Showing 9 of 675 Publications by Geoffrey A. Manne

ICLE Comments Re: Request for Information on Consolidation in Health Care Markets

Regulatory Comments I. Introduction/Summary We appreciate the opportunity to respond to this request for information on consolidation in health-care markets, Docket No. ATR-102, issued by the U.S. . . .

I. Introduction/Summary

We appreciate the opportunity to respond to this request for information on consolidation in health-care markets, Docket No. ATR-102, issued by the U.S. Justice Department (DOJ), the U.S. Department of Health and Human Services (HHS), and the Federal Trade Commission (FTC) (collectively, “the agencies”).[1] We agree wholeheartedly that robust competition in health-care markets is critical to consumer welfare and the U.S. economy. As an FTC staff policy paper summarized, “[c]ompetition in health care markets benefits consumers by helping to control costs and prices, improve quality of care, promote innovative products, services, and service delivery models, and expand access to health care services and goods.”[2] Conversely, anticompetitive health-care acquisitions can harm competition and consumer welfare; there is a substantial economic literature to that effect.[3] The agencies, over decades, have done well to oppose such mergers and, as a related matter, attempts to insulate such mergers from federal antitrust scrutiny.[4] Hence, the agencies have important roles to play in protecting health-care competition, as they enforce key federal statutes relevant to it, including the general antitrust laws that are enforced by the FTC and the DOJ,[5] and that apply across health-care markets, among others. Effective and accurate antitrust enforcement is a key component of health-care policy, and one that tends to benefit patients and other health-care consumers, including both private and public payers.

We recognize that the FTC also has a distinctive research and reporting mission assigned by Congress under Section 6 of the FTC Act, and that it has decades of experience engaging in policy and economic research, both internally and in cooperation with DOJ and HHS.[6] Acknowledging express statutory limits—such as the restriction on studies and reports regarding the business of insurance under Section 6(l) of the FTC Act[7]—the fulfillment of that research mission across health-care markets has been wide-ranging; and, as described below, it has often found salutary application in antitrust enforcement.

Our response to the agencies’ RFI comprises, at the highest level of generality, one observation and one recommendation. The observation is that, while health-care-provider acquisitions remain an extremely important domain of merger scrutiny, neither enforcement experience nor the economic literature support any fundamental changes in procedural or substantive antitrust law or regulation, whether for provider acquisitions generally or any of the categories of acquirers specified in the RFI. Competition policy is not, and should not be, static. At the same time, sound policy reform is a difficult, stepwise process, and one that requires a firm foundation in both research and enforcement experience, along with attention to established precedent. Information submitted in response to the RFI may well contribute to the agencies’ aggregate knowledge base on provider transactions. But the present inquiry does not appear designed to move that body of knowledge much beyond the margin. Indeed, as explained below, FTC research and enforcement experience underscore not just the importance of health-care competition, but also how complex the tasks of merger scrutiny and reform are.

Correspondingly, our overarching recommendation is that the agencies build on the substantial body of research regarding mergers and acquisitions in the health-care sector that has been conducted over the course of several decades by agency staff and others. That body of research includes, notably, contributions made by the staff of the FTC Bureau of Economics (BE).[8] More specifically, to that end, we recommend that economic and policy staff at the agencies synthesize the extant body of research at their disposal. To be sure, market developments, and developments in research methods and available data, may suggest new avenues of research, as well as those in need of significant updates. But a serious, critical synthesis of the available literature will only help to sharpen the agencies’ sense of new research demands, just as it will provide a basis on which to contemplate new enforcement initiatives. Such a synthesis can also ground more focused and productive requests for information on critically important issues in health-care competition going forward.

Our recommendation of such a research synthesis or review is consistent with the agencies’ acknowledgement that they may require “additional proceedings, including workshops or other public engagement, to learn more about [concerns identified in response to the RFI].”[9] While such workshops and other engagements have been a useful component of the agencies’ understanding of health-care-competition policy, we stress that they should be seen as complements to rigorous systematic research. And both should be seen as complements to building on case-specific agency enforcement experience, which typically scrutinizes the specific facts and circumstances of transactions and other firm conduct.

For example, in many smaller markets, independent providers of hospital-based services, such as anesthesiology, may be highly concentrated on any standard for “highly concentrated” markets.[10] Further research might aid the agencies in examining highly concentrated provider markets to develop filters or screens for provider acquisitions below the Hart-Scott-Rodino filing threshold, so that the agencies might identify and investigate those sub-threshold filings that are the most likely candidates for investigations and, depending on the results of those investigations, enforcement actions. Efficient matter-selection tools will be critical to that effort, less the agencies commit scarce resources to small, unpromising investigations and impose undue costs on health-care providers.

In our comments below, we recognize that the agencies themselves have established models for building the sort of “policy R&D” contemplated by the RFI in a way that complements their enforcement mandates. Also, we understand that the RFI is but one of the tools the agencies use to further their understanding of provider consolidation.[11] And, indeed, the RFI may contribute to the larger health-care-competition R&D programs at the agencies, if only at the margin.

At the same time, we write to note certain concerns about the agencies’ framing of their RFI, including, specifically, elements of the RFI that appear to be in tension with learning from agency-sponsored research and agency-enforcement experience.

Some of our concerns may be summarized as follows:

First, evidence and enforcement experience do not identify categories of health-care acquisitions that “always or almost always” impede competition and reduce output. That militates against per se prohibitions. Absent an express charge from Congress, new competition regulations regarding health-care acquisitions are not justified.

Second, agency experience—and, indeed, the FTC’s landmark success in the 2nd U.S. Circuit Court of Appeals in the American Medical Association case[12]—suggests legitimate competition concerns about undue restraints on “the corporate practice of medicine.” More broadly—and consistent with longstanding agency practice—the agencies should be cautious about drawing general conclusions about whole industries, business models, or methods of health-care delivery, and more cautious still in condemning them.

Third, while there is no doubt that provider consolidation can be anticompetitive, the relationship between concentration and competition is complex, both as a general matter and—on current understanding—across the various sectors and subsectors identified in the RFI. That general point has been illustrated by BE staff research, even as that research has helped to refine and strengthen appropriate antitrust scrutiny of health-care provider mergers and acquisitions.

Fourth, it is well-understood that vertical acquisitions can harm competition and consumers under certain conditions. At the same time, vertical mergers are not generally—or even typically—anticompetitive. Vertical mergers may entail certain efficiencies, and are commonly procompetitive or benign on net, as research by agency personnel and the larger academic community has demonstrated. Analogously, while conglomerate mergers may raise competition concerns, they are not generally anticompetitive.

Fifth, the RFI’s framing seems problematic—both uncharacteristic of open inquiry and in tension with antitrust experience and its economic foundations. For example, we question a statement at the outset of the RFI: “Given recent trends, we are concerned that transactions may generate profits for those firms at the expense of patients’ health, workers’ safety, and affordable health care for patients and taxpayers.”[13] To be sure, some provider transactions, under particular facts and circumstances, may harm competition and consumer welfare, in violation of the antitrust laws.[14] But the agencies understand that antitrust law and economics do not recognize any general or fundamental tension between firm profits on the one hand, and the consumer benefits typically associated with competition on the other. Indeed, FTC research and enforcement have specifically undermined the notion that not-for-profit provider mergers should be treated differently under the antitrust laws.[15]

More generally, health-care acquisitions can prove anticompetitive, procompetitive, or benign, but the RFI pointedly does not request information on potential patient or payor benefits that may be associated with consolidation. More generally, the RFI does not seem to recognize that health-care acquisitions commonly entail tradeoffs of benefits and costs. Such tradeoffs are well-documented in the literature and are recognized in U.S. merger jurisprudence.

As a related matter, the FTC’s recent workshop, “Private Capital, Public Impact: An FTC Workshop on Private Equity in Health Care,”[16] seemed uncharacteristically lax and imbalanced. The workshop was announced and timed to make it appear a complement to the RFI.[17] While sponsored by the FTC, leadership from the DOJ and HHS also participated. But there were no participants from the industry in question or from insurers, large health plans, or other private payors. Instead, participants—including those providing largely anecdotal evidence—appear to have been chosen exclusively for the purpose of representing agency and third-party criticism of private equity in health care. That is, the workshop seems to have been conclusory by design.

II. Discussion

A. Economic Research and Other Forms of Policy R&D Provide a Critical Foundation for Enforcement Policy

A 2009 report by then-FTC Chairman William Kovacic defines “policy R&D” broadly in a way that comprises, but is not limited to, original, author-initiated academic research by BE staff.[18] It also incorporates diverse forms of policy inquiries, including, e.g., hearings,[19] workshops,[20] conferences,[21] and, indeed, requests for public information.[22] These can all be mutually reinforcing, providing expert input that range from issue-spotting to literature review, to the presentation of new data and studies, as well as diverse perspectives on agency interests and activities. They can, in turn, help to inform case selection and enforcement, just as enforcement experience can yield data and other inputs into subsequent policy R&D. But one need not gainsay concerns about health-care competition or specific types of acquisitions to appreciate the difficulty of grounded, systematic reform of enforcement policy in these areas. We appreciate the agencies’ recent extension of the deadline for submissions in response to the RFI; that will likely increase the utility of the inquiry. Still, while the present RFI may be a useful endeavor, it is just one tool—in itself, a limited one—in the agencies’ “policy R&D” toolbox.

Below, we sketch some of the long-running developments in the agencies’ policy R&D pertinent to provider acquisitions and health-care consolidation. Our description of the many pertinent agency endeavors focuses on work by FTC staff and leadership, in large part because of the FTC’s enforcement experience with provider mergers and its sustained health-care-competition research program. We recognize, of course, that DOJ and HHS have also made substantial contributions of their own and, in turn, that the empirical literature regarding health-care consolidation is considerable, if not vast. We recognize, too, that inquiries are ongoing, and not restricted to the RFI. Our sketch is an abridged one, partly because the agencies—and, certainly, the FTC Bureau of Economics—are well familiar with their own research programs, just as they are familiar with the challenges of building lasting enforcement reforms.

On the one hand, we mean to underscore the advances made in understanding the competitive effects of provider consolidation and its potential—both realized and residual—for application in rigorous enforcement. At the same time—and based in no small part on their own contributions to understanding health-care consolidation—the agencies should appreciate the complexity and challenge of the Policy R&D project, both across health-care sectors and within them. That complexity and challenge militate against hasty conclusions about, e.g., specific sectors, business models, and competitive trends.

1. Policy investigations

Varied hearings, workshops, RFIs, and other agency policy tools have played a significant role in developing competition policy at the agencies, even if no single agency workshop or RFI is likely to generate a record adequate to justify a significant change in enforcement policy. For example, from February through October 2003, FTC and DOJ jointly conducted 27 days of hearings on health-care-competition issues, with testimony from diverse stakeholders from academia, industry, and, not incidentally, agency staff.[23] Although HHS did not cosponsor those hearings, representatives from various HHS agencies—including the Centers for Medicare & Medicaid Services (CMS) and the Agency for Healthcare Research and Quality (AHRQ)—provided testimony and otherwise consulted on the hearings.

Based on the record of those hearings, an FTC-sponsored workshop in September 2002, and independent research (including applied-industrial-organization research conducted within and without the agencies), FTC and DOJ jointly published a substantial policy report in 2004.[24] The report reviewed systematic research, diverse stakeholder perspectives, and numerous health-care-competition policy issues. It also presented concrete policy recommendations by FTC and DOJ, drawn from that review.[25]

Follow-up workshops conducted by FTC staff, such as the 2008 workshop “Innovations in Health Care Delivery,”[26] also included participation by HHS personnel, including that of the national coordinator for health information technology and the deputy director for health information privacy at the HHS Office for Civil Rights.[27] A 2014 FTC workshop[28] and 2015 joint FTC/DOJ workshop[29] on health-care-competition issues both similarly involved officials and other personnel from HHS, FTC, DOJ, and other agencies, as well as academics, practitioners, and diverse industry stakeholders.

More focused health-care-competition and policy workshops have also informed agency enforcement policy. For example, a 2010 workshop on accountable care organizations (ACOs) jointly conducted by the FTC, the DOJ, and HHS, together with a 2011 FTC workshop on ACOs (with participation from DOJ staff),[30] informed the joint FTC/DOJ enforcement-policy statement on ACOs,[31] which was developed in consultation with the HHS Centers for Medicare and Medicaid Services, and which applied to specific forms of provider collaborations (not mergers) under the Medicare Shared Savings program.

2. Economic research on provider consolidation

The wide-ranging policy inquiries described above were not conducted in a vacuum. Rather, they build on a larger body of economic research and enforcement experience, including, notably, research on health-care competition from within and without BE, coupled with enforcement by the FTC Bureau of Competition. Staff and management in BE have made substantial contributions to the study of competition in health-care markets, with a focus on the study of provider consolidation;[32] and the FTC’s longstanding, multi-pronged investigation of provider consolidation represents a signal model of the application of applied-industrial-organization research to policy development and law enforcement.[33]

Many—including current leadership at the antitrust agencies,[34] among others[35]—have recognized that BE research, specifically, has had a significant impact on the courts’ treatment of provider mergers. Between 1993 and 2000, antitrust enforcers challenged eight hospital mergers, losing all eight challenges.[36] Hospital-merger challenges waned, and might have been abandoned, but the losing streak spurred renewed research efforts, both within the bureau and across the academy.[37] Critically, BE staff undertook a series of merger-retrospective studies, with then-FTC Chairman Timothy Muris initiating a program of merger-review studies that built on, for example, Vita & Sacher’s 2001 study, “The Competitive Effects of Not-for-Profit Hospital Mergers: A Case Study.”[38]

Subsequent provider-merger-retrospective studies have ranged from individual case studies to reviews of dozens of consummated provider mergers.[39] These are, in essence, forensic investigations, aiming “to determine ex post how, if at all, a particular merger affected equilibrium behavior in one or more markets.”[40] The retrospectives have helped to refine merger-screening methods employed within the FTC; and they have been widely credited with reversing the way that hospital mergers are viewed in the courts.[41] As Michael Salinger observes in a recent article in the Review of Industrial Organization, the retrospective studies grounded testimony in, e.g.:

the FTC’s successful challenge to Evanston-Northwestern Healthcare’s acquisition of the Highland Park Hospital . . . and the empirical methods the Bureau of Economics developed (in conjunction with noted academic health care economists) were essential to subsequent success of the Agencies in challenging hospital mergers.[42]

Especially important to litigation challenges were results on the price effects of not-for-profit hospital mergers (which some courts had supposed were generally benign) and on methods of geographic-market definition (where some courts had been inclined toward very broad geographic markets).[43] Subsequent provider retrospectives have extended the scope of the body of work, considering, e.g., nonprice effects,[44] and merger-screening methods more broadly.[45] Indeed, subsequent studies have not been confined to hospital mergers, but have examined, for example, mergers of physician practice groups[46] and the acquisition of physician practices by hospitals.[47]

Of course, retrospective studies of provider mergers at the enforcement margin have limitations, as well as advantages.[48] Critically, the retrospective studies are not conducted or considered in isolation; rather, they complement methodologically diverse studies of hospital mergers and other forms of provider consolidation, including observational studies based on panel data and cross-sectional data,[49] event studies,[50] and theoretical work.[51] Research has also examined the interaction between providers and third-party payors, as it shapes the nature of competition in health-care-provider markets,[52] as well as vertical[53] and cross-market acquisitions.[54]

Several of the annual review papers published by BE (first, by the FTC, and subsequently by the Review of Industrial Organization) provide brief reviews and, importantly, sketch the application of the academic research to provider merger reviews.[55] Learning from the body of research has, in turn, informed investigations of transactions involving, e.g., outpatient kidney-dialysis centers and specialty surgical centers, as well as physician and hospital mergers.[56]

B. There Is No Sound Basis for New Substantive Competition Regulations Regarding Health-Care Acquisitions, and the RFI Seems Unlikely to Provide One

The agencies state that the RFI will inform, inter alia, “new regulations aimed at promoting and protecting competition in health care markets.”[57] Absent a notice of proposed rulemaking (NPRM), it is unclear what is being contemplated, and correspondingly unclear how the RFI might lead to an NPRM from any of the three agencies. Certainly, FTC and HHS already enforce consumer-protection regulations, issued under express congressional charges, that may have procompetitive effects.[58] These include, for example, the FTC’s Contact Lens Rule (CLR),[59] implementing the Fairness to Contact Lens Consumers Act,[60] and FDA regulations regarding over-the-counter (OTC) hearing aids,[61] implementing certain provisions of the FDA Reauthorization Act of 2017 (FDARA).[62] We would welcome reporting from the FTC on the question of whether it has brought any cases to enforce the central prescription-release provision of the CLR, initially adopted in 2004. More broadly, study of the competitive effects of these regulations may be salutary, to the extent that it might inform proposals to amend the rules. Still, we recognize that enforcement of these regulations is a proper part of the congressional charges to the agencies, and we do not propose changes to either rule.

The prospect of new competition regulations seems, at best, premature. First, the agencies may lack the authority to promulgate such competition rules. The question of whether Congress has granted the FTC substantive or “legislative” competition-rulemaking authority is contentious;[63] and we are aware of no legal basis on which the DOJ could adopt substantive competition regulations. We also are unaware of any amenable statutory charge to HHS. Certainly, HHS can and should consider competitive effects when implementing health-care statutes, but statutory charges for health-care regulations to HHS tend to be specific ones—as was the charge to promulgate regulations for OTC hearing aids noted in the preceding paragraph—and not commonly related to merger scrutiny.

Cognizant of the agencies’ substantial enforcement experience and a significant body of academic literature regarding health-care consolidation, it is difficult to imagine how submissions to the RFI could establish the prerequisites to competition rulemaking regarding health-care acquisitions, even if FTC were deemed to have the requisite rulemaking authority. At present, the agencies do not enforce any such health-care regulations and, to the best of our knowledge, none of the agencies has ever adopted a rule regarding health-care acquisitions under a general grant of legislative rulemaking authority. Specific health-care acquisitions, whether proposed or consummated, can, of course, be blocked, if found anticompetitive, under an administrative ruling, by the Federal Trade Commission, or a judicial ruling, by a federal court. To the best of our knowledge, such decisions have always been case-specific.

Contemporary antitrust law reserves broad rule-like prohibitions for a very limited number of “naked” restraints on trade, such as horizontal price-fixing. For more than 40 years, the U.S. Supreme Court has been clear that general, per se, prohibitions are reserved for the types of matters that “always or almost always tend to restrict competition and decrease output.”[64] None of the types of acquisitions listed in the RFI can demonstrably meet that standard and, absent an express statutory charge from Congress, there is no evident ground for regulating categories of health-care acquisitions under a lesser standard.

Again, we do not—and cannot—impugn ex ante competition concerns that may be raised by specific health-care acquisitions. But, for example, a given study suggesting that certain private-equity acquisitions of hospitals are associated with poorer quality in-patient care, at least on certain measures (chiefly, falls and central-line infections for Medicare beneficiaries)[65] may indeed inform merger scrutiny, but such average effects from a single noncausal study,[66] driven by select effects in a select patient population, cannot suffice to establish that such acquisitions are anticompetitive on net, on average, much less that they “always or almost always tend to restrict competition and reduce output.”

Of course, by noting the potential for new regulations, the agencies might contemplate not only—or even primarily—the promulgation of regulations sui generis, but research and advocacy reported to lawmakers that could inform subsequent and specific statutory charges for regulations.[67] Such research and advocacy can indeed have a salutary effect on policy, although, again, we caution that the present RFI seems unlikely to lead to well-founded policy recommendations, even if it does advance agency learning at the margin.

C. Vertical Transactions Are Not Generally Anticompetitive

The RFI raises broad questions about vertical acquisitions, both in questioning the impact of “[t]ransactions conducted by private equity funds or other alternative asset managers,”[68] (some of which might be considered conglomerate mergers) and in questioning the impact of “[t]ransactions conducted by health systems.”[69]

There is no doubt that vertical mergers can be anticompetitive under certain circumstances. For example, an integrated firm may have an incentive to exclude rivals,[70] and a vertical merger can have an anticompetitive effect if the upstream firm has market power and the ability, post-acquisition, to foreclose its competitors’ access to a key input.[71] In that regard, raising rivals’ costs can “represent[] a credible theory of economic harm” if other conditions of exclusionary conduct are met.”[72] But the implications of vertical mergers are theoretically ambiguous: anticompetitive effects are possible, but they are neither necessary nor, for that matter, typical: “The circumstances… in which [raising rivals’ costs] can occur are usually so limited that [it] almost always represents a minimal threat to competition.”[73] Moreover:

[a] major difficulty in relying principally on theory to guide vertical enforcement policy is that the conditions necessary for vertical restraints to harm welfare generally are the same conditions under which the practices increase consumer welfare.[74]

This structural ambiguity weighs against any presumption against vertical mergers, and it suggests the importance of empirical research in formulating standards to evaluate vertical transactions.

The economics literature is, to borrow a phrase from Leegin, “replete with procompetitive justifications” for vertical integration. Vertical integration typically confers benefits, such as eliminating double marginalization,[75] increasing R&D investment,[76] and creating operational and transactional efficiencies.[77]

Empirical evidence further supports the established legal distinctions between horizontal mergers and vertical mergers (as well as other forms of vertical integration), indicating that vertical integration tends to be procompetitive or benign. For example, a meta-analysis of more than 70 studies of vertical transactions analyzed groups of studies for their implications for various theories or models of vertical integration, and for the effects of vertical integration. From that analysis:

a fairly clear empirical picture emerges. The data appear to be telling us that efficiency considerations overwhelm anticompetitive motives in most contexts. Furthermore, even when we limit attention to natural monopolies or tight oligopolies, the evidence of anticompetitive harm is not strong.[78]

On the contrary, “under most circumstances, profit-maximizing vertical integration decisions are efficient, not just from the firms’ but also from the consumers’ points of view.”[79] And “[a]lthough there are isolated studies that contradict this claim, the vast majority support it….”[80] Lafontaine & Slade accordingly concluded that “faced with a vertical arrangement, the burden of evidence should be placed on competition authorities to demonstrate that that arrangement is harmful before the practice is attacked.”[81] Another study of vertical restraints finds that, “[e]mpirically, vertical restraints appear to reduce price and/or increase output. Thus, absent a good natural experiment to evaluate a particular restraint’s effect, an optimal policy places a heavy burden on plaintiffs to show that a restraint is anticompetitive.”[82]

Subsequent research has reinforced these findings. Reviewing the more recent literature from 2009-18, John Yun concluded “the weight of the empirical evidence continues to support the proposition that vertical mergers are less likely to generate competitive concerns than horizontal ones.”[83]

Leading contributors to the empirical literature, reviewing both new studies and critiques of the established view of vertical mergers, maintain a consistent view. For example, testifying at a 2018 FTC hearing, former FTC Bureau of Economics Director Francine Lafontaine acknowledged that some of the early empirical evidence is less than ideal, in terms of data and methods, but reinforced the overall conclusions of her earlier research “that the empirical literature reveals consistent evidence of efficiencies associated with the use of vertical restraints (when chosen by market participants) and, similarly, with vertical integration decisions.”[84]

The empirical literature regarding vertical acquisitions involving health-care providers, specifically, remains unclear.[85] One study of hospital acquisitions of large physician groups, employing Medicare claims data, finds significant changes at the physician level, with acquired physicians delivering substantially more care in the acquirers’ hospitals post-acquisition (and less at other hospitals and via office-based care).[86] It also finds increased billing at the hospital level, although observed hospital-level effects are smaller, and estimates less precise, than those at the physician level.[87] Here, increased costs—at least for these acquisitions on these measures—appear to be “consistent with the hypothesis that acquired physicians are responding to CMS’ location-based billing policy, which provides higher compensation for care delivered in hospital settings relative to doctors’ offices.”[88] Another study fails to find systematic clinical benefits to vertical integration across diverse quality-of-care metrics.[89]

Such studies may tend to impugn the notion that vertical acquisitions of physician practices by hospitals tend to provide efficiencies that offset cost or price increases, but they cannot be regarded as comprehensive. Further, they suggest the role that public health-care programs and regulations may play in distorting competitive dynamics for both utilization and costs. That raises the question of where policy reform might best be located, supposing that it is called for.

Finally, such studies do not resolve the larger question of why so many physicians—both individually and through their practice groups—are leaving independent practice for hospital- and system-based employment. While the extant literature can certainly inform provider-merger scrutiny in individual matters, it does not appear to implicate general policy reforms for vertical acquisitions of health-care providers and, indeed, suggests equal concern with the design of federal programs and regulations beyond antitrust.

In short, empirical research confirms that the law properly does not presume that vertical mergers have anticompetitive effects; rather, it requires specific evidence of both harms and efficiencies.

The preceding comments apply a fortiori to conglomerate mergers. Whereas vertical mergers combine firms in the same supply chain, conglomerate mergers combine firms that are neither engaged in head-to-head competition nor operating in the same supply chain. Such mergers thus do not inherently reduce competition in any market. The government has explained that conglomerate mergers can produce many of the same “procompetitive benefits” of vertical mergers if the combined firms’ “production or distribution uses the same assets, inputs, or know-how.”[90] That is so “even if the merged firm will become a more effective competitor or gain [market] share.”[91] The resulting economies of scope can increase consumer welfare.

Conglomerate mergers between large, established firms and smaller innovators also play an important role in fostering innovation—and, thus, product competition—in, for example, the pharmaceutical industry. As the Congressional Budget Office (CBO) explains:

The acquisition of a small company by a larger one can create efficiencies that might increase the combined value of the firms by allowing drug companies of different sizes…to specialize in activities in which they have a comparative advantage. Small companies—with relatively fewer administrative staff, less expertise in conducting clinical trials, and less physical and financial capital to manage—can concentrate primarily on research. For their part, large drug companies are much better capitalized and can more easily finance and manage clinical trials. They also have readier access to markets through established drug distribution networks and relationships with buyers.[92]

Conglomerate mergers in the pharmaceutical industry thus can realize the procompetitive effects of vertical combinations (creating efficiencies) while avoiding the anticompetitive effects of horizontal mergers (eliminating competition).

That is not to say conglomerate mergers can never lead to higher prices. Recent research on bargaining models indicates it is possible for cross-market acquisitions to facilitate a price increase. Such models do not, however, suggest that is a likely result. Instead, empirical research indicates that, generally, “cross-market acquisitions by larger companies do not have a significant effect on price.”[93] Moreover, a common theory of competitive harm holds only “as long as” the parties’ “products have common customers.”[94] Hospital acquisitions may provide a special case, in that they may cross geographic markets even if they do not cross product markets. Moreover, geographic markets may have different boundaries from the perspectives of patients and third-party payors.[95] Put another way, certain provider mergers that may be deemed cross-market transactions from a patient perspective may also alter provider bargaining with health plans for whom the providers are substitutes (or complements); hence, from another perspective, they are within the same geographic market. In that regard, additional research[96] and additional enforcement experience may, in time, lead to further refinements in hospital-merger scrutiny.

Of course, none of this is to say that the agencies should not scrutinize vertical or conglomerate mergers involving health-care providers. Further research might sharpen the agencies’ understanding of specific industries in which, or circumstances under which, provider acquisitions may be more or less likely to raise competitive concerns. Ongoing research by the agencies—including BE staff research[97]—will no doubt further that goal.

But, as we note above, a single study is not a body of literature, much less one that is mature or settled. Indeed, a single study suggesting that certain private-equity acquisitions of hospitals are associated with poorer quality in-patient care, at least on certain measures (chiefly, falls and central-line infections for Medicare beneficiaries)[98] may, indeed, inform merger scrutiny, but such average effects from a single noncausal study, driven by select effects in a select patient population, cannot suffice to establish that such acquisitions are anticompetitive on net, on average, much less that they “always or almost always tend to restrict competition and reduce output.” More plausibly, it may be pertinent to questions of, e.g., when to issue a second request or commence a formal investigation, at least at the margin.

Similarly, future research may sharpen the agencies’ understanding of the conditions under which vertical (or conglomerate) acquisitions by third-party payors are more, or less, likely to raise competition concerns. Such research might be bolstered by additional enforcement experience with such acquisitions, but we expect that the present RFI is not well-designed to further agency understanding much beyond that already available to agency staff.

D.  A Provider’s For-Profit or Not-for-Profit Status May Say Little About the Likely Competitive Effects of Mergers or Acquisitions Involving that Provider

As noted above, we were struck by a statement at the outset of the RFI: “Given recent trends, we are concerned that transactions may generate profits for those firms at the expense of patients’ health, workers’ safety, and affordable health care for patients and taxpayers.”[99] To be sure, some transactions do just that. There is no doubt that the antitrust laws are broadly applicable to health-care transactions or that particular provider mergers, under particular facts and circumstances, may violate the antitrust laws, harming competition and consumer welfare.[100] And nonprice effects, such as quality of care, may factor in antitrust scrutiny of a provider merger.[101]

Nonetheless, the agencies understand that antitrust law and economics do not recognize any general or fundamental tension between firm profits, on the one hand, and the consumer benefits typically associated with price and nonprice competition in goods and services markets, on the other. Moreover, considerable research militates against the suggestion that for-profit and not-for-profit providers should be distinguished for the purposes of merger scrutiny. As Martin Gaynor—former director of the FTC Bureau of Economics and presently special advisor to the assistant U.S. attorney general for antitrust—summarized in testimony before the Senate Judiciary Committee: “Research evidence shows not-for-profit hospitals exploit market power just as much as for-profits.”[102]

In fact, as noted above, two early foci for the FTC’s hospital-merger retrospective studies were, one, the question of how best to approach geographic-market definition (not least, because some courts were inclined toward very broadly drawn hospital markets, at odds with established methods) and, two, the question of whether not-for-profit hospitals were less likely than for-profit hospitals to exploit market power, when they had it, within the relevant geographic boundaries (not least, because some courts were accepting what amounted to a “not-for-profit defense” to hospital-merger challenges).[103] The merger retrospectives consistently demonstrated that not-for-profit status did not make a difference.[104]

Our point is not that the extant literature is definitive or that it is easily generalized across different types of providers. Rather, there are good reasons to think that not-for-profit providers are not special from a competition standpoint, and substantial evidence on that point in a well-investigated provider domain.

Further research and enforcement experience might suggest a different perspective on one or more specific subcategories of provider acquisitions. Still, the agencies should be mindful of the hospital findings as a background matter. And as the agencies’ research staff are likely aware, research regarding the question of whether for-profit provider status in, e.g., hemodialysis treatment is associated with different treatment quality has provided mixed results, with some agency research failing to find any statistically significant indication that it is.[105] Results also are observed to vary across empirical specifications and available datasets.[106] In addition, given the large number of hemodialysis acquisitions nationwide associated with two acquiring firms, there is an open question, even with regard to kidney dialysis, how best to parse for-profit status from the management practices of two very large for-profit acquirers.[107]

E. Various Models of Health-Care Delivery May Be Associated with Complex Tradeoffs

Whereas some of the interests or concerns in the RFI focus on transactions’ structural features—e.g., on the competitive effects of horizontal, vertical, or conglomerate mergers—an overlapping set of questions focuses on the type of acquiring firm. For example, the RFI notes “concerning trends” in, e.g., “transactions in the health care market conducted by private equity funds or other alternative asset managers, health systems, and private payers .”[108] The RFI suggests that there is “recent research indicating these categories of transactions may harm health care quality, access, and/or costs.”[109]

But the suggestion about “recent research” has no attached citation. An earlier footnote substantiates the claim that “[a]cademic research and agency experience in enforcement actions has shown that patients, health care workers, and others may suffer negative consequences as a result of horizontal and vertical consolidation of a range of different types of providers—including not-for-profit providers.”[110] While the agencies cite only two primary research articles and one policy review for that general proposition—and while one of the articles suggests limited results[111]—we take it that the far more general claim is (or should be) uncontroversial. The dozens of papers cited above in Section II.A.2. of these comments tend to substantiate those broad claims. That is, diverse provider acquisitions can raise competitive concerns; and, moreover, competitive concerns can be raised equally by transactions (or other conduct) involving not-for-profit and for-profit providers.

Not incidentally, many provider markets are highly concentrated, pre-acquisition, on any notion of “highly concentrated.”[112] For example, the FTC’s defense of its authority in the Phoebe-Putney matter concerned what was, in effect, a merger to monopoly;[113] and several surrounding counties—like many outside metropolitan areas across the country—had no general hospital at all.[114] No policy reform is needed to provide that two-to-one hospital mergers will be carefully scrutinized by antitrust authorities. Similarly high provider concentration can be observed across diverse specialty practices in many rural and other small markets.[115]

But the research base and enforcement experience regarding specific types of acquiring (or target) firms is considerably less well-developed. We do not mean to impugn specific studies, so much as to place available results in context. For example, as we also discuss above, there are studies suggesting negative health-care-quality effects—cognizable harms—associated with certain for-profit acquisitions of hemodialysis-treatment facilities.[116] But as we note there, results in that space are somewhat mixed, varying across empirical specifications and available data, and there is some research that fails to find any statistically significant indication that acquisitions by for-profit firms are associated with different treatment quality.[117]

Turning to private-equity acquisitions, the RFI cites a single study suggesting that certain private-equity acquisitions of hospitals are associated with poorer quality in-patient care, at least on certain measures (chiefly, falls and central-line infections for Medicare beneficiaries).[118] We cannot gainsay competition concerns about such acquisitions, and the study may, indeed, inform merger scrutiny going forward. Such average effects from a single non-causal study, driven by select effects in a select patient population, cannot, however, suffice to establish that such acquisitions are anticompetitive on net, on average, much less ground a fundamentally different approach to private-equity acquisitions of health-care providers. We note, too, that another study (with two of the same coauthors) found more mixed results, including some suggesting improved quality of care:

In our main analysis, we observed greater improvements in process quality measures among private equity–acquired hospitals relative to controls, which may reflect better care for patients. However, it could also be consistent with better adherence to compliance standards or efforts to maximize opportunities for quality bonuses under pay-for-performance contracts.[119]

Positive income and profitability were also observed. Both studies evidence some heterogeneity of findings across the private-equity and control hospitals. Our point is not that the agencies should be unconcerned about nonprice effects, such as quality of care. Rather, it is that the understanding of this class of transactions is incomplete, and unlikely to be resolved by submissions in response to this RFI. Also, the research does not resolve the fundamental question of when or under what conditions hospitals may be targets for private-equity acquisitions. And to the extent it suggests new management practices, it may suggest not just concerns but tradeoffs in the management capacity associated with different acquirers.[120]

Given mixed results and a lacunae in the literature, further research is warranted, as well as case-specific investigation using established methods. To the extent that specific findings on specific categories of provider mergers are mixed, unclear, or conspicuously limited, more general economic learning and precedent regarding, e.g., horizontal, vertical, and conglomerate mergers may be especially informative. So, too, may be agency experience with undue restraints on the “corporate practice of medicine” or other undue restraints on new models of distribution for health care, dating at least to the FTC’s landmark 1980 case against the American Medical Association, which addressed various restraints on physician and nonphysician contracting.[121] Analogously, in 1992, based on its research regarding the eye-care industry, FTC staff advocated for the repeal of “prohibitions against practicing in retail settings and against corporate affiliations.”[122]

Finally, given results suggesting the confounding effects of health-care programs and regulations, from Medicare reimbursement policies to state-based certificate-of-public-advantage and certificate-of-need regulations, the agencies should be ever alert to the question of the best locus for policy reform.

F. The Framing of a Request for Information Can Influence the Quality of the Response

As we explain above, we appreciate the importance of the agencies’ efforts to protect and foster competition in diverse health-care markets; and we appreciate the mutually reinforcing roles that policy studies and enforcement experience can play in health-care and antitrust policy. Still, one need not gainsay concerns about health-care competition or specific types of acquisitions to appreciate the difficulty of grounded, systematic reform of enforcement policy in these areas. The agencies’ extension of the deadline for submissions in response to the RFI will, no doubt, increase the utility of the inquiry. But while recognizing that the present RFI may be a useful endeavor, it is just one tool—in itself, a limited one—in the agencies “policy R&D”[123] toolbox. Moreover, the RFI’s framing seems, in many ways, unfortunate: not conducive to the most constructive use of agency resources or third-party contributions.

First, the scope of the RFI is unclear. The agencies note that they are:

particularly interested in information on transactions in the health care market conducted by private equity funds or other alternative asset managers, health systems, and private payers, especially those transactions that would not be noticed to the Department of Justice and the Federal Trade Commission under the Hart-Scott-Rodino Antitrust Improvements Act, 15 USC 18(a).[124]

Such transactions may be both numerous and diverse, with or without a restriction on HSR-reportable transactions. The scope and heterogeneity of agency interests is only underscored by the RFI’s elaboration on the transactions of interest:

These transactions could involve dialysis clinics, nursing homes, hospice providers, primary care providers, hospitals, home health agencies, home- and community-based services providers, behavioral health providers, billing and collections services, revenue cycle management services, support for value-based care, data/analytics services, and other types of health care payers, providers, facilities, Pharmacy Benefit Managers (PBMs), Group Purchasing Organizations (GPOs), or ancillary products or services.[125]

If the RFI is meant to be but one inquiry in a much larger project—say, for example, something akin to the 2003 FTC/DOJ health-care hearings that led to the 2024 Dose of Competition report—some sense of the scope of the larger project would be helpful to the public. On its own, the reference to acquisitions across such diverse health-care entities seems extremely broad, and not well-suited to the production of usefully focused submissions.

Whatever the scope of the RFI, its framing is critical to its utility. Given the agencies’ considerable contributions to health-care competition over the course of several decades,[126] we regret to comment on a conspicuous deficit in the RFI’s framing. As we note in our introductory summary, the agencies’ RFI seems, at times, to prejudge the answers to its own questions. That may be unproductive for research purposes and, specifically, may bias submissions to the public record.

The most egregious example of this may be the FTC’s press release announcing the RFI, which informed both the press and stakeholders that the FTC, DOJ, and HHS “Launch Cross-Government Inquiry on Impact of Corporate Greed in Health Care.”[127] That framing would seem overly dramatic if it announced allegations of antitrust violations; it seems an especially poor way to announce a request for information from the diverse stakeholders constituting “the public.”

Of course, a press release is just that, but the language is repeated in the FTC’s May 1 announcement that the agencies had extended the RFI comment period;[128] and at least some readers may have noticed that the language in the FTC’s press release mirrors that of a White House “fact sheet” noting that the administration was “[l]aunching a cross-government public inquiry into corporate greed in health care.”[129] Some individuals may be “greedy,” in a colloquial sense, whether in their personal capacities or acting as corporate agents. But “corporate greed” has no clear meaning in antitrust law or industrial-organization economics. It is hardly a subject for systematic investigation by expert agencies; it seems, at best, an atmospheric distraction.

Announcements of the RFI from DOJ and HHS seem similarly, if less steeply, slanted, describing a “cross-government public inquiry into private-equity and other corporations’ increasing control over health care.”[130] Identifying legitimate competition concerns is not, in itself, problematic. But suggesting such concerns about broad categories of transactions, without any acknowledgment of potential merger benefits, and without any acknowledgment that most provider mergers and acquisitions are not challenged, much less blocked, and should be presumed lawful until established otherwise, seems to suggest a general hostility to provider acquisitions with no basis in legal precedent, economic research, or agency practice. Similarly, any suggestion that profits in highly differentiated product and service markets are inconsistent with the fruits of vigorous health-care competition—lower prices, higher quality, and greater availability of health care—would appear fundamentally at odds with both established antitrust law and economic learning.

Similarly, as we note above, we were struck by a statement at the outset of the RFI itself: “Given recent trends, we are concerned that transactions may generate profits for those firms at the expense of patients’ health, workers’ safety, and affordable health care for patients and taxpayers.”[131] To be sure, some transactions do just that. But, as we discuss at some length above,[132] antitrust law and economics do not recognize any general or fundamental tension between firm profits, on the one hand, and the consumer benefits typically associated with price and nonprice competition in goods and services markets, on the other.

And while there is some research suggesting that some categories of for-profit provider acquisitions may be associated with competitive harms, at least in some circumstances, a considerable body of research, reinforced by agency-enforcement experience, militates against the suggestion that for-profit and not-for-profit providers should be distinguished for the purposes of merger scrutiny. As Martin Gaynor—former director of the FTC Bureau of Economics and presently special advisor to the assistant U.S. attorney general for antitrust—summarized in testimony before the Senate Judiciary Committee: “Research evidence shows not-for-profit hospitals exploit market power just as much as for-profits.”[133]

We wonder, too, about the design of a March 5, 2024, workshop titled “Private Capital, Public Impact: An FTC Workshop on Private Equity in Health Care,”[134] which was hosted by the FTC with leadership from the DOJ and HHS also participating. The workshop was timed to coincide with the RFI and, not incidentally, was noted identically by all three agencies in their press releases for the RFI.[135] The posted agenda specifies a brief event—less than half a day—at which roughly half of the participants represented the agencies themselves, and none obviously worked in or for the industry in question or, e.g., for large health plans or other private payors. Several participants were individual practitioners relating their own perceptions of specific acquisitions.[136]

To be sure, providers and other stakeholders might well be interested in the perspectives of agency officials. At the same time, the airing of agency views and concerns seems ill-timed, given the timing of the RFI itself, as submissions in response to the RFI were initially due just one day after the workshop.

Moreover, the FTC’s announcement of its workshop echoes the apparent imbalance of the RFI itself:

In recent years, the Commission has become increasingly concerned about the effects of private equity investment in this sector. We are convening a workshop bringing together experts and affected individuals to discuss their insights. The workshop will consist of several panels and feature remarks from government officials, academics, economists, and practitioners, as well as members of the public who have experienced, first-hand, the effects of private equity investment in the health care system.[137]

Again, we do not take any issue with the identification of legitimate competition concerns. Merger scrutiny is the proper purview and, indeed, obligation of the antitrust agencies; and we do not write to opine on open matters or potential acquisitions. But the workshop design, description, and timing suggest an information-gathering exercise distinct from an open-minded public inquiry, if not the prejudgment of myriad fact-dependent potential enforcement matters.

III. Conclusion

Health-care-provider consolidation is an important area of concern for antitrust enforcers, and there is no doubt that specific provider acquisitions can prove anticompetitive. For those reasons, the RFI may indeed prompt the submission of useful materials to the antitrust agencies and, perhaps, to HHS. To the extent that the RFI is considered but one more step in the agencies’ ongoing competition R&D program, it may be salutary. At the same time, the RFI does not seem designed to move agency learning much beyond the margin—certainly not across the broad swath of issues it raises; and the RFI’s framing seems likely to skew, rather than focus, the information submitted.

Further, while competition concerns may be important to how the agencies implement various congressional charges to promulgate specific regulations (and, by statute, are implicated in any FTC rulemaking regarding unfair or deceptive acts or practices), neither enforcement experience nor economic literature militate in favor of new competition regulations regarding provider mergers and acquisitions.

While there may be ample reasons for diverse competitive concerns, such concerns do not establish categories of acquisitions that warrant per se condemnation, via regulation or otherwise. To the contrary, agency experience and expertise with, e.g., restraints on the “corporate practice of medicine” and with other regulatory restraints on diverse methods or models of health-care delivery illustrate the competitive (and welfare) tradeoffs implicated by many types of provider acquisitions and, indeed, by specific transactions. Such tradeoffs can have—and have had—directionally different competition implications on a case-by-case basis.

