khan was sworn in as chair of the Federal Trade Commission (FTC) on June 15, 2021. On July 9 of that year, the . . .

Lina M. Khan was sworn in as chair of the Federal Trade Commission (FTC) on June 15, 2021. On July 9 of that year, the FTC withdrew the commission’s 2015 “Statement of Enforcement Principles Regarding ‘Unfair Methods of Competition’ Under Section 5 of the FTC Act.” That three-week lag was, in practical terms, nothing. Even ignoring the many practical/ministerial/managerial things that come with assuming the chair, there’s a certain amount of process required of policy decisions at the commission.

As many noted at the time, rescinding the 2015 statement, “absent any new guidance about how the Commission interprets its authority,” did little to signal the commission’s new view of its authority. That is, apart from the vague signal that a far more expansive statement was forthcoming and, of course, a not-so-tepid statement that…

Read the full piece here.

Labor-Market Monopsony

tl;dr Background: Concerns have been raised that the proposed merger of grocers Kroger and Albertsons may leave the combined firm with monopsony power in the . . .

tl;dr

Background: Concerns have been raised that the proposed merger of grocers Kroger and Albertsons may leave the combined firm with monopsony power in the markets for wholesale produce  and for grocery workers. This follows last year’s ruling by a federal court blocking the merger of Penguin Random House LLC and Simon & Schuster, similarly on grounds of labor-market monopsony. The argument is that the company would dominate in areas where the merging firms formerly competed for employees and other inputs. The combined firm could then use that power to suppress wages, reduce employment, or impose unreasonable working conditions on workers.

This isn’t the first time U.S. antitrust regulators have targeted monopsony in labor markets. In addition to merger review, other recent efforts have included lawsuits against “no-poach” agreements, as well as the Federal Trade Commission’s (FTC) recent proposal to ban  noncompete agreements in employment contracts.

But… Monopsony power often derives from labor-market frictions that antitrust can’t address. Most labor markets aren’t highly concentrated and most workers have multiple potential employers from which to choose. In other words, labor markets are generally poor targets for antitrust enforcement. As explained below, this raises several significant challenges for antitrust enforcers.

KEY TAKEAWAYS

MOST LABOR MARKETS ARE COMPETITIVE

So-called “company towns,” in which one firm dominates or actually owns a community, are rare. Most workers in the labor force have a broad range of employment opportunities across occupations, industries, and locations. A supermarket cashier can find employment at another supermarket, at another retail outlet, or shift their occupation to hospitality, food service, or distribution and logistics. They can also move to locations with better employment opportunities. A merger of supermarkets won’t suppress those opportunities.

The most compelling monopsony claims tend to concern labor markets that demand extensive or idiosyncratic skills, which couldn’t easily be transferred to other occupations or industries. For example, the Penguin/Simon & Schuster case centered on writers of bestsellers whose book advances exceed $250,000.

PRODUCT MARKETS IN ANTITRUST

All antitrust claims require defining a relevant market, but the endeavor is significantly more complicated in the context of labor markets. 

For example, what is the relevant labor market for supermarket employees? Surely, Costco employees should be included, even if Costco does not technically qualify as a “supermarket,” but what about employees of other retailers? What about hospitality and fast-food workers? When examining the labor market for workers who lack extensive or idiosyncratic skills, just about any reasonable definition of the relevant market would be too large to allege that any one firm possesses market power.

In a perfect world, these questions could be tested empirically. Unfortunately, antitrust enforcers often don’t have the requisite data and must rely on anecdotal evidence to delineate labor markets.

GEOGRAPHIC MARKETS IN ANTITRUST

U.S. workers are highly mobile. Roughly half of American adults live in a state other than the one in which they were born. Indeed, much of U.S. demographic history concerns people relocating for work. This makes it especially challenging to define a relevant geographic market for labor-monopsony claims. 

This is particularly true in urban environments, where there are many employment opportunities within commuting distance, especially for workers with fewer skills or less experience. Hence, stronger claims of labor-market monopsony tend to concern rural markets with limited job opportunities. It’s much easier to claim that Walmart holds labor-monopsony power in a small town than in even a medium-sized city.

UNION POWER AND ANTITRUST

Antitrust enforcers also need to account for the countervailing market power held by labor unions. Obtaining and exerting market power is unions’ raison d’être. As the old song says: “There is power in a union.”

For instance, if the FTC challenges the Kroger-Albertsons merger (as is expected) by alleging labor-market monopsony, the agency will have to contend with the fact that roughly 60% of the merged company’s workforce will be unionized. Attempts to exercise monopsony power would likely be dampened by the effects of unions collectively bargaining to maintain high wages and prevent layoffs.

BALANCING CONSUMER & WORKER WELFARE

The final challenge to labor-monopsony cases is that the primary purpose of antitrust enforcement is widely accepted to be protecting against harms to competition or to consumers. In labor cases, this will almost inevitably require important tradeoffs. 

While a merger might suppress the wages that would otherwise be paid by the merging companies, these wage reductions may then be passed on to consumers in the form of lower prices. Reduced labor input for a particular type of worker or workers does not mechanically translate into reduced output for consumers. This can be the case, for example, when a merger results in restructuring. 

In evaluating a merger, the agencies and the courts must balance the anticipated harms to employees against the potential benefits to consumers. This is a daunting task that may prove insurmountable in many cases.

For more on this issue, see the International Center for Law & Economics (ICLE) issue brief “Five Problems with a Potential FTC Challenge to the Kroger/Albertsons Merger.” See also, “FTC Should Allow Kroger-Albertsons Merger to Go Through” by Eric Fruits and Geoffrey Manne.

 

The Modern Video Marketplace Does Not Need Help From the FCC

The Federal Communications Commission (FCC) is no stranger to undertaking controversial and potentially counterproductive regulatory projects. The commission’s digital-discrimination proceeding is expected to continue in November, and . . .

The Federal Communications Commission (FCC) is no stranger to undertaking controversial and potentially counterproductive regulatory projects. The commission’s digital-discrimination proceeding is expected to continue in November, and FCC Chair Jessica Rosenworcel just announced that the FCC will revive the warmed-over corpse of the 2015 Open Internet Order. This latter item highlights how the FCC’s Democratic majority has been emboldened to pursue risky regulatory adventures with the addition of recently confirmed Commissioner Anna Gomez.

But given that the FCC will already have a plate full of difficult docket items, it should continue to avoid a further landmine that some advocates have been pressing to take up this year: reopening former Chair Tom Wheeler’s proceeding on multichannel video programming distributors (MVPDs). First proposed in late 2014 but ultimately not adopted by the commission, the Wheeler FCC’s notice of proposed rulemaking (NPRM) would bring over-the-top linear-video providers like YouTube TV and Hulu Live under the FCC’s program access and carriage rules.

Read the full piece here.

The FTC’s (and DOJ’s) Merger Aversion

There is mounting evidence that both the Federal Commission (FTC) and the U.S. Justice Department’s (DOJ) Antitrust Division (DOJ) are, under their current leadership, hostile . . .

There is mounting evidence that both the Federal Commission (FTC) and the U.S. Justice Department’s (DOJ) Antitrust Division (DOJ) are, under their current leadership, hostile to mergers. There are multiple elements to this evidence.

Read the full piece here.

Health Care and Health Insurance Merger Retrospective: A Personal Law & Economics Experience

My colleagues at the International Center for Law & Economics (ICLE) often engage not only in excellent analysis of proposed mergers and acquisitions, but also . . .

My colleagues at the International Center for Law & Economics (ICLE) often engage not only in excellent analysis of proposed mergers and acquisitions, but also have been known to offer retrospectives on past mergers. Today, I want to offer a very personal version of this.

Read the full piece here.

AMICUS BRIEFS

ICLE Amicus to US Supreme Court in Apple v Epic

Amicus respectfully submits this brief in support of Petitioner Apple Inc.[1] INTEREST OF AMICUS CURIAE The International Center for Law & Economics (“ICLE”) is a . . .

Amicus respectfully submits this brief in support of Petitioner Apple Inc.[1]

INTEREST OF AMICUS CURIAE

The International Center for Law & Economics (“ICLE”) is a nonprofit, non-partisan global research and policy center aimed at building the intellectual foundations for sensible, economically grounded policy. ICLE promotes the use of law and economics methodologies and economic learning to inform policy debates and has longstanding expertise evaluating antitrust law and policy.

ICLE has an interest in ensuring that antitrust law promotes the public interest by remaining grounded in sensible rules informed by sound economic analysis. That includes fostering consistency between antitrust law and other laws that proscribe unfair methods of competition, such as California’s Unfair Competition Law, and advising against far-reaching injunctions that could deteriorate the quality of mobile ecosystems, thereby harming the interests of consumers and app developers.

INTRODUCTION AND SUMMARY OF ARGUMENT

This Court has admonished that “injunctive relief should be no more burdensome to the defendant than necessary to provide complete relief to the plaintiffs.” Califano v. Yamasaki, 442 U.S. 682, 702 (1979); see also Application for a Stay at 5, Murthy v. Missouri, No. 23-411 (Sept. 14, 2023) (review granted on government’s stay motion arguing “the injunction sweeps far beyond what is necessary to address any cognizable harm to respondents”). The nationwide injunction issued in this case, which applies to millions of non-party app developers, cannot be reconciled with that principle.

The lower court’s use of a nationwide injunction to address narrow alleged injuries has severe consequences that are best understood through the lens of law and economics principles. The district court recognized that Apple’s walled-garden ecosystem yields procompetitive consumer benefits, including greater privacy and data security, and that such benefits are cognizable under federal antitrust law. Pet. App. 261a-270a. Yet the district court’s nationwide injunction undercuts precisely those benefits. Apple’s practice of vetting unsafe payment systems and malware on its App Store depends on its ability to prevent third parties from “steering” consumers towards purchase mechanisms other than Apple’s secure in-app purchasing (“IAP”) system. In addition, the anti-steering policy prevents free-riding and protects Apple’s incentive to invest in its platform to improve the curation of apps, privacy, safety, and security.

These harms to Apple’s platform are not offset by benefits to consumers, or even to developers taken as a whole. All the injunction does is alter the allocation of app store fees between developers, because even if Apple’s ability to collect a commission through its IAP is limited, Apple would still have the right to collect a commission in other ways for the use of its proprietary software and technology. It could do so by readjusting whom it charges for access to the App Store, and how much it charges.

For instance, rather than charge a commission to developers on paid downloads of apps and on in-app purchases of digital goods and services, as it does now, Apple could instead charge all developers a fee for accessing the App Store. While this might ostensibly benefit big developers who rely heavily on in-app purchases and paid downloads to monetize their apps, it is not at all clear that the net effects would be positive. One thing does seem clear, however: The current model, in which small, free apps pay few fees, would likely cease to be tenable under a nationwide federal injunction.

Put differently, despite not violating federal antitrust law, the district court’s sweeping remedy risks harming the vast majority of app developers, who have not requested the injunction and are now operating on the iOS for free. And it may ultimately harm tens of millions of consumers using Apple’s App Store and iOS.

ARGUMENT

I.              The Injunction Is Unnecessarily Broad and Would Affect Millions of Developers, Not Just Epic

The district court imposed an injunction that affects Apple’s anti-steering provisions across the board, and thus redefines Apple’s relationship with many developers—not just Epic. As it stands, the injunction is overly broad and at odds with established jurisprudence. Gill v. Witford, 138 S. Ct. 1916, 1933-34 (2018); Califano, 442 U.S. at 702. And it reduces consumer welfare by precluding more beneficial conduct than the harmful behavior it deters.

There are about thirty million registered app developers of native iOS apps. Pet. App. 10a. There are about two million apps  available in the United States storefront for the App Store, and most of them were created by third-party developers. See Apple Inc. v. Pepper, 139 S. Ct. 1514, 1519 (2019). All the developers have signed Apple’s guidelines regarding the exclusive use of Apple’s IAP and the related anti-steering provisions. By contrast, the trial evidence established that a little over 100 developers use Epic’s Epic Store. See Pet. App. 115a (citing Trial Tr. 1220:18-20). Yet, the anti-steering injunction would affect all App Store developers. The plaintiff is not even among these developers, because Epic was jettisoned from the App Store in 2020 for introducing an in-app payment system that bypassed Apple’s IAP. Epic has only one subsidiary that is active on the App Store. See Pet. App. 12a; D.Ct. ECF No. 825-8.

It is thus unclear why the district court found it necessary to issue an injunction covering all developers who are licensed to make iOS apps for the App Store’s U.S. storefront, not just Epic’s subsidiary and the approximately 100 developers who use the Epic Store.

Two considerations are especially pertinent. First, Califano precludes the Ninth Circuit’s erroneous assertion that an injunction need only be “tied to Epic’s injuries.” Pet. App. 82a; Califano, 442 U.S. at 702. Indeed, as the government argued in a recently granted petition that raises similar issues, an overbroad injunction cannot be justified on the theory that the non-parties are simply incidental beneficiaries of the injunction for the prevailing parties. Application for a Stay, supra, at 34-36; see Order Granting Review & Order Granting Stay, Murthy v. Missouri, No. 23-411 (Oct. 20, 2023). Instead, “[i]njunctive relief may ‘be no more burdensome to the defendant than necessary to provide complete relief to the plaintiffs.’” Id. at 34-35 (quoting Califano, 442 U.S. at 702).

Second, Apple already settled a class-action lawsuit with developers regarding developer-consumer communications. As part of the Cameron v. Apple Inc. settlement, Apple deleted a prohibition on targeted communication between developers and consumers outside of the app, meaning that developers are now free to communicate outside the apps about external purchasing options (or anything else). See Order: Granting Motion for Final Approval of Class Action Settlement; Granting in Part and Denying in Part Mot. for Attorney’s Fees, Costs, and Service Award; and Judgment at 13, Cameron v. Apple Inc., No. 19-cv-03074 (N.D. Cal. June 10, 2022), ECF No. 491. That settlement, spurred by a properly certified Rule 23 class action representing around 6,700 app developers, did not, however, require Apple to modify or remove the anti-steering provision at issue here (links and buttons within apps). See Declaration of Steve W. Berman in Support of Developer Plaintiffs’ Motion for Preliminary Approval of Settlement with Defendant Apple Inc. at 7-41, Cameron v. Apple Inc., No. 19-cv-3074 (Aug. 26, 2021), ECF No. 396-1.

It is jarring that the courts would now issue a much broader injunction in a case involving a single plaintiff. This could cause serious harm to nonparties who had no opportunity to argue for more limited relief. Zayn Siddique, Nationwide Injunctions, 117 Colum. L. Rev. 2095, 2125 (2017). And it also raises the question whether such a blanket remedy is even necessary given that Cameron v. Apple strikes a balance between Apple’s ability to safeguard its investments and maintain the safety and security of its ecosystem, and app developers’ ability to steer users to alternative payment systems. That agreement was found acceptable by Apple and some 6,700 app developers. Why should it be overridden by an injunction in a case involving a single plaintiff, when app developers have already had the opportunity to join a properly certified class action before, and have either chosen not to do so or have agreed to a different settlement? Further, if a single plaintiff’s allegations of harm can undercut a court-approved, negotiated settlement involving a much larger number of plaintiffs, that diminishes the incentives of parties to fashion and negotiate reasonable settlements in the first instance.

A broad injunction may well be warranted when it is difficult to separate the parties affected by the enjoined conduct from those that are not. But this is not the case here. The identity of the parties that have supposedly been harmed is clear—they are, at most, Epic’s subsidiary and the approximately 100 developers that use the Epic Store. Even if the district court’s conclusions regarding harm to Epic’s subsidiary and other developers with apps on the Epic Store were correct, it would be easy—and necessary—to carve a much narrower remedy than the one the district court imposed. See Barr v. Am. Ass’n of Pol. Consultants, Inc., 140 S. Ct. 2335, 2354-55 (2020).

Overly broad injunctions represent a Constitutional threat, as several members of this Court have warned. See, e.g., United States v. Texas, 143 S. Ct. 1964, 1980 (2023) (Gorusch, J., concurring); Trump v. Hawaii, 138 S. Ct. 2392, 2425 (2018) (Thomas, J., concurring); see also Lewis v. Casey, 518 U.S. 343, 360 (1996). “[G]ranting a remedy beyond what [is] necessary to provide relief to [the plaintiff is] improper.” Lewis, 518 U.S. at 360. In addition to such constitutional implications, overly broad injunctions also raise problems from a law and economics perspective such as hindering and even destroying beneficial conduct. If an injunction is not properly tailored, the beneficial conduct which it precludes may be greater than the harmful conduct which it prevents, resulting in a loss to both total social welfare and consumer welfare.

II.           Platforms Have Legitimate Business Reasons for Anti-Steering Provisions

By casting an overly wide net, the district court’s injunction throws the proverbial baby out with the bathwater. Anti-steering provisions are commonly used by digital platforms and other businesses because they serve a series of legitimate aims, such as allowing for the recoupment of investments. They also result in tangible procompetitive benefits, such as increased privacy, security, and market-wide output. These rules can be procompetitive, as this Court has recognized. Ohio v. Am. Express Co., 138 S. Ct. 2274, 2289 (2018) [hereinafter Amex].

Absent intervention by this Court, Apple will have to comply with a nationwide injunction that risks diminishing these benefits. If the decision is not corrected, the precedent could have a harmful ripple effect, subjecting other platforms to overly broad injunctions against anti-steering provisions, even though those anti-steering provisions may help sustain and improve the overall quality of those platforms.

A.            The framework for assessing competitive effects in a two-sided market requires a broad examination of the market as a whole

The district court properly found that Apple’s procompetitive justifications for the anti-steering provisions in its IAP system outweighed any anticompetitive effects of those provisions. In fact, Epic failed to make even a prima facie case under the requisite rule-of-reason analysis, as Epic failed to show that Apple’s app distribution and IAP system caused the significant, market-wide competitive harm that the Supreme Court deemed necessary to a showing of anticompetitive harm in Amex.

In Amex, the Court recognized the importance of platform economics and network effects to understanding the market and competitive effects at issue. Two-sided platforms intermediate between two groups, offering a different product or service to each. 138 S. Ct. at 2280 (citing e.g., David Evans & Richard Schmalensee, Markets with Two-Sided Platforms, 1 Issues in Competition L. & Pol’y 667 (2008); David Evans & Michael Noel, Defining Antitrust Markets When Firms Operate Two-Sided Platforms, 2005 Colum. Bus. L. Rev. 667 (2005)).

The Court noted that two-sided platforms are characterized by indirect network effects, where the value of the platform to each group depends on the scale of, or number of members in, the other. Id. at 2280-81. Specifically, the Court observed that “two-sided transaction platforms exhibit more pronounced indirect network effects and interconnected pricing and demand.” Id. at 2286 (emphasis added) (citing Benjamin Klein et al., Competition in Two-Sided Markets: The Antitrust Economics of Payment Card Interchange Fees, 73 Antitrust L.J. 571, 583 (2006)). Hence, “[e]valuating both sides of a two-sided transaction platform is . . . necessary to accurately assess competition.” Id. at 2287.

B.            Anti-steering provisions can be procompetitive

At issue in Amex were various anti-steering provisions American Express had placed in its contracts with merchants. The plaintiffs had alleged that the anti-steering provisions violated Section 1 of the Sherman Act. 138 S. Ct. at 2283. But in Amex, the Court recognized that “there is nothing inherently anticompetitive about . . . antisteering provisions.” Id. at 2289. Those vertical provisions can, among other things, prevent merchants from free-riding, thereby increasing the availability of “‘tangible or intangible services or promotional efforts’ that enhance competition and consumer welfare.” Id. at 2290 (quoting Leegin Creative Leather Prods., Inc. v. PSKS, Inc., 551 U.S. 877, 890-91 (2007)).

As in Amex, understanding the competitive effects of conduct between the platform and parties on either side of the platform—for example, vertical agreements between the IAP and app developers—requires examining effects on the system as a whole. And just as in Amex, there are legitimate, procompetitive reasons for anti-steering provisions.

First, as discussed above, anti-steering provisions help prevent free-riding. Simply put, “free-riding” occurs when someone uses a valuable resource without paying for it. Free-riding—even the potential for free-riding—tends to undermine incentives to provide the resource in the first place, as well as incentives to improve it in later development. It presents an especially serious challenge to the provision of goods or services where it is difficult to exclude those who have not paid, as with city parks or policing. Everyone, even those who would be willing to pay if they had to, has an incentive to avoid fees. Thus, where free-riding is possible, desirable goods and services tend to be underfunded, reducing their provision (or, in antitrust terms, output), or, in the alternative, are provided dependent on government subsidy. The most common solution to free-rider problems is to create ways to exclude those who are unwilling to pay.

In this case, Apple owns a valuable resource that it has created and steadily improved—the iPhone and iOS ecosystem, including the App Store. Apple currently charges commissions between 15% and 30% for digital goods sold through the App Store, including for certain in-app purchases. Epic would like to access that ecosystem without paying. But while Epic may benefit from its long-term strategy to reduce the fees it pays to Apple, consumers might not. If reductions in revenue from the iOS ecosystem mean that Apple has less incentive to invest in it, Epic’s gain may come at the consumer’s expense.

The district court correctly rejected Epic’s main claim, as Epic failed to establish cognizable harm under the antitrust laws. That foreclosed Epic’s ability to directly circumvent the App Store and pay a lower commission, or none at all. In granting a nationwide injunction against Apple’s anti-steering provisions, the district court facilitated precisely the type of free-riding that failed to gain traction under federal antitrust law. Doing so will greatly exacerbate any free-rider problem Epic itself might have caused Apple, to the likely detriment of many developers and most consumers.

The situation is further complicated by the fact that the district court’s injunction is vaguely written and is thus likely to be interpreted quite differently by different parties. Ultimately, if it allows app developers to link users outside of the in-app payments flow, and bypass Apple’s IAP fees, it will further enable free-riding and undermine Apple’s incentives to invest in iOS, iPhones, and iPads. And the injunction could undermine the incentive for Apple’s competitors to develop whatever products might someday displace the current ones through competition.

Second, as a two-sided market, the App Store is valuable only because it is used by both consumers and developers, and Apple has to balance both sides of the market. The risk of developers leaving the iOS ecosystem creates a built-in ceiling on the prices Apple can charge, as users will be less inclined to pay for Apple products if valuable apps are not there. The commission-fee business model gives Apple and other platforms significant incentives to develop new distribution mediums (like smart TVs, for example) and to improve existing ones. Such development expands the audience that software can reach.

Apple’s “closed” distribution model also allows the company to curate the App Store’s apps and payment options. For example, Apple’s guidelines exclude apps that pose data security threats, threaten to impose physical harm on users, or undermine child-safety filters. These rules increase trust between users and previously unknown developers, because users do not have to fear their apps contain malware. They also reduce user fears about payment fraud. Rivals could free-ride on Apple’s curation by mimicking its decisions and undercutting it on price. Doing so does not enhance competition on the merits: It eviscerates it by eroding Apple’s incentives to enforce such rules.

Apple’s closed business model also enables it to maintain a high standard of performance on iOS devices by excluding apps and payment systems that might slow devices or crash frequently. Users may not know when device performance is affected by a given app or purchase mechanism, so an open system would mean the potential for apps that crash the entire device. Apple’s closed model ensures that unscrupulous developers cannot impose negative externalities on the entire ecosystem.

By increasing the total value of the platform, these benefits also increase the number of market-wide transactions. In a two-sided market, it is output—not prices—that tells us what is happening on the market as a whole, and it is therefore output that should be used as the relevant parameter to determine whether conduct is procompetitive or anticompetitive. “What is material is whether Apple’s overall pricing structure reduces output by deterring app developers from participating in the market or users from purchasing apps (or iOS devices at all) because of the amount of the app developer commission.” Geoffrey A. Manne & Kristian Stout, The Evolution of Antitrust Doctrine After Ohio v. Amex and the Apple v. Pepper Decision That Should Have Been, 98 Neb. L. Rev. 425, 457 (2019). Notably, the district court found that it could not ascertain whether Apple’s alleged restrictions had “a negative or a positive impact on game transaction volume.” Pet. App. 253a; see also id. (“no evidence that a substantial number of developers actually forego making games because of Apple’s commission.”); id. at 319a (finding Epic failed to show reduction in output and that “[t]he record contains substantial evidence that output has increased.”).

Ultimately, however, the benefits of anti-steering provisions are obvious only if one adopts the correct, holistic vision of app stores as a two-sided market; conversely, they appear less relevant if one applies “one-sided logic in two-sided markets.” Julian Wright, One-sided Logic in Two-sided Markets, 3 Review of Network Econ. 44, 45-51 (2004). In this sense, in a two-sided market, anti-steering provisions can reduce transaction friction and bolster security and safety, thereby improving the platform’s overall quality and, ultimately, attracting more users. See Amex, 138 S. Ct. at 2889 (sustaining similar anti-circumvention rules as procompetitive for these reasons). Developers may get a smaller share of revenues, but it is a smaller slice of a much larger pie. Thus, while the ability to circumvent Apple’s commission fee can, on the surface, appear to benefit some developers, in the longer term most developers and consumers will be worse off.

Apple’s anti-steering provisions increase safety and curation, and an injunction against them can reduce the overall value of Apple’s platform. That would in turn discourage developers and users from using the iOS ecosystem, and would prompt a downward spiral in quality and choice for both sides of the market—which would depreciate the value of the platform even further.

C.            Open and closed platforms are not inherently good or bad: They represent alternative business models with potential advantages and disadvantages

Any comparison between “open” and “closed” platforms should account for the fact that there are tradeoffs between the two; it should not simply assume that “open” equals “good” while “closed” equals “bad.” Such analysis also must consider tradeoffs among consumers, and among developers, in addition to tradeoffs between developers and consumers. 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 especially benefits smaller developers and those whose apps fall outside the popular gaming sector. In short, the IAP fee cross-subsidizes the delivery of services to the approximately 80% of apps on the App Store that are free to consumers and pay no IAP fees.

Centralized app distribution and Apple’s “walled garden” model (including IAP) increase interbrand competition because they are at the core of what differentiates Apple from Android, the other major competing platform. 1-ER-148–49. They play into Apple’s historical business model, which focuses on being user-friendly, reliable, safe, private, and secure. 1?ER-86; see also 1-ER-107 (recognizing that the safety and security of Apple’s closed system is a “competitive differentiator for its devices and operating system”). Even Epic recognized that Apple would lose its competitive advantage if it were to compromise its safety and security features. 1-ER-48 n.250 (noting Epic’s expert, Susan Athey, testified that “privacy and security are competitive differentiators for Apple”).

For Apple and its users, the touchstone of a good platform is not “openness,” but carefully curated selection and security, understood broadly as encompassing the removal of objectionable content, protection of privacy, and protection from “social engineering,” and the like. 1-ER-148–49. By contrast, Android’s bet is on the open platform model, which sacrifices some degree of security for the greater variety and customization associated with more open distribution. These are legitimate differences in product design and business philosophy. See Andrei Hagiu, Proprietary vs. Open Two-Sided Platforms and Social Efficiency 2-3 (AEI-Brookings Joint Ctr. for Regul. Stud., Working Paper No. 06-12, 2006) [hereinafter Proprietary vs. Open Platforms] (explaining that there is a “fundamental welfare tradeoff between two-sided proprietary . . . platforms and two-sided open platforms, which allow ‘free entry’ on both sides of the market” and thus “it is by no means obvious which type of platform will create higher product variety, consumer adoption and total social welfare”) (emphasis omitted); Jonathan M. Barnett, The Host’s Dilemma: Strategic Forfeiture in Platform Mkts. for Informational Goods, 124 Harv. L. Rev. 1861, 1927 (2011) (“Open systems may yield no net social gain over closed systems, can impose a net social loss under certain circumstances, and . . . can impose a net social gain under yet other circumstances.”).

Because consumers and developers could reasonably prefer either ecosystem, it is not clear that loosening Apple’s control over the App Store would necessarily lead to more app transactions market wide. Indeed, in a two-sided market context, a proprietary platform like Apple’s “may in fact induce more developer entry (i.e. product variety), user adoption and higher total social welfare than an open platform.” Proprietary vs. Open Platforms, at 15-16. In other words, preventing certain apps from accessing the App Store, and preventing certain transactions from taking place on it, may ultimately have increased the number of apps and transactions on Apple’s platform, because doing so made it attractive to a wider set of consumers and developers.

Yet the injunction brings Apple’s iOS closer to an “open” system, effectively rendering Apple’s platform more similar to Android’s. The district court found that Apple did not have a monopoly, yet under the guise of fostering competition on Apple’s platform the injunction eliminates competition where it matters most—at the interbrand, systems level between Apple and Android. See Michael L. Katz & Carl Shapiro, Systems Competition and Network Effects, 8 J. Econ. Persps. 93, 110 (1994), (“[T]he primary cost of standardization is a loss of variety: consumers have fewer differentiated products to pick from, especially if standardization prevents the development of promising but unique and incompatible new systems”). By limiting intrabrand competition, in other words, Apple ultimately promotes interbrand competition. 1-ER-148–49. Again, Amex provides useful insight here, because the Court noted that the business model had “spurred robust interbrand competition,” while increasing both the quality and quantity of transactions. Amex, 138 S. Ct. at 2290.

D.            Anti-steering provisions are a legitimate way of recouping a platform’s investments

Anti-steering provisions are a legitimate way for a platform to recoup its investments. Epic has argued that Apple could simply lift restrictions on the use of third-party IAP processors (e.g., Visa and MasterCard), but still be appropriately compensated for the use of its intellectual property, ensure that iPhone users’ IAP are sufficiently secure, and guarantee quality. 1-ER-153; Epic 9th Cir. Br. 44-47. But exactly how Apple could achieve these ends without increasing its costs is a question Epic has not even tried to answer. See, e.g., 1-ER-151–52 (noting that Epic’s requests for relief “leave unclear whether Apple can collect licensing royalties and, if so, how it would do so”); 1-ER-153 & n.617 (noting it would “be more difficult” and more costly for Apple to collect commission without the IAP system). Nor did Epic, the Epic amici, or the district court properly address the effect of the proposed less restrictive alternatives on consumers rather than competing developers. See 1-ER-148 n.605 (noting it is “unclear the extent or degree to which developers would pass on any savings to consumers”).

Consistent with Epic’s proposed approach, Apple could allow independent payment processors to compete, and charge an all-in fee of 30% when Apple’s IAP is chosen. To recoup the costs of developing and running its App Store, Apple could then charge app developers a reduced, mandatory per-transaction fee (on top of developers’ “competitive” payment to a third-party IAP provider) when Apple’s IAP is not used. Indeed, where a similar remedy has been imposed already, Apple has taken similar steps. In the Netherlands, for example, where Apple is required by the Authority for Consumers and Markets to uncouple distribution and payments for dating apps, Apple has adopted a policy under which any apps that want to use a non-Apple payment provider must still “pay Apple a commission on transactions” that is 3% less than normal (so 27% for most transactions), a slightly “reduced rate that excludes value related to payment processing and related activities.” Apple, Distributing Dating Apps in the Netherlands, (last visited Oct. 26, 2023).

III.         A State Law Should Not Undermine the Fundamental Goals of Federal Antitrust Policy

When assessing the effects of Apple’s anti-steering provisions, the courts should not ignore Federal antitrust law and, especially, the effects on competition and consumers. In other words, the fact that anti-steering provisions are procompetitive should be a relevant factor in whether a federal court grants nationwide injunctive relief. To interpret California’s Unfair Competition Law (“UCL”) as the district court has done—in a way that is at loggerheads with federal antitrust law but yet permits a nationwide injunction—is to undermine the fundamental goal of antitrust policy, and to do so on a national level. As the Court has observed, “The heart of our national economic policy long has been faith in the value of competition.” Nat’l Soc’y of Prof. Eng’rs v. United States, 435 U.S. 679, 695 (1978) (quoting Standard Oil Co. v. FTC, 340 U.S. 231, 248 (1951)).

The district court recognized Apple’s security arguments as a key procompetitive factor that determines Apple’s success and increases output across the platform, ultimately benefitting both consumers and developers. Yet the court issued an unnecessarily broad injunction against Apple’s anti-steering provisions that risks chilling procompetitive conduct by deterring investment in efficiency-enhancing business practices, such as Apple’s “walled-garden” iOS (see sections II.B and II.D on the procompetitive benefits of anti-steering provisions). See also Verizon Commc’ns, Inc. v. Law Offices of Curtis V. Trinko, LLP, 540 U.S. 398, 414 (2004) (“[F]alse condemnations ‘are especially costly, because they chill the very conduct the antitrust laws are designed to protect.’”) (quoting Matsushita Elec. Indus. Co. v. Zenith Radio Corp., 475 U.S. 574, 594 (1986)).

Even more egregiously, perhaps: It risks undermining federal antitrust law by enjoining conduct under state unfair competition law that is recognized as benign—and even beneficial—under federal antitrust law. If the district court’s remedy is left to stand, state laws will be stretched beyond their territorial remit and used to contradict federal antitrust laws nationally, thus eviscerating federal antitrust policy from the bottom-up. This is not a hypothetical threat, either: California has already expressed its intent to use the UCL to “seek nationwide injunctions” on the same theory as the ruling below. Michael Acton, Epic Games-Apple US Appeals Court Ruling Shows Power of California’s Competition Law, Blizzard Says, MLex (May 10, 2023).

The district court erred in finding Apple’s anti-steering provision “unfair” despite a concurrent finding that there is no incipient antitrust violation. And a nationwide injunction based on that finding lifts what could have been a relatively contained mistake to the national level, and thereby magnifies it.

This is misguided from an antitrust perspective because it undermines some of the procompetitive benefits that anti-steering provisions in closed two-sided platforms can give to consumers and app developers. A national injunction that subverts Apple’s ability to charge a commission for the use of its software and technology through paid apps and in-app payments might also alter the current balance between the two sides of the App Store, to the detriment of smaller developers of free apps. In this zero-sum game, the gain of a handful of developers who rely on paid downloads of apps and frequent in-app purchases by users will come at the expense of the majority who do not.

CONCLUSION

For the foregoing reasons, this Court should grant Apple’s petition for a writ of certiorari.

[1] Amicus notified counsel for the parties of its intent to file this brief more than ten days before the deadline. No counsel in this matter for any party authored this brief in whole or in part, and no person other than amicus or its counsel have made any monetary contribution intended to fund the preparation or submission of this brief.

Amicus of Legal and Economic Scholars to the 5th Circuit in Tesla v Louisiana Auto Dealers Association

STATEMENT OF AMICI INTEREST Amici are law professors, economists, or other academics with expertise in competition law and economic regulation. Amici do not work for . . .

STATEMENT OF AMICI INTEREST

Amici are law professors, economists, or other academics with expertise in competition law and economic regulation. Amici do not work for Tesla, nor have they been compensated in any way for their participation in this brief.[1]

SUMMARY OF ARGUMENT

Amici appear in support of Tesla on two issues with a common thread.[2] The district court’s opinion erred in insulating the actions of the Louisiana legislature and the Louisiana Motor Vehicle Commission (“LMVC”) from antitrust and constitutional review under a flawed framework for scrutinizing state regulations that suppress competition and favor economic special interests.

First, Amici submit that the district court erred in holding that commissioners of the LMVC were protected by Noerr-Pennington immunity when they “agreed with [the Louisiana Automobile Dealers Association (“LADA”)] to use the regulatory power of the Commission to investigate Tesla.” Op. at 27. Although public officials may enjoy Noerr-Pennington immunity when they act in a purely private capacity, a public official who is also a market participant and agrees with others to utilize public power in a manner designed to suppress competition in order to further his own economic interests should not be immunized from antitrust scrutiny. The Noerr-Pennington doctrine protects the rights of citizens to petition the government for redress of grievance. It does not protect governmental officials who conspire to use governmental power to favor their own economic interests. The district court’s approach would create a loophole in the antitrust laws permitting actors wielding state power to avoid responsibility for abuses of official power.