More specifically, while extant research and enforcement experience may identify or heighten competitive concerns about certain transactions, they militate against, rather than for, new policies regarding for-profit providers, overly simple structural approaches to health-care-merger screening, and the conflation of considerations for horizontal, vertical, and conglomerate acquisitions.

Emerging concerns may prompt reallocation of screening resources and priorities within the agencies, although the importance of building experience cumulatively may suggest caution there, too.

As a related matter, concerns about provider acquisitions—single transactions or clusters of them—below the HSR reporting threshold may be justified in many markets, especially in rural or other underserved areas. That suggests a complex of inquiries, however, and not new rules or general policies. Given the myriad factors driving consolidation—especially in small (and, often, shrinking) markets—and given the fact that the large majority of mergers, above or below the threshold, are not anticompetitive, how can further research and enforcement experience identify filters by which the agencies might identify and screen those sub-threshold acquisitions most likely to raise competitive concerns?

Finally, as we suggest in the introduction to these comments, further policy inquiries—from RFIs to workshops to systematic research—might best be served by agency economists conducting a serious critical synthesis of the extant body of research regarding health-care-provider acquisitions. That is a nontrivial project, but it should be prologue to consideration of or recommendations regarding policy reforms in the area.


[1] Request for Information on Consolidation in Health Care Markets, Docket No. ATR-102, Dep’t Justice, Dep’t Health & Human Servs., & Fed. Trade Comm’n (Mar. 5, 2024), [hereinafter “RFI”].

[2] Daniel J. Gilman & Tara Isa Koslov, Policy Perspectives: Competition and the Regulation of Advanced Practice Nurses, Fed. Trade Comm’n, 1 (Mar. 2014), available at

[3] For an overview, see, e.g., Hearing on Antitrust Applied: Hospital Consolidation Concerns and Solutions, Testimony before the Subcomm. on Competition Pol’y, Antitrust, and Consumer Rights, S. Comm. on the Judiciary, 117th Cong. (2021) (statement of Martin Gaynor, E.J. Barone University Professor of Economics and Public Policy Heinz College, Carnegie Mellon University),

[4] For successful cases against provider mergers, see, e.g., Fed. Trade Comm’n v. Hackensack Meridian Health, Inc., 30 F.4th 160 (3d Cir. 2022); ProMedica Health Sys., Inc., FTC Docket No. 9346, 2012 WL 2450574 (Jun. 25, 2012), aff’d, ProMedica Health Sys., Inc., v. FTC, 749 F.3d 559 (6th Cir. 2014); FTC v. Penn State Hershey Med. Ctr., 185 F. Supp. 3d 552 (M.D. Pa. 2016), rev’d, 838 F.3d 327, 343 (3d Cir. 2016); Saint Alphonsus Med. Ctr.-Nampa Inc. v. St. Luke’s Health Sys., Ltd., No. 1:12-cv-00560, 2014 WL 407446 (D. Idaho Jan. 24, 2014), aff’d, 778 F.3d 775 (9th Cir. 2015). Regarding the authority to review provider mergers, see Fed. Trade Comm’n v. Phoebe Putney Health Sys., Inc., 586 U.S. 216 (2013) (acquisition not immune from scrutiny under state-action doctrine).

[5] We refer to the Sherman and Clayton Acts and, by extension, the Federal Trade Commission Act, and recognize that other parties, including state attorneys general and private parties, may sue to enforce certain provisions of the antitrust laws, while recognizing that there is no private right of action under the FTC Act.

[6] 15 U.S.C. § 46. See infra., text accompanying notes 23-31, for constructive examples.

[7] Id. at § 46(l).

[8] Links to economic research, including reports, working papers, issue papers, and articles in peer-reviewed journals can be found at Fed. Trade Comm’n, Bureau of Econ., Research in the Bureau of Economics, See also infra. Section II.A.2.

[9] RFI at 11.

[10] Sub-threshold acquisitions may well be 3-to-2 or merger-to-monopoly transactions for critical services. “Any standard” would include, for example, those described in any or all editions of the horizontal merger guidelines.

[11] We note that, e.g., in January 2021, the FTC issued orders, under its FTC Act Section 6(b) authority to six health-insurance companies to provide information to facilitate the agency’s study of the effects of physician group and health-care-facility consolidation from 2015 through 2020. See Press Release, FTC to Study the Impact of Physician Group and Healthcare Facility Mergers, Fed. Trade Comm’n (Jan. 14, 2021), While the information collected under such orders is limited partly by restrictions imposed under the Paperwork Reduction Act, and not merely available data and methodological concerns, it may nonetheless help advance understanding of provider consolidation. We assume that this project, initiated at the tail end of the last administration, is ongoing.

[12] AMA v FTC, 638 F.2d 443 (2d Cir. 1980).

[13] Id. at 1.

[14] See generally, e.g., Fed. Trade Comm’n v. Phoebe Putney Health Sys., Inc., 568 U.S. 216 (2013). In its unanimous decision, the Court noted that the 11th U.S. Circuit Court of Appeals had, as an initial matter, “‘agreed with the [FTC] that, on the facts alleged, the joint operation of Memorial and Palmyra would substantially lessen competition or tend to create, if not create, a monopoly’” 568 U.S. at 222-3 (quoting Fed. Trade Comm’n v. Phoebe Putney Health Sys., Inc., 663 F.3d 1369, 1375 (2011). The Court’s holding in Phoebe Putney upheld the FTC’s jurisdiction over the hospital merger, notwithstanding the grant of certain powers to hospital authorities by the state of Georgia. 568 U.S. at 224. For a discussion of various FTC research, advocacy, and enforcement activities in health care, including scrutiny of provider mergers, see, e.g., Maureen K. Ohlhausen, The First Wealth is Health: Protecting Competition in Healthcare Markets, Remarks at the 2017 ABA Fall Forum (Nov. 16, 2017), available at Although the FTC and the DOJ have concurrent jurisdiction over mergers under Section 7 of the Clayton Act, health-care-provider mergers are typically assigned to the FTC under the FTC/DOJ clearance process. For a list of health-care-enforcement matters, see FTC, The FTC’s Health Care Work: Cases, (last accessed May 1, 2024).

[15] See infra. Section II.D.

[16] Private Capital, Public Impact: An FTC Workshop on Private Equity in Health Care, Fed. Trade Comm’n (May 5, 2024), The workshop webpage includes a description, along with links to the agenda, participant biographies, and a transcript of the proceedings.

[17] Press Release, Federal Trade Commission, the Department of Justice and the Department of Health and Human Services Launch Cross-Government Inquiry on Impact of Corporate Greed in Health Care, Fed. Trade Comm’n (Mar. 5, 2024), (noting that, “[i]n addition to the launch of the RFI, all three agencies will also be participating today in a virtual public workshop that will explore the impact of private equity in health care and will discuss what the government is doing to address any harmful effects.”). The announcement of the FTC workshop was repeated verbatim in DOJ and HHS announcements of the RFI. Press Release, Justice Department, Federal Trade Commission and Department of Health and Human Services Issue Request for Public Input as Part of Inquiry into Impacts of Corporate Ownership Trend in Health Care, Dep’t Justice (Mar. 5, 2024),; Press Release, HHS, DOJ, and FTC Issue Request for Public Input as Part of Inquiry into Impacts of Corporate Ownership Trend in Health Care, Dep’t Health & Human Servs. (Mar. 5, 2024),

[18] William e. Kovacic, The Federal Trade Commission at 100: Into Our Second Century, 91-92 (Jan. 2009), available at

[19] Public Hearings: Health Care and Competition Law and Policy Hearings, Fed. Trade Comm’n & Dep’t Justice (2023), (hearings page with links to agendas and transcripts); Hearings on Competition and Consumer Protection in the 21st Century, Fed. Trade Comm’n (2018-19), (hearings page with links to agendas, transcripts, and submissions).

[20] See, e.g., Now Hear This: Competition, Innovation, and Consumer Protection Issues in Hearing Health Care, Fed. Trade Comm’n (Apr. 18, 2017), (FTC Workshop); Examining Health Care Competition, Fed. Trade Comm’n (Mar. 2014), (FTC Workshop); Innovations in Health Care Delivery, Fed. Trade Comm’n (Apr. 24, 2008), (FTC Workshop).

[21] See, e.g., 16th Annual Microeconomics Conf., Fed. Trade Comm’n (Nov. 2023), (annual conference hosted by FTC Bureau of Economics; 2023 conference was cosponsored by FTC and Tobin Ctr., Yale Univ.).

[22] See, e.g., RFI; FTC Seeks Comment on Contact Lens Rule Review, 16 CFR Part 315, Fed. Trade Comm’n (May 28, 2019),

[23] Public Hearings: Health Care and Competition Law and Policy, Dep’t Justice (last updated Aug. 21, 2023), (describing hearings jointly conducted by DOJ and FTC, and providing links to agendas and transcripts for individual hearings, submissions to the public record, and various supporting materials).

[24] Fed. Trade Comm’n & U.S. Dep’t of Justice (“DOJ”), Improving Health Care: A Dose of Competition (2004), available at

[25] Id. at exec. summ., 20-29.

[26] Innovations in Health Care Delivery (workshop), Fed. Trade Comm’n (Apr. 24, 2008),

[27] The workshop agenda is available at

[28] Examining Health Care Competition (workshop), Fed. Trade Comm’n (Mar. 2014),

[29] Examining Health Care Competition, Fed. Trade Comm’n & Dep’t Justice (Feb. 2015),

[30] Another Dose of Competition: Accountable Care Organizations and Antitrust, FTC Workshop (May 2011),

[31] See, e.g., Statement of Antitrust Enforcement Policy Regarding Accountable Care Organizations Participating in the Medicare Shared Savings Program, 76 Fed. Reg. 67026 (Oct. 28, 2011), (Final Policy Statement, Fed. Trade Comm’n and Dep’t Justice Antitrust Div.).

[32] See, e.g., Devesh Raval et al., Using Disaster Induced Closures to Evaluate Discrete Choice Models of Hospital Demand, 53 RAND J. Econ. 561 (2022); Thomas Koch & Shawn W. Ulrick, Price Effects of a Merger: Evidence from a Physicians’ Market, 59 Econ. Inquiry 790 (2021); Keith Brand & Ted Rosenbaum, A Review of the Economic Literature on Cross-Market Healthcare Mergers, 82 Antitrust L.J. 533 (2019); Thomas Koch et al., Physician Market Structure, Patient Outcomes, and Spending: An Examination of Medicare Beneficiaries, 53 Health Servs. Res. 3549 (2018); Thomas G. Koch, Brett W. Wendling, & Nathan E. Wilson, How Vertical Integration Affects the Quantity and Cost of Care for Medicare Beneficiaries, 52 J. Health Econ. 19 (2017); Julie A. Carlson et al., Economics at the FTC: Physician Acquisitions, Standard Essential Patents, and Accuracy of Credit Reporting, 43 Rev. Indus. Org. 303 (2013); See also, e.g., Martin Gaynor & Robert J. Town, The Impact of Hospital Consolidation—Update, Robert Wood Johnson Foundation, The Synthesis Project (2012) (Gaynor is a former director of the FTC’s Bureau of Economics who serves presently as a special advisor to the assistant U.S. attorney general for antitrust); Martin Gaynor & William B. Vogt, Competition Among Hospitals, 34 RAND J. Econ. 764 (2003); Maximillian J. Pany, Michael E. Chernew, & Leemore S. Dafny, Regulating Hospital Prices Based on Market Concentration Is Likely to Leave High-Price Hospitals Unaffected, 40 Health Aff. 1386 (Sept. 2021) (Dafny was deputy director for health care antitrust in the FTC’s Bureau of Economics from 2012-13); Leemore S. Dafny, Hospital Industry Consolidation—Still More to Come?, 370 New Eng. J. Med. 198 (2014).

[33] We focus here on research associated with the FTC’s Bureau of Economics, which comprises a significant body of pertinent research. We recognize, of course, that diverse empirical research from DOJ economists and, indeed, various HHS agencies, may be pertinent to provide antitrust scrutiny as well. Stepping back, the larger and still-developing body of academic literature regarding health-care competition is considerable and complex. We do not attempt to review it here.

[34] See, e.g., Letter from Lina Khan, Chair, Fed. Trade Comm’n & Jonathan Kanter, Asst. Atty. General, Antitrust Div., Dept. Justice to the Hon. François-Philippe Champagne, Canada Ministry Innovation, Sci. & Indus. (Mar. 31, 2023), See also id. at 3, n. 11 and 9, n. 40 (highlighting specific hospital-merger retrospective studies and merger retrospectives more generally).

[35] See, e.g., Michael A. Salinger, The 2023 Merger Guidelines and the Role or Economics, Rev. Indus. Org. (May 3, 2024),; see also, Prepared Opening Remarks of Chairman Joseph J. Simons, Hearings on Competition and Consumer Protection in the 21st Century, Merger Retrospectives, Fed. Trade Comm’n (Apr. 12, 2019), available at Numerous injunctions obtained by the FTC in provider matters since commencement of the hospital-merger retrospective study program can be found at

[36] See, e.g., Thomas L. Greaney, Whither Antitrust? The Uncertain Future of Competition Law in Health Care, 21 Health Affs. 185 (2002); Christopher Garmon, Hospital Mergers—Retrospective Studies to Improve Prediction, CPI Antitrust Chronicle (Jul. 2017).

[37] Orly Ashenfelter, Daniel Hosken, Michael Vita, & Matthew Wienberg, Retrospective Analysis of Hospital Mergers, 18 Int. J. Econ. & Bus. 5, 6 (2011).

[38] Michael G. Vita & Seth Sacher, The Competitive Effects of Not?For?Profit Hospital Mergers: A Case Study, 49 J. Indus. Econ. 63 (2001).

[39] See, e.g., Christopher Garmon & Laura Kmitch, Hospital Mergers and Antitrust Immunity: The Acquisition of Palmyra Medical Center by Phoebe Putney Health, 14 J. Comp. L. & Econ. 433 (2018); Christopher Garmon, The Accuracy of Hospital Screening Methods, 48 RAND J. Econ. 1068 (2017) (reviewing post-merger price changes for 28 hospital mergers, initially published as BE Working Paper); Deborah Haas?Wilson & Christopher Garmon, Hospital Mergers and Competitive Effects: Two Retrospective Analyses, 18 Int. J. Econ. Bus. 17 (2011); Steven Tenn, The Price Effects of Hospital Mergers: A Case Study of the Sutter–Summit Transaction, 18 Int. J. Econ. Bus. 65 (2011) (originally published as BE Working Paper); Aileen Thompson, The Effect of Hospital Mergers on Inpatient Prices: A Case Study of the New Hanover-Cape Fear Transaction, 18 Int. J. Econ. Bus. 91 (2011) (originally published as BE Working Paper); Ashenfelter et al., supra note 37; Patrick S. Romano & David J. Balan, A Retrospective Analysis of the Clinical Quality Effects of the Acquisition of Highland Park Hospital by Evanston Northwestern Healthcare, 18 Int. J. Econ. Bus. 45 (2010); John Simpson, Geographic Markets in Hospital Mergers: A Case Study, 10 Int. J. Econ. Bus. 291 (2003); Vita & Sacher, supra note 38. A bibliography of merger-retrospective studies compiled by the Bureau of Economics comprises more than 30 provider-merger retrospectives, with contributors from within and without BE. Those, in turn, inform and are informed by the larger body of research regarding health-care merger retrospectives. Fed. Trade Comm’n, Merger Retrospectives Bibliography, (last visited May 10, 2024).

[40] Joseph Farrell, Paul Pautler, & Michael Vita, Economics at the FTC: Retrospective Merger Analysis with a Focus on Hospitals, 35 Rev. Indus. Org. 369 (2009).

[41] See, Overview of the Merger Retrospective Program in the Bureau of Economics, Fed. Trade Comm’n (last visited Apr. 12, 2023),; see also, Simons, supra note 35; Khan & Kanter, supra note 34.

[42] Salinger, supra note 35, at note 10.

[43] Ashenfelter et al., supra note 37, at 6-7.

[44] See, e.g., Romano & Balan, supra note 39 (regarding impact on clinical quality).

[45] See, e.g., Hass-Wilson & Garmon, supra note 39.

[46] Koch & Ulrick, supra note 32.

[47] See, e.g., Thomas G. Koch, Brett W. Wendling, & Nathan E. Wilson, The Effects of Physician and Hospital Integration on Medicare Beneficiaries’ Health Outcomes, 103 Rev. Econ. & Stats. 725 (2021) (initially published as BE Working Paper).

[48] See, e.g., Dennis W. Carlton, Why We Need to Measure the Effect of Merger Policy and How to Do It, 5 Comp. Pol’y Int. 77 (2009); Ashenfelter et al., supra note 37; Farrell, Pautler, & Vita, supra note 40.

[49] Matthew Panhans, Ted Rosenbaum, & Nathan E. Wilson, Prices for Medical Services Vary Within Hospitals, But Vary More Across Them, 78 Med. Care Res. Rev. 157 (2021, initially published as BE Working Paper); Koch, Wendling, & Wilson, Medicare Beneficiaries’ Health Outcomes, supra note 47 (initially published as BE Working Paper); Asako S. Moriya, William B. Vogt, & Martin Gaynor, Hospital Prices and Market Structure in the Hospital and Insurance Industry, 5 Health Econ, Pol’y & Law 1 (2010) (Martin Gaynor is a former director of the FTC’s Bureau of Economics presently serving as a special advisor to the assistant U.S. attorney general for antitrust); Martin Gaynor & William B. Vogt, Competition Among Hospitals, 34 RAND J. Econ. 764 (2003); Maximillian J. Pany, Michael E. Chernew, & Leemore S. Dafny, Regulating Hospital Prices Based on Market Concentration Is Likely to Leave High-Price Hospitals Unaffected, 40 Health Aff. 1386 (September 2021) (Dafny was deputy director for health care antitrust in the FTC’s Bureau of Economics from 2012-13); Leemore S. Dafny, Hospital Industry Consolidation—Still More to Come?, 370 New Eng. J. Med. 198 (2014); Devesh Raval et al., Using Disaster Induced Closures to Evaluate Discrete Choice Models of Hospital Demand, 53 RAND J. Econ. 561 (2022) (initially published as BE Working Paper); Nathan E. Wilson, Market Structure as a Determinant of Patient Care Quality, 2 Amer. J. Health Econ. 241 (2016) (studying hemodialysis care) (initially published as BE Working Paper).

[50] See, e.g., Devesh Raval, Ted Rosenbaum, & Nathan E. Wilson, Using Disaster-Induced Closures to Evaluate Discrete Choice Models of Hospital Demand, 53 RAND J. Econ. 561 (2022). Event studies are, of course, also observational studies, even if they—and merger retrospectives—may be considered in some regards “quasi-experimental.”

[51] David J. Balan & Keith Brand, Simulating Hospital Merger Simulations, 71 J. Indus. Econ. 47 (2023) (initially published as BE Working Paper); see also Leemore Dafny, Katherine Ho, & Robin Lee, The Price Effects of Cross-Market Mergers: Theory and Evidence from the Hospital Industry, 50 RAND J. Econ. 286 (2019) (theoretical analysis with empirical extension).

[52] See, e.g., Carlson et al., supra note 32 (citing Cory Capps, David Dranove, & Mark Satterthwaite, Competition and Market Power in Option Demand Markets, 34 RAND J. Econ. 737 (2003); Robert Town & Gregory Vistnes, Hospital Competition in HMO Networks, 20 J. Health Econ. 753 (2001)).

[53] Koch, Wendling, & Wilson, Quantity and Spending, supra note 32; Koch, Wendling, & Wilson, Health Outcomes, supra note 47.

[54] Dafny, Ho, & Lee, supra note 51; see also, Keith Brand & Ted Rosenbaum, A Review of the Economic Literature on Cross-Market Healthcare Mergers, 82 Antitrust L.J. 533 (2019).

[55] Keith Brand, Martin Gaynor, Patrick McAlvanah, David Schmidt, & Elizabeth Scheirov, Economics at the FTC: Office Supply Retailers Redux, Healthcare Quality Efficiencies Analysis, and Litigation of an Alleged Get-Rich-Quick Scheme, 45 Rev. Indus. Org. 325 (2014); Julie A. Carlson, Leemore S. Dafny, Beth A. Freeborn, Pauline M. Ippolito, & Brett W. Wendling, Economics at the FTC: Physician Acquisitions, Standard Essential Patents, and Accuracy of Credit Reporting, 43 Rev. Indus. Org. 303 (2013); Joseph Farrell, David J. Balan, Keith Brand, & Brett W. Wendling, Economics at the FTC: Hospital Mergers, Authorized Generic Drugs, and Consumer Credit Markets, 39 Rev. Indus. Org. 271 (2011). Cf. Martin Gaynor, Kate Ho, & Robert J. Town, The Industrial Organization of Health-Care Markets, 53 J. Econ. Lit. 235 (2015); Brand & Rosenbaum, supra note 54 (review of cross-market health-care mergers literature).

[56] See Carlson et al., supra note 32.

[57] RFI at 4.

[58] See, e.g., FTC Staff Comment to the Food and Drug Administration in Docket No. FDA-2021-N-0555 Concerning Over-the-Counter Hearing Aids, Fed. Trade Comm’n (Jan. 28, 2022), available at (noting the likely procompetitive effect of rule).

[59] 16 C.F.R. § 315.

[60] 15 U.S.C. 7601-7610.

[61] 21 C.F.R. § 800.30.

[62] Pub. L. 115-52, 131 Stat. 1005, Aug. 18, 2017.

[63] See, e.g., Thomas W. Merrill, Antitrust Rulemaking: the FTC’s Delegation Deficit, 75 Admin Law Rev. 277 (2023) (“the FTC has no legal authority to engage in legislative rulemaking on competition matters.” Id. at 278); see also, Thomas W. Merrill et al., Agency Rules with the Force of Law: The Original Convention, 116 Harv. L. Rev. 467 (2002).

[64] Broadcast Music, Inc. v. Columbia Broad. Sys., Inc., 441 U.S. 1, 19-20 (1979) (citations omitted).

[65] Sneha Kannan, Joseph Dov Bruch, & Zirui Song, Changes in Hospital Adverse Events and Patient Outcomes Associated with Private Equity Acquisition, 330 JAMA 2365 (2023).

[66] As discussed below, other studies suggest mixed results. See, e.g., infra. note 118, and accompanying text.

[67] Regarding competition advocacy generally, see, e.g., James C. Cooper, Paul A. Pautler, & Todd J. Zywicki, Theory and Practice of Competition Advocacy at the FTC, 72 Antitrust L.J. 1091 (2005); Maureen K. Ohlhausen, Identifying, Challenging, and Assigning Political Responsibility for State Regulation Restricting Competition, 2 Comp. Pol’y Int. 151 (2006); Daniel J. Gilman, Advocacy, in SAGE Encyclopedia of Political Behavior 8 (Fathali M. Moghaddam ed., 2017). Links to numerous studies, reports, and advocacy documents by the FTC and its staff are at We note that FTC and DOJ jointly issued many such documents. See, e.g., Joint Statement of the Federal Trade Commission and the Antitrust Division of the U.S. Department of Justice Regarding Certificate-of-Need (CON) Laws and Alaska Senate Bill 62, Which Would Repeal Alaska’s CON Program (Apr. 12, 2017), available at

[68] RFI at 5.

[69] RFI at 6.

[70] Steven C. Salop & David T. Scheffman, Cost-Raising Strategies, 36 J. Indus. Econ. 19 (1985).

[71] Janusz A. Ordover, Garth Saloner, & Steven C. Salop, Equilibrium Vertical Foreclosure, 80 Am. Econ. Rev. 127 (1990).

[72] Malcolm B. Coate & Andrew N. Kleit, Exclusion, Collusion, and Confusion: The Limits of Raising Rivals’ Costs (FTC Bureau of Economics, Working Paper No. 179, 1990).

[73] Id. at 3.

[74] James C. Cooper et al.Vertical Antitrust Policy as a Problem of Inference, 23 Int’l. J. Indus. Org. 639, 643 (2005).

[75] David Reiffen & Michael Vita, Comment: Is There New Thinking on Vertical Mergers? 63 Antitrust L. J. 917 (1995); see also, e.g., Daniel O’Brien, The Antitrust Treatment of Vertical Restraint: Beyond the Possibility Theorems, in THE PROS AND CONS OF VERTICAL RESTRAINTS 22, 36 (Konkurrensverket ed., 2008); Michael A. Salinger, Vertical Mergers and Market Foreclosure, 103 Q. J. Econ. 345 (1988).

[76] Henry Ogden Armour & David Teece, Vertical Integration and Technological Innovation, 62 Rev. Econ. & Stat. 470 (1980).

[77] Dennis W. Carlton, Transaction Costs and Competition Policy, 73 Int’l J. Indus. Org. 1 (2020).

[78] Francine Lafontaine & Margaret Slade, Vertical Integration and Firm Boundaries: The Evidence, 45 J. Econ. Lit. 629, 677 (2007).

[79] Id.

[80] Id.

[81] Id.

[82] Cooper et al., supra note 74, at 639.

[83] John M. Yun, Vertical Mergers and Integration in Digital Markets, in THE GAI REPORT ON THE DIGITAL ECONOMY (Joshua D. Wright & Douglas H. Ginsburg eds., 2020) at 245.

[84] Francine Lafontaine, Vertical Mergers (Presentation Slides), in FTC, Competition and Consumer Protection in the 21st Century; FTC Hearing #5: Vertical Merger Analysis and the Role of the Consumer Welfare Standard in U.S. Antitrust Law, Presentation Slides 93 (Nov. 1, 2018), available at See also Francine Lafontaine & Margaret E. Slade, Presumptions in Vertical Mergers: The Role of Evidence, 59 Rev. Indus. Org. 255 (2021).

[85] See Koch, Wendling, & Wilson, Outcomes, supra note 47 (discussing research challenges and mixed results in the literature).

[86] Koch, Wendling, & Wilson, Quantity and Cost of Care, supra note 32

[87] Id. at 20.

[88] Id. at 20.

[89] Koch, Wendling, & Wilson, Outcomes, supra note 47.

[90] Conglomerate Effects of Mergers – Note by the United States 2, OECD (Jun. 10, 2020), available at (“Conglomerate Effects”).

[91] Id. at 2-3.

[92] Research and Development in the Pharmaceutical Industry, Cong. Budget Off. (April 2021),

[93] Josh Feng et al., Mergers that Matter: The Impact of M&A Activity in Prescription Drug Markets 6 (SSRN Working Paper, Jul. 25, 2023),

[94] Id. at 5-6.

[95] Leemore Dafny, Katherine Ho, & Robin Lee, The Price Effects of Cross-Market Mergers: Theory and Evidence from the Hospital Industry, 50 RAND J. Econ. 286 (2019) (finding price effects for mergers across geographic markets, but within state boundaries); see also Keith Brand & Ted Rosenbaum, A Review of the Economic Literature on Cross-Market Healthcare Mergers, 82 Antitrust L.J. 533 (2019) (reviewing several studies and noting observed competitive effects and issues for further study).

[96] See, e.g., Brand & Rosenbaum, supra note 95 (regarding possible application to hospital-merger cases, among others, as well as issues for further research).

[97] We note that, e.g., in January 2021, the FTC issued orders under its FTC Act Section 6(b) authority to six health-insurance companies to furnish information in order to facilitate the agency’s study of the effects of physician group and health-care-facility consolidation from 2015 through 2020. Press Release, FTC to Study the Impact of Physician Group and Healthcare Facility Mergers, Fed. Trade Comm’n (Jan.14, 2021), While the information collected under such orders is limited partly by restrictions imposed under the Paperwork Reduction Act, and not merely available data and methodological concerns, it may nonetheless help advance understanding of provider consolidation.

[98] Kannan et al., supra note 65.

[99] RFI at 1 (the second sentence of the summary).

[100] See generally, e.g., FTC v. Phoebe Putney Health Sys., Inc., 568 U.S. 216 (2013). In its unanimous decision, the Court noted that the 11th U.S. Circuit Court of Appeals had, as an initial matter, “‘agreed with the [FTC] that, on the facts alleged, the joint operation of Memorial and Palmyra would substantially lessen competition or tend to create, if not create, a monopoly’” 568 U.S. at 222-3 (quoting FTC v. Phoebe Putney Health Sys., Inc., 663 F.3d 1369, 1375 (2011). The Court’s holding in Phoebe Putney upheld the FTC’s jurisdiction over the hospital merger, notwithstanding the grant of certain powers to hospital authorities by the State of Georgia. 568 U.S. at 224. For a discussion of various FTC research, advocacy, and enforcement activities in health care, including scrutiny of provider mergers, see, e.g., Maureen K. Ohlhausen, The First Wealth is Health: Protecting Competition in Healthcare Markets, Remarks at the 2017 ABA Fall Forum (Nov. 16, 2017), available at While the FTC and the DOJ have concurrent jurisdiction over mergers under Section 7 of the Clayton Act, health-care-provider mergers are typically assigned to the FTC under the FTC/DOJ clearance process. For a list of health-care-enforcement matters, see FTC, The FTC’s Health Care Work: Cases, (last visited May 1, 2024).

[101] See, e.g., David J. Balan & Patrick S. Romano, A Retrospective Analysis of the Clinical Quality Effects of the Acquisition of Highland Park Hospital by Evanston Northwestern Healthcare, 18 Int. J. Econ. Bus. 45 (2011) (initially published as a BE Working Paper, available at

[102] Gaynor testimony, supra note 3.

[103] Ashenfelter et al., supra note 37, at 12.

[104] Id.

[105] See, e.g., Wilson, supra note 49 (studying hemodialysis care and finding no statistically significant indication that for-profit status is associated with a different quality of care; and comparing, e.g., Paul Grieco, & Ryan C. McDevitt, Productivity and Quality in Health Care: Evidence from the Dialysis Industry, 84 Rev. Econ. Studs. 1071 (2006) with John M. Brooks et al., Effect of Dialysis Center Profit-Status on Patient Survival: A Comparison of Risk-Adjustment and Instrumental Variable Approaches, 41 Health Servs. Res. (2006)). As we note below, it may be difficult to generalize observations from the U.S. dialysis industry because of both variation in the quality of care and the degree to which two firms account for for-profit acquisitions of independent facilities.

[106] See Wilson, supra note 49.

[107] For example, in a 2020 paper, Eliason et al. observed that only 21% of dialysis facilities were independently owned, and that two large, publicly traded companies owned 60% of the facilities and 90% of the revenue in the space. Paul J. Eliason et al., How Acquisitions Affect Firm Behavior and Performance: Evidence from the Dialysis Industry, 135 Q. J. Econ. 221, 222 (220). We note, too, that the FTC already has consent orders in place with both of those firms. Under one such order, DaVita, Inc. was required to divest certain facilities and limit its use of noncompete agreements; it must also get prior approval for future acquisitions from the FTC. See, In the Matter of DaVita, Inc., and Total Renal Care, FTC File No. 211-0013 (Oct. 25, 2021), (agreement containing consent orders).

[108] RFI at 3.

[109] RFI at 5.

[110] RFI at 4-5.

[111] Elena Praeger & Matt Schmitt, Employer Consolidation and Wages: Evidence from Hospitals, 111 Am. Econ. Rev. 397–427 (2021). Praeger & Schmitt examine whether hospital employees’ wage growth slows following consolidation. While they observe some slowing wage growth under limited conditions (large increases in concentration, plus industry-specific skills), they fail to reject zero wage effects in most cases.

[112] See supra note 10.

[113] FTC v. Phoebe Putney Health Sys., Inc., 568 U.S. 216 (2013).

[114] See, e.g., FTC Staff Comment Before the Georgia Department of Community Health Regarding the Certificate of Need Application Filed by Lee County Medical Center, Fed. Trade Comm’n (2017), available at (discussing ongoing dearth of competition for hospital services in surrounding five-county area).

[115] For example, in a 2019 letter to the Texas Medical Board, FTC staff noted that most of the critical-access hospitals in Texas were located in counties where there were no practicing anesthesiologists, with 37 of those hospitals located in counties where certified-registered-nurse anesthetists were the only licensed, specialized providers of anesthesia and anesthesia-related services. FTC Comment to Texas Medical Board on Its Proposed Rule 193.13 to Add Supervision Requirements for Texas-Certified Nurse Anesthetists, 2, Fed. Trade Comm’n (2019), available at

[116] See supra text accompanying notes 105-107.

[117] Id. (citing Wilson, supra note 49).

[118] Kannan et al., supra note 65.

[119] Joseph D. Bruch, Suhas Gondi, & Zirui Song, Changes in Hospital Income, Use, and Quality Associated with Private Equity Acquisition, 180 JAMA Intern. Med. 1 (2020).

[120] Agency staff have no doubt also noticed that the studies regard limited numbers of private-equity acquirers. For example, the Bruch, Gondi, & Song study, id., incorporates numerous acquisitions by the Hospital Corporation of America (HCA), which may provide a sharper picture of HCA acquisitions, but may or may not generalize across the industry.

[121] American Medical Assn. v. FTC, 638 F.2d 443 (2d Cir. 1980) (aff’d per curiam American Medical Assn. v. FTC, 455 U.S. 676 (1982)); cf., e.g., Matthew Mandelberg et al., Reconsidering the Ban on Physician-Owned Hospitals to Combat Consolidation, and Matthew Mandelberg, Michael Smith, Jesse Ehrenfeld, & Brian Miller, Reconsidering the Ban on Physician-Owned Hospitals to Combat Consolidation (Feb. 5, 2023). Forthcoming in N.Y.U. J. Leg. & Pub. Pol’y, available at SSRN:

[122] Statement on L.D. 1866 to the Committee on Bus. Leg., Maine House of Representatives (Jan. 8, 1992), available at; see also, FTC Staff Comment Before the North Carolina State Board of Opticians Concerning Proposed Regulations for Optical Goods and Optical Goods Businesses, Fed. Trade Comm’n (2011), Cf. FTC Staff Comment to the Food & Drug Admin. in Docket No. FDA-2021-N-055 Concerning Over-the-Counter Hearing Aids, Fed. Trade Comm’n (Jan. 8, 2022), available at

[123] A 2009 report by then-FTC Chair William Kovacic defines “policy R&D” broadly in a way that comprises, but is not limited to, original, author-initiated academic research by BE staff. It also includes various review, issue-spotting, and synthetic endeavors, such as policy workshops and, indeed, requests for public information. William E. Kovacic, The Federal Trade Commission at 100: Into Our Second Century, 91-92 (Jan. 2009), available at

[124] RFI at 3.

[125] Id. at 3-4.

[126] For a review of diverse endeavors, see, e.g., Ohlhausen, supra note 3.

[127] Press Release, Federal Trade Commission, Department of Justice, and Department of Health and Human Services Launch Cross-Government Inquiry on Impact of Corporate Greed in Health Care, Fed. Trade Comm’n (Mar. 5, 2024),

[128] Press Release, FTC, DOJ, and HHS Extend Comment Period on Cross-Government Inquiry on Impact of Corporate Greed in Health Care, Fed. Trade Comm’n (May 1, 2024),

[129] Press Release, Fact Sheet: Biden-Harris Administration Announces New Actions to Lower Health Care and Prescription Drug Costs by Promoting Competition, The White House (Dec. 7, 2023),

[130] See, e.g., Press Release, Justice Department, Federal Trade Commission and Department of Health and Human Services Issue Request for Public Input as Part of Inquiry into Impacts of Corporate Ownership Trend in Health Care, Dep’t Justice, (Mar. 5, 2024),

[131] RFI at 3.

[132] See supra Section II.D.

[133] Gaynor statement, supra note 3.

[134] Private Capital, Public Impact: An FTC Workshop on Private Equity in Health Care, Fed. Trade Comm’n (May 5, 2024), The workshop webpage includes a description, along with links to the agenda, participant biographies, and a transcript of the proceedings.

[135] Id.; Press Release, Justice Department, Federal Trade Commission and Department of Health and Human Services Issue Request for Public Input as Part of Inquiry into Impacts of Corporate Ownership Trend in Health Care, Dep’t Justice (Mar. 5, 2024),; Press Release, HHS, DOJ, and FTC Issue Request for Public Input as Part of Inquiry into Impacts of Corporate Ownership Trend in Health Care, Dep’t Health & Human Servs. (Mar. 5, 2024),

[136] Their testimony is confined to their own perceptions of, as the agencies themselves put it in the RFI, “how their experiences . . . changed after a facility or other provider where they work or receive treatment or services was acquired or underwent a merger.” Such perceptions may help make certain policy concerns vivid or accessible, but there is no credible argument that they were either randomly selected or representative of practitioner experience, much less that they represent legal or economic analyses of the acquisitions under discussion. That they may be considered as part of a larger policy inquiry is uncontroversial. That three such participants were selected for such a brief workshop—absent industry participants, and given the dearth of economic evidence and legal perspectives beyond those of enforcers—strains credulity.

[137] See supra note 135.

Continue reading
Antitrust & Consumer Protection

New York, Listen to California: Antitrust Legislation Threatens Our Innovation Economy

Popular Media California does not have a reputation for business-friendly legislation. This makes it all the more surprising that a California legislative report rejected a New York . . .

California does not have a reputation for business-friendly legislation. This makes it all the more surprising that a California legislative report rejected a New York bill as too anti-business for the Golden State. That bill, the 21st Century Antitrust Act, championed by New York State Senate Deputy Majority Leader Michael Gianaris (D-Queens), would import European competition-policy principles and expand on them, ultimately making New York an outlier in U.S. antitrust enforcement.

In its current form, Gianaris’ bill would lead enforcers to punish the mere possession of monopoly power, rather than anti-competitive behavior that harms consumers. This marks a firm rejection of longstanding U.S. antitrust principles. Indeed, not punishing monopolization has been a longstanding concern of U.S. antitrust law. As Albany native and Second Circuit Court of Appeals Judge Learned Hand wrote in 1945: “The successful competitor, having been urged to compete, must not be turned upon when he wins.”

Continue reading
Antitrust & Consumer Protection

Dynamic Competition in Broadband Markets: A 2024 Update

ICLE White Paper I. Introduction In mid-2021, the International Center for Law & Economics (ICLE) published a white paper on the state of broadband competition in the United . . .

I. Introduction

In mid-2021, the International Center for Law & Economics (ICLE) published a white paper on the state of broadband competition in the United States,[1] which concluded that:

  • The U.S. broadband market was generally healthy and competitive, with 95.6% of the population having access to high-speed broadband;
  • Concentration metrics are poor predictors of competitiveness—broadband markets can be dynamic and competitive even with only a few providers. Indeed, in some cases, increased concentration can result from efficiency gains and innovation, benefiting consumers through better services; and
  • Municipal broadband often requires significant taxpayer subsidies or cross-subsidies from other municipal enterprises, and is thus an example of “predatory entry,” rather than market competition.[2]

Rather than repeat the analysis conducted in the 2021 report, in this report, we investigate the extent to which broadband competition has evolved over the past three years. We find that it has been a rapid evolution:

  • More households are connected to the internet;
  • Broadband speeds have increased, while prices have fallen;
  • More households are served by multiple providers; and
  • New technologies like satellite and 5G have expanded internet access and intermodal competition among providers.

When the 2021 ICLE white paper was published, the worst of the COVID-19 pandemic appeared to be over, but the virus’ Delta variant was surging.[3] With pandemic precautions keeping people at home to work, go to school, visit health-care providers, or be entertained, broadband access and use was seen by many as a necessity, rather than a luxury. At the time, Congress considered whether to devote significant federal resources toward promoting broadband access in underserved communities. Toward this end, in November 2021, Congress passed the Infrastructure Investment and Jobs Act (IIJA), which includes three key provisions to foster greater broadband access:[4]

  1. The COVID-era Emergency Broadband Benefit’s temporary subsidy was extended indefinitely and renamed the Affordable Connectivity Program (ACP). The IIJA allocated an additional $14 billion to provide subsidies of $30 a month to eligible households;
  2. The IIJA also created and funded the Broadband Equity, Access, and Deployment Program (BEAD), which provides $42 billion to expand high-speed internet access to “unserved” and “underserved” locations; and
  3. The law required the Federal Communications Commission (FCC) to adopt final rules to prevent “digital discrimination” in broadband access based on income level, race, ethnicity, color, religion, or national origin, while also instructing the commission to consider issues of technical and economic feasibility.

These three policies were intended to intertwine in order to foster greater broadband competition. ACP subsidies are intended to boost consumer demand for broadband and generate revenue to support providers’ profitable deployment of broadband investments.[5] BEAD investments are intended to reduce the costs of broadband deployment.[6] The law’s digital-discrimination provisions were intended to prevent discrimination by broadband providers that serves to deny or limit consumers’ access to broadband internet.[7]

Alas, today, we find that each of these provisions faces headwinds. With Congress failing to extend appropriations beyond a May 31 deadline, the ACP has run out of funding.[8] States attempting to implement the BEAD program have complained of tight timelines, restrictive rules, limited coordination, and administrative burdens that may undermine effectiveness.[9] Providers and local jurisdictions report that BEAD’s Buy America rules are particularly onerous.[10] Smaller internet service providers say BEAD’s financial requirements exclude them from projects they would otherwise be able to complete successfully.[11] Complying with Buy America rules regarding attaching equipment to utility poles and railroad crossings also threatens deployment timelines.[12] And, in November 2023, the FCC approved rules to apply a disparate-impact approach toward the IIJA’s digital-discrimination mandate, which could raise constitutional issues over the major questions doctrine.[13]

In addition to these programs, the FCC appears dead set to regulate more stringently much of the broadband-internet industry. First, the agency’s sweeping digital-discrimination rules cover nearly every aspect of the deployment and delivery of internet services and nearly every entity associated—even tangentially—with deployment and delivery.[14] Next, the agency approved Title II common-carrier regulation with its recently adopted Safeguarding and Securing the Open Internet Order.[15],[16]

The current state of broadband competition policy appears to be one of confusion. Some policies foster competition, while others hinder it. Programs such as the ACP and BEAD could do much to encourage competition by simultaneously generating demand for broadband and helping to build out supply. At the same time, these programs—especially BEAD—attempt to micromanage competition with stifling conditions and de facto rate regulation. Similarly, the FCC’s digital-discrimination rules explicitly subject broadband pricing to ex post scrutiny and enforcement. The FCC’s reclassification of broadband internet-access services under Title II of the Communications Act raises the specter of common-carrier rate regulation that will hang over the industry unless either vacated by the courts, or a future administration once again reverses course.

Put simply, broadband competition in the United States is currently robust, innovative, and successful. But this state of vibrant competition is at risk from recent and forthcoming regulations. Without a course correction, we are likely to see slowing or shrinking broadband investment, reduced innovation, and the exit of small and rural providers.

II. The Broadband Market Is Competitive and Dynamic

By all relevant measures, U.S. broadband competition is vibrant and has increased dramatically since the COVID-19 pandemic. Since 2021, more households are connected to the internet, broadband speeds have increased, prices have fallen, more households are served by more than a single provider, and new technologies like satellite and 5G have expanded internet access and intermodal competition among providers.

A. Access and Adoption

By any reasonable measure today’s U.S. broadband market is an incredible success. Nearly the entire country has access to at-home internet, a vast majority has access to high-speed internet, and much of the country has access to these speeds from three or more providers. Nevertheless, criticisms of the current state of broadband deployment claim that too few Americans have affordable access to adequate broadband speed and capacity and that this, in turn, is the result of insufficient competition among broadband providers.[17] For example, in her speech announcing the FCC’s most recent process to regulate internet services under Title II, Chair Rosenworcel claimed that 80% of the country faces a monopoly or duopoly for download speeds of 100 Mbps or greater.[18] These claims are belied by widespread broadband adoption and competitive markets.

FIGURE 1: US At-Home Internet Access and Adoption, 2021

SOURCE: U.S. Census Bureau, American Community Survey

The U.S. Census Bureau’s American Community Survey reports that 97.6% of households have access to at-home internet and 92.6% use the internet at home (Figure 1).[19] While a large majority with at-home internet get it through a broadband subscription, a substantial minority access the internet from their mobile wireless providers. A small number (2.3%) claim they can access the internet at home without paying for a subscription. This likely includes multi-family units, as well as student and senior housing in which broadband access is included in the rent. Among the 7.4% who do not use an at-home internet connection, two-thirds indicate that internet access is available, but they have chosen not to adopt it.[20]

In 2021, approximately 97 percent of 3- to 18-year-olds had home internet access, according to the National Center for Education Statistics. This represents a five-percentage-point increase since 2016.[21]

Until March 2024, the FCC defined high-speed broadband as internet service that offered speeds of at least 25/3 Mbps.[22] The IIJA defines a location as “unserved” if it has no internet connection available or only has a connection offering speeds of less than 25/3 Mbps.[23] A location is considered “underserved” if the only options available offer speeds of less than 100/20 Mbps.[24]

As shown in Figure 2, smaller households with relatively simple needs can generally access the internet productively with download speeds of less than 100 Mbps, or even 25 Mbps. The third iteration of the National Broadband Map, released in November 2023, indicated:[25]

  • 8% of locations have access to connections of 25/3 Mbps or greater;
  • 5% of locations have access to speeds of 200/25 Mbps or greater;
  • 5% of locations have access to 1000/100 Mbps speeds; and
  • Only 6.2% of locations are unserved, and 2.6% are “underserved” with connections of less than 100/20 Mbps, as those terms are defined in the IIJA.

FIGURE 2: FCC Recommended Internet Speeds and US Household Access, 2021

SOURCE: Allconnect, ‘Everything You Need to Know;’ FCC, ‘Fixed Broadband Deployment’

FIGURE 3: Typical Maximum Download Speed by Connection Type, 2021 (Mbps)

SOURCE:, ‘What Type of Internet Do You Have?’

The FCC reports that more than 90% of U.S. households have access to speeds of 100 Mbps or greater, and nearly 90% have access to 1 Gbps or greater (Table 1).[26] Fewer than 4% of U.S. households lack access to at least 30 Mbps download speeds via fixed broadband.

TABLE 1: US Household Internet Access by Download Speed, 2021

SOURCE: FCC, ‘International Broadband Data Report’[27]

Some note that, while high-speed connections are available across nearly the entire country, in many cases, only a single provider offers such speeds. This, such critics assert, suggests insufficient competition among providers of high-speed internet. For example, regarding 100 Mbps service, FCC Chair Rosenworcel claimed that “only half of us can get it from more than a single provider. Only one-fifth of the country has more than two choices at this speed.”[28]

This provides a misleading sense of the rate of high-speed broadband deployment and the scope of availability. The most recent information from the FCC on broadband deployment across the United States suggests that 90% of the population in 2021 was served by one or more providers offering 250/25 Mbps or higher speeds (Table 2).[29] That is more than double the population share five years earlier, when only 44% of the population had access to such speeds.[30] In 2019, the FCC did not report the share of population with access to 1,000/100 Mbps speeds or greater. By 2021, 28% of the population had access to gigabit download speeds.[31]

Moreover, Table 2 shows that, in 2021, more than 85% of the population was covered by two or more fixed-broadband providers offering 25/3 Mbps or greater speeds, and more than 60% of the country was covered by three or more providers providing such speeds. Moreover, if satellite and 5G providers are included, close to 100% of the country is served by two or more high-speed providers.

TABLE 2: US Population Fixed-Broadband Access by Number of Providers, 2021

SOURCE: FCC, ‘Fixed Broadband Deployment’

At the same time, the evidence indicates that broadband competition has increased over time, as measured by the number of competing high-speed providers (Figure 4).[32]

  • In 2018, 73.0% of households had access to 25/3 Mbps speeds from only one or two fixed-broadband providers, and only 21.6% had access from three or more providers. In 2021, only 29.1% of households had access from one or two providers while 69.3% were served by three or more providers. Thus, the number of households served by three or more providers increased by 47.7 percentage points from 2018 through 2021.
  • In 2018, 11.6% of households had no access to 100/20 Mbps speeds and 14.8% had access from three or more fixed broadband providers. In 2021, 5.4% of households had no access, while 21.3% were served by three or more providers. Thus, the number households served by three or more providers increased by 6.5 percentage points from 2018 through 2021.

FIGURE 4: Percentage of US Households Living in Census Blocks with Multiple Provider Options for Fixed-Terrestrial Services (2018 vs 2021)

SOURCE: FCC, ‘2022 Communications Marketplace Report’

Additionally, intermodal competition among providers is only improving. Starlink satellite service has been made available to all locations in the United States.[33] Starlink’s reported speeds are between 25/5 Mbps and 220/25 Mbps.[34] And Project Kuiper has successfully launched its first test satellites,[35] with commercial service expected to begin in the second half of 2024.[36]

B. Broadband Prices Continued to Fall, Even as Speeds Increased and Demand Grew During the Pandemic

After accounting for speed and data usage, the United States has some of the lowest broadband prices in the world. Even so, critics of the current state of U.S. broadband competition claim that U.S. prices are among the highest in the developed world because, they claim, the U.S. market is not as competitive as other jurisdictions. For example, the Community Tech Network asks rhetorically, “[s]o why does the internet cost so much more in the U.S. than in other countries? One possible answer is the lack of competition.”[37] Their article included a graphic in which U.S. internet service is described as “expensive and slow” while Australia is categorized as “fast and cheap.” Yet none of these claims hold up under scrutiny, such as adjusting for consumption and download speeds.

It’s true the United States has the third-highest average monthly broadband costs among OECD countries, according to (Figure 5). Australia, however, has the seventh-highest.[38] On a cost-per-megabit basis, Australia has the second-highest costs in the OECD, while the United States is in the bottom third of the distribution (Figure 6).[39] Speedtest’s Global Index of median speeds reports that the United States has the second-fastest median speed, and Australia the third-slowest median speed, among OECD countries (Figure 7).[40]

FIGURE 5: Average Monthly Cost of Broadband (OECD, in $US)

SOURCE:, ‘Global Broadband Pricing League Table 2023’

FIGURE 6: Average Monthly Cost of Broadband (OECD, Per Megabit $US)

SOURCE:, ‘Global Broadband Pricing League Table 2023’

FIGURE 7: Median Download Speed (OECD, Mbps)

SOURCE: Speedtest, Global Index

Cross-country comparisons of broadband pricing are especially fraught, due to country-by-country variations in factors that drive the costs of delivering broadband and the prices paid by consumers.[41] Deployment costs are driven largely by population density and terrain, as well as each country’s unique regulatory and tax policies.[42] Consumer choices often drive the prices paid by subscribers. These include choices regarding the mix of fixed broadband and mobile, speed preferences, and data consumption.[43]

For example, Figure 8 demonstrates a clear relationship between the average monthly cost for broadband and the monthly cost per megabit; a higher monthly cost tends to be associated with a higher cost per megabit. But there are outliers. The United States is well below the trendline, but Canada is well above it. While the average monthly cost in the two countries is similar, the information provided by suggests that U.S. consumers use 9-10 times more megabits per month than Canadian consumers. In addition, as shown in Figure 7, the median U.S. download speed is about 35% faster than the median in Canada.

FIGURE 8: Relationship Between Average Monthly Cost of Broadband and Cost Per-Megabit Per-Month (OECD, in $US)

SOURCE:, ‘Global Broadband Pricing League Table 2023’

FIGURE 9: Relationship Between Average Monthly Cost of Broadband and Median Download Speed

SOURCE:, ‘Global Broadband Pricing League Table 2023’; Speedtest, Global Index

A broadband-pricing index published annually by USTelecom reports that inflation-adjusted broadband prices for the most popular speed tiers fell by 54.7% from 2015 to 2023, or 5.6% annually.[44] Prices for the highest speed tiers fell by 55.8% over the same period. The Producer Price Index for residential internet-access services fell by 11.2% from 2015 through July 2023.[45] The median fixed-broadband connection in the United States delivers more than 207 Mbps download service, an 80% increase over pre-pandemic median speeds (Figure 10).[46]

FIGURE 10: Median Download Speed in the US (Mbps)

SOURCE: Speedtest, Global Index (July of each year)

Evidence from large surveys suggests that price is not a dominant factor driving adoption for the currently unconnected. For example, among the 7% of households who do not use the internet at home, more than half of Current Population Survey respondents indicated that they “don’t need it or [are] not interested.”[47] About one-third of respondents indicated that price is a factor, with responses such as “can’t afford it” or “not worth the cost.”[48]

Of course, cost and interest are not mutually exclusive factors.[49] A common response to CPS surveys among those who do not subscribe to internet service is that it is “not worth the cost.” This is an unhelpful response to guide policymakers because it doesn’t answer whether the cost is “too high,” the value is “too low,” or a combination of both. Another common response is “not interested.” This, too, is unhelpful, as it does not identify the price at which a potential consumer might become interested, if such a price exists. For example, surveys suggest that some nonadopters may become interested in subscribing to internet services or find it worth the cost at a price of zero.

  • A National Telecommunications and Information Administration (NTIA) survey of internet use reported the average monthly price that offline households wanted to pay for internet access was approximately $10 per month; roughly 75% of households gave $0 or “none” as their answer.[50]
  • Another NTIA publication reports that households with “no need/interest” in home internet are willing to pay about $6 a month, while those who indicate it is “too expensive” are willing to pay approximately $16 a month.[51]

In addition, as shown in Figure 1, about a quarter of households without a broadband or smartphone subscription claim that they can access the internet at home without paying for a subscription.

Jamie Greig & Hannah Nelson note that low-income households are more likely to use smartphones than computers for internet access.[52] According to Pew Research, 19% of adults who do not have at-home broadband report that their smartphone does everything they need to do online.[53] Colin Rhinesmith et al. summarize the response of a Detroit focus group participant: “[I]f he had to choose between home access and mobile access, the latter is more desirable as it allows him to be reachable and flexible for job interviews and the like”[54]

C. Investment by Broadband Providers Has Remained High

When the FCC issued the Open Internet Order (OIO) in 2015 to reclassify broadband internet-access service under Title II, opponents claimed the policy would diminish broadband investment. Similarly, when the FCC repealed the reclassification in 2018, opponents claimed the repeal would diminish broadband investment. While U.S. broadband capital expenditures have been relatively stable for the past two decades, there was a noticeable drop in the wake of the 2015 OIO (Figure 11).[55]

FIGURE 11: US Broadband Provider Capital Expenditures ($B)


Recent peer-reviewed econometric research from economist Wolfgang Briglauer and his coauthors—indicates that net-neutrality rules do, in fact, slow broadband investment, as measured by the number of fiber connections deployed.[56] The study analyzed 2000-2021 data across OECD countries. Thus, it includes both 2015’s imposition of Title II regulations in the United States and the 2017 repeal. It found that introducing net-neutrality rules was associated with a 22-25% decrease in fiber investments.

Briglauer’s study isolated the effects of net neutrality from other factors that might have affected investment, such as general economic conditions. It focused on new fiber connections as representing growth in network capacity, rather than short-term fluctuations in spending. Even controlling for other variables, net neutrality had an independent negative relationship with fiber deployments.

ICLE’s 2021 white paper argued that broadband markets are dynamic and characterized by ongoing innovation in technologies and business models. Investment and innovation do not solely come from new entrants, as incumbents often are important sources of innovation while they try to stay competitive and avoid disruption. In this way, providers compete through new product introductions and disruption, not just on price. Because of these dynamics, mergers and increased concentration can sometimes be associated with increased investment, in that they may allow firms to achieve greater economies of scale and scope.[57] In addition, firms make long-term investments to upgrade networks and deploy new technologies even amid just a few competitors.[58]

Since ICLE’s white paper, Kenneth Flamm & Pablo Varas published research examining the relationship between the change in a territory’s number of providers and changes in service-plan quality (e.g., upload and download speeds).[59] They examine Census blocks that were served by only two “legacy” broadband providers in 2014, which they define as cable and digital subscriber line (DSL) providers. Their study tracked entry and exit of providers in these blocks through 2018, and evaluated the change in maximum download speeds available in those blocks over time. They find that blocks with no entry or exit (what they call “unchanged duopoly”) experienced an increase of 750 Mbps in maximum download speeds (Figure 12). Blocks that transitioned from duopoly to monopoly experienced a relatively modest 430 Mbps increase, while blocks that transitioned from two to three providers experienced an 810 Mbps increase. Blocks that transitioned from three to four providers experienced an 854 Mbps increase.

They also noted that internet providers may be highly motivated to introduce new, higher-quality speed tiers as technology improves. These results comport with research summarized in the 2021 ICLE white paper, which found the most significant incremental benefits in broadband quality came from adding a second service provider (relative to monopoly), with some marginal benefit from adding a third provider, and a much smaller benefit from adding a fourth.

FIGURE 12: Increase in Maximum Download Speed Associated with Cable or Digital Subscriber Line Provider Entry or Exit, 2014-2018 (Mbps)

SOURCE: Flamm & Varas (2022)

Another recent study is Andrew Kearns’ analysis of the Seattle market.[60] In contrast to Flamm & Varas, Kearns concluded that competition among broadband providers might weaken the incentive to increase quality, which he measured as a provider upgrading a Census block to fiber. He argued that improvements in quality often require significant investment, and the returns on this investment may be uncertain in a competitive market. Thus, in a competitive market, providers may prioritize attracting customers with lower prices and a wider range of product options, rather than investing in improvements to the quality of their service. Even so, Kearns concluded that increased competition offers substantial benefits to consumers related to increased product choice and lower prices.

The latest published research supports ICLE’s earlier observation that whether adding or removing a competitor is associated with more or less investment depends greatly on various factors, including the market’s initial conditions.[61] Thus, a case can be made that competition (as judged by counting the number of competitors in a market) may be, in and of itself, of only lesser importance relative to other factors that guide investment decisions, such as population density, terrain, and demand, as well as the local regulatory and tax environment.[62]

III. Current and Anticipated Policies Affecting Broadband Competition

Broadband internet has become a service that many Americans—and U.S. policymakers—consider essential. But new and forthcoming regulations imposed in an effort to promote equal access to broadband may actually risk dampening innovation and investment in this critical sector. In this section, we discuss the Affordable Connectivity Program and Broadband Equity, Access, and Deployment subsidy programs, which could foster broadband competition by stimulating both demand and supply. Even so, administration of both of these programs have erected significant hurdles that may damage their effectiveness if not remedied by Congress or the regulatory agencies.

We also discuss other programs that are likely to reduce broadband competition by diminishing the incentives to invest and innovate. Though motivated by a desire to prevent discriminatory access, rigid rules to correct “disparate impact” in broadband-deployment decisions fail to account for the dynamic efficiencies of differentiated service models calibrated to consumer demand. At the same time, attempts to impose common-carrier obligations on broadband providers ignore the truly competitive nature of modern broadband markets, which are thriving under light-touch regulation.

Going forward, policymakers should resist the temptation to micromanage a sector as dynamic as U.S. broadband internet. Instead, they should focus their attention on interventions to address genuinely unfair or anticompetitive conduct, while trusting that innovation and investment will be maximized when companies retain the flexibility to respond to consumer demand, while constrained by economic and technical realities.

A. ACP More Effective at Reducing Broadband Costs Than Connecting the Unconnected

The ACP is a federal subsidy program that provides eligible low-income households with monthly broadband-service discounts of up to $30, or up to $75 for households on tribal lands.[63] It also provides a one-time $100 discount for the purchase of a computer or tablet. ICLE has argued that well-designed subsidies targeted to underserved consumers can be an effective way to increase broadband deployment and adoption.[64] Subsidies help make providing service in high-cost, low-density areas more financially viable for providers. They also make broadband more affordable for lower-income consumers, stimulating demand.[65]

Proponents of the ACP identify two main goals for the program:

  1. to increase at-home internet adoption by unconnected households; and
  2. to maintain internet connections for low-income households at risk of “unadoption” due to unaffordability.[66]

Through the ACP, the federal government absorbs part of the cost of providing broadband service to these households, making them more financially attractive customers for broadband providers. The program also creates an incentive for providers to expand their networks to reach eligible households, as they can now potentially recover more revenue from serving those users.[67] For example, if ACP subsidies stimulate consumer demand, providers may find it profitable to deploy broadband to areas that would not otherwise generate a sufficient return on investment to justify deployment. In some cases, a new provider might be able to offer services to a market currently served by a single incumbent firm.

To date, however, the ACP and its predecessors do not appear to have been as successful in increasing at-home internet adoption by unconnected households as was hoped when such programs were created. Due to what appears to be inelastic demand, ACP has faced difficulties in stimulating sufficient interest among the 5% of unconnected households who could access the internet, but fail to take up service.[68] These households may not be aware of the program or may lack digital literacy; may be able to access the internet without a subscription; or may have no interest in subscribing to an internet service at any price.

On the other hand, the ACP’s subsidies appear to have successfully enabled already-subscribed households to maintain at-home internet service through the COVID-19 pandemic, thereby proving effective in enabling economically vulnerable inframarginal consumers to remain connected. More than 23 million U.S. households (about 17%) were enrolled in the ACP before the program lapsed at the end of May 2024.[69] It is currently unknown how many of these households will unsubscribe now that ACP subsidies are unavailable. In turn, it’s also unknown how providers will respond should large number of households unsubscribe from their internet services.

In March 2024, the FCC announced that April 2024 would be the program’s last fully funded month, with partial subsidies through May 2024.[70] Without ACP subsidies, one expects some households will unsubscribe from internet service, and the decreased demand may even lead to consolidation in some markets through exits or mergers. Moreover, Congress’ failure to renew the ACP risks other long-term policy responses that could waste already-invested funds.

In the face of another economic downturn, the inframarginal households that unadopt internet service will likely spur future rounds of congressional appropriations to bring these households back online. This turmoil, meanwhile, stands to erode providers’ investment incentives, due to lack of demand. This threatens to create a vicious cycle that requires periodic reinvestment from Congress just to stand these programs back up. Over the long term, it would almost certainly be more efficient to extend and focus the ACP program to ensure that truly needy households receive the subsidy (including those that would otherwise unadopt), rather than construing the program as strictly focused on convincing the last 5% of households with inelastic demand to adopt.

B. Red Tape and Regulation May Stymie BEAD’s Efforts to Expand Broadband Access

In 2023, the NTIA awarded more than $42 billion in grants to state governments under the Broadband Equity, Access, and Deployment (BEAD) program,[71] whose primary purpose is to expand high-speed internet access in areas that currently lack it.[72] Congress focused the BEAD program on connecting “unserved” and “underserved” territories. The law requires that those areas lacking connections with speeds of at least 100/20 Mbps must be helped first before addressing other priorities, such as upgrades, adoption programs, and middle-mile infrastructure.[73] Funding is distributed directly to states, which are required to develop plans tailored to connect their unserved and underserved locations.[74]

But much of that congressional intent got muddled in the NTIA’s implementation of BEAD funding. The NTIA’s notice of funding opportunity (“NOFO”) introduced conflicting priorities beyond connecting the unserved. These additional priorities include “middle-class affordability” requirements, the provision of “low cost” plans, and a ban on data caps.[75] The NOFO also gave clear preference to fiber networks over wireless and satellite providers, and to governmental and municipal providers over private companies.[76]

The NTIA’s NOFO prompted 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).[77] The notice provided several examples of how this could be achieved, including:

  1. Requiring providers to offer low-cost, high-speed plans to all middle-class households using the BEAD-funded network; and
  2. Providing consumer subsidies to defray subscription costs for households ineligible for the Affordable Connectivity Benefit or other federal subsidies.

Despite the IIJA’s explicit prohibition of price regulation, the NTIA’s approval process appears to envision exactly this. The first example provided above is clear rate regulation. It specifies a price (“low-cost”); a quantity (“all middle-class households”); and imposes a quality mandate (“high-speed”). Toward these ends, the notice provides an example of a “low-cost” plan that would be acceptable to NTIA:

  • Costs $30 per month or less, inclusive of all taxes, fees, and charges, with no additional non-recurring costs or fees to the consumer;
  • Allows the end user to apply the Affordable Connectivity Benefit subsidy to the service price;
  • Provides download speeds of at least 100 Mbps and upload speeds of at least 20 Mbps, or the fastest speeds the infrastructure is capable of if less than 100 Mbps/20 Mbps;
  • Provides typical latency measurements of no more than 100 milliseconds; and
  • Is not subject to data caps, surcharges, or usage-based throttling.[78]

A policy bulletin published by the Phoenix Center for Advanced Legal & Economic Public Policy Studies notes that the NTIA did not conclude that broadband was unaffordable for middle-class households.[79] George Ford, the bulletin’s author, collected data on broadband adoption by income level. The data indicate that, in general, internet-adoption rates increase with higher income levels (Figure 12). Higher-income households have higher adoption rates (97.3%) than middle-income households (92.9%), which in turn have higher adoption rates than lower-income households (78.1%).

FIGURE 13: Internet Adoption and Income

SOURCE: Adapted from Ford (2022), Table 2 and Figure 2.

For each of the 50 states and the District of Columbia, the Phoenix bulletin finds that middle-income internet-adoption rates are, to a statistically significant degree, higher than lower-income adoption. Thus, the Phoenix bulletin concludes that broadband currently is “affordable” to middle-class households and that “no direct intervention is required” to ensure affordability to the middle class.[80]

John Mayo, Greg Rosston, & Scott Wallsten point out that BEAD’s key purpose of providing high-speed internet access to locations that lack it (presumably because it’s too expensive to deploy to these areas without investment subsidies) conflicts with NTIA’s focus on affordability:

A substantial portion of the unserved and underserved areas of the country that are the likely targets of the BEAD program, however, are rural, low-population density areas where deployment costs will be high. These high deployment costs may seem to indicate that even “cost-based” rates—normally seen as an attractive competitive benchmark—may be high, violating the IIJA’s “affordability” standard.[81]

The only effective way to simultaneously reduce broadband prices, increase access, and improve quality is to increase supply. But the NTIA’s attempts at rate regulation work at cross-purposes with BEAD’s objective to increase supply. Therefore, attempts to use BEAD funding to impose price controls may act to reduce broadband competition, rather than preserve or increase it.

The potential harm to competition is worsened by NTIA’s preference for government or municipal providers over private providers, which we discuss in more detail in Section III.G. The NTIA’s funding notice required states to ensure the participation of “non-traditional broadband providers,” such as municipalities and cooperatives. Municipal broadband networks might make sense in some rare cases where private providers are unable to deploy, but such systems have generally mired taxpayers in expensive projects that failed to deliver on promises.

In addition to these challenges, BEAD applications must come with a letter of credit issued by a qualified bank for 25% of the grant amount.[82] This is a guarantee to the grant administrator (e.g., a state broadband office) that there is liquid cash in an account that it can claw back should the applicant not deliver on their grant requirements. To receive a letter of credit, applicants will be required by the issuing bank to provide collateral—which could be cash or cash equivalents equal to the full value of the letter of credit. The letter-of-credit requirement is separate and in addition to BEAD’s match requirement, which demands that applicants contribute a minimum 25% of the total build cost. The letter-of-credit and matching requirements may hinder competition by favoring large and well-capitalized providers over smaller internet-service providers (ISPs) that may be better positioned to serve rural areas.

In November 2023, NTIA released a waiver for the letter-of-credit requirement because of industry concerns about how the rule may prevent smaller ISPs from participating in the BEAD program.[83] The “programmatic waiver” describes several alternatives to the letter of credit. For example, subgrantees can obtain the letter of credit from a credit union instead of a bank. The expectation is that credit unions would offer lower interest rates for loans and lower fees. Alternatively, applicants can provide a performance bond “equal to 100% of the BEAD subaward amount.” In addition, the NTIA is allowing states and territories to reduce the percentage requirement of the performance bond or letter of credit over time, as service providers meet certain project milestones.

Congress set an ambitious goal with BEAD: To expand high-speed internet access in areas that currently lack it. The $42 billion appropriated for the program could have been used to deploy broadband to underserved areas and to foster broadband implementation. However, NTIA’s implementation of the program appears designed to dampen private investment and stifle competition among broadband, wireless, and satellite providers.

C. Digital-Discrimination Rules

One of the most problematic new regulations to hit the broadband sector is the FCC’s digital-discrimination rules. While well-intentioned, these rules are virtually certain to curtail broadband investment and adoption. In late 2023, the FCC adopted final rules facilitating equal access to broadband internet under Section 60506 of the IIJA.[84] The statutory text directs the FCC to prevent discrimination in broadband access based on income level, race, ethnicity, color, religion, or national origin, while also directing the commission to consider issues of technical and economic feasibility.

The rules prohibit digital discrimination of access, which is defined as policies or practices that differentially affect or are intended to differentially affect consumers’ access to broadband internet-access service based on their income level, race, ethnicity, color, religion or national origin, unless justified by genuine issues of technical or economic feasibility.[85] The are two key provisions that will disrupt broadband competition, namely:

  1. Adopting a disparate-impact standard to define “digital discrimination of access;” and
  2. Subjecting a “broad range” of service characteristics to digital-discrimination rules, including pricing, promotional conditions, terms of service, and quality of service.

The rules apply to entities that provide, facilitate, and/or affect consumer access to broadband internet-access service. This includes typical broadband providers, as well as entities that “affect consumer access to broadband internet access service.”[86] Under this broad definition, local governments, nonprofits, and even apartment-building owners all may be subject to the FCC’s digital-discrimination rules.

The rules also revise the commission’s informal consumer-complaint process to accept complaints of digital discrimination of access, and to authorize the commission to initiate investigations and impose penalties and remedies for violations of the rules.[87]

The FCC also proposed additional rules that would require providers to submit annual reports on their major deployment, upgrade, and maintenance projects, and to establish and maintain internal compliance programs to assess whether their policies and practices advance or impede equal access to broadband internet-access service within their service areas.[88] In essence, these proposed rules would require providers to prepare their own disparate-impact analysis every year.

Because of the expansive definition of covered entities and services subject to the digital-discrimination rules, providers will face legal uncertainty and litigation risks.[89] The most obvious of these involve the likelihood of complaints or investigations based on allegations of disparate impact, which may be difficult to disprove. Comments to the FCC from the U.S. Chamber of Commerce highlight these concerns:[90]

These policies would render it impossible for businesses and the marketplace to make rational investment decisions. The scope of the services that the Draft covers is so broad that it does not provide meaningful guidance for how to comply. And because the Draft fails to grant sufficient guidance, it does not give fair notice of how to avoid liability. Consequently, investment in broadband innovation would disappear and consumers would have to pay higher costs for less efficient services.

The digital-discrimination rules also may discourage innovation and differentiation in broadband service offerings, as providers could avoid service offerings that may be perceived as discriminatory or having a differential impact on certain consumers or communities. Providers could also be reluctant to invest in new technologies or platforms that, while improving broadband service quality or availability, might also create disparities in service characteristics among consumers or areas. As FCC Commissioner Brendan Carr has noted:[91]

Another telling last minute addition is a new advisory opinion process. This is the very definition of swapping out permissionless innovation for a mother-may-I pre-approval process. What’s more? The FCC undermines whatever value that type of process could provide because, to the extent the FCC does—at some point in the future—authorize your conduct, the Order says that the agency reserves the right to rescind an advisory opinion at any time and on a moment’s notice. At that time, the covered provider “must promptly discontinue” the practice or policy. That does not provide the confidence necessary to invest and innovate.

Private, public, and nonprofit entities may even face allegations of intentional discrimination for policies and practices designed to increase internet adoption and use by protected groups. In particular, programs intended to increase broadband adoption among low-income and price-sensitive consumers could run afoul of the digital-discrimination rules. George Ford provides an example of such a program:[92]

For example, Cox Communications offers 100 Mbps broadband service for $49.99 per month, but ACP eligible households can get the same service for $30 per month. Higher-income households may not avail themselves of the discounted price.

In Tennessee, Hamilton County Schools’ EdConnect program offers free high-speed internet access to eligible students, where eligibility is based on income level—i.e., students who receive free or reduced-cost lunch, attend any school where every student receives free or reduced-cost lunch, or whose family participates in the Supplemental Nutrition Assistance Program (SNAP) or other economic-assistance programs.[93] Both the school district and the nonprofit that runs the program would also be covered entities. The fact that the price (free) is available only to those of a certain income level is explicit, intentional discrimination.

The FCC’s digital-discrimination rules will almost surely increase the regulatory burden and compliance costs for providers. Small and rural providers may be disproportionately burdened, as these providers tend to have more limited resources and face technical and economic challenges in deploying and maintaining broadband networks in unserved and underserved areas. The FCC’s proposal that broadband providers submit an annual report on their substantial broadband projects could likewise give larger providers an advantage, as they are more likely to have the resources to comply with this requirement. For example, the Wireless Internet Service Providers Association commented to the FCC:[94]

Annual reporting and record retention rules and the requirement to adopt and certify to the existence and compliance with an internal digital discrimination compliance plan would impose significant burdens on broadband providers, especially smaller providers that may not track investment data and lack the resources to develop a compliance program with ongoing obligations. The burdens are overly egregious given that smaller providers do not have any record of engaging in digital discrimination.

Further complicating the evaluation of digital-discrimination claims based on income is that, not only is income a key factor influencing whether a given consumer will adopt broadband, but it is also highly correlated with race, ethnicity, national origin, age, education level, and home-computer ownership and usage. The FCC’s digital-discrimination rules fail to recognize this “income conundrum” and will invite costly and time-consuming litigation based on allegations of digital discrimination either where it does not exist or where it is excused by economic-feasibility considerations. Moreover, by specifying pricing as an area subject to digital-discrimination scrutiny, the FCC’s rules allow for ex-post regulation of rates, prompting Commissioner Carr to characterize the agency’s digital-discrimination rules and Title II rules as “fraternal twins.”[95]

D. Title II and Net Neutrality

In 2015, the FCC issued the Open Internet Order (OIO), which reclassified broadband internet-access service as a telecommunications service subject to Title II of the Communications Act. Proponents of the OIO contend that the Title II classification was necessary to ensure net neutrality—that is, that internet service providers (ISP) would treat all internet traffic equally. In 2018, the Title II classification was repealed by the FCC’s Restoring Internet Freedom Order (RIFO).

One month after ICLE’s white paper was published in 2021, President Joe Biden issued an executive order that “encouraged” the FCC to “[r]estore Net Neutrality rules undone by the prior administration.” Last year, Anna Gomez was confirmed as an FCC commissioner, providing the commission a 3-2 Democratic majority. One day after her confirmation, FCC Chair Rosenworcel announced the agency’s proposal to reimpose Title II regulation on internet services. Soon thereafter, the FCC issued its “Notice of Proposed Rulemaking for the Safeguarding and Securing the Open Internet Order,” which would again reclassify broadband under Title II.[96] On April 25, 2024, the commission approved the order on a 3-2 party-line vote.[97]

While the FCC provides several reasons for reclassifying broadband, most of the justifications are built on the same underlying premise: That broadband is an essential public utility and should be regulated as such. Of course, many other essentials—shelter, food, clothing—are provided by various suppliers in competitive markets. Utilities are considered distinct because they tend to have significant economies of scale such that:

  1. a single monopoly provider can provide the goods or services at a lower cost than multiple competing firms; and/or
  2. market demand is insufficient to support more than a single supplier.[98]

Under this definition, water, sewer, electricity, and natural gas constitute examples typically cited as “natural” monopolies.[99] In some cases, not only are these industries treated as regulated monopolies, but their monopoly status is solidified by laws forbidding competition.

At one time, local and long-distance telephone services were similarly treated as natural monopolies, as was cable television.[100] Various innovations eroded the “natural” monopolies in telephone and cable service over time.[101] As of the year 2000, 94% of U.S. households had a landline telephone, while only 42% had a mobile phone.[102] By 2018, those numbers flipped. In 2015, 73% of households subscribed to cable or satellite television service.[103] Today, fewer than half of U.S. households subscribe.[104] Much of that transition has been due to the enormous improvements in broadband speed, reliability, and affordability discussed in Section II. Similarly, innovations in 5G, fixed wireless, and satellite are eroding the already-tenuous claims that broadband internet service is akin to a utility.

The FCC’s latest reclassification of broadband under Title II prohibits blocking, throttling, or engaging in paid or affiliated prioritization arrangements.[105] In addition, it imposes “a general conduct standard that would prohibit unreasonable interference or unreasonable disadvantage to consumers or edge providers.” Under the OIO, the FCC invoked the general conduct standard to scrutinize providers’ “zero rating” programs.[106] Although Title II regulation explicitly allows for rate regulation of covered entities, the 2024 order forebears rate regulation.[107]

Critics of Title II regulation have argued that some of the conduct prohibited under the FCC’s proposal may be pro-competitive practices that benefit consumers. For example, Hyun Ji Lee & Brian Whitacre found that low-income consumers were willing to pay for an extra GB of data each month, but were not willing to pay extra for a higher speed.[108] This data-speed tradeoff suggests those consumers would benefit from a plan that offered a larger data allowance, but throttled speeds if the allowance is exceeded. In 2014 comments to the FCC, ICLE and TechFreedom described a pro-competitive benefit of paid prioritization:[109]

Prioritization at least requires content providers to respond to incentives—to take congestion into account instead of using up a common resource without regard to cost. It also allows the gaming company to buy better service, which isn’t an option at all with neutrality, under which it just has to suffer congestion. The truth is that, if the game developer can’t afford to pay for clear access, then it may have a bad business model if it is built on an expectation that it will have unfettered, free access to a scarce, contestable resource.

Aside from the likely pro-competitive effects of the conduct the FCC seeks to prohibit, in the face of robust competition, consumers can readily switch away from providers who charge anticompetitive prices or impose harmful terms and conditions. In its 2019 Mozilla decision, the U.S. Circuit Court of Appeals for the D.C. Circuit concluded:[110]

[M]any customers can access edge provider’s content from multiple sources (i.e., fixed and mobile). In this way, there is no terminating monopoly. Additionally, the Commission argued that even if a terminating monopoly exists for some edge providers the commenters did not offer sufficient evidence in the record to demonstrate that the resulting prices will be inefficient. Given these reasons, we reject Petitioners’ claim that the Commission’s conclusion on terminating monopolies is without explanation.

In addition, the court noted:[111]

More importantly, the Commission contends that low churn rates do not per se indicate market power. Instead, they could be a function of competitive actions taken by broadband providers to attract and retain customers. And such action to convince customers to switch providers, the Commission argues, is indicia of material competition for new customers.

Regardless of the FCC’s intent in imposing Title II regulation, the effect will be a stifling of innovation in the delivery and pricing of broadband-internet service. In tandem with the agency’s digital-discrimination rules, the proposed “net neutrality” rules attempt to transition broadband to a commodity service with little differentiation between providers. In so doing, the FCC is eliminating, piece-by-piece, the dimensions among which broadband providers compete, resulting in both higher prices for consumers and lower returns for providers. Rather than a “virtuous cycle” of growth and innovation, the U.S. broadband market may instead experience a “doom loop” of stagnant internet adoption, depressed investment in deployment, and diminished broadband competition.

E. De-Facto Rate Regulation

Rate regulation—any mechanism whereby government intervenes in the pricing process—has long been a contentious issue in the realm of broadband services.[112] Historically, the FCC has been deeply involved in rate regulation, tasked with ensuring fair rates, reliable service, and universal access to telecommunications since 1934.[113] As the telecommunications landscape has evolved, however, so too has the FCC’s approach, increasingly moving toward deregulatory approaches. That is, until recently.[114] Unfortunately, there are multiple ways that rates can be regulated, and—despite public disavowals—policymakers already appear to be implementing some forms of rate regulation on broadband providers.

Explicit rate regulation manifests primarily in two forms: price ceilings and floors.[115] Price ceilings limit the maximum price that can be charged, a common example being rent control. Price floors, on the other hand, set a minimum price, akin to minimum wage laws. Each of these forms impacts the broadband sector differently, potentially altering market dynamics and influencing consumer access and provider revenues.[116]

Policymakers can also resort to less-obvious means of regulating prices—de-facto rate regulation—such as rent stabilization or inflation-linked wage increases, which control the rate of price changes rather than the prices themselves.[117] Moreover, as discussed further infra, price controls are sometimes introduced laterally as requirements to participate in various federal programs, with the effect remaining that government agents assume broad control over prices. Still other regulations may not explicitly regulate rates, but act in much the same way as direct rate regulation, as explained by Jonathan Nuechterlein and Howard Shelanski:[118]

Finally, but no less important, the line between “price” and “non-price” regulation is thin, and regulatory obligations can amount to rate regulation even when regulators do not perceive themselves as setting rates at either the retail or wholesale level.

The FCC’s 2015 OIO, while explicitly eschewing rate regulation, indirectly influenced pricing strategies in the broadband market.[119] By imposing common-carriage obligations, the OIO impacted how ISPs invested and priced their services. In this respect, the FCC’s 2024 rules are identical to the 2015 rules. But this time, Title II regulation will work hand-in-hand with the agency’s digital-discrimination rules. While the proposed common-carrier rules explicitly eschew ex-ante rate regulation through forbearance, the digital-discrimination rules explicitly subject pricing policies and practices to ex-post discrimination scrutiny.

In some ways, the FCC may be imposing among the worst of possible rate-regulation regimes. Under an ex-ante approach to rate regulation, providers have—at a minimum—a framework to form their expectations about whether and how rates will be regulated. As discussed in Section III.C, however, under the ex-post approach that the FCC has adopted in its digital-discrimination rules, providers and any other “covered entity” lack any meaningful framework regarding how the agency may regulate rates or how to avoid liability.

Specifically, the FCC’s Digital Discrimination Order states:

The Commission need not prescribe prices for broadband internet access service, as some commenters have cautioned against, in order to determine whether prices are “comparable” within the meaning of the equal access definition. The record reflects support for the Commission ensuring pricing consistency as between different groups of consumers. We also find that the Commission is well situated to analyze comparability in pricing, as we must already do so in other contexts.[120]

While assessing the comparability of prices is not explicit rate regulation, a policy that holds entities liable for those disparities, such that an ISP must adjust its prices until it matches the FCC definitions of “comparable” and “consistency,” is tantamount to setting that rate.[121]

In addition to the FCC digital-discrimination and Title II rules, recent developments in broadband policy have introduced other forms of de-facto rate regulation. The BEAD program itself mandates a “low-cost” option be made available to recipients of the Affordable Connectivity Program by providers that receive a BEAD grant.[122] The NTIA’s NOFO for the BEAD program further mandates that participating states include an affordability plan that ensures access to affordable high-speed internet for all middle-class consumers.[123] This initiative might require providers to offer low-cost plans or to provide consumer subsidies. Similarly, the U.S. Department of Agriculture’s (USDA) ReConnect Loan and Grant Program awards funding preferences to applicants that adhere to net-neutrality rules and offer “affordable” options.[124] New York’s Affordable Broadband Act is another example of broadband rules that mandate ISPs provide low-cost internet-access plans to qualifying low-income households.[125]

Rate regulation, de facto or otherwise, has a major effect on providers’ ability to enter new markets and to improve service in those markets in which they already operate. Rate regulations lead to market distortions. By capping prices below the market rate, such regulations can increase demand without a corresponding increase in supply, potentially leading to shortages and discouraging providers from making output-improving investments.[126] For broadband providers, this can translate into reduced investment in network expansion and quality improvement, particularly in less profitable or more challenging areas. Moreover, binding rate regulations can lower the returns on investment, thereby discouraging deployments and slowing overall broadband expansion. Quality and service also may suffer under rate regulation. A regulated provider, constrained by price ceilings, cannot fully reap the benefits of service-quality improvements, leading to a reduced incentive to enhance that service quality.[127]

F. Pole Attachments

The importance of pole attachments cannot be overstated in the context of expanding broadband connectivity, even if utility-pole issues often fly under the radar. This is particularly true due to their implications for competition in the relevant local broadband markets. Access to physical infrastructure is critical, and where providers cannot readily access this physical infrastructure, it can delay deployment or make it more costly.

The FCC has recognized the crucial role of pole attachments in a pending proceeding that seeks to address inefficiencies in access to pole attachments that lead to cost overruns and delays in deployment.[128] In December 2023, in an effort to expedite broadband deployment, the commission adopted several important pole-attachment reform measures.[129] These included introducing a streamlined process to resolve utility-pole attachment disputes, which could be pivotal to hasten broadband rollouts, especially in underserved areas.[130] The FCC also mandated that utilities provide comprehensive pole-inspection information to broadband attachers, which is expected to facilitate more informed planning and to reduce delays.[131] The commission has also refined its procedural rules to foster quicker resolutions through mediation and expedited adjudication via the Accelerated Docket.[132]

The FCC is on the right track: ensuring timely access to pole infrastructure is crucial to ensure that broadband markets remain competitive, and that the substantial investments in broadband infrastructure directed by programs like BEAD yield the intended benefits.

The goal of pole-attachment rules should be to equitably assess costs in ways that avoid inefficient rent extraction and ensure the smooth deployment of broadband infrastructure.[133] The FCC’s current rules, however, can impose on a requesting attacher the entire cost of pole replacement, which is economically suboptimal.[134] There is therefore a need to revisit the current formula to ensure that the incremental costs and benefits are appropriately allocated to each relevant party. In its recent order, the FCC expanded the definition of what constitutes as a “red tagged” pole in need of replacement.[135] The extent to which this works in practice will, however, depend on how the FCC processes applications under its new “red tag” policy.

One critical concern is the emergence of hold-up and hold-out problems.[136] Section 224 of the Communications Act authorizes the FCC to ensure that the costs of pole attachments are just and reasonable.[137] This provision, however, also allows pole owners to deny access when there is insufficient capacity, creating a potential imbalance in bargaining power.[138] This imbalance is exacerbated by the pole owners’ superior knowledge of their cost structures and their ability to impose “take it or leave it” offers on prospective attachers.[139] Consequently, attachers might be, at the margin, discouraged from deploying in areas with capacity-constrained poles. Further, the “last attacher pays” model can inadvertently create a disincentive for pole owners to replace or upgrade poles until a new attacher is obligated to bear the full cost. This scenario may lead to delays in broadband deployment, especially in areas where the cost of deployment is already high. The recent FCC order aims to address these concerns by clarifying cost-causation principles and ensuring more equitable cost sharing for pole replacements and modifications.[140] But there again remains interpretive room within the framework the commission has established. Thus, it remains to be seen how effectively the new rules will mitigate the problem.