Second, Amici dispute the district court’s finding that Louisiana’s direct sales ban had a rational basis in consumer protection. As Amici explain below, direct sales bans in automotive retailing were historically focused on the exclusive goal of protecting dealers in franchise relationships with manufacturers. Thus, in the cases in which this Court upheld such statutes against constitutional challenge—Ford Motor Co. v. Texas Dep’t of Transp., 264 F.3d 493 (5th Cir. 2001); Int’l Truck & Engine Corp. v. Bray, 372 F.3d 717 (5th Cir. 2004)—the ostensible rational basis of the legislation was the protection of dealers against the superior bargaining power of their franchising manufacturers. But that logic can have no bearing on the application of Louisiana’s 2017, anti-Tesla direct sales prohibition, for the simple reason that Tesla (and other new electric vehicle manufacturers) do not use franchised dealers at all, but sell directly to the consuming public. In such circumstances, dealers are not being protected as franchisees, they are protected from economic competition by companies using a different business model—exactly what this Court held does not count as a rational basis in St. Joseph Abbey v. Castille, 712 F.3d 215 (5th Cir. 2013). Further, efforts to justify direct sales bans as consumer protection rather than dealer protection have no support in economic theory or evidence. Such arguments are mere pretexts for the economic protectionism that this Court has held does not survive equal protection scrutiny

[1] Amici join this brief solely in their individual capacities and express only their individual views. Institutional affiliations are listed for identification purposes only.

[2] Amici take no position on other arguments raised by Tesla’s appeal.

 

COMMENTS & STATEMENTS

ICLE Comments on Artificial Intelligence and Copyright

Introduction We thank you for the opportunity to comment on this important notice of inquiry (NOI)[1] on artificial intelligence (AI) and copyright. We appreciate the . . .

Introduction

We thank you for the opportunity to comment on this important notice of inquiry (NOI)[1] on artificial intelligence (AI) and copyright. We appreciate the U.S. Copyright Office undertaking a comprehensive review of the policy and copyright-law issues raised by recent advances in generative AI systems. This NOI covers key areas that require attention, from legal questions regarding infringement and fair use, to questions about how policy choices could shape opportunities for creators and AI producers to engage in licensing.

At this early date, AI systems have already generated some incredible visual art and impressive written texts, as well as a good deal of controversy. Some artists have banded together as part of an anti-AI campaign;[2] lawsuits have been filed;[3] and policy experts have attempted to think through the various legal questions raised by these machine-learning systems.

The debates over the role of AI in creative industries have particular salience for intellectual-property rights. Copyright is notoriously difficult to protect online, and the emergence of AI may exacerbate that difficulty. AI systems also potentially pose an additional wrinkle: it is at least arguable that the outputs they produce can themselves be considered unique creations. There are, of course, other open questions whose answers are relevant here, not the least being whether it is fair to assert that only a human can be “creative” (at least, so far).[4]

But leaving these questions aside, we can say that at least some AI systems produce unique outputs and are not merely routinely duplicating other pieces of work in a digital equivalent of collage. That is, at some level, the machines are engaged in a rudimentary sort of “learning” about how humans arrange creative inputs when generating images, music, or written works. The machines appear to be able to reconstruct this process and produce new sets of words, sounds, or lines and colors that conform to the patterns found in human art, in at least a simulacrum of “creativity.”

But that conclusion isn’t the end of the story. Even if some of these AI outputs are unique and noninfringing, the way that AI systems learn—by ingesting massive quantities of existing creative work—raises a number of thorny copyright-law issues. Indeed, some argue that these systems inherently infringe copyright during the learning phase and that, as discussed below, such processes may not survive a “fair use” analysis.

But nor is that assertion the end of the analysis. Rather, it raises the question of whether applying existing doctrine in this novel technological context yields the best results for society. Moreover, it heightens the need for a comprehensive analytical framework to help parse these questions.

A.            The Law & Economics of Copyright and AI

Nearly all would agree that it is crucial that law and public policy strike the appropriate balance between protecting creators’ existing rights and enabling society to enjoy the potentially significant benefits that could arise from the development of AI systems. Indeed, the subject is often cast as a dramatic conflict between creative professionals struggling to make ends meet and innovative firms working to provide cutting-edge AI technology. For the moment, however, it is likely more important to determine the right questions to ask and the proper analytical framework to employ than it is to identify any precise balancing point.

What is important to remember is that copyright policy is foremost economic in nature and “can be explained as a means for promoting efficient allocation of resources.”[5] That is to say, the reason that property rights in creative expression exist is to guarantee the continued production of such works.[6] The fundamental tradeoff in copyright policy is between the costs of limiting access to creative works, and the value obtained by encouraging production of such works.[7] The same applies in the context of AI: identifying the key tradeoffs and weighing the costs and benefits of restricting access to protected works by the producers (and users) of AI systems.[8]

This entails examining the costs and benefits of relatively stronger or weaker forms copyright protection in terms of their effects on both incentives and access, and as they relate to both copyright holders and AI-system developers. It also requires considering where the transaction costs should be allocated for negotiating access to both copyright and, as discussed infra,[9] the use of name/image/likeness, as well as how those allocations are likely to shape outcomes.

At root, these questions center on how to think about the property rights that limit access to protected works and, possibly even more importantly, how to assign new property rights governing the ability to control the use of a name/image/likeness. As we know from the work of the late Nobel laureate Ronald Coase, the actual demarcation of rights affects parties’ abilities to negotiate superior solutions.[10] The development of nuisance law provides a good example of the problem at hand. When a legal regime provides either strict liability or no-liability rules around pollution, parties have little incentive to minimize harmful conduct:

The factory that has the absolute right to pollute will, if transaction costs are prohibitive, have no incentives to stop (or reduce) pollution even if the cost of stopping would be much less than the cost of pollution to the homeowners. Conversely, homeowners who have an absolute right to be free from pollution will, if transaction costs are prohibitive, have no incentive to take steps of their own to reduce the effects of pollution even if the cost to them of doing so (perhaps by moving away) is less than the cost to the factory of not polluting or of polluting less.[11]

As Coase observed, this class of problem is best regarded as reciprocal in nature, and the allocation of rights matters in obtaining an efficient outcome. This is necessarily so because, when fully considered, B’s ability to restrain A from the pollution-generating activity can itself be conceived of as another kind of harm that B can impose on A. Therefore, the problem requires a balancing of the relative harms generated by both A and B in exercising conflicting claims in a particular context.

When thinking about how to minimize harms—whether from pollution or other activity that generates social costs (which is to say, nearly every activity)—the aim is to decide whether “the gain from preventing the harm is greater than the loss which would be suffered elsewhere as a result of stopping the action which produces the harm.”[12] Theoretically, in a world without transaction costs, even assignments of no-liability or strict-liability rules could be bargained around. But we do not live in such a world.[13] Thus, “[i]n a world in which there are costs of rearranging the rights established by the legal system [common law and statutory assignments of liability] are, in effect, making a decision on the economic problem and determining how resources are to be employed.”[14]

While pollution rules, unlicensed uses of intellectual property, and a host of other activities subject to legal sanction are not typically framed as resource-allocation decisions, it is undeniable that they do have this character. This is true even where legislation attempts to correct deficiencies in the system. We experience a form of blindness when we focus on correcting what may be rightly perceived as problems in a liability regime. Such analysis tends to concentrate attention on particular deficiencies of the system and to nourish the belief that any measure that removes the deficiency is necessarily desirable. It diverts attention from other changes inevitably associated with the corrective measure—changes that may well produce more harm than the original deficiency.[15]

All of this is to say that one solution to the costs generated by the need for AI systems to process a massive corpus of expensive, copyright-protected material is neither to undermine property rights, nor to make AI impossible, but to think about how new property rights could make the system work. It may be that some entirely different form or allocation of property right would facilitate bargaining between rightsholders and AI creators, optimizing resource allocation in a way the existing doctrinal regime may not be able to.

A number of other questions flow from this insight into the allocative nature of copyright. How would the incentives for human creators change under different copyright rules for AI systems, or in the face of additional rights? And how would access to copyrighted works for AI training change with different rules, and what effects would that access have on AI innovation?

Above all, our goal today should be to properly frame the AI and copyright debate by identifying tradeoffs, quantifying effects (where possible), and asking what rules best serve the overall objectives of the copyright system and the social goal of encouraging AI innovation. The best chance of striking the right balance will come from a rigorous framing of the questions and from the use of economic analysis to try to answer them.

B.            Copyright Law and AI: Moving Forward

As the Copyright Office undertakes this inquiry, it is important to recognize that, regardless of how the immediate legal questions around AI and copyright are resolved, the growing capabilities and adoption of generative AI systems will likely necessitate some changes in the long term.

The complex questions surrounding the intersection of AI and copyright law admit reasonable arguments on both sides. But AI is here to stay, regardless, and if copyright law is applied in an unduly restrictive manner that substantially hinders socially beneficial AI innovation, it could provoke a broader public-policy backlash that does more to harm copyright’s ability to protect creative works than it does to stanch AI’s ability to undermine it. Copyright law risks being perceived as an obstruction to technological progress if it is used preemptively to kill AI in the cradle. Such an outcome could galvanize calls for recalibrating copyright’s scope and protections in the name of the public interest.

This illustrates the precarious balancing act that copyright law faces in the wake of rapidly evolving technologies like AI. Aggressive copyright restrictions that curtail AI development could instigate a public-policy counter-reaction before Congress and the courts that ultimately undermines copyright’s objectives. The judicious course is to adapt copyright law cautiously to enable AI’s responsible evolution, while resolutely preserving the incentives for human creativity.

In the remainder of this analysis, we offer our perspective on the likely outcomes of the AI-copyright issues raised in this NOI, given the current state of the law. These assessments reflect our perspective formed through the rigorous application of established copyright principles and precedent to the novel technological context of generative AI systems. Reasonable arguments rooted in existing doctrine could be made to support different conclusions. We submit these comments not as definitive predictions or normative preferences, but rather as informed appraisals of how courts may analyze AI under present copyright law, absent legislative intervention.

We appreciate the Copyright Office starting this process to modernize copyright law for the AI age. This inquiry is an important first step, but openness to further evolution will be key to promoting progress in both AI and the arts. We believe an open, evidence-based discussion of these issues will lead to balanced solutions that uphold copyright’s constitutionally mandated purpose, while allowing responsible AI innovation for the public benefit.

II.            The Training of AI Systems and the Applicability of Fair Use

In the NOI, the Copyright Offices asks: “[u]nder what circumstances would the unauthorized use of copyrighted works to train AI models constitute fair use?”[16]

To answer this question, it would be useful to first briefly walk through a high-level example of how AI systems work, in order to address the most relevant points of contact between AI systems and copyright law.

A.            A Brief Technical Description of AI Training

AI-generated content is not a single “thing,” but a collection of differing processes, each with different implications for the law. For the purposes of this discussion, we will discuss image generation using “generated adversarial networks” (GANs) and diffusion models. Although different systems and different types of content generation will vary, the basic concepts discussed below are nonetheless useful at a general level.[17]

A GAN is a type of machine-learning model that consists of two parts: a generator and a discriminator.[18] The generator is trained to create new images that look like they come from a particular dataset, while the discriminator is trained to distinguish the generated images from real images in its original dataset.[19] The two parts are trained together in an adversarial manner, with the generator trying to produce images that can fool the discriminator and the discriminator trying to correctly identify the generated images.[20]

A diffusion model, by contrast, analyzes the distribution of information in an image, as noise is progressively added to it.[21] This kind of algorithm analyzes characteristics of sample images, like the distribution of colors or lines, in order to understand what counts as an accurate representation of a subject (i.e., what makes a picture of a cat look like a cat, and not like a dog).[22]

For example, in the generation phase, diffusion-based systems start with randomly generated noise, and work backward in “denoising” steps to essentially “see” shapes:

The sampled noise is predicted so that if we subtract it from the image, we get an image that’s closer to the images the model was trained on (not the exact images themselves, but the distribution – the world of pixel arrangements where the sky is usually blue and above the ground, people have two eyes, cats look a certain way – pointy ears and clearly unimpressed).[23]

While it is possible that some implementations might be designed in a way that saves copies of the training images,[24] for at least some systems, once the network is trained using these techniques, it will not need to rely on saved copies of input work in order to produce outputs. The models that are produced during training are, in essence, instructions to a different piece of software about how to start with a prompt from a user, a palette of pure noise, and progressively “discover” signal in that image until some new image emerges.

B.            Fair Use

The creator of some of the most popular AI tools, OpenAI, is not shy about their use of protected works in the training phase of the algorithms. In comments to the U.S. Patent and Trademark Office (PTO), OpenAI noted that:

Modern AI systems require large amounts of data. For certain tasks, that data is derived from existing publicly accessible “corpora”… of data that include copyrighted works. By analyzing large corpora (which necessarily involves first making copies of the data to be analyzed), AI systems can learn patterns inherent in human-generated data and then use those patterns to synthesize similar data which yield increasingly compelling novel media in modalities as diverse as text, image, and audio. (emphasis added).[25]

Thus, at the training stage, the most popular forms of AI systems require making copies of existing works. And where that material is either not in the public domain or is not licensed, an infringement can occur. Thus, the copy must not be infringing (say, because it is transient), or some affirmative defense is needed to excuse the infringement. Toward this end, OpenAI believes that this use should qualify as fair use,[26] as do most or all the other major producers of generative AI systems.[27]

But as OpenAI has framed the fair-use analysis, it is not clear that these uses should qualify. There are two major questions in this respect: will the data used to train these systems count as “copies” under the Copyright Act, and, if so, is the use of these “copies” sufficiently “transformative” to qualify for the fair-use defense?

1.              Are AI systems being trained with ‘copies’ of protected works?

Section 106 of the Copyright Act grants the owner of a copyright the exclusive right “to reproduce… copyrighted work in copies” and to authorize others to do so.[28] If an AI system makes a copy of a file to a computer during training, this would likely constitute a prima facie violation of the copyright owner’s exclusive right of reproduction under Section 106. This is fairly straightforward.

But what if the “copy” is “transient” and/or only partial pieces of content are used in the training? For example, what if a training program merely streamed small bits of a protected work into temporary memory as part of its training, and retained no permanent copy?

As the Copyright Office has previously observed, even temporary reproductions of a work in a computer’s memory can constitute “copies” under the Copyright Act.[29] Critically, this includes even temporary reproductions made as part of a packet-switching network transmission, where a particular file is broken into individual packets, because the packets can be reassembled into substantial portions or even entire works.[30] On the topic of network-based transmission, the Copyright Office further observed that:

Digital networks permit a single disk copy of a work to meet the demands of many users by creating multiple RAM copies. These copies need exist only long enough to be perceived (e.g., displayed on the screen or played through speakers), reproduced or otherwise communicated (e.g., to a computer’s processing unit) in order for their economic value to be realized. If the network is sufficiently reliable, users have no need to retain copies of the material. Commercial exploitation in a network environment can be said to be based on selling a right to perceive temporary reproductions of works.[31]

This is a critical insight that translates well to the context of AI training. The “transience” of the copy matters with respect to the receiver’s ability to perceive the work in a way that yields commercial value. Under this reasoning, the relevant locus of analysis is on the AI system’s ability to “perceive” a work for the purposes of being trained to “understand” the work. In this sense, you could theoretically find the existence of even more temporary copies than that necessary for human perception to implicate the reproduction right.

Even where courts have been skeptical of extending the definition of “copy” to “fleeting” copies in computer memory, this underlying logic is revealed. In Cartoon Network LP, LLLP v. CSC Holdings, Inc., 536 F.3d 121 (2008), the 2nd U.S. Circuit Court of Appeals had to determine whether buffered media sent to a DVR device was too “transient” to count as a “copy”:

No bit of data remains in any buffer for more than a fleeting 1.2 seconds. And unlike the data in cases like MAI Systems, which remained embodied in the computer’s RAM memory until the user turned the computer off, each bit of data here is rapidly and automatically overwritten as soon as it is processed. While our inquiry is necessarily fact-specific, and other factors not present here may alter the duration analysis significantly, these facts strongly suggest that the works in this case are embodied in the buffer for only a “transitory” period, thus failing the duration requirement.[32]

In Cartoon Network, the court acknowledged both that the duration analysis was fact-bound, and also that the “fleeting” nature of the reproduction was important. “Fleeting” is a relative term, based on the receiver’s capacities. A ball flying through the air may look “fleeting” to a human observer, but may appear to go much more cognizable to a creature with faster reaction time, such as a house fly. So, too, with copies of a work in a computer’s memory and the ability to “perceive” what is fixed in a buffer: what may be much too quick for a human to perceive may very well be within an AI system’s perceptual capabilities.

Therefore, however the training copies are held, there is a strong possibility that a court will find them to be “copies” for the purposes of the reproduction right—even with respect to partial copies that exist for very small amounts of time.

2.              The purpose and character of using protected works to train AI systems

Fair use provides for an affirmative defense against infringement when the use is, among other things, “for purposes such as criticism, comment, news reporting, teaching…, scholarship, or research.”[33] When deciding whether a fair-use defense is applicable, a court must balance a number of factors:

  1. the purpose and character of the use, including whether such use is of a commercial nature or is for nonprofit educational purposes;
  2. the nature of the copyrighted work;
  3. the amount and substantiality of the portion used in relation to the copyrighted work as a whole; and
  4. the effect of the use upon the potential market for or value of the copyrighted work.[34]

The fair-use defense that AI creators have advanced is rooted in the first factor: the nature and character of the use. Although a full analysis of all the factors is ultimately necessary, analysis of the first factor is sufficiently complicated to warrant full attention here. In particular, the complex issue at hand is whether uses of protected works to train AI systems are sufficiently “transformative” or not.[35]

Whether the use of a copyrighted work to train an AI is “transformative” is certainly a novel question, but it is one that will likely be answered in light of an observation the U.S. Supreme Court made in Campbell v. Acuff Rose Music:

[W]hen a commercial use amounts to mere duplication of the entirety of an original, it clearly “supersede[s] the objects,”… of the original and serves as a market replacement for it, making it likely that cognizable market harm to the original will occur… But when, on the contrary, the second use is transformative, market substitution is at least less certain, and market harm may not be so readily inferred.[36]

Moreover, “[t]he word ‘transformative’ cannot be taken too literally as a sufficient key to understanding the elements of fair use. It is rather a suggestive symbol for a complex thought, and does not mean that any and all changes made to an author’s original text will necessarily support a finding of fair use.”[37] A key question, then, is whether training AI systems on copyrighted works amounts to a mere “duplication of the entirety of an original” or is sufficiently “transformative” to support a fair-use defense. As noted above, OpenAI believes that its use is transformative. According to its comments:

Training of AI systems is clearly highly transformative. Works in training corpora were meant primarily for human consumption for their standalone entertainment value. The “object of the original creation,” in other words, is direct human consumption of the author’s ?expression.? Intermediate copying of works in training AI systems is, by contrast, “non-expressive” the copying helps computer programs learn the patterns inherent in human-generated media. The aim of this process—creation of a useful generative AI system—is quite different than the original object of human consumption. The output is different too: nobody looking to read a specific webpage contained in the corpus used to train an AI system can do so by studying the AI system or its outputs. The new purpose and expression are thus both highly transformative.[38]

This framing, however, works against OpenAI’s interests. As noted above, and reinforced in the immediately preceding quote, generative AI systems are made of at least two distinct pieces. The first is a piece of software that ingests existing works and creates a file that can serve as instructions to the second piece of software. The second piece of software takes the output of the first and can produce independent results. Thus, there is a clear discontinuity in the process whereby the ultimate work created by the system is disconnected from the creative inputs used to train the software.

Therefore, the protected works are arguably ingested into the first part of the system “for their standalone entertainment value.” That is to say, the goal of copying and showing a protected work to an AI system is for the analog of “direct human consumption of the author’s expression” in order for the system to learn about that expression.

The software is learning what counts as “standalone entertainment value” and therefore the works must be used in those terms. Surely, a computer is not sitting on a couch and surfing for its pleasure. But it is solely for the very “standalone entertainment value” that the first piece of software is being shown copyrighted material. By contrast, parody or “remixing” uses incorporate a work into some secondary expression that directly transforms the input. The way these systems work is to learn what makes a piece entertaining and then to discard that piece altogether. Moreover, this use for the art qua art most certainly interferes with the existing market, insofar as this use is in lieu of reaching a licensing agreement with rightsholders.

A good analogy is art students and art textbooks. Art students view protected works in an art textbook in order to learn how to reproduce the styles contained therein. The students would not be forgiven for pirating the textbooks merely because they intend to go on to make new paintings. They would still be liable for copyright infringement if they used unlicensed protected works as part of their education.

The 2nd U.S. Circuit Court of Appeals dealt with a case that operates similarly to this dynamic. In American Geophysical Union v. Texaco, 60 F.3d 913 (2d Cir. 1994), the 2nd Circuit considered whether Texaco’s photocopying of scientific articles produced by the plaintiffs qualified for a fair-use defense. Texaco employed between 400 and 500 research scientists and, as part of supporting their work, maintained subscriptions to a number of scientific journals.[39]

It was common practice for Texaco’s scientists to photocopy entire articles and save them in a file.[40] The plaintiffs sued for copyright infringement.[41] Texaco asserted that photocopying by its scientists for the purposes of furthering scientific research—that is to train the scientists on the content of the journal articles—should count as a fair use. The argument was, at least in part, that this was sufficiently “transformative,” because the scientists were using that knowledge to invent new products.[42] The 2nd Circuit disagreed:

The “transformative use” concept is pertinent to a court’s investigation under the first factor because it assesses the value generated by the secondary use and the means by which such value is generated. To the extent that the secondary use involves merely an untransformed duplication, the value generated by the secondary use is little or nothing more than the value that inheres in the original. Rather than making some contribution of new intellectual value and thereby fostering the advancement of the arts and sciences, an untransformed copy is likely to be used simply for the same intrinsic purpose as the original, thereby providing limited justification for a finding of fair use….[43]

The 2nd Circuit thus observed that copies of the scientific articles were made solely to consume the material itself. AI developers often make an argument analogous to that made by Texaco: that training AI systems surely advances scientific research, and therefore fosters the “advancement of the arts and sciences.” But in American Geophysical Union, the initial copying of copyrighted content, even where it was ultimately used for the “advancement of the arts and sciences,” was not held to be sufficiently “transformative.”[44] The case thus stands for the proposition that one cannot merely identify a social goal down that would be advanced at some future date in order to permit an exception to copyright protection. As the court put it:

[T]he dominant purpose of the use is a systematic institutional policy of multiplying the available number of copies of pertinent copyrighted articles by circulating the journals among employed scientists for them to make copies, thereby serving the same purpose for which additional subscriptions are normally sold, or… for which photocopying licenses may be obtained.[45]

The use itself must be transformative and different, and copying is not transformative merely because it may be used as an input into a later transformative use. By the same token, therefore, it seems likely that where an AI system ingests (copies) copyrighted works, that use is similarly not transformative, despite its ultimate use as an input in the creation of other original works.

Comparing the American Geophysical Union analysis with the search-engine “snippets” and “thumbnails” cases provides a useful comparison relevant to the AI analysis. In Kelly v. Arriba Soft Corp., 336 F.3d 811 (9th Cir. 2002), the 9th U.S. Circuit Court of Appeals ruled that a search engine’s creation of thumbnail images from original copies was a transformative fair use.[46] Arriba’s search-engine crawler made full-sized copies of Kelly’s images and stored them temporarily on Arriba’s server to generate thumbnail versions. After the thumbnails were created, the full-sized originals were deleted. The thumbnails were used to facilitate Arriba’s image-based search engine. In reaching its fair-use conclusion, the 9th Circuit opined that:

Arriba’s use of Kelly’s images promotes the goals of the Copyright Act and the fair use exception. The thumbnails do not stifle artistic creativity because they are not used for illustrative or artistic purposes and therefore do not supplant the need for the originals.[47]

Further, although “Arriba made exact replications of Kelly’s images, the thumbnails were much smaller, lower-resolution images that served an entirely different function than Kelly’s original images.”[48]

The court found it important that the search engine did not use the protected works for their intended “aesthetic experience,” but rather for the purpose of constructing a search index.[49] Indeed, the entire point of a search engine is not to “supersede” the original, but in many or most cases to provider users an efficient means to find that original online.[50]

The court discussed, but only briefly, the benefit to the public of Arriba’s transformative use,[51] noting that “[Arriba’s thumbnails] benefit the public by enhancing information-gathering techniques on the internet.”[52] Five years later, in Perfect 10 Inc. v. Amazon.com Inc., 487 F.3d 701 (2007), the 9th Circuit expanded on this question somewhat.[53] There, in holding that the novelty of the use was of crucial importance to the analysis,[54] the court also stressed that the value of that use was a function of its newness:

[A] search engine provides social benefit by incorporating an original work into a new work, namely, an electronic reference tool. Indeed, a search engine may be more transformative than a parody [the use at issue in Campbell] because a search engine provides an entirely new use for the original work, while a parody typically has the same entertainment purpose as the original work.[55]

Indeed, even in light of the commercial nature of Google’s use of copyrighted content in its search engine, its significant public benefit carried the day: “We conclude that the significantly transformative nature of Google’s search engine, particularly in light of its public benefit, outweighs Google’s superseding and commercial uses of the thumbnails in this case.”[56] And, of particular relevance to these questions in the context of AI, the court in Perfect 10 went on to “note the importance of analyzing fair use flexibly in light of new circumstances.”[57]

Ultimately, the Perfect 10 decision tracked Kelly fairly closely on the rest of the “transformativeness” analysis in finding fair use, because “[a]lthough an image may have been created originally to serve an entertainment, aesthetic, or informative function, a search engine transforms the image into a pointer directing a user to a source of information.”[58]

The core throughline in this line of cases is the question of whether a piece of content is being used for its expressive content, weighed against the backdrop of whether the use is for some new (and, thus, presumptively valuable) purpose. In Perfect 10 and Kelly, the transformative use was the creation of a search index.

“Snippets” fair-use cases track a similar line of reasoning. For example, in Authors Guild v. Google Inc., 804 F.3d 202 (2d Cir. 2015), the 2nd Circuit ruled that Google’s use of “snippets” of copyrighted books in its Library Project and Google Books website was a “transformative” fair use.[59] Holding that the “snippet view” of books digitized as part of the Google Books project did not constitute an effectively competing substitute to the original works, the circuit court noted that copying for the purpose of “criticism” or—as in that case—copying for the purpose of “provision of information about” the protected work, “tends most clearly to satisfy Campbell’s notion of the ‘transformative’ purpose.”[60]

Importantly, the court emphasized the importance of the public-benefit aspect of transformative uses: “[T]ransformative uses tend to favor a fair use finding because a transformative use is one that communicates something new and different from the original or expands its utility, thus serving copyright’s overall objective of contributing to public knowledge.”[61]

Underscoring the idea that the “transformativeness” analysis weighs whether a use is merely for expressive content against the novelty/utility of the intended use, the court observed:

Google’s division of the page into tiny snippets is designed to show the searcher just enough context surrounding the searched term to help her evaluate whether the book falls within the scope of her interest (without revealing so much as to threaten the author’s copyright interests). Snippet view thus adds importantly to the highly transformative purpose of identifying books of interest to the searcher.[62]

Thus, the absence of use of the work’s expressive content, coupled with a fairly circumscribed (but highly novel) use was critical to the outcome.

The entwined questions of transformative use and the public benefit it confers are significantly more complicated in the AI context, however. Unlike the incidental copying involved in search-engine indexing or thumbnails, training generative AI systems directly leverages copyrighted works for their expressive value. In the Google Books and Kelly cases, the defendant systems extracted limited portions of works or down-sampled images solely to identify and catalog their location for search purposes. The copies enabled indexing and access, and they expanded public knowledge through a means unrelated to the works’ protected aesthetics.

But in training AI models on copyrighted data, the systems necessarily parse the intrinsic creative expression of those works. The AI engages with the protected aesthetic elements themselves, not just superficial markers (like title, length, location on the internet, etc.), in order to internalize stylistic and compositional principles. This appropriates the heart of the works’ copyright protection for expressive ends, unlike the more tenuous connections in search systems.

The AI is thus “learning” directly from the protected expression in a manner akin to a human student studying an art textbook, or like the scientists learning from the journals in American Geophysical Union. The subsequent AI generations are built from mastery of the copyrighted training materials’ creative expression. Thus, while search-engine copies only incidentally interact with protected expression to enable unrelated innovation, AI training is predicated on excavating the protected expression itself to fuel iterative creation. These meaningfully different purposes have significant fair-use implications.

This functional difference is, as noted, central to the analysis of a use’s “purpose and character.” Indeed, “even making an exact copy of a work may be transformative so long as the copy serves a different function than the original work.”[63] But the benefit to the public from the new use is important, as well, particularly with respect to the possible legislative response that a restrictive interpretation of existing doctrine may engender.

If existing fair-use principles prohibit the copying required for AI, absent costly item-by-item negotiation and licensing, the transaction costs could become prohibitive, thwarting the development of technologies that promise great public value.[64] Copyright law has faced similar dilemmas before, where the transaction costs of obtaining permission for socially beneficial uses could frustrate those uses entirely.[65] In such cases, we have developed mechanisms like compulsory licensing to facilitate the necessary copying, while still attempting to compensate rightsholders. An unduly narrow fair-use finding for AI training could spur calls for similar interventions in service of enabling AI progress.

In other words, regardless of the veracity of the above conclusion that AI’s use of copyrighted works may not, in fact, serve a different function than the original, courts and legislators may be reluctant to allow copyright doctrine to serve as an absolute bar against self-evidently valuable activity like AI development. Our aim should be to interpret or recalibrate copyright law to permit such progress while upholding critical incentives for creators.

C.            Opt-In vs. Opt-Out Use of Protected Works

The question at the heart of the prior discussion—and, indeed, at the heart of the economic analysis of copyright—is whether the transaction costs that accompany requiring express ex ante permission for the use of protected works are so high that they impedes socially beneficial conduct whose value would outweigh the social cost of allowing permissionless and/or uncompensated use.[66] The NOI alludes to this question when it asks: “Should copyright owners have to affirmatively consent (opt in) to the use of their works for training materials, or should they be provided with the means to object (opt out)?”[67]

This is a complex problem. Given the foregoing thoughts on fair use, it seems quite possible that, at present, the law requires creators of AI systems to seek licenses for protected content, or else must resort to public-domain works for training. Given the volume of copyrighted works that AI developers currently use to train these systems, such requirements may be broadly infeasible.

On one hand, requiring affirmative opt-in consent from copyright holders imposes significant transaction costs on AI-system developers to identify and negotiate licenses for the vast amounts of training data required. This could hamper innovation in socially beneficial AI systems. On the other hand, an opt-out approach shifts more of the transaction-cost burden to copyright holders, who must monitor and object to unwanted uses of their works. This raises concerns about uncompensated use.

Ultimately, the question is where the burden should lie: with AI-system developers to obtain express consent, or with copyright holders to monitor and object to uses? Requiring some form of consent may be necessary to respect copyright interests. Yet an opt-out approach may strike the right balance, by shifting some of the burden back to AI developers while avoiding the infeasibly high transaction costs of mandatory opt-in consent. The optimal approach likely involves nuanced policymaking to balance these competing considerations. Moreover, as we discuss infra, the realistic outcome is most likely going to require rethinking the allocation of property rights in ways that provide for large-scale licensing. Ideally, this could be done through collective negotiation, but perhaps at a de minimis rate, while allowing creators to bargain for remuneration on the basis of other rights, like a right of publicity or other rights attached to the output of AI systems, rather than the inputs.[68]

1.              Creator consent

Relatedly, the Copyright Office asks: “If copyright owners’ consent is required to train generative AI models, how can or should licenses be obtained?”[69]

Licensing markets exist, and it is entirely possible that major AI developers and large groups of rightsholders can come to mutually beneficial terms that permit a sufficiently large body of protected works to be made available as training data. Something like a licensing agency for creators who choose to make their works available could arise, similar to the services that exist to provide licensed music and footage for video creators.[70] It is also possible for some to form collective-licensing organizations to negotiate blanket permissions covering many works.

It’s important to remember that our current thinking is constrained by our past experience. All we know today are AI models trained on vast amounts of unlicensed works. It is entirely possible that, if firms were required to seek licenses, unexpected business models would emerge to satisfy both sides of the equation.

For example, an AI firm could develop its own version of YouTube’s ContentID, which would allow creators to control when their work is used in AI training. For some well-known artists, this could be negotiated with an upfront licensing fee. On the user side, any artist who has opted in could then be selected as a “style” for the AI to emulate—triggering a royalty payment to the artist when a user generates an image or song in that style. Creators could also have the option of removing their influence from the system if they so desire.

Undoubtedly, there are other ways to structure the relationship between creators and AI systems  that would facilitate creators’ monetization of the use of their work in AI systems, including legal and commercial structures that create opportunities for both creators and AI firms to succeed.

III.          Generative AI Outputs: Protection of Outputs and Outputs that Infringe

The Copyright Office asks: “Under copyright law, are there circumstances when a human using a generative AI system should be considered the ‘author’ of material produced by the system?”[71]

Generally speaking, we see no reason why copyright law should be altered to afford protection to purely automatic creations generated by AI systems. That said, when a human makes a nontrivial contribution to generative AI output—such as editing, reframing, or embedding the AI-generated component within a larger work—the resulting work should qualify for copyright protection.

Copyright law centers on the concept of original human authorship.[72] The U.S. Constitution expressly limits copyright to “authors.”[73] As of this writing, however, generative AI’s capacities do not rise to the level of true independent authorship. AI systems remain tools that require human direction and judgment.[74] As such, when a person provides the initial prompt or framing, makes choices regarding the iterative development of the AI output, and decides that the result is satisfactory for inclusion in a final work, they are fundamentally engaging in creative decision making that constitutes authorship under copyright law.

As Joshua Gans has observed of recent Copyright Review Board decisions:

Trying to draw some line between AI and humans with the current technology opens up a massive can of worms. There is literally no piece of digital work these days that does not have some AI element to it, and some of these mix and blur the lines in terms of what is creative and what is not. Here are some examples:

A music artist uses AI to denoise a track or to add an instrument or beat to a track or to just get a composition started.

A photographer uses Photoshop or takes pictures with an iPhone that already uses AI to focus the image and to sort a burst of images into one that is appropriate.

A writer uses AI to prompt for some dialogue when stuck at some point or to suggest a frame for writing a story.[75]

Attempting to separate out an “AI portion” from the final work, as the Copyright Review Board proposed, fundamentally misunderstands the integrated nature of the human-AI collaborative process. The AI system cannot function without human input, and its output remains raw material requiring human creativity to incorporate meaningfully into a finished product.

Therefore, when a generative AI system is used as part of a process guided by human creative choices, the final work should be protected by copyright, just as a work created using any other artistic tool or collaborator would be. Attenuating copyrightability due to the use of AI would undermine basic copyright principles and fail to recognize the essentially human nature of the creative process.

A.            AI Outputs and Infringement

The NOI asks: “Is the substantial similarity test adequate to address claims of infringement based on outputs from a generative AI system, or is some other standard appropriate or necessary?” (Question 23)

The outputs of AI systems may or may not violate IP laws, but there is nothing inherent in the processes described above that dictates that they must. As noted, the most common AI systems do not save copies of existing works, but merely “instructions” (more or less) on how to create new work that conforms to patterns found by examining existing work. If we assume that a system isn’t violating copyright at the input stage, it’s entirely possible that it can produce completely new pieces of art that have never before existed and do not violate copyright.

They can, however, be made to violate copyrights. For example, these systems can be instructed to generate art, not just in the style of a particular artist, but art that very closely resembles existing pieces. In this sense, it would be making a copy that theoretically infringes. The fact of an AI’s involvement would not change the analysis: just as with a human-created work, if it is substantially similar to a copyrighted work, it may be found infringing.

There is, however, a common bug in AI systems that leads to outputs that are more likely to violate copyright in this way. Known as “overfitting,” the training leg of these AI systems can be presented with samples that contain too many instances of a particular image.[76] This leads to a dataset that contains too much information about the specific image, such that—when the AI generates a new image—it is constrained to producing something very close to the original. Similarly, there is evidence that some AI systems are “memorizing” parts of protected books.[77] This could lead to AI systems repeating copyright-protected written works.

1.              The substantial-similarity test

The substantial-similarity test remains functionally the same when evaluating works generated using AI. To find “substantial similarity,” courts require evidence of copying, as well as an expression that is substantially similar to a protected work.[78] “It is now an axiom of copyright law that actionable copying can be inferred from the defendant’s access to the copyrighted work and substantial similarity between the copyrighted work and the alleged infringement.”[79] In many or most cases, it will arguably be the case that AI systems have access to quite a wide array of protected works that are posted online. Thus, there may not be a particularly high hurdle to determine that an AI system actually copied a protected work.