Any reconsideration of pole-attachment rules also must account for the fact that the pole market is highly regulated.[141] The actual cost for pole replacements in a free market, without regulatory intervention, would likely be some middle ground between the total replacement cost and the new rental price charged to attachers. The FCC must judiciously leverage its ability to set reasonable rental rates to approach the ideal price that would otherwise be discovered through market mechanisms.

Toward this end, the upfront “make-ready” charges for pole replacement should be limited to a pole owner’s incremental cost.[142] This approach acknowledges that early replacements simply shift the timing of the expense, rather than adding additional costs. The formula could incorporate the depreciated value of the pole being replaced and allocate the costs associated with increased capacity across all beneficiaries, including new attachers as well as the pole owner, who may realize additional revenue from the increased capacity.

Beyond disputes over privately owned poles, a lacuna in the FCC’s authority over poles owned by certain public entities threatens to erect large roadblocks to deployment. This is particularly the case for poles owned by the Tennessee Valley Authority (TVA).[143]  Such common TVA practices as refusing reasonable and nondiscriminatory pole-attachment agreements risk significantly slowing the deployment of broadband, especially in the rural areas the TVA services.[144]

The source of this problem is a provision of Section 224 of the Communications Act that exempts municipal and electric-cooperative (coop) pole owners from FCC oversight.[145] This exemption allows the TVA to set its own rates for pole attachments, which are notably higher than FCC rates, and often sidestep access requirements typically mandated by states and the FCC.[146]

Municipally owned electricity distributors constitute what economists call state-owned enterprises. As such, they face significantly different restraints than privately owned enterprises.[147] Private businesses must pass the profit-and-loss test on the market, while state-owned enterprises are not similarly constrained. Municipally owned electricity distributors are usually monopolies, either because private competitors are not allowed to compete, or because they receive government benefits not available to potential private competitors. As a result, they may pursue other goals in the “public interest,” such as providing their products and services at below-market prices.[148] This includes the ability to leverage their electricity monopolies to enter into broadband provision. The problem is that these municipally owned electricity distributors also have strong incentives to refuse to deal with private competitors in the broadband market who need access to the electric poles they own.[149]

Rural electric cooperatives (RECs), particularly those distributing electricity from the TVA, also hold a privileged position that allows them to act in potentially anticompetitive ways toward broadband providers seeking pole attachments. Unlike municipally owned electricity distributors, RECs need to earn sufficient revenues to remain operational. They are also, however, much more like state-owned enterprises in the governmental benefits they receive, including the immense difficulty of normal oversight from the market for corporate control.[150] This similarly incentivizes them to act anticompetitively, particularly as many enter or plan to enter the broadband market.[151]

These circumstances often lead RECs to refuse to deal with private broadband providers, thereby stifling competition and deployment in rural areas.[152] Furthermore, RECs often face little oversight from rate regulators regarding pole attachments, leading to significantly higher costs for broadband companies seeking to attach to poles owned by co-ops and municipalities outside FCC jurisdiction.[153]

This regulatory loophole not only leads to higher costs for broadband providers, but also raises concerns about the application of antitrust laws to these entities. Sen. Mike Lee (R-Utah) has argued that the U.S. Justice Department (DOJ) should examine the antitrust implications of these practices, emphasizing that these government-owned entities should be subject to antitrust laws when acting as market participants.[154] And FCC Commissioner Brendan Carr has noted ongoing concerns about delays and costs associated with attaching to poles owned by municipal and cooperative utilities.[155] Addressing this loophole is crucial to bridge the digital divide and ensure that the IIJA’s goals are met effectively.

G. Municipal/Co-Op Broadband

As previously noted, despite persistent interest in some quarters to promote municipal broadband,[156] there are many challenges that contribute to such projects’ poor record. In particular, the financial prospects of municipal networks are typically dim, as many such projects generate negative cash flow and are unsustainable without substantial improvements in operations.[157] Only a small subset of municipalities—usually those with existing municipal-power utilities—might be well-positioned to venture into municipal broadband, due to potential cross-subsidization opportunities.[158] Even among those municipal-broadband projects that have been deemed successful, however, the repayment of project costs is daunting, often requiring substantial subsidies and cross-subsidization.[159] The prospects for municipal broadband have not improved since ICLE’s 2021 white paper.

In a study by Christopher Yoo et al., the authors examine the financial performance of every municipal fiber project operating in the United States from 2010 through 2019 that provided annual financial reports for its fiber operations.[160] Each of the 15 projects was located in an urban area, as defined by the U.S. Census Bureau. In addition, each project was built in areas already served by one or more private broadband providers—none were designed to serve previously unserved areas. In every case, the municipality issued revenue bonds to fund construction and initially expected the projects to repay their construction and operating costs from project revenues, rather than from taxes or interfund transfers. In some cases, the cities anticipated the projects would generate surpluses that would, in turn, allow the cities to lower taxes.

In contrast to these expectations, every project either needed infusions of cash from outside sources or debt relief through refinancing. Three projects defaulted on their debt, two of which were liquidated at significant losses.

Yoo et al. employed two measures of financial performance:

  1. adjusted net cash flow (ANCF), which measures the actual cash collected and spent by a fiber project; and
  2. net present value of cash flow from operations (NPV), which discounts cash flow using the project’s weighted average cost of capital.

Based on ANCF, only two of the 15 projects have broken even or are expected to break even by the time their initial debt matures. Based on NPV, more than half of the projects were not on track to break even—even assuming a theoretical best-case performance in terms of capital expenditures and debt service.

Municipalities that are unable to cover their broadband projects’ costs of debt and operations must make up the shortfall from general tax revenues or default on their debt. Making up a shortfall from tax revenues means the city must enact some combination of tax increases or service cuts. A default will result in a downgrade in the municipality’s bond rating, which will increase the costs of financing all of the city’s operations, not just the broadband project. These additional costs must ultimately be paid the municipality’s taxpayers.

In a separate analysis, George Ford notes that many municipal-broadband projects are located in cities that operate their own electric utilities.[161] Such an arrangement allows the broadband network’s debt and other expenses to be placed on the electric utility’s books, thereby improving the apparent financial condition of the broadband network. As electricity rates are based on cost of service, Ford argues that a shift of broadband costs to the electric utility would be expected to increase electricity rates.

To evaluate this hypothesis, he compares municipal electricity rates among four Tennessee cities that own and operate municipal broadband. Two cities financed the projects with general-obligation bonds funded by tax revenues and other sources of the municipality’s income. The other two cities used electric-utility profits to cover the broadband project’s financial losses. One of these cities is Chattanooga, which received $111 million in subsidies and in which the city’s electric utility assumed $162 million of debt to construct the broadband network and made $50 million of loans to the broadband division.

Ford’s statistical analysis calculates broadband projects are associated with a 5.4% increase in electricity rates in cities with utility-funded projects, relative to cities that issued general-obligation bonds. It should be emphasized that the higher rates are imposed on all electricity ratepayers, not just those who subscribe to the city’s broadband. These higher electricity rates are used to cross-subsidize municipal-broadband subscribers. For example, Ford reports that, in Chattanooga, the average monthly revenue per broadband subscriber was $147 in 2015. In addition, the average subscriber was associated with a monthly subsidy of $30. Thus, cross-subsidies from electricity ratepayers account for about 17% of the average monthly broadband-subscriber cost.

The conclusions from ICLE’s 2021 white paper remain valid today. Proposals to offer municipal broadband as a means to increase broadband adoption—either by attempting to increase supply, or to suppress prices—put the cart before the horse. That’s because private supply and demand conditions are usually sufficient to guarantee creation of adequate broadband networks throughout most of the country.

Some uneconomic locations (i.e., the unserved areas) may require interventions to ensure broadband access. In some cases, municipal broadband may be an effective option to subsidize hard-to-reach consumers. Municipal broadband should not, however, be considered the best or only option. Indeed, the evidence demonstrates that municipal broadband might best be considered a solution of last resort, used only when no private provider finds it economically viable to serve a particular area.

IV. Conclusion

By most measures, U.S. broadband competition is vibrant and has increased dramatically since the COVID-19 pandemic. Since 2021, more households are connected to the internet; broadband speeds have increased, while prices have declined; more households are served by more than a single provider, and new technologies like satellite and 5G have expanded internet access and intermodal competition among providers.

Broadband competition policy currently appears to be in a state of confusion: Some policies foster competition, while others hinder it. Programs such as the ACP and BEAD could do much to encourage competition by simultaneously increasing the demand for broadband and facilitating the buildout of supply. At the same time, some facets of these programs’ implementation act to stifle competition with onerous rules, reporting requirements, and—in some cases—de-facto rate regulation.

In addition, the FCC’s digital-discrimination rules explicitly subject broadband pricing and other dimensions of competition to ex-post scrutiny and enforcement. In reclassifying broadband internet-access services under Title II of the Communications Act, the FCC has rendered nearly every aspect of broadband deployment and delivery subject to its regulation or scrutiny.

Put simply, today, U.S. broadband competition is robust, innovative, and successful. At the same time, new and forthcoming regulations threaten broadband competition by eliminating or proscribing the policies and practices by which providers compete. As a result, the United States is at risk of slowing or shrinking broadband investment—thereby reducing innovation and harming the very consumers that policymakers claim they seek to help.

[1] Geoffrey A. Manne, Kristian Stout, & Ben Sperry, A Dynamic Analysis of Broadband Competition: What Concentration Numbers Fail to Capture, Int’l Ctr. for L. & Econ. (Jun. 2021), available at

[2] See id. at 2-3; 35-37.

[3] CDC Museum COVID-19 Timeline, Ctr. for Disease Control and Prevention (Mar. 15, 2023),

[4] H.R. 3684, 117th Cong. (2021).

[5] Eric Fruits & Kristian Stout, Finding Marginal Improvements for the ‘Good Enough’ Affordable Connectivity Program, Int’l Ctr. for L. & Econ. (Sep. 15, 2023), available at

[6] Eric Fruits & Geoffrey A. Manne, Quack Attack: De Facto Rate Regulation in Telecommunications, Int’l Ctr. for L. & Econ. (Mar. 30, 2023), available at

[7] 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

[8] Wireline Competition Bureau Announces the Final Month of the Affordable Connectivity Program, WC Docket No. 21-450 (Mar. 4, 2024), available at; see also Brian Fung, FCC Ends Affordable Internet Program Due to Lack of Funds, CNN (May 31, 2024),

[9] Anthony Hennen, More Money, More Problems for National Broadband Expansion, The Center Square (Aug. 15, 2023),

[10] Lindsay McKenzie, BEAD Waiver Information Coming This Summer, NTIA Says, StateScoop (Aug. 17, 2023),

[11] BEAD Letter of Credit Concerns, $4.3M in ACP Outreach Grants, FCC Waives Rules for Hawaii Wildfires, Broadband Breakfast (Aug. 21, 2023),

[12] Eric Fruits, Red Tape and Headaches Plague BEAD Rollout, Truth on the Market (Aug. 17, 2023),

[13] Fruits & Stout, supra note 6; 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, Int’l Ctr. for L. & Econ. (Apr. 20, 2023),

[14] FCC, Report and Order and Further Notice of Proposed Rulemaking on Implementing the Infrastructure Investment and Jobs Act: Prevention and Elimination of Digital Discrimination, GN Docket No. 18-238, FCC 19-44 (Nov. 20, 2023), available at [hereinafter “Digital Discrimination Order”]. See also Eric Fruits, Everyone Discriminates Under the FCC’s Proposed New Rules, Truth on the Market (Oct. 30, 2023), (reporting that, under the rules, “broadband service” includes every element of a consumer’s broadband-internet experience, including speeds, data caps, pricing, and discounts, and that the rules broadly apply to broadband providers as well as to “entities outside the communications industry” that “provide services that facilitate and affect consumer access to broadband,” which may include municipalities and property owners).

[15] Notice of Proposed Rulemaking, Safeguarding and Securing the Open Internet, WC Docket No. 23-320 (Sep. 28, 2023). [hereinafter “Title II NPRM”]

[16] Declaratory Ruling, Order, Report and Order, and Order on Reconsideration, Safeguarding and Securing the Open Internet, WC Docket No. 23-320, WC Docket No. 17-108 (adopted Apr. 25, 2024), available at [hereinafter “SSOIO” or “2024 Order”].

[17] See, e.g., Karl Bode, Colorado Eyes Killing State Law Prohibiting Community Broadband Networks, TechDirt (Mar. 30, 2023), (local broadband monopolies are a “widespread market failure that’s left Americans paying an arm and a leg for what’s often spotty, substandard broadband access.”).

[18] FCC Chair Rosenworcel on Reinstating Net Neutrality Rules, C-Span (Sep. 25, 2023), (“Only one-fifth of the country has more than two choices at [100 Mbps download] speed. So, if your broadband provider mucks up your traffic, messes around with your ability to go where you want and do what you want online, you can’t just pick up and take your business to another provider. That provider may be the only game in town.”).

[19] U.S. Census Bureau, 2021 American Community Survey 1-Year Estimates, Table Id. S2801 (2021); U.S. Census Bureau, ACS 1-Year Estimates Public Use Microdata Sample 2021, Access to the Internet (ACCESSINET) (2021).

[20] In contrast, a 2021 NTIA survey reports that 14.4% of households do not use the internet at home, with three-quarters of these households indicating they have “no need/interest” and one quarter indicating it is “too expensive.” See, Michelle Cao & Rafi Goldberg, Switched Off: Why Are One in Five U.S. Households Not Online?, National Telecommunications and Information Administration (2022),

[21] National Center for Education Statistics, Children’s Internet Access at Home, Condition of Education, (U.S. Department of Education, Institute of Education Sciences, Aug. 2023),

[22] See FCC, 2015 Broadband Progress Report (2015), (upgrading the standard speed from 4/1 Mbps to 25/3 Mbps). In March 2024, the FCC approved a report increasing the fixed-speed benchmark to 100/20 Mbps and setting an “aspirational goal” of 1 Gbps/500 Mbps. See, FCC, In the Matter of Inquiry Concerning the Deployment of Advanced Telecommunications Capability to All Americans in a Reasonable and Timely Fashion, GN Docket No. 22-270 (Mar. 14, 2024), available at In November 2023, FCC Chair Jessica Rosenworcel proposed reaching a 1 Gbps/500 Mbps benchmark by the year 2030. See Eric Fruits, Gotta Go Fast: Sonic the Hedgehog Meets the FCC, Truth on the Market (Nov. 3, 2023),

[23] Infrastructure Investment and Jobs Act, Pub. L. No. 117-58, § 60102 (a)(1)(A)(ii), 135 Stat. 429 (Nov. 15, 2021), available at; Jake Varn, What Makes a Community “Unserved” or “Underserved” by Broadband?, Pew Charitable Trusts (May 3, 2023), available at–and-underserved-definitions-ta-memo-pdf.pdf.

[24] Id., IIJA.

[25] Mike Conlow, New FCC Broadband Map, Version 3, Mike’s Newsletter (Nov. 20, 2023),

[26] FCC, Communications Marketplace Report, GN Docket No 22-203, FCC 22-103, Appendix G (Dec. 20, 2022),

[27] Pursuant to the IIJA, the FCC and providers are working to provide new broadband-coverage maps. These numbers will change over time, but FCC Chair Jessica Rosenworcel noted: “Looking ahead, we expect that any changes in the number of locations will overwhelmingly reflect on-the-ground changes such as the construction of new housing.” See Brad Randall, FCC’s Updated Broadband Map Shows Increasing National Connectivity, Broadband Communities (Nov. 27, 2023),

[28] FCC Chair Rosenworcel on Reinstating Net Neutrality Rules, C-Span (Sep. 26, 2023),

[29] FCC, Fixed Broadband Deployment (Jun. 2021),

[30] FCC, 2019 Broadband Deployment Report, GN Docket No. 18-238, FCC 19-44 at Fig. 4 (May 29, 2019), available at

[31] The FCC does not explain the differences between the information summarized in Table 1 and Table 2. The differences likely reflect different methodologies. For example, Table 1 may be at the household level and Table 2 at the population level.

[32] 2022 Communications Marketplace Report, GN Docket No. 22-203 (Dec. 30, 2022) at Fig. II.A.28, available at

[33] Dan Heming, Starlink No Longer Has a Waitlist for Standard Service, and 10 MPH Speed Enforcement Update, Mobile Internet Resource Center (Oct. 3, 2023),,order%20anywhere%20in%20the%20USA.

[34] Starlink Specifications, Starlink,

[35] Amazon Shares an Update on How Project Kuiper’s Test Satellites Are Performing, Amazon (Oct. 16, 2023),

[36] Kuiper Service to Start by End of 2024: Amazon, Communications Daily (Oct. 12, 2023),

[37] Why Is the Internet More Expensive in the USA than in Other Countries?, Community Tech Network (Feb. 2, 2023),

[38] Dan Howdle, Global Broadband Pricing League Table 2023, (2023),, data available at

[39] This is qualitatively consistent with the FCC’s finding that United States has the seventh-lowest prices per gigabit of data consumption, and that Australia has the 12th-lowest among OECD countries. FCC, 2022 Communications Marketplace Report, Docket No. 22-103, Appendix G (Dec. 30, 2022), available at

[40] Median Country Speeds, Speedtest Global Index (Oct. 2023), (last visited Dec. 7, 2023).

[41] See Christian Dippon, et al., Adding a Warning Label to Rewheel’s International Price Comparison and Competitiveness Rankings (Nov. 30, 2020), available at

[42] Fruits & Stout, supra note 6; see also Giuseppe Colangelo, Regulatory Myopia and the Fair Share of Network Costs: Learning from Net Neutrality’s Mistakes, Int’l Ctr. for L. & Econ. (Comments to European Commission Exploratory Consultation, The Future of the Electronic Communications Sector and Its Infrastructure, May 18, 2023),

[43] Id. at 14.

[44] Arthur Menko Business Planning Inc., 2023 Broadband Pricing Index, USTelecom (Oct. 2023), available at

[45] U.S. Bureau of Labor Statistics, Producer Price Index by Commodity: Telecommunication, Cable, and Internet User Services: Residential Internet Access Services [WPU374102], retrieved from FRED, Federal Reserve Bank of St. Louis (Aug. 29, 2023),

[46] United States Median Country Speeds July 2023, Speedtest Global Index (2023), Prior years retrieved from Internet Archive. See also Camryn Smith, The Average Internet Speed in the U.S. Has Increased by Over 100 Mbps since 2017, Allconnect (Aug. 4, 2023), (average download speed in the United States was 30.7 Mbps in 2017 and 138.9 Mbps in the first half of 2023).

[47] George S. Ford, Confusing Relevance and Price: Interpreting and Improving Surveys on Internet Non-adoption, 45 Telecomm. Pol’y 102084 (2021).

[48] Smaller surveys and focus groups that allow more opportunities for follow-up questions, however, suggest that price may be more important than is suggested by Census Bureau surveys. For example, one study in Detroit, Michigan, used surveys and focus groups to examine internet adoption and use in three low-income urban neighborhoods. Participants who reported lacking at-home internet mentioned lack of interest and high costs at roughly equal rates. See, Colin Rhinesmith, Bianca Reisdorf, & Madison Bishop, The Ability to Pay For Broadband, 5 Comm. Res. Pract. 121 (2019).

[49] Ford, supra note 9.

[50] Michelle Cao & Rafi Goldberg, New Analysis Shows Offline Households Are Willing to Pay $10-a-Month on Average for Home Internet Service, Though Three in Four Say Any Cost Is Too Much, National Telecommunications and Information Administration (Oct. 6, 2022),

[51] Michelle Cao & Rafi Goldberg, Switched Off: Why Are One in Five U.S. Households Not Online?, National Telecommunications and Information Administration (2022),

[52] Jamie Greig & Hannah Nelson, Federal Funding Challenges Inhibit a Twenty-First Century “New Deal” for Rural Broadband, 37 Choices 1 (2022).

[53] Andrew Perrin, Mobile Technology and Home Broadband 2021, Pew Research Center (Jun. 3, 2021),

[54] Rhinesmith, et al., supra note 10.

[55] 2022 Broadband Capex Report, USTelecom (Sep. 8, 2023), available at

[56] Wolfgang Briglauer, Carlo Cambini, Klaus Gugler, & Volker Stocker, Net Neutrality and High-Speed Broadband Networks: Evidence from OECD Countries, 55 Eur. J. L. & Econ. 533 (2023).

[57] Eric Fruits, Justin (Gus) Hurwitz, Geoffrey A. Manne, Julian Morris, & Alec Stapp, Static and Dynamic Effects of Mergers: A Review of the Empirical Evidence in the Wireless Telecommunications Industry, (OECD Directorate for Financial and Enterprise Affairs Competition Committee, Global Forum on Competition, DAF/COMP/GF(2019)13, Dec. 6, 2019), available at

[58] Manne, Stout, & Sperry, supra note 1.

[59] Kenneth Flamm & Pablo Varas, Effects of Market Structure on Broadband Quality in Local U.S. Residential Service Markets, 12 J. Info. Pol’y 234 (2022).

[60] Andrew Kearns, Does Competition From Cable Providers Spur the Deployment of Fiber? (Jul. 27, 2023), or

[61] Manne, Stout, & Sperry, supra note 1.

[62] Fruits, et al., supra note 55.

[63] FCC, Affordable Connectivity Program (Oct. 2, 2023),

[64] Eric Fruits & Kristian Stout, Finding Marginal Improvements for the ‘Good Enough’ Affordable Connectivity Program (Int’l. Ctr. for L. & Econ. Issue Brief, Sep. 15, 2023), available at

[65] See Paul Winfree, Bidenomics Goes Online: Increasing the Costs of High-Speed Internet, Econ. Pol’y Innovation Ctr (Jan. 8, 2024), available at (Finding ACP subsidies are associated with higher prices for all broadband plans, especially lower-speed plans, but these costs are more than offset by the subsidies for those who receive them. Thus, the ACP provides lower prices net of subsidy to ACP beneficiaries, but higher prices for those who are not.).

[66] Id.

[67] Id.

[68] Fruits & Stout, supra note 4.

[69] Universal Service Administrative Co., ACP Enrollment and Claims Tracker (Feb. 8, 2024), Beginning Feb. 8, 2024, the ACP ceased enrollment.

[70] Wireline Competition Bureau Announces the Final Month of the Affordable Connectivity Program, WC Docket No. 21-450 (Mar. 4, 2024), available at

[71] Biden-Harris Administration Announces State Allocations for $42.45 Billion High-Speed Internet Grant Program as Part of Investing in America Agenda, Nat’l Telecomms and Info. Admin. (Jun. 26, 2023),

[72] Id.

[73] U.S. Dep’t of Com., Internet For All Frequently Asked Questions and Answers Draft Answers Version 2.0 Broadband, Equity, Access, and Deployment (BEAD) Program, Nat’l Telecomms and Info. Admin. (Sep. 2022), available at

[74] Infrastructure Investment and Jobs Act Overview, BroadbandUSA, (last visited Dec. 7, 2023).

[75] U.S. Dep’t of Com., Notice of Funding Opportunity, Broadband Equity, Access, and Deployment Program, NTIA-BEAD-2022, Nat’l Telecomms and Info. Admin. (May 2022), available at [hereinafter “BEAD NOFO”]

[76] Id. See also, Ted Cruz, Red Light Report, Stop Waste, Fraud, and Abuse in Federal Broadband Funding, U.S. S. Comm. on Com., Science, and Transp. (Sep. 2023),

[77] U.S. Dep’t of Com., Notice of Funding Opportunity, Broadband Equity, Access, and Deployment Program, NTIA-BEAD-2022, NTIA (May 2022), available at (note that the IIJA itself did not include this requirement, as it was an addition 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).

[78] Id. at 67.

[79] George S. Ford, Middle-Class Affordability of Broadband: An Empirical Look at the Threshold Question, Phoenix Ctr. for Adv. Leg. & Econ. Pub. Pol’y Stud., Pol’y Bull. No. 61 (Oct. 2022), available at

[80] Id.

[81] John W. Mayo, Gregory L. Rosston & Scott J. Wallsten, From a Silk Purse to a Sow’s Ear? Implementing the Broadband, Equity, Access and Deployment Act, Geo. U. McDonough Sch. of Bus. Ctr. for Bus. & Pub. Pol’y (Aug. 2022),

[82] BEAD Letter of Credit Concerns, $4.3M in ACP Outreach Grants, FCC Waives Rules for Hawaii Wildfires, Broadband Breakfast (Aug. 21, 2023),

[83] NTIA, Ensuring Robust Participation in the BEAD Program (Nov. 1, 2023),

[84] FCC, Report and Order and Further Notice of Proposed Rulemaking, GN Docket No. 22-69, FCC 23-100 (Nov. 20, 2023), available at

[85] Id. at 3.

[86] Id.

[87] Id.

[88] Id.

[89] Fruits, supra note 13.

[90] U.S. Chamber of Commerce, In the Matter of Implementing the Infrastructure Investment and Jobs Act: Prevention and Elimination of Digital Discrimination, GN Docket No. 22-69 (Nov. 6, 2023), (citations omitted).

[91] FCC, Dissenting Statement of Commissioner Brendan Carr Regarding the Implementing the Infrastructure Investment and Jobs Act: Prevention and Elimination of Digital Discrimination, GN Docket No. 22-69, Report and Order and Further Notice of Proposed Rulemaking, FCC 23-100 (2023), available at

[92] George S. Ford, Will Digital Discrimination Policies End Discount Plans for Low-Income Consumers? (Phoenix Ctr. for Advanced Legal & Econ. Pub. Pol’y Stud., Nov. 1, 2023),

[93] HCS EdConnect, Welcome to HCS EdConnect (2023),

[94] WISPA, In the Matter of Implementing the Infrastructure Investment and Jobs Act: Prevention and Elimination of Digital Discrimination, GN Docket No. 22-69 (Nov. 8, 2023),

[95] Testimony of Brendan Carr, Commissioner, Federal Communications Commission, Before the Subcommittee on Communications and Technology of the United States House of Representatives Committee on Energy and Commerce, “Oversight of President Biden’s Broadband Takeover” (Nov. 30, 2023), available at

[96] Title II NPRM, supra note 14.

[97] SSOIO, supra note 16.

[98] See, Paul Krugman & Robin Wells, Economics (4th ed. 2015) at 389 (“So the natural monopolist has increasing returns to scale over the entire range of output for which any firm would want to remain in the industry—the range of output at which the firm would at least break even in the long run. The source of this condition is large fixed costs: when large fixed costs are required to operate, a given quantity of output is produced at lower average total cost by one large firm than by two or more smaller firms.”)

[99] Id. (“The most visible natural monopolies in the modern economy are local utilities—water, gas, and sometimes electricity. As we’ll see, natural monopolies pose a special challenge to public policy.”)

[100] Richard H. K. Vietor, Contrived Competition (1994) at 167 (“[I]n the early part of the twentieth century, American Telephone and Telegraph (AT&T) set itself the goal of providing universal telephone services through an end-to-end national monopoly. … By [the 1960s], however, the distortions of regulatory cross-subsidy had diverged too far from the economics of technological change.”). Thomas W. Hazlett, Cable TV Franchises as Barriers to Video Competition, 2 Va. J.L. & Tech. 1 (2007) (“Traditionally, municipal cable TV franchises were advanced as consumer protection to counter “natural monopoly” video providers. …  Now, marketplace changes render even this weak traditional case moot. … [V]ideo rivalry has proven viable, with inter-modal competition from satellite TV and local exchange carriers (LECs) offering “triple play” services.”)

[101] Id.

[102] Share of United States Households Using Specific Technologies, Our World in Data (n.d.),

[103] Edward Carlson, Cutting the Cord: NTIA Data Show Shift to Streaming Video as Consumers Drop Pay-TV, NTIA (2019),

[104] Karl Bode, A New Low: Just 46% Of U.S. Households Subscribe to Traditional Cable TV, TechDirt (Sep. 18, 2023), See also, Shira Ovide, Cable TV Is the New Landline, New York Times (Jan. 6, 2022),

[105] SSOIO, supra, note 16.

[106] FCC, Wireless Telecommunications Bureau Report: Policy Review of Mobile Broadband Operators’ Sponsored Data Offerings for Zero-Rated Content and Services (Jan. 2017), available at

[107] SSOIO, supra, note 16.

[108] Hyun Ji Lee & Brian Whitacre, Estimating Willingness-to-Pay for Broadband Attributes among Low-Income Consumers: Results from Two FCC Lifeline Pilot Projects, 41 Telecomm. Pol’y. 769 (Oct. 2017).

[109] Geoffrey A. Manne, Ben Sperry, Berin Szóka, & Tom Struble, ICLE & TechFreedom Policy Comments (Jul. 14, 2014), available at

[110] Mozilla Corp. v. Fed. Commc’ns Comm’n, 940 F.3d 1 (D.C. Cir. 2019) (citations omitted).

[111] Id.

[112] In 2015, when the FCC voted to enact the 2015 Open Internet Order, Chair Tom Wheeler promised to forebear from applying such rate regulation, stating flatly that “we are not trying to regulate rates.” FCC Reauthorization: Oversight of the Commission, Hearing Before the Subcommittee on Communications and Technology, Committee on Energy and Commerce, House of Representatives, 114 Cong. 27 (Mar. 19, 2015) (Statement of Tom Wheeler). Standing as a nominee to the FCC, Gigi Sohn was asked during a 2021 confirmation hearing before the U.S. Senate Commerce Committee if she would support the agency’s regulation of broadband rates. She responded: “No. That was an easy one.” David Shepardson, FCC Nominee Does Not Support U.S. Internet Rate Regulation, Reuters (Dec. 1, 2021), In September 2023, 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.” FCC Chair Rosenworcel on Reinstating Net Neutrality Rules, C-Span (Sep. 26, 2023),

[113] Vietor, supra note 89.

[114] Id. See also, Illinois Economic and Fiscal Commission, Telecommunications Deregulation Issues and Impacts: A Special Report (Apr. 2001), available at and Kevin J. Martin, Balancing Deregulation and Consumer Protection, 17 Commlaw Conspectus (2008), available at

[115] Fruits & Manne, supra note 5, at 1.

[116] Id.

[117] Id. at 7.

[118] Jonathan E. Nuechterlein & Howard Shelanski, Building on What Works: An Analysis of U.S. Broadband Policy, 73 Fed. Comm. L.J. 219 (2021)

[119] Fruits & Manne, supra note 5, at 13.

[120] Digital Discrimination Order, supra note 15 [emphasis added].

[121] Brief of the International Center for Law & Economics and the Information Technology & Innovation Foundation as Amici Curiae in Support of Petitioners and Setting Aside the Commission’s Order, Minnesota Telecom Alliance v. FCC, No. 24-1179 (8th Cir. Apr. 29, 2024) available at

[122] IIJA 60102 (h)(4)(B).

[123] U.S. Dep’t of Com., supra note 66, at 66. States have begun to follow this lead by prescribing obligations to local providers for quality and price on deployments that have speeds and capabilities far above what BEAD and the FCC consider as the baseline for a “served” household. See, e.g., ConnectLA, BEAD Initial Proposal, vol. 2 (Aug. 2023), available at (prescribing a complex system for preferencing providers that deploy “affordable” fiber and other high-speed service to middle-class homes).

[124] RUS Vol. 87, No. 149, Notice of Availability of the Draft Programmatic Environmental Assessment for the Partnerships for Climate-Smart Commodities Funding Opportunity, Docket No. NRCS–2022–0009 (U.S.D.A., Aug. 4, 2022), and RD, Preparing for ReConnect Round 4, (USDA) available at

[125] New York State Telecommunications Association, Inc. v. James, No. 21-1075 (2nd Cir. Apr. 26, 2024), available at See also, Randolph J. May & Seth L. Cooper, Second Circuit Hears Preemption Challenge to New York’s Broadband Rate Regulation Law, FedSoc Blog (Feb. 7, 2023),

[126] Fruits & Manne, supra note 5, at 16.

[127] Id. at 1.

[128] FCC, Fourth Report and Order, Declaratory Ruling, and Third Further Notice of Proposed Rulemaking Accelerating Wireline Broadband Deployment by Removing Barriers to Infrastructure Investment, WC Docket No. 17-84 (Dec. 15, 2023), available at [hereinafter “Poles Order”].

[129] Id.

[130] Id. at ¶ 7.

[131] Id.

[132] Id.

[133] Kristian Stout & Eric Fruits, Reply Comments of the International Center for Law & Economics, In the Matter of Accelerating Wireline Broadband Deployment by Removing Barriers to Infrastructure Investment, WC Docket No. 17-84 at 4 (submitted Aug. 26, 2022), available at

[134] Id.

[135] See Poles Order at ¶ 42.

[136] Id.

[137] Id. at 8.

[138] Id. at 9.

[139] Id.

[140] See Poles Order at ¶ 42.

[141] Id.

[142] Id. at 10.

[143] Ben Sperry, Geoffrey A. Manne, & Kristian Stout, The Role of Antitrust and Pole-Attachment Oversight in TVA Broadband Deployment (Int’l Ctr. for L. & Econ. Issue Brief 2023-09-04, 2023), available at

[144] Id. at 2.

[145] Id. at 3.

[146] Id.

[147] Id. at 4.

[148] Id.

[149] Id.

[150] Id. at 6-9.

[151] Id. at 10.

[152] Id.

[153] Id. at 11.

[154] Sen. Michael S. Lee, Letter to DOJ Re: Tennessee Valley Authority (TVA) – Supporting Broadband Deployment (June 22, 2023), in Ben Sperry, Geoffrey A. Manne, & Kristian Stout, The Role of Antitrust and Pole-Attachment Oversight in TVA Broadband Deployment (Int’l. Ctr. for L. & Econ. Issue Brief, Sep. 4, 2023) available at

[155] Sperry, Manne, & Stout, supra note 124, at 16.

[156] See, e.g., BEAD NOFO, supra note 71.

[157] Manne, Stout, & Sperry, supra note 1.

[158] Id.

[159] Id.

[160] Christopher S. Yoo, Jesse Lambert & Timothy P. Pfenninger, Municipal Fiber in the United States: A Financial Assessment, 46 Telecomm. Pol. 102292 (Jun. 2022).

[161] George S. Ford, Electricity Rates and the Funding of Municipal Broadband Networks: An Empirical Analysis, 102 Energy Econ. 105475 (2021).

Continue reading
Telecommunications & Regulated Utilities

Live Nation Breakup: Are Mergers Really to Blame for Ticketmaster’s Problems?

TOTM The U.S. Justice Department (DOJ) announced yesterday that it has filed suit, along with 29 states and the District of Columbia, charging Live Nation Entertainment Inc. . . .

The U.S. Justice Department (DOJ) announced yesterday that it has filed suit, along with 29 states and the District of Columbia, charging Live Nation Entertainment Inc. and its subsidiary Ticketmaster LLC with monopolizing the live-events industry in violation of Section 2 of the Sherman Act.

The suit, filed in the U.S. District Court for the Southern District of New York, alleges that Live Nation’s so-called “flywheel” (its bundle of concert promotions, artist management, venue ownership, and ticketing services) allows it to extend its dominance in one market into the other adjacent markets. It seeks both a jury trial and structural relief, with U.S. Attorney General Merrick Garland declaring that: “It is time to break up Live Nation-Ticketmaster.”

Read the full piece here.

Continue reading
Antitrust & Consumer Protection

Against the ‘Europeanization’ of California’s Antitrust Law

Popular Media The California State Legislature is considering amendments to the state’s antitrust laws that would enable more stringent antitrust scrutiny of technology companies, particularly so-called “Big Tech.” A . . .

The California State Legislature is considering amendments to the state’s antitrust laws that would enable more stringent antitrust scrutiny of technology companies, particularly so-called “Big Tech.” A preliminary report on single-firm conduct authored by a group of experts recruited by the California Law Revision Commission suggests this could be achieved by mimicking several features of European competition law. Unfortunately, this “Europeanization” of Californian antitrust law would benefit neither California’s economy nor its consumers.

Read the full piece here.

Continue reading
Antitrust & Consumer Protection

Regulate for What? A Closer Look at the Rationale and Goals of Digital Competition Regulations

ICLE White Paper For more on this topic, see the ICLE Issue Spotlight “Digital Competition Regulations Around the World.” Executive Summary Inspired by the European Union’s Digital Markets . . .

For more on this topic, see the ICLE Issue Spotlight “Digital Competition Regulations Around the World.”

Executive Summary

Inspired by the European Union’s Digital Markets Act (DMA), a growing number of jurisdictions around the globe either have adopted or are considering adopting a framework of ex-ante rules to more closely regulate the business models and behavior of online platforms.

These digital competition regulations (“DCRs”) share two key features. The first is that they target so-called “gatekeepers” who control the world’s largest online platforms. Such regulations assume that these firms have accumulated a degree of economic and political power that allows them to harm competition, exclude rivals, exploit users, and possibly inflict a broader range of social harms in ways that cannot be adequately addressed through existing competition laws. Typically cited as examples of gatekeepers are the main platforms of Google, Amazon, Facebook/Meta, Apple, and Microsoft.

The second common features of these DCR regimes is that they impose similar, if not identical, per-se prohibitions and obligations on gatekeepers. These often include prohibitions on self-preferencing and the use of third-party party data, as well as obligations for interoperability and data sharing. These two basic characteristics set DCRs apart from other forms of “digital regulation”—e.g., those that concern with AI, privacy, or content moderation and misinformation.

This paper seeks to understand what digital competition regulations aim to achieve and whether a common rationale underpins their promulgation across such a broad swatch of territories.

A. Multiple and Diverging Goals?

We find that DCRs pursue multiple goals that may vary across jurisdictions. Some DCRs are guided by the same goals as competition law, and may even be embedded into such laws. Such is the case, e.g., in Germany and Turkey. Other regulations address competition concerns under differing or modified standards. Examples here include the “material-harm-to-competition” standard in the United States and, arguably, digital competition regulation in the UK and Australia—where traditional competition-law goals such as the protection of competition and consumer welfare comingle with an increased emphasis on “fairness.”

DCRs sometimes pursue a much broader set of goals. For instance, a prospective digital competition regulation in South Africa seeks greater visibility and opportunities for small South African platforms and increased inclusivity of historically disadvantaged peoples, along with other more competition-oriented objectives (this duality is a common feature of South African legislation). Similarly, a bill proposed in Brazil attempts to reduce regional and social inequality, as well as to widen social participation in matters of public interest, alongside its stated effort to protect competition.

In the United States, apart from protection of competition, proponents of the (now-stalled) DCR bills have invoked a broad set of potential benefits, including fairness; fair prices; a more level playing field; reduced gatekeeper power; protections for small and medium-sized enterprises (“SMEs”); reduced costs for consumers; and boosts to innovation.

Some DCRs, however, are not promulgated in pursuit of competition-oriented objectives at all—at least, not explicitly or not in the sense in which such objectives are understood in traditional competition law. The clearest example is the EU’s DMA itself, which openly eschews traditional competition-related goals and instead seeks to make digital markets “fair” and “contestable.”

B. A New Form of Competition Regulation

Regardless of the overarching goals, it is evident that DCRs incorporate themes and concepts familiar to the competition lawyer, such as barriers to entry, exclusionary conduct, competitive constraints, monopolistic outcomes, and, in some cases, even market power. This may, at first blush, hint at a close relationship between digital competition regulation and competition law. While not entirely incorrect, that assessment must come with a number of caveats.

DCRs diverge in subtle but significant ways from mainstream notions of competition law. We posit that DCRs are guided by three fundamental goals: wealth redistribution among firms, the protection of competitors of incumbent digital platforms, and the “leveling down” of those same digital platforms.

C. Rent Redistribution Among Firms

The notion of “gatekeepers” itself presumes asymmetrical power relations between digital platforms and other actors, which are further presumed both to lead to unfair outcomes and to be insurmountable without regulatory intervention. Thus, the first commonality among the DCRs we study is that they all seek to transfer rents directly from gatekeepers to rival firms, complementors, and, to a lesser extent, consumers. This conclusion follows inexorably from the DCRs’ stated goals, the prohibitions and obligations they promulgate, and the public statements of those who promote them.

While the extent to which various groups are intended to benefit from this rent re-allocation might not always be identical, all DCRs aim to redistribute rents generated on digital platforms away from gatekeepers and toward some other group or groups—most commonly the business users active on those platforms.

D. Protection of Competitors

Another important feature that DCRs share is the common goal not just to protect business users, but to directly benefit competitors—including, but not limited to, via rent redistribution. DCRs are concerned with ensuring that competitors—even if they are less efficient—enter or remain on the market. This is evidenced by the lack of overarching efficiency or consumer-welfare goals—at the very least, for those regulations not based on existing competition laws—that would otherwise enable enforcers to differentiate anticompetitive exclusion of rivals from those market exits that result from rivals’ inferior product offerings.

This focus on protecting competitors can also be seen in DCRs’ pursuit of “contestability.” As defined by DCRs, promoting contestability entails diminishing the benefits of network effects and the data advantages enjoyed by incumbents because they make it hard for other firms to compete, not because they are harmful in and of themselves or because they have been acquired illegally or through deceit. In other words, DCRs pursue contestability—understood as other firms’ ability to challenge incumbent digital platforms’ position—regardless of the efficiency of those challengers or the ultimate effects on consumers.

E. ‘Leveling Down’ Gatekeepers

The other way that DCRs seek to balance power relations and achieve fairness is by “leveling down” the status of the incumbent digital platforms. DCRs directly and indirectly worsen gatekeepers’ competitive position in at least three ways:

  1. By imposing costs on gatekeepers not borne by competitors;
  2. By negating gatekeepers’ ability to capitalize on key investments; and
  3. By facilitating third parties’ free riding on those investments.

For example, prohibitions on the use of nonpublic (third-party) data benefit competitors, but they also negate the massive investments that incumbents have made in harvesting that data. Similarly, data-sharing obligations impose a cost on gatekeepers because data-tracking and sharing is anything but free. Gatekeepers are expected to aid and subsidize competitors and third parties at little or no cost, thereby diminishing their competitive position and dissipating their resources (and investments) for the benefit of another group. The same can be said, mutatis mutandis, for other staples of digital competition regulation, such as prohibitions on self-preferencing and sideloading mandates.

F. The Perils of Redistributive and Protectionist Competition Regulation

It should be noted, of course, that direct rent redistribution among firms is generally not the goal of competition law. Rent redistribution entails significant risks of judicial error and rent seeking. Regulators may require firms to supply their services at inefficiently low prices that are not mutually advantageous, and may diminish those same firms’ incentives to invest and innovate. Those difficulties are compounded in the fast-moving digital space, where innovation cycles are faster, and yesterday’s prices and other nonprice factors may no longer be relevant today. In short, rent redistribution is difficult to do well in traditional natural-monopoly settings and may be impossible to do without judicial error in the digital world.

Protecting competitors at the expense of competition, as DCRs aim to do, is equally problematic. Competition depresses prices, increases output, leads to the efficient allocation of resources, and encourages firms to innovate. By facilitating competitors—including those that may have fallen behind precisely because they have not made the same investments in technology, innovation, or product offerings—DCRs may dampen incentives to strive to become a so-called gatekeeper, to the ultimate detriment of consumers. Protecting competition benefits the public, but protecting competitors safeguards their special interests at the public’s expense.

This is not only anathema to competition law but also to free competition. As Judge Learned Hand observed 80 years ago in his famous Alcoa decision: “the successful competitor, having been urged to compete, must not be turned upon when he wins.” Critiques of digital competition regulation’s punitive impulse against incumbent platforms flow from this essential premise—which, we contend, is the cornerstone of good competition regulation. The multiplicity of alternative justifications put forward by proponents of such regulations are generally either pretextual or serve as a signal to the voting public. To paraphrase Aldous Huxley: “several excuses are always less convincing than one.”

We end by speculating that digital competition regulation could signal more than just a digression from established principles in a relatively niche, technical field such as competition law. If extended, the DCR approach could mark a new conception of the roles of companies, markets, and the state in society. In this “post-neoliberal” world, the role of the state would not be limited to discrete interventions to address market failures that harm consumers, invoking general, abstract, and reactive rules—such as, among others, competition law. It would instead be free to intercede aggressively to redraw markets, redesign products, pick winners, and redistribute rents; indeed, to function as the ultimate ordering power of the economy.

Ultimately, however, we conclude that it is too early to make any such generalizations, and that only time will tell whether digital competition regulation was truly a sign of things to come, or merely a small but ultimately insignificant abrupt dirigiste turn in the zig-zagging of antitrust history.


Inspired by the European Union’s Digital Markets Act (“DMA”),[1] a growing number of jurisdictions around the globe either have adopted or are considering adopting a framework of ex-ante rules to more closely regulate the business models and behavior of online platforms.

These “digital competition regulations”[2] (“DCRs”) share two key features. The first is that they target so-called “gatekeepers” who control the world’s largest online platforms. Such regulations assume that these firms have accumulated a degree of economic and political power that allows them to harm competition, exclude rivals, exploit users, and possibly inflict a broader range of social harms in ways that cannot be adequately addressed through existing competition laws.[3] Typically cited as examples of gatekeepers are the main platforms of Google, Amazon, Facebook/Meta, Apple, and Microsoft.

The second common feature these DCR regimes share is that they impose similar, if not identical, per-se prohibitions and obligations on gatekeepers. These often include prohibitions on self-preferencing and the use of third-party party data, as well as obligations for interoperability and data sharing. These two basic characteristics set DCRs apart from other forms of “digital regulation”—e.g., those dealing with AI,[4] privacy,[5] or content moderation and misinformation.[6]

It is not, however, always entirely clear what DCRs aim to achieve. A cursory survey suggests that these rules pursue different goals, without an immediately apparent unifying theme. For example, some DCRs have been integrated into existing competition laws and ostensibly pursue the same goals: the protection of competition and consumer welfare. Others aim for a range of goals—including, but not limited to, competition—such as the protection of small and medium-sized enterprises (“SMEs”); regional equality; social participation; and improving the lot of business users who operate on online platforms. Some DCRs purposefully and explicitly sidestep competition-oriented considerations, aiming instead for such adjacent but ultimately distinct goals as “fairness” and “contestability.”[7]

What emerges is a seeming patchwork of goals and objectives. In this paper, we seek to assess those disparate goals and objectives, drawing on many of the major proposed and enacted DCRs.

Part I examines the goals that DCRs claim to pursue. It takes those goals at face value and offers a largely descriptive account of the objectives offered. Where necessary (such as, for example, where those goals are cryptic or not clearly articulated), reference is made to public statements by those who promulgated them.

Part II argues that DCRs are best understood as a new form of law, grounded in ideas that have found limited success in competition law itself. To some extent, DCRs are based on a common narrative that has transformed some of the core principles and themes of antitrust law. As such, DCRs partially jibe with antitrust law, but ultimately diverge from it in subtle but consequential ways.

Part III argues that, despite superficial differences, DCRs share three common goals. The first is a desire to redistribute rents from some companies to others. At the most fundamental level, DCRs all seek to address what are perceived to be extreme power imbalances between digital platforms and the rest of society—especially business users and competitors. Thus, they seek to redistribute rents away from so-called “gatekeepers” and toward the business users that operate on those platforms, and to promote competitors (including, but not limited to, via rent redistribution).

DCRs are particularly concerned with ensuring that competitors, even if they are less efficient, enter or remain in the market. This is evidenced by a lack of overarching efficiency or consumer-welfare goals—even in those regulations that are based on existing competition laws—that would otherwise enable enforcers to differentiate between anticompetitive exclusion of rivals and market exit that results from rivals’ inferior product offerings. The focus on protecting competitors also stems from DCRs’ pursuit of “contestability.” In this context, promoting contestability entails diminishing the benefits of the network effects and the data advantages enjoyed by incumbents on the theory that they make it difficult for other firms to compete—not because they are harmful to consumers or because they have been acquired illegally or through deceit.

The third way that DCRs seek to balance power relations and achieve fairness is by “leveling down” the status of the incumbent digital platforms. DCRs worsen the competitive position of gatekeepers in at least three ways:

  1. By imposing costs on gatekeepers not borne by competitors;
  2. By negating their ability to capitalize on key investments; and
  3. By helping third parties to free ride on those investments.

Essentially, gatekeepers are expected to aid and subsidize competitors and third parties at little or no cost. This, in turn, diminishes their competitive position and dissipates their resources (and investments) for the benefit of another group.

Part IV concludes. It speculates that DCRs might signal the advent of a new paradigm in political economy: a redrawing of the existing lines and roles between states, markets, and firms, with greater emphasis on the role of the state as the ultimate ordering power of the economy. In hindsight, one expression of this could turn out to be the overturning (if only partial) of the essential principles of modern competition policy: the protection of competition rather than competitors, a policy emphasis on maximizing economic output rather than rent redistribution among firms, and a commitment to merit, rather than fairness and equity. It is difficult to overstate how deeply at loggerheads this conception of the role of competition is from the existing, predominant paradigm long found in competition law.

I. A Cacophony of Goals in Digital Competition Regulation

Most DCRs pursue multiple overlapping objectives. The global picture is even more complex, as there is only partial overlap among the various goals pursued by DCRs in different jurisdictions.

Some DCRs are an extension of competition-law frameworks and are sometimes even formally embedded into existing competition laws. In principle, this means that the standard goals and rationale of competition law apply. Germany, for instance, recently amended its Competition Act, emphasizing the need to “intervene at an early stage in cases where competition is threatened by certain large digital companies.”[8] According to the Bundeskartellamt:

The newly introduced Section 19a probably represents the most important change as the Bundeskartellamt will now be able to intervene at an early stage in cases where competition is threatened by certain large digital companies. As a preventive measure the Bundeskartellamt can prohibit certain types of conduct by companies which, due to their strategic position and their resources, are of paramount significance for competition across markets.[9]

Similarly, Turkey currently is looking to amend the Turkish Competition Act with the objectives of promoting competition and innovation in digital markets; protecting consumer and business rights; and ensuring that gatekeepers do not engage in anticompetitive practices.[10] Proponents argue that the current Turkish Competition Act is not adequately equipped to address anticompetitive conduct in digital markets—such as, e.g., that the process of defining relevant markets is inappropriate for dynamic and global digital ecosystems and that specific regulations are needed due to the network effects that digital platforms confer.[11] These are all nominally competition-related concerns.[12] Other proposed changes to the Turkish Competition Act similarly reflect an increased emphasis on competition. For instance, in merger analysis, the current “dominance test” would be substituted with a “significant impediment to effective competition test,” similar to that in the EU merger-control regime. A “de minimis” rule would also be added to Article 41 to exempt agreements “that do not significantly impede competition.”

Other DCRs appear, at least to some extent, to pursue competition-law-inspired goals, despite not being formally incorporated into existing competition laws. In South Korea, for example, the Korean Fair Trade Commission (“KFTC”) recently proposed a draft DMA-style bill, the Platform Competition Promotion Act,  whose purpose is establish ex-ante rules to restore competition rapidly in designated markets “without the tedious process of defining a relevant market through economic analysis.”[13] According to the KFTC, digital competition regulation is necessary to combat monopolization in digital markets, where monopolies tend to become entrenched.[14]  As some observers have noted,[15] the Platform Competition Promotion Act covers conduct already addressed by South Korea’s existing Monopoly Regulation and Fair Trade Act.[16] Thus, while the draft bill is likely to be passed as a separate piece of legislation, there appears to be a continuum between it and South Korean competition law.

In the United Kingdom, the 2023 Digital Markets, Competition, and Consumer Bill (“DMCC”) is in the final stages of legislative approval.[17] The DMCC aims to “provide for the regulation of competition in digital markets” and, in theory, dovetails with goals pursued by competition law (it even invokes familiar competition-law themes, such as market power).[18] The DMCC would grant the UK antitrust enforcer, the Competition and Markets Authority (“CMA”), power to take “pro-competition interventions” where it has reasonable grounds to believer there may be an adverse effect on competition.[19]

The DMCC has, however, also been touted as a tool to “stamp out unfairness in digital markets.”[20] This could refer to the bill’s consumer-protection provisions, which would prohibit, inter alia, unfair commercial practices.[21] But it may also suggest that the DMCC goes beyond the remit of traditional competition law, in which “unfairness” is generally not central, except within the relatively narrow confines of the abuse-of-dominance provision under S.18 of the Competition Act.[22]

Further, in a press release welcoming the DMCC draft, the CMA enumerated the bill’s benefits as falling into the three categories of “consumer protection,” “competition,” and “digital markets.”[23] The second category grants the CMA increased powers to “identify and stop unlawful anticompetitive conduct more quickly.”[24] The third, however, proposes that the bill will “[enable] all innovating businesses to compete fairly.”[25] This could imply that competition rules in “digital markets” would be governed by different principles than those that apply in “traditional” markets—that is, those that do not involve the purchase or sale of goods over the internet, or the provision of digital content.[26] The DMCC’s provisions on “digital markets” are also formally separate from those on “competition.”[27]

In Australia, the Australian Competition and Consumers Commission (“ACCC”) is conducting a five-year digital-platform-services inquiry (“DPS Inquiry”), set to be finalized in March 2025.[28] The ACCC recommended, as part of the inquiry’s fifth interim report, service-specific obligations (similar to the UK’s proposed ex-ante rules) for “designated” digital platforms.[29] These would serve to address “anticompetitive conduct, unfair treatment of business users and barriers to entry and expansion that prevent effective competition in digital platform markets.”[30] Thus, alongside competition law’s traditional concerns (e.g., harms and benefits to consumers, innovation, efficiency, and “effective competition”), the ACCC would also incorporate concerns over “fairness” and, especially, the protection of business users.

In the United States, several bills have been put forward that are formally separate from existing antitrust law, but cover some of the same conduct as would typically be addressed under U.S. antitrust law—albeit with seemingly different goals and standards. Some of these new goals and standards represent only slight variations on the usual goals of competition law. Three main pieces of legislation have so far been put forward: the American Innovation and Choice Online Act (“AICOA”),[31] the Open App Market Act (“OAMA”),[32] and the Augmenting Compatibility and Competition by Enabling Service Switch Act (“ACCESS Act”)[33] (together, “U.S. tech bills”).

Although the U.S. tech bills largely fail to describe their underlying goals, the titles of the bills and statements made by their sponsors suggest a set of overlapping concerns, such as preventing “material harm to competition,”[34] reducing “gatekeeper power in the app economy,”[35] and “increasing choice, improving quality, and reducing costs for consumers.”[36] These goals appear to fall relatively well within the traditional remit of antitrust law.

But there are others. According to U.S. Sen. Amy Klobuchar (D-Minn.), the primary sponsor or cosponsor of several of the U.S. tech bills, AICOA is intended to “restore competition online by establishing commonsense rules of the road,” “ensure small businesses and entrepreneurs still have the opportunity to succeed in the digital marketplace,” and “create a more even playing field,” all “while also providing consumers with the benefit of greater choice online.”[37] “Fairness,” “fair prices,” and “innovation” all have also been invoked by the bills’ supporters.[38]

At the same time, for three out of the 10 types of challenged conduct, AICOA would require demonstrating “material harm to competition,” which would suggest that one of that bill’s goals is to protect competition. As the American Bar Association’s Antitrust Section has observed, however, there is no “material harm to competition” standard in U.S. antitrust law.[39] This suggests that AICOA may posit a different interpretation of what it means to protect competition, or of what sort of competition should be protected, than does traditional U.S. antitrust law.

OAMA, on the other hand, aims to open competitive avenues for startup apps, third-party app stores, and payment services in existing digital ecosystems.[40] Its title reads: “to promote competition and reduce gatekeeper power in the app economy, increase choice, improve quality, and reduce costs for consumers.” Unlike AICOA, however, OAMA would not require a showing of harm to competition—material or otherwise—to establish liability, which appears to suggest that competition might be less of a concern than the bill’s title implies.

Finally, the ACCESS Act is intended to “promote competition, lower entry barriers and reduce switching costs for consumers and businesses online.”[41] U.S. Sen. Mark Warner (D-Va.), the bill’s primary sponsor, has said that the ACCESS Act will promote competition, allow startups to “compete on equal terms with the biggest social media companies,” and “level the playing field between consumers and companies” by giving them more control over who manages their privacy.[42] Again, these are antitrust-adjacent objectives, but with a flavor (“equal terms,” “level playing field,” etc.) that is largely foreign to U.S. antitrust law.

Other DCRs pursue a mix of competition and noncompetition goals. The South African Competition Commission’s (“SACC”) Final Report on the Online Intermediation Platforms Market Inquiry, for example, found that remedial actions similar to the ex-ante rules contemplated in the DMA and elsewhere are needed to grant “[g]reater visibility and opportunity for smaller South African platforms” to compete with international players; “[e]nabl[e] more intense platform competition,” offer “more choice and innovation”; reduce prices for consumers and business users; “[p]rovid[e] a level playing field for small businesses selling through these platforms, including fairer pricing and opportunities”; and “[p]rovid[e] a more inclusive digital economy” for historically disadvantaged peoples.[43]

In a similar vein, Brazil’s proposed law PL 2768/2022 (“PL 2768”) pursues an expansive grab-bag of social and economic goals.[44] Article 4 states that targeted digital platforms must operate based on the following principles: freedom of initiative, free competition, consumer protection, a reduction in regional and social inequality, combatting the abuse of economic power, and widening social participation in matters of public interest.[45] In addition, PL 2768 also states as objectives that it will enable access to information, knowledge, and culture; foster innovation and mass access to new technologies and access models; promote interoperability among apps; and enable data portability.[46]

Finally, there are those DCRs that claim not to pursue competition-oriented goals at all. The DMA has two stated goals: “fairness” and “contestability,”[47] and explicitly denies being bound by, or even pursuing, the traditional goals of competition law: protecting competition and consumer welfare.[48] According to the DMA, competition, consumer welfare, and efficiency considerations such as those that underpin antitrust law are not relevant under the new framework. This is, according to the DMA’s text, because the goals of competition law and the DMA “are complimentary but ultimately distinct.”[49]

Interestingly, however, few other DCRs have so steadfastly disavowed competition considerations, even those that copy the DMA’s provisions verbatim. India is a case in point. In 2023, a report by the Standing Committee on Finance argued that, if digital competition regulation was not passed, “interconnected digital markets will rapidly demonstrate monopolistic outcomes that prevent fair competition. This will restrict consumer choice, inhibit business users, and prevent the rise of dynamic new companies.”[50] These concerns jibe with traditional antitrust goals, as indicated inter alia by the report’s title (“anti-competitive practices by big tech companies”). Later, another report—the Report of the Committee on Digital Competition Law (“CDC Report”)—proposed a Draft Digital Competition Bill (“DCB”).[51] According to the CDC Report, DMA-style digital competition regulation was needed to supplement the 2002 Indian Competition Act (“ICA”),[52] which—and here is the interesting part—supposedly also aims to promote “fairness and contestability.”[53]

But the ICA’s stated aims were the protection of competition, the interests of consumers, and free trade.[54] The Report of the High-Powered Expert Committee on Competition Law and Policy (“Raghavan Committee Report”),[55] which served as the basis for the ICA, modernized Indian competition law by moving it away from the structure-based paradigm of the earlier Anti-Monopolies and Restrictive Trade Practices Act of 1969 and toward an economic-effects-based analysis. The Raghavan Committee Report was unequivocal in its support of consumer welfare as the system’s ultimate goal.[56] Moreover, the report advised against a plurality of goals, including, specifically, “bureaucratic perceptions”[57] of equity and fairness, which, it argued, were mutually contradictory, difficult to quantify, and potentially opposed to the sustenance of free, unfettered competition.[58] It is therefore curious, to say the least, that the CDC Report would now, in hindsight, recast the ICA’s goals to support essentially the opposite idea.

The multiplicity of goals and their unclear, partially overlapping relationship with competition law raises questions about how we should think about these laws and, indeed, whether we can even think of them as a coherent, unified group. In the next section, we seek to untangle the nature and classification of digital competition regulation.

II. A New Form of Competition Regulation

DCRs are likely best understood as a new form of competition regulation. As some authors have noted, the precise relationship between competition law and the EU’s DMA is difficult to pinpoint.[59] In a similar vein, it is evident that many DCRs incorporate themes and concepts familiar to the competition lawyer, such as barriers to entry, exclusionary conduct, competitive constraints, monopolistic outcomes, and, in some cases, even market power. At first blush, this may suggest a direct relationship between digital competition regulation and competition law. While not entirely incorrect, that assessment comes with considerable caveats.

In this section, we argue that DCRs are a new form of competition regulation that diverges in subtle but definitive ways from mainstream notions of competition law. In essence, DCRs take plausible competition-law themes and alter and subvert them in fundamental ways, creating what could be described as sector-specific[60] or enforcer-friendly[61] competition laws. Due to their blend of competition principles and prescriptive, top-down regulatory provisions, we have opted for the term “digital competition regulation.” To understand their nature, we must start with their underlying assumptions and the ills they claim to address.

A. The DCR Narrative

A starting assumption of all DCRs is that there is an extreme imbalance of power between large digital platforms and virtually every other stakeholder with whom they deal—from other industries to the businesses that operate on digital platforms to their competitors to, finally, end-users.[62] Even governments are often presumed to be virtually powerless in the face of the depredations of so-called “Big Tech.”[63] The adage that “big tech has too much power” has been almost universally endorsed by proponents of DCRs and strong antitrust enforcement;[64] is explicitly or implicitly embedded into those DCRs;[65] and now also permeates popular discourse, media, and entertainment.[66] The corollary is that asymmetric regulation is needed to help those other actors that have been “dispossessed” by big-tech platforms.

This notion is widespread and underpins a range of other policy proposals, not just DCRs. For example, the EU is considering a “Fair Share” regulation that would address the supposed power imbalance between tech companies and telecommunications operators, by forcing the former to pay for the infrastructure of the latter.[67] Similarly, various “bargaining codes” either already have been adopted or are currently under consideration to force tech companies to pay news publishers. In Australia, the Treasury Laws Amendment (News Media and Digital Platforms Mandatory Bargaining Code) Act 2021 (“Bargaining Code”) was put in place to address the supposed bargaining-power imbalance between digital platforms and news-media businesses.[68]  According to the ACCC, digital-advertisement regulation was necessary to support the sustainability of the Australian news-media sector, “which is essential to a well-functioning democracy.”[69] Laws with a similar rationale have also been passed or are under consideration in other jurisdictions.[70]

All these initiatives originate from the same foundational assumption, which is that tech companies are more powerful than anyone else, and are therefore able to get away with imposing draconian conditions unilaterally that allow them to benefit disproportionately at the expense of all other parties, business users, complementors, and consumers. While it is not always easy to identify a coherent thread running through the rules and prohibitions contained in DCRs and other initiatives to regulate “Big Tech,” a good rule of thumb to understand the unifying logic behind these initiatives is that digital platforms should have less “power,” and other stakeholders should have more “power.”

Sometimes—but by no means always—this also encompasses familiar notions of “market power,” i.e., firms’ ability to profitably raise prices because of the absence of sufficient competition. In fact, in most DCRs, “power” stems from the fact that an online platform is an important gateway for business users to reach consumers.[71] This is considered manifestly evident by the platform’s size, turnover, or “strategic” importance.[72] As Bundeskartellamt (the German competition authority) President Andreas Mundt has put it: “we shouldn’t talk about this narrow issue of price, we should talk about power.”[73]

DCRs embody this principle. They seek to extract better deals for the party or parties that are considered to suffer from an imbalance of bargaining power vis-à-vis digital platforms—such as, for instance, through interoperability and data-sharing mandates. As we argue in Section III, these beneficiaries are intended to be the platform’s business users and competitors.

The reasoning is as follows. The asymmetrical power relations between digital platforms and other actors are presumed to lead to unfair outcomes in how these stakeholders are treated and the ways that rents are allocated across the supply chain. As the DMA explains in its preamble:

The combination of those features of gatekeepers is likely to lead, in many cases, to serious imbalances in bargaining power and, consequently, to unfair practices and conditions for business users, as well as for end users of core platform services provided by gatekeepers, to the detriment of prices, quality, fair competition, choice and innovation in the digital sector.[74]

Once it is accepted that power relations between digital platforms and other stakeholders are unfairly skewed, any outcome resulting from the interaction of the two groups must also, by definition, be “unfair.” For example, under the DMA, “unfairness” is broadly defined as “an imbalance between the rights and obligations of business users where the gatekeeper obtains a disproportionate advantage.”[75] A “fair” outcome would be one in which market participants—including, but not limited to, business users—“adequately” capture the benefits from their innovations or other efforts, something the DMA assumes is currently not taking place due to gatekeepers’ superior bargaining power.

In the world of digital competition regulation, “unfairness” is a foregone conclusion. And, sure enough, the concept of “fairness” is the central normative value driving these regulations. Proponents liberally invoke it[76] and it features prominently in DCRs.[77] This narrative, however, is built on premises that differ markedly from those of antitrust law. We discuss these below.

B. Key Differences in First Principles

The DMA is the original blueprint for all digital competition regulation that has followed in its wake. The DMA’s text states that it is distinct from competition law:

This Regulation pursues an objective that is complementary to, but different from that of protecting undistorted competition on any given market, as defined in competition-law terms, which is to ensure that markets where gatekeepers are present are and remain contestable and fair, independently from the actual, potential or presumed effects of the conduct of a given gatekeeper covered by this Regulation on competition on a given market. This Regulation therefore aims to protect a different legal interest from that protected by those rules and it should apply without prejudice to their application.[78]

Other DCRs are rarely so candid about their break with competition law. On the contrary, some are even outwardly couched in competition-based terms. But in the end, DCRs replicate all or most of the prohibitions and obligations pioneered by the DMA.[79] DCRs also apply largely to the same companies as the DMA or, at the very least, use the same thresholds to establish which companies should be subject to regulation.[80]

This leads to a curious “Schrödinger’s DCR” scenario, where the same substantive rules simultaneously are and are not competition law. In the EU, for example, they are not; but in Turkey and Germany, they are. India’s DCB is a verbatim copy of the DMA, yet it is presented as a specific competition law.[81] This apparent contradiction is salvageable only if one thinks of digital competition regulation neither as competition law, strictu sensu, nor as an entirely separate regulation, but rather, as a partially overlapping tool that regulates competition and competition-related conduct in a different—and sometimes fundamentally different—manner.

Consider the example of the EU. EU competition law seeks to protect competition and consumer welfare. The DMA, on the other hand, is guided by the twin goals of “fairness” and “contestability.” As such, under the DMA (as under all other DCRs) the relevant standards are inverted. Under most DCRs, market power—understood as a firm’s ability to raise praises profitably—is either immaterial or not essential to establish whether a firm is a gatekeeper.[82] The competition-law practice of defining relevant markets on a case-by-case basis to determine whether a company has market power is, therefore, likewise moot.[83]

That approach is instead substituted for a list of pre-determined “core platform services,” which are thought to be sufficiently unique that they necessitate special and more stringent regulation.[84] Notably, and unlike in competition law, this presumption admits no evidence to the contrary. Once a good or service is marked as a core platform service, all a company can do to escape digital competition regulation is to argue either that it is not a gatekeeper, or that its services do not fall into the definition of a core platform service.

A corollary of this is that it is typically irrelevant whether a firm is dominant, or even a monopolist. Instead, DCRs apply to companies with high turnover and many business- or end-users—in other words, to “big” companies or companies people currently rely on or like to use.

Lastly, consumer-welfare considerations, which are central under competition law,[85] play only a marginal role in digital competition regulation, both in imposing prohibitions and mandates and in exempting companies from fulfilling those prohibitions or obligations.[86] While DCR supporters applaud this shift toward a broader conception of power,[87] it is important to understand how this approach differs from competition law.[88]

Competition law generally does not engage companies for being big or “important”—even if they are of “paramount importance”—except in very narrow instances, such as those prescribed by the essential-facilities doctrine.[89] Rather, antitrust targets conduct that restricts competition to the ultimate detriment of consumers. To establish whether a company has the ability and incentive to restrict competition, an assessment of market power is typically required, and definitions of relevant product and geographic markets are instrumental to that end.

Even the concept of dominance in competition law eschews crude arithmetic in favor of evidence-based analysis of market power, including the dynamics of the specific market; the extent to which products are differentiated; and shifts in market-share trends over time.[90] As one leading EU competition-law textbook puts it:

The assessment of substantial market power calls for a realistic analysis of the competitive pressure both from within and from outside the relevant market. A finding of a dominant position derives from a combination of several factors which, taken separately, are not necessarily determinative.[91]

Well-established competition-law principles—such as the prevention of free-riding,[92] the protection of competition rather than competitors,[93] and the freedom of even a monopolist to set its own terms and choose with whom it does business[94]—all preclude the imposition of hard-and-fast prohibitions and obligations without a robust case-by-case analysis or consideration of countervailing efficiencies. The narrow exceptions are those few cases where (substantive) experience shows that per-se prohibitions are warranted. But note that even cartels, “the cancers of the market economy,”[95] can generally be exempted under EU competition law.[96]

There exists no such consensus about the harms inflicted by the sort of gatekeeper conduct covered by DCRs.[97] Yet in digital competition regulation, strict (often per-se) prohibitions and obligations based on a company’s size are the norm.

C. The Transformation of Familiar Antitrust Themes

Even those DCRs that explicitly allude to competition-related objectives—such as the protection of competition and consumers—modify those objectives in subtle, but important ways. The U.S. tech bills are a case in point. AICOA would introduce a new “material harm to competition” standard. This facially sounds like it could be an existing standard under U.S. antitrust law, but it is not.[98]

DCRs also combine traditional competition-law objectives with considerations that would not be cognizable under antitrust law. For example, Brazilian competition law is guided by the constitutional principles of free competition, freedom of initiative, the social role of property, consumer protection, and prevention of the abuse of economic power.[99] PL 2768, however, would add two exogenous elements to these relatively mainstream antitrust goals: a reduction in regional and social inequality and increased social participation in matters of public interest.[100]

Other DCRs—like the UK’s or Australia’s prospective efforts to regulate digital platforms—also combine “fairness” goals with consumer welfare and competition considerations.[101] India’s DCB even offers an ex-post rationalization of competition law that brings it in line with the “fairness and contestability” goals of the new digital competition regulation.[102]

It is also questionable whether the protection of consumers and business users under DCRs accords with antitrust notions of “consumer welfare.” It should be noted that competition law, unlike consumer-protection law, protects consumers only indirectly, through the suppression of anticompetitive practices that may affect them through increased prices or decreased quality. Thus, antitrust law is generally uninterested in a company’s deceptive practices, unless they stem directly from a competitive restraint or the misuse of market power.[103] In this scenario, market power acts as a filter to determine where a company’s conduct can be corrected by market forces, and where intervention may be necessary.[104]

By contrast, most DCRs that claim to protect consumers[105] seek to do so through mandates of increased transparency, explicit consent, choice screens, and the like, imposed independently of market power.[106] While some of the focus on consumers remains (at least nominally), the ways in which DCRs protect consumers are more in line with consumer-protection law than competition law.

As for the protection of business users, according to some interpretations, antitrust law protects both consumers and other trading parties (customers).[107] This could, in principle, also include “business users.” Unlike digital competition regulation, however, antitrust law does not generally protect a predetermined group of businesses such that, for example, business users of online platforms would be afforded special protection. Any trading party—regardless of size, industry, or position in the supply chain, and whether a small developer or a large online platform—could theoretically benefit from the protection afforded by antitrust law to those harmed by the misuse of market power.

D. Partial Conclusion: When Failed Antitrust Doctrine Becomes ‘Groundbreaking’ New Regulation

While digital competition regulation’s approach to competition diverges from that of mainstream competition law, and may even be anathema to it, the arguments it espouses are not new. To the contrary, digital competition regulation, in many ways, codifies ideas that have been repeatedly tried and spurned by competition law.

The fountainhead of these ideas is that size alone should be the determining factor for antitrust action and liability.[108] On this historically recurring view—which is championed today most fervently by American “neo-Brandeisians” and European “ordoliberals”—big business inherently harms smaller companies, consumers, and democracy. It is therefore the role of antitrust law to combat this pernicious influence through structural remedies, merger control, and other interventions intended to disperse economic power.[109]

In a similar vein, digital competition regulation targets companies that, a priori, have little in common. Digital competition regulation applies to information-technology firms that specialize in online advertising, such as Google and Meta, but also to electronics companies that focus on hardware, such as Apple.[110] It covers voice assistants and social media, which are vastly different products. Cloud computing, another “core platform service,” is arguably not even a platform; yet, it was included in the DMA at the 11th hour.[111] In the end, what these “gatekeepers” have in common is that they all enjoy significant turnover, large user bases, are disruptors of legacy industries (such as, for example, news media), and are—possibly for these precise reasons—politically convenient targets.[112]

One corollary of this school of thought is that antitrust law should abandon (or, at least, drastically reduce) its reliance on the consumer-welfare standard as the lodestar of competition.[113] The law’s fixation on consumer welfare, the argument goes, has turned a blind eye to rampant economic concentration and to any form of abuse or exploitation that does not result in decreased output or higher prices.[114] Instead of this “myopic” focus on economic efficiency, proponents argue, antitrust law should strive to uphold a pluralistic market structure, which necessarily implies protecting companies from more efficient competitors.[115] This, they claim, was the Sherman Act’s original intent, which was subverted, in time, by the Chicago School’s emphasis on economic efficiency.[116]

Shunning consumer welfare also has implications for the role of market power in antitrust analysis. At the most fundamental level, competition law is concerned with controlling market power.[117] However, on the neo-Brandeisian view, antitrust’s historical concern with delineating efficient and inefficient market exit gives way to the unitary goal of controlling size and maintaining a certain market structure, regardless of companies’ ability to restrict competition and profitably raise prices.[118] This disenfranchises market power or, at the very least, redefines it as synonymous with size and market concentration.[119] This is familiar ground for digital competition regulation, which, as we have seen, generally does not target companies with market power, but companies with a certain size and “economic significance.”

Throughout antitrust law’s storied history, it has often been argued that antitrust law pursues, or should pursue, a plurality of goals and values.[120] Today, these arguments posit that antitrust law must look beyond a “narrow focus” on consumer welfare,[121] which is still enshrined as the dominant paradigm in most jurisdictions. Some of the alternative goals posited to inform the adjudication of competition-law cases include, but are not limited to, democracy, protection of competitors (especially SMEs), pluralism, social participation, combating undue corporate size, and equality. In turn, many of these goals are mentioned in digital competition regulation. In Section III, we argue that wealth redistribution (equality), the protection of competitors, and combatting size are truly shared goals of DCRs.

Digital competition regulation is a bridge between competition law and regulation. That bridge is built on old but persistent ideas that have found limited success in antitrust law and that have largely been precluded by decades of case-law and the progressively mounting exigencies of robust, effects-based economic analysis.[122] It is therefore perhaps unsurprising that digital competition regulation spurns both in favor or new legislation and per-se rules.

Its break with antitrust law, however, is not total, and was arguably never intended to be. Instead, digital competition regulation revises modern competition law to bring it in line with the regulatory philosophy it seeks to resuscitate, selectively plucking those bits and pieces that conform to that vision, and discarding those that do not.

The partial continuity between competition law and digital competition regulation is not merely hypothetical, either. Consider the example of the DMA. According to EU Commissioner of Competition Margrethe Vestager, “the Digital Markets Act is very different to antitrust enforcement under Article 102 TFEU. First, the DMA is not competition law. Its legal basis is Article 114 TFEU. Therefore, it pursues objectives pertaining to the internal market.”[123]

But observe that the DMA covers conduct identical to that which the Commission has pursued under EU competition law. For instance, Google Shopping is a self-preferencing case that would fall under Article 6(5) DMA.[124] Cases AT.40462 and AT.40703, which related to Amazon’s use of nonpublic trader data when competing on Marketplace, and its supposed bias when awarding the “Buy Box,” would now be caught by Articles 6(2) and 6(5) DMA.[125] The fine issued against Apple for its anti-steering provisions, which would be prohibited by Article 5(4) DMA, mere days before the law’s entry into force, is another case in point.[126]

This casts doubt on the assertion that the DMA and EU competition law are two distinctly different regimes. It suggests instead that the DMA is simply a more stringent, targeted, and enforcer-friendly form of competition regulation, intended specifically to cover certain products, certain companies, and certain markets. Or, as some have put it, “the DMA is just antitrust law in disguise.”[127] Indeed, Australia’s ACCC may have said the quiet part out loud when it contended that its proposed DCR would be both a “compliment to, and an expansion of, existing competition rules.”[128]

Or consider the example of India. In India, digital competition regulation would also be implemented though separate legislation. According to a 2023 report of the Standing Committee on Finance, a “Digital Competition Act”[129] is needed to prevent monopolistic outcomes and anticompetitive practices in “digital markets,” which are thought to differ in important ways from “traditional” markets:

India’s competition law must be enhanced so that it can meet the requirements of restraining anti-competitive behaviours in the digital markets. To that end, it is also necessary to strengthen the Competition Commission of India to take on the new responsibilities. India needs to enhance its competition law to address the unique needs of digital markets. Unlike traditional markets, the economic drivers that are rampant in digital markets quickly result in a few massive players dominating vast swathes of the digital ecosystem.[130]

But it seems that, based on the relevant Report of the Standing Committee on Finance, this new regime would be inspired by goals similar to Indian competition law. One important difference is that, according to Indian ministers, the new Digital Competition Act would adopt a “whole government approach.”[131] Pursuant to the  Digital Competition Act, the government would have the power to override any decisions taken by the Competition Commission of India on public-policy grounds. This, again, underscores the “subtle” but significant differences between the competition regimes that would essentially apply in parallel to digital platforms and all other companies, as India’s Competition Act does not otherwise adopt a “whole government approach” to anticompetitive conduct.[132]

A separate question, beyond the scope of this paper, is whether the sui generis logic of digital competition regulation will eventually be transferred to standard competition law. Now that they have the weight of the law—in jurisdictions like Turkey and Germany, even formally incorporated into competition law—ideas that have hitherto remained at the fringes of mainstream competition law may start to be seen as more respectable. Further, the goals of competition law may even be reconfigured, a posteriori, in accordance with the rationale of digital competition regulation.

This possibility cannot be discarded as entirely hypothetical. For example, Andreas Mundt recently remarked that competition law “has always been about fairness and contestability,”[133] thus de facto extrapolating the logic of the DMA’s sector-specific competition regulation to all competition law.

When populist arguments about equality, fairness, and “anti-bigness” previously have reared their head in competition law, they have largely (though not entirely) failed. It is thus somewhat ironic that such ideas should now be spurred by passage of the DMA, a regulation that is—by its own terms—not even a competition law, sensu proprio.

III. The Real Goals of Digital Competition Regulation

Notwithstanding certain differences, DCRs are largely animated by a common narrative and seek to achieve, on the whole, similar goals. At the most basic level, DCRs seek to tip the balance of power away from digital platforms (see Section IIA); to scatter rents, especially toward app developers and complementors; and to make it easier for potential competitors to contest incumbents’ positions. In this context, traditional antitrust conceptions of competition and consumer welfare are afforded, at best, a ceremonial role.

A. Redistributing Rents Among Firms

Despite the apparent discrepancies identified in Section I, it becomes evident on closer examination that DCRs share a common set of assumptions, rationales, and goals. The first of these goals is direct rent redistribution among firms.

The central conceit of DCRs is that asymmetrical power relations between digital platforms and virtually everyone else produce “unfair” outcomes where, in a zero-sum game, “big tech” gets a big slice of the piece of the pie at the expense of every other stakeholder.[134] Thus, DCRs must step in to reallocate rents across the supply chain, so that other actors receive a share of benefits in line with regulators’ understanding of what constitutes a “fair” distributive outcome.

Indeed, as the OECD has noted, the concept of “fairness” is strongly tied to redistribution.[135] As Pablo Ibanez Colomo wrote of the then-draft DMA: “the proposal is crafted to grant substantial leeway to restructure digital markets and re-allocate rents.”[136] This notion is accepted even by DCR proponents, who have admitted that “the regime is not designed to regulate infrastructure monopolies, but rather to create competition as well as to redistribute some rents.”[137]

As to whom should benefit principally from such interventions, the answer varies across jurisdictions, and may depend on the effectiveness of various groups’ rent-seeking efforts, or the particular country’s political priorities.[138] In countries like Korea and South Africa, there has been an explicit emphasis on SMEs, with attempts made to “equalize” their bargaining position vis-à-vis large digital platforms.[139] Other jurisdictions, such as the EU, emphasize competitors (see Section IIIB) and companies that “depend” on the digital platform to do business—such as, e.g., app developers and complementors that “depend” on access to users through iOS; logistics operators that “depend” on Amazon to reach customers; and shops that “depend” on Google for exposure.[140] Granted, these companies may also be SMEs, but they need necessarily not be.[141] In fact, many of the DMA’s expected beneficiaries, including Spotify,, Epic, and Yelp,[142] are not small companies at all.[143]

Elsewhere, it is explicitly recognized that DCRs seek to abet the market position of national companies. Prior to the DMA’s adoption, many leading European politicians touted the act’s text as a protectionist industrial-policy tool that would hinder U.S. firms to the benefit of European rivals. As French Minister of the Economy Bruno Le Maire stated:

Digital giants are not just nice companies with whom we need to cooperate, they are rivals, rivals of the states that do not respect our economic rules, which must therefore be regulated…. There is no political sovereignty without technological sovereignty. You cannot claim sovereignty if your 5G networks are Chinese, if your satellites are American, if your launchers are Russian and if all the products are imported from outside.[144]

This logic dovetails neatly with the EU’s broader push for digital and technology sovereignty, a strategy intended to reduce the continent’s dependence on technologies that originate abroad. This strategy has already been institutionalized at different levels of EU digital and industrial policy.[145] In fact, the European Parliament’s 2020 briefing on “Digital Sovereignty for Europe” explicitly anticipated an ex-ante regulatory regime similar to the DMA as a central piece of that puzzle.[146]

The fact that no European companies were designated as gatekeepers lends credence to theories about the DMA’s protectionist origins.[147] But while protectionism is not explicitly embedded in EU law, it likely will be in South Africa’s digital competition regulation. The understanding of “free competition” that underpins the SACC’s DCR proposal hinges on forcing large, foreign digital platforms to elevate local competitors and complementors, even if it means granting them unique advantages.[148] Moreover, unlike other DCRs, SACC’s proposal explicitly notes that its proposed remedies are designed to redistribute wealth from the targeted digital companies or downstream business users toward certain social groups—namely, South African companies, historically disadvantaged peoples (“HDPs”), and SMEs, especially those owned by HDPs.

For instance, to address the “unfair” advantage enjoyed by larger competitors who are displayed more prominently in Google’s search results and are able to invest in search-engine optimization,[149] the SACC would oblige Google to introduce “new platform sites unit (or carousel) to display smaller SA platforms relevant to the search (e.g., travel platforms in a travel search) for free and augment organic search results with a content-rich display.”[150] In addition, Google would be forced to add a South African flag identifier and South African platform filter to “aid consumers to easily identify and support local platforms in competition to global ones.”[151]

The SACC’s proposal is chock full of similar, blatantly redistributive policies that—despite being formally integrated into competition law—flip its logic on its head by requiring distortions of competition in order to (putatively) preserve undistorted competition. Thus, the SACC’s proposal would require gatekeepers to give free credit to South African SMEs; offer promotional rebates; waive fees; provide direct funding for the identification, onboarding, promotion, and growth of SMEs owned by HDPs; force app stores to have a “local curation of apps” aimed at circumventing “automated curation based on sales and downloads for the SA storefronts, and some geo-relevance criteria”; and ban both volume-based discounts that benefit larger companies (relative to SMEs) and promotions that would otherwise “decimate” local competitors.[152]

One reading is that the SACC’s report deviates from the “standard” in digital competition regulation. Another is that the SACC is simply more forthright about accomplishing the goals implicit in the DMA. Indeed, the SACC targets the same types of digital platforms as the DMA, includes many of the same prohibitions and obligations (e.g., self-preferencing, interoperability, cross-use of data, price parity clauses), and openly references the DMA.[153]

In conclusion, despite some distributional differences, the overarching implication of digital competition regulation is generally the same: competitors and business-users (e.g., app store and app developers in the case of Apple’s iOS; sellers and logistics operators in the case of Amazon’s marketplace; competing search and service providers in the case of Google search) should be propped up by gatekeepers. These parties, DCR proponents argue, should get more and easier access to the platforms, feature more prominently therein, be entitled to a larger slice of the transactions facilitated by those platforms,[154] and pay gatekeepers less (or nothing at all).

In some countries, the beneficiaries are intended to be primarily national companies or SMEs. Ultimately, like many other questions surrounding digital competition regulation, the question of cui bono—who benefits?—is not an economic, but a political one, hinging on whatever parties lawmakers want to favor, and at the expense of whatever parties they wish to disfavor.[155] The bottom line, however, goes back to the same, simple idea: gatekeepers should get less, and other businesses should get more.

Consider, for example, the reaction to Apple’s DMA-compliance plan.[156] Most of the backlash concerned the (frustrated) expectations that Apple would, as a result of the obligations imposed by the DMA, take a smaller cut from in-app payments and paid downloads on its platform.[157] If one strips away the rhetoric, the reaction was not about competitive bottlenecks, competition, fairness, contestability, or any other such lofty ambitions, but about the very simple arithmetic of rent seeking, whereby those who invest in lobbying legislators expect a return on their investments.[158]

Or consider the UK’s DMCC. The DMCC includes a “final offer mechanism” that the CMA can use in some cases where a conduct requirement relating to fair and reasonable payment has been breached, and where the CMA considers other powers would not resolve the breach within a reasonable time period.[159] A key aspect of the mechanism is that the two parties to a transaction (at least one of them being a gatekeeper, or a firm with “strategic market status”) submit suggested payment terms for the transaction. The CMA then decides between the two offers, with no option to take a third or intermediate course.