There is, however, one potential problem for the first prong of this analysis. Models produced during a system’s training process do not (usually) contain the original work, but are the “ideas” that the AI systems generated during training. Thus, where the provenance of works contained in a training corpus is difficult to source, it may not be so straightforward to make inferences about whether a model “saw” a particular work. This is because the “ideas” that the AI “learns” from its training corpus are unprotected under U.S. copyright law, as it is permissible to mimic unprotected elements of a copyrighted work (such as ideas).[80]

Imagine a generative AI system trained on horror fiction. It would be possible for this system to produce a new short story that is similar to one written by Stephen King, but the latent data in the model almost certainly would not violate any copyrights that King holds in his work. The model would contain “ideas” about horror stories, including those learned from an array of authors who were themselves influences on Stephen King, and potentially some of King’s own stories. What the AI system “learns” in this case is the relationship between words and other linguistic particularities that are commonly contained in horror fiction. That is, it has “ideas” about what goes into a horror story, not (theoretically) the text of the horror story itself.

Thus, when demonstrating indirect proof of copying in the case of a Stephen King story, it may pose a difficulty that an AI system has ingested all of H.P. Lovecraft’s work—an author who had a major influence on King. The “ideas” in the model and the output it subsequently produces may, in fact, produce something similar to a Stephen King work, but it may have been constructed largely or entirely on material from Lovecraft and other public-domain horror writers. The problem becomes only more complicated when you realize that this system could also have been trained on public-domain fan fiction written in the style of Stephen King. Thus, for the purposes of the first prong of this analysis, courts may place greater burden on plaintiffs in copyright actions against model producers to demonstrate more than merely that a work was merely available online.

Assuming that plaintiffs are able to satisfy the first prong, once an AI system “expresses” those ideas, that expression could violate copyright law under the second prong of the substantial-similarity test. The second prong inquires whether the final work appropriated the protected original expression.[81] Any similarities in unprotectable ideas, facts, or common tropes are disregarded.[82] So, in both traditional and AI contexts, the substantial-similarity test ultimately focuses on the protected components of creative expression, not surface similarity.

The key determination is whether the original work’s protected expression itself has been impermissibly copied, no matter the process that generated the copy. AI is properly viewed as simply another potential tool that could be used in certain acts of copying. It does not require revisiting settled principles of copyright law.

B.            Direct and Secondary Liability

The NOI asks: “If AI-generated material is found to infringe a copyrighted work, who should be directly or secondarily liable—the developer of a generative AI model, the developer of the system incorporating that model, end users of the system, or other parties?”[83]

Applying traditional copyright-infringement frameworks to AI-generated works poses unique challenges in determining direct versus secondary liability. In some cases, the AI system itself may create infringing content without any direct human causation.

1.              Direct liability

If the end user prompts an AI system in a way that intentionally targets copyrighted source material, they may meet the threshold for direct infringement by causing the AI to reproduce protected expression.[84] Though many AI prompts contain only unprotected ideas, users may sometimes input copyrightable material as the basis for the AI output. For example, a user could upload a copyrighted image and request the AI to make a new drawing based on the sample. In such cases, the user is intentionally targeting copyrighted works and directly “causing” the AI system to reproduce output that is similar. If sufficiently similar, that output could infringe on the protected input. This would be a question of first impression, but it is a plausible reading of available cases.

For example, in CoStar Grp. Inc. v. LoopNet Inc., 373 F.3d 544 (4th Cir. 2004), the 4th U.S. Circuit Court of Appeals had to consider whether an internet service provider (ISP) could be directly liable when third parties reposted copyrighted material owned by the plaintiff. In determining that merely owning the “machine” through which copies were made or transmitted was not enough to “cause” a direct infringement, the court held that:

[T]o establish direct liability under §§ 501 and 106 of the Act, something more must be shown than mere ownership of a machine used by others to make illegal copies. There must be actual infringing conduct with a nexus sufficiently close and causal to the illegal copying that one could conclude that the machine owner himself trespassed on the exclusive domain of the copyright owner. The Netcom court described this nexus as requiring some aspect of volition or causation… Indeed, counsel for both parties agreed at oral argument that a copy machine owner who makes the machine available to the public to use for copying is not, without more, strictly liable under § 106 for illegal copying by a customer. The ISP in this case is an analogue to the owner of a traditional copying machine whose customers pay a fixed amount per copy and operate the machine themselves to make copies. When a customer duplicates an infringing work, the owner of the copy machine is not considered a direct infringer. Similarly, an ISP who owns an electronic facility that responds automatically to users’ input is not a direct infringer.[85]

Implied in the 4th Circuit’s analogy is that, while the owner of a copying machine might not be a direct infringer, a user employing such a machine could be a direct infringer. It’s an imperfect analogy, but a user of an AI system prompting it to create a “substantially similar” reproduction of a protected work could very well be a direct infringer under this framing. Nevertheless, the analogy is inexact, because the user feeds an original into a copying machine in order to make a more-or-less perfect copy of the original, whereas an AI system generates something new but similar. The basic mechanism of using a machine to try to reproduce a protected work, however, remains essentially the same. Whether there is an infringement would be a question of “substantial similarity.”

2.              Secondary liability

As in the case of direct liability, the nature of generative AI makes the secondary-liability determination slightly more complicated, as well. That is, paradoxically, the basis for secondary liability could theoretically arise even where there was no direct infringement.[86]

The first piece of this analysis is relatively easier. If a user is directly liable for infringing a protected work, as noted above, the developer and provider of a generative AI system may face secondary copyright liability. If the AI developer or distributor knows the system can produce infringing outputs, and provides tools or material support that allows users to infringe, it may be liable for contributory infringement.[87] Critically, merely designing a system that is capable of infringing is not enough to find contributory liability.[88]

An AI producer or distributor may also have vicarious liability, insofar as it has the right and ability to supervise users’ activity and a direct financial interest in that activity.[89] AI producers have already demonstrated their ability to control users’ behavior to thwart unwanted uses of the service.[90] Thus, if there is a direct infringement by a user, a plausible claim for vicarious liability could be made so long as there is sufficient connection between the user’s behavior and the producer’s financial interests.

The question becomes more complicated when a user did not direct the AI system to infringe. When the AI generates infringing content without user direction, it’s not immediately clear who would be liable for the infringement.[91] Consider the case where, unprompted by either the user or the AI producer, an AI system creates an output that would infringe under the substantial-similarity test. Assuming that the model has not been directed by the producer to “memorize” the works it ingests, the model itself consists of statistical information about the relationship between different kinds of data. The infringer, in a literal sense, is the AI system itself, as it is the creator of the offending output. Technically, this may be a case of vicarious liability, even without an independent human agent causing the direct infringement.

We know that copyright protection can only be granted to humans. As the Copyright Review Board recently found in a case deciding whether AI-generated outputs can be copyrighted:

The Copyright Act protects, and the Office registers, “original works of authorship fixed in any tangible medium of expression.” 17 U.S.C. § 102(a). Courts have interpreted the statutory phrase “works of authorship” to require human creation of the work.[92]

But can an AI system directly violate copyright? In his Aereo dissent, Justice Clarence Thomas asserted that it was a longstanding feature of copyright law that violation of the performance right required volitional behavior.[93] But the majority disagreed with him, holding that, by running a fully automated system of antennas intended to allow users to view video at home, the system gave rise to direct copyright liability.[94] Thus, implied in the majority’s opinion is the idea that direct copyright infringement does not require “volitional” conduct.

It is therefore plausible that a non-sentient, fully automated AI system could infringe copyright, even if, ultimately, there is no way to recover against the nonhuman agent. That does, however, provide an opportunity for claims of vicarious liability against the AI producer or distributor— at least, where the producer has the power to control the AI system’s behavior and that behavior appears to align with the producer’s financial interests.

3.              Protecting the ‘style’ of human creators

The NOI asks: “Are there or should there be protections against an AI system generating outputs that imitate the artistic style of a human creator (such as an AI system producing visual works ‘in the style of’ a specific artist)?”[95]

At the federal level, one candidate for protection against AI imitating some aspects of a creator’s works can currently be found in trademark law. Trademark law, governed by the Lanham Act, protects names, symbols, and other source identifiers that distinguish goods and services in commerce.[96] Unfortunately, a photograph or likeness, on its own, typically does not qualify for trademark protection, unless it is consistently used on specific goods.[97] Even where there is a likeness (or similar “mark”) used consistently as part of branding a distinct product, many trademark-infringement claims would be difficult to establish in this context, because trademark law does little to protect many aspects of a creator’s work.

Moreover, the Supreme Court has been wary about creating a sort of “mutant copyright” in cases that invoke the Lanham Act as a means to enforce a sort of “right of attribution,” which would potentially give creators the ability to control the use of their name in broader contexts.[98] In this context, the Court has held that the relevant parts of the Lanham Act were not designed to “protect originality or creativity,”[99] but are focused solely on “actions like trademark infringement that deceive consumers and impair a producer’s goodwill.”[100]

In many ways, there is a parallel here to the trademark cases involving keyword bidding in online ads. At a high level, search engines and other digital-advertising services do not generally infringe trademark when they allow businesses to purchase ads triggered by a user’s search for competitor trademarks (i.e., rivals’ business names).[101] But in some contexts, this can be infringing—e.g., where the use of trademarked terms in combination with advertising text can mislead consumers about the origin of a good or service.[102]

Thus, the harm, when it arises, would not be in a user asking an AI system to generate something “in the style of” a known creator, but when that user subsequently seeks to release a new AI-generated work and falsely claims it originated from the creator, or leaves the matter ambiguous and misleading to consumers.

Alternative remedies for creators could be found in the “right of publicity” laws in various states. A state-level right of publicity “is not merely a legal right of the ‘celebrity,’ but is a right inherent to everyone to control the commercial use of identity and persona and recover in court damages and the commercial value of an unpermitted taking.”[103] Such rights are recognized under state common law and statutes, which vary considerably in scope across jurisdictions—frequently as part of other privacy statutes.[104] For example, some states only protect an individual’s name, likeness, or voice, while others also cover distinctive appearances, gestures, and mannerisms.[105] The protections afforded for right-of-publicity claims vary significantly based on the state where the unauthorized use occurs or the individual is domiciled.[106] This creates challenges for the application of uniform nationwide protection of creators’ interests in the various aspects that such laws protect.

In recent hearings before the U.S. Senate Judiciary Subcommittee on Intellectual Property, several witnesses advocated creating a federal version of the right of publicity.[107] The Copyright Office has also previously opined that it may be desirable for Congress to enact some form of a “right of publicity” law.[108] If Congress chose to enact a federal “right of privacy” statute, several key issues would need to be addressed regarding the scope of protection, effect on state laws, constitutional authority, and First Amendment limitations.

Congress would have to delineate the contours of the federal right of publicity, including the aspects of identity covered and the types of uses prohibited. A broad right of privacy could protect names, images, likenesses, voices, gestures, distinctive appearances, and biographical information from any unauthorized commercial use. Or Congress could take a narrower approach focused only on particular identity attributes, like name and likeness. Congress would also need to determine whether a federal right-of-publicity statute preempts state right-of-publicity laws or sets a floor that would allow state protections to exceed the federal standards.

4.              Bargaining for the use of likenesses

A federal right of publicity could present an interesting way out of the current dispute between rightsholders and AI producers. Most of the foregoing comment attempts to pull apart different pieces of potential infringement actions, but such actions are only necessary, obviously, if a mutually beneficial agreement cannot be struck between creators and AI producers. The main issue at hand is that, given the vast amount of content necessary to train an AI system, it could be financially impractical for even the largest AI firms to license all the necessary content. Even if the comments above are correct, and fair use is not available, it could very well be the case that AI producers will not license very much content, possibly relying on public-domain material, and choosing to license only a very small selection.

Something like a “right of publicity,” or an equivalent agreement between creators and AI producers, could provide alternative licensing and monetization strategies that encourage cooperation between the parties. If creators had the opportunity to opt into the use of their likeness (or the relevant equivalent for the sort of AI system in question), the creators could generate revenue when the AI system actually uses the results of processing their content. Thus, the producers would not need to license content that contributes an unknown and possibly de minimis value to their systems, and would only need to pay for individual instances of use.

Indeed, in this respect, we are already beginning to see some experimentation with business models. The licensing of celebrity likenesses for Meta’s new AI chatbots highlights an emerging opportunity for creators to monetize their brand through contractual agreements that grant usage rights to tech companies that commercialize conversational AI.[109] As this technology matures, there will be more opportunities for collaborations between AI producers—who are eager to leverage reputable and recognizable personalities—and celebrities or influencers seeking new income streams.

As noted, much of the opportunity for creators and AI producers to reach these agreements will depend on how rights are assigned.[110] It may be the case that a “right of publicity” is not necessary to make this sort of bargaining happen, as creators could—at least theoretically—pursue litigation on a state-by-state basis. This disparate-litigation strategy could deter many creators, however, and it could also be the case that a single federal standard outlining a minimal property right in “publicity” could help to facilitate bargaining.

Conclusion

The advent of generative AI systems presents complex new public-policy challenges centered on the intersection of technology and copyright law. As the Copyright Office’s inquiry recognizes, there are open questions around the legal status of AI-training data, the attribution of AI outputs, and infringement liability, which all require thoughtful analysis.

Ultimately, maintaining incentives for human creativity, while also allowing AI systems to flourish, will require compromise and cooperation between stakeholders. Rather than an outright ban on the unauthorized use of copyrighted works for training data, a licensing market that enables access to a large corpora could emerge. Rightsholders may need to accept changes to how they typically license content. In exchange, AI producers will have to consider how they can share the benefit of their use of protected works with creators.

Copyright law retains flexibility to adapt to new technologies, as past reforms reacting to photography, sound recordings, software, and the internet all demonstrate. With careful balancing of interests, appropriate limitations, and respect for constitutional bounds, copyright can continue to promote the progress of science and the useful arts even in the age of artificial intelligence. This inquiry marks a constructive starting point, although ongoing reassessment will likely be needed as generative AI capabilities continue to advance rapidly.

[1] Artificial Intelligence and Copyright, Notice of Inquiry and Request for Comments, U.S. Copyright Office, Library of Congress (Aug. 30, 2023) [hereinafter “NOI”].

[2] Tim Sweeney (@TimSweeneyEpic), Twitter (Jan. 15, 2023, 3:35 AM), https://twitter.com/timsweeneyepic/status/1614541807064608768?s=46&t=0MH_nl5w4PJJl46J2ZT0Dw.

[3] Pulitzer Prize Winner and Other Authors Accuse OpenAI of Misusing Their Writing, Competition Policy International (Sep. 11, 2023), https://www.pymnts.com/cpi_posts/pulitzer-prize-winner-and-other-authors-accuse-openai-of-misusing-their-writing; Getty Images Statement, Getty Images (Jan. 17, 2023), https://newsroom.gettyimages.com/en/getty-images/getty-images-statement.

[4] See, e.g., Anton Oleinik, What Are Neural Networks Not Good At? On Artificial Creativity, 6 Big Data & Society (2019), available at https://journals.sagepub.com/doi/full/10.1177/2053951719839433#bibr75-2053951719839433.

[5] William M. Landes & Richard A. Posner, An Economic Analysis of Copyright Law, 18 J. Legal Stud. 325 (1989).

[6] Id. at 332.

[7] Id. at 326.

[8] Id.

[9] See infra, notes 102-103 and accompanying text.

[10] See generally R.H. Coase, The Problem of Social Cost, 3 J. L. & Econ. 1, 2 (1960).

[11] Richard Posner, Economic Analysis of Law (Aspen 5th ed 1998) 65, 79.

[12] Coase, supra note 9, at 27.

[13] Id.

[14] Id. at 27.

[15] Id. at 42-43.

[16] U.S. Copyright Office, Library of Congress, supra note 1, at 14.

[17] For more detailed discussion of GANs and Stable Diffusion see Ian Spektor, From DALL E to Stable Diffusion: How Do Text-to-image Generation Models Work?, Tryo Labs Blog (Aug. 31, 2022), https://tryolabs.com/blog/2022/08/31/from-dalle-to-stable-diffusion.

[18] Id.

[19] Id.

[20] Id.

[21] Id.

[22] Id.

[23] Jay Alammar, The Illustrated Stable Diffusion, Blog (Oct. 4, 2022), https://jalammar.github.io/illustrated-stable-diffusion.

[24] Indeed, there is evidence that some models may be trained in a way that they “memorize” their training set, to at least some extent. See, e.g., Kent K. Chang, Mackenzie Cramer, Sandeep Soni, & David Bamman, Speak, Memory: An Archaeology of Books Known to ChatGPT/GPT-4, arXiv Preprint (Oct. 20, 2023), https://arxiv.org/abs/2305.00118; OpenAI LP, Comment Regarding Request for Comments on Intellectual Property Protection for Artificial Intelligence Innovation, Before the USPTO, Dep’t of Com. (2019), available at https://www.uspto.gov/sites/default/files/documents/OpenAI_RFC-84-FR-58141.pdf.

[25] OpenAI, LP, Comment Regarding Request for Comments on Intellectual Property Protection for Artificial Intelligence, id. (emphasis added).

[26] 17 U.S.C. § 107.

[27] See, e.g., Blake Brittain, Meta Tells Court AI Software Does Not Violate Author Copyrights, Reuters (Sep. 19, 2023), https://www.reuters.com/legal/litigation/meta-tells-court-ai-software-does-not-violate-author-copyrights-2023-09-19; Avram Piltch, Google Wants AI Scraping to be ‘Fair Use.’ Will That Fly in Court?, Tom’s Hardware (Aug. 11, 2023), https://www.tomshardware.com/news/google-ai-scraping-as-fair-use.

[28] 17 U.S.C. § 106.

[29] Register of Copyrights, DMCA Section 104 Report (U.S. Copyright Office, Aug. 2001), at 108-22, available at https://www.copyright.gov/reports/studies/dmca/sec-104-report-vol-1.pdf.

[30] Id. at 122-23.

[31] Id. at 112 (emphasis added).

[32] Id. at 129–30.

[33] 17 U.S.C. § 107.

[34] Id.; see also Campbell v. Acuff-Rose Music Inc., 510 U.S. 569 (1994).

[35] Critically, a fair use analysis is a multi-factor test, and even within the first factor, it’s not a mandatory requirement that a use be “transformative.” It is entirely possible that a court balancing all of the factors could indeed find that training AI systems is fair use, even if it does not hold that such uses are “transformative.”

[36] Campbell, supra note 22, at 591.

[37] Authors Guild v. Google, Inc., 804 F.3d 202, 214 (2d Cir. 2015).

[38] OpenAI submission, supra note 13, at 5.

[39] Id. at 915.

[40] Id.

[41] Id.

[42] Id. at 933-34.

[43] Id. at 923. (emphasis added)

[44] Id.

[45] Id. at 924.

[46] Kelly v. Arriba Soft Corp., 336 F.3d 811 (9th Cir. 2002).

[47] Id.

[48] Id. at 818.

[49] Id.

[50] Id. at 819 (“Arriba’s use of the images serves a different function than Kelly’s use—improving access to information on the internet versus artistic expression.”).

[51] The “public benefit” aspect of copyright law is reflected in the fair-use provision, 17 U.S.C. § 107. In Campbell v. Acuff-Rose Music, Inc., 510 U.S. 569, 579 (1994), the Supreme Court highlighted the “social benefit” that a use may provide, depending on the first of the statute’s four fair-use factors, the “the purpose and character of the use.”

[52] Supra note 46, at 820.

[53] Perfect 10 Inc. v. Amazon.com Inc., 487 F.3d 701 (9th Cir., 2007)

[54] Id. at 721 (“Although an image may have been created originally to serve an entertainment, aesthetic, or informative function, a search engine transforms the image into a pointer directing a user to a source of information.”).

[55] Id. at 721.

[56] Id. at 723 (emphasis added).

[57] Id. (emphasis added).

[58] Id.

[59] Supra note 37, at 218.

[60] Id. at 215-16.

[61] Id. at 214. See also id. (“The more the appropriator is using the copied material for new, transformative purposes, the more it serves copyright’s goal of enriching public knowledge and the less likely it is that the appropriation will serve as a substitute for the original or its plausible derivatives, shrinking the protected market opportunities of the copyrighted work.”).

[62] Id. at 218.

[63] Perfect 10, 487 F.3d at 721-22 (citing Kelly, 336 F.3d at 818-19). See also Campbell, 510 U.S. at 579 (“The central purpose of this investigation is to see, in Justice Story’s words, whether the new work merely ‘supersede[s] the objects’ of the original creation, or instead adds something new, with a further purpose or different character….”) (citations omitted).

[64] See supra, notes 9-14 and accompanying text.

[65] See, e.g., the development of the compulsory “mechanical royalty,” now embodied in 17 U.S.C. § 115, that was adopted in the early 20th century as a way to make it possible for the manufacturers of player pianos to distribute sheet music playable by their instruments.

[66] See supra notes 9-14 and accompanying text.

[67] U.S. Copyright Office, Library of Congress, supra note 1, at 15.

[68] See infra, notes at 102-103 and accompanying text.

[69] U.S. Copyright Office, Library of Congress, supra note 1, at 15.

[70] See, e.g., Copyright Free Music, Premium Beat By Shutterstock, https://www.premiumbeat.com/royalty-free/licensed-music; Royalty-free stock footage at your fingertips, Adobe Stock, https://stock.adobe.com/video.

[71] U.S. Copyright Office, Library of Congress, supra note 1, at 19.

[72] Id.

[73] U.S. Const. art. I, § 8, cl. 8.

[74] See Ajay Agrawal, Joshua S. Gans, & Avi Goldfarb, Exploring the Impact of Artificial Intelligence: Prediction Versus Judgment, 47 Info. Econ. & Pol’y 1, 1 (2019) (“We term this process of understanding payoffs, ‘judgment’. At the moment, it is uniquely human as no machine can form those payoffs.”).

[75] Joshua Gans, Can AI works get copyright protection? (Redux), Joshua Gans’ Newsletter (Sept. 7, 2023), https://joshuagans.substack.com/p/can-ai-works-get-copyright-protection.

[76] See Nicholas Carlini, et al., Extracting Training Data from Diffusion Models, Cornell Univ. (Jan. 30, 2023), available at https://arxiv.org/abs/2301.13188.

[77] See Chang, Cramer, Soni, & Bamman, supra note 24; see also Matthew Sag, Copyright Safety for Generative AI, Working Paper (May 4, 2023), available at https://ssrn.com/abstract=4438593.; Andrés Guadamuz, A Scanner Darkly: Copyright Liability and Exceptions in Artificial Intelligence Inputs and Outputs, 25-27 (Mar. 1, 2023), available at https://papers.ssrn.com/sol3/papers.cfm?abstract_id=4371204.

[78] Laureyssens v. Idea Grp. Inc., 964 F.2d 131, 140 (2d Cir. 1992), as amended (June 24, 1992).

[79] Id. at 139.

[80] Harney v. Sony Pictures Television Inc., 704 F.3d 173, 178 (1st Cir. 2013). This assumes, for argument’s sake, that a given model is not “memorizing,” as noted above.

[81] Id. at 178-79.

[82] Id.

[83] U.S. Copyright Office, Library of Congress, supra note 1, at 25.

[84] Notably, the state of mind of the user would be irrelevant from the point of view of whether an infringement occurs. All that is required is that a plaintiff owns a valid copyright, and that the defendant infringed it. 17 U.S.C. 106. There are cases where the state of mind of the defendant will matter, however. For one, willful or recklessly indifferent infringement by a plaintiff will open the door for higher statutory damages. See, e.g., Island Software & Computer Serv., Inc. v. Microsoft Corp., 413 F.3d 257, 263 (2d Cir. 2005). For another, a case of criminal copyright infringement will require that a defendant have acted “willfully.” 17 U.S.C. § 506(a)(1) (2023), 18 U.S.C. § 2319 (2023).

[85] Id. at 550.

[86] Legally speaking, it would be incoherent to suggest that there can be secondary liability without primary liability. The way that AI systems work, however, could prompt Congress to modify the law in order to account for the identified situation.

[87] See, e.g., Metro-Goldwyn-Mayer Studios Inc. v. Grokster Ltd., 380 F.3d 1154, 1160 (9th Cir. 2004), vacated and remanded, 545 U.S. 913, 125 S. Ct. 2764, 162 L. Ed. 2d 781 (2005).

[88] See BMG Rts. Mgmt. (US) LLC v. Cox Commc’ns Inc., 881 F.3d 293, 306 (4th Cir. 2018); Sony Corp. of Am. v. Universal City Studios Inc., 464 U.S. 417, 442 (1984).

[89] A&M Recs. Inc. v. Napster Inc., 239 F.3d 1004, 1022 (9th Cir. 2001), as amended (Apr. 3, 2001), aff’d sub nom. A&M Recs. Inc. v. Napster Inc., 284 F.3d 1091 (9th Cir. 2002), and aff’d sub nom. A&M Recs. Inc. v. Napster Inc., 284 F.3d 1091 (9th Cir. 2002).

[90] See, e.g., Content Filtering, Microsoft Ignite, available at https://learn.microsoft.com/en-us/azure/ai-services/openai/concepts/content-filter (last visited Oct. 27, 2023).

[91] Note that, if an AI producer can demonstrate that they used no protected works in the training phase, there may in fact be no liability for infringement at all. If a protected work is never made available to the AI system, even an output very similar to that protected work might not be “substantially similar” in a legal sense.

[92] Copyright Review Board, Second Request for Reconsideration for Refusal to Register Théâtre D’opéra Spatial (SR # 1-11743923581; Correspondence ID: 1-5T5320R), U.S. Copyright Office (Sep. 5, 2023), available at https://fingfx.thomsonreuters.com/gfx/legaldocs/byprrqkqxpe/AI%20COPYRIGHT%20REGISTRATION%20decision.pdf.

[93] Am. Broad. Companies Inc. v. Aereo Inc., 573 U.S. 431, 453 (2014). (Thomas J, dissenting).

[94] Id. at 451.

[95] U.S. Copyright Office, Library of Congress, supra note 1, at 21.

[96] See 5 U.S.C. § 1051 et seq. at § 1127.

[97] See, e.g., ETW Corp. v. Jireh Pub. Inc., 332 F.3d 915, 923 (6th Cir. 2003).

[98] Dastar Corp. v. Twentieth Century Fox Film Corp., 539 U.S. 23, 34 (2003).

[99] Id. at 37.

[100] Id. at 32.

[101] See, e.g., Multi Time Mach. Inc. v. Amazon.com Inc., 804 F.3d 930, 938 (9th Cir. 2015); EarthCam Inc. v. OxBlue Corp., 49 F. Supp. 3d 1210, 1241 (N.D. Ga. 2014); Coll. Network Inc. v. Moore Educ. Publishers Inc., 378 F. App’x 403, 414 (5th Cir. 2010).

[102] Digby Adler Grp. LLC v. Image Rent a Car Inc., 79 F. Supp. 3d 1095, 1102 (N.D. Cal. 2015).

[103] J. Thomas McCarthy, The Rights of Publicity and Privacy § 1:3. Introduction—Definition and History of the Right of Publicity—Simple Definition of the Right of Publicity, 1 Rights of Publicity and Privacy § 1:3 (2d ed).

[104] See id. at § 6:3.

[105] Compare Ind. Code § 32-36-1-7 (covering name, voice, signature, photograph, image, likeness, distinctive appearance, gesture, or mannerism), with Ky. Rev. Stat. Ann. § 391.170 (limited to name and likeness for “public figures”).

[106] See Restatement (Third) of Unfair Competition § 46 (1995).

[107] See, e.g., Jeff Harleston, Artificial Intelligence and Intellectual Property – Part II: Copyright, U.S. Senate Comm. on the Judiciary Subcomm. on Intellectual Property (Jul.12, 2023), available at https://www.judiciary.senate.gov/imo/media/doc/2023-07-12_pm_-_testimony_-_harleston1.pdf; Karla Ortiz, “AI and Copyright”, U.S. Senate Comm. on the Judiciary Subcomm. on Intellectual Property (Jul. 7, 2023), available at https://www.judiciary.senate.gov/imo/media/doc/2023-07-12_pm_-_testimony_-_ortiz.pdf; Matthew Sag, “Artificial Intelligence and Intellectual Property – Part II: Copyright and Artificial Intelligence”, U.S. Senate Comm. on the Judiciary Subcomm. on Intellectual Property (Jul. 12, 2023), available at https://www.judiciary.senate.gov/imo/media/doc/2023-07-12_pm_-_testimony_-_sag.pdf.

[108] Authors, Attribution, and Integrity: Examining Moral Rights in the United States, U.S. Copyright Office (Apr. 2019) at 117-119, https://www.copyright.gov/policy/moralrights/full-report.pdf.

[109] Benj Edwards, Meta Launches Consumer AI Chatbots with Celebrity Avatars in its Social Apps, ArsTechnica (Sep. 28, 2023), https://arstechnica.com/information-technology/2023/09/meta-launches-consumer-ai-chatbots-with-celebrity-avatars-in-its-social-apps; Max Chafkin, Meta’s New AI Buddies Aren’t Great Conversationalists, Bloomberg (Oct. 17, 2023), https://www.bloomberg.com/news/newsletters/2023-10-17/meta-s-celebrity-ai-chatbots-on-facebook-instagram-are-surreal.

[110] See supra, notes 8-14 and accompanying text.

ICLE Statement on FCC’s Net Neutrality Vote

PORTLAND, Ore. (Oct. 19, 2023) – The International Center for Law & Economics (ICLE) offers the following statement from ICLE Director of Innovation Policy Kristian Stout . . .

PORTLAND, Ore. (Oct. 19, 2023) – The International Center for Law & Economics (ICLE) offers the following statement from ICLE Director of Innovation Policy Kristian Stout in response to today’s vote by the Federal Communications Commission (FCC) to initiate a rulemaking proceeding reinstating Title II regulations on broadband providers:

This attempt by the Federal Communications Commission (FCC) to reintroduce Title II regulations (often referred to as “net neutrality”) stands on even shakier ground now than it did in 2015. The wisdom behind the commission’s 2018 Restoring Internet Freedom Order is more evident with each passing year, as the broadband market continues to demonstrate no need for utility-style regulations.

The justifications the FCC has put forward for imposing Title II obligations on broadband providers not only lack merit, but emphasize the likely legal challenges this order will face. The commission’s rulemaking present a “major questions doctrine” issue that will leave the FCC in an uphill battle before the U.S. Supreme Court in the face of inevitable legal challenges.

After considering the record in the pending proceeding, one hopes the commission will realize that Title II is simply a mistake for the broadband industry, and that it will abandon this ill-fated effort.

To schedule an interview with Kristian about the FCC’s planned regulations, contact ICLE Media and Communications Manager Elizabeth Lincicome at [email protected] or (919) 744-8087.

ICLE White Paper Finds Divestitures Should Resolve Antitrust Concerns in Kroger-Albertsons Merger

PORTLAND, Ore. (Oct. 17, 2023) – The proposed $24.6 billion merger between supermarkets Kroger Co. and Albertsons Cos. Inc. has reportedly drawn regulatory scrutiny, including from . . .

PORTLAND, Ore. (Oct. 17, 2023) – The proposed $24.6 billion merger between supermarkets Kroger Co. and Albertsons Cos. Inc. has reportedly drawn regulatory scrutiny, including from the Federal Trade Commission (FTC) and California Attorney General Rob Bonta.

But according to a new International Center for Law & Economics (ICLE) white paper, attempts to block the transaction would go against the analytical framework historically used to evaluate similar mergers, as well as historical precedent of accepting divestures as a remedy to address localized problems where they arise.

The paper finds that only one supermarket merger has been challenged in court since American Store’s acquisition of Lucky Stores in 1988: the Whole Foods/Wild Oats merger in 2007. Over the last 35 years, authors Brian C. Albrecht, Dirk Auer, Eric Fruits and Geoffrey A. Manne note, the FTC has allowed every other supermarket merger and most retail-store transactions to proceed with divestitures.

Moreover, the authors argue, critics of the deal fail to consider the significant changes over the past quarter-century in how consumers shop for food and groceries, including the growth of wholesale clubs, delivery services, e-commerce, and other retail formats. Supermarkets’ share of retail sales have fallen from 81% in 1994 to 56% in 2021, while warehouse clubs and supercenters grew from 14% to 42% over that same period, they note.

“The product-market definition that the FTC has employed in its consent orders over the past more than two decades is likely to be—and should be—challenged to include warehouse clubs, in addition to accounting for online retail and delivery,” the authors write.

Kroger is currently the fourth-largest food and grocery retailer in the United States, behind Walmart, Amazon, and Costco. If the merger goes through, the combined firm will move into third place in market share, but would still account for just 9% of nationwide sales, the authors note.

In September 2023, Kroger and Albertsons announced a $1.9 billion divestiture plan that would see the firms sell 413 stores, eight distribution centers, and three store brands to C&S Wholesale Grocers, who would retain the right to purchase up to 237 additional stores if needed to resolve antitrust concerns.

“With the FTC’s knowledge of the industry and of its own past successes and failures, divestitures remain an appropriate and adequate remedy for this merger,” the authors write. “The parties appear committed to working cooperatively with regulators to craft divestitures that fully resolve competitive concerns. Rather than blocking the deal outright, the FTC can allow the merger to proceed, conditioned on acceptable divestitures that protect consumers, while permitting efficiency gains across the majority of stores.”

The full white paper can be downloaded here. To schedule an interview with one of the authors, contact Elizabeth Lincicome at [email protected] or (919) 744-8087.

ICLE Ex Parte on Digital Discrimination

We write to offer our thoughts in response to the National Telecommunications and Information Administration’s (NTIA) recently submitted comments to the Federal Communications Commission (FCC) . . .

We write to offer our thoughts in response to the National Telecommunications and Information Administration’s (NTIA) recently submitted comments to the Federal Communications Commission (FCC) in response to the notice of proposed rulemaking (NPRM) in this matter. Among its recommendations, the NTIA proposes that the FCC:[1]

  1. Adopt a disparate-impact standard to define “digital discrimination of access;” and
  2. Subject a “broad range” of service characteristics to digital-discrimination rules, including pricing, promotional conditions, terms of service, and quality of service.

We urge the FCC to reject both of these recommendations. As we note in a recent International Center for Law & Economics (ICLE) issue brief, a disparate-impact approach would likely run afoul of the U.S. Supreme Court’s tests for when such an approach is appropriate.[2] In addition, the NTIA’s recommendation to use digital-discrimination rules to regulate prices and quality of service would subject broadband-internet providers to rate regulation that the FCC has historically said it eschews.

A.      Section 60506 Does Not Support Disparate Impact Analysis

Section 60506 of the Infrastructure Investment and Jobs Act (IIJA) requires the FCC to adopt final rules facilitating equal access to broadband internet:

[T]he Commission shall adopt final rules to facilitate equal access to broadband internet access service, taking into account the issues of technical and economic feasibility presented by that objective, including … preventing digital discrimination of access based on income level, race, ethnicity, color, religion, or national origin; and… identifying necessary steps for the Commissions to take to eliminate discrimination …[3]

The U.S. Supreme Court has established tests governing when it is appropriate to conduct an effects-based “disparate impact” analysis in the context of discrimination law. First, the presence of language like “otherwise make unavailable” is critical to construing a statute as demanding an effects-based analysis.[4] Such phrases, the Court found, “refer[] to the consequences of an action rather than the actor’s intent.” Second, the structure of a statute’s language matters:

The relevant statutory phrases … play an identical role in the structure common to all three statutes: Located at the end of lengthy sentences that begin with prohibitions on disparate treatment, they serve as catchall phrases looking to consequences, not intent. And all [of these] statutes use the word “otherwise” to introduce the results-oriented phrase. “Otherwise” means “in a different way or manner,” thus signaling a shift in emphasis from an actor’s intent to the consequences of his actions.[5]

Thus, as Section 60506 was drafted without “results-oriented language” and instead frames the prohibition against digital discrimination as “based on income level, race, ethnicity, color, religion, or national origin,” this would put the rule squarely within the realm of prohibitions on intentional discrimination. That is, to be discriminatory, the conduct in question must have been made intentionally because of the protected characteristic. Mere statistical correlation between outcomes and protected characteristics is insufficient to demonstrate discrimination under Section 60506.