Under the DMCC, however, this mechanism could be applied to any SMS business relationship with third parties. While, as the British government says, this does not involve “direct price setting,”[160] it does mean the CMA would be empowered to decide between two alternative offers and, thus, will determine the distribution of revenues between gatekeepers and, potentially, any third party.[161]

B. Facilitating Competitors and the Duality of Contestability

DCRs share a common aim not just to protect business users, but to benefit competitors directly.[162] In contrast with modern notions of competition law, which readily accept that protecting competition often forces less-efficient competitors to depart the market,[163] DCRs are chiefly concerned with ensuring that even inferior competitors enter or remain on the market. Simply put: if a designated digital platform acts “unfairly,” its actions are illegal. But it is generally—save limited exceptions—irrelevant whether its behavior is efficient or if it enhances consumer welfare. These are the very questions that typically serve to delineate pro-competitive from anti-competitive conduct in the context of competition law (and competition on the merits from anti-competitive conduct).[164]

This makes sense if one recognizes that digital competition regulation and competition law have fundamentally different goals: the former seeks to make it easier for nonincumbent digital platforms to succeed and stay on the market, regardless of the costs either to consumers or to the regulated platforms; the latter seeks to protect competition to the ultimate benefit of consumers, which often implies (and requires) weeding out laggard competitors (see Section II).[165]

As former Federal Trade Commission (“FTC”) Commissioner Maureen Ohlhausen has observed:

Some recent legislative and regulatory proposals appear to be in tension with this basic premise. Rather than focusing on protection of competition itself, they appear to impose requirements on some companies designed specifically to facilitate their competitors, including those competitors that may have fallen behind precisely because they had not made the same investments in technology, innovation or product offerings. For example, the Digital Markets Act (DMA) would force a ‘gatekeeper’ company to provide business users of its service, as well as those who provide complementary services, access to and interoperability with the same operating system, hardware, or software features that are available to or used by the gatekeeper. While this would restrain gatekeepers and presumably facilitate the interests of the gatekeeper’s rivals, it is not clear how this would protect consumers, as opposed to competitors.[166]

That is because the two kinds of legislation pursue mutually exclusive goals. DCRs aim to facilitate competitors by making covered digital markets more “contestable.” The assumption is that, because consumers consistently use certain dominant platforms, “digital markets” must not contestable, or not sufficiently contestable.[167] The putative reason for this low level of contestability allegedly lies in certain advantages that have accrued to incumbent platforms and that competitors purportedly cannot reasonably replicate, such as network effects, data accumulation, and data-driven economies of scale. Consumer cognitive biases and lock-in are asserted as further cementing incumbents’ positions. Because digital markets are also said to be “winner-takes-all,” the corollary is that currently dominant firms will remain dominant unless regulators intercede swiftly and decisively to bolster contestability.

DCRs seek to achieve this state of contestability by “equalizing” the positions of gatekeepers and competitors in two interconnected ways: by diminishing incumbents’ advantages and by forcing them to share some of those advantages with competitors. Making digital markets more contestable therefore requires undercutting the benefits of network effects and advantages enjoyed by “data-rich” incumbents,[168] not because data harvesting is inherently bad or because incumbents have acquired such data illegally or through deceit; but because it makes it hard for other firms to compete. Contestability—understood as other firms’ ability to challenge incumbent digital platforms’ positions—is therefore put forward as a goal in itself, regardless of those challengers’ relative efficiency or what effects the contestability-enhancing obligations have on consumers (see Section IIID).

It is not hard to see how the deontological focus on contestability is narrowly connected to the protection of competitors. Many, if not most, of the obligations and prohibitions in DCRs are best understood as attempts to improve contestability by facilitating competitors, while stifling incumbents. For instance, data-sharing obligations—such as those included in Article 19a of the German Competition Act and Art.6(j) DMA—make it harder for incumbents to accumulate data, while also forcing them to share the data they harvest with competitors. The objective is clearly not to tackle data harvesting because it is noxious, but to disperse users and data across smaller competitors and thereby make it easier for those competitors to stay on the market and contest the incumbents’ position.

Similarly, so-called “self-preferencing” provisions seek to prohibit designated companies from preferencing their own products’ position ahead of that granted to competitors, even if consumers ultimately benefit from such positioning (e.g., because the incumbent’s package is more convenient).[169]

Interoperability obligations likewise require incumbents to make their products and services compatible with those offered by competitors, often with very limited scope for affirmative defenses grounded solely in objective security and privacy considerations. The logic is that interoperability reduces switching costs and allows competitors to attract more easily the previously “locked-in” users.

There are also prohibitions on the use of data generated by a platform’s business users, which essentially ignore the potentially pro-competitive cost reductions and product improvements that may result from the cumulative use of such data. Instead, the goal is to preclude gatekeepers from outperforming—including through more vigorous competition, such as better products or more relevant offers—the third parties who have generated such data on gatekeepers’ platforms.

Ultimately, what all these provisions have in common is that they primarily seek to increase the number of competitors on the market and to enhance their ability to gain market share at the incumbent’s expense, regardless of the effects on the quality of competition, end products, or concerns related to free-riding on incumbents’ legitimate business investments, superior management decisions, or product design (all of which are considerations that would be cognizable under antitrust law—on which, see Section II).[170] “Contestability” in digital competition regulation thus means an erosion, through regulatory means, of incumbents’ competitive advantages, regardless of how those competitive advantages have been achieved.

Digital competition regulation is therefore inherently competitor-oriented, regardless of its stated goals, and this focus is often enshrined in law in other, subtler ways. For instance, the DMCC explicitly invites potential or actual competitors to provide testimony to the CMA before it imposes or revokes a conduct requirement. It requires the CMA to initiate consultations on the imposition or removal of such conduct requirements (S. 24), as well as on “procompetitive interventions” (S. 48).

The proposed ACCESS Act in the United States likewise gives competitors a privileged seat at the table.[171] According to Sec.4(e) of the bill, if a covered platform wishes to make any changes to its interoperability interface, it must ask the FTC for permission. In deciding the question, the FTC is to consult with a “technical committee” formed by, among others, representatives of businesses that utilize or compete with the covered platform.[172] Representatives of the covered platform also would sit on the technical committee, but have no vote.[173]

Importantly, the FTC’s decision in these matters would be dependent on whether competitors’ interests have been harmed—i.e., “that the change is not being made with the purpose or effect of unreasonably denying access or undermining interoperability for competing businesses or potential competing businesses.”[174] This is tantamount to asking competitors for permission to make product-design decisions on a company’s own platform, based on the vested interests of those competitors.

Finally, less than a month after the DMA’s entry into force, the European Commission launched investigations into four gatekeepers for noncompliance. Critical to the Commission’s decision to investigate these companies was feedback received from stakeholders,[175] most of whom are competing firms who hoped to benefit from its provisions.

C. ‘Levelling Down’ Gatekeepers

There are two ways to promote equality: one is to lift up Party A, the other is to drag down Party B.[176] DCRs typically do both, all in service of suppressing the presumably illegitimate levels of gatekeeper power. In the previous subsection, we argued that DCRs facilitate competitors. But it is just as important to note that they also—sometimes concomitantly and sometimes separately—seek to worsen gatekeepers’ competitive position in at least three ways: by imposing costs on gatekeepers that are not borne by competitors, by negating their ability to capitalize on key investments, and by facilitating free riding by third parties on those investments.

For example, prohibitions on the use of nonpublic third-party data benefit competitors, but they also negate the massive investments made by incumbents to harvest that data. They preclude gatekeepers from monetizing the investments made in their platforms by, say, using that data to improve their own products and product lineup in response to new information about users’ changing tastes. This directly undermines gatekeepers’ competitive position, which depends on their ability to improve and adapt their products (see Section IIID). But this is a feature, not a bug, of DCRs. DCRs seek to dissipate gatekeepers’ “power,” where power does not necessarily mean “market power,” but simply their ability to compete effectively. For example, even if allowing gatekeepers to use nonpublic data would improve their products, to consumers’ ultimate benefit, it would also “harm” competitors in the sense that it would make it harder for them to compete with the gatekeeper. In other words, it would not be anticompetitive, but it would be “unfair.” By contrast, in the moral lexicon of digital competition regulation, free riding and effectively expropriating gatekeepers’ investments is not considered “unfair.”

Nor are data-sharing obligations. Data-sharing obligations clearly impose costs on gatekeepers: tracking and sharing data is anything but free. Nonetheless, gatekeepers are expected to aid and subsidize competitors and third parties at little or no cost,[177] thereby diminishing their competitive position and dissipating their resources (and investments) for the benefit of another group.

Similar arguments can be made about the other prohibitions and obligations that form part of the standard DCR package. Sideloading mandates allow third parties to free ride on gatekeepers’ investments in developing popular and functioning operating systems.[178] Insofar as they worsen gatekeepers’ ability to curate content and monitor safety and privacy risks, they also deprecate platforms’ overall quality and integrity, thereby potentially harming even the very companies they seek to aid.[179] Sideloading and interoperability mandates also essentially turn closed platforms into open ones (or, at the very least, they bring the two much closer together), thus forcing closed platforms to forfeit their competitive benefits relative to the primary alternatives.[180]

Antitrust law is unequivocal in its preference for inter-brand over intra-brand competition.[181] But under digital competition regulation, this principle gives way to a de-facto harmonization toward the model preferred by regulators—i.e., the one that makes every successful platform as open and accessible to competitors as possible, regardless of tradeoffs.

For example, self-preferencing prohibitions destroy one of the primary incentives for (and benefits of) vertical integration, which is the ability to prioritize a company’s own upstream or downstream products.[182] Such prohibitions also allow third parties that without the foresight to invest in a platform to accrue the same benefits as those that have. They also limit a platform’s ability to offer goods whose quality and delivery it can more readily guarantee,[183] another bane for competitiveness recast as a desirable symptom of “fairness and contestability.”

Some DCRs are considerably more candid than others about their intent to hamstring gatekeepers. The Turkish E-Commerce Law includes some provisions that differ from the DMA, despite being evidently inspired by it.[184]  Among those provisions are regulations that would not only prevent electronic-commerce intermediary-service providers (“ECISPs”) from gaining significant market power, but also require that those already in a dominant position must lose this power.[185] Moreover:

Another example of atypical regulations envisaged in the E-Commerce Law is the limitations imposed on the advertising and discount budgets of large-scale ECISPs. Under Additional Article 2/3(a), the annual advertising budget of large-scale ECISPs is limited to the sum of 2% of the amount of 45 Billion Turkish Liras of the net transaction volume of the previous calendar year applied to the twelve-month average Consumer Price Index change rate for the same calendar year and 0.03% of the amount above 45 Billion Turkish Liras. This limit constitutes the total advertising budget for all ECISPs within the same economic unit and for all ECSPs operating in the e-commerce marketplace within the same economic unity.[186]

According to Kadir Bas and Kerem Cen Sanli:

The amended E-Commerce Law goes beyond prohibiting gatekeepers’ behavior that restricts fair and effective competition, and introduces provisions that prevent undertakings in the e-commerce sector from gaining market power through organic internal growth without distorting competition or committing any unfair practices. In this context, the E-Commerce Law gradually imposes obligations and restrictions on undertakings based on their transaction volumes, which are not directly related to market power, and some restrictions significantly limiting the ability to compete are imposed on all undertakings in the sector. When those features of the E-Commerce Law are evaluated together, it can be said that the legislator aims to structurally design the competition conditions and business models in the Turkish e-commerce sector.[187]

Bas and Sanli argue that this distinguishes the E-Commerce Directive from the DMA. While it technically true that the DMA does not impose measures that would, e.g., directly limit a firm’s advertising expenditure or tax additional transactions beyond a certain threshold, it does nevertheless “level down” gatekeepers’ ability to compete and grow organically in other ways. On this view, the Turkish E-Commerce Directive takes the DMA’s logic to its natural conclusion and, much like the SACC’s proposal, simply says the quiet part out loud.

Similarly, the UK’s DMCC is designed to foreclose activities that would otherwise bolster gatekeepers’ “strategic significance.”[188] A company with strategic significance is defined as one that fulfills one or several of the following conditions: has achieved significant size or scale; is used by a significant number of other undertakings in carrying out their business; has a position that allows it to determine or substantially influence the ways in which other undertakings conduct themselves; or is in a position to extend its market power to different activities. At least three of these conditions (the first three) can easily result from organic growth or procompetitive behavior. There are many investments and innovations that would, if permitted, benefit consumers—either immediately or over the longer term—but which may enhance a platform’s “strategic significance,” as defined by the DMCC.[189] Indeed, improving a firm’s products and thereby increasing its sales will often naturally lead to increased size or scale.

The inverse is also true: product improvements, innovation, and efficiencies can result from size or scale.[190] This is especially relevant in the context of digital platforms, where a product’s attractiveness often comes precisely from its size and scope. In two-sided markets like digital platforms, product quality often derives from the direct and indirect network effects that result from adding an additional user to the network. In other words, the more consumers use a product or service, the more valuable that product or service becomes to consumers on both sides of the platform.[191] Capping scale and size thus curtails one of the primary (if not the main) spurs of digital platforms’ growth and competitiveness.

Which, of course, arguably was the intent behind DCRs all along. In this context, some DCRs contain provisions that allow enforcers to impose a moratorium on mergers and acquisitions involving a gatekeeper, even where such concentrations would not ordinarily fall within the scope of merger-control rules.[192]

This degree of animosity may seem puzzling.[193] but one’s priors matter quite a bit here. If one accepts, tout court, the dystopian narrative that casts digital platforms as uniquely powerful, unfair, and abusive (see Section IIA),[194] this punitive approach[195] is understandable and, in a sense, even required.

D. Consumers as an Afterthought

DCRs affect wealth transfers from gatekeepers to other firms (see Section IIIA). But DCRs also affect—or, at least, tacitly accept—wealth transfers from consumers to other firms. First, DCRs generally do not require a finding of consumer harm to intercede. Second, DCRs provide limited scope for efficiency defenses. Generally, only defenses rooted in objective privacy and security concerns are allowed,[196] and even these are subject to a high evidentiary burden.[197]

On the other hand, justifications related to product quality, curation, or that otherwise seek to preserve the consumers’ experience are not typically permitted. For example, the quality-of-life improvements that may come from better curation and selection of apps in a closed platform (e.g., one that does not allow for the sideloading of apps or third-party app stores) are not relevant under the DMA, nor is any other dimension of consumer welfare, including price, quality, aesthetics, or curation. The Turkish DCR goes even further than the DMA, in that does not appear to allow for any exemptions (even on the basis of safety and privacy).[198] The SACC’s proposal likewise does not appear to provide scope for affirmative defenses.

In Australia, the DPI states that exemptions should be put in place to mitigate “unintended consequences.” This could, in principle, include consumer-welfare considerations, but the DPI’s explicit reference to the DMA[199] and various public statements by the ACCC suggest that this is unlikely to be the case. The ACCC said in its Fifth Interim Report that “[t]he drafting of obligations should consider any justifiable reasons for the conduct (such as necessary and proportionate privacy or security justifications).”[200]

The narrow and strict exceptions to the above DCRs confirm the downgraded status of consumer welfare in digital competition regulation (vis-à-vis competition law). German Article 19a, for example, allows for exemptions where there is an “objective justification.”[201] But unlike in every other instance under the German Competition Act, Article 19a reverses the burden of proof and requires the gatekeeper, not the Bundeskartellamt, to prove that the prohibited conduct is objectively justified.

In a similar vein, the AICOA bill in the United States would only require that the plaintiff show “material harm to competition” in provisions related to self-preferencing and service discrimination provisions.[202] The remaining provisions do not contain affirmative-effects requirements, but would not apply if the defendant shows a lack of “material harm to competition.” In other words, the burden of proof is shifted from the plaintiff to the defendant — who must prove a negative.[203]

The UK’s DMCC allows for a “countervailing benefits exception,”[204] which would apply when behavior that breaches a conduct requirement is found to provide sufficient other benefits to consumers without making effective competition impossible, and is “indispensable and proportionate” (s. 29(2)(c)) to the achievement of the benefit.[205] Again, this sets a high bar to clear.[206] For example, a limitation on interoperability might provide a benefit to user security and safety. But the exemption would apply only if the CMA were persuaded that this limitation was the only way to achieve such protection, which could be very hard or impossible to demonstrate.

The marginality of consumer welfare as a relevant policy factor is compounded in the UK by the fact that CMA decisions would only be subject to judicial review. Firms will thus be unable to challenge the authority’s factual assessments on questions such as indispensability and proportionality.[207] Even the chance that such a thing could be shown will be of little value to affected firms since the exemption can apply only once an investigation into a breach of a conduct requirement is underway.[208]

Finally, the Brazilian proposal states that costs, benefits, and proportionality should be observed when establishing an obligation under Art.10, [209] although there is no telling what this would mean in practice, or whether it encompasses consumer welfare (Arts. 10 and 11 of PL 2768 do not mention consumer welfare).[210]

The broader question, however, is whether a pro-consumer approach is even compatible with the overarching goals of digital competition regulation. A corollary of facilitating competitors and levelling down gatekeepers is that successful companies and their products are made worse—often at consumers’ expense. For instance, choice screens may facilitate competitors, but at the expense of the user experience, in terms of the time taken to make such choices. Not integrating products might give a leg up to competing services, but consumers might resent the diminished functionality.[211] Interoperability may similarly reduce the benefits an incumbent enjoys from network effects, but users may prefer the improved safety, privacy, and curation that typically comes with closed or semi-closed “walled-garden” ecosystems, like Apple’s iOS.[212]

In sum, proponents of DCRs appear to see losses in consumer welfare as a valid and potentially even desirable tradeoff for competitors’ increased ability to contest the incumbents’ position, as well as for wealth transfers across the supply chain that are seen as inherently just, equitable, and fair.

E. Partial Conclusion: The Perils of Redistributive and Protectionist Competition Regulation

While competition enforcement can affect the allocation of rents among firms, this is generally not the goal of competition policy. The only rent redistribution that is, in principle, relevant in competition law is the one between companies that misuse their market power and consumers (or, in some cases, trading parties). But the overarching goal is to prevent distortions of competition that result in deadweight loss and transfer consumer surplus to the monopolist, not to allocate resources among a set of hand-picked “big” firms and their smaller rivals in way that legislators or regulators consider “fair.” It is the market, not the government, that determines what is “fair.” Competition laws exist to preserve, not to rewrite, that outcome.

Indeed, even some advocates of incorporating political goals into antitrust law, such as Robert Pitofsky, have opposed using the law to protect small businesses and redistribute income to achieve social goals.[213] This is for good reason. Rent redistribution among firms entails significant risks of judicial error and rent seeking. Regulators may require firms to supply their services at inefficiently low prices that are not mutually advantageous, which may in turn diminish those firms’ incentives to invest and innovate.

DCR backers may retort that rent redistribution is the goal of most natural-monopoly regulations (such as those in the telecommunications and energy-distribution industries), which generally rely on both price regulation and access regimes to favor downstream firms and (ultimately) consumers.[214] But digital markets tend to be very different to those traditionally subject to price regulation and access regimes. And even in those industries, price regulation and access regimes raise many difficulties—such as identifying appropriate price/cost ratios and fleshing out the nonprice aspects of the goods/services or regulated firms.

Those difficulties are compounded in the fast-moving digital space, where innovation cycles are faster and yesterday’s prices and other nonprice factors may no longer be relevant today.[215] In short, rent redistribution is difficult to do well in traditional natural-monopoly settings, and may be impossible to do without judicial error in the digital world.

Assuming that such redistribution was to take place, what would a fair redistribution entail? “Fairness” is subjective and, as such, in the eye of the beholder.[216] Moreover, reasonable people may and often do disagree on what is and is not fair. What is “unfair” for the app distributor who pays a commission to use in-app payments may seem “fair” to the owner of the operating system and the app store that makes significant investments to maintain them.[217] Because fairness is such an inherently elusive concept,[218] DCRs ultimately define “fair” and “unfair” by induction—i.e., from the bottom up, in a “you know it when you see it” approach that is difficult to square with any cogent normative theory or limiting principle.[219]

For example, in response to claims that Apple must allow competing in-app purchases (“IAPs”) on its App Store in order to make its 30% IAP fee more competitive (cheaper), Apple could allow independent payment processors to compete, charge an all-in fee of 30% when Apple’s IAP is chosen and, in order to recoup the costs of developing and running its App Store, 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 such a remedy has already been imposed, that is exactly what Apple has done. In the Netherlands, where Apple is required by the Authority for Consumers and Markets (“ACM”) to uncouple distribution and payments for dating apps, Apple adopted the following policy:

Developers of dating apps who want to continue using Apple’s in-app purchase system may do so and no further action is needed…. Consistent with the ACM’s order, dating apps that . . . use a third-party in-app payment provider will pay Apple a commission on transactions. Apple will charge a 27% commission on the price paid by the user, net of value-added taxes. This is a reduced rate that excludes value related to payment processing and related activities.[220]

The company responded similarly to the DMA.[221] It is not hard to see the fundamental problem with this approach. If a 27% commission plus competitive payment-provider fee permits more “competition,” or is fairer, than complete exclusion of third-party providers, then surely a 26% fee would permit even more competition, or be even fairer. And a 25% fee fairer still. Such a hypothetical exercise logically ends only where a self-interested competitor or customer wants it to end, and is virtually impossible to measure.[222]

Even if it were possible, it would entail precisely the kind of price management that antitrust law has long rejected as being at loggerheads with a free market.[223] Without a measurable market failure, what is the frontier of fairness? When does a complaint stop being a competition or gatekeeper issue and become a private dispute about wanting to pay less—or nothing—for a service?[224]

Another obvious problem with facilitating competitors and levelling down gatekeepers is that it discourages investment, innovation, and competition on the merits. Having been encouraged to bring new, innovative products to market and compete for consumers’ business, successful companies—now branded with the “gatekeeper” epithet—are subject to punitive regulation.[225] The benefits that they have legitimately and arduously acquired are dissipated across the supply chain and their competitors, who lacked the foresight and business acumen to make the same or similar investments, are rewarded for their sluggishness.[226] This stifles the mechanisms that propel competition. As Justice Learned Hand observed almost 80 years ago, “the successful competitor, having been urged to compete, must not be turned upon when he wins.”[227] There is no reason why digital competition regulation should be impervious to that logic.

The abrupt shift from competition law to digital competition regulation also sends investors the wrong message by creating commitment issues.

Commitment issues arise where a government commits itself in one period to behaving in certain ways in the future but, when it comes to a future point in time, reneges on the earlier commitment to reflect its preferences at that later point in time.[228]

For example, today’s gatekeepers have made significant investments in data processing, vertical integration, scaling, and building ecosystems. Many of these investments are sunk, meaning that they can no longer be recouped or can be recouped only partially. With the various DCRs’ entry into force, however, gatekeepers can no longer fully utilize those investments. For instance, they cannot self-preference and thereby reap the full benefits of vertical integration;[229] they cannot use third-party data generated on the platforms they have built and in which they have invested; and they must now allow third-parties and competitors to free ride on those investments in a plethora of ways, ranging from allowing sideloading to mandated data sharing (see Section IIIB).

In dynamic contexts, time-inconsistency can obviously affect firms’ actions and decisions, leading to diminished investments.[230] From a less consequentialist and more deontological perspective, however, it is also questionable how “fair” (to use the mot du jour of digital competition regulation) it is to expropriate a company’s sunk-cost investments by abruptly shifting the regulatory goalposts under a new paradigm of competition regulation that essentially subverts the logic of the previous one, and penalizes what was until recently seen as permissible and even desirable conduct (see Section II).

IV. Conclusion: Beyond Digital Competition Regulation

Aldous Huxley once wrote that several excuses are always less convincing than one.[231] His point was that multiple justifications may often conceal the fact that none of them are entirely convincing in their own right. This maxim aptly captures the doubts that persist surrounding DCRs.

On the surface, DCRs pursue a variety of sometimes overlapping, sometimes contradictory, and sometimes disparate goals and objectives (Section I). Some of these goals and objectives hearken back to familiar antitrust themes, but it would be a mistake to treat DCRs as either an appendix to or extension of competition law (Section II). Unlike mainstream competition laws, DCRs address a moral, rather than an economic failure. DCRs emphasize notions of power that are foreign to competition law, essentially promulgating a new form of competition regulation that subverts the logic, rationale, and goals of the existing paradigm.

This approach to regulating competition may be new, but it is not original. To the contrary, the use of antitrust law to castigate concerns seen as “too big and powerful,” promote visions of social justice, and facilitate laggard competitors (even if it comes at the expense of competition, total, or consumer welfare) have been around since the inception of the Sherman Act.[232] In this sense, those who say that digital competition regulation is not competition law, and should therefore not be judged by competition-law standards, [233] are correct on the form but wrong on the substance.[234] They miss the bigger and more important point, which is that—regardless of its legal classification—digital competition regulation is competition regulation, just not the kind we have known for (at least) the last half a century.

The rationale that underpins digital competition regulation can be explained as follows. (Section III). Competition is no longer about consumer-facing efficiency, but about fairness, equality, and inclusivity. In practice, this means improving the lot of some, while “levelling down” others—regardless of the respective merits or demerits of each group (or their products). In this world, “contestability” is not so much the ability to displace an incumbent through competition on the merits, but very much the reverse. It is about lowering the competitive bar to increase the number of companies on the market—full stop. Whether or not this benefits consumers is largely immaterial, as the normative lodestars of digital competition regulation—fairness and contestability—are seen as having inherent and deontological value and thus removed from any utilitarian calculi of countervailing efficiencies (except, arguably, increases in competitors’ market shares).

Ultimately, however, this “new” approach to competition will have to reckon with the same problems and contradictions as the erstwhile antitrust paradigms from which it draws inspiration. The minefield of redistributive policies is likely to hamstring investment and innovation by targeted digital platforms significantly, while simultaneously encouraging rent-seeking behavior by self-interested third parties. Enabling competitors and purposefully harming incumbents also sends the message that equitable outcomes are preferred to excellence, which could encourage even more free riding and rent seeking and further stifle procompetitive conduct.

Finally, the irony is likely not lost on even the most casual observer that, for regulations so obsessed with “fairness,” it is fundamentally unfair for DCRs to syphon rents away from some companies and into others by fiat; and to force those companies to share their hard-earned competitive advantages with others who have not had the foresight or business acumen to make the same investments in a timely fashion.

It is difficult to overstate how big of a departure from competition law this approach to competition regulation is. But digital competition regulation is potentially more than just a digression from established principles in a relatively niche, technical field like competition law. Under the most expansive version of this interpretation, digital competition regulation heralds a new conception of the role and place of companies, markets, and the state in society.

In this “post-neoliberal” world,[235] the role of the state would not be to address market failures that harm consumers through discrete interventions guided by general, abstract, and reactive rules (such as competition law). Instead, it would be to intercede aggressively to redraw markets, redesign products, pick winners, and redistribute rents; indeed, to act as the ultimate ordering power of the economy.[236] It is not difficult to see how “old” competition-law principles, such as the consumer-welfare standard, effects-based analysis, and the procedural safeguards designed to cabin enforcers’ discretion could disrupt this system.

But for now, this remains just a hypothesis, and some would say—perhaps rightly so—an alarmist one. Yet there are unmistakable signs—as unmistakable as social science will allow—that a new paradigm of political philosophy is in the making: from the rehabilitation of once-maligned industrial policy to the rise of neo-Brandeisianism to recurrent proclamations of the “death of neoliberalism”[237] and its “idols,” including the consumer-welfare standard in antitrust law.[238]

Only time will tell if the digital competition regulation is truly sign of things to come, or merely a small but ultimately insignificant and abrupt dirigiste turn in the zig-zagging of antitrust history.[239] And only time will tell whether the approach to competition regulation promulgated by digital competition regulation will stay confined to the activities of a few large concerns and a handful of core platform services, or whether its logic will, in the end, seep into other spheres of policy and social life.

[1] Regulation (EU) 2022/1925 of the European Parliament and of the Council of 14 September 2022 on Contestable and Fair Markets in the Digital Sector and amending Directives (EU) 2019/1937 and (EU) 2020/1828, 2022 O.J. (L 265) 1 (hereinafter “Digital Markets Act” or “DMA”).

[2] “Digital competition regulation” or “DCR” will be used throughout to refer both to rules already in place and to rules currently under consideration. Context on legislative status will be given where available and appropriate.

[3] The terms “competition law” and “antitrust law” will be used interchangeably.

[4] See, e.g., Proposal for a Regulation of the European Parliament and of the Council Laying Down Harmonised Rules on Artificial Intelligence (Artificial Intelligence Act) and Amending Certain Union Legislative Acts, COM (2021) 206 final (Apr. 21, 2021).

[5] Regulation (EU) 2022/2065 of the European Parliament and of the Council of 19 October 2022 on a Single Market for Digital Services and Amending Directive 2000/31/EC, 2022, O.J. (L 277) 1 (hereinafter “Digital Services Act” or “DSA”).

[6] 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 O.J. (L 119) 1 (hereinafter “General Data Protection Regulation” or “GDPR”).

[7] See, e.g., DMA, supra note 1.

[8] Press Release, Amendment of the German Act Against Restraints of Competition, Bundeskartellamt (Jan. 19, 2021),

[9] Id.

[10] The Act on the Protection of Competition No. 4054, Official Gazette (Dec. 13, 1994) (Turk.).

[11] See, E-Pazaryeri Platformari Sektor Incelemesi Nihai Raporu, Turkish Competition Authority (2022), available at (Turkish language only).

[12] Arguably, however, there is an increased emphasis on “business rights.”

[13] See, KFTC Proposes Ex-Ante Regulation of Platforms Under the “Platform Competition Promotion Act,Legal 500 (Jan. 4, 2024),

[14] Park So-Jeong & Lee Jung-Soo, S. Korea Speeds Up to Regulate Platform Giants Such as Google or Apple, The Chosun Daily (Feb. 4, 2024),

[15] Id.

[16] Monopoly Regulation and Fair Trade Act, Act. No, 3320, Dec. 31, 1980, amended by Act No. 18661, Dec. 28, 2021 (S. Kor.).

[17] Digital Markets Competition and Consumer Bill, 2023-24, H.L. Bill (53) (U.K.)  (hereinafter “DMCC”).

[18] See, e.g., id. at Part 1, S. 2, which defines companies with “strategic market status” as those with “substantial and entrenched market power.” By contrast, Recital 5 of the DMA states: “Although Articles 101 and 102 of the Treaty on the Functioning of the European Union (TFEU) apply to the conduct of gatekeepers, the scope of those provisions is limited to certain instances of market power, for example dominance on specific markets and of anti-competitive behaviour, and enforcement occurs ex post and requires an extensive investigation of often very complex facts on a case by case basis. Moreover, existing Union law does not address, or does not address effectively, the challenges to the effective functioning of the internal market posed by the conduct of gatekeepers that are not necessarily dominant in competition-law terms.”

[19] DMCC, supra note 18, at Part 1, Chapter 4.

[20] Press Release, New Bill to Stamp Out Unfair Practices and Promote Competition in Digital Markets, UK Competition and Markets Authority (Apr. 25, 2023),

[21] DMCC, supra note 18, at Part 4.

[22] Competition Act 1998 c.41 (U.K.).

[23] See Press Release, supra note 21.

[24] Id.

[25] Id. (emphasis added).

[26] The DMCC defines “digital activities” as those involving the purchase or sale of goods over the internet, or the provision of digital content. DMCC, Part 1, S.3.

[27] The provisions on digital markets are covered in Part 1 of the DMCC. DMCC, Part 2 covers competition.

[28] Digital Platform Services Inquiry 2020-25, Australian Competition and Consumer Commission, (last accessed May 13, 2024).

[29] Digital Platform Services Inquiry, Interim Report 5, Australian Competition and Consumer Commission (2022), at 5 (“The ACCC recommends a new regulatory regime to promote competition in digital platform services. The regime would introduce new competition measures for digital platforms.”). The term “digital regime” has also been used to describe the authority granted to the UK’s newly created Digital Markets Unit. See Moritz Godel, Mayumi Louguet, Paula Ramada, & Rhys Williams, Monitoring and Evaluating the Digital Markets Unit (DMU) and New Pro-Competition Regime for Digital Markets, London Economics (Jan. 2023), available at

[30] Digital Platform Services Inquiry, Interim Report 5, id. at 5.

[31] American Innovation and Choice Online Act, S. 2992, 117th Cong. (2022), (hereinafter “AICOA”).

[32] Open App Market Act, S. 2710, 117th Cong. (2022), (hereinafter “OAMA”).

[33] ACCESS Act of 2021, H.R. 3849, 117th Cong. (2021).

[34] AICOA, § 3.

[35] OAMA.

[36] Id.

[37] Press Release, Klobuchar, Grassley, Colleagues to Introduce Bipartisan Legislation to Rein in Big Tech, U.S. Sen. Amy Klobuchar (Oct. 14, 2021), The bill’s title is somewhat ambiguous, as it reads: “to provide that certain discriminatory conduct by covered platforms shall be unlawful, and for other purposes.” See AICOA, supra note 36.

[38] See id.

[39] Comments of the American Bar Association Antitrust Law Section Regarding the American Innovation and Choice Online Act (S. 2992), American Bar Association Antitrust Law Section (Apr. 27, 2022) at 5, available at (hereinafter “ABA Letter”).

[40] Press Release, Klobuchar Statement on Judiciary Passage of Legislation to Set App Store Rules of the Road, U.S. Senator Amy Klobuchar (Feb. 3, 2022),

[41] This is stated in the title of the bill. The ACCESS Act also claims to “encourage entry by reducing or eliminating the network effects that limit competition with the covered platform.” See ACCESS ACT at § 6(c).

[42] Press Release, Lawmakers Reintroduce Bipartisan Legislation to Encourage Competition in Social Media, U.S. Sen. Mark R. Warner (May 25, 2022),; see also, The ACCESS Act of 2022, U.S. Senator Mark R. Warner, available at

[43] Online Intermediation Platforms Market Inquiry, Summary of Final Report and Remedial Actions, South African Competition Commission (2023), 13, available at

[44] Projeto de Lei PL 2768/2022, (Braz.) (Portuguese language only).

[45] Id. at Art. 4.

[46] Id. at Art. 5.

[47] DMA, supra note 2 at recitals 2, 31. On the two objectives being intertwined, see Recital 34.

[48] Id., at Recital 10.

[49] Id.

[50] Anti-Competitive Practices by Big Tech Companies, Fifty Third Report, Standing Committee on Finance, 17th Lok Sahba (India), (2022-23), available at, at 29.

[51] Report of the Committee on Digital Competition Law (India), Annexure IV: Draft Digital Competition Bill (2024),

[52] The Competition Act, No. 12 of 2003, INDIA CODE (1993).

[53] CDC Report, at 4, 42.

[54] ICA, preamble. The ICA does not mention “contestability.”

[55] Report of the High-Powered Expert Committee on Competition Law and Policy (India) (1999), available at

[56] Raghavan Committee Report, at 1.1.9. “The ultimate raison d’être of competition is the interest of the consumer”; see also at 1.2.0.

[57] Raghavan Committee Report, at 2.4.1.

[58] Raghavan Committee Report, at 3.2.8. “If multiple objectives are allowed to rein in the Competition Policy, conflicts and inconsistent results may surface detriment to the consumers… In addition, such concerns as community breakdown, fairness, equity and pluralism cannot be quantified easily or even defined acceptably… it needs to be underscored that attempts to incorporate such concerns may result in inconsistent application and interpretation of Competition Policy, besides dilution of competition principles. The peril is that the competitive process may be undermined, if too many objectives are built into the Competition Policy and too many exemptions/exceptions are laid down in dilution of competition principles.”

[59] See, e.g., Pelle Beems, The DMA in the Broader Regulatory Landscape of the EU: An Institutional Perspective, 19 Eur. Comp. J. 1, 27 (2023); Pierre Larouche & Alexandre De Streel, The European Digital Market: A Revolution Grounded on Traditions, 12 J. Eur. Comp. L. & Practice 542, 542 (2021) (arguing that the DMA’s conceptual nature is in a “difficult epistemological position”).

[60] See Nicolas Petit, The Proposed Digital Markets Act (DMA): A Legal and Policy Review, 12 J. Eur. Competition L. & Practice 529, 530 (2021) (“The DMA is essentially sector-specific competition law.”). The DMA’s competition-law DNA is also explicitly reflected in Section 1.4.1 of the Legislative Financial Statement, which is annexed to the DMA proposal. See id. (“The general objective of this initiative is to ensure the proper functioning of the internal market by promoting effective competition in digital markets.”). See also Beems, supra note 62, at 27 (“In my view, it could be desirable to qualify the DMA as a specific branch of competition law that applies to gatekeepers.”).

[61] See Giuseppe Colangelo, The European Digital Markets Act and Antitrust Enforcement: A Liaison Dangereuse, 5 Eur. L. Rev., 597, 610 (2022) (“In service of this goal of speedier enforcement, the DMA dispenses with economic analysis and the efficiency-oriented consumer welfare test, substituting lower legal standards and evidentiary burdens.”). See also Pablo Iba?n?ez Colomo, The Draft Digital Markets Act: A Legal and Institutional Analysis, 12 J. Eur. Competition L. & Practice 561, 566 (2021).

[62] It should be underscored that “power” here means something much broader and general than the narrow concept of “market power” under competition law. Unlike “market power,” assertions that so-called “Big Tech” wield “power” are not intended to invoke a state-of-the art term, but rather are general references to companies’ size, resources, and capacity. Neo-Brandeisians like Lina Khan and Tim Wu often refer to the “power” of Big Tech in such terms. See generally Tim Wu, The Curse of Bigness: Antitrust in the Gilded Age (2018). For Wu, like Khan, the harmful “power” of Big Tech refers not just to concentrated economic power or market power, but to a range of other mechanisms by which these firms allegedly hold sway over democracy, elections, and society at-large. See also Zephyr Teachout & Lina Khan, Market Structure and Political Law: A Taxonomy of Power, 9 Duke J. Const. L. & Pub. Pol’y 37,74 (2014).

[63] See, e.g., Joshua Q. Nelson, Joe Concha: “Big Tech is More Powerful than Government” in Terms of Speech, Fox News (Jan. 27 2021),; How 5 Tech Giants Have Become More like Governments than Companies, Fresh Air (Oct. 26, 2017), (“New York Times tech columnist Farhad Manjoo warns that the ‘frightful five’—Amazon, Google, Apple, Microsoft and Facebook—are collectively more powerful than many governments.”).

[64] See, e.g., Press Release, Klobuchar, Grassley Statements on Judiciary Committee Passage of First Major Technology Bill on Competition to Advance to Senate Floor Since the Dawn of the Internet, U.S. Sen. Amy Klobuchar (Jan. 20, 2022), (“Everyone acknowledges the problems posed by dominant online platforms.”).

[65] See, e.g., DMA recitals 3, 4, 33, and 62.

[66] See, e.g., The Social Dilemma (Exposure Labs, Argent Pictures & The Space Program, 2020); Tech Monopolies: Last Week Tonight with John Oliver (HBO, 2022); Yanis Varoufakis, Technofeudalism: What Killed Capitalism (2023); Shoshana Zuboff, The Age of Surveillance Capitalism: The Fight for a Human Future at the New Frontier of Power (2019).

[67] See Luca Bertuzzi, EU Commission Launches Connectivity Package with ‘Fair Share‘ Consultation, EurActiv (Feb. 28, 2023),; see also Daniele Condorelli, Jorge Padilla, & Zita Vasas, Another Look at the Debate on the “Fair Share” Proposal: An Economic Viewpoint, Compass Lexecon (2023), available at On the supposed bargaining-power imbalance between large traffic originators and telecommunications companies, see id. at point 1.34(d). “There is a risk that the current unregulated arrangements result in no payments from LTOs due to asymmetries of information between industry participants, free-riding among LTOs, and the large imbalance in bargaining power between LTOs and TELCOs.” See also id. at points 3.77, 3.78 and 3.79-3.84 for the argument that the power imbalances require intervention. For a different view of “fair share,” see Giuseppe Colangelo, Fair Share of Network Costs and Regulatory Myopia: Learning from Net Neutrality Mistakes, Int’l. Ctr. for Law & Econ. (Jul. 18, 2023) (forthcoming in Law, Innovation and Technology), available at

[68] See Treasury Laws Amendment (News Media and Digital Platforms Mandatory Bargaining Code) Act 2021 (Austl.); for a defense of legislation forcing digital platforms to compensate media companies, see Zephyr Teachout, The Big Unfriendly Tech Giants, The Nation (Dec. 25, 2023),

[69] News Media Bargaining Code, Australian Competition and Consumer Commission, (last accessed May 14, 2024).

[70] See, e.g., Journalism and Competition Preservation Act of 2023, S. 1094, 118th Cong. (2023); Online News Act (S.C. 2023, c.23) (Can.).

[71] See, e.g., DMA at recitals 1, 15, 20, 62, and Art.1(b); DMCC at s.6(b); PL 2768 at Art. 2, which defines the regulation’s targets as companies with the “power to control essential access”; Competition Act in the version published on 26 June 2013 (Bundesgesetzblatt (Federal Law Gazette) I, 2013, p. 1750, 3245), as last amended by Article 1 of the Act of 25 October 2023 (Federal Law Gazette I, p. 294), Art.19(a) (1)5 (Ger.) (hereinafter “German Competition Act”).

[72] See, e.g., DMA at Recital 23, Art.3 and Art.3(8)(a); DMCC at s.6(a); German Competition Act, Art.19(a); but see DSA, Section 5, which imposes special obligations on “very large online platforms.”

[73] “From Price to Power”? Reorienting Antitrust for the New Political Economy, panel at Antitrust, Regulation and the Next World Order conference, (Feb. 2, 2024),

[74] DMA, at recital 4 (emphasis added).

[75] DMA, at recital 33.

[76] See Press Release, Digital Markets Act: Commission Welcomes Political Agreement on Rules to Ensure Fair and Open Digital Markets, European Commission (Mar. 25, 2022), (“What we want is simple: Fair markets also in digital. We are now taking a huge step forward to get there—that markets are fair, open and contestable…. This regulation, together with strong competition law enforcement, will bring fairer conditions to consumers and businesses for many digital services across the EU.”) (emphasis added); see also Press Release, Klobuchar, Grassley, Colleagues to Introduce Bipartisan Legislation to Rein in Big Tech, U.S. Sen. Amy Klobuchar (Oct. 14, 2021), (joint statement by Sens. Amy Klobuchar and Chuck Grassley with references to “fair competition,” “fair prices,” “unfairly preferencing their own products,” “fairer prices,” “unfairly limiting consumer choices,” “fair rules for the road”).

[77] For example, the DMA mentions the term “fairness,” or some variation thereof, 90 times in 66 pages.

[78] DMA, at Recital 11 (emphasis added).

[79] See DMA, Arts. 5-7.

[80] See DMA, Art. 3.

[81] CDC Report, at 2.

[82] See, e.g., German Competition Act, at Section 19a(1), stating that, in determining the paramount significance for competition across an undertaking’s markets, there shall be particular account taken of its dominant position; financial strength or access to other resources; vertical integration; access to data relevant to competition; and the relevance of its activities for third-party access to supply and sales markets. See also DMCC, at S. 5 and S.6 (substantial and entrenched market power is a cumulative criterion, together with a position of strategic significance); DMA, at Recital 5 and Art. 3 (market power is irrelevant because the criteria for designation are (a) having a significant impact on the internal market; (b) providing a core platform service that is an important gateway for business users to reach end users; and (c) enjoying an entrenched and durable position). PL 2768 does not mention market power, and instead references control of essential access; the U.S. tech bills do not define covered platforms on the basis of market power either.