B.      Broadband Rate Regulation Is at Odds With Longstanding FCC Policy

The FCC has for years been explicit about its apprehension to impose direct rate regulation on broadband-internet providers. Obama-era FCC Chair Tom Wheeler promised to forebear from rate regulation under the 2015 Open Internet Order (OIO), declaring “we are not trying to regulate rates.”[6] Last month, in a speech announcing the FCC’s proposal to regulate broadband internet under Title II of the Communications Act, Chair Jessica Rosenworcel was emphatic: “They say this is a stalking horse for rate regulation. Nope. No how, no way.”[7]

And yet, the NTIA recommends precisely that: a stalking horse for rate regulation under the guise of preventing digital discrimination: “Without addressing pricing as a possible source of discrimination, the Commission will be hard pressed to meet its statutory mandate to prevent digital discrimination of access.”[8] Indeed, in addition to prices, the NTIA recommends addressing promotions, as well as terms of service and quality of service—both of which are inextricably intertwined with pricing. Section 60506 is explicit that the FCC must “tak[e] into account the issues of technical and economic feasibility” in addressing digital discrimination. The NTIA, however, appears to suggest ignoring that mandate.[9]

C.      The Problem With De Facto Rate Regulation

In a competitive market, prices allow for the successful coordination of supply and demand, and the market price reflects both consumer demand and the costs of production. Of course, for those on the demand side of the equation, the price of a good or service is a cost to them, and they would prefer falling prices to rising prices. For suppliers, the price represents the revenue from selling the good or service, and they would prefer rising prices to falling prices.

Due to this inherent tension, there is a natural inclination on the part of both consumers and producers to seek government intervention in the competitive process to either halt or slow price changes. The most obvious way the government can intervene is through rate regulation, such as price controls. It is well known and widely accepted that price controls can make both consumers and sellers worse off.[10] Consequently, policymakers often pitch policies to control prices under another name (e.g., “second-generation rent relief” instead of “rent control”) or introduce policies that are not explicit price controls. These de facto rate regulations  (e.g., quality-of-service mandates)  have substantially the same effects as direct price controls.

For example, some agricultural products are subject to “marketing orders,” which are legal cartels than can dictate the price and quality of produce.[11] Consider an apple market subject to a marketing order that specifies fresh apples must be of a certain shape and size, such that only large, round apples can be sold as fresh produce.

Consumers presumably prefer large apples to small apples and prefer round apples to misshapen apples. Thus, the order that only large, round apples can be sold as fresh has the effect of increasing/shifting the demand curve. Consumers would be willing to pay more for the seemingly better fruit, and they’d be willing to buy more. But the order also increases the cost to apple growers. They have to find a way to dispose of their smaller or misshapen apples, perhaps by making apple sauce or juicing the fruit. They also incur higher costs of managing their crop to produce more of the higher-quality fruit. This has the effect of decreasing/shifting the supply curve for fresh fruit. Growers will supply less fruit at a higher cost.

Combining the effects from both the shift in supply and the shift in demand shows that the marketing order unambiguously results in a higher price for apples. What is not known, however, is whether more or fewer apples are sold. That will depend on the elasticities of demand and supply. Because the order results in a higher price, however, it has created a de facto price floor without explicitly setting one. Consumers are not aware that they are paying a higher price, because they do not know what type of fruit would otherwise be available, and at what price, absent the quality restrictions.

Similarly, broadband quality-of-service mandates simultaneously increase demand while increasing costs. Were all other things held constant, the result would be a higher price for broadband. What is not known, however, is whether more or fewer households will subscribe to broadband. That will depend on the elasticities of demand and supply. Because the mandate results in a higher price, however, it has created a de facto price floor without explicitly setting one. Consumers are not aware that they are paying a higher price because they do not know what quality of service would otherwise be available—and at what price—absent the quality-of-service mandate.

A recent ICLE issue brief explores in detail how these sorts of  terms-of-service and quality-of-service mandates often amount to de facto rate regulation.[12] Sadly, we have seen many recent attempts—including by the NTIA itself—to introduce these sorts of de facto rate regulations:

  • The NTIA’s notice of funding opportunity (NOFO) under the Broadband Equity, Access, and Deployment (BEAD) program requires each participating U.S. state or territory to include a “middle-class affordability plan to ensure that all consumers have access to affordable high-speed internet” (emphasis in original).[13] The NOFO specifies a price (“affordable”); a quantity (“all middle-class households”); and imposes a quality mandate (“high-speed”).
  • In its third and fourth funding rounds, the U.S. Department of Agriculture’s ReConnect Loan and Grant Program included provision of a “low-cost option” as a point criteria in award decisions. It also included a requirement that projects must provide broadband access at speeds of at least 100/100 Mbps (e., 100 Mbps symmetrical speed).
  • The FCC’s 2015 Open Internet Order outright prohibited “paid prioritization”—that is, seeking payments for network utilization from edge providers like Google, Facebook, and Netflix—while casting suspicion on other pricing schemes under its Internet Conduct Standard.

Moreover, as we note in our Income Conundrum issue brief, the evaluation of digital-discrimination claims based on income level can yield highly complicated analyses due to income’s correlation with a host of factors, both protected (e.g., race and national origin) and unprotected (e.g., home-computer ownership).  Adoption of Section 60506 rules that do not recognize this “income conundrum” will invite costly and time-consuming disparate-impact litigation that alleges digital discrimination, both where no such discrimination exists and where it is excused by economic-feasibility considerations. Even worse, regulatory interventions on price, quality of service, and terms of service would extend this damage further by creating de facto utility regulation on providers that completely distorts investment incentives.

D.     Conclusion

The FCC must be cautious when promulgating rules under Section 60506. In particular, the commission should adopt an intent-based discriminatory-treatment standard, rather than one that opens the doors to disparate-impact claims. And FCC rules should articulate a presumption of nondiscrimination, in which allegations of digital discrimination must be demonstrated, rather than a presumption of discrimination that must be rebutted for each deployment, service, and pricing decision.

If the commission has good evidence of intentional discrimination in the deployment of broadband, it has a role to play in preventing it. But without strong, compelling evidence of intentional discrimination, the FCC will run the risk of a constitutional challenge to its rules and waste scarce resources chasing bogeymen.

Above all, the FCC should resist calls to engage in rate regulation, either through direct intervention on broadband prices or through interventions on quality of service or terms of service. The longstanding policy to avoid rate regulation has been an important factor leading to increased broadband deployment in the United States.

[1] Ex Parte Comments of the National Telecommunications and Information Administration, In the Matter of Implementing the Infrastructure Investment and Jobs Act: Prevention and Elimination of Digital Discrimination, GN Docket No. 22-69 (Oct. 6, 2023), available at  https://www.ntia.gov/sites/default/files/publications/ntia_digital_discrimination_ex_parte_comment_10.6.23.pdf at 3 and 8.

[2] Eric Fruits & Kristian Stout, The Income Conundrum: Intent and Effects Analysis of Digital Discrimination, Int’l. Ctr. for L. & Econ. (Nov. 14, 2022), available at https://laweconcenter.org/wp-content/uploads/2022/11/The-Income-Conundrum-Intent-and-Effects-Analysis-of-Digital-Discrimination.pdf; see also Eric Fruits, Kristian Stout, & Ben Sperry, ICLE Reply Comments on Prevention and Elimination of Digital Discrimination, Notice of Proposed Rulemaking, In the Matter of Implementing the Infrastructure, Investment, and Jobs Act: Prevention and Elimination of Digital Discrimination, No. 22-69, at Part III (Apr. 20, 2023), https://laweconcenter.org/resources/icle-reply-comments-on-prevention-and-elimination-of-digital-discrimination.

[3] 47 U.S.C. § 1754 (emphasis added).

[4] Texas Dep’t of Hous. & Cmty. Affs., 576 U.S. at 534.

[5] Id. at 534-535 [emphasis added].

[6] Tom Wheeler, Hearing on FCC Reauthorization: Oversight of the Commission, U.S. House Energy and Commerce Subcommittee on Communications and Technology, (Mar. 19, 2015), https://www.govinfo.gov/content/pkg/CHRG-114hhrg95817/html/CHRG-114hhrg95817.htm.

[7] FCC Chair Rosenworcel on Reinstating Net Neutrality Rules, C-Span (Sep. 26, 2023), https://www.c-span.org/video/?530731-1/fcc-chair-rosenworcel-reinstating-net-neutrality-rule.

[8] Supra note 1, at 10.

[9] Id. (“Congress set out a specific list of demographic groups protected by this statute, including … racial and ethnic minorities (who as previously noted are disproportionately likely to live in environments where networks are costlier to maintain, among other challenges).”) [emphasis added].

[10] See, e.g., N. Gregory Mankiw, Principles of Microeconomics, 4th ed., Thomson South-Western (2007); Paul Krugman & Robin Wells, Economics, 6th ed., MacMillan (2021); Steven A. Greenlaw & David Shapiro, Principles of Microeconomics 2nd ed., OpenStax (2017).

[11] See Darren Filson, Edward Keen, Eric Fruits, & Thomas Borcherding, Market Power and Cartel Formation: Theory and an Empirical Test, 44 J. L. & ECON. 465 (2001).

[12] Eric Fruits & Geoffrey A. Manne, Quack Attack: De Facto Rate Regulation in Telecommunications, Int’l. Ctr. for L. & Econ. (Mar. 30, 2023), available at https://laweconcenter.org/wp-content/uploads/2023/03/De-Facto-Rate-Reg-Final-1.pdf.

[13] Notice of Funding Opportunity, Broadband Equity, Access, and Deployment Program, NTIA-BEAD-2022, NTIA (May 2022), available at https://broadbandusa.ntia.doc.gov/sites/default/files/2022-05/BEAD%20NOFO.pdf (note that the IIJA itself did not include this requirement, but it was added by NTIA as part of the NOFO process; thus, it is unclear the extent to which this represents a valid requirement by NTIA under the BEAD program).

LONG FORM WRITING

Food-Retail Competition, Antitrust Law, and the Kroger/Albertsons Merger

Executive Summary In October 2022, the Kroger Co. and Albertsons Cos. Inc. announced their intent to merge in a deal valued at $24.6 billion.[1] Given . . .

Executive Summary

In October 2022, the Kroger Co. and Albertsons Cos. Inc. announced their intent to merge in a deal valued at $24.6 billion.[1] Given the Federal Trade Commission’s (FTC) increasingly aggressive enforcement stance against mergers and acquisitions, as well as Chair Lina Khan’s previous writings on food retail specifically,[2] the agency appears poised to try to block the transaction—even with divestitures.[3] The FTC and U.S. Justice Department’s (DOJ) recently unveiled draft revisions to the agencies’ merger guidelines further suggest that they plan to challenge more mergers—and to do so more aggressively than under past administrations.[4]

But attempting to block this transaction would go against the analytical framework the FTC has historically used to evaluate similar transactions, as well as the agency’s historical precedent of accepting divestures as a remedy to address localized problems where they arise. Such breaks with the past sometimes happen; our understanding of the law and economics evolves. But in the case at hand, these breaks from tradition would reflect a failure to consider relevant and significant changes in how consumers shop for food and groceries in today’s world.

The FTC has a long history of assessing retail mergers in a manner significantly at odds with the aggressive approach it is currently signaling. Only one supermarket merger has been challenged in court since American Store’s acquisition of Lucky Stores in 1988: the Whole Foods/Wild Oats merger in 2007.[5] Over the last 35 years, the FTC has allowed every other supermarket merger and most retail-store transactions to proceed with divestitures. Within the last two years alone, these have included Tractor Supply/Orschlein and 7-Eleven/Speedway.[6] The FTC’s historic approach recognizes the reality that competitive concerns regarding supermarket mergers can be readily and adequately remedied by divestitures in geographic markets of concern; indeed, even Whole Foods/Wild Oats was ultimately resolved with a divestiture agreement following a fractured circuit court decision.[7]

The retail food and grocery landscape has changed dramatically since American/Lucky and Whole Foods/Wild Oats. But the ways in which the market has changed point toward its becoming more competitive, further undermining a possible FTC case. With the growth of wholesale clubs, delivery services, e-commerce, and other retail formats, the industry is no longer dominated by traditional supermarkets. In addition, these changing dynamics have made geographic distance, traffic patterns, and population density decreasingly relevant in a consumer’s choice of where they purchase food and groceries. Today, Kroger is only the fourth-largest food and grocery retailer in the United States, behind Walmart, Amazon, and Costco. If the merger goes through, the combined firm will move into third place in market share but would still account for just 9% of nationwide sales.[8]

The upshot is that the food and grocery industry is arguably as competitive as it has ever been. Unfortunately, recent developments suggest the FTC may well ignore or dismiss the economic realities of this rapid transformation of the food and grocery industry, substituting instead the outdated approach to market definition and industry concentration signaled by the draft guidelines.[9]

In light of these developments in the food-retail market and the FTC’s likely break from precedent, this paper highlights important areas where both the commission and commentators’ stances appear to run headlong into legal precedent that mandates an evidence-based approach to merger review, even as the best available evidence points to a dynamic and competitive grocery industry. The correct understanding of the law and the industry appears entirely at odds with a challenge to the proposed merger.

A.     The FTC’s Merger-Enforcement Policy Is on a Collision Course with the Law

The Kroger/Albertsons merger proceeds against a backdrop of tough merger-enforcement rhetoric and actions from the FTC. Recent developments include the publication of aggressive revised merger guidelines, a string of cases brought to block seemingly benign mergers, process revisions that burden even unproblematic mergers, and FTC leadership’s contentious and expansive interpretation of the merger laws. The FTC’s ambition to remake U.S. merger law is likely to falter before the courts, but not before imposing a substantial tax on all corporate transactions—and, ultimately, on consumers.

The retail food and grocery market has changed substantially since the last time a supermarket merger was challenged in court. If the merger goes to trial, the court will need to address issues that have not been litigated in decades, if ever. Depending how the court rules, the market definition for future supermarket mergers may be substantially revised. Moreover, if the FTC attempts to litigate allegations of labor-market or input-market monopsony, the agency runs the risk of a humiliating loss that could stymie future attempts to expand the role of monopsony in competition enforcement and policy. The FTC thus would do well to even-handedly assess the Kroger Albertsons merger, remaining open to new evidence and sensible remedies. This is especially true given the agency’s losing streak in court—culminating with its unsuccessful challenges of the Meta/Within and Microsoft/Activision Blizzard deals.[10]

B.     The Product Market Is Broader Than Local Supermarkets

Because of recent changes in market dynamics, it no longer makes sense to limit the relevant market to supermarkets alone. Rather, consumer behavior in the face of omnipresent wholesale clubs, e-commerce, and local delivery platforms significantly constrains supermarkets’ pricing decisions.

Recent FTC consent orders involving supermarket mergers have limited the relevant product market to local brick-and-mortar supermarkets and food and grocery sales at nearby hypermarkets (e.g., Walmart supercenters), while excluding wholesale-club stores (e.g., Costco), e-commerce (e.g., Amazon), and further-flung stores accessible through online-delivery platforms (e.g., Instacart). This is based on an assertion that the relevant market includes only those retail formats in which a consumer can purchase nearly all of a household’s weekly food and grocery needs from a single stop, at a single retailer, in the shopper’s neighborhood. This is, however, no longer how most of today’s consumers shop. Instead, shoppers purchase different bundles of groceries from multiple sources, often simultaneously.[11] This pattern has substantial implications for supermarkets’ competitive environment, and underscores why the FTC should not rely on outdated market definitions.

Past FTC consent orders have defined the relevant geographic markets to be areas that range from a two- to ten-mile radius around each of the merging parties’ supermarkets.[12] The radius depends on such factors as population density, traffic patterns, and the unique characteristics of each market. It would, however, be reasonable to expand the relevant geographic market when club stores are present, as these have a larger catchment area than supermarkets. Finally, the rapid growth of e-commerce and delivery services make distance, traffic patterns, and population density decreasingly relevant in a consumer’s choice of where they purchase food and groceries. As with product-market definitions, this is a crucial empirical issue that should be evaluated in the FTC’s merger review and any litigation.

C.     Labor Monopsony Concerns Are Unlikely to Hold Up in Court

More than in any previous retail merger, opponents of the Kroger/Albertsons deal have raised the specter of potential monopsony power in labor markets. But these concerns reflect a manifestly unrealistic conception of labor-market competition. Fundamentally, the market for labor in the retail sector is extremely competitive, and workers have a wide range of alternative employment options—both in and out of the retail sector. At the same time, both Kroger and Albertsons are highly unionized, providing a counterbalance to any potential exercise of monopsony power by the merged firm.

D.    The Alleged ‘Waterbed Effect’ Is Not Borne Out by Evidence

Some critics of the merger have speculated that the merged company would be able to exercise monopsony power against its food and grocery suppliers (i.e., wholesalers and small manufacturers), often invoking an economic concept called the “waterbed effect.” The intuition is that the largest buyers may use their monopsony power to negotiate lower input prices from suppliers, leading the suppliers to make up the lost revenue by raising prices for their smaller, weaker buyers.

But these arguments are far from compelling. It is very difficult, for example, to hypothesize any relevant market for purchasing a good where the merged firm would have market power. Critics also often fail to consider the ability of many producers—both small and large—to sell directly to consumers, as demonstrated by the rise of online shopping, with its low entry barriers entry and low-cost structure.

Much of the discussion of the waterbed effect focuses on harm to competing retailers, rather than consumers. But this is not the harm that U.S. antitrust law seeks to prevent. It is thus not surprising that at least one U.S. court has rejected waterbed-effect claims on grounds that there was no harm to consumers.

E.     Divestitures Historically Have Proven an Appropriate and Adequate Remedy

Historically, the FTC has allowed most grocery-store transactions to proceed with divestitures, such as Ahold/Delhaize (81 stores divested), Albertsons/Safeway (168 stores), and Price Chopper/Tops (12 stores). The extent of the remedies sought depends on the extent of post-merger competition in the relevant local markets, as well as the likelihood of significant entry by additional competitors into the relevant markets. The benefit of selling off stores is that you can allow the vast majority of stores—where there is no worry about anticompetitive effect—to merge, while targeting the areas that have the highest concern.

Despite a long history of divestitures serving as an appropriate and adequate remedy in supermarket mergers, some point to the Albertsons/Safeway merger divestitures to Haggen as evidence that divestitures are no longer an appropriate remedy. But several factors idiosyncratic to Haggen and its acquisition strategy led to the failure of that divestiture, and it does not properly stand for the claim that all supermarket divestitures are doomed.

In September 2023, Kroger and Albertsons announced a $1.9 billion divestiture proposal to sell 413 stores, eight distribution centers, two offices, and five private-label brands to C&S Wholesale Grocers LLC.[13] If consummated, the deal would cover operations spanning 17 states and the District of Columbia, and C&S has committed to maintain collective-bargaining agreements with labor.[14] As antitrust enforcers review whether these proposed divestitures are adequate, they should learn from the Haggen experience, rather than view it as justification to reject reasonable divestiture options that have worked for other mergers.

I.  Introduction

In October 2022, the Kroger Co. and Albertsons Cos. announced their intent to merge the two companies in a deal valued at $24.6 billion.[15] Kroger is the fourth-largest food and grocery retailer in the United States—behind Walmart, Amazon, and Costco—while Albertsons is fifth.[16] Both chains trail market-leading Walmart  by a considerable margin. Kroger operates 2,726 stores under the Kroger, Harris Teeter, and Smith’s banners, while Albertsons operates 2,278 stores under the Safeway, Albertsons, and Von’s grocery banners.[17] By contrast, Walmart and Sam’s Club combined store count is greater than 5,300, and the company’s grocery revenue is more than twice that of Kroger and Albertsons combined.[18]

While the proposed Kroger/Albertsons merger is a large transaction in terms of dollar valuation and the combined firm would move into third place in market share, it would still account for just 9% of nationwide sales.[19] In some localities, the market share would be much larger, however, raising questions regarding whether the merger would increase Kroger/Albertson’s monopoly power in those retail markets and convey monopsony power in wholesale markets and local labor markets. To address such questions, Kroger and Albertsons have announced a $1.9 billion divestiture proposal that would include the sale of 413 stores and eight distribution centers across 17 states to C&S Wholesale Grocers LLC.[20]

But given the Federal Trade Commission’s (FTC) recent disposition toward proposed mergers and Chair Lina Khan’s previous writings about food retail, it is widely expected that the FTC will not be satisfied with any remedy offers from the companies—including offers to divest stores—and will instead attempt to block the merger.[21] California Attorney General Rob Bonta has also signaled that he is likely to challenge the deal.[22] In anticipation, Kroger’s chief executive officer announced that both companies are “committed to litigate” if the enforcers act to block the merger.[23]

What can we expect of such a court battle? The precedent is slightly complicated. While several retail mergers have been challenged in court, including Staples/Office Depot in 2015[24] and Whole Foods/Wild Oats in 2007,[25] no supermarket mergers have been litigated since the State of California’s 1988 challenge to American Store’s acquisition of Lucky Stores.[26] Both the supermarket business and antitrust analysis have changed dramatically over the intervening 35 years. In this paper, we describe some of the changes that have occurred within food retail over the past 35 years, and how they should be addressed by the merging parties, the FTC, and—if litigated—by the courts.

As with most merger analysis, many of the most important questions hinge on a proper definition of the relevant market. The most obvious changes we have observed in food retail in recent decades, including the rise of wholesale club stores and e-commerce, are directly relevant to the question of market definition. The most recent FTC consent orders involving supermarket mergers defined the relevant product market to include “traditional” brick-and-mortar supermarkets, as well as food and grocery sales at hypermarkets (e.g., Walmart supercenters), while excluding wholesale club stores (e.g., Costco) and e-commerce (e.g., Amazon and home-delivery services).[27] This is based on the longstanding assertion that the relevant market includes only those retail formats in which a consumer can purchase all or nearly all of their household’s weekly food and grocery needs during a single stop at a single retailer.

Any attempt by the FTC to maintain this outdated market definition will likely be challenged in court. Research has shown that consumer behavior has changed over time, in that the typical consumer no longer makes once-a-week shopping trips to a single food and grocery retailer.[28] Instead, the typical consumer makes multiple weekly trips and multi-homes across several different retailers and retail formats. This change alone blurs the line between traditional supermarkets and other retail formats. The extent to which wholesale clubs, e-commerce, and delivery services should be included in the relevant market is a key empirical issue that surely will—and should—be evaluated in the FTC’s merger review and any ensuing litigation.

Geographic market is also an issue. The most recent FTC consent orders have defined the relevant geographic markets to be areas that range from a two- to ten-mile radius surrounding each of the merging parties’ supermarkets.[29] The radius depends on such factors as population density, traffic patterns, and the unique characteristics of each market. That, too, needs revision. Based on academic research, industry surveys, and reports from companies that find club stores compete with supermarkets and have larger catchment areas, we conclude that it would be reasonable to expand the relevant geographic market when club stores are present. In addition, the rapid growth of e-commerce and delivery services make distance, traffic patterns, and population density decreasingly relevant in a consumer’s choice of where they purchase food and groceries. As with product-market definition, this is a crucial empirical issue that should be evaluated in the FTC’s merger review and any ensuing litigation.

More than in any previous retail merger, opponents of the Kroger/Albertsons merger have raised the specter of potential monopsony power in labor markets. We argue that these concerns are likely overblown and will be nearly impossible to demonstrate if the merger were to be litigated. Fundamentally, the market for labor in the retail sector is highly competitive, with workers having a wide range of alternative employment if a particular employer attempted to exploit any claimed monopsony power. In addition, both Kroger and Albertsons are highly unionized. Through their collective-bargaining agreements, unions exercise monopoly power in labor negotiations that likely counterbalances any attempted exercise of monopsony power by the merged firm.

Lastly, some critics of the merger have speculated that the merged company may have and exercise monopsony power among its food and grocery suppliers (e.g., wholesalers and small manufacturers). In particular, critics invoke a concept colloquially called the “waterbed effect,” where pushing input prices down for some retailers ends up raising the price for other retailers. Why prices are “pushed” down is not always clear in popular discussions, nor is it clear that it qualifies as an antitrust harm in any way. Being the easiest trading partner would also result in lower prices.

We conclude this may be the weakest argument raised in opposition to the merger. The United Kingdom has evaluated “waterbed effect” allegations in at least two supermarket mergers and found no evidence indicating any anticipated effects from the mergers on input prices that would harm consumers.[30] More importantly, much of the discussion of waterbed effects focuses on harm to competing retailers, rather than to consumers. At least one U.S. court has rejected waterbed-effect claims on the grounds that (1) the plaintiffs did not demonstrate any harm to consumers, and (2) firms can substitute to other suppliers, thereby mitigating any anticipated waterbed effect.[31]

Given the size of a merged Kroger and Albertsons, it would be easy, but naïve, to conclude that the merger should be blocked.[32] The retail food and grocery market has changed substantially since the last time a supermarket merger was challenged in court. If the merger goes to trial, the court will address issues that have not been litigated in decades, if ever. Depending how the court rules, the market definition for future supermarket mergers may be substantially revised. Moreover, if the FTC attempts to litigate allegations of labor-market or input-market monopsony, the agency runs the risk of a humiliating loss that could stymie future attempts to expand the role of monopsony in competition enforcement and policy.

II.     The Agencies Are Trying to Rewrite Merger-Review Standards

The recently published FTC-DOJ draft merger guidelines are a particularly notable backdrop for the Kroger/Albertsons merger, leading many commentators to expect the FTC to take a hardline stance on the deal.[33] Merger case law, however, has not changed much in recent years. Given the merging parties’ apparent willingness to litigate the case, if necessary, the likelihood of a protracted legal battle appears to be high. As we explain below, at least at first sight, any case against the merger would be largely built on sand, and the commission’s chances of succeeding in court appear slim.

The Clayton Act of 1914 grants the U.S. government authority to review and challenge mergers that may substantially lessen competition. The FTC and DOJ are the two antitrust agencies that share responsibility to enforce this law. Traditionally, the FTC investigates retail mergers, while the DOJ oversees other sectors, such as telecommunications, banking, and transportation.

Before the FTC and DOJ officials appointed by the current administration came into office, the settled practice was for the antitrust agencies to follow the 2010 Horizontal Merger Guidelines, which outline the analytical framework and evidence they use to evaluate mergers. The 2010 guidelines describe four major steps of merger analysis:

  1. The first step is to define the relevant product and geographic markets affected by the merger. The goal is to identify the set of products and regions that are close substitutes to the merging parties’ products and regions.
  2. The second step is to assess the merger’s competitive effects, or how the merger may harm competition in the defined markets.
  3. The third step is to examine the role of market entry as a potential counterbalance to the merger’s competitive effects. For entry to be sufficient to deter or undo the anticompetitive effects of a merger, it must be timely, likely, and sufficient in scale and scope.
  4. The fourth and final step is to evaluate the efficiencies the merger would generate, or how the merger may benefit consumers by reducing costs and improving quality.

The antitrust agencies weigh all these factors to determine whether a merger is likely to harm competition and consumers. If they find that a merger raises significant competitive concerns, they may seek to block it or require remedies such as divestitures or behavioral commitments from the merging parties.

Several factors, however, suggest that authorities are unlikely to follow this measured approach when reviewing the Kroger/Albertsons merger. Primarily, the FTC and DOJ have recently issued draft revised merger guidelines. The 2023 guidelines have not yet been adopted, although the public comment period is closed. Compared to the previous iteration, which guided recent consent decrees, the new guidelines contain more stringent structural presumptions—that is, a presumption that a merger that merely increases concentration (as all horizontal mergers do) by a certain amount violates the law, rather than deferring to more nuanced economic analysis connecting specific market attributes to a likelihood of actual consumer harm.[34] These new, more stringent structural presumptions are not justified by new economic learnings about the economic effects of concentration. As now-FTC Bureau of Economics Director Aviv Nevo and colleagues wrote in 2022 (just before he joined the commission):

If the agencies were to substantially change the presumption thresholds, they would also need to persuade courts that the new thresholds were at the right level. Is the evidence there to do so? The existing body of research on this question is, today, thin and mostly based on individual case studies in a handful of industries. Our reading of the literature is that it is not clear and persuasive enough, at this point in time, to support a substantially different threshold that will be applied across the board to all industries and market conditions.[35]

Although elements of Nevo and coauthors’ proposed framework are present in the new proposed guidelines, the guidelines also incorporate new language that reflects a persistent thumb on the scale, systematically undermining merging parties’ ability to justify their merger. For example, while a presumption of harm is triggered at a certain level of concentration (an HHI of 1800), in markets where there has previously been consolidation (over an unspecified timeframe), an impermissible “trend [toward concentration] can be established by… a steadily increasing HHI [that] exceeds 1,000 and rises toward 1,800.”[36] Traditionally, an HHI under 1500 would be considered “unconcentrated” and presumed to raise no competitive concerns.[37]

While the FTC will likely point to the renewed focus on concentration measures as capturing the Clayton Act’s focus on lessening competition and the tendency to create a monopoly, the draft guidelines make clear that commission now views concentration as problematic in itself, regardless of whether it lessens competition. For example, the draft guidelines state “efficiencies are not cognizable if they will accelerate a trend toward concentration.”[38] Such a statement effectively negates any efficiency defense available to all but the very smallest firms. Efficiencies will almost always increase concentration—especially if those efficiencies come from economies of scale. If a merger creates efficiencies, the merged firm can lower costs, cut prices, and attract more customers. Attracting more customers with better products and prices will likely increase competition.

The economic evidence is quite strong that efficiency increases concentration.[39] If no efficiency defense is possible, any horizontal merger could accelerate a trend toward concentration (if it had been previously becoming more concentrated). Spinning in these circles is why the notion of a “trend” toward concentration raising particular concern hasn’t been reflected in guidelines since 1968,[40] and reached its apotheosis in Von’s Grocery in 1966[41]—one of the most thoroughly reviled merger cases in U.S. history.[42]

Before updating the merger guidelines, the FTC had already started to tighten its merger-enforcement policy. Among other actions, the agency brought high-profile cases against the Illumina/GRAIL, Meta/Within, and Microsoft/Activision Blizzard deals.[43] So far, all three challenges have resulted in defeat for the FTC in adjudication. Taken together, these cases suggest the agency is willing to push the law beyond its limits in an attempt to limit corporate consolidation, whatever the actual competitive effect. The courts have thus far shown themselves unwilling to buy the agencies more speculative claims of harm. In contrast, the DOJ recently was able to block Penguin-Random House from merging with Simon & Schuster.[44] On the surface, this may seem like a novel case similar to those the FTC has been pursuing; it involved monopsony power against authors. But in this case, the parties agreed that, if there was harm to the authors, there would be fewer books, thereby harming consumers.[45] Fighting over harms to consumers (not concentration) is textbook antitrust litigation.[46]

Finally, the FTC’s leadership has been particularly bearish about the potential consumer benefits of corporate mergers and acquisitions. This inclination is reflected in Chair Khan’s assertion that the Clayton Act embodies a “broad mandate aimed at prohibiting mergers even when they do not constitute monopolization and even when their tendency to lessen competition is not certain.”[47] One way to prohibit mergers is to make them more costly without even going to court. This is what will happen under the agencies’ proposed changes to the premerger notification rules (“Hart-Scott-Rodino Act”).[48] Despite no evidence presented of anticompetitive mergers slipping through the cracks due to the current reporting being too lax, the revised guidelines would greatly increase the cost of merging, thereby reducing the number of mergers.

Even the FTC estimates a massive increase in compliance costs of approximately $350 million, to more than $470 million per year. But that estimate is likely a serious underestimate, as it is based on, among other things, an unscientific “estimate” of the time involved and a dated assumption about the average hourly costs imposed on filers’ senior executives and firms’ counsel.

A survey that the U.S. Chamber of Commerce conducted of 70 antitrust practitioners about the proposed HSR revisions found that the new rules would increase compliance costs by $1.66 billion, almost five times the FTC’s $350 million estimate.”[49] While this general approach to blocking more mergers will not be directly applicable in any particular case, it highlights the FTC’s willingness to not follow “the old rules.”

All of these factors—in concert with the merging parties’ claim that they are prepared to go to court if the FTC decides to block the transaction outright[50]—suggest that there is a particularly high likelihood that the Kroger/Albertsons merger will be challenged and litigated, rather than approved, or challenged and settled.

For the reasons outlined in the following sections, however, the FTC is unlikely to prevail in court. The market overlaps between the merging parties are few and can be resolved by relatively straightforward divestiture remedies, as already proposed by the parties—which, even if disfavored by the agency, are routinely accepted by courts. Likewise, the FTC’s likely market definition and potential theories of harm pertaining to labor monopsony and purchasing power more generally appear speculative at best. The upshot is that the FTC’s desire to bring tougher merger enforcement appears to be on a collision course with the law as it is currently enforced by U.S. courts.

III.   The Relevant Market Is Broader Than Hyper-Local Supermarkets

The retail food and grocery landscape has changed dramatically since the last litigated supermarket merger. Consumer-shopping behavior has shifted toward more frequent shopping trips across a wide variety of formats, which include warehouse clubs (e.g., Costco); e-commerce (e.g., Amazon); online-delivery platforms (e.g., Instacart); limited-assortment stores (e.g., Trader Joe’s and Aldi); natural and organic markets (e.g., Whole Foods); and ethnic-specialty stores (e.g., H Mart); in addition to traditional supermarkets. Because of these enormous changes, the market definition assumed in previous FTC consent orders likely will be—and should be—challenged, given the empirical evidence.

A.     Recent Trends in Retailing Have Upended the ‘Traditional’ Grocery Market Definition

The FTC is likely to find the relevant product market to be supermarkets, which the agency has previously defined as retail stores that enable consumers to purchase all of their weekly food and grocery requirements during a single shopping visit. This product-market definition has remained mostly unchanged for at least a quarter of a century. In both the Albertsons/Safeway merger and the Ahold/Delhaize merger, consent orders between the FTC and the merging parties defined the relevant market to be “the retail sale of food and other grocery products in supermarkets.”[51] The orders defined supermarkets as “full-line grocery stores” that provide “one-stop shopping” that enables consumers “to shop in a single store for all of their food and grocery needs.”[52]

On the one hand, the consent orders’ product-market definitions included supermarkets located within hypermarkets, such as Walmart supercenters. Hypermarkets sell both products that are not typically sold in traditional supermarkets, as well as a sufficient range of food and grocery products such that consumers can “purchase all of their weekly grocery requirements in a single shopping visit.”[53]

On the other hand, the consent orders excluded club stores—such as Costco, Sam’s Club, and BJ’s Wholesale Club—as well as “hard discounters, limited assortment stores, natural and organic markets, [and] ethnic specialty stores.”[54] The orders claim that these stores are excluded from the relevant product market because “they offer a more limited range of products and services than supermarkets and because they appeal to a distinct customer type.”[55] In addition, the orders indicate that “supermarkets do not view them as providing as significant or close competition as traditional supermarkets.”[56] Prior consent orders omitted any discussion of whether online retailers or delivery services should be included or excluded from the relevant market. This product-market definition has remained mostly unchanged—and mostly unchallenged—since the Ahold/Giant merger a quarter-century ago.[57]

Figure 1 shows that retail sales by supermarkets, warehouse clubs, supercenters, and other grocery stores have been relatively stable at 5-6% of U.S. gross domestic product (GDP).[58] Figure 2 shows that supermarkets’ share of retail sales dropped sharply from the early 1990s through the mid-2000s, with that share shifting to warehouse clubs and supercenters. These figures are consistent with the conclusion that warehouse clubs and supercenters successfully compete against traditional grocery stores. Indeed, it would be reasonable to conclude that the rise of warehouse clubs and supercenters at the expense of traditional supermarkets is one of the most significant long-run trends in retail.