[83] The DMA explicitly rejects it. See Recital 23.

[84] Examples include online-intermediation services, online search engines, online social-networking services, and video-sharing platform services. See DMA, at Art. 2.

[85] See Elise Dorsey, Geoffrey A. Manne, Jan M. Rybnicek, Kristian Stout, and Joshua D. Wright, Consumer Welfare & the Rule of Law: The Case Against the New Populist Antitrust Movement, 47 Pepp. L. Rev. 861, 916 (2020).

[86] There are some exceptions to this. Some digital competition regulations seem to incorporate consumer-welfare considerations. One example is the KFTC’s recently proposed digital competition regulation, which is putatively aimed at protecting business users and consumers, and would allow for an efficiency defense. See Lee & Ko, supra note 21.

[87] See supra note 76.

[88] See infra Sections II.C and II.D.

[89] On the essential-facilities doctrine in the United States, see Philip K. Areeda, Essential Facilities: An Epithet in Need of Limiting Principles, 58 Antitrust L.J. 841 (1990); ever since the Supreme Court’s ruling in Trinko, no plaintiff has successfully litigated an essential-facilities claim to judgment. See, Verizon Communications, Inc. v. Law Offices of Curtis V. Trinko, LLP, 540 U.S. 398 (2003) (“As a general matter, the Sherman Act “does not restrict the long recognized right of [a] trader or manufacturer engaged in an entirely private business, freely to exercise his own independent discretion as to parties with whom he will deal.’”) (citations omitted).

[90] Communication from the Commission — Guidance on the Commission’s enforcement priorities in applying Article 82 of the EC Treaty to abusive exclusionary conduct by dominant undertakings (2009), at recital 13.

[91] Richard Whish & David Bailey, Competition Law (10th ed. 2021), at 142-3.

[92] See, e.g., Christopher M. Seelen, The Essential Facilities Doctrine: What does it Mean to be Essential? 80 Marq. L. Rev. 1117, 1123 (1997), discussing free-riding and the moral-hazard considerations implicit in defining essential facilities as essential to a competitor, rather than to competition. (“[A]pplication of the doctrine often focuses unduly on the effect of the denial of access on the plaintiff’s ability to compete-not on the infringement of competition which is the objective of the antitrust law.” (citations omitted), and at 1124 (“There exists a moral hazard when plaintiffs bring an essential facility claim against a single competitor. Indeed, firms might try to use the doctrine to take a ‘free ride’ on the efforts of a competitor.”). See also, Verband Deutscher Wetterdienstleister v. Google, Reference No. 408 HKO 36/13, Court of Hamburg (Apr. 4, 2013), 4, available at (“[A]pplicant’s members have been participating and will continue to participate in Google Search as ‘free riders.’ They demand favorable positioning without offering compensation.”); Continental T.V., Inc. v. GTE Sylvania Inc., 433 U.S. 36 (1977) (applying the rule of reason to territorial restrictions because they might be imposed by a manufacturer who wishes to prevent dealers from free-riding on point-of-sale services provided by another dealer).

[93] See, e.g., Brown Shoe Co. v. United States, 370 U.S. 294, 344 (1962) (“It is competition, not competitors, which the Act protects.”). See also Donna E. Patterson and Carl Shapiro, Transatlantic Divergence in GE/Honeywell: Causes and Lessons, Antitrust 18 (2001); Maureen K. Ohlhausen & John M. Taladay, Are Competition Officials Abandoning Competition Principles?, 13 J. Comp. L. & Practice 463 (2022).

[94] See, e.g., Trinko at 408; Pac. Bell Tel. Co. v. Linkline Commc’ns, Inc., 555 U.S. 438, 448 (2009); Chavez v. Whirlpool Corp., 113 Cal. Rptr. 2d, 182-83 (Ct. App. 2001); Foremost Pro Color, Inc. v. Eastman Kodak Co., 703 F.2d 534, 545 (9th Cir. 1983) (“The antitrust laws [do] not impose a duty on [firms] . . . to assist [competitors] . . . to ‘survive or expand.’”) (citations omitted).

[95] Mario Monti, Speech at the Third Nordic Competition Policy Conference, Stockholm: Fighting Cartels Why and How? Why Should We be Concerned with Cartels and Collusive Behaviour? (Sept. 11, 2000); see also Trinko at 408 (characterizing cartels as “the supreme evil of antitrust”).

[96] Although there is a rebuttable presumption to the contrary, undertakings can argue that agreements containing hardcore restrictions should benefit from an individual exemption under Article 101(3) TFEU. See Judgment of 13 October 2011, Pierre Fabre, C?439/09, ECLI:EU:C:2011:649. Moreover, “hardcore restrictions,” like cartels, need to be restrictions of competition “by object,” within the meaning of Art. 101(1) TFEU. Undertakings can hence try to demonstrate that a given hardcore restriction, examined in its economic and legal context, is objectively justified and does not fall within the prohibition laid down in Article 101(1) TFEU. See Opinion of Advocate General Wahl delivered on 16 July 2017, Coty, C-230/16, ECLI:EU:C:2017:603.

[97] For an extensive set of views opposing those endorsed by proponents of digital competition regulations, see, e.g., The Global Antitrust Institute Report on the Digital Economy (Joshua D. Wright & Douglas H. Ginsburg, eds., Nov. 11, 2020),

[98] See ABA letter, supra note 41.

[99] Law No. 12.529 of 30 November, 2011 (Braz.), available at

[100] PL 2768, art. 4.

[101] See Section I.

[102] See Section I.

[103] See, e.g., Rambus v. Fed. Trade Comm’n, 522 F.3d 456, 459 (D.C. Cir. 2008) (“[D]eceit merely enabling a monopolist to charge higher prices than it otherwise could have charged—would not in itself constitute monopolization.”). See also Judgment of 4 August 2023, Meta Platforms v. Bundeskartellamt, Case C 252-21, ECLI:EU:C:2023:537.

[104] For example, where a small company increases prices or downgrades its product, this can generally be corrected through competition, as the company will lose market share and be forced out of the market unless it changes its behavior. But when the same outcome is achieved through restrictions of competition or the misuse of market power, the market may be unable to respond effectively, and intervention may become necessary.

[105] We question whether this was ever the true intent behind digital competition regulation, see Section IIII.C.

[106] See also Section IIB.

[107] See, e.g., Svend Albaek, Consumer Welfare in EU Competition Policy, in Aims and Values in Competition Law, 67, 75 (Caroline Heide-JørgensenUlla NeergaardChristian Bergqvist, & Sune T. Poulsen eds., 2013) (“In practice it turns out that we should understand ‘consumers’ as customers rather than ‘real’ or ‘final’ consumers. Paragraph 84 of the General Guidelines takes a first step towards clarifying this: ‘[C]onsumers within the meaning of Article 81(3) are the customers of the parties to the agreement and subsequent purchasers.”); see also Article 102 (c) TFEU, which prohibits dominant companies from “applying dissimilar conditions to equivalent transactions with other trading parties, thereby placing them at a competitive disadvantage” (emphasis added). For a U.S. perspective, see, e.g., Kenneth Heyer, Welfare Standards and Merger Analysis: Why Not the Best? 2 Comp. Pol’y Int’l 29 (2006).

[108] In the United States, the clearest exponent of these ideas was Justice Louis D. Brandeis, who coined the term “curse of bigness” to refer to the material, social, and political ills that accompanied large corporations. See, e.g., Louis D. Brandeis, The Curse of Bigness: Miscellaneous Papers of Louis D. Brandeis (Osmond K. Fraenkel ed., 1934); in Europe, the notion is associated with the ordoliberal school. See, e.g., Wilhelm Roepke, A Humane Economy: The Social Framework of the Free Market (2014) at 32 (“If we want to name a common denominator for the social disease of our times, then it is concentration”).

[109] See, e.g. Wu, 2018 supra note 65; Sally Lee, Tim Wu Explains How Big Tech is Crippling Democracy, Columbia Magazine (Spring 2019) Asked whether bigness must be bad by its very nature, Tim Wu replies: “well, it’s designed to put its own interests over human interests, to grow like a cancer, and to never die. I once heard someone say that if a corporation were a person, it would be a sociopath. Which brings us to the real question: who is this country for? For humans or these artificial entities?”; See also Khan & Teachout, 2014, supra note 65, at 37. “Ever-increasing corporate size and concentration undercut democratic self-governance by disproportionately influencing governmental actors, as recognized by campaign finance reformers.”; and at 40-1. “Antitrust means, for us, government power to limit company size and concentration; this incarnation is an ethos, not a legal term.”

[110] See, e.g., Amanda Lotz, “Big Tech” Isn’t One Big Monopoly — It’s 5 Companies All in Different Businesses, The Conversation (Mar. 23, 2018),; Isobel A. Hamilton, Tim Cook Says He‘s Tired of Big Tech Being Painted as a Monolithic Force That Needs Tearing Apart, Business Insider (May 7, 2019), (“Tech is not monolithic. That would be like saying ‘all restaurants are the same,’ or ‘all TV networks are the same.’”) See also Nicolas Petit, Big tech and the Digital Economy: The Moligopoly Scenario (2022); Frank H. Easterbrook, Cyberspace and the Law of the Horse, 1996 U. Chi. Leg. Forum 207 (1996).

[111] See Friso Bostoen, Understanding the Digital Markets Act, 68 Antitrust Bull. 263, 282 (2023) (“It is difficult to find a common thread here. For starters, NIICS and cloud services are one-sided rather than multisided, so they can hardly be core platform services”).

[112] See Lazar Radic, Gatekeeping, the DMA, and the Future of Competition Regulation, Truth on the Market (Nov. 8, 2023), On tech disruption of traditional industries, see Adam Hayes, 20 Industries Threatened by Tech Disruption, Investopedia (Jan. 23, 2022),; on the bipartisan hostility toward “Big Tech” in the United States, see Nitasha Tiku, How Big Tech Became a Bipartisan Whipping Boy, Wired (Oct. 23, 2017),

[113] See, e.g., Lina Khan, Amazon’s Antitrust Paradox, 126 Yale L.J. 710 (2017); Zephyr Teachout & Lina Khan, Market Power and Political Law: A Taxonomy of Power, 9 Duke J. Const. L. & Pub. Pol’y 37 (2014); Kirk Ott, Event Notes: The Consumer Welfare Standard is Dead, Long Live the Standard, ProMarket (Nov.1, 2022),; Zephyr Teachout, The Death of the Consumer Welfare Standard, ProMarket (Nov. 7, 2023),

[114] See, e.g., Rana Foohar, The Great US-Europe Antitrust Divide, Financial Times (Feb. 5, 2024),

[115] Neo-Brandeisians often argue that antitrust law should strive to uphold a dispersed market structure and protect small business. See, e.g., Lina Khan & Sandeep Vaheesan. Market Power and Inequality: The Antitrust Counterrevolution and its discontents, 11 Harv. L. & Pol’y Rev. 235 (2017), at 237. “Antitrust laws must be reoriented away from the current efficiency focus toward a broader understanding that aims to protect consumers and small suppliers from the market power of large sellers and buyers, maintain the openness of markets, and disperse economic and political power.”

[116] See Khan & Vaheesan, supra note 122 at 236-7 (2017). “Antitrust laws historically sought to protect consumers and small suppliers from noncompetitive pricing, preserve open markets to all comers, and disperse economic and political power. The Reagan administration—with no input from Congress—rewrote antitrust to focus on the concept of neoclassical economic efficiency”; and, at 294, “It is important to trace contemporary antitrust enforcement and the philosophy underpinning it to the Chicago School intellectual revolution of the 1970s and 1980s, codified into policy by President Reagan. By collapsing a multitude of goals into the pursuit of narrow ‘economic efficiency,’ both scholars and practitioners ushered in standards and analyses that have heavily tilted the field in favor of defendants.”

[117] See, e.g., Nicolas Petit, Understanding Market Power (Robert Schuman Centre for Advanced Studies Working Paper No. RSC 14, 1, 2022) (“Antitrust laws are concerned with undue market power. In an economic conception of the law, antitrust rules of liability strike down anticompetitive business conduct or mergers that represent illegitimate market power strategies.”).

[118] On inefficient and efficient market exit, see Dirk Auer & Lazar Radic, The Growing Legacy of Intel, 14 J. Comp. L. & Prac. 15 (2023).

[119] According to some, the interpretation of market power as synonymous with size and concentration is the European reading of the concept. See Petit, supra note 124, at 1 (“When European antitrust lawyers think about market power, they do not direct their attention to consumer prices. They think about corporate size and industrial concentration, see giant American firms, and deduce that they have a domestic market power problem.”).

[120] See, e.g., Or Brook, Non-Competition Interests in EU Antitrust Law: An Empirical Study of Article 101 TFEU (1st ed. 2023), discussing the different goals and values of EU competition law throughout the years; Konstantinos Stylianou & Marcos Iacovides, The Goals of EU Competition Law: A Comprehensive Empirical Investigation, 42 Legal Studies 1, 17-8 (2020). “EU competition law is not monothematic but pursues a multitude of goals historically and today;” In the United States, see Robert H. Bork, The Antitrust Paradox: A Policy at War with Itself, 7 (1978) (finding the collection of socio-political goals at the time to be “mutually incompatible”); Joshua D. Wright & Douglas H. Ginsburg, The Goals of Antitrust: Welfare Trumps Choice, 81 Fordham L Rev 2405, 2405 (2013). “The Court interpreted the Sherman and Clayton Acts to reflect a hodgepodge of social and political goals…”; Thomas A. Lambert & Tate Cooper, Neo-Brandeisianism’s Democracy Paradox, University, 49 Journal of Corporation Law, 18 (2023).“In the mid-Twentieth Century, U.S. courts embraced the sort of multi-goaled deconcentration agenda Neo-Brandeisians advocate;” and Joshua D. Wright, Elyse Dorsey, Jonathan Klick, & Jan M. Rybnicek, Requiem for a Paradox: The Dubious Rise and Inevitable Fall of Hipster Antitrust, 51 Ariz. St. L. J. 293, 300-1 (2019) (discussing multi-goaled approach of mid-20th-century antitrust).

[121] See, e.g., Ioannis Lianos, Polycentric Competition Law, 71 Current Legal Problems 161 (2019); Maurice E. Stucke, Reconsidering Antitrust’s Goals, 53 B.C.L. Rev. 551, 551 (2012), “[t]he quest for a single economic goal has failed…this article proposes how to integrate antitrust’s multiple policy objectives into the legal framework.”; The Consumer Welfare Standard in Antitrust: Outdated or a Harbor in a Sea of Doubt?: Hearing Before the Subcomm. on Antitrust, Competition and Consumer Rights of the S. Comm. on the Judiciary, 115th Cong. (2017) (statement of Barry Lynn), arguing for the return to a “political antitrust”; Dina I. Walked, Antitrust as Public Interest Law: Redistribution, Equity, and Social Justice, 65 Antitrust Bull 87, 87 (2020), “[o]nce we frame antitrust as public interest law, in its broadest sense, we are empowered to use it to address inequality;” Saksham Malik, Social Justice as a Goal of Competition Policy, Kluwer Competition Law Blog (Feb. 23, 2024),

[122] It is no coincidence that critics of the “status quo” consistently attempt to cast economic analysis and (certain) antitrust case-law as a mistake brought about by judges adhering to the ideology of “neoliberalism,” rather than as the result of organic, piecemeal progression. See Khan & Vaheesan, supra note 122.

[123] Magrethe Vestager, Keynote of EVP Vestager at the European Competition Law Tuesdays: A Principles Based Approach to Competition Policy (Oct. 25, 2022),; See also Ohlhausen & Taladay, supra note 70 at 465.

[124] See, supra note 8.

[125] See also Press Release, Antitrust: Commission Accepts Commitments by Amazon Barring It from Using Marketplace Seller Data and Ensuring Equal Access to BuyBox and Prime, European Commission (Dec. 20, 2022),

[126] For commentary, see Lazar Radic, Apple Fined at the 11th Hour Before DMA Enters into Force, Truth on the Market (Mar. 5 2024),

[127] Giuseppe Colangelo (@GiuColangelo), Twitter (Oct. 5, 2023, 2:37 PM),

[128] Digital Platform Services Inquiry, supra note 18 at 14.

[129] Standing Committee on Finance, supra note 22, at 28, 38-39.

[130] Id. at 30.

[131] Shivi Gupta & Mansi Raghav, Digital Competition Law Committee to Finalise Report by August 2023, Lexology (Jul. 31, 2023),; Whole Government Approach to be Adopted for Digital Competition Laws, The Economic Times (Jul. 4 2023),

[132] The Competition Act, 2002, No.12 of 2003 (India), available at

[133] Antitrust, Regulation, and the Next World Order, supra note 53.

[134] See, e.g., the DMA’s definition of “fairness.” DMA, Recital 4.

[135] Ex Ante Regulation in Digital Markets – Background Note, DAF/COMP(2021)15, 16, OECD (Dec. 1, 2021) (“Framing regulations in terms of fairness may therefore also refer to redistribution, better treatment of users, or a host of other goals”). See also id. at 19.

[136] Pablo Ibanez Colomo, The Draft Digital Markets Act: A Legal and Institutional Analysis, 12 Journal of Competition Law & Practice 561, 562 (2021). See also id. at 565(“The driver of many disputes that may superficially be seen as relating to leveraging can be more rationalised, more convincingly, as attempts to re-allocate rents away from vertically-integrated incumbents to rivals”).

[137] See, e.g., Fiona Scott Morton & Cristina Caffarra, The European Commission Digital Markets Act: A Translation, VoxEU (Jan. 5, 021), We contest the assertion that the DMA and other digital competition regulations aim to create competition, rather than aid competitors, in Section IIIB.

[138] On the relationship between rent seeking and ex-ante regulation, see generally Thom Lambert, Rent-Seeking and Public Choice in Digital Markets, in The Global Antitrust Institute Report on the Digital Economy (Joshua D. Wright & Douglas H. Ginsburg, eds., Nov. 11, 2020).

[139] See, e.g., Making the Digital Market Easier to Use: The Act on Improving Transparency and Fairness of Digital Platforms (TFDPA), Japanese Ministry of Economy, Trade, and Industry (Apr. 23, 2021), The Ministry of Economy, Trade, and Industry specifically links the TFDPA to benefits for SMEs; see also Ebru Gokce Dessemond, Restoring Competition in ”Winner-Took-All” Digital Platform Markets, UNCTAD (Feb. 4, 2020), (“Competition law provisions on unfair trade practices and abuse of superior bargaining position, as found in competition laws of Japan and the Republic of Korea, would empower competition authorities in protecting the interests of smaller firms vis-à-vis big platforms”).

[140] See DMA, Recital 2, referring to a significant degree of dependence of both consumers and business users. See, in a similar vein, DMA Recitals 20, 43, 75. On self-inflicted dependence, see Geoffrey A. Manne, The Real Reason Foundem Foundered, Int’l. Ctr. for Law & Econ. (2018), available at

[141] For commentary on how bans on self-preferencing benefit large, but less-efficient competitors, see Lazar Radic & Geoffrey A. Manne, Amazon Italy’s Efficiency Offense, Truth on the Market (Jan. 11, 2022),

[142] Adam Kovacevich, The Digital Markets Act’s “Statler & Waldorf” Problem, Medium (Mar. 7 2024), (arguing that the companies who lobbied for the DMA are content aggregators like Yelp, Tripadvisor, and; big app makers like Spotify, Epic Games, and; and rival search engines like Ecosia, Yandex, and DuckDuckGo).

[143] For example, Epic Games’ revenue in 2023 was roughly $5.6 billion. In 2023, Epic Games employed about 4,300 workers. See, respectively, and According to the OECD, a small and medium-sized enterprise is one that employs fewer than 250 people. Enterprises by Business Size (Indicator), OECD,More, (last accessed May 14, 2024).

[144] Mathieu Pollet, France to Prioritise Digital Regulation, Tech Sovereignty During EU Council Presidency, EurActiv (Dec. 14, 2021),

[145] See, e.g., Matthias Bauer & Fredrik Erixon, Europe’s Quest for Technology Sovereignty: Opportunities and Pitfalls, ECIPE (2020),; see also Dennis Csernatoni et al., Digital Sovereignty: From Narrative to Policy?, EU Cyber Direct (2022),

[146] Digital Sovereignty for Europe, European Parliament (2020), available at For further discussion, see Lazar Radic, Gatekeeping, the DMA, and the Future of Competition Regulation, Truth on the Market (Nov. 8, 2023),

[147] Press Release, Digital Markets Act: Commission Designates Six Gatekeepers, European Commission (Sep. 6, 2023),

[148] Online Intermediation Platforms Market Inquiry, Summary of Final Report, South African Competition Commission (2023),

[149] Note that the unfairness here stems from having the resources to invest in search-engine optimization.

[150] Id. at 3.

[151] Id.

[152] Id. at 10.

[153] Id, at 6, 9, 23, 32, and 67.

[154] It is becoming clearer and clearer that the test for compliance with DMA’s rules will be whether competitors and complementors enjoy an increase in market share. See Foo Yun Chee & Martin Coulter, EU’s Digital Markets Act Hands Boost to Big Tech’s Smaller Rivals, Reuters (Mar. 11 2024) The public-policy chief of Ecosia, one of Google’s competitors in search, had this to say about the implementation of the DMA: “the implementation of these new rules is a step in the right direction, but the proof of the pudding is always in the eating, and whether we see any meaningful changes in market share.”

[155] Even the DMA’s supporters accept that the regulation is not grounded in economics. Cristina Caffarra, Europe’s Tech Regulation is Not Economic Policy, Project Syndicate (Oct. 11, 2023),

[156] Press Release, Apple Announces Changes to iOS, Safari, and the App Store in the European Union, Apple Inc., (Jan. 25, 2024),

[157] Andy Yen, Apple’s DMA Compliance Plan Is a Trap and a Slap in the Face for the European Commission, Proton (2024),; Press Release, Apple’s Proposed Changes Reject the Goals of the DMA, Spotify (Jan. 26, 2024),; Morgan Meaker, Apple Isn’t Ready to Release Its Grip on the App Store (Jan. 26, 2024),

[158] See, supra note 125 (discussing who lobbied for the DMA).

[159] DMCC, S. 38-45.

[160] See, A New Pro-competition Regime for Digital Markets: Policy Summary Briefing, UK Department for Business & Trade & Department for Science Innovation & Technology (2023),; see also, A New Pro-Competition Regime for Digital Markets. Consultation Document, UK Department for Culture, M. S. and Department for Business Energy & Industrial Strategy (2022), available at

[161] Dirk Auer, Matthew Lesh, & Lazar Radic, Digital Overload: How the Digital Markets, Competition and Consumers Bill’s Sweeping New Powers Threaten Britain’s Economy, Institute of Economic Affairs (Sep. 18, 2023),

[162] See also Alfonso Lamadrid & Pablo Ibáñez Colomo, The DMA – Procedural Afterthoughts, Chillin’ Competition (Sep. 5, 2022), (“Unlike competition law, the DMA is not so much about protecting consumers, but competitors/ third parties”); Chee & Coulter, supra note 137. “As the world’s biggest tech companies revamp their core online services to comply with the European Union’s landmark Digital Markets Act, the changes could give some smaller rivals and even peers a competitive edge.”

[163] See, e.g., Judgment of 6 September 2016, Intel v. Commission, Case C?413/14 P, EU:C:2017:632, para. 134 (“Thus, not every exclusionary effect is necessarily detrimental to competition. Competition on the merits may, by definition, lead to the departure from the market or the marginalisation of competitors that are less attractive to consumers from the point of view of, among other things, price, choice, quality or innovation”) (emphasis added).

[164] See, Auer & Radic, supra note 91.

[165] See, e.g., Competition on the Merits, DAF/COMP(2005)27, 9, OECD (2005), available at (“It is widely agreed that the purpose of competition policy is to protect competition, not competitors”).

[166] Ohlhausen & Taladay, supra note 70 at 465.

[167] See e.g., Questions and Answers, Digital Markets Act: Ensuring Fair and Open Digital Markets, European Commission (Sep. 6, 2023), (“[Gatekeepers] will therefore have to proactively implement certain behaviours that make the markets more open and contestable”).

[168] Jan Krämer & Daniel Schnurr, Big Data and Digital Markets Contestability: Theory of Harm and Data Access Remedies, 18 Journal of Competition Law & Economics 255 (2021).

[169] On self-preferencing in the context of antitrust, see Radic & Manne, supra note 113.

[170] On data portability and free-riding, see Sam Bowman, Data Portability: The Costs of Imposed Openness, Int’l. Ctr. for Law & Econ. (2020), available at

[171] H.R. 3849, supra note 29.

[172] Id. at § 7.

[173] Id. at § 7(b)(4).

[174] Id. at § 4(e)(1).

[175] Remarks by Executive-Vice President Vestager and Commissioner Breton on the Opening of Non-Compliance Investigations under the Digital Markets Act, European Commission (Mar. 25 2024), “Stakeholders provided feedback on the compliance solutions offered. Their feedback tells us that certain compliance measures fail to achieve their objectives and fall short of expectations.”

[176] The terms “levelling down” and “levelling up” are, to our knowledge, not normally deployed in the fields of antitrust law and digital competition regulation. They are, however, used frequently in areas of constitutional law, such as equality and free speech. In the context of equality law, see generally Deborah L. Brake, When Equality Leaves Everyone Worse Off: The Problem of Levelling Down in Equality Law, 46 Wm. & Mary L. Rev. 513 (2004). Examples include achieving equality between men and women by levelling down men’s opportunities until they reach parity with women’s, or levelling down public spending in wealthier school districts to reach equality with poorer districts.

[177] See, e.g., DMA Art. 6(7), establishing a duty to provide interoperability with the gatekeepers’ services, free of charge; see also arts.5(4), 5(10), 6(8), 6(9), and 7(1).

[178] See, e.g., Dirk Auer & Geoffrey A. Manne, TL;DR: Apple v Epic: The Value of Closed Systems, Int’l. Ctr. for Law & Econ. (Apr. 20, 2021), available at

[179] This argument was accepted in the context of in-app payment systems by the U.S. District Court in Epic Games, Inc. v. Apple Inc., 4:20-cv-05640-YGR (N.D. Cal. Nov. 9, 2021). On the security and privacy risks posed by sideloading and interoperability, see, e.g., Mikolaj Barczentewicz, Privacy and Security Implications of Regulation of Digital Services in the EU and in the U.S., Stanford-Vienna Transatlantic Technology Law Forum TTLF Working Papers No. 84 (2022); Bjorn Lundqvist, Injecting Security into European Tech Policy, CEPA (2023),

[180] “Open” and “closed” platforms are not synonymous with “good” and “bad” platforms. These are legitimate differences in product design and business philosophy, and neither is inherently more restrictive than the other. Andrei Haigu, Proprietary vs. Open Two-Sided Platforms and Social Efficiency, Harvard Business School Strategy Unit Working Paper No. 09-113 (2007), 2-3 (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”); see also Jonathan M. Barnett, The Host’s Dilemma: Strategic Forfeiture in Platform Markets for Informational Goods, 124 Harv. L. Rev. 1861, 1927 (2011).

[181] See, e.g., Bus. Elecs. Corp. v. Sharp Elecs. Corp., 485 U.S. 717, 748– 49 (1988) (Stevens, J., dissenting) (“A demonstrable benefit to interbrand competition will outweigh the harm to intrabrand competition that is caused by the imposition of vertical nonprice restrictions on dealers”); Leegin Creative Leather Products, Inc. v. PSKS, Inc., 551 U.S. 877 (2007) (“For, as has been indicated already, the antitrust laws are designed primarily to protect interbrand competition, from which lower prices can later result”).

[182] Issue Spotlight: Self-Preferencing, Int’l. Ctr. for Law & Econ. (last updated Nov. 10, 2022),

[183] Sam Bowman & Geoffrey A. Manne, Platform Self-Preferencing Can be Good for Consumers and Even Competitors, Truth on the Market (Mar. 4, 2021),

[184] Kadir Bas & Kerem Cem Sanli, Amendments to E-Commerce Law: Protecting or Preventing Competition?, Marmara University Law School Journal (2024) (forthcoming).

[185] Id. at 10.

[186] Id. at 21.

[187] Id. at 5 (emphasis added).

[188] DMCC, S. 20(3)(c).

[189] Auer, Lesh, & Radic, supra note 128.

  • [190] Joseph A. Schumpeter, Capitalism, Socialism, and Democracy, 100-1 (Harper and Row, New York 1942), 100-1 (“[t]here cannot be any reasonable doubt that under the conditions of our epoch such [technological] superiority is as a matter of fact the outstanding feature of the typical large-scale unit of control”); Hadi Houlla & Aurelien Portuese, The Great Revealing: Taking Competition in America and Europe Seriously, ITIF 23 (2023). (“In highly innovative industries, greater firm size and concentration lower industry-wide costs. A European study shows that larger high-tech firms could increase technological knowledge better than smaller ones…When economies of scale or network effects are large, firms must be sufficiently large to be efficient”); William Baumol, The Free Market Innovation Machine (2002), 196 (“Oligopolistic competition among large, high-tech, business firms, with innovation as a prime competitive weapon, ensures continued innovative activities and, very plausibly, their growth. In this market form, in which a few giant firms dominate a particular market, innovation has replaced price as the name of the game in a number of important industries.”).

[191] Two-sided markets connect distinct sets of users whose demands for the platform are interdependent—i.e., consumers’ demand for a platform increases as more products are available and, conversely, sellers’ demand for a platform increases as additional consumers use the platform, increasing the overall potential for transactions. These network effects can be direct (more consumers on one side attract more consumers on the same side), or indirect (more consumers on one side attract more consumers on the other side). See Bruno Jullien, Alessandro Pavan, & Marc Rysman, Two-Sided Markets, Pricing and Network Effects, 4 Handbook of Industrial Organization 485, 487 (2021)(“A central aspect of platform economics is the role of network effects, which apply when a product is valued based on the extent to which other market participants adopt or use the same product”); OECD Policy Roundtables, Two-Sided Markets 11 (Dec. 17, 2009), available at

[192] Art. 14 DMA establishes a duty to report mergers that would ordinarily fall under the relevant EU merger-control rules threshold. Art. 18(2) also empowers the Commission to prohibit gatekeepers from entering into future concentrations concerning core platform services or any digital products or services, in cases where gatekeepers have engaged in “systematic non-compliance.” Systemic noncompliance occurs when a gatekeeper receives as few as three noncompliance decisions within eight years (Art. 18(3)); S. 55 of the DMCC mandates companies with SMS to notify certain mergers, even though the UK does not have a compulsory notification regime.

[193] For a tongue-in-cheek remark, see Herbert Hovenkamp (@Sherman1890), Twitter (Jan. 15, 2024, 7:22 AM),; see also Robert Armstrong & Ethan Wu, What Big Tech Antitrust Gets Wrong, An Interview with Herbert Hovenkamp, Financial Times (Jan. 19, 2024), (“With Big Tech, we’re looking at probably the most productive part of the economy. The rate of innovation is high. They spend a lot of money on R&D. They are among the largest patent holders. There’s very little evidence of collusion. They seem to be competing with each other quite strongly. They pay their workers relatively well and have fairly educated workforces. None of this is a sign that these are industries we should be pursuing. That doesn’t mean they don’t do some anti-competitive things. But the whole idea that we should be targeting Big Tech strikes me as fundamentally wrong-headed”). It should be noted that Hovenkamp’s comment is made within the context of antitrust law. But the general sentiment about the unique hostility of certain regulators and legislatures toward certain tech companies could be extrapolated, mutatis mutandis, to digital competition regulation, especially with respect to the competition-oriented elements of DCRs (see Section II).

[194] See also Dirk Auer & Geoffrey Manne, Antitrust Dystopia and Antitrust Nostalgia: Alarmist Theories of Harm in Digital Markets and their Origins, 28(4) Geo. Mason L. Rev. 1279 (2023),

[195] Oles Andriychuk, Do DMA Obligations for Gatekeepers Create Entitlements for Business Users?, 11 Journal of Antitrust Enforcement 123 (2022) (Referring to the DMA as “punitive” and “interventionist,” and suggesting that exceptionally demanding obligations are put in place to slow down gatekeepers). See also at 127 (“the means for allowing the second-tier ersatz-Big Tech to scale up is punitive: to slow down the current gatekeepers by imposing upon them a catalogue of exceptionally demanding obligations”) and at 131 (“This punitive nature of the DMA also means that the obligations can be blatantly arduous and interventionist”) (emphasis added).

[196] DMA, Art. 7(9). There is also a limited exemption in which the gatekeeper can show that, due to exceptional circumstances beyond its control, complying with the obligations of the DMA endangers the economic viability of its operation in the EU. DMA, Art. 9(1).

[197] Id. (“…provided that such measures are strictly necessary and proportionate and are duly justified by the gatekeeper”) (emphasis added).

[198] Digital Markets Regulation Handbook, Cleary Gottlieb (Thomas Graf, et al., eds. 2022), 59,

[199] See Digital Platform Services Inquiry, supra note 18 at § 7.2.4.

[200] Id. at 123 (emphasis added).

[201] German Competition Act, supra note 50 at Art. 19a(7).

[202] As discussed in Section II, “material harm to competition” already establishes a lower (but also fundamentally different) threshold for the plaintiff than the standard typically applied in antitrust law, as it implies a showing of harm to competitors, rather than to competition.

[203] Cleary Gottlieb, supra note 162 at 76.

[204] DMCC at S. 29; see also, A New Pro-Competition Regime for Digital Markets: Advice for the Digital Markets Taskforce section 4.40, CMA (2020), available at (“Conduct which may in some circumstances be harmful, in others may be permissible or desirable as it produces sufficient countervailing benefits”).

[205] Auer, Lesh, & Radic, supra note 128.

[206] Id.

[207] Id.

[208] This is implied by the fact such an exemption arises only in S. 29, which concerns investigations into breaches of conduct requirements.

[209] PL 2768, supra note 32 at Art. 11.

[210] As discussed in Section I, PL 2768 pursues a multiplicity of goals, and there is no telling how much weight would be afforded to consumer protection under Art. 10.

[211] There is some evidence that this has already happened with Google and Google Maps. See Edith Hancock, “Severe Pain in the Butt”: EU’s Digital Competition Rules Make New Enemies on the Internet, Politico (Mar. 25 2024), (“Before [the DMA], users could search for a location on Google by simply clicking on the Google Map link to expand it and navigate it easily. That feature doesn’t work in the same way in Europe anymore and users are irritated.”).

[212] For the importance of interbrand competition between closed and open platforms, see ICLE Brief for the 9th Circuit in Epic Games v. Apple, No. 21-16695 (9th Cir.), ID: 12409936, Dkt Entry: 98, Int’l. Ctr. for Law & Econ. (Mar. 31, 2022), See also id. at 26 (“Even if an open platform led to more apps and IAP options for all consumers, some consumers may be better off as a result and others may be worse off. More vigilant users may avoid downloading apps and using IAP systems that are unreliable or which impose invasive data-sharing obligations, but less vigilant users will fall prey to malware, spyware, and other harmful content invited by an open system. The upshot is, “a more competitive market may be better at delivering to vigilant consumers what they want, but may end up exploiting more vulnerable consumers”). See also Mark Armstrong, Interactions Between Competition and Consumer Policy, Competition Policy International (2008),

[213] Robert Pitofsky, The Political Content of Antitrust, 127 Penn. L. Rev. 1051, 1058 (1979).

[214] See, generally, Christopher Decker, Modern Economic Regulation (2014).

[215] In the context of the DMCC, see Auer, Lesh, & Radic, supra note 128.

[216] Pinar Akman, Regulating Competition in Digital Platform Markets: A Critical Assessment of the Framework and Approach of the EU Digital Markets Act, 47(1) European Law Review 85, 110 (2023) ( “The description of “(un)fairness” as provided for in the DMA cannot be said to improve upon the position of the concept in competition law, as it, too, relies on an assessment that is ultimately subjective and involves a value judgement”). See also id. at n. 134 (“This is because it involves establishing what counts as an ‘imbalance of rights and obligations’ on the business users of a gatekeeper and what counts as an ‘advantage’ obtained by the gatekeeper from its business users that is ‘disproportionate’ to the service provided by the gatekeeper to its business users’; see DMA (n 2) Article 10(2)(a) DMA. On the vagueness of the ‘fairness’ concept embodied in the DMA from an economics perspective, see also Monopolkommission (n 38) [23].”). The report Akman references is: Recommendations for an Effective and Efficient Digital Markets Act, Special Report 82, Monopolkommission (2021),

[217] On the in-app payment commission being a legitimate way to recoup investments, see ICLE Brief in Epic Games v. Apple, supra note 177.

[218] Giuseppe Colangelo, In Fairness we (Should Not) Trust. The Duplicity of the EU Competition Policy Mantra in Digital Markets, 68 Antitrust Bulletin 618, 622 (2023) (“Despite its appealing features, fairness appears a subjective and vague moral concept, hence useless as a tool in decisionmaking”).

[219] As an example, Chapter III of the DMA is appropriately entitled: “Practices of Gatekeepers that Limit Contestability or are Unfair.” The chapter sets out practices that are, by definition, unfair.

[220] Distributing Dating Apps in the Netherlands, Apple Developer Support, (last visited Mar. 10, 2024),

[221] Press Release, Apple Announces Changes to IOS, Safari, and the App Store in the European Union, Apple Inc. (Jan. 25, 2024),

[222] Adam Kovacevich has referred to this as the “Stalter and Waldorf” problem. See, supra note 125.

[223] Trinko at 407 (2003) (“The mere possession of monopoly power, and the concomitant charging of monopoly prices, is not only not unlawful; it is an important element of the free-market system […] Firms may acquire monopoly power by establishing an infrastructure that renders them uniquely suited to serve their customers […] Enforced sharing also requires antitrust courts to act as central planners, identifying the proper price, quantity, and other terms of dealing—a role for which they are ill-suited”); see also Brian Albrecht, Imposed Final Offer Arbitration: Price Regulation by Any Other Name, Truth on the Market (Dec. 7, 2022),

[224] See, ICLE Brief in Epic Games v. Apple, supra note 174 (“In essence, Epic is trying to recast its objection to Apple’s 30% commission for use of Apple’s optional IAP system as a harm to consumers and competition more broadly”); on a similar trend in antitrust that we believe is even more relevant in the context of DCRs, see Jonathan Barnett, Antitrustifying Contract: Thoughts on Epic Games v. Apple and Apple v. Qualcomm, Truth on the Market (Oct. 26 2020)

[225] Andriychuk, supra note 159.

[226] See also Ohlhausen & Taladay supra note 69.

[227] United States v. Aluminum Co. of America, 148 F.2d 416, 430 (2d Cir. 1945).

[228] Decker, supra note 176.

[229] Many companies vertically integrate to have the ability to preference their own downstream or upstream products or services. See generally Eric Fruits, Geoffrey Manne, & Kristian Stout, The Fatal Economic Flaws of the Contemporary Campaign Against Vertical Integration, 68 Kan. L. Rev. 5 (2020),; Sam Bowman & Geoffrey Manne, Tl;DR: Self-Preferencing: Building an Ecosystem, Int’l. Ctr. for Law & Econ. (Jul. 21, 2020).

[230]Decker, supra note 176 at 190-1.

[231] Aldous Huxley, Point Counter Point (1928).

[232] As discussed, these ideas are, at least to some extent, redolent of the neo-Brandeisian school of thought in the United States and ordoliberalism in Europe. See e.g., Joseph Coniglio, Why the “New Brandeis Movement Gets Antitrust Wrong, Law360 (Apr. 24, 2018), (“The [neo-Brandeisian movement] is not a new entrant in the marketplace of ideas”); see also Daniel Crane, How Much Brandeis Do the Neo-Brandeisians Want?, 64 Antitrust Bulletin 4 (2019).

[233] See, e.g., Rupprecht Podszun, Philipp Bongartz, & Sarah Langenstein, Proposals on How to Improve the Digital Markets Act, 3, (Feb. 18, 2021), (“Critics who wish to place the tool into the realm of competition law miss the point that this is a fundamentally different approach”).

[234] In the EU, for example, the DMA was proposed on the basis of Article 114 TFEU, rather than Article 352 TFEU. The consequence is that, for the purpose of EU law, the DMA is considered internal market regulation, rather than competition legislation. It has been argued that Article 352 TFEU, or Article 114 TFEU in conjunction with Article 103 TFEU, would have been the more appropriate legal mechanism. See, e.g., Alfonso Lamadrid de Pablo & Nieves Bayón Fernández, Why the Proposed DMA Might be Illegal Under Article 114 TFEU, and How to Fix It, 12 J. Competition L. & Prac. 7, (2021). One reason why the Commission might have preferred to use Art.114 TFEU over Art.352 TFEU is that the process under Art.114 TFEU is less cumbersome. Unlike Art. 114 TFEU, Article 352 TFEU requires unanimity among EU member states and would not enable the European Parliament to function as co-legislator. Alfonso Lamadrid de Pablo, The Key to Understand the Digital Markets Act: It’s the Legal Basis, Chilling Competition (Dec. 03, 2020),

[235] The term is used often in the literature and media. For an example of the former, see William Davies & Nicholas Gane, Post-Neoliberalism? An Introduction, 38 Theory, Culture & Soc’y 3 (2021); for an example of the latter, see Rana Fohar, The New Rules for Business in a Post-Neoliberal World, Financial Times (Oct. 9, 2022),

[236] Thomas Biebricher and Frieder Vogelmann have used the term in describing the views of ordoliberals on the role of the market and the state. Thomas Biebricher and Frieder Vogelmann, The Birth of Austerity: German Ordoliberalism and Contemporary Neoliberalism, 138-139 (2017).

[237] Davies and Gane, supra note 198 at 1 (“While events of 2020–21 have facilitated new forms of privatization of many public services and goods, they also signal, potentially, a break from the neoliberal orthodoxies of the previous four decades, and, in particular, from their overriding concern for the market”); see also Edward Luce, It’s the End of Globalism As We Know It, Financial Times (May 8, 2020), (“The past 40 years have been predicated on a complex system of neoliberalism that is slowly but surely coming undone, but as of yet, we don’t have any global replacement”); Paolo Gerbaudo, A Post-Neoliberal Paradigm is Emerging: Conversation with Felicia Wong, El Pais (Nov. 4, 2022),

[238] Zephyr Teachout, The Death of the Consumer Welfare Standard, ProMarket (Nov. 07, 2023),

[239] After Hillary Clinton lost the 2016 U.S. presidential election to Donald Trump, Barack Obama referred to history and progress in the United States as zig-zagging, rather than moving in a straight line. See, Statement by the President, White House Office of the Press Secretary (Nov. 09, 2016),

Continue reading
Antitrust & Consumer Protection

ICLE Comments to UK Competition and Markets Authority on AI Partnerships

Regulatory Comments Executive Summary We thank the Competition and Markets Authority (CMA) for this invitation to comment (ITC) on partnerships and other arrangements involving artificial intelligence (AI).[1] . . .