The retail food and grocery industry has changed dramatically. In particular, a great deal of attention has been paid to consolidation in the industry. Lina Khan & Sandeep Vaheesan, for example, note that “[t]he share of groceries sold by the four biggest food retailers more than doubled between 1997 and 2009, from seventeen percent in 1994 to twenty-eight percent in 1999 and thirty-four percent in 2004.”[59]

Below, we note that the average consumer shops for food and groceries more than once a week and shops at more than one retail format in a given week. Competition in groceries is not just between supermarkets. While Khan & Vaheesan recognize that supermarkets started to compete with warehouse clubs and supercenters, they fail to update their market definition to reflect that competition.[60] Over the period of increasing concentration within groceries, warehouse clubs and supercenters were steadily eroding supermarkets’ share of retail sales.

Figure 2 shows that, in 1994, supermarkets accounted for 81% of retail sales, which fell to 61% by 2004.[61] Over the same period warehouse clubs and supercenters grew from 14% of retail sales to 35%. In 2021, supermarkets accounted for 56% of retail sales and warehouse clubs and supercenters accounted for 42%. Since the Ahold/Giant merger in 1998, warehouse clubs and supercenters have doubled their share of retail sales, while supermarkets’ share has dropped by more than 25%.[62] Put simply, the four largest supermarket retailers were occupying a larger share of a shrinking segment.

Based on these observations, the product-market definition that the FTC has employed in its consent orders over the past more than two decades is likely to be—and should be—challenged to include warehouse clubs, in addition to accounting for online retail and delivery.

B.     The Once-a-Week Shopper Is No Longer the Norm

In the past, the FTC has specified that, for a firm to be in the relevant market of “supermarkets,” it must be able to “enable[e] consumers to purchase substantially all of their weekly food and grocery shopping requirements in a single shopping visit.”[63] This definition suffers from several deficiencies.

The first deficiency is that this hypothetical consumer behavior is at odds with how many or most consumers behave today.

  • Surveys conducted by the Food Marketing Institute and The Hartman Group report the average shopper makes 1.6 weekly trips to buy groceries.[64]
  • Earlier research by FMI and Hartman find, in addition, other household members take another 0.6 trips, implying that the total number of trips per household each week is about 2.2, or approximately one trip every three days.[65]
  • Surveys conducted by Drive Research show the average household makes an average of 8.1 grocery shopping trips a month, or around two trips a week F(Table 1).[66]

There is no evidence that consumers view retailers that provide one-stop shopping for an entire week’s food and grocery needs as distinct from other retailers who provide food and groceries. In fact, evidence suggests that many consumers “multi-home” across several different retail categories.

  • Survey data published by Drive Research indicate that many households spread their shopping across grocery stores, mass merchants, warehouse clubs, independent grocery stores, natural and specialty grocery stores, dollar stores, and online retailers (Table 1).[67]
  • Acosta, a sales and marketing consulting firm, reports that 76% of consumers shop at more than one retailer a week and about one third “retail hop” among three or more retailers a week for groceries and staples.[68]
  • Research from the University of Florida found that, in 2017, an average consumer visited 3.2 to 4.3 different formats of food outlets a month, depending on income level.[69]
  • Survey results from PYMNTS, a data-analysis firm, demonstrate a significant shift in consumer spending away from traditional supermarkets. For example, in 2020, approximately 98% of consumers who bought at least one common household product each week—“center aisle” products, such as paper towels, cleaning supplies, and canned goods—made the purchase at a grocery store. In 2023, more than a third of consumers (37%) say they purchase none of these products from a traditional grocery store, with online purchases accounting for much of the shift.[70]

Thus, while one-stop weekly food and grocery shopping at single retailers was once typical, the evidence indicates such a phenomenon is much less common today.

C.     Supermarkets Compete with Warehouse Clubs

The FTC’s earlier consent orders provide four reasons to exclude warehouse clubs from the relevant market that includes supermarkets:

  1. They offer a “more limited range of products and services” than supermarkets;
  2. They “appeal to a distinct customer type” from supermarket customers;
  3. Shoppers do not view warehouse clubs as “adequate substitutes for supermarkets;” and
  4. Supermarkets do not view warehouse clubs as “significant or close competition,” relative to other supermarkets.[71]

In contrast to these conclusions, there is widespread recognition today that warehouse clubs impose significant competitive pressure on supermarkets.[72]

  • The National Academies of Sciences concludes that, over time, the entry and growth of warehouse clubs, superstores, and online retail has “blurred” the distinctions among retail formats.[73] More importantly for merger-review analysis, the National Academies concludes that the retail sector is “highly competitive,” in part because of the entry and growth of warehouse clubs, superstores, and online retail.[74]
  • Based on their empirical analysis, Paul Ellickson and coauthors conclude that warehouse clubs are “relevant substitutes” for supermarkets, even when the club stores are outside the geographic area typically used by the FTC in merger reviews.3[75]
  • Prior to her appointment as FTC chair, Lina Khan and her coauthor concluded that competition from warehouse clubs “fueled” grocery mergers in the late 1990s.[76]

The FTC’s consent orders note that warehouse clubs offer a “more limited range of products and services” than supermarkets. The orders identify products sold at supermarket as “including, but not limited to, fresh meat, dairy products, frozen foods, beverages, bakery goods, dry groceries, detergents, and health and beauty products.”[77]

The annual reports for Costco, Walmart (which include Sam’s Club), and BJ’s, however, indicate that each company offers the same range of products the FTC consent orders identify as being offered by supermarkets.[78] Indeed, the reports indicate that supermarkets do consider wholesale clubs to be competitors and vice versa. BJ’s annual report goes to great lengths to explain its efforts to compete with supermarkets by offering similar manufacturer-branded products at lower prices.[79] The role of warehouse clubs cannot be assumed or asserted away. Indeed, any identification of the relevant market for retail food and grocery sales should begin with a presumption that warehouse clubs provide competitive pressure, since that is what the economics research finds.

It is true that warehouse stores typically carry fewer stock-keeping units (“SKUs”) than supermarkets. Costco’s annual report indicates the company carries fewer than 4,000 SKUs in its warehouse stores and offers 10,000 to 11,000 SKUs for online purchases.[80] BJ’s annual report says that the company carries “approximately 7,000 core active stock keeping units.”[81] In contrast, the BJ’s report notes, “supermarkets normally carry an average of 40,000 SKUs, and supercenters may stock 100,000 SKUs or more.” But these differences in SKU counts do not in themselves demonstrate fundamental differences in product offerings between the two store formats.[82]

The main difference between the two formats is that warehouse clubs tend to offer a smaller range of sizes and packaging and a smaller variety of brands. For example, Costco notes that many of the products it stocks “are offered for sale in case, carton, or multiple-pack quantities only.”[83] Another difference can be observed in the variety of brands offered. For example, Costco has invested in developing its Costco Wholesale and Kirkland Signature private labels.[84] In contrast, CFRA Research analyst Arun Sundaram observes that Sam’s Club “typically offers a bigger national-brand product selection than its club cohorts.”[85] Such differences are insufficient to exclude wholesale clubs from the relevant market. If warehouse clubs offer similar products, see themselves as competing with supermarkets, and customers view them as substitutes, warehouse clubs must be in the same market.

D.    E-Commerce Has Changed the Food Landscape

In the years since the Ahold/Giant merger in 1998, online shopping and home delivery have grown from niche services serving only 10,000 households nationwide to a landscape where approximately one-in-eight consumers purchase groceries “exclusively” or “mostly” online.[86] This shift has increased competition and innovation in the supermarket industry, as traditional supermarkets have adapted to changing consumer preferences and behaviors by offering more delivery and pickup options, expanding their online assortments, and enhancing their digital capabilities.[87] Some have invested in their own e-commerce platforms and many have partnered with such third-party providers as Instacart, Shipt, and Peapod.[88]

One might surmise that e-commerce simply replaced in-person shopping, but with the same stores competing. This is not what has been observed. E-commerce has also increased competition by bringing in new companies with which traditional stores need to compete (e.g., Amazon) and by increasing the options available to consumers through services like Instacart, which allow for direct price and product comparisons among many stores. Each of these innovations has blurred the lines between brick-and-mortar food and grocery retailers and e-commerce, as well as the lines between supermarkets and other retail formats.

In 2023, more than a third of consumers (37%) say they purchase no center-aisle products (products such as paper towels, cleaning supplies, and canned goods) from a traditional grocery store, with online purchases accounting for much of the shift.[89] The fluidity between supermarkets, grocery stores, and online purchases makes the distinction nearly meaningless.

If a consumer uses Instacart to purchase and deliver groceries from Safeway, is that a supermarket purchase or e-commerce? What if the same consumer uses Instacart to purchase and deliver the same goods from Costco or Target? Does the consumer care which retailer the food and groceries came from when competition is just a click away? Indeed, the National Academies of Sciences concludes it is “often impossible” to distinguish between brick-and-mortar retail sales and e-commerce.[90]

E-commerce and club stores also matter for defining the geographic market. Past FTC consent orders have defined the relevant geographic markets to be areas that range from a two- to ten-mile radius around each of the merging parties’ supermarkets. We conclude that, because club stores have a larger catchment area than supermarkets, it would be reasonable to expand the relevant geographic market in localities where club stores are present.

Combined, the rapid growth of e-commerce and delivery services make distance, traffic patterns, and population density decreasingly relevant in a consumer’s choice of where they purchase food and groceries. Dimitropoulos and coauthors note (1) the presence of warehouse clubs expands the relevant geographic market, (2) online-delivery options expand the geographic market “far away,” and (3) online food and grocery purchases can be delivered from fulfillment centers, as well as from traditional stores.[91]

Because of these observations, the product market-definition that has been employed in the FTC’s consent orders over the past more than two decades is likely to be—and should be—challenged and should be revised to include warehouse clubs and to account for online retail and delivery.

IV.   The Merger Is Unlikely to Increase Labor Monopsony Power

In recent years, there has been an increasing emphasis in antitrust discussions on labor markets and potential harms to workers. The recent draft merger guidelines added an explicit section on mergers that “May Substantially Lessen Competition for Workers or Other Sellers.”[92] Even before the guidelines’ publication, some observers predicted that the FTC was set to push a case on labor competition.[93] While, in theory, antitrust harms can occur in labor markets, just as in product markets, proving that harm is more difficult.

An important fact about the proposed Kroger/Albertsons merger is that both companies have many unionized workers. Around two-thirds of Kroger employees[94] and a majority of Albertsons employees[95] are part of the United Food and Commercial Workers International Union (UFCW), which represents 1.3 million members. Even if the merger would increase labor monopsony power in the absence of unions, the FTC will have to acknowledge the reality of the unions’ own bargaining power.

Delegates of the UFCW unanimously voted to oppose the merger[96] Rather than monopsony power or lower wages, however, the union’s stated reason for their opposition was lack of transparency.[97] While lack of transparency may be problematic for the UFCW members, it does not constitute an antitrust harm. Kroger, for its part, has contended that the merger will benefit employees, citing a commitment to invest an additional $1 billion toward increased wages and expanded benefits, starting from the day the deal closes.[98] Albertsons claims that no store closures or frontline associate layoffs will result from the transaction and that the merger will “secure the long-term future of union jobs by establishing a more competitive alternative to large, non-union retailers.”[99]?As with most announced goals, however, there is no enforcement mechanism for this commitment at present, although one could be litigated.

Rather than relying on proclamations from any of the parties, we need economic analysis of the relevant labor markets, asking the types of questions raised above regarding the output markets. A policy report from Economic Policy Institute estimates that “workers stand to lose over $300 million annually” from the merger,[100] but the report arrives at that estimate by using an estimate of the correlation between concentration (HHI) in labor markets and wages. While academic research may benefit from such an estimate, it is unhelpful in merger analysis. As a long list of prominent antitrust economists recently wrote, “regressions of price on HHI should not be used in merger review… [A] regression of price on the HHI does not recover a causal effect that could inform the likely competitive effects of a merger.”[101]

While HHI regressions are of little practical help in this context, according to standard economic theory, it is possible that the average worker would be harmed for traditional labor-monopsony reasons. It is, however, more difficult to identify anticompetitive labor-market harms than to identify analogous product-market harms. For the product market, if the merger simply enhanced monopoly power without producing efficiency gains, the quantity sold would decrease, either because the merging parties raise prices or because quality declines. A merger that creates monopsony power will necessarily reduce the prices and quantity purchased of inputs like labor and materials. A strong union could counteract a firm’s monopsony power to some extent, by collectively advocating for higher wages, fewer layoffs, and other worker benefits. Indeed, obtaining and exerting labor market power is a union’s raison d’être.

Of course, the presence of a unionized workforce does not render monopsony impossible; unions’ ability to offset the effects of a monopsony or monopoly may also be limited, and monopsony power could increase under a merger, even with unions. Still, the existence of union bargaining power makes any monopsony case more difficult and is an important factor to consider in evaluating a merger’s likely labor-market effects—particularly in this case, given the high rates of union membership at both companies.

Moreover, the FTC needs to be careful with any labor case. For labor markets, a decline in the number of workers employed (which harms the workers) may not be anticompetitive. The reduction in input purchases may be because of efficiency gains.[102] For example, if two merging hospitals integrate their information-technology resources, therefore requiring fewer overlapping workers, the merged firm will employ fewer IT workers. This may even reduce the wages of specialized IT workers, even if the newly merged hospital does not exercise any market power to suppress wages.

The same applies for any inputs from an efficiency-enhancing merger: inputs may decrease. But using fewer inputs is not an antitrust harm. The key point is that monopsony needs to be treated differently than monopoly. The antitrust agencies cannot simply look at the quantity of inputs purchased in the monopsony case as the flip side of the quantity sold in the monopoly case, because the efficiency-enhancing merger can look like the monopsony merger in terms of the level of inputs purchased.

Another difficulty with pursuing a labor monopsony case is that the usual antitrust tools, such as merger simulation, cannot be easily applied to the labor market. Unlike the DOJ’s recent success in blocking Penguin-Random House from merging with Simon & Schuster on grounds that the merger would hurt authors with advances above $250,000,[103] the labor market for most employees of Kroger and Albertsons is much larger than those two companies, or even the largest definition of grocery stores. Indeed, it cannot be narrowed down to a handful of companies.

Any labor case would require showing that the merger would harm workers by reducing their bargaining power. For most workers involved, there are still many potential employers competing. One relevant piece of evidence for this is that press releases Kroger issued during the COVID-19 pandemic highlighted that the company was hiring workers from a wide variety of firms and industries—from hospitality (Marriott International) to restaurants (Waffle House) to food distribution (Sysco).[104] While we are not aware of publicly available data that would more comprehensively illustrated worker flows among different companies, such flows of retail workers into and out of roughly adjacent labor markets makes intuitive sense. As economist Kevin Murphy has explained:

If you look at where people go when they leave a firm or where people come from when they go to the firm, often very diffuse. People go many, many different places. If you look at employer data and you ask where do people go when they leave, often you’ll find no more than 5 percent of them go to any one firm, that they go all over the place. And some go in the same industry. Some go in other industries. Some change occupations. Some don’t. You look at plant closings, where people go. Again, not so often a big concentration of where they go to. If you look at data on where people are hired from, you see much the same patterns. That’s kind of a much more diffuse nature.[105]

If, as is likely, an overwhelming majority of Kroger’s workers’ next best option (what they would do if a store closed) was not an Albertsons store, but something completely outside of the market for grocery-store labor (or even outside the retail-food industry more broadly), the merger would not take away those workers’ next best option. If true, the merger cannot be said to increase labor monopsony power to the extent necessary to justify blocking a merger.

V.     ‘Waterbed Effects’ Are Highly Speculative

One potential antitrust harm that has been discussed frequently in recent years is the so-called “waterbed effect,” in which “a large and powerful firm improves its own terms of supply by exercising its bargaining power, [but] the terms of its competitors can deteriorate sufficiently so as ultimately to increase average retail prices and, thereby, reduce total consumer surplus.”[106] The waterbed effect is not unique to mergers, but can apply any time there is differential buyer-market power. The firm with more market power gets a better deal from suppliers and its competitors are ostensibly harmed because they cannot get the same deal. Long before the proposed merger, but still in the context of retail, people were speculating about a waterbed effect regarding Walmart.[107]

In the context of the Kroger/Albertsons merger, critics have again raised the possibility of a waterbed effect. Michael Needler Jr.—the president and chief executive of Fresh Encounter, a chain of 98 grocery stores based in Findlay, Ohio—raised the possibility in a U.S. Senate hearing on the merger.[108] He was also quoted by The New York Times as saying:

When the large power buyers demand full orders, on time and at the lowest cost, it effectively causes the water-bed effect. They push down, and the consumer packaged goods companies have no option but to supply them at their demands, leaving rural stores with higher costs and less availability to products.[109]

In a letter to the FTC, the American Antitrust Institute raised several concerns about the merger, arguing that:

The waterbed effect is likely to worsen with Kroger-Albertsons enhanced buyer power post-merger, with adverse effects on the ability of independent grocers to compete in a tighter oligopoly of large grocery chains.[110]

The implied argument in this version of the waterbed effect goes as follows: A merged Kroger and Albertsons would have additional market power over some of its suppliers. It could then exercise that market power to extract discounts from those suppliers, which would be unavailable to its competitors. The merged firm could then pass those cost savings on to consumers in the form of lower retail prices, thereby increasing Kroger/Albertsons’ sales. Some of these sales would come at the expense of smaller competitors, who could no longer compete on price. And because of these reduced sales, they would purchase less from their suppliers, further eroding their bargaining power with suppliers. Ultimately, consumers of the smaller retailers may face higher retail prices. Thus, under this theory, consumers of the merged firm would pay lower retail prices, while consumers of the smaller retailers would pay more.

Even if all of that were true, however, what remains unknown (and unaddressed under this argument) is whether consumers as a whole would be better or worse off. That, of course, is precisely the result that would be required to establish harm under antitrust law.

Roman Inderst & Tommaso Valletti are credited with the first formal theoretical model of a waterbed effect and how it could potentially harm consumers as a whole (as opposed to merely certain competitors or the subset of consumers who continue to buy from them).[111] In order to establish consumer harm under their model, three key assumptions must be met:

  1. Retailers buy their inputs from a supplier who can price discriminate among retailers;
  2. Retailers can access an alternative source of supply, but must incur a fixed switching cost, which is the same across all retailers; and
  3. Retailers compete on price.

Because a larger retailer can spread the fixed switching costs across more units, its per-unit costs will be lower. This provides the larger retailer with a better bargaining position with its suppliers to extract lower input prices. If the large retailer reduces its prices to consumers, the reduced sales to smaller competing retailers results in those competitors having higher per-unit switching costs, thus reducing their ability to change suppliers, reducing their bargaining power with the initial supplier, and increasing the price they pay to the supplier for inputs.

While the model shows how the effect is possible and that it could harm consumers, it does not imply that the waterbed effect necessarily harms consumers. In fact, the same waterbed effect would also occur if a merger generated efficiency gains (as the authors point out), but with considerably different welfare and antitrust implications. Setting aside mergers, in Inderst & Valletti’s model, if one firm discovers a cheaper importer, for example, it would give that firm more buyer power, because it presents a more credible threat of leaving for a competitor. Recognizing that the firm has a better “outside option,” the wholesaler offers better terms. This, too, generates a waterbed effect, but it is clearly pro-competitive, as it would help consumers. Unless we are willing to declare finding another source of supply to be anticompetitive, we should be hesitant about jumping to the conclusion that the waterbed effect is anticompetitive.

Inderst & Valletti’s model also demonstrates that, with relatively low supplier-switching costs, the supplier has little scope to price discriminate among retailers. As a result, “any further growth of the large buyer… will reduce all retail prices.”[112] This is true even in the presence of a waterbed effect. Thus, for a waterbed effect to result in higher average retail prices for consumers, the large retailer’s buying advantage must be “sufficiently larger in size,” and smaller retailers must face much high switching costs, with those switching costs serving as the reason why the supplier can effectively price discriminate across the retailers.[113] For many products, this simply won’t be the case.

A competition authority that pursued a waterbed theory to block a merger must first demonstrate that a waterbed effect exists. Because each product sold in a food and grocery retailer may have its own idiosyncratic manufacturing, wholesale, and distribution characteristics, this evaluation likely must be conducted on a product-by-product basis. Then, for each market, the authority must evaluate the suppliers’ abilities to price discriminate (which could raise additional antitrust issues). Last, the authority must evaluate competing firms’ anticipated price response to any identified waterbed effect. While Inderst & Valletti provide a seemingly straightforward theoretical approach to evaluating allegations of a waterbed effect, applying their model to the real world of food and grocery mergers would likely amount to a costly and time-consuming wild goose chase.

That is likely why finding empirical demonstration of a waterbed effect has been so elusive. Indeed, we are not aware of any empirical literature indicating the existence of a waterbed effect in retail markets, let alone any evidence of consumer harms from such an effect.[114] Indeed, UK competition authorities have been unconvinced of any waterbed effects in the food and grocery mergers in which the issue has been raised. In 2006, the UK Office of Fair Trading concluded:

[T]here are theoretical questions that would need to be resolved before concluding that the price differentials observed are evidence of a waterbed effect. For example, it is not clear how suppliers would be able to charge significantly above cost to smaller retailers without rivals undercutting them in the market; similarly, it is not clear why suppliers would price persistently below cost to the large supermarkets.[115]

In the United States, only one district court has issued an opinion on the waterbed effect. In DeHoog, consumers sued to block AB InBev’s acquisition of SABMiller.[116] The consumers alleged that the merged firm would be a “powerful buyer” that “demands lower prices or other concessions from its suppliers, causing the supplier to, in turn, increase prices to smaller buyers.” The district court rejected the consumers’ waterbed claim because (1) the alleged harm was to competing brewers, not to consumers, and (2) competing brewers could switch to different hops, thereby avoiding any waterbed effect.

DeHoog highlights the high hurdles an antitrust authority or private plaintiff would need to clear in order to successfully allege a waterbed effect. Challengers must demonstrate that switching costs are insurmountably high and that a waterbed effect exists. They must demonstrate then that the waterbed effect harms consumers, rather than competitors.

Demonstrating a waterbed effect in the Kroger/Albertsons merger may be especially challenging. Although the notion has been invoked by several critics of the merger, we are not aware of any specific product or product category in which a potential waterbed effect has been alleged.[117] If the FTC chooses to pursue the waterbed-effect theory, it must identify the relevant products that would be subject to the effect and demonstrate the anticipated consumer harm associated with it. If the agency relies on the waterbed effect in an effort to block the Kroger/Albertsons merger, then it would be reasonable to conclude that its “traditional” claims of horizontal market power are especially weak.

VI.   Remedies Can Resolve Any Remaining Competitive Concerns

While the above sections argue that the FTC will (and should) have a hard time making a case that the Kroger/Albertsons merger is overall anticompetitive, there may be some specific geographic markets where concerns remain. In the face of such concerns, the FTC historically has allowed most supermarket transactions to proceed with divestitures, such as Ahold/Delhaize (81 stores divested), Albertsons/Safeway (168 stores), and Price Chopper/Tops (12 stores).[118] The extent of the remedies sought depends on the extent of post-merger competition in the relevant markets, as well as the likelihood of entry by additional competitors.[119] Dimitropoulos and coauthors have noted that most divestitures required by consent orders in recent supermarket mergers have occurred in geographic markets with fewer than five remaining competitors.[120]

There is good reason (and a long history of examples in previous grocery-merger settlements) to think that targeted divestitures in certain markets—as have been proposed from the start of this process by the merging parties[121]—should be sufficient to address any geographic-market-specific concerns that may arise.

One reason that divestiture—instead of outright blocking—should be appropriate in this case is that the vast majority of Kroger and Albertsons stores are in geographic markets where the other is not located (Figure 3). As such, there is no antitrust concerns from a product-market perspective. The merger does not affect competition in the South (where Kroger is focused) or in the Northeast (where Albertsons is focused). In these regions, the merger generates all of the efficiencies without the possible downside of a loss to competition.

In some other geographic locations, however, the companies do currently compete, and antitrust concerns could therefore arise. This is where divestiture comes in. By most measures, there appear to be some 1,400 overlapping stores; resolving this overlap entails divestiture of no more than 700, or 14% of the two companies’ more than 5,000 stores.[122] By the same token, only a limited number of geographic markets have Kroger and Albertsons stores in close proximity, suggesting that targeted divestitures could address those concerns, while allowing the merger to proceed unimpeded in the great majority of markets.[123]

Previous remedies sought by the FTC in merger cases have generally been successful in achieving their goals. The FTC’s most recent merger-remedies study, covering 89 orders from 2006-2012, provides additional support for the feasibility of divestitures as an effective remedy.[124] The study found that the vast majority of divestitures succeeded in maintaining competition in the affected markets. All remedies involving divestiture of an ongoing business were successful. Divestitures of more limited “selected assets” also largely succeeded, although at a lower rate. Overall, the FTC concluded that more than 80% of the orders examined achieved the goal of maintaining or restoring competition post-merger.[125]

Nonetheless, despite a long history of divestitures serving as an appropriate and adequate remedy in food-retail mergers, some advocates for stronger antitrust are extremely skeptical of divestiture remedies. As authors from the American Economic Liberties Project and the Open Markets Institute put it in one recent article:

It should not fall on our overburdened antitrust enforcers to pore over the individual assets changing hands in service of coming up with a carve-out that somehow brings a merger into technical compliance with an arbitrary Reaganite standard devised by bad-faith ideologues.[126]

Such concerns are leveled at the grocery industry, in particular, with critics consistently pointing to the Albertsons/Safeway merger divestitures to Haggen as evidence that, in this industry (if not elsewhere), divestiture is no longer an appropriate merger remedy.[127] But these arguments ring hollow. Several factors idiosyncratic to Haggen and its acquisition strategy led to that divestiture’s failure.

A.     Distinguishing the 2014 Haggen Divestiture

In 2014, the parent company of Albertsons announced plans to purchase rival food and grocery chain Safeway for $9.4 billion.[128] Prior to the merger, Albertsons was the fifth-largest grocer in the United States and operated approximately 1,075 supermarkets in 29 U.S. states. At the time, Safeway was the second-largest and operated more than 1,300 stores nationwide.[129] During its merger review, the FTC identified 130 local markets in Western and Midwestern states where it alleged the merger would be anticompetitive.[130] In response, Albertsons and Safeway agreed to divest 168 supermarkets in those geographic markets.[131] Haggen Holdings LLC was the largest buyer of the divested stores, acquiring 146 Albertsons and Safeway stores in Arizona, California, Nevada, Oregon, and Washington.

Following the acquisition, Haggen almost immediately encountered numerous problems at the converted stores. Consumers complained of high prices, and sales plummeted at some stores. The company struggled and began selling some of its stores. Less than a year after the FTC announced the divestiture agreement, Haggen filed for bankruptcy. Following the bankruptcy, Albertsons bought back 33 of the stores it had divested in its merger with Safeway.

In retrospect, Haggen may not have been an appropriate buyer for the divested stores. Before acquiring the divested stores, Haggen was a small regional chain with only 18 stores, mostly in Washington State. The acquisition represented a tenfold increase in the number of stores the company would operate. While Haggen was once an independent, family-owned firm, at the time of the acquisition, the company was owned by a private investment firm that used a sale-leaseback scheme to finance the purchase. Christopher Wetzel notes that Haggen failed to invest sufficiently in the marketing necessary to create brand awareness in regions where Haggen had not previously operated.[132] Such issues would need to be avoided in any future divestitures, and experience shows they can be.

Around the same time that it filed for bankruptcy, Haggen also filed a lawsuit against Albertsons in federal district court, arguing that Albertsons engaged in “coordinated and systematic efforts to eliminate competition and Haggen as a viable competitor.”[133] Haggen claimed that Albertsons made false representations to both Haggen and the FTC about its commitment to providing a smooth transition that would allow Haggen to be a viable competitor. Among other allegations, Haggen claimed that Albertsons overstocked the divested stores with perishable meat and produce, provided inaccurate and misleading price information that led to inflated prices, and failed to perform maintenance on stores and equipment.

None of these claims were demonstrated, as the matter settled months after the complaint was filed. Even so, FTC consent orders typically provide asset-maintenance agreements to address the kinds of issues raised by Haggen. David Balto reports that, after the 1995 Schnucks/National merger, the FTC sued Schnucks, alleging that it had violated a provision of the asset-maintenance agreement in the consent order.[134] The suit resulted in a settlement in which Schnucks paid a $3 million civil penalty and was required to divest two additional properties. These two properties were stores that had been closed by Schnucks, but that presumably could be reopened by a new buyer.

The problems with the Haggen divestiture need not be repeated. In particular, there are many companies of various sizes that have the capabilities and desire to expand. In recent merger-consent orders, divested stores have been acquired by both retail supermarkets and wholesalers with retail outlets, including Publix, Supervalu, Big Y, Weis, Associated Wholesale Grocers, Associated Food Stores, and C&S Wholesale Grocers.[135] Several of these companies have successfully expanded—in some cases outside of their “home” territories. For example, Publix is a Florida-based chain that operates nearly 1,350 stores in seven southeastern states.[136] Publix expanded to North Carolina in 2014, Virginia in 2017, and has announced plans to expand into Kentucky this year.[137] Weis Markets is a Pennsylvania-based chain that operates more than 200 stores in seven northeastern states.[138] Last year, the company announced plans to spend more than $150 million on projects, including new retail locations and upgrades to existing facilities.[139] And Rochester, New York-based Wegmans has successfully entered Delaware, Virginia, and the District of Columbia in recent years.[140]

While the relevant product and geographic markets for supermarket mergers has shifted enormously over the past few decades, divestitures remain an appropriate and adequate remedy for any competitive concerns. The FTC has knowledge and experience with divestiture remedies and should have a good understanding of what works. In particular—and, perhaps, unlike in the Haggen example—firms acquiring divested assets should have an adequate cushion of capital, experience with the market conditions in which the stores are located, and the operational and marketing expertise to transition customers through the change.

B.     Proposed C&S Divestiture

As noted, Kroger and Albertsons have contemplated divestitures from the beginning, even including a provision in their merger agreement preemptively agreeing to divest up to 650 stores.[141] More recently, however, the companies have made their divestiture plans more concrete. In September 2023, the companies presented a proposal (both publicly and to the FTC) proposing to divest 413 stores, eight distribution centers, and three store brands to C&S Wholesale Grocers for $1.9 billion.[142] The agreement also allows C&S to purchase up to 237 additional stores if needed to resolve antitrust concerns. C&S also has committed to maintain any existing collective-bargaining agreements with labor unions.[143]

The specific characteristics of the proposed buyer of the divested stores suggests that it is unlikely to fall prey to the limitations that scuttled the Haggen divestiture. Unlike Haggen, the purchasing party here has experience operating more than 160 supermarkets under brands like Grand Union. This existing operation of stores makes C&S better positioned as a buyer than Haggen was when it attempted to rapidly expand from 18 to 168 stores.[144]

While C&S is primarily a wholesaler, its Grand Union retail operations and the transition support offered under the divestiture agreement should position it to successfully operate the divested stores. In that way, the divestiture does not just spin off or increase the size of a horizontal competitor. Rather, the plan jumpstarts greater vertical integration by C&S, whose wholesale operations include the production of private-label products.

Indeed, by enabling C&S to take better advantage of the benefits of vertical integration, the divestiture appears to ensure that C&S will emerge with a structure more in line with the rest of the food-retail market. Over the past decade, many retailers (including Kroger and Albertsons) have shifted toward “bringing private label production in-house.”[145] This move by firms (even without any market power) likely reflects competitive advantages gained from vertical integration.

The targeted nature of the divestiture would allow the merger to proceed in the majority of geographic markets where there are no competitive concerns between Kroger and Albertsons. Divesting stores only where localized overlaps in specific regions exist enables the realization of efficiencies and benefits in the many remaining markets. The FTC will still need to closely scrutinize the buyer and the proposed divestiture package. But the announced plan demonstrates that the merging parties are taking seriously the need for divestitures.

Of course, as with any complex business transaction, there is always some possibility that aspects of a divestiture may not fully go according to plan. The recent piece by Maureen Tkacik & Claire Kelloway throws out many of these possibilities.[146] It’s possible that C&S turns out to not want to run grocery stores but only wants to resell the properties. It’s possible that C&S will be unable to afford the leases. Regulators and merging parties alike operate under inherent uncertainty when predicting competitive outcomes. Antitrust analysis does not deal with certainties, but rather with probabilistic assessments of likely competitive effects.

The relevant question is whether the divestiture is likely to effectively maintain competition in the markets of concern, not whether it can be guaranteed to perfectly do so in all scenarios. When we take more episodes than Haggen’s into account, despite the uncertainty, the FTC’s experience shows that targeted divestitures with an experienced buyer are likely to adequately protect consumers post-merger. The possibility that some unforeseen complication may arise does not negate the high probability that competition will be preserved. Antitrust regulation requires reasonable predictive judgments, acknowledging that business transactions inherently carry risks.

With the FTC’s knowledge of the industry and of its own past successes and failures, divestitures remain an appropriate and adequate remedy for this merger. The parties appear committed to working cooperatively with regulators to craft divestitures that fully resolve competitive concerns. Rather than blocking the deal outright, the FTC can allow the merger to proceed, conditioned on acceptable divestitures that protect consumers, while permitting efficiency gains across the majority of stores.

 

[1] Press Release, Kroger and Albertsons Companies Announce Definitive Merger Agreement, Kroger (Oct. 14, 2022), https://ir.kroger.com/CorporateProfile/press-releases/press-release/2022/Kroger-and-Albertsons-Companies-Announce-Definitive-Merger-Agreement/default.aspx.

[2] In an article written with Sandeep Vaheesan before she became chair of the FTC, Lina Khan expressed disdain for grocery-industry consolidation and deep skepticism of even the best divestiture packages. See Lina Khan & Sandeep Vaheesan, Market Power and Inequality: The Antitrust Counterrevolution and Its Discontents, 11 Harv. L. & Pol’y Rev. 235, 254 (2017) (“Retail consolidation has enabled firms to squeeze their suppliers… and led to worse outcomes for consumers.”) & 289 (“Even if divestitures could be perfectly tailored and if they preserved competition in narrow markets in every instance, they would fail to advance the citizen interest standard.”).

[3] See, e.g., David Dayen, Proposed Kroger-Albertsons Merger Would Create a Grocery Giant, The American Prospect (Oct. 17, 2022), https://prospect.org/power/proposed-kroger-albertsons-merger-would-create-grocery-giant; Richard Smoley, Kroger, Albertsons, and Lina Khan, Blue Book Services (May 2, 2023), https://www.producebluebook.com/2023/05/02/kroger-albertsons-and-lina-khan.

[4] U.S. Dep’t of Justice & Fed. Trade Comm’n, Draft Merger Guidelines (Jul. 19, 2023), available at https://www.justice.gov/d9/2023-07/2023-draft-merger-guidelines_0.pdf. See also Gus Hurwitz & Geoffrey Manne, Antitrust Regulation by Intimidation, Wall St. J. (Jul. 24, 2023), https://www.wsj.com/articles/antitrust-regulation-by-intimidation-khan-kanter-case-law-courts-merger-27f610d9.

[5] Prior to Whole Foods/Wild Oats, the last litigated supermarket merger was the State of California’s 1988 challenge to American Store’s acquisition of Lucky Stores. Several retail mergers have been challenged in court, however, such as Staples/Office Depot in 2015. See Press Release, FTC Challenges Proposed Merger of Staples, Inc. and Office Depot, Inc., Federal Trade Commission (Dec. 7, 2015), https://www.ftc.gov/news-events/news/press-releases/2015/12/ftc-challenges-proposed-merger-staples-inc-office-depot-inc.

[6] This includes approving Albertsons/Safeway (2015), Ahold/Delhaize (2016), and Price Chopper/Tops (2022). See Analysis of Agreement Containing Consent Order to Aid Public Comment, In the Matter of Cerberus Institutional Partners V, L.P., AB Acquisition, LLC, and Safeway Inc. (File No. 141 0108) (Jan. 27, 2015) available at https://www.ftc.gov/system/files/documents/cases/150127cereberusfrn.pdf; Analysis of Agreement Containing Consent Order to Aid Public Comment, In the Matter of Koninklijke Ahold N.V. and Delhaize Group NV/SA (File No. 151-0175) (Jul. 22, 2016), available at https://www.ftc.gov/system/files/documents/cases/160722koninklijkeanalysis.pdf; Analysis of Agreement Containing Consent Order to Aid Public Comment, In the Matter of The Golub Corporation and Tops Markets Corporation (File No. 211-0002, Docket No. C-4753) (Nov. 8, 2021), available at https://www.ftc.gov/system/files/documents/cases/2110002pricechoppertopsaapc.pdf.

[7] Decision and Order, In the Matter of Whole Foods Market, Inc., (Docket No. 9324) (May 28, 2009), available at https://www.ftc.gov/sites/default/files/documents/cases/2009/05/090529wfdo.pdf; FTC v. Whole Foods Market, 548 F.3d 1028 (D.C. Cir. 2008).

[8] Number based on authors’ calculations, using data from 90th Annual Report, Progressive Grocer (May 2023), https://progressivegrocer.com/crossroads-progressive-grocers-90th-annual-report.

[9] See Draft Merger Guidelines, supra note 4.

[10] FTC v. Meta Platforms Inc., U.S. Dist. LEXIS 29832 (2023); FTC v. Microsoft Corporation et al., No. 23-cv-02880-JSC (N.D. Cal. Jul. 10, 2023), available at https://s3.documentcloud.org/documents/23870711/ftc-v-microsoft-preliminary-injunction-opinion.pdf.

[11] See, e.g., George Kuhn, Grocery Shopping Consumer Segmentation, Drive Research (2002), available at https://www.driveresearch.com/media/4725/final-2022-grocery-segmentation-report.pdf.

[12] See, e.g., In the Matter of Cerberus Institutional Partners, supra note 6, at 3.

[13] Press Release, Kroger and Albertsons Companies Announce Comprehensive Divestiture Plan with C&S Wholesale Grocers, LLC in Connection with Proposed Merger, The Kroger Co. (Sep. 8, 2023), https://www.prnewswire.com/news-releases/kroger-and-albertsons-companies-announce-comprehensive-divestiture-plan-with-cs-wholesale-grocers-llc-in-connection-with-proposed-merger-301921933.html.

[14] Id.

[15] See Press Release, supra note 1.

[16] Progressive Grocer, supra note 8.

[17] Who Are the Top 10 Grocers in the United States?, Foodindustry.com (last visited Oct. 10, 2023), https://www.foodindustry.com/articles/top-10-grocers-in-the-united-states-2019.

[18] Id.

[19] Number based on authors’ calculations, using data from Progressive Grocer Staff, 90th Annual Report, Progressive Grocer (May 2023), https://progressivegrocer.com/crossroads-progressive-grocers-90th-annual-report.

[20] Kroger, supra note 13.

[21] See Khan & Vaheesan, supra note 2.

[22] See Leah Nylen & Jeannette Neumann, California Preparing Lawsuit to Block Merger of Kroger, Jewel Parent, Bloomberg (Oct. 12, 2023), https://www.chicagobusiness.com/retail/california-preparing-lawsuit-block-kroger-albertsons-deal.

[23] Alexander Coolidge, Report: Kroger CEO Is “Committed to Litigate” If FTC Regulators Fight Albertsons Merger, Cincinnati Enquirer (May 11, 2023), https://www.cincinnati.com/story/money/2023/05/11/kroger-commited-to-litigate-if-ftc-blocks-albertsons-deal/70206692007.

[24] Press Release, FTC Challenges Proposed Merger of Staples, Inc. and Office Depot, Inc., Federal Trade Commission (Dec. 7, 2015), https://www.ftc.gov/news-events/news/press-releases/2015/12/ftc-challenges-proposed-merger-staples-inc-office-depot-inc.

[25] Jesse Greenspan, FTC To Challenge Whole Foods, Wild Oats Merger, Law360 (Jun. 5, 2007), https://www.law360.com/texas/articles/26191/ftc-to-challenge-whole-foods-wild-oats-merger.

[26] State of Cal. v. American Stores Co., 872 F.2d 837 (9th Cir. 1989) (granting preliminary injunction).

[27] See, e.g., In the Matter of Cerberus Institutional Partners, supra note 6, at 2-3.

[28] See, e.g., Food Marketing Institute & The Hartman Group, Consumers’ Weekly Grocery Shopping Trips in the United States from 2006 to 2022 (Average Weekly Trips per Household), Statista (May 2022), available at https://www.statista.com/statistics/251728/weekly-number-of-us-grocery-shopping-trips-per-household.

[29] See, e.g., In the Matter of Cerberus Institutional Partners, supra note 6, at 3.

[30] Safeway Merger Report, UK Competition Commission (2003), available at https://webarchive.nationalarchives.gov.uk/ukgwa/20120119163858/http:/www.competition-commission.org.uk/inquiries/completed/2003/safeway/index.htm (“Overall, therefore, there is little evidence of an immediate or short-term ‘waterbed’ effect. … [O]ur surveys produced insufficient evidence on this point for us to conclude that any waterbed effect would be exacerbated by any of the mergers.”); Anticipated Merger between J Sainsbury PLC and ASDA Group Ltd: Summary of Final Report, UK Competition & Markets Authority (Apr. 25, 2019), available at https://assets.publishing.service.gov.uk/media/5cc1434ee5274a467a8dd482/Executive_summary.pdf (“Overall, it seems unlikely that many retailers will raise their prices in response to the Merger; and even if some individual retailers do, the overall effect on UK households is unlikely to be negative. On that basis, our finding is that the Merger is unlikely to lead to customer harm through a waterbed effect.”).

[31] DeHoog v. Anheuser-Busch InBev, SA/NV, No. 1:15-CV-02250-CL, 2016 U.S. Dist. LEXIS 137759, at *13-16 (D. Or. July 22, 2016).

[32] Leading to truculent statements like that of California Attorney General Rob Bonta that “[r]ight now, there’s not a lot of reason not to sue [to block the merger].” See Nylen & Neumann, supra note 19.

[33] See, e.g., Dayen, supra note 3; Smoley, id.

[34] For instance, the Herfindahl–Hirschman Index (HHI) at which mergers are deemed problematic has been lowered from 2500 (and a post-merger increase of 200) to 1800 (and a post-merger increase of 100). Likewise, combined market shares of more than 30% are generally deemed problematic under the new guidelines (if a merger also increase the market’s HHI by 100 or more). The revised guidelines also focus more heavily on monopsony and labor-market issues. See Draft Merger Guidelines, supra note 4, at 6-7.

[35] John Asker et al, Comments on the January 2022 DOJ and FTC RFI on Merger Enforcement, available at https://www.regulations.gov/comment/FTC-2022-0003-1847 at 15-16 (emphasis added).

[36] Draft Merger Guidelines, supra note 4, at 21.

[37] See U.S. Dep’t of Justice & Fed. Trade Comm’n, 2010 Horizontal Merger Guidelines (Aug. 19, 2010) at §5.3, available at https://www.justice.gov/atr/horizontal-merger-guidelines-08192010#5c.

[38] Draft Merger Guidelines, supra note 4, at 26.

[39] Chad Syverson, Macroeconomics and Market Power: Context, Implications, and Open Questions 33 J. Econ. Persp. 23, 27 (2019).

[40] See U.S. Dep’t of Justice, Merger Guidelines (1968) at 6-7, available at https://www.justice.gov/archives/atr/1968-merger-guidelines.

[41] United States v. Von’s Grocery Co., 384 U.S. 270 (1966).

[42] See, e.g., Robert H. Bork, The Goals of Antitrust Policy, 57 Am. Econ. Rev. Papers & Proceedings 242 (1967) (“In the Von’s Grocery case a majority of the Supreme Court was willing to outlaw a merger which did not conceivably threaten consumers in order to help preserve small groceries in the Los Angeles area against the superior efficiency of the chains.”).

[43] Supra note 10; FTC v. Illumina, Inc., U.S. Dist. LEXIS 75172 (2021).

[44] United States v. Bertelsmann SE & Co. KGaA, No. CV 21-2886-FYP, 2022 WL 16949715 (D.D.C. Nov. 15, 2022).

[45] Id. (“The defendants do not dispute that if advances are significantly decreased, some authors will not be able to write, resulting in fewer books being published, less variety in the marketplace of ideas, and an inevitable loss of intellectual and creative output.”)

[46] Brian Albrecht, Business as Usual for Antitrust, City Journal (Nov 22, 2022), available at https://www.city-journal.org/article/business-as-usual-for-antitrust.

[47] Lina M. Khan, Rohit Chopra, & Kelly Slaughter, Comm’rs, Fed. Trade Comm’n, Statement on the Withdrawal of the Vertical Merger Guidelines (Sep. 15, 2021) at 3, available at https://www.ftc.gov/system/files/documents/public_statements/1596396/statement_of_chair_lina_m_khan_commissioner_rohit_chopra_and_commissioner_rebecca_kelly_slaughter_on.pdf.

[48] Premerger Notification Rules, 88 Fed. Reg. 42178 (RIN 3084-AB46), proposed Jun. 29, 2023 (to be codified at 16 C.F.R. Parts 801 and 803).

[49] Antitrust Experts Reject FTC/DOJ Changes to Merger Process, U.S. Chamber of Commerce (Sep. 19, 2023), https://www.uschamber.com/finance/antitrust/antitrust-experts-reject-ftc-doj-changes-to-merger-process. The surveyed group was made up seasoned antitrust veterans from across a variety of backgrounds: 80% had been involved in more than 50 mergers and 59% in more than 100.

[50] Supra note 23.

[51] Analysis of Agreement Containing Consent Order to Aid Public Comment, In the Matter of Cerberus Institutional Partners V, L.P., AB Acquisition, LLC, and Safeway Inc. (File No. 141 0108) (Jan. 27, 2015) available at https://www.ftc.gov/system/files/documents/cases/150127cereberusfrn.pdf. Analysis of Agreement Containing Consent Order to Aid Public Comment, In the Matter of Koninklijke Ahold N.V. and Delhaize Group NV/SA (Jul. 22, 2016) (File No. 151-0175) available at https://www.ftc.gov/system/files/documents/cases/160722koninklijkeanalysis.pdf.

[52] In the Matter of Cerberus Institutional Partners, supra note 6, at 2.

[53] In this paper, the terms “hypermarket” and “supercenter” are used synonymously. See Richard Volpe, Annemarie Kuhns, & Ted Jaenicke, Store Formats and Patterns in Household Grocery Purchases, Economic Research Service Economic Information Bulletin No. 167 (Mar. 2017), https://www.ers.usda.gov/webdocs/publications/82929/eib-167.pdf?v=0 (supercenters are also known as hypermarkets or superstores).

[54] In the Matter of Cerberus Institutional Partners, supra note 6, at 3.

[55] Id.

[56] Id.

[57] See In the Matter of Koninklijke Ahold, supra note 6.

[58] Data obtained from: U.S. Census Bureau, Report on Retail Sales and Trends: Annual Retail Trade Survey: 2021, https://www.census.gov/data/tables/2021/econ/arts/annual-report.html.

[59] Khan & Vaheesan, supra note 2, at 255.

[60] Id. (“Grocers sought to bulk up in order to compete with the scale of warehouse clubs and large discount stores, fueling further mergers and leading many local grocers to close….”).

[61] U.S. Census Bureau, supra note 58.

[62] Id.

[63] In the Matter of Cerberus Institutional Partners, supra note 6, at 2.

[64] Food Marketing Institute & The Hartman Group, Consumers’ Weekly Grocery Shopping Trips in the United States from 2006 to 2022 (Average Weekly Trips per Household), Statista (May 2022), available at https://www.statista.com/statistics/251728/weekly-number-of-us-grocery-shopping-trips-per-household.

[65] Michael Browne, Grocery Shopping Has a Hold on Consumers, Study Finds, Supermarket News (Jun. 27, 2018), https://www.supermarketnews.com/issues-trends/grocery-shopping-has-hold-consumers-study-finds.

[66] Kuhn, supra note 11.

[67] Id.

[68] Trip Drivers: Top Influencers Driving Shopper Traffic, Acosta (2017), available at https://acostastorage.blob.core.windows.net/uploads/prod/newsroom/publication_phetw_0rzq.pdf.

[69] Lijun Angelia Chen & Lisa House, US Food Shopper Trends in 2017, Univ. of Fla, IFAS Extension Pub. No. FE1126 (Dec. 7, 2022), https://edis.ifas.ufl.edu/publication/FE1126.

[70] Karen Webster, Consumer Shopping Data Shows Troubling Signs for Grocery Stores’ Future, PYMNTS (Feb. 6, 2023), https://www.pymnts.com/news/retail/2023/consumer-shopping-data-shows-troubling-signs-for-grocery-stores-future.

[71] See, e.g., In the Matter of Cerberus Institutional Partners, supra note 6, at 3.

[72] See Paul B. Ellickson, Paul L.E. Grieco, & Oleksii Khvastunov, Measuring Competition in Spatial Retail, 51 RAND J. Econ. 189 (2020) (“[C]lub stores are able to draw revenue from a significantly larger geographic area than traditional grocers. Hence, club stores are relevant substitutes for grocery stores, even if they are located even several miles away, a fact that could easily be overlooked in an analysis in which stores are simply clustered by geographic market.”).

[73] National Academies of Sciences, Engineering, and Medicine, A Satellite Account to Measure the Retail Transformation: Organizational, Conceptual, and Data Foundations (2021), available at https://www.bls.gov/evaluation/a-satellite-account-to-measure-the-retail-transformation.pdf (“[T]he restructuring that started first with the warehouse clubs and superstores and then moved on to e-commerce has begun to blur the lines between the retail industry and several other sectors….”).

[74] Id. at 25 (“[C]hanges experienced by retail over the past few decades suggest that the sector is highly competitive and is undergoing substantial change and reorganization. As discussed earlier, the changes described involve warehouse clubs and superstores … e-commerce … digital goods, imports, and large firms….”).

[75] Ellickson et al., supra note 72, (“Due to their size and attractiveness for larger purchases, club stores represent strong competitors to grocery stores even, when they are a significant distance away.”).

[76] Khan & Vaheesan, supra note 2, at 255 (“Grocers sought to bulk up in order to compete with the scale of warehouse clubs and large discount stores, fueling further mergers and leading many local grocers to close….”).

[77] See, e.g., In the Matter of Cerberus Institutional Partners, supra note 6, at 2.

[78] Costco Wholesale Corporation, Annual Report (Form 10-K) (Aug. 28, 2022), https://www.sec.gov/ix?doc=/Archives/edgar/data/0000909832/000090983222000021/cost-20220828.htm; BJ’s Wholesale Club Holdings, Inc., Annual Report (Form 10-K) (Mar. 16, 2023), https://www.sec.gov/ix?doc=/Archives/edgar/data/1531152/000153115223000026/bj-20230128.htm; Walmart Inc., Annual Report (Form 10-K) (Mar. 27, 2023), https://www.sec.gov/ix?doc=/Archives/edgar/data/104169/000010416923000020/wmt-20230131.htm.

[79] BJ’s Wholesale Club Holdings, Inc, id.

[80] Costco Wholesale Corporation, id.

[81] BJ’s Wholesale Club Holdings, Inc, id.

[82] Id.

[83] Costco Wholesale Corporation, id.

[84] Id.

[85] Russell Redman, Report: Club Stores Absorbing Grocery Market Share from Supermarkets, Winsight Grocery Business (Apr. 20, 2023), https://www.winsightgrocerybusiness.com/retailers/report-club-stores-absorbing-grocery-market-share-supermarkets.

[86] Hean Tat Keh & Elain Shieh, Online Grocery Retailing: Success Factors and Potential Pitfalls, 44 Bus. Horizons 73 (Jul.-Aug., 2001); Appinio & Spryker, Share of Consumers Purchasing Groceries Online in the United States in 2022, by Channel, Statista (Sep. 2002).

[87] Navigating the Market Headwinds: The State of Grocery Retail 2022, McKinsey & Co. (May 2022), available at https://www.mckinsey.com/~/media/mckinsey/industries/retail/how%20we%20help%20clients/the%20state%20of%20grocery%20retail%202022%20north%20america/mck_state%20of%20grocery%20na_fullreport_v9.pdf.

[88] Id.; Dimitri Dimitropoulos, Renée M. Duplantis, & Loren K. Smith, Trends in Consumer Shopping Behavior and Their Implications for Retail Grocery Merger Reviews, CPI Antitrust Chron. (Dec. 2021), available at https://www.brattle.com/wp-content/uploads/2022/01/Trends-in-Consumer-Shopping-Behavior-and-their-Implications-for-Retail-Grocery-Merger-Review.pdf.

[89] See Webster, supra note 70.

[90] National Academies of Sciences, Engineering, and Medicine, supra note 73 (“As e-commerce has grown in recent years, it has become increasingly difficult to separate out the e-commerce portion of the industry. Most e-commerce could be identified within the nonstore retailer category as of 2013, but e-commerce is becoming so pervasive that it is now not only difficult to clearly identify individual firms as predominantly e-commerce firms, but also often impossible to clearly classify individual retail sales as either e-commerce or not.” citations omitted).

[91] Dimitropoulos, et al., supra note 88 (“Of course, adjustments to geographic market definition likely would need to be factored into the analysis, as club stores tend to have larger catchment areas than traditional grocery stores, and online delivery can reach as far away as can be travelled by truck from a central fulfilment center.”)

[92] Draft Merger Guidelines, supra note 4, at 25-7.

[93] Maeve Sheehey & Dan Papscun, Kroger-Albertsons Merger Tests FTC’s Focus on Labor Competition, Bloomberg Law (Dec. 2, 2022) https://news.bloomberglaw.com/antitrust/kroger-albertsons-merger-tests-ftcs-focus-on-labor-competition.

[94] Kroger Union, UFCW (last accessed Jul. 26, 2023), https://www.ufcw.org/actions/campaign/kroger-union.

[95] Albertsons and Safeway Union, UFCW (last accessed Jul. 26, 2023), https://www.ufcw.org/actions/campaign/albertsons-and-safeway-union.

[96] Press Release, America’s Largest Union of Essential Grocery Workers Announces Opposition to Kroger and Albertsons Merger, UFCW (May 5, 2023), https://www.ufcw.org/press-releases/americas-largest-union-of-essential-grocery-workers-announces-opposition-to-kroger-and-albertsons-merger.

[97] Id. (“Given the lack of transparency and the impact a merger between two of the largest supermarket companies could have on essential workers – and the communities and customers they serve – the UFCW stands united in its opposition to the proposed Kroger and Albertsons merger”).

[98] Press Release, Kroger and Albertsons Companies Announce Definitive Merger Agreement, Kroger (Oct. 14, 2022), https://ir.kroger.com/CorporateProfile/press-releases/press-release/2022/Kroger-and-Albertsons-Companies-Announce-Definitive-Merger-Agreement/default.aspx.

[99] Bill Wilson, Teamsters Union Says ‘No’ to Kroger, Albertsons Merger, Supermarket News (Jun. 13, 2023),  https://www.supermarketnews.com/retail-financial/teamsters-union-says-no-kroger-albertsons-merger.

[100] Ben Zipperer, Kroger-Albertsons Merger Will Harm Grocery Store Worker Wages, Economic Policy Institute (May 1, 2023), https://www.epi.org/publication/kroger-albertsons-merger.

[101] Nathan Miller et al., On the Misuse of Regressions of Price on the HHI in Merger Review, 10 J. Antitrust Enforcement 248 (2022).

[102] See Geoffrey A. Manne, Dirk Auer, Brian C. Albrecht, Eric Fruits & Lazar Radi?, Comments of the International Center for Law and Economics on the DOJ-FTC Request for Information on Merger Enforcement (2022), at 29, available at https://papers.ssrn.com/sol3/papers.cfm?abstract_id=4090844.

[103] See United States v. Bertelsmann SE & Co. KGaA, 1:21-CV-02886, 2022 WL 16949715 (D.D.C. Nov. 15, 2022).

[104] Press Release, The Kroger Family of Companies Provides New Career Opportunities to 100,000 Workers, Kroger (May 14, 2020), https://ir.kroger.com/CorporateProfile/press-releases/press-release/2020/The-Kroger-Family-of-Companies-Provides-New-Career-Opportunities-to-100000-Workers/default.aspx. While the exact job-to-job switches are unknown, at least during the pandemic we know that some workers at non-grocery employers viewed at least some grocery-industry jobs as substitutes.

[105] Transcript of Proceedings at the Public Workshop Held by the Antitrust Division of the United States Department of Justice, U.S. Justice Department (Sep. 23, 2019), available at https://www.justice.gov/atr/page/file/1209071/download.

[106] Roman Inderst & Tommaso M. Valletti, Buyer Power and the ‘Waterbed Effect’ 59 J. Ind. Econ. 1, 2 (2011).

[107] Albert Foer, Mr. Magoo Visits Wal-Mart: Finding the Right Lens for Antitrust, American Antitrust Institute Working Paper No. 06-07, https://papers.ssrn.com/sol3/papers.cfm?abstract_id=1103609.

[108] Michael Needler Jr., Senate Hearing on Kroger and Albertsons Grocery Store Chains, at 1:43:00, available at https://www.c-span.org/video/?524439-1/senate-hearing-kroger-albertsons-grocery-store-chains.

[109] Julie Creswell, Kroger-Albertsons Merger Faces Long Road Before Approval, New York Times (Jan. 23, 2023), https://www.nytimes.com/2023/01/23/business/kroger-albertsons-merger.html.

[110] Diana Moss, The American Antitrust Institute to the Honorable Lina M. Khan, American Antitrust Institute (Feb. 7, 2023), available at https://www.antitrustinstitute.org/wp-content/uploads/2023/02/Kroger-Albertsons_Ltr-to-FTC_2.7.23.pdf.

[111] Inderst & Valletti, supra note 106. For a short history of the development of the waterbed model, see Eric Fruits, Sloshing Around with the “Waterbed Effect,” Truth on the Market (Sep. 5, 2023), https://truthonthemarket.com/2023/09/05/sloshing-around-with-the-waterbed-effect.

[112] Inderst & Valletti, supra note 106, at 9.

[113] Id. at 10.

[114] There has been some investigation of the waterbed effect in two-sided markets in telecommunications, but these markets are very different from retail food and grocery. See Christos Genakos & Tommaso Valletti, Testing the “Waterbed” Effect in Mobile Telephony, 9 J. Eur. Econ. Assoc. 1114 (Dec. 2011) (evaluating the effect of cutting mobile-termination fees on mobile-subscription prices).

[115] UK Competition & Markets Authority, supra note 30.

[116] DeHoog v. Anheuser-Busch InBev, supra note 31.

[117] See, e.g., Albert Foer, Mr. Magoo Visits Wal-Mart: Finding the Right Lens for Antitrust, American Antitrust Institute Working Paper No. 06-07 (Nov. 30, 2006), available at https://ssrn.com/abstract=1103609 (alleging a Walmart waterbed effect without identifying any product or product category with the relevant characteristics that would make it subject to the effect).

[118] See In the Matter of Cerberus Institutional Partners, supra note 6; In the Matter of Koninklijke Ahold, id., In the Matter of the Golub Corporation, id.

[119] See Dimitropoulos, et al., supra note 88.

[120] See id.

[121] See, e.g., Abigail Summerville and Anirban Sen, Analysis: Kroger, Albertsons Spin-Off Is Extra Ammunition in Regulatory Battle, Reuters (Oct. 17, 2022), https://www.reuters.com/business/retail-consumer/kroger-albertsons-spin-off-is-extra-ammunition-regulatory-battle-2022-10-17 (“Kroger Co and Albertsons Cos Inc are willing to divest up to 650 supermarket stores to secure regulatory clearance for their $24.6 billion deal….”); Dan Papscun, Kroger-Albertsons Divestiture Bid Aims to Head Off Challenge, Bloomberg Law (Oct. 14, 2022) https://news.bloomberglaw.com/antitrust/kroger-albertsons-divestiture-plan-is-bid-to-deflect-regulators (“The FTC must factor the divestiture proposal in its deal analysis, now that the companies themselves have built it into their own proposal, said Steven Cernak, a Bona Law partner. The companies’ divestiture proposal makes the tie-up ‘a tougher deal for the FTC to challenge,’ Cernak said.”).

[122] See Kroger/Albertsons: Companies Have Overlap of More Than 1,400 Stores; Khan Highly Critical of Failed Supermarket Divestitures, The Capitol Forum (Nov. 2, 2022), https://thecapitolforum.com/kroger-albertsons-companies-have-overlap-of-more-than-1400-stores-khan-highly-critical-of-failed-supermarket-divestitures.

[123] Id.

[124] The FTC’s Merger Remedies 2006-2012: A Report of the Bureaus of Competition and Economics, Fed. Trade Comm’n (Jan. 2017), available at https://www.ftc.gov/system/files/documents/reports/ftcs-merger-remedies-2006-2012-report-bureaus-competition-economics/p143100_ftc_merger_remedies_2006-2012.pdf.

[125] Id. at 2.

[126] Maureen Tkacik & Claire Kelloway, The No Spin-Off Zone, The American Prospect (Oct. 11, 2023), https://prospect.org/power/2023-10-11-no-spin-off-zone-kroger-albertsons-merger/.

[127] See, e.g., Dayen, supra note 3 (“As the Haggen affair makes clear, the whole idea of using conditions to allow high-level mergers and competition simultaneously has been a failure.”). See also Tkacik & Kelloway, id. (“The Kroger-Albertsons merger shows us why regulators need to permanently divest the concept of, well, divesting.”).

[128] Scott Neuman, Grocery Chains Safeway and Albertson’s Announce Merger Deal, The Two Way (Mar. 6, 2014), https://www.npr.org/sections/thetwo-way/2014/03/06/286935900/grocery-chains-safeway-and-albertsons-announce-merger-deal.

[129] See In the Matter of Cerberus Institutional Partners, supra note 6, at 2.

[130] See id., at 3-5.

[131] See id., at 5.

[132] Christopher A. Wetzel, Strict(er) Scrutiny: The Impact of Failed Divestitures on U.S. Merger Remedies, 64 Antitrust Bull. 341 (2019).

[133] Jon Talton, Haggen: What Went Wrong?, Seattle Times (Mar. 15, 2016), https://www.seattletimes.com/business/economy/haggen-what-went-wrong.

[134] David A. Balto, Supermarket Merger Enforcement, 20 J. Pub. Pol’y & Marketing 38 (Spr. 2001).

[135] See In the Matter of Cerberus Institutional Partners, supra note 6; In the Matter of Koninklijke Ahold, id., In the Matter of the Golub Corporation, id.

[136] See Facts and Figures, Publix (last visited Oct. 10, 2023), https://corporate.publix.com/about-publix/company-overview/facts-figures.

[137] See Caroline A., The History of Publix: Entering New States, The Publix Checkout (Jan. 4, 2018), https://blog.publix.com/publix/the-history-of-publix-entering-new-states; Press Release, Publix Breaks Ground on First Kentucky Store and Announces Third Location, Publix (Jun. 23, 2022), https://corporate.publix.com/newsroom/news-stories/publix-breaks-ground-on-first-kentucky-store-and-announces-third-location.

[138] Weis Markets, LinkedIn https://www.linkedin.com/company/weis-markets/about, (last accessed Jul. 26, 2023).

[139] Sam Silverstein, Weis Markets Unveils $150M Expansion and Upgrade Plan, Grocery Dive (May 2, 2022), https://www.grocerydive.com/news/weis-markets-unveils-150m-expansion-and-upgrade-plan/623015.

[140] Russell Redman, Wegmans lines up its next new store locations, Winsight Grocery Business (Dec. 1, 2022) https://www.winsightgrocerybusiness.com/retailers/wegmans-lines-its-next-new-store-locations.

[141] See, e.g., Summerville & Sen, supra note 121.

[142] See Catherine Douglas Moran & Petyon Giora, Mapping Kroger and Albertson’s Store Divestiture Deal with C&S, Grocery Dive (Sept. 12, 2023), https://www.grocerydive.com/news/mapping-kroger-and-albertsons-store-divestiture-deal-with-cs/693186.

[143] Kroger, supra note 13.

[144] Grocery Dive Staff, The Friday Checkout: C&S Would Catapult to Major Retailer Status with Kroger-Albertsons Deal, Grocery Dive (Sept. 8, 2023), available at https://www.grocerydive.com/news/cs-wholesale-grocers-major-grocer-kroger-albertsons-merger-deal/693127.

[145] Catherine Douglas Moran, Why More Grocers Are Bringing Private Label Production In-House, Grocery Dive (June 13, 2023), available at https://www.grocerydive.com/news/grocers-private-label-production-in-house-sales-manufacturing/651986.

[146] Tkacik & Kelloway, supra note 119

The Privacy-Antitrust Curse: Insights from GDPR Application in EU Competition Law

The integrated approach that many competition and privacy regulators have endorsed for oversight of the major online platforms, whose business models rely on collecting . . .

Abstract

The integrated approach that many competition and privacy regulators have endorsed for oversight of the major online platforms, whose business models rely on collecting and processing large troves of personal data, has often been justified on grounds that competition and data protection are complementary ends. In this respect, Europe represents a testing ground for evaluating how privacy breaches may inform antitrust investigations. Indeed, the European Union’s General Data Protection Regulation (GDPR) and the recent German antitrust decision concerning Facebook may be considered polestars for this emerging regulatory approach that links market power and data power. This paper tests the degree to which such an approach is viable in concrete terms by analyzing how the European Commission and national competition authorities have applied data-protection rules and principles in antitrust proceedings. Notably, the paper aims to demonstrate the fallacy of characterizing the relationship between privacy and antitrust in terms of synergy and complementarity. Further, the paper maintains that the principles the European Court of Justice recently affirmed in its Meta decision do not appear to address the issue conclusively. The tension between these areas of law is illustrated by allegations raised in the numerous Apple ATT investigations concerning the strategic use of privacy as a business justification to pursue anticompetitive advantages. Rather than strengthening antitrust enforcement against gatekeepers and their data strategies, the inclusion of privacy harms in antitrust proceedings may turn out to be a potential curse for competition authorities, as it allows firms opportunities for regulatory gaming that can serve to undermine antitrust enforcement.

I.       Introduction

A significant share of the past decade’s academic literature on the role of data in digital markets has focused on the intersection of what had been previously thought of as the separate domains of privacy and antitrust. Given that data serves as a significant input for many of the major online platforms’ services and products, digital firms are eager to collect and process as much of it as possible. Such firms also use data-sharing agreements to obtain further data (i.e., information collected and provided by external suppliers) in order to improve their products and services. This is particularly true for those platforms whose business models rely on monetizing consumer information by selling targeted advertising and personalized sponsored content. In a market where platforms’ data-acquisition strategies are driven by the objective of granting sellers preferential access to consumer attention, personal data can represent an especially valuable portion of platforms’ information assets.[1] Moreover, given the social dimension of personal data, one user’s choice to share personal information with an online platform may generate externalities on other non-disclosing users (or non-users) by revealing information about them. Recent advances in machine learning may magnify the extent of these externalities, and raise questions about the effectiveness of data-protection regulations more generally.[2]

These dynamics have moved policymakers to take a greater interest in the degree to which data-accumulation strategies undermine individual privacy and entrench platforms’ market power. Some contend that the peculiar features of digital markets and the potential adverse uses of data in the digital economy require a regulatory approach that integrates privacy into antitrust enforcement and ensures close cooperation between antitrust authorities and data-protection regulators.[3]

According to this account, as network effects strengthen online firms’ market power, it becomes progressively more difficult to structure incentives for firms to compete on offering privacy-friendly products and services.[4] Conversely, these advocates claim, more competition in digital markets would lead to more privacy.[5]

Particular scrutiny is directed toward advertising-funded platforms that offer free services to attract users and thereby feed users’ data to the other side of the platform (i.e., advertisers), whose willingness to pay is strictly dependent on being able to deliver effective marketing through granular targeting or personalization. For their part, however, end users may not be aware of the value of their own data or may be induced to disclose private information. This could happen because users are attracted by zero-price services’ offers or, given the lack of available and comparable alternatives, in order to remain connected to their social, family, or work networks, users may feel compelled to accept take-it-or-leave-it terms that include the unwanted collection and use of their data.[6]

Some suggest that privacy should be included in antitrust assessments because suboptimal privacy offerings may be the result of anti-competitive behavior leading to decreased quality of products and services.[7] In this sense, privacy would represent a particularly significant factor to be taken into account in the merger-review process, as market concentration among companies that hold big data could further expand the merging firms’ tools to profile consumers and potentially invade their privacy.[8]

Finally, some advocates propose commingling antitrust and privacy regulation as part of a broader agenda to realign competition policy away from pure efficiency-oriented antitrust enforcement and instead toward a holistic approach that combines competition law with other fields of law, in order to take account of a broader swath of social interests.[9] In essence, privacy and antitrust would each help to cover the other’s purported Achilles heel.[10] While end users’ privacy interests would become relevant in investigating data-accumulation strategies that antitrust might otherwise fail to tackle, antitrust authorities would be more effective in ensuring data protection.[11]

Against the integrationist perspective, however, some scholars warn of risks that would attend transforming privacy infringements into per se antitrust violations.[12] Indeed, competition law and privacy regulation pursue different aims and deploy different tools. While privacy is not irrelevant to competition law and may constitute an important component of nonprice competition, the goals of competition and privacy are often at odds. Pushing these regulatory regimes to converge threatens to confuse, rather than strengthen, the enforcement of either.[13]

Further, the widely recognized “privacy paradox” illustrates that assessments of privacy are extremely subjective. Different consumers in differing contexts often express starkly different sensitivities about the protection of their personal data, rendering it challenging to provide accurate quality-driven assessments or even to set broadly acceptable baseline rules and policies.[14] More generally, an expansive approach that would treat privacy violations as sources of competitive harm potentially implies the need for antitrust investigations whenever dominant firms potentially violate any law, as they would acquire an advantage by saving costs or raising rivals’ costs.[15] Antitrust authorities would therefore become economy-wide regulators.

While some recent cases brought by U.S. antitrust authorities have also placed privacy concerns in a prominent position,[16] there are two reasons that Europe appears to represent the primary testing ground for an integrated approach for privacy and antitrust. First, European policymakers long have prided themselves as leaders in regulating digital markets, notably for a broad array of heterogeneous legislative initiatives that have in common their strenuous efforts to foster data sharing and their sponsors’ belief that the emergence of large technology platforms requires a bespoke approach.[17] In this sense, the initiative that blazed the path for the emerging integrationist perspective was the EU’s General Data Protection Regulation (GDPR), which assigned control rights over data to individuals and, in light of the emerging regulatory convergence of privacy and antitrust, introduced a general data-portability right for individuals, the rationale of which was inherently pro-competitive.[18]

Second, on the antitrust side of the ledger, the decision handed down by the German competition authority in the Facebook case was the first (and remains the primary) example of the trend toward enforcers asserting that competition law should be informed by data-protection principles and that data protection should enforced outside its usual legal context, with the goal of remedying the shortcomings of privacy law.[19]

Despite the purported synergies underpinning the respective policy goals of competition and data-protection law, however, their interests and objectives are not necessarily aligned.[20] In particular, there are signs that some major digital firms may interpret data-protection requirements in ways that risk distorting competition.[21] Namely, once privacy harms are included among the interests ostensibly protected in antitrust proceedings, platforms may have incentive to adjust their strategies to invoke data protection as a business justification for allegedly anticompetitive conduct.[22]

For example, some platforms justify their decisions to deny rivals access to their facilities on grounds that doing so would risk violating their users’ privacy.[23] App-store providers in particular have described some restrictions that may be interpreted as anticompetitive self-preferencing (e.g., requiring in-app purchases to be routed through their own in-app payment processor, limiting sideloading, and limiting app developers’ ability to communicate with end users about the availability of alternative payment options) as necessary to guarantee users’ security and privacy.[24]

The most debated example illustrating the growing tension between data protection and antitrust is Apple’s adoption of its “app tracking transparency” (ATT) policy, which creates new consent and notification requirements that change the way app developers can collect and use consumer data for mobile advertising on iOS. There very well could be privacy benefits associated with the new Apple framework, as it may enhance users’ privacy and control over their personal data. But ATT also would now differentiate between a user’s consent for Apple’s advertising services and consent for third-party advertising services. The ATT policy might therefore represent a form of discrimination that benefits Apple’s own advertising services and reinforces its position in app distribution to the detriment of rivals. For these reasons, the ATT policy is under investigation by several antitrust authorities.[25]

Given this backdrop, this paper seeks to investigate the intersection of privacy and competition law and to analyze how data-protection rules and principles have been applied in antitrust proceedings by the European Commission and by EU national competition authorities (NCAs). The analysis of the case law will illustrate how data protection has been progressively transformed from a weapon used by antitrust authorities to limit data accumulation to a shield exploited by digital platforms to justify potentially anticompetitive strategies and to game antitrust rules.

As a result, the paper aims to demonstrate the fallacy of the narrative that describes the relationship between privacy and antitrust in terms of synergy and complementarity. Such a paradigm, indeed, does not provide useful insights to solve the growing conflicts between the interests protected and the goals pursued by these different fields of law.

As has already happened with regard to the traditional intersection of intellectual-property protection and competition law, invoking a convergence of aims does not in itself sketch out a pragmatic solution. Notably, competition authorities’ cooperation with data-protection regulators may help to ensure a coherent and uniform interpretation and application of the GDPR, it will not help antitrust authorities to strike the balance between privacy benefits and anticompetitive restrictions. In such a scenario, competition law enforcers risk being forced, like Buridan’s Ass, to make a choice that cannot be made.[26]

The remainder of the paper is structured as follows. Section II examines the European cases in which privacy concerns have been addressed in antitrust proceedings to tackle data-accumulation strategies by large online platforms. Section III deals with the strategic use of privacy as a business justification for potential anticompetitive conduct, which emerges as a byproduct of promoting the integration of privacy and antitrust. Taking stock of the German Facebook case recently addressed by the Court of Justice of the European Union (CJEU),[27] Section IV illustrates how the intrinsic conflict between data-protection and competition law cannot be solved merely by invoking a purported synergy or complementarity. Section V concludes.

II.     Privacy as an Antitrust Sword Against Data-Accumulation Strategies

While data-protection and competition law serve different goals, it is commonly argued that the emergence of business models involving the collection and commercial use of personal data creates inevitable linkages between market power and data protection.[28] Notably, given that the key goal of the GDPR was to enable individuals to have control of their own personal data,[29] applying competition rules to digital markets could, it is asserted, promote precisely that control.[30] As a consequence, “previously separate policy areas become interlinked, and different regulatory authorities are increasingly required to consider a given set of issues from the perspective of contrasting policy aims and objectives.”[31]

From this perspective, combining data-protection and competition law is justified on grounds that a common aim they share is to avoid exploitation of personal data and restrictions on consumers’ privacy.[32] Since end users may experience less privacy and autonomy as a result of excessive data collection and use:

Reductions in privacy could also be a matter of abuse control, if an incumbent collects data by clearly breaching data protection law and if there is a strong interplay between the data collection and the undertaking’s market position.[33]

Indeed, from the standpoint of competition law, the idea has been advanced that the acquisition and exploitation of user information is itself the result of, or evidence of, market failure.[34] In particular, users of dominant advertiser-based platforms are said to suffer both from significant information asymmetries as a result of opaque data policies, and from platform lock-in, with no choice other than to consent to the harvesting and use of their data because of the lack of viable alternatives.[35]

On the data-protection side of the ledger, it is bears noting that, according to the GDPR, consent means any “freely given, specific, informed and unambiguous” indication of a data subject’s wishes—whether by statement or some other clear affirmative action—that signifies agreement to the processing of his or her personal data.[36] Further, the GDPR specifies the conditions for consent, which include that: the request for consent be presented in a manner clearly distinguishable from other matters; that it be in an intelligible and easily accessible form; that it use clear and plain language; that the data subject has the right to withdraw consent at any time; and that, when assessing whether consent is freely given, utmost account shall be taken of whether, inter alia, the performance of a contract—including the provision of a service—is conditional on consent to processing personal data not actually needed for the performance of that contract.[37]

A. Privacy Harm as an Antitrust Abuse

As the French and German competition authorities have argued in a joint paper:

[L]ooking at excessive trading conditions, especially terms and conditions which are imposed on consumers in order to use a service or product, data privacy regulations might be a useful benchmark to assess an exploitative conduct, especially in a context where most consumers do not read the conditions and terms of services and privacy policies of the various providers of the services that they use.[38]

From this perspective, privacy concerns support the use of antitrust intervention to limit data-accumulation strategies by treating the restriction on privacy as a form of exploitative abuse.

Another way that privacy interests can be leveraged by antitrust authorities to address competitive concerns about data accumulation is through the merger-review process. Indeed, “firms that gain a powerful position through a merger may be able to gain further market power through the collection of more consumer data and privacy degradation.”[39] The use of merger review is expected to be more effective to achieve privacy-policy goals given that, while an antitrust abuse investigation may at best neutralize or alleviate exploitation of data gathered by a dominant player, merger proceedings would prevent data accumulation in the first place.

  1. The German Facebook case: Users’ privacy-exploitation claim

The Bundeskartellamt’s decision in Facebook undoubtedly represents the apex, to date, of enforcers’ application of the integrationist perspective.[40] According to the German competition authority, Facebook unlawfully exploited its dominant position in the German market for social networks by making the use of its social-networking service conditional on users granting extensive permission to collect and process their personal data. Notably, Facebook failed to make its users fully aware of the fact that it collected their personal data from sources other than the Facebook platform and then merged those data with personal information gathered through its own platform.[41] Further, Facebook put its users in the difficult position of either accepting this data policy or refraining from use of the social network in its entirety.

Indeed, even well-informed users would have not been able to voluntarily consent to such data collection and combination, as they would fear the alternative of no longer being able to access the social network.[42] Therefore, according to the German competition authority, when the data controller is in a dominant position, its users’ consent is insufficient under the GDPR, because the platform’s market power always puts users in the position of having to either take or leave any offers made.

Considering these findings, the Bundeskartellamt established a link between market power and privacy concerns. In its view, Facebook’s terms and conditions were neither justified under data-protection principles nor appropriate under competition-law standards. To comply with the GDPR, users should have been asked whether they voluntarily consent to the practice of combining data in their Facebook user accounts, which could not consist merely of ticking a box. Indeed, given Facebook’s superior market power, the user’s choice to either accept comprehensive data combination or to refrain from using the social network could not be regarded as voluntary consent.[43] The Bundeskartellamt therefore concluded that Facebook had infringed GDPR rules by depriving its users of the human right to control the processing of their personal data and of the constitutional right of informational self-determination.

This form of coercion is, however, also relevant to competition law, as it was the result of Facebook’s dominant position. Hence, Facebook’s conduct could be considered exploitative within the meaning of the general clause of Section 19(1) of the German Competition Act (GWB), according to which competition law applies in every case where one bargaining party is so powerful that it can dictate the terms of the contract, with the end result being the abolition of the contractual autonomy of the other bargaining party. From the Bundeskartellamt’s standpoint, if a dominant firm collects and analyzes users’ data pursuant to terms and conditions that do not comply with EU data-protection rules, it also violates antitrust law by acquiring an unfair competitive advantage over firms that do adhere to the GDPR.

In summary, while the primary concern in the Facebook case was an antitrust issue (i.e., the excessive quantity of data that Facebook accumulated in its unique dataset),[44] the Bundeskartellamt elaborated a theory of harm based primarily on protecting the constitutional right to informational self-determination. In other words, the competition authority invoked the right under which data-protection law affords individuals the power to decide freely and without coercion how their personal data is processed. Such reasoning is consistent with the case law of Section 19(1) GWB, which allows an antitrust authority to consider the protection of constitutional values and interests in assessing the practices of dominant firms. While the Bundeskartellamt contended that its proceedings against Facebook would also generally be possible under the EU’s antitrust provision on exploitative abuses (Article 102(a) TFEU),[45] Section 19 GWB offered a broader (and, hence, more legally convenient) general clause.[46]

This privacy-focused approach also manifested in the remedy that Meta presented, and which the Bundeskartellamt welcomed. To implement the German antitrust authority’s decision, Meta proposed several changes to the accounts center that would allow customers to decide whether they wanted to use all services separately, each with their own circumscribed functions, or to use additional functions across accounts, which would require sharing more personal data.[47] In the Bundeskartellamt’s view, this solution would allow Meta’s customers to make a largely free and informed decision.

The Bundeskartellamt’s approach in the Facebook case therefore appears quite distinctive and essentially German-specific, as well as particularly controversial with respect to the scope and boundaries of competition and data-protection enforcement.[48] Indeed, in ascertaining a privacy violation previously undetected by any data-protection authority, the Bundeskartellamt acted as a self-appointed enforcer of data-protection rules.

It also interpreted data-protection rules in ways that far exceed the limits of its legal competence, given that there is nothing in the GDPR that makes the quality of a user’s consent agreement contingent on the data controller’s market power. Indeed, the GDPR makes no distinction at all on the basis of a firm’s market power. Size does not matter when it comes to data-protection law; a dominant firm is just as bound by privacy rules as its smaller rivals. At the same time, from the perspective of competition law, following the Bundeskartellamt’s expansive stance, virtually every legal infringement by a dominant firm could amount to an antitrust violation.

Because of the thorny implications for the interface between antitrust and data-protection law, the Facebook decision unsurprisingly sparked a heated debate not only in the literature, but also between German courts.

The Higher Regional Court (Oberlandesgericht, or OLG) of Du?sseldorf suspended the landmark decision, expressing serious doubts about its legal basis and complaining that the Bundeskartellamt was “merely discussing a data protection issue, and not a competition problem.”[49] Pursuant to both European and German antitrust provisions, a charge of abuse of market power by a dominant undertaking requires a finding of anticompetitive conduct and, hence, damage to competition—namely, to the freedom of competition, that is “safeguarding competition and the openness of market access.”[50] Therefore, dominant undertakings carry a special responsibility only in the domain of competition, rather than for compliance with the entire legal system by avoiding any violation of the law.[51] Further, in the appellate court’s view, no influence was exerted on users, as Facebook’s terms of service simply require them to weigh the benefits of using an ad-financed (and, therefore, free) social network against the consequences of Facebook’s use of the additional data that it gathers.

However, the Federal Supreme Court (Bundesgerichtshof, or BGH) overturned the OLG’s judgment and held that Facebook must comply with the Bundeskartellamt’s decision.[52] The BGH’s reasoning did, however, differ from the Bundeskartellamt’s. According to the Federal Supreme Court,  it is inconclusive whether Facebook’s processing and use of personal data complied with the GDPR. The court’s decision turned instead on Facebook’s terms of service, which the BGH found are abusive if they deprive Facebook users of any choice in whether they wish to use the network in a more personalized manner (thus, linking their experience to Facebook’s potentially unlimited access to characteristics that include their off-Facebook use of the internet more generally) or whether they wanted a level of personalization that was based solely on data that they themselves share on Facebook.[53]

Notably, the BGH found that Facebook’s data processing constitutes an “imposed extension of services,” as users receive an indispensable service only in combination with another undesired service.[54] Accordingly, such a practice was evaluated as both an exploitative and an exclusionary abuse. The lack of options available to users affects their personal autonomy and the exercise of their right to informational self-determination, as protected by the GDPR. Given lock-in effects that serve as barriers for network users who would otherwise like to switch providers, the BGH found that this lack of options exploits users in a manner relevant under competition law since, under effective competition, one would expect more diverse market offerings for social networks.[55] Further, the terms of service could also impede competition for online advertising, allowing Facebook to protect its dominant position against rivals, as they would be able to improve their offerings due to privileged access to a considerably larger database.[56]

As a result of this clash among the German courts, the Higher Regional Court of Du?sseldorf decided to refer the case to the CJEU, adding a new twist to the Facebook saga.[57] In particular, the OLG of Du?sseldorf raised seven questions about the interpretation of the GDPR, fundamentally asking the CJEU to untie the knot and clarify the competence of a competition authority to determine and penalize a GDPR breach; the prohibition on processing sensitive personal data and the conditions applicable to consenting to their use; the lawfulness of processing personal data in light of certain justification; and the validity of a user’s consent to processing personal data given to an undertaking in a dominant position.[58]

It is also worth noting the different approaches taken by other authorities concerning the very same Facebook conduct. Notably, the Italian competition authority evaluated such practices as violations of the Consumer Code (instead of the competition law),[59] while in Belgium, the Court of First Instance of Brussels found a violation of privacy rules.[60]

  1. The Digital Markets Act: Rivals’ exclusion and primacy of data-protection interests over competition-policy goals

The Facebook case has already influenced the broader debate about the limits of competition law to address certain features of digital markets effectively. The EU’s Digital Markets Act (DMA)—which was explicitly grounded in the assumption that competition law alone is unfit to tackle certain challenges and systemic problems posed by the platform economy—specifically prohibits combining personal data across a gatekeeper’s services, a provision clearly inspired by the German investigation.[61]

Notably, pursuant to Article 5(2) DMA, a gatekeeper shall not: (a) process—for the purpose of providing online-advertising services—end users’ personal data using third-party services that themselves make use of the gatekeeper’s core platform services; (b) combine personal data from the relevant core platform service with personal data from any further core platform services, or from any other services provided by the gatekeeper, or with personal data from third-party services; (c) cross-use personal data from the relevant core platform service in other services provided separately by the gatekeeper, including other core platform services, and vice versa; and (d) sign end users into the gatekeeper’s other services in order to combine personal data, “unless the end user has been presented with the specific choice and has given consent” within the meaning of the GDPR.

Further, according to Recital 36—given that gatekeepers process personal data from a significantly larger number of third parties than other undertakings—data processing for the purpose of providing online-advertising services gives gatekeeper platforms potential “advantages in terms of accumulation of data,” thereby “raising barriers to entry.” To ensure that gatekeepers do not unfairly undermine the “contestability” of core platform services, gatekeepers should enable end users to “freely choose to opt-in” to such data processing and sign-in practices. This may be accomplished by offering a less-personalized but equivalent alternative, and without making the use of (or certain functions of) the core platform service conditional on the end user’s consent.[62]

Moreover, in light of Recital 37, when a gatekeeper does request consent, it should proactively present a “user-friendly solution” to the end user to provide, modify, or withdraw consent in an explicit, clear, and straightforward manner. In particular, consent should be given by a clear affirmative action or statement establishing a freely given, specific, informed and unambiguous indication of agreement by the end user, as defined in the GDPR.

Lastly, it should be as easy to withdraw consent as to give it. Gatekeepers should not design, organize, or operate their online interfaces in a way that deceives, manipulates, or otherwise materially distorts or impairs end users’ ability to freely give or withdraw consent.[63] In particular, gatekeepers should not be allowed to prompt end users more than once a year to give consent for a data-processing purpose for which the user either did not initially give consent or actively withdrew consent.

The idea that only opt-in mechanisms can produce effective consent within the meaning of the GDPR is confirmed by the obligation under Article 6(10) DMA, which imposes on gatekeepers the duty to provide business users, or third parties authorized by a business user, access to aggregated and non-aggregated data (including personal data) generated in the context of using the relevant core platform services.[64]

The provision under Article 5(2) DMA provides interesting insights into the relationship between data-protection and competition law. By emphasizing that the primary concern is online gatekeepers’ data-accumulation strategies, the DMA’s approach differs from the one the Bundeskartellamt pursued in Facebook. Rather than focusing on potential harms to users’ self-determination and digital identity, the DMA points to a pure antitrust harm related to market contestability. Therefore, even if “[t]he data protection and privacy interests of end users are relevant to any assessment of potential negative effects of the observed practice of gatekeepers to collect and accumulate large amounts of data from end users,”[65] the primary interest protected is a competitive one—namely to avoid foreclosure against rivals.

From this perspective, it may be argued that the DMA adopts an integrated approach that takes data-protection principles into account within a competitive assessment of gatekeepers’ conduct. The very last part of the provision, however, demonstrates the opposite. By subordinating the prohibitions to respect the GDPR, European authorities arguably acknowledge the potential tensions between data-protection interests and competition-policy goals. Moreover, in the event of such a conflict, the DMA affirms the primacy of the former. Indeed, all the forms of conduct listed in Article 5(2) are forbidden “unless” the end user has been presented with a specific choice and given consent within the meaning of the GDPR.

  1. New German platform-specific antitrust rules and the Google case

There is another interesting and ongoing German investigation regarding Google’s data-processing terms. Notably, in January 2023, the Bundeskartellamt issued a statement of objections against Google claiming that, under the company’s current terms, users are not given “sufficient choice” as to how their data are processed across services.[66]

The antitrust authority noted that Google’s business model relies heavily on processing user data and that its current terms allow the company to combine various data from various services and use them, for example, to create very detailed user profiles that the company can exploit for advertising and other purposes, or to train functions provided by Google services. Google may, for various purposes, collect and process data across services, which include both its own widely used services (Google Search, YouTube, Google Play, Google Maps, and Google Assistant), as well as numerous third-party websites and apps. Bundeskartellamt President Andreas Mundt stated that this grants Google a “strategic advantage” over other companies.[67]

According to the Bundeskartellamt’s preliminary assessment, the choices offered to users are too general and insufficiently transparent. The authority contends that sufficient choice would require that users be able to limit data processing to the specific service used. In addition, they also must be able to differentiate between the purposes for which the data are processed. Moreover, the choices must not be devised in a way that would make consenting to data processing across services easier than not consenting to it.

The framing of the Google investigation is similar to that of the Facebook case. The antitrust authority is fundamentally concerned with a data-accumulation strategy that it contends confers to Google a critical competitive advantage. And given that having access to more user data than rivals have cannot in itself be considered anticompetitive, privacy concerns are exploited to limit such a strategy.

There is, however, a significant difference worth highlighting. In the Google case, the Bundeskartellamt’s position benefits from a new provision of Section 19a GWB,[68] which empowers national competition authorities to tackle platform-specific practices that are similar and functionally equivalent to those prohibited under the DMA.[69] Notably, since January 2021, the Bundeskartellamt has had the power to designate undertakings of “paramount significance for competition across markets.” The factors relevant to this designation include a platform’s dominant position in one or more markets; financial strength or access to other resources; vertical integration and activities in otherwise related markets; access to data relevant for competition; and the importance of the activities for third parties’ access to supply and sales markets and related influence on third parties’ business activities. Google has been the first platform to be designated as of paramount significance for competition across markets.[70]

Once the designation is completed, the Bundeskartellamt can prohibit such undertakings from engaging in anticompetitive practices. In particular, the new provision introduces a list of seven types of abusive practices that are prohibited, unless the undertaking is able to demonstrate that the conduct at issue is objectively justified. While the targeted practices are similar to those captured by the DMA, the main differences are that the German list is considered exhaustive and the practices at issue are not prohibited per se. Instead, it introduces a reversal of the burden of proof, allowing firms to provide objective justifications for their conduct, which is not allowed under the DMA.

For the sake of this analysis, pursuant to paragraph 4 of Section 19a GWB, the Bundeskartellamt may prohibit an undertaking of paramount significance for competition across markets from creating or appreciably raising barriers to market entry (or otherwise impeding other undertakings) by processing data relevant for competition that have been collected by the undertaking, or demanding terms and conditions that permit such processing—in particular, making the use of its services conditional on a user agreeing to data processing by the undertaking’s other services or by a third-party provider without “sufficient choice” as to whether, how, and for what purpose such data are processed.

As mentioned, while the Google investigation resembles the background of the Facebook decision, the introduction of Section 19a(4) GWB has relevant implications. The new provision is clearly inspired by the strategy investigated in Facebook and, as already enshrined in the DMA, essentially aims to ease enforcement, avoiding the hurdles and burdens of standard antitrust analysis. Practically speaking, the Bundeskartellamt therefore does not need to struggle to find a proper theory of harm and can easily avoid the odyssey it experienced in Facebook. Moreover, the new provision’s wording changes the legal landscape, distinguishing the Google investigation from both the parallel DMA provision and the Facebook decision. Indeed, by relying on the lack of “sufficient choice” for users, Section 19a(4) GWB does not include any reference to the GDPR, thus allowing the Bundeskartellamt to provide an autonomous interpretation. With regard to the comparison with Facebook, on the other hand, Section 19a(4) GWB—just like the DMA—aims to promote contestability in the market (“creating or appreciably raising barriers to market entry”). Hence, data accumulation is prohibited to the extent that it excludes rivals, rather than whether it exploits users’ privacy.

That the German provision is effective has been confirmed by Google’s decision to end the proceeding by submitting commitments.[71] Under those commitments, Google will give its users the option to grant free, specific, informed, and unambiguous consent to have their data processed across services.[72] Google will also offer corresponding choice options for particular combinations of data and services, and will design selection dialogues to avoid dark patterns, thus not guiding users manipulatively towards cross-service data processing.

It is worth noting that Google’s commitments involve more than 25 services, with only those services that the European Commission has since designated as core platform services under the DMA (i.e., Google Shopping, Google Play, Google Maps, Google Search, YouTube, Google Android, Google Chrome and Google’s online-advertising services) excluded from the list. While this was intended to avoid practical conflicts with application of the DMA, it also represents an acknowledgment that the DMA and German antitrust law pursue the very same goals. Indeed, as stated in the decision, Google’s commitments “are intended to correspond in substance to an extension of Google’s obligations under Article 5(2) DMA” to further services and, therefore, “in case of doubt, the terms used in the Commitments are to be interpreted in accordance with their meaning in the DMA.”[73]

B. Privacy Harm in Merger Analysis: The European Commission’s Case Law

Given this broad consensus regarding synergies between data-protection and competition law in digital markets, it is somewhat surprising how reluctant the European Commission has been to implement this integrated approach in the context of merger analysis.[74] Indeed, while acknowledging privacy’s role as a parameter of competition between online platforms, the Commission has to date not blocked any merger on the grounds of protecting individuals’ control over personal data, and it has nearly always approved unconditionally those mergers that raised privacy concerns.

Notably, in the days before the GDPR, the Commission authorized the Google/DoubleClick merger, in the process affirming that antitrust and data-protection rules had wholly separate scopes.[75] While it could have determined that the combined data-collection activities of two players active in the online-advertising industry raised concentration concerns and a possible unfair advantage in producing targeted advertising, the Commission’s assessment, under pure antitrust criteria, was that it was unlikely that the new entity would obtain a competitive advantage unmatchable by its rivals.[76] Further, the Commission underlined that its decision exclusively concerned an appraisal of the operation under competition rules, without prejudice to other obligations imposed on the parties by data-protection and privacy laws.[77]

This stance of maintaining separate regulatory spheres of inquiry was even more clearcut in the 2014 Facebook/WhatsApp merger.[78] Assessing the potential edge the combined entity might derive from controlling huge amounts of data, the Commission found that, regardless whether the merged entity would start using WhatsApp user data to improve targeted advertising on Facebook, there continued to be large troves of valuable internet user data that were not within Facebook’s exclusive control.[79] More importantly, the Commission stated that:

Any privacy-related concerns flowing from the increased concentration of data within the control of Facebook as a result of the Transaction do not fall within the scope of the EU competition law rules but within the scope of the EU data protection rules.[80]

The outcome and reasoning were the same in Microsoft/LinkedIn.[81] Consistent with the findings in Facebook/WhatsApp, the results of the Commission’s market investigation revealed that privacy is an important parameter of competition and a driver of customer choice.[82] But not only did the transaction not raise serious antitrust concerns in online advertising, given that combining the firms’ respective datasets did not appear to result in raising rivals’ barriers to entry or expansion,[83] but also:

[S]uch data combination could only be implemented by the merged entity to the extent it is allowed by applicable data protection rules. … Microsoft and LinkedIn are subject to relevant national data protection rules with respect to the collection, processing, storage and usage of personal data, which, subject to certain exceptions, limit their ability to process the dataset they maintain.[84]

Moreover, the Commission noted that the GDPR “may further limit Microsoft’s ability to have access to, and process, its users’ personal data in the future since the new rules will strengthen the existing rights and empower individuals with more control over their personal data.”[85]

In a nutshell, the Commission again chose to defer to privacy rules for protecting individuals’ personal data and analyzed the transaction’s antitrust issues while “[a]ssuming such data combination [was] allowed under the applicable data protection legislation.”[86] The Commission did not discuss whether the relevant markets under consideration were sufficiently competitive to provide users with the optimal level of privacy-friendly options. It didn’t establish any link between the merging firms’ market power and the variety of privacy-friendly tools and services they provided. Nor did it find any connection between such market power and the optimal quantity of personal data that the firms under scrutiny should have collected.

In Apple/Shazam, despite some concern that the acquisition would grant Apple access to commercially sensitive information about competitors of its Apple Music service, the Commission regarded it as unclear whether the merged entity would be able to put competing providers of digital-music streaming apps at a competitive disadvantage. And they again stressed that personal-data processing remained subject to the GDPR.[87]

The recent Google/Fitbit merger offered the Commission another opportunity to interrogate overlaps among data protection and antitrust. Ultimately, the Commission’s analysis focused on the data collected via Fitbit’s wearable devices and the interoperability of wearable devices with Google’s Android operating system for smartphones.[88] While some market participants complained that, in combining those databases, Google could obtain a competitive advantage in the digital health-care sector that would leave competitors unable to compete, others (including the European Data Protection Board) raised privacy concerns on grounds that the merger would make it increasingly difficult for users to track the purposes for which their health data would be used.[89]

To address such issues, Google offered (and the Commission accepted) commitments to maintain a technical separation of Fitbit user data by storing them in a data silo separate from any Google data used for advertising; that it will not use the health and wellness data collected from users’ wrist-worn wearable devices and other Fitbit devices for Google Ads; and it will ensure that users have an effective choice to grant or deny the use of health and wellness data stored in their Google Account or Fitbit Account by other Google services.

With regard to privacy concerns, the Commission reminded those involved that the parties are held accountable to implement appropriate technical and organizational measures to ensure that data processing is performed in accordance with the GDPR.[90] More specifically, the Commission noted that the GDPR is designed to enhance transparency over data processing, accountability by data controllers and, ultimately, users’ control over their data.[91] The Commission found no evidence that privacy was an important parameter of competition in wearables and underlined that any privacy or data-protection decision or initiative the parties might adopt would have to comply with the data-protection rules set out by the GDPR.[92]

The Commission addressed similar privacy issues arising from the combination of datasets in Microsoft/Nuance[93] and Meta/Kustomer,[94] each time noting that GDPR served as the appropriate safeguard.

Moreover, the Commission appears to retain this “separatist” stance, as confirmed recently by its unconditional approval of a joint venture among Deutsche Telekom, Orange, Telefo?nica, and Vodafone, which will offer a platform to support brands and publishers’ digital-marketing and advertising activities in France, Germany, Italy, Spain, and the United Kingdom.[95] Subject to a user’s consent (i.e., on an opt-in basis only), the joint venture will generate a unique digital code derived from the user’s mobile or fixed-network subscription that will allow brands and publishers to recognize users on their websites or applications on a pseudonymous basis, group them under various categories, and tailor their content to specific user groups.

Whatever privacy and security benefits or harms might arise from the operation, the Commission was ultimately guided in its decision by the lack of competition concerns. Moreover, the Commission declared that it has been in contact with data-protection authorities during its investigation and that data-protection rules are fully applicable, irrespective of the merger’s clearance.

III.   Privacy as a Shield Against Antitrust Allegations

Amid these limited and somewhat confused attempts to address privacy concerns in digital markets by integrating data-protection rules and competition-law enforcement, a novel and challenging phenomenon has emerged. Taking stock of some authorities’ willingness to grant primacy to data protection in the context of antitrust interventions, some platforms have implemented changes to their ecosystems with the declared aim of ensuring increased privacy to end users. For instance, Apple and Google have developed policies to restrict third parties from sharing user data through apps in the platforms’ respective operating systems and websites in their respective browsers.[96] These policies include Apple’s ATT, Intelligent Tracking Prevention, and iCloud Private Relay, and Google’s Android Privacy Sandbox and Chrome Privacy Sandbox. To a certain extent, the DMA may have even encouraged some of these design choices by apparently endorsing the view that only opt-in systems can ensure effective consent within the meaning of the GDPR.

The suspicion is that such facially noble intentions may actually conceal a goal of achieving anticompetitive advantages at the expense of rivals and business users. Therefore, it appears that a new form of regulatory gaming is on the horizon. Particularly in online-advertising markets, privacy may be weaponized as a business justification for potentially anticompetitive conduct and data-protection requirements may be leveraged to distort competition. The relevance and dangerousness of such hypotheses are confirmed by certain antitrust investigations launched recent years, which the following paragraphs will analyze.

A. Apple’s ATT Policy

As illustrated above, data represents a primary input for platforms whose business models rely on monetizing consumer information by selling targeted advertising and personalized sponsored content. In digital markets, advertisers benefit from access to detailed (and hence, highly valuable) user data, such as browsing behavior, profiles on company websites, demographic information, shopping habits, and past purchase history, especially given the potential to use that data across advertising platforms.[97] Therefore, the effectiveness of targeted advertising and the overall profitability of advertising-based business models rely on data tracking.

To enhance users’ privacy protection, however, regulatory interventions like the GDPR aim to reduce data collection and mitigate platforms’ tracking by requiring explicit consent for users’ individual-behavior data to be used for targeted advertising.[98] In addition, some platforms have adopted (or announced) privacy-centric policies that would limit third parties’ ability to track data, thus affecting the profitability and revenues of their advertising strategies.[99]

Apple’s ATT policy is a paramount example of such product changes. With the iOS 14.5 privacy update, Apple introduced an opt-in mechanism that imposes more restrictive rules on competing app developers than those the company applies to itself. The differential treatment mostly concerns features that prompt users to grant apps permission to track them. Without consumers opting into this prompt, developers cannot access their identifiers for advertisers (IDFA), which are used to monitor users’ activity across apps.

The wording of the prompts ATT offers for user consent may unduly influence users to withhold consent from third-party apps. For apps developed by Apple itself, the consent prompt focuses on the positive aspects of personalized services, rather than the tracking of users’ browsing activity. In contrast, the prompt for third-party app developers places greater emphasis on other companies’ app and website tracking activities (without explaining the term “track”) and does not provide information about the benefits that users could derive from personalized advertising. Moreover, even if the user gives consent to be tracked, third-party app developers remain unable to share the same data that would allow for the personalization of ads, and measure their effectiveness, on another app. Indeed, for third-party app developers, the ATT framework introduces a double opt-in, requiring the user to consent to being tracked for each access to different apps, even if these apps are linked.

This model illustrates an apparent tension between data-protection interests and antitrust goals. While the ATT policy has been framed as a privacy-protecting measure, it is not just the level of privacy chosen by Apple in its digital ecosystem that is at issue, but also the competitive implications that arise from the choice to adopt discriminatory privacy policies. Indeed, the differentiated treatment imposed on third-party app developers appears likely to reduce their advertising revenues, and hence their level of competitiveness vis-à-vis Apple, and could eventually enhance the dominance of the iOS ecosystem.

Notably, the ATT framework may hinder competitors’ ability to sell advertising space, in ways that redound to Apple’s own advantage—in particular, benefiting the company’s own direct sales and advertising-intermediation platforms. Further, limiting third parties’ ability to profile users may reduce business-model differentiation. The advertising-based monetization model used by free and freemium apps may be rendered less sustainable, causing these apps to exit the market or gradually shift to the fee-supported model. This would come at the expense of end consumers, for whom the possibility of choosing free or lower-priced apps could be reduced.[100]

For these reasons, the ATT framework is currently under scrutiny by antitrust authorities in France,[101] Germany,[102] Italy,[103] and Poland,[104] who suspect that Apple is masking an anticompetitive strategy under the guise of privacy protection. Similar doubts have been raised by the UK Competition and Markets Authority in its market study on mobile ecosystems.[105]

Given these kinds of market responses, it is difficult to see how an integrated approach to data-protection and competition law could be implemented in practice. Contrasting the Italian and French investigations may provide useful insights into this conundrum. The Italian competition authority correctly stated that the case does not implicate the level of privacy chosen by Apple, but rather its decision to adopt a differentiated policy at the expense of its rivals.[106] Conversely, in evaluating whether to issue an interim measure against Apple, France’s Autorité de la Concurrence solicited input from the domestic data-protection regulator (the Commission Nationale de L’Informatique et des Liberte?s, or CNIL), which de facto prevented the competition authority from ordering interim measures. Indeed, in the CNIL’s view, the changes proposed by Apple could be of genuine benefit to both users and app publishers.[107] In particular, the ATT prompt would give users more control over their personal data by allowing them to make choices in a simple and informed manner,[108] and would allow app publishers to collect informed consent as required by the applicable regulation.

It is worth noting, however, that while all the other competition authorities are investigating Apple’s policy as a potential form of discriminatory self-preferencing, the French authority has initially evaluated whether the introduction of the ATT prompt would result in imposing unfair trading conditions or a supplementary obligation, in breach of Article 102(a) and (d) TFEU. The complaint’s investigation on the merits of the case will allow the French authority to assess whether ATT does or does not result in a form of discrimination.

B. Google’s Privacy Sandbox

Concerns regarding the potential impact of privacy policies on digital-advertising competition and publishers’ ability to generate revenue have also been against Google’s proposals to remove third-party cookies and other functionalities from its Chrome browser. In particular, Google’s Privacy Sandbox project would disable third-party cookies on the Chrome browser and Chromium browser engine, with the stated goal of better protecting consumer privacy. The project would replace those cookies with a new set of tools for targeting advertising and other functionalities. Therefore, similar to Apple’s ATT policy, Google’s planned privacy changes raise concerns about anticompetitive discrimination against rivals.

Indeed, in 2021, the European Commission initiated antitrust proceedings to investigate the effects of Google’s privacy policies on online display advertising and online display advertising-intermediation markets. The inquiry focused on whether Google had violated EU competition rules by favoring—through a broad range of practices—its own online display advertising-technology services in the ad tech supply chain, to the detriment of competing providers of advertising-technology services, advertisers, and online publishers.[109] Notably, the Commission also examined restrictions on third parties’ ability to access data about user identity or user behavior, which remained available to Google’s own advertising-intermediation services, as well as Google’s announced plans to cease making advertising identifiers available to third parties on Android mobile devices whenever a user opts out of personalized advertising.

The Commission declared that it would “take into account the need to protect user privacy, in accordance with EU laws in this respect,” underscoring that “[c]ompetition law and data protection laws must work hand in hand to ensure that display advertising markets operate on a level playing field in which all market participants protect user privacy in the same manner.”[110]

A similar investigation was launched that same year by the UK Competition and Markets Authority (CMA).[111] The CMA subsequently accepted commitments from Google designed to ensure consistent use of data by both third parties and Google’s own digital-advertising businesses through the use of safeguards to support privacy without self-preferencing.[112] In considering how best to address legitimate privacy concerns without distorting competition, the CMA highlighted the relevance of the close partnership with the UK Information Commissioner’s Office (ICO), the public body tasked with the enforcement of the Data Protection Act 2018, which is the UK’s implementation of the GDPR.[113]

IV.   The Failure of the Integrated Approach

The call for integrating privacy into antitrust enforcement reflects the policy goal of curbing ever-increasing personal-data collection and processing by a few large online platforms, who monetize such data by selling targeted advertising. Toward this aim, competition and data-protection laws are described as synergistic, as the economic features of digital markets generate connections between market power and data power. Against this background, rather than relying on the GDPR, scholars and policymakers ask competition law to step in to address the perceived problem of data-protection authorities lacking capacity to address privacy concerns effectively, as well as the extreme difficulty of forbidding data accumulation under antitrust provisions. Therefore, rather than reflecting a natural connection, data-protection and competition laws are fundamentally obtorto collo complementary, as each are considered weak in isolation.

Four primary theories of harm have been advanced to bring antitrust and privacy issues together.[114]

According to the first theory, there is a close relationship between (the lack of) competition in digital markets and privacy violations. In a competitive market, this theory asserts, firms would compete to offer privacy-friendly products and services, but the economic features of digital markets strengthen gatekeepers’ power, regardless of their willingness to deliver privacy-enhancing solutions.[115]

The second theory centers on risks arising from potential “databases of intentions” and primarily invokes the role of merger control.[116] Under this view, mergers among companies that hold significant data assets require more stringent scrutiny, as such mergers would grant the new entity tools to better profile individuals and invade their privacy.

A further attempt to justify commingling antitrust and privacy relies on assessing the quality of products and services as privacy-friendly.[117] As consumer welfare is not solely dependent on prices and output, products and services viewed as not privacy-friendly or that intrude into users’ privacy may be considered low-quality and therefore harm consumer welfare.

Finally, it has been argued that privacy policies could be applied by antitrust enforcers when they are implemented by dominant players that rely on data as a primary input of their products and services—e.g., by forcing individuals to accept take-it-or-leave-it terms involving the unwanted collection and use of their data.[118]

This overview of EU antitrust proceedings, however, demonstrates that none of these four theories of harm has been successful and that the much-invoked integrated approach is more proclaimed than adopted in practice. Indeed, neither other NCAs nor the European Commission have ever shared the Bundeskartellamt’s stance of considering a GDPR violation as a benchmark for finding a dominant firm’s practice to be abusive. Further, in the context of merger analysis, the Commission has systematically stated that any privacy-related concerns resulting from data collection and processing are within the scope of the GDPR enforcement.

Even in Germany, the Bundeskartellamt’s approach has been sufficiently controversial to spark a clash among courts and a request for clarification from the CJEU. The recent update of the GWB seems to confirm the limits of such an approach, as the new Section 19a provides an antitrust authority with a convenient shortcut to target Facebook-like data-accumulation strategies on grounds of market contestability—namely, prohibiting rivals’ foreclosure rather than users’ privacy exploitation.

In addition, these EU antitrust proceedings demonstrate that twisting competition-law enforcement may be counterproductive. Indeed, the growing phenomenon of digital platforms adopting privacy policies as justification for potentially anticompetitive conduct does not fit the narrative of the complementarity of antitrust and privacy.[119] Emerging as a byproduct of the Facebook investigation, the Apple ATT case illustrates the intrinsic tension between these areas of law, highlighting the urgency of determining how to strike a balance between conflicting interests. From this perspective, the Facebook and Apple ATT cases are two faces of the same coin. Each results from the strategic use of privacy in antitrust proceedings by both competition authorities and digital platforms, respectively.

Moreover, the French episode of Apple ATT shows that proposing cooperation between authorities is just rhetoric unfit to resolve these tensions. It is regularly affirmed that any tension between competition and data protection law “can be reconciled through careful consideration of the issues on a case-by-case basis, with consistent and appropriate application of competition and data protection law, and through continued close cooperation” between the authorities.[120] Nonetheless, in the French Apple ATT case, the data-protection regulator’s intervention actually jeopardized the antitrust investigation, demonstrating how the different goals pursued under antitrust and privacy provisions may be irreconcilable in practice.

Finally, the EU’s solution to alleged failures by antitrust and privacy regulators in addressing data accumulation in digital markets has ultimately been crafted outside the traditional competition-law framework and according to a regulation that resolves any potential conflict between competition and data-protection policy goals once and for all. Even the DMA, however, does not fully square with any of the aforementioned theories of harm, as it introduces a pure privacy exception.[121] Indeed, tackling data collection and processing by digital gatekeepers, Article 5(2) DMA prohibits personal-data accumulation strategies unless they are compliant with the GDPR—namely, unless users have been presented with the specific choice and given consent according to data-protection rules. Therefore, rather than providing criteria to evaluate case by case how to strike a balance among the interests involved, the DMA establishes competition-policy deference to privacy, finding that, where personal-data collection and processing by large online platforms are involved, privacy is the greater good.

A. The CJEU’s Judgment in Meta

Given this background, the CJEU’s July 2023 judgment in Meta was much-awaited, representing the season finale of the German Facebook saga.[122]

The decision is in line with the opinion delivered by the Advocate General (AG) Athanasios Rantos.[123] As Rantos had argued, “conduct relating to data processing may breach competition rules even if it complies with the GDPR; conversely, unlawful conduct under the GDPR does not automatically mean that it breaches competition rules.”[124] Therefore, the lawfulness of conduct under antitrust provisions “is not apparent from its compliance or lack of compliance with the GDPR or other legal rules.”[125] Further, according to well-settled CJEU principles, the antitrust assessment requires demonstrating that a dominant undertaking used means other than those within the scope of competition on the merits and, toward this aim, the court must take account of the circumstances of the case, including the relevant legal and economic context.[126] “In that respect, the compliance or non-compliance of that conduct with the provisions of the GDPR, not taken in isolation but considering all the circumstances of the case, may be a vital clue as to whether that conduct entails resorting to methods prevailing under merit-based competition.”[127] Indeed, “access to personal data and the fact that it is possible to process such data have become a significant parameter of competition between undertakings in the digital economy. Therefore, excluding the rules on the protection of personal data from the legal framework to be taken into consideration by the competition authorities when examining an abuse of a dominant position would disregard the reality of this economic development and would be liable to undermine the effectiveness of competition law.”[128]

It follows that. “in the context of the examination of an abuse of a dominant position by an undertaking on a particular market, it may be necessary for the competition authority of the Member State concerned also to examine whether that undertaking’s conduct complies with rules other than those relating to competition law, such as the rules on the protection of personal data laid down by the GDPR.”[129]

Rantos more explicitly distinguished the hypothesis under which an antitrust authority, when prosecuting a breach of competition provisions, rules “primarily” on an infringement of the GDPR from cases in which such evaluations are merely “incidental”:

[T]he examination of an abuse of a dominant position on the market may justify the interpretation, by a competition authority, of rules other than those relating to competition law, such as those of the GDPR, while specifying that such an examination is carried out in an incidental manner and is without prejudice to the application of that regulation by the competent supervisory authorities.[130]

Given the differing objectives of competition and data-protection law, however, where an antitrust authority identifies an infringement of the GDPR in the context of finding of abuse of a dominant position, it does not replace the data-protection supervisory authorities.[131] Therefore, when examining whether an undertaking’s conduct is consistent with the GDPR, competition authorities are required to consult and cooperate sincerely with the competent data-protection authority in order to ensure consistent application of that regulation.[132] In addition, where the data-protection authority has ruled on the application of certain provisions of the GDPR with respect to the same practice or similar practices, the competition authority cannot deviate from that interpretation, although it remains free to draw its own conclusions from the perspective of applying competition law.[133]

While these principles are compelling, they do not appear conclusive in addressing the issue, for two main reasons.

First, as competition authorities have significant leeway in framing their investigations, it will be extremely difficult in practice to demonstrate that they are primarily—rather than incidentally—tackling a data-protection breach. In this regard, the German Facebook investigation represents an illustrative example. In the press release announcing the launch of the proceedings, the Bundeskartellamt stated that Facebook’s terms and conditions violated data-protection law and may “also” be regarded as abuses of a dominant position.[134] Later in the press release, however, in a section concerning the preliminary assessment, the authority changed that perspective, asserting that Facebook’s contractual terms were unfair, quite apart from any privacy infringement, and that, in assessing the competitive impact of such a strategy, it was “also” applying data-protection principles. Further, the Bundeskartellamt ascertained a privacy violation previously undetected by any data-protection authority. If the Facebook case fulfills both requirements of an incidental assessment of a privacy breach and sincere cooperation with the data-protection authority, it will be difficult to imagine any antitrust investigation not passing the bar.[135]

Second, the judgment only examines a scenario in which a GDPR infringement may occur, while not being useful to unraveling the very different situation in which the adoption of a privacy-enhancing solution is invoked as justification for anticompetitive conduct. In that case, cooperation between competition and data-protection authorities has thus far proven to be a harbinger of new issues and conflicts, rather than a panacea for all of the problems.

Finally, the CJEU also addressed another crucial topic of the integration between antitrust and privacy—that being the meaning of “consent” under the GDPR, and especially the requirement of freedom of consent. Supporters of an integrated approach find the legal basis of the privacy/antitrust marriage in the GDPR to be pivotally centered on the role assigned to freely given consent.[136] Notably, they imagine that the GDPR provides the legal basis for a link between data power and market power by stating that, among other things, there is no freely given consent to personal-data processing where there is a “clear imbalance” between the data subject and the controller.[137] In this respect, if the controller holds a dominant position on the market, it is argued that such market power could lead to a clear imbalance in the sense described in the GDPR.

According to the CJEU, however, while it may create such an imbalance, the existence of a dominant position alone cannot, in principle, render the consent invalid.[138] Notably, the fact that the operator of an online social network holds a dominant position on the social-network market does not, as such, prevent users of that social network from validly giving their consent, within the meaning of the GDPR, to the processing of their personal data by that operator. Consequently, the validity of consent should be examined on a case-by-case basis.

Moreover, as observed by Rantos, this does not imply that for market power to be relevant for GDPR enforcement, it needs to be regarded as a dominant position within the meaning of competition law.[139] Therefore, the relationship between data-protection and competition law is not one of mutual respect. While a competition authority is required to cooperate with a data-protection regulator in the case of a privacy breach, and is bound by the interpretation the latter gives of the GDPR, the converse does not apply with regard to the notion of “clear imbalance” under the GDPR. Data-protection authorities are granted significant leeway to establish market power under the GDPR.[140]

V.     Conclusion

The features of digital markets and the emergence of a few large online gatekeepers whose business models revolve around collecting and processing large amounts of data may suggest a link between market power and data power. Accordingly, scholars and policymakers have supported regulatory measures intended to promote data sharing and to empower individuals with more control over their personal data. From a different perspective, this also has led to the idea that competition and data-protection are intertwined and therefore require an integrated approach where, despite holding different objectives, antitrust enforcement should also protect privacy interests.

The integrationist movement claims that unity makes strength. According to this view, while competition and data-protection laws are, in isolation, considered unfit to safeguard their respective interests, the inclusion of privacy harms into antitrust assessments would allow competition authorities to better tackle data-accumulation strategies, and that the enforcement of antitrust rules would be more effective in ensuring data protection.

The purported complementarity, or even synergy, between competition and data-protection law appears, however, difficult to detect in practice. The only case in which a GDPR breach has been considered a proper legal basis for an antitrust intervention is the rather controversial Bundeskartellamt Facebook decision. Further, recent legislative initiatives that have introduced provisions clearly inspired by Facebook and essentially motivated by the aim of bypassing the traditional antitrust analysis (e.g., Article 5(2)DMA and Section 19a GWB) confirm the failure of the integrationist narrative and awareness that it would be impossible to endorse the Bundeskartellamt’s stance. Moreover, whether or not one would argue that the DMA represents a concrete and advanced attempt at integrating data-protection concerns in competition policy, it is worth pointing out that Article 5(2)DMA actually establishes antitrust deference toward privacy.

As if this were not enough, the idea of commingling antitrust and privacy has generated a significant side effect. As a reaction to Facebook and the DMA, some platforms have, indeed, adopted policy changes to restrict user-data tracking on their ecosystems in ways that undermine the effectiveness of rivals’ targeted advertising. The strategic use of privacy as a business justification to pursue anticompetitive advantages testifies once again to the tension between these fields of law. Further, as shown by the French Apple ATT investigation, the call for close cooperation between the authorities is often just a useless and rhetorical expedient.

The proposal to integrate competition and data-protection law in digital markets has been submitted as a much-needed boost to strengthen antitrust enforcement against gatekeepers and their data strategies. Moving away from pure efficiency-oriented assessments to embrace broader social interests, advocates claim, would help ensure more aggressive and effective antitrust enforcement. Including privacy harms in antitrust proceedings turns out, instead, to be a potential curse for competition authorities, providing the major digital players with an opportunity for regulatory gaming to undermine antitrust enforcement.

This should serve as a cautionary tale about the risks of twisting rules to achieve policy outcomes and the importance of respecting the principles and scope of different areas of law.

 

[1] See Jacques Cre?mer, Yves-Alexander de Montjoye, & Heike Schweitzer, Competition Policy for the Digital Era, (2019) Report for the European Commission, 4, available at https://ec.europa.eu/competition/publications/reports/kd0419345enn.pdf (referring to the possibility that a dominant platform could have incentives to sell “monopoly positions” to sellers by showing buyers alternatives that do not meet their needs).

[2] See Alessandro Bonatti, The Platform Dimension of Digital Privacy, forthcoming in The Economics of Privacy, (Avi Goldfard & Catherine Tucker, eds.), University of Chicago Press; Daron Acemoglu, Ali Makhdoumi, Azarakhsh Malekian, & Asu Ozdaglar, Too Much Data: Prices and Inefficiencies in Data Markets, 14 Am Econ J Microecon 218 (2022); Shota Ichihashi, The Economics of Data Externalities, 196 J. Econ. Theory 105316 (2021); Omri Ben-Shahar, Data Pollution, 11 J. Leg. Anal. 104 (2019); Jay Pil Choi, Doh-Shin Jeon, & Byung-Cheol Kim, Privacy and Personal Data Collection with Information Externalities, 173 J. Public Econ. 113 (2019); see also Jeanine Miklós-Thal, Avi Goldfarb, Avery M. Haviv, & Catherine Tucker, Digital Hermits, NBER Working Paper No. 30920 (2023), (arguing that, as advances in machine learning allow firms to infer more accurately sensitive data from data that appears otherwise innocuous, users’ data-sharing decisions polarize between a group of users choosing to share no data and another group choosing to share all their data (sensitive or not sensitive)).

[3] See, e.g., Competition and Data Protection in Digital Markets: A Joint Statement Between the CMA and the ICO, UK Competition and Markets Authority and Information Commissioner’s Office, (2021) 5, https://www.gov.uk/government/publications/cma-ico-joint-statement-on-competition-and-data-protection-law [hereinafter “CMA-ICO Joint Statement”]; Privacy and Competitiveness in the Age of Big Data: The Interplay Between Data Protection, Competition Law and Consumer Protection in the Digital Economy, European Data Protection Supervisor (2014) https://edps.europa.eu/data-protection/our-work/publications/opinions/privacy-and-competitiveness-age-big-data_en.

[4] See, e.g., Investigation of Competition in Digital Markets’, Majority Staff Reports and Recommendations, U.S. House Energy and Commerce Subcommittee on Antitrust, Commercial, and Administrative Law (2020), 28, available at https://www.govinfo.gov/content/pkg/CPRT-117HPRT47832/pdf/CPRT-117HPRT47832.pdf [hereinafter, “Antitrust Subcommittee Report”]; Frank Pasquale, Privacy, Antitrust, and Power, 20 George Mason Law Rev. 1009 (2013); Pamela J. Harbour & Tara I. Koslov, Section 2 in a Web 2.0 World: An Expanded Vision of Relevant Product Markets, 76 Antitrust Law J. 769 (2010).

[5] See, e.g., Antitrust Subcommittee Report, supra note 4, 39, citing Howard A. Shelanski, Information, Innovation, and Competition Policy for the Internet, 161 U. Pa. L. Rev. 1663 (2013), to argue that “[t]he persistent collection and misuse of consumer data is an indicator of market power in the digital economy”; European Data Protection Supervisor, supra note 3, 35, stating that, where there are a limited number of operators or when one operator is dominant, “the concept of consent becomes more and more illusory;” see also, Online Platforms and Digital Advertising, UK Competition and Markets Authority (2020) para. 6.26, available at https://assets.publishing.service.gov.uk/media/5fa557668fa8f5788db46efc/Final_report_Digital_ALT_TEXT.pdf, stating that “[i]n a more competitive market, we would expect that it would be clear to consumers what data is collected about them and how it is used and, crucially, the consumer would have more control. We would then expect platforms to compete with one another to persuade consumers of the benefits of sharing their data or adopt different business models for more privacy-conscious consumers.” However, see also James C. Cooper & John M. Yun, Antitrust & Privacy: It’s Complicated, J. Law Technol. Policy 343 (2022), finding no systematic relationship between privacy ratings and market concentration.

[6] See, e.g., Report on Social Media Services, Australian Competition & Consumer Commission (2023), 128, https://www.accc.gov.au/media-release/accc-report-on-social-media-reinforces-the-need-for-more-protections-for-consumers-and-small-business; Rebecca Kelly Slaughter, The FTC’s Approach to Consumer Privacy, Federal Trade Commission (2019) 3, available at https://www.ftc.gov/system/files/documents/public_statements/1513009/slaughter_remarks_at_ftc_approach_to_consumer_privacy_hearing_4-10-19.pdf.

[7] Antitrust Subcommittee Report, supra note 4, 28; Maurice E. Stucke & Ariel Ezrachi, When Competition Fails to Optimise Quality: A Look at Search Engines, 18 Yale J. Law Technol. 70 (2016).

[8] Pamela J. Harbour, Dissenting Statement in the Matter of Google/DoubleClick, Federal Trade Commission (2007), 4, available at https://www.ftc.gov/sites/default/files/documents/public_statements/statement-matter-google/doubleclick/071220harbour_0.pdf.

[9] For a critical perspective, see Giuseppe Colangelo, In Fairness We (Should Not) Trust: The Duplicity of the EU Competition Policy Mantra in Digital Markets, Antitrust Bulletin (forthcoming).

[10] See Cristina Caffarra & Johnny Ryan, Why Privacy Experts Need a Place at the Antitrust Table, ProMarket (2021) https://www.promarket.org/2021/07/28/privacy-experts-antitrust-data-harms-digital-platforms, arguing that “[t]here is a market power crisis and a privacy crisis, and they compound each other.”

[11] See, e.g., Wolfgang Kerber & Karsten K. Zolna, The German Facebook Case: The Law and Economics of the Relationship Between Competition and Data Protection Law, 54 Eur. J. Law Econ. 217 (2022), arguing that digital markets exhibit two types of market failure (i.e., competition problems on the one hand, and information and behavioral problems on the other) and suggesting that the effectiveness of enforcement should also be an important criterion for determining which policy should deal with a case if both laws can be applied. Accordingly, if data-protection law is uncapable of dealing effectively with privacy issues and competition law appears better able to overcome this challenge, then the competition authority should step in as the lead enforcer. On the enforcement failure of old and new data-protection regimes, see Filippo Lancieri, Narrowing Data Protection’s Enforcement Gap, 74 Maine Law Rev. 15 (2022).

[12] For an overview of various theories that have emerged in the literature, see Erika M. Douglas, The New Antitrust/Data Privacy Law Interface, Yale L.J. F. 647 (2021); Giuseppe Colangelo & Mariateresa Maggiolino, Data Protection in Attention Markets: Protecting Privacy Through Competition? 8 J. Eur. Compet. Law Pract. 363 (2017). See also, Consumer Data Rights and Competition Background: Note by the Secretariat, OECD (2020), available at https://one.oecd.org/document/DAF/COMP(2020)1/en/pdf, and Geoffrey A. Manne & Ben Sperry, The Problems and Perils of Bootstrapping Privacy and Data into an Antitrust Framework, CPI Antitrust Chronicle 2 (2015), exploring the difficulties associated with incorporating consumer-data considerations into competition policy and enforcement.

[13] See Noah Joshua Phillips, Remarks at the Mentor Group Paris Forum, Federal Trade Commission (2019), 13-15, https://www.ftc.gov/news-events/news/speeches/remarks-commissioner-noah-joshua-phillips-mentor-group-paris-forum; and Maureen K. Ohlhausen & Ben Rossen, Privacy and Competition: Discord or Harmony? 67 Antitrust Bulletin 552 (2022).

[14] See, e.g., Susan Athey, Christian Catalini, & Catherine E. Tucker, The Digital Privacy Paradox: Small Money, Small Costs, Small Talk, NBER Working Paper No. 23488 (2017); Alessandro Acquisti, Curtis Taylor, & Liad Wagman, The Economics of Privacy, 54 J Econ Lit 442 (2016). See also, Avi Goldfarb & Catherine Tucker, Shifts in Privacy Concerns, 102 Am Econ Rev: Papers and Proceedings 349 (2012), noting that individuals’ privacy preferences evolve over time; notably, as people grow older. they get more privacy-conscious. See also Jeffrey T. Prince & Scott Wallsten, How Much Is Privacy Worth Around the World and Across Platforms?, 31 J Econ Manag Strategy. 841 (2022), estimating individuals’ valuation of online privacy across countries (United States, Mexico, Brazil, Colombia, Argentina, and Germany) and data types (personal information on finances, biometrics, location, networks, communications, and web browsing), and finding that Germans value privacy more than people in the United States and Latin American countries do and that, across countries, people most value privacy for financial and biometric information.

[15] Giuseppe Colangelo & Mariateresa Maggiolino, Antitrust Über Alles. Whither Competition Law After Facebook?, 42 World Competition Law and Economics Review 355 (2019).

[16] See, e.g., Federal Trade Commission v. Facebook, Case No. 1:20-cv-03590 (D.D.C. 2021), para. 163, arguing that “[t]he benefits to users of additional competition include some or all of the following: … variety of data protection privacy options for users, including, but not limited to, options regarding data gathering and data usage practices”; and U.S. et al. v. Google, No. 1:20-cv-03010 (D.D.C. 2020), para. 167, arguing that “[b]y restricting competition in general search services, Google’s conduct has harmed consumers by reducing the quality of general search services (including dimensions such as privacy, data protection, and use of consumer data), lessening choice in general search services, and impeding innovation.” See also, Executive Order on Promoting Competition in the American Economy, The White House (2021), https://www.whitehouse.gov/briefing-room/presidential-actions/2021/07/09/executive-order-on-promoting-competition-in-the-american-economy, urging federal agencies to pay closer attention to “unfair data collection and surveillance practices that may damage competition, consumer autonomy, and consumer privacy.”

[17] See Margrethe Vestager, Tearing Down Big Tech’s Walls, Project Syndicate (2023) https://www.project-syndicate.org/commentary/eu-big-tech-legislation-digital-services-markets-by-margrethe-vestager-2023-03, stating that “[w]e are proud that Europe has become the cradle of tech regulation globally.”

[18] Regulation (EU) 2016/679 of the European Parliament and of the Council of 27 April 2016 on the protection of natural persons with regard to the processing of personal data and on the free movement of such data, and repealing Directive 95/46/EC, [2016] OJ L 119/1, Article 20. See Bert-Jaap Koops, The Trouble with European Data Protection Law, 4 Int. Data Priv. Law 4, 44 (2014), arguing that “[b]y its nature, data portability would be more at home in the regulation of unfair business practices or electronic commerce, or perhaps competition law—all domains that regulate abuse of power by commercial providers to lock-in consumers.”

[19] Bundeskartellamt, 7 February 2019, Case B6-22/16.

[20] CMA-ICO Joint Statement, supra note 3, 18-19.

[21] Ibid., 23.

[22] Douglas, supra note 12.

[23] See, e.g., hiQ Labs v. LinkedIn, 938 F.3d 985 (9th Cir. 2019), affirmed 31 F.4th 1180 (9th Cir. 2022), allowing hiQ continued access to LinkedIn users’ profile information in the name of competition. Notably, the court pointed out that hiQ’s entire business depends on being able to access public LinkedIn member profiles and that, at the same time, there is little evidence that LinkedIn users who choose to make their profiles public actually maintain an expectation of privacy with respect to the information that they post publicly. Therefore, “even if some users retain some privacy interests in their information notwithstanding their decision to make their profiles public, we cannot, on the record before us, conclude that those interests—or more specifically, LinkedIn’s interest in preventing hiQ from scraping those profiles—are significant enough to outweigh hiQ’s interest in continuing its business, which depends on accessing, analyzing, and communicating information derived from public LinkedIn profiles.”

[24] See, e.g., Epic Games v. Apple, 559 F. Supp. 3d 898, 922–23 (N.D. Cal. 2021), affirmed in part and reversed in part 2023 U.S. App. LEXIS 9775 (9th Cir. 2023), finding that Apple’s restrictions are designed to improve device security and user privacy; and District Court (Rechtbank) of Rotterdam, 24 December 2021, Case No. ROT 21/4781 and ROT 21/4782, dismissing the arguments that Apple’s in-app payment system is needed for security and privacy.

[25] See, e.g., Autorità Garante della Concorrenza e del Mercato, 11 May 2023, Case A561; Press Release, Bundeskartellamt Reviews Apple’s Tracking Rules for Third-Party Apps, Bundeskartellamt (2022), https://www.bundeskartellamt.de/SharedDocs/Meldung/EN/Pressemitteilungen/2022/14_06_2022_Apple.html; Autorité de la Concurrence, 17 March 2021, Decision 21-D-07, Apple, https://www.autoritedelaconcurrence.fr/en/decision/regarding-request-interim-measures-submitted-associations-interactive-advertising-bureau; Apple – The President of UOKiK Initiates an Investigation, Urz?d Ochrony Konkurencji i Konsumentów (2021), https://uokik.gov.pl/news.php?news_id=18092. See also, Mobile Ecosystems: Market Study Final Report, UK Competition and Markets Authority (2022) Chapter 6 and Appendix J, https://www.gov.uk/cma-cases/mobile-ecosystems-market-study.

[26] Phillips, supra note 13, 15.

[27] CJEU (Grand Chamber), 4 July 2023, Case C-252/21, Meta Platforms v. Bundeskartellamt, EU:C:2023:537.

[28] See, e.g., European Data Protection Supervisor, supra note 3, 26, stating that “clearly power is achieved through control over massive volumes of data on service users.”

[29] See GDPR, supra note 18, Recital 7.

[30] European Data Protection Supervisor, supra note 3, 26.

[31] CMA-ICO Joint Statement, supra note 3, 5.

[32] Nicholas Economides & Ioannis Lianos, Restrictions on Privacy and Exploitation in the Digital Economy: A Market Failure Perspective, 17 J. Competition Law Econ. 765 (2021).

[33] Competition Law and Data, Autorité de la Concurrence and Bundeskartellamt (2016), 25, available at https://www.bundeskartellamt.de/SharedDocs/Publikation/DE/Berichte/Big%20Data%20Papier.pdf?__blob=publicationFile&v=2.

[34] Economides & Lianos, supra note 32.

[35] Ibid., 770-771.

[36] GDPR, supra note 18, Article 4(11).

[37] Ibid., Article 7.

[38] Autorité de la Concurrence and Bundeskartellamt, supra note 33, 25. See also Australian Competition & Consumer Commission, supra note 6, 41, arguing that exploitative conduct involves the use of market power to “give less and charge more” and that, for consumers, this may involve lower-quality services or the excessive costs of providing personal data to access services.

[39] Autorité de la Concurrence and Bundeskartellamt, supra note 33, 24.

[40] Facebook, supra note 19. For a comment on the different episodes of the Facebook saga, see, e.g., Kerber and Zolna, supra note 11; Anne C. Witt, Excessive Data Collection as a Form of Anticompetitive Conduct: The German Facebook Case, 66 Antitrust Bulletin 276 (2021); Marco Botta and Klaus Wiedemann, The interaction of EU competition, consumer, and data protection law in the digital economy: the regulatory dilemma in the Facebook odyssey, 64 Antitrust Bulletin 428 (2019); Colangelo and Maggiolino, supra note 15.

[41] Facebook, supra note 19, paras. 778-780 and 792, stating that users could not have expected that the platform would analyse data emanating from other websites and, when they had the opportunity to read Facebook’s terms of service, users could barely understand the reasons why Facebook was processing and combining their data since Facebook’s terms of service were very complex, replete with links to other explanations, and significantly too opaque to allow ordinary users to understand its data policy.

[42] Ibid., section B(II), stating that voluntary consent to users’ information being processed cannot be assumed if their consent is a prerequisite for using the Facebook service in the first place.

[43] Ibid., para. 645, highlighting that GDPR’s Recitals 42 and 43 state that consent is not freely given where consumers have no alternative options, or where there are clear power imbalances. See also Inge Graef & Sean Van Berlo, Towards Smarter Regulation in the Areas of Competition, Data Protection and Consumer Law: Why Greater Power Should Come with Greater Responsibility, 12 Eur. J. Risk Regul. 674 (2021), arguing that, in formulating this two-way interaction between data-protection law and competition law, the Bundeskartellamt has not only incorporated data-protection principles into its competition analysis, but similarly transferred elements of competition law into data protection; and Orla Lynskey, Grappling With ‘Data Power’: Normative Nudges From Data Protection and Privacy, 20 Theor. Inq. Law 189 (2019), supporting the view that the GDPR provides a normative foundation for imposing a special responsibility on controllers holding data power, analogous to the special responsibility that competition law imposes on dominant firms.

[44] See Press Release, Bundeskartellamt Prohibits Facebook From Combining User Data From Different Sources, Bundeskartellamt (2019), https://www.bundeskartellamt.de/SharedDocs/Publikation/EN/Pressemitteilungen/2019/07_02_2019_Facebook.html;jsessionid=8A581062B36687451A3D1E7A5C256390.2_cid378?nn=3600108, arguing that “[t]he combination of data sources substantially contributed to the fact that Facebook was able to build a unique database for each individual user and thus to gain market power.”

[45] Facebook FAQs, Bundeskartellamt (2019), 6, https://www.bundeskartellamt.de/SharedDocs/Publikation/EN/Pressemitteilungen/2019/07_02_2019_Facebook_FAQs.pdf?__blob=publicationFile&v=6.

[46] See Colangelo & Maggiolino, supra note 15.

[47] Press Release, Meta (Facebook) Introduces New Accounts Center – An Important Step in the Implementation of the Bundeskartellamt’s Decision, Bundeskartellamt (2023), https://www.bundeskartellamt.de/SharedDocs/Meldung/EN/Pressemitteilungen/2023/07_06_Meta_Daten.html.

[48] Colangelo & Maggiolino, supra note 15.

[49] OLG Du?sseldorf, 26 August 2019, Case VI-Kart 1/19 (V), 10.

[50] Ibid., 11.

[51] Ibid., 12.

[52] Bundesgerichtshof, 23 June 2020, Case KVR 69/19.

[53] Ibid., para. 58.

[54] Ibid..

[55] Ibid., para. 86.

[56] Ibid., para. 94.

[57] OLG Du?sseldorf, 24 March 2021, Case Kart 2/19 (V).

[58] Meta, supra note 27.

[59] Autorità Garante della Concorrenza e del Mercato, 10 December 2018, Case PS11112, Facebook-Condivisione dati con terzi.

[60] Nederlandstalige Rechtbank van Eerste Aanleg te Brussel, 16 February 2018.

[61] Regulation (EU) 2022/1925 on contestable and fair markets in the digital sector and amending Directives (EU) 2019/1937 and (EU) 2020/1828 (Digital Markets Act) [2022] OJ L 265/1, Article 5(2).

[62] Ibid., Recital 36.

[63] Ibid., Recital 37.

[64] For critical analysis of this issue and more generally on the controversial relationship between the DMA and the GDPR, see Alba Ribera Marti?nez, The Circularity of Consent in the DMA: A Close Look into the Prejudiced Substance of Articles 5(2) and 6(10), Concorrenza e Mercato (forthcoming). See also Marco Botta & Danielle Da Costa Leite Borges, User’s Consent Under Art. 5(2) Digital Markets Act (DMA): Exploring the Complex Relationship Between the DMA and the GDPR, EUI RSC Working Paper (forthcoming), arguing that, while respecting the general criteria indicated by Art. 7 GDPR, the users’ consent under Art. 5(2) DMA should be adjusted to the DMA peculiarity and that the DMA should be considered as a lex specialis, taking precedence over the GDPR in case of conflict. Previously, the revised e-Privacy Directive introduced an opt-in system for website cookies: see Directive 2009/136/EC amending Directive 2002/22/EC on universal service and users’ rights relating to electronic communications networks and services, Directive 2002/58/EC concerning the processing of personal data and the protection of privacy in the electronic communications sector and Regulation (EC) No 2006/2004 on cooperation between national authorities responsible for the enforcement of consumer protection laws, (2009) OJ L 337/11, Article 5(3).

[65] DMA, supra note 61, Recital 72.

[66] Press Release, Statement of Objections Issued Against Google’s Data Processing Terms, Bundeskartellamt (2023), https://www.bundeskartellamt.de/SharedDocs/Meldung/EN/Pressemitteilungen/2023/11_01_2023_Google_Data_Processing_Terms.html.

[67] Ibid.

[68] Entwurf Eines Gesetzes zur A?nderung des Gesetzes Gegen Wettbewerbsbeschra?nkungen fu?r ein Fokussiertes, Proaktives und Digitales Wettbewerbsrecht 4.0 und Anderer Wettbewerbsrechtlicher Bestimmungen, Bundestag (2020), available at https://dserver.bundestag.de/btd/19/234/1923492.pdf.

[69] See Giuseppe Colangelo, The European Digital Markets Act and Antitrust Enforcement: A Liaison Dangereuse, 47 Eur. Law Rev. 597 (2022).

[70] Bundeskartellamt, 30 December 2021, Case B7-61/21, https://www.bundeskartellamt.de/SharedDocs/Entscheidung/EN/Entscheidungen/Missbrauchsaufsicht/2022/B7-61-22.html.

[71] Bundeskartellamt, 5 October 2023, Case B7-70/21.

[72] The Bundeskartellamt identified four main deficiencies to support its prohibition of Google’s data-processing terms (ibid., paras. 50-54). Namely, because of a lack of sufficient granularity in the settings options, users could not opt out of cross-service data processing or limit data processing to the Google service in which the data were generated. End users could only choose between accepting personalization across all services or opting out of personalization altogether. Further, users were not given sufficient choice within the meaning of Section 19a GWB, as in some cases, Google offers users no choice at all as to data-processing options. Furthermore, the settings options that Google offered lacked sufficient transparency—i.e., sufficiently concise and comprehensible indications providing users with sufficient information as to whether, how, and for what purpose Google processes data across services. Finally, when creating a Google account, a user’s options consent or reject consent were not equivalent.

[73] Ibid., para. 78.

[74] See, e.g., Inge Graef, Damian Clifford, & Peggy Valcke, Fairness and Enforcement: Bridging Competition, Data Protection, and Consumer Law, 8 Int. Data Priv. Law 200, 219-220 (2018).

[75] European Commission, 11 March 2008, Case COMP/M.4731. Previously, in a different setting (i.e., discussing an exchange-of-information case), the CJEU (23 November 2006, Case C-238/05, Asnef-Equifax, EU:C:2006:734, para. 63) affirmed that “any possible issues relating to the sensitivity of personal data are not, as such, a matter for competition law, they may be resolved on the basis of the relevant provisions governing data protection.”

[76] Google/DoubleClick, supra note 75, para. 364. See also para. 365, where the Commission noted that “that the combination of data about searches with data about users’ web surfing behaviour [wa]s already available to a number of Google’s competitors.”

[77] Ibid., para. 368.

[78] European Commission, 3 October 2014, Case COMP/M.7217.

[79] Ibid., para. 189.

[80] Ibid., para. 164.

[81] European Commission, 6 December 2016, Case COMP/M.8124.

[82] Ibid., fn 330.

[83] Ibid., para. 180.

[84] Ibid., para. 177.

[85] Ibid., para. 178.

[86] Ibid., para. 179.

[87] European Commission, 6 September 2018, Case COMP/M.8788, paras. 221 and 314.

[88] European Commission, 17 December 2020, Case COMP/M.9660.

[89] See, Statement on Privacy Implications of Mergers, European Data Protection Board (2020), available at https://edpb.europa.eu/sites/default/files/files/file1/edpb_statement_2020_privacyimplicationsofmergers_en.pdf, arguing that “(t)here are concerns that the possible further combination and accumulation of sensitive personal data regarding people in Europe by a major tech company could entail a high level of risk to the fundamental rights to privacy and to the protection of personal data.”

[90] Google/Fitbit, supra note 84, para. 410.

[91] Ibid., fn. 299.

[92] Ibid., fn. 300.

[93] European Commission, 21 December 2021, Case COMP/M.10290.

[94] European Commission, 27 January 2022, Case COMP/M.10262.

[95] Press Release, Commission Clears Creation of a Joint Venture by Deutsche Telekom, Orange, Telefo?nica and Vodafone, European Commission (2023), https://ec.europa.eu/commission/presscorner/detail/en/IP_23_721. Previously, in a similar vein, see European Commission, 4 September 2012, Case COMP/M.6314, Telefo?nica UK/Vodafone UK/ Everything Everywhere/ JV.

[96] UK Competition and Markets Authority, supra note 25, Appendix J.

[97] See, e.g., Nils Wernerfelt, Anna Tuchman, Bradley Shapiro, & Robert Moakler, Estimating the Value of Offsite Data to Advertisers on Meta, SSRN (2022) https://papers.ssrn.com/sol3/papers.cfm?abstract_id=4176208, finding that the costs to acquire new consumers through targeted advertisements increases tremendously without access to offsite data. On the value of external data and on the relevance (especially for small and medium-sized players) of gaining access to external data from large players in the marketplace, see also Xiaoxia Lei, Yixing Chen, & Ananya Sen, The Value of External Data for Digital Platforms: Evidence from a Field Experiment o