Executive Summary

We thank the Competition and Markets Authority (CMA) for this invitation to comment (ITC) on partnerships and other arrangements involving artificial intelligence (AI).[1] The International Center for Law & Economics (ICLE) is a nonprofit, nonpartisan global research and policy center founded with the goal of building the intellectual foundations for sensible, economically grounded policy. ICLE promotes the use of law & economics methodologies to inform public-policy debates and has longstanding expertise in the evaluation of competition law and policy. ICLE’s interest is to ensure that competition law remains grounded in clear rules, established precedent, a record of evidence, and sound economic analysis.

In our comments, we express concern that policymakers’ current concerns about competition in AI industries may be unwarranted. This is particularly true of the notion that incumbent digital platforms may use strategic partnerships with AI firms to insulate themselves from competition, including the three transactions that are central to the current ITC:

  1. Amazon’s partnership with Anthropic;
  2. Microsoft’s partnership with Mistral AI; and,
  3. Microsoft’s hiring of former Inflection AI employees (including, notably, founder Mustafa Suleyman) and related arrangements with the company.

Indeed, publicly available information suggests these transactions may not warrant merger-control investigation, let alone the heightened scrutiny that comes with potential Phase II proceedings. At the very least, given the AI industry’s competitive landscape, there is little to suggest these transactions merit closer scrutiny than similar deals in other sectors.

Overenforcement in the field of generative AI paradoxically could engender the very harms that policymakers currently seek to avert. As we explain in greater detail below, preventing so-called “big tech” firms from competing in these markets (for example, by threatening competition intervention as soon as they build strategic relationships with AI startups) may thwart an important source of competition needed to keep today’s leading generative-AI firms in check. In short, competition in AI markets is important[2], but trying naïvely to hold incumbent (in adjacent markets) tech firms back out of misguided fears they will come to dominate this space is likely to do more harm than good.

At a more granular level, there are important reasons to believe these agreements will have no negative impact on competition and they may, in fact, benefit consumers—e.g., by enabling those startups to raise capital and deploy their services at an even larger scale. In other words, they do not bear any of the prima facie traits of “killer acquisitions” or even of the acquisition of “nascent potential competitors”.[3]

Most importantly, these partnerships all involve the acquisition of minority stakes that do not entail any change of control over the target companies. Amazon, for instance, will not have “ownership control” of Anthropic. The precise amount of shares acquired has not been made public, but a reported investment of $4 billion in a company valued at $18.4 billion does not give Amazon a majority stake or sufficient voting rights to control the company or its competitive strategy. [4] It has also been reported that the deal will not give Amazon any seats on the Anthropic board or special voting rights (such as the power to veto some decisions).[5] There is thus little reason to believe Amazon has acquired indirect or de facto control over Anthropic.

Microsoft’s investment in Mistral AI is even smaller, in both absolute and relative terms. Microsoft is reportedly investing only $16 million in a company valued at $2.1 billion.[6] This represents less than 1% of Mistral’s equity, making it all but impossible for Microsoft to exert any significant control or influence over Mistral AI’s competitive strategy. Likewise, there have been no reports of Microsoft acquiring seats on Mistral AI’s board or special voting rights. We can therefore be confident that the deal will not affect competition in AI markets.

Much of the same applies to Microsoft’s dealings with Inflection AI. Microsoft hired two of the company’s three founders (which currently does not fall under the scope of merger laws), and also paid $620 million for nonexclusive rights to sell access to the Inflection AI model through its Azure Cloud.[7] Admittedly, the latter could entail (depending on deal’s specifics) some limited control over Inflection AI’s competitive strategy, but there is currently no evidence to suggest this will be the case.

Finally, none of these deals entails any competitively significant behavioral commitments from the target companies. There are no reports of exclusivity agreements or other commitments that would restrict third parties’ access to these firms’ underlying AI models. Again, this means the deals are extremely unlikely to negatively impact the competitive landscape in these markets.

At a more macro level, how the CMA deals with these proposed partnerships could have important ramifications for the UK economy. On the one hand, competition authorities (including the CMA) may be tempted to avoid the mistakes they arguably made during the formative years of what have become today’s largest online platforms.[8] The argument is that tougher enforcement may have reduced the high levels of concentration we see in these markets (the counterpoint is that these markets present features that naturally lead to relatively high levels of concentration and that this concentration benefits consumers in several ways[9]).

Unfortunately, this urge to curtail false negatives may come at the expense of judicial errors that hobble the UK economy. Discussing the EU’s AI Act during a recent interview, French President Emmanuel Macron implicitly suggested the UK is in a unique position to attract AI (and other tech) investments away from the European Union. In his words:

We can decide to regulate much faster and much stronger than our major competitors. But we will regulate things that we will no longer produce or invent. This is never a good idea…

When I look at France, it is probably the first country in terms of artificial intelligence in continental Europe. We are neck and neck with the British. They will not have this regulation on foundational models. But above all, we are all very far behind the Chinese and the Americans. [10]

To capitalise on this opportunity, however, the UK must foster a fertile environment for startup activity. The CMA’s approach to merger review in the AI industry is a small, but important, part of this picture. Looking at AI partnerships in an even-handed manner would signal a commitment to evidence-based policymaking that creates legal certainty for startups. For instance, sound merger-review principles would assure founders that corporate acquisition will remain a viable exit strategy in all but exceptional circumstances.

Of course, none of this is to say that established competition-law principles should play second fiddle to broader geopolitical ambitions. It does, however, suggest that the cost of false positives is particularly high in key industries like AI.

In short, how the CMA approaches these AI partnerships is of pivotal importance for both UK competition policy and the country’s broader economic ambitions. The CMA should therefore look at these partnerships with an open mind, despite the important political and reputational pressure to be seen as “doing something” in this cutting-edge industry. Generative AI is already changing the ways that many firms do business and improving employee productivity in many industries.[11] The technology is also increasingly useful in the field of scientific research, where it has enabled new complex models that expand scientists’ reach.[12] And while sensible enforcement is of vital importance to maintain competition and consumer welfare, it must be grounded in empirical evidence.

In the remainder of these comments, we will discuss the assumptions that underpin calls for heightened competition scrutiny in AI industries, and explain why they are unfounded. The big picture is that AI markets have grown rapidly, and new players are thriving. This would suggest that competition is intense. If incumbent firms could easily leverage their dominance into burgeoning generative-AI markets, we would not have seen the growth of such AI “unicorns” as OpenAI, Midjourney, and Anthropic, to name but a few. Furthermore, AI platforms developed by incumbent data collectors—such as Meta’s Llama or Google’s Bard, recently relaunched as Gemini—have struggled to gain traction. Of course, this is not to say that competition enforcers shouldn’t care about generative AI markets, but rather that there is currently no apparent need for increased competition scrutiny in these markets.

The comments proceed as follows. Section I summarises recent calls for competition intervention in generative-AI markets. Section II argues that many of these calls are underpinned by fears of data-related incumbency advantages (often referred to as “data-network effects”), including in the context of mergers. Section III explains why these effects are unlikely to play a meaningful role in generative-AI markets. Section IV concludes by offering five key takeaways to help policymakers better weigh the tradeoffs inherent to competition intervention (including merger control) in generative-AI markets.

I. Calls for Intervention in AI Markets

It was once (and frequently) said that Google’s “data monopoly” was unassailable: “If ‘big data’ is the oil of the information economy, Google has Standard Oil-like monopoly dominance—and uses that control to maintain its dominant position”.[13] Similar claims of data dominance have been attached to nearly all large online platforms, including Facebook (Meta), Amazon, and Uber.[14]

While some of these claims continue even today (for example, “big data” is a key component of the U.S. Justice Department’s (DOJ) Google Search and adtech antitrust suits),[15] a shiny new data target has emerged in the form of generative artificial intelligence (AI). The launch of ChatGPT in November 2022, as well as the advent of AI image-generation services like Midjourney and Dall-E, have dramatically expanded the public’s conception of what is—and what might be—possible to achieve with generative-AI technologies built on massive datasets.

While these services remain both in the early stages of mainstream adoption and in the throes of rapid, unpredictable technological evolution, they nevertheless already appear to be on the radar of competition policymakers around the world. Several antitrust enforcers appear to believe that, by acting now, they can avoid the “mistakes” that were purportedly made during the formative years of Web 2.0.[16] These mistakes, critics assert, include failing to appreciate the centrality of data in online markets, as well as letting mergers go unchecked and allowing early movers to entrench their market positions.[17] As Lina Khan, chair of the U.S. Federal Trade Commission (FTC), put it: “we are still reeling from the concentration that resulted from Web 2.0, and we don’t want to repeat the mis-steps of the past with AI”.[18]

This response from the competition-policy world is deeply troubling. Rather than engage in critical self-assessment and adopt an appropriately restrained stance, the enforcement community appears to be champing at the bit. Rather than assessing their prior assumptions based on the current technological moment, enforcers’ top priority appears to be figuring out how to rapidly and almost reflexively deploy existing competition tools to address the presumed competitive failures presented by generative AI.[19]

It is increasingly common for competition enforcers to argue that so-called “data-network effects” serve not only to entrench incumbents in those markets where the data is collected, but also to confer similar, self-reinforcing benefits in adjacent markets. Several enforcers have, for example, prevented large online platforms from acquiring smaller firms in adjacent markets, citing the risk that they could use their vast access to data to extend their dominance into these new markets.[20]

They have also launched consultations to ascertain the role that data plays in AI competition. For instance, in an ongoing consultation, the European Commission asks: “What is the role of data and what are its relevant characteristics for the provision of generative AI systems and/or components, including AI models?”[21] Unsurprisingly, the FTC has likewise been bullish about the risks posed by incumbents’ access to data. In comments submitted to the U.S. Copyright Office, for example, the FTC argued that:

The rapid development and deployment of AI also poses potential risks to competition. The rising importance of AI to the economy may further lock in the market dominance of large incumbent technology firms. These powerful, vertically integrated incumbents control many of the inputs necessary for the effective development and deployment of AI tools, including cloud-based or local computing power and access to large stores of training data. These dominant technology companies may have the incentive to use their control over these inputs to unlawfully entrench their market positions in AI and related markets, including digital content markets.[22]

Certainly, it stands to reason that the largest online platforms—including Alphabet, Meta, Apple, and Amazon—should have a meaningful advantage in the burgeoning markets for generative-AI services. After all, it is widely recognised that data is an essential input for generative AI.[23] This competitive advantage should be all the more significant, given that these firms have been at the forefront of AI technology for more than a decade. Over this period, Google’s DeepMind and AlphaGo and Meta’s NLLB-200 have routinely made headlines.[24] Apple and Amazon also have vast experience with AI assistants, and all of these firms use AI technology throughout their platforms.[25]

Contrary to what one might expect, however, the tech giants have, to date, been largely unable to leverage their vast data troves of data to outcompete startups like OpenAI and Midjourney. At the time of writing, OpenAI’s ChatGPT appears to be, by far, the most successful chatbot,[26] despite the large tech platforms’ apparent access to far more (and more up-to-date) data.

In these comments, we suggest that there are important lessons to glean from these developments, if only enforcers would stop to reflect. The meteoric rise of consumer-facing AI services should offer competition enforcers and policymakers an opportunity for introspection. As we explain, the rapid emergence of generative-AI technology may undercut many core assumptions of today’s competition-policy debates, which have focused largely on the rueful after-effects of the purported failure of 20th-century antitrust to address the allegedly manifest harms of 21st-century technology. These include the notions that data advantages constitute barriers to entry and can be leveraged to project dominance into adjacent markets; that scale itself is a market failure to be addressed by enforcers; and that the use of consumer data is inherently harmful to those consumers.

II. Data-Network Effects Theory and Enforcement

Proponents of more extensive intervention by competition enforcers into digital markets often cite data-network effects as a source of competitive advantage and barrier to entry (though terms like “economies of scale and scope” may offer more precision).[27] The crux of the argument is that “the collection and use of data creates a feedback loop of more data, which ultimately insulates incumbent platforms from entrants who, but for their data disadvantage, might offer a better product”.[28] This self-reinforcing cycle purportedly leads to market domination by a single firm. Thus, it is argued, e.g., that Google’s “ever-expanding control of user personal data, and that data’s critical value to online advertisers, creates an insurmountable barrier to entry for new competition”.[29]

But it is important to note the conceptual problems these claims face. Because data can be used to improve products’ quality and/or to subsidise their use, treating the possession of data as an entry barrier suggests that any product improvement or price reduction made by an incumbent could be a problematic entry barrier to any new entrant. This is tantamount to an argument that competition itself is a cognizable barrier to entry. Of course, it would be a curious approach to antitrust if competition were treated as a problem, as it would imply that firms should under-compete—i.e., should forego consumer-welfare enhancements—in order to inculcate a greater number of firms in a given market, simply for its own sake.[30]

Meanwhile, actual economic studies of data-network effects have been few and far between, with scant empirical evidence to support the theory.[31] Andrei Hagiu and Julian Wright’s theoretical paper offers perhaps the most comprehensive treatment of the topic to date.[32] The authors ultimately conclude that data-network effects can be of differing magnitudes and have varying effects on firms’ incumbency advantage.[33] They cite Grammarly (an AI writing-assistance tool) as a potential example: “As users make corrections to the suggestions offered by Grammarly, its language experts and artificial intelligence can use this feedback to continue to improve its future recommendations for all users”.[34]

This is echoed by other economists who contend that “[t]he algorithmic analysis of user data and information might increase incumbency advantages, creating lock-in effects among users and making them more reluctant to join an entrant platform”.[35] Crucially, some scholars take this logic a step further, arguing that platforms may use data from their “origin markets” in order to enter and dominate adjacent ones:

First, as we already mentioned, data collected in the origin market can be used, once the enveloper has entered the target market, to provide products more efficiently in the target market. Second, data collected in the origin market can be used to reduce the asymmetric information to which an entrant is typically subject when deciding to invest (for example, in R&D) to enter a new market. For instance, a search engine could be able to predict new trends from consumer searches and therefore face less uncertainty in product design.[36]

This possibility is also implicit in Hagiu and Wright’s paper.[37] Indeed, the authors’ theoretical model rests on an important distinction between within-user data advantages (that is, having access to more data about a given user) and across-user data advantages (information gleaned from having access to a wider user base). In both cases, there is an implicit assumption that platforms may use data from one service to gain an advantage in another market (because what matters is information about aggregate or individual user preferences, regardless of its origin).

Our review of the economic evidence suggests that several scholars have, with varying degrees of certainty, raised the possibility that incumbents may leverage data advantages to stifle competitors in their primary market or in adjacent ones (be it via merger or organic growth). As we explain below, however, there is ultimately little evidence to support such claims. Policymakers have, however, been keenly receptive to these limited theoretical findings, basing multiple decisions on these theories, often with little consideration given to the caveats that accompany them.[38]

Indeed, it is remarkable that, in its section on “[t]he data advantage for incumbents”, the “Furman Report” created for the UK government cited only two empirical economic studies, and they offer directly contradictory conclusions with respect to the question of the strength of data advantages.[39] Nevertheless, the Furman Report concludes that data “may confer a form of unmatchable advantage on the incumbent business, making successful rivalry less likely”,[40] and adopts without reservation “convincing” evidence from non-economists that have no apparent empirical basis.[41]

In the Google/Fitbit merger proceedings, the European Commission found that the combination of data from Google services with that of Fitbit devices would reduce competition in advertising markets:

Giving [sic] the large amount of data already used for advertising purposes that Google holds, the increase in Google’s data collection capabilities, which goes beyond the mere number of active users for which Fitbit has been collecting data so far, the Transaction is likely to have a negative impact on the development of an unfettered competition in the markets for online advertising.[42]

As a result, the Commission cleared the merger on the condition that Google refrain from using data from Fitbit devices for its advertising platform.[43] The Commission also appears likely to focus on similar issues in its ongoing investigation of Microsoft’s investment into OpenAI.[44]

Along similar lines, in its complaint to enjoin Meta’s purchase of Within Unlimited—makers of the virtual-reality (VR) fitness app Supernatural—the FTC relied on, among other things, the fact that Meta could leverage its data about VR-user behavior to inform its decisions and potentially outcompete rival VR-fitness apps: “Meta’s control over the Quest platform also gives it unique access to VR user data, which it uses to inform strategic decisions”.[45]

The DOJ’s twin cases against Google also implicate data leveraging and data barriers to entry. The agency’s adtech complaint charges that “Google intentionally exploited its massive trove of user data to further entrench its monopoly across the digital advertising industry”.[46] Similarly, in its Google Search complaint, the agency argues that:

Google’s anticompetitive practices are especially pernicious because they deny rivals scale to compete effectively. General search services, search advertising, and general search text advertising require complex algorithms that are constantly learning which organic results and ads best respond to user queries; the volume, variety, and velocity of data accelerates the automated learning of search and search advertising algorithms.[47]

Finally, updated merger guidelines published in recent years by several competition enforcers cite the acquisition of data as a potential source of competition concerns. For instance, the FTC and DOJ’s 2023 guidelines state that “acquiring data that helps facilitate matching, sorting, or prediction services may enable the platform to weaken rival platforms by denying them that data”.[48] Likewise, the CMA itself warns against incumbents acquiring firms in order to obtain their data and foreclose other rivals:

Incentive to foreclose rivals…

7.19(e) Particularly in complex and dynamic markets, firms may not focus on short term margins but may pursue other objectives to maximise their long-run profitability, which the CMA may consider. This may include… obtaining access to customer data….[49]

In short, competition authorities around the globe have been taking an increasingly aggressive stance on data-network effects. Among the ways this has manifested is in enforcement decisions based on fears that data collected by one platform might confer a decisive competitive advantage in adjacent markets. Unfortunately, these concerns rest on little to no empirical evidence, either in the economic literature or the underlying case records.

III. Data-Incumbency Advantages in Generative-AI Markets

Given the assertions detailed in the previous section, it would be reasonable to assume that firms such as Google, Meta, and Amazon should be in pole position to dominate the burgeoning market for generative AI. After all, these firms have not only been at the forefront of the field for the better part of a decade, but they also have access to vast troves of data, the likes of which their rivals could only dream when they launched their own services. Thus, the authors of the Furman Report caution that “to the degree that the next technological revolution centres around artificial intelligence and machine learning, then the companies most able to take advantage of it may well be the existing large companies because of the importance of data for the successful use of these tools”.[50]

To date, however, this is not how things have unfolded—although it bears noting these markets remain in flux and the competitive landscape is susceptible to change. The first significantly successful generative-AI service was arguably not from either Meta—which had been working on chatbots for years and had access to, arguably, the world’s largest database of actual chats—or Google. Instead, the breakthrough came from a previously unknown firm called OpenAI.

OpenAI’s ChatGPT service currently holds an estimated 60% of the market (though reliable numbers are somewhat elusive).[51] It broke the record for the fastest online service to reach 100 million users (in only a couple of months), more than four times faster than previous record holder TikTok.[52] Based on Google Trends data, ChatGPT is nine times more popular worldwide than Google’s own Bard service, and 14 times more popular in the United States.[53] In April 2023, ChatGPT reportedly registered 206.7 million unique visitors, compared to 19.5 million for Google’s Bard.[54] In short, at the time we are writing, ChatGPT appears to be the most popular chatbot. The entry of large players such as Google Bard or Meta AI appear to have had little effect thus far on its market position.[55]

The picture is similar in the field of AI-image generation. As of August 2023, Midjourney, Dall-E, and Stable Diffusion appear to be the three market leaders in terms of user visits.[56] This is despite competition from the likes of Google and Meta, who arguably have access to unparalleled image and video databases by virtue of their primary platform activities.[57]

This raises several crucial questions: how have these AI upstarts managed to be so successful, and is their success just a flash in the pan before Web 2.0 giants catch up and overthrow them? While we cannot answer either of these questions dispositively, we offer what we believe to be some relevant observations concerning the role and value of data in digital markets.

A first important observation is that empirical studies suggest that data exhibits diminishing marginal returns. In other words, past a certain point, acquiring more data does not confer a meaningful edge to the acquiring firm. As Catherine Tucker put it following a review of the literature: “Empirically there is little evidence of economies of scale and scope in digital data in the instances where one would expect to find them”.[58]

Likewise, following a survey of the empirical literature on this topic, Geoffrey Manne and Dirk Auer conclude that:

Available evidence suggests that claims of “extreme” returns to scale in the tech sector are greatly overblown. Not only are the largest expenditures of digital platforms unlikely to become proportionally less important as output increases, but empirical research strongly suggests that even data does not give rise to increasing returns to scale, despite routinely being cited as the source of this effect.[59]

In other words, being the firm with the most data appears to be far less important than having enough data. Moreover, this lower bar may be accessible to far more firms than one might initially think possible. Furthermore, obtaining sufficient data could become easier still—that is, the volume of required data could become even smaller—with technological progress. For instance, synthetic data may provide an adequate substitute to real-world data,[60] or may even outperform real-world data.[61] As Thibault Schrepel and Alex Pentland surmise:

[A]dvances in computer science and analytics are making the amount of data less relevant every day. In recent months, important technological advances have allowed companies with small data sets to compete with larger ones.[62]

Indeed, past a certain threshold, acquiring more data might not meaningfully improve a service, where other improvements (such as better training methods or data curation) could have a large impact. In fact, there is some evidence that excessive data impedes a service’s ability to generate results appropriate for a given query: “[S]uperior model performance can often be achieved with smaller, high-quality datasets than massive, uncurated ones. Data curation ensures that training datasets are devoid of noise, irrelevant instances, and duplications, thus maximizing the efficiency of every training iteration”.[63]

Consider, for instance, a user who wants to generate an image of a basketball. Using a model trained on an indiscriminate range and number of public photos in which a basketball appears surrounded by copious other image data, the user may end up with an inordinately noisy result. By contrast, a model trained with a better method on fewer, more carefully selected images could readily yield far superior results.[64] In one important example:

[t]he model’s performance is particularly remarkable, given its small size. “This is not a large language model trained on the whole Internet; this is a relatively small transformer trained for these tasks,” says Armando Solar-Lezama, a computer scientist at the Massachusetts Institute of Technology, who was not involved in the new study…. The finding implies that instead of just shoving ever more training data into machine-learning models, a complementary strategy might be to offer AI algorithms the equivalent of a focused linguistics or algebra class.[65]

Platforms’ current efforts are thus focused on improving the mathematical and logical reasoning of large language models (LLMs), rather than maximizing training datasets.[66]

Two points stand out. The first is that firms like OpenAI rely largely on publicly available datasets—such as GSM8K—to train their LLMs.[67] Second, the real challenge to create cutting-edge AI is not so much in collecting data, but rather in creating innovative AI-training processes and architectures:

[B]uilding a truly general reasoning engine will require a more fundamental architectural innovation. What’s needed is a way for language models to learn new abstractions that go beyond their training data and have these evolving abstractions influence the model’s choices as it explores the space of possible solutions.

We know this is possible because the human brain does it. But it might be a while before OpenAI, DeepMind, or anyone else figures out how to do it in silicon.[68]

Furthermore, it is worth noting that the data most relevant to startups in a given market may not be those held by large incumbent platforms in other markets, but rather data specific to the market in which the startup is active or, even better, to the given problem it is attempting to solve:

As Andres Lerner has argued, if you wanted to start a travel business, the data from Kayak or Priceline would be far more relevant. Or if you wanted to start a ride-sharing business, data from cab companies would be more useful than the broad, market-cross-cutting profiles Google and Facebook have. Consider companies like Uber, Lyft and Sidecar that had no customer data when they began to challenge established cab companies that did possess such data. If data were really so significant, they could never have competed successfully. But Uber, Lyft and Sidecar have been able to effectively compete because they built products that users wanted to use—they came up with an idea for a better mousetrap. The data they have accrued came after they innovated, entered the market and mounted their successful challenges—not before.[69]

The bottom line is that data is not the be-all and end-all that many in competition circles make it out to be. While data often may confer marginal benefits, there is little sense that these benefits are ultimately decisive.[70] As a result, incumbent platforms’ access to vast numbers of users and troves of data in their primary markets might only marginally affect their competitiveness in AI markets.

A related observation is that firms’ capabilities and other features of their products arguably play a more important role than the data they own.[71] Examples of this abound in digital markets. Google overthrew Yahoo in search, despite initially having access to far fewer users and far less data; Google and Apple overcame Microsoft in the smartphone operating system market, despite having comparatively tiny ecosystems (at the time) to leverage; and TikTok rose to prominence despite intense competition from incumbents like Instagram, which had much larger user bases. In each of these cases, important product-design decisions (such as the PageRank algorithm, recognizing the specific needs of mobile users,[72] and TikTok’s clever algorithm) appear to have played a far more significant role than initial user and data endowments (or lack thereof).

All of this suggests that the early success of OpenAI likely has more to do with its engineering decisions than with what data it did or did not possess. Going forward, OpenAI and its rivals’ ability to offer and monetise compelling stores offering custom versions of their generative-AI technology will arguably play a much larger role than (and contribute to) their ownership of data.[73] In other words, the ultimate challenge is arguably to create a valuable platform, of which data ownership is a consequence, but not a cause.

It is also important to note that, in those instances where it is valuable, data does not just fall from the sky. Instead, it is through smart business and engineering decisions that firms can generate valuable information (which does not necessarily correlate with owning more data).

For instance, OpenAI’s success with ChatGPT is often attributed to its more efficient algorithms and training models, which arguably have enabled the service to improve more rapidly than its rivals.[74] Likewise, the ability of firms like Meta and Google to generate valuable data for advertising arguably depends more on design decisions that elicit the right data from users, rather than the raw number of users in their networks.

Put differently, setting up a business so as to extract and organise the right information is more important than simply owning vast troves of data.[75] Even in those instances where high-quality data is an essential parameter of competition, it does not follow that having vaster databases or more users on a platform necessarily leads to better information for the platform.

Indeed, if data ownership consistently conferred a significant competitive advantage, these new firms would not be where they are today. This does not, of course, mean that data is worthless. Rather, it means that competition authorities should not assume that the mere possession of data is a dispositive competitive advantage, absent compelling empirical evidence to support such a finding. In this light, the current wave of decisions and competition-policy pronouncements that rely on data-related theories of harm are premature.

IV. Five Key Takeaways: Reconceptualizing the Role of Data in Generative-AI Competition

As we explain above, data network effects are not the source of barriers to entry that they are sometimes made out to be. The picture is far more nuanced. Indeed, as economist Andres Lerner demonstrated almost a decade ago (and the assessment is only truer today):

Although the collection of user data is generally valuable for online providers, the conclusion that such benefits of user data lead to significant returns to scale and to the entrenchment of dominant online platforms is based on unsupported assumptions. Although, in theory, control of an “essential” input can lead to the exclusion of rivals, a careful analysis of real-world evidence indicates that such concerns are unwarranted for many online businesses that have been the focus of the “big data” debate.[76]

While data can be an important part of the competitive landscape, incumbents’ data advantages are far less pronounced than today’s policymakers commonly assume. In that respect, five main lessons emerge:

  1. Data can be (very) valuable, but beyond a certain threshold, those benefits tend to diminish. In other words, having the most data is less important than having enough;
  2. The ability to generate valuable information does not depend on the number of users or the amount of data a platform has previously acquired;
  3. The most important datasets are not always proprietary;
  4. Technological advances and platforms’ engineering decisions affect their ability to generate valuable information, and this effect swamps effects stemming from the amount of data they own; and
  5. How platforms use data is arguably more important than what data or how much data they own.

These lessons have important ramifications for policy debates over the competitive implications of data in technologically evolving areas.

First, it is not surprising that startups, rather than incumbents, have taken an early lead in generative AI (and in Web 2.0 before it). After all, if data-incumbency advantages are small or even nonexistent, then smaller and more nimble players may have an edge over established tech platforms. This is all the more likely given that, despite significant efforts, the biggest tech platforms were unable to offer compelling generative-AI chatbots and image-generation services before the emergence of ChatGPT, Dall-E, Midjourney, etc.

This failure suggests that, in a process akin to Clayton Christensen’s “innovator’s dilemma”,[77] something about the incumbent platforms’ existing services and capabilities was holding them back in those markets. Of course, this does not necessarily mean that those same services or capabilities could not become an advantage when the generative-AI market starts addressing issues of monetisation and scale.[78] But it does mean that assumptions about a firm’s market power based on its possession of data are off the mark.

Another important implication is that, paradoxically, policymakers’ efforts to prevent Web 2.0 platforms from competing freely in generative-AI markets may ultimately backfire and lead to less, not more, competition. Indeed, OpenAI is currently acquiring a sizeable lead in generative AI. While competition authorities might like to think that other startups will emerge and thrive in this space, it is important not to confuse desires with reality. While there currently exists a vibrant AI-startup ecosystem, there is at least a case to be made that the most significant competition for today’s AI leaders will come from incumbent Web 2.0 platforms—although nothing is certain at this stage.

Policymakers should beware not to stifle that competition on the misguided assumption that competitive pressure from large incumbents is somehow less valuable to consumers than that which originates from smaller firms. This is particularly relevant in the context of merger control. An acquisition (or an “acqui-hire”) by a “big tech” company does not only, in principle, entail a minor risk to harm competition (it is not a horizontal merger[79]) but could create a stronger competitor to the current market leaders.

Finally, even if there were a competition-related market failure to be addressed in the field of generative AI (which is anything but clear), the remedies under contemplation may do more harm than good. Some of the solutions that have been put forward have highly ambiguous effects on consumer welfare. Scholars have shown that, e.g., mandated data sharing—a solution championed by EU policymakers, among others—may sometimes dampen competition in generative-AI markets.[80] This is also true of legislation like the General Data Protection Regulation (GDPR), which makes it harder for firms to acquire more data about consumers—assuming such data is, indeed, useful to generative-AI services.[81]

In sum, it is a flawed understanding of the economics and practical consequences of large agglomerations of data that lead competition authorities to believe that data-incumbency advantages are likely to harm competition in generative AI markets—or even in the data-intensive Web 2.0 markets that preceded them. Indeed, competition or regulatory intervention to “correct” data barriers and data network and scale effects is liable to do more harm than good.


[1] CMA Seeks Views on AI Partnerships and Other Arrangements, Competition and Markets Authority (24 Apr. 2024),

[2] AI, of course, is not a market (at least not a relevant antitrust market). Within the realm of what is being called “AI”, companies can offer myriad products and services, and specific relevant markets would need to be defined before assessing harm to competition in specific cases.

[3] OECD, Start-ups, Killer Acquisitions and Merger Control (2020), available at

[4] Kate Rooney & Hayden Field, Amazon Spends $2.75 Billion on AI Startup Anthropic in Its Largest Venture Investment Yet, CNBC (27 Mar. 2024),

[5] Id.

[6] Tom Warren, Microsoft Partners with Mistral in Second AI Deal Beyond OpenAI, The Verge (26 Feb. 2024),

[7] Mark Sullivan, Microsoft’s Inflection AI Grab Likely Cost More Than $1 Billion, Says An Insider (Exclusive), Fast Company  (26 Mar. 2024),; see also, Mustafa Suleyman, DeepMind and Inflection Co-Founder, Joins Microsoft to Lead Copilot, Microsoft Corporate Blogs (19 Mar. 2024),; Krystal Hu & Harshita Mary Varghese, Microsoft Pays Inflection $ 650 Mln in Licensing Deal While Poaching Top Talent, Source Says, Reuters (21 Mar. 2024),; The New Inflection: An Important Change to How We’ll Work, Inflection (Mar. 19, 2024),; Julie Bort, Here’s How Microsoft Is Providing a ‘Good Outcome’ for Inflection AI VCs, as Reid Hoffman Promised, Tech Crunch (21 Mar. 2024),

[8] See Rana Foroohar, The Great US-Europe Antitrust Divide, Financial Times (5 Feb. 2024), (quoting FTC Chair Lina Khan “we are still reeling from the concentration that resulted from Web 2.0, and we don’t want to repeat the mis-steps of the past with AI”).

[9] See, e.g., Geoffrey Manne & Dirk Auer, Antitrust Dystopia and Antitrust Nostalgia: Alarmist Theories of Harm in Digital Markets and Their Origins, 28 Geo. Mason L. Rev. 1279, 1294 (2021). (“But while these increasing returns can cause markets to become more concentrated, they also imply that it is often more efficient to have a single firm serve the entire market. For instance, to a first approximation, network effects, which are one potential source of increasing returns, imply that it is more valuable-not just to the platform, but to the users themselves-for all users to be present on the same network or platform. In other words, fragmentation—de-concentration—may be more of a problem than monopoly in markets that exhibit network effects and increasing returns to scale. Given this, it is far from clear that antitrust authorities should try to prevent consolidation in markets that exhibit such characteristics, nor is it self-evident that these markets somehow produce less consumer surplus than markets that do not exhibit such increasing returns”.)

[10] Javier Espinoza & Leila Abboud, EU’s New AI Act Risks Hampering Innovation, Warns Emmanuel Macron, Financial Times (11 Dec. 2023),

[11] See, e.g., Michael Chui, et al., The Economic Potential of Generative AI: The Next Productivity Frontier, McKinsey (14 Jun. 2023),

[12] See, e.g., Zhuoran Qiao, Weili Nie, Arash Vahdat, Thomas F. Miller III, & Animashree Anandkumar, State-Specific Protein–Ligand Complex Structure Prediction with a Multiscale Deep Generative Model, 6 Nature Machine Intelligence, 195-208 (2024); see also Jaemin Seo, Sang Kyeun Kim, Azarakhsh Jalalvand, Rory Conlin, Andrew Rothstein, Joseph Abbate, Keith Erickson, Josiah Wai, Ricardo Shousha, & Egemen Kolemen, Avoiding Fusion Plasma Tearing Instability with Deep Reinforcement Learning, 626 Nature, 746-751 (2024).

[13] Nathan Newman, Taking on Google’s Monopoly Means Regulating Its Control of User Data, Huffington Post (24 Sep. 2013),

[14] See, e.g., Lina Khan & K. Sabeel Rahman, Restoring Competition in the U.S. Economy, in Untamed: How to Check Corporate, Financial, and Monopoly Power (Nell Abernathy, Mike Konczal, & Kathryn Milani, eds., 2016), at 23 (“From Amazon to Google to Uber, there is a new form of economic power on display, distinct from conventional monopolies and oligopolies…, leverag[ing] data, algorithms, and internet-based technologies… in ways that could operate invisibly and anticompetitively”.); Mark Weinstein, I Changed My Mind—Facebook Is a Monopoly, Wall St. J. (1 Oct. 2021), (“[T]he glue that holds it all together is Facebook’s monopoly over data…. Facebook’s data troves give it unrivaled knowledge about people, governments—and its competitors”.).

[15] See, generally, Abigail Slater, Why “Big Data” Is a Big Deal, The Reg. Rev. (6 Nov. 2023),; Amended Complaint at ¶36, United States v. Google, 1:20-cv-03010- (D.D.C. 2020); Complaint at ¶37, United States v. Google, 1:23-cv-00108 (E.D. Va. 2023), (“Google intentionally exploited its massive trove of user data to further entrench its monopoly across the digital advertising industry”.).

[16] See, e.g., Press Release, Commission Launches Calls for Contributions on Competition in Virtual Worlds and Generative AI, European Commission (9 Jan. 2024),; Krysten Crawford, FTC’s Lina Khan Warns Big Tech over AI, SIEPR (3 Nov. 2020), (“Federal Trade Commission Chair Lina Khan delivered a sharp warning to the technology industry in a speech at Stanford on Thursday: Antitrust enforcers are watching what you do in the race to profit from artificial intelligence”.) (emphasis added).

[17] See, e.g., John M. Newman, Antitrust in Digital Markets, 72 Vand. L. Rev. 1497, 1501 (2019) (“[T]he status quo has frequently failed in this vital area, and it continues to do so with alarming regularity. The laissez-faire approach advocated for by scholars and adopted by courts and enforcers has allowed potentially massive harms to go unchecked”.); Bertin Martins, Are New EU Data Market Regulations Coherent and Efficient?, Bruegel Working Paper 21/23 (2023), (“Technical restrictions on access to and re-use of data may result in failures in data markets and data-driven services markets”.); Valéria Faure-Muntian, Competitive Dysfunction: Why Competition Law Is Failing in a Digital World, The Forum Network (24 Feb. 2021),

[18] See Foroohar, supra note 8.

[19] See, e.g., Press Release, European Commission, supra note 16.

[20] See infra, Section II. Commentators have also made similar claims; see, e.g., Ganesh Sitaram & Tejas N. Narechania, It’s Time for the Government to Regulate AI. Here’s How, Politico (15 Jan. 2024) (“All that cloud computing power is used to train foundation models by having them “learn” from incomprehensibly huge quantities of data. Unsurprisingly, the entities that own these massive computing resources are also the companies that dominate model development. Google has Bard, Meta has LLaMa. Amazon recently invested $4 billion into one of OpenAI’s leading competitors, Anthropic. And Microsoft has a 49 percent ownership stake in OpenAI — giving it extraordinary influence, as the recent board struggles over Sam Altman’s role as CEO showed”.).

[21] Press Release, European Commission, supra note 16.

[22] Comment of U.S. Federal Trade Commission to the U.S. Copyright Office, Artificial Intelligence and Copyright, Docket No. 2023-6 (30 Oct. 2023), at 4, (emphasis added).

[23] See, e.g., Joe Caserta, Holger Harreis, Kayvaun Rowshankish, Nikhil Srinidhi, & Asin Tavakoli, The Data Dividend: Fueling Generative AI, McKinsey Digital (15 Sep. 2023), (“Your data and its underlying foundations are the determining factors to what’s possible with generative AI”.).

[24] See, e.g., Tim Keary, Google DeepMind’s Achievements and Breakthroughs in AI Research, Techopedia (11 Aug. 2023),; See, e.g., Will Douglas Heaven, Google DeepMind Used a Large Language Model to Solve an Unsolved Math Problem, MIT Technology Review (14 Dec. 2023),; see also, A Decade of Advancing the State-of-the-Art in AI Through Open Research, Meta (30 Nov. 2023),; see also, 200 Languages Within a Single AI Model: A Breakthrough in High-Quality Machine Translation, Meta, (last visited 18 Jan. 2023).

[25] See, e.g., Jennifer Allen, 10 Years of Siri: The History of Apple’s Voice Assistant, Tech Radar (4 Oct. 2021),; see also Evan Selleck, How Apple Is Already Using Machine Learning and AI in iOS, Apple Insider (20 Nov. 2023),; see also, Kathleen Walch, The Twenty Year History Of AI At Amazon, Forbes (19 July 2019),

[26] See infra Section III.

[27] See, e.g., Cédric Argenton & Jens Prüfer, Search Engine Competition with Network Externalities, 8 J. Comp. L. & Econ. 73, 74 (2012).

[28] John M. Yun, The Role of Big Data in Antitrust, in The Global Antitrust Institute Report on the Digital Economy (Joshua D. Wright & Douglas H. Ginsburg, eds., 11 Nov. 2020) at 233,; see also, e.g., Robert Wayne Gregory, Ola Henfridsson, Evgeny Kaganer, & Harris Kyriakou, The Role of Artificial Intelligence and Data Network Effects for Creating User Value, 46 Acad. of Mgmt. Rev. 534 (2020), final pre-print version at 4, (“A platform exhibits data network effects if, the more that the platform learns from the data it collects on users, the more valuable the platform becomes to each user”.); see also, Karl Schmedders, José Parra-Moyano, & Michael Wade, Why Data Aggregation Laws Could be the Answer to Big Tech Dominance, Silicon Republic (6 Feb. 2024),

[29] Nathan Newman, Search, Antitrust, and the Economics of the Control of User Data, 31 Yale J. Reg. 401, 409 (2014) (emphasis added); see also id. at 420 & 423 (“While there are a number of network effects that come into play with Google, [“its intimate knowledge of its users contained in its vast databases of user personal data”] is likely the most important one in terms of entrenching the company’s monopoly in search advertising…. Google’s overwhelming control of user data… might make its dominance nearly unchallengeable”.).

[30] See also Yun, supra note 28 at 229 (“[I]nvestments in big data can create competitive distance between a firm and its rivals, including potential entrants, but this distance is the result of a competitive desire to improve one’s product”.).

[31] For a review of the literature on increasing returns to scale in data (this topic is broader than data-network effects) see Geoffrey Manne & Dirk Auer, Antitrust Dystopia and Antitrust Nostalgia: Alarmist Theories of Harm in Digital Markets and Their Origins, 28 Geo Mason L. Rev. 1281, 1344 (2021).

[32] Andrei Hagiu & Julian Wright, Data-Enabled Learning, Network Effects, and Competitive Advantage, 54 RAND J. Econ. 638 (2023).

[33] Id. at 639. The authors conclude that “Data-enabled learning would seem to give incumbent firms a competitive advantage. But how strong is this advantage and how does it differ from that obtained from more traditional mechanisms…”.

[34] Id.

[35] Bruno Jullien & Wilfried Sand-Zantman, The Economics of Platforms: A Theory Guide for Competition Policy, 54 Info. Econ. & Pol’y 10080, 101031 (2021).

[36] Daniele Condorelli & Jorge Padilla, Harnessing Platform Envelopment in the Digital World, 16 J. Comp. L. & Pol’y 143, 167 (2020).

[37] See Hagiu & Wright, supra note 32.

[38] For a summary of these limitations, see generally Catherine Tucker, Network Effects and Market Power: What Have We Learned in the Last Decade?, Antitrust (2018) at 72, available at; see also Manne & Auer, supra note 31, at 1330.

[39] See Jason Furman, Diane Coyle, Amelia Fletcher, Derek McAuley, & Philip Marsden (Dig. Competition Expert Panel), Unlocking Digital Competition (2019) at 32-35 (“Furman Report”), available at

[40] Id. at 34.

[41] Id. at 35. To its credit, it should be noted, the Furman Report does counsel caution before mandating access to data as a remedy to promote competition. See id. at 75. That said, the Furman Report maintains that such a remedy should certainly be on the table because “the evidence suggests that large data holdings are at the heart of the potential for some platform markets to be dominated by single players and for that dominance to be entrenched in a way that lessens the potential for competition for the market”. Id. In fact, the evidence does not show this.

[42] Case COMP/M.9660 — Google/Fitbit, Commission Decision (17 Dec. 2020) (Summary at O.J. (C 194) 7), available at at 455.

[43] Id. at 896.

[44] See Natasha Lomas, EU Checking if Microsoft’s OpenAI Investment Falls Under Merger Rules, TechCrunch (9 Jan. 2024),

[45] Amended Complaint at 11, Meta/Zuckerberg/Within, Fed. Trade Comm’n. (2022) (No. 605837), available at

[46] Amended Complaint (D.D.C), supra note 15 at ¶37.

[47] Amended Complaint (E.D. Va), supra note 15 at ¶8.

[48] Merger Guidelines, US Dep’t of Justice & Fed. Trade Comm’n (2023) at 25, available at

[49] Merger Assessment Guidelines, Competition and Mkts. Auth (2021) at  ¶7.19(e), available at–_.pdf.

[50] Furman Report, supra note 39, at ¶4.

[51] See, e.g., Chris Westfall, New Research Shows ChatGPT Reigns Supreme in AI Tool Sector, Forbes (16 Nov. 2023),

[52] See Krystal Hu, ChatGPT Sets Record for Fastest-Growing User Base, Reuters (2 Feb. 2023),; Google: The AI Race Is On, App Economy Insights (7 Feb. 2023),

[53] See Google Trends,,%2Fg%2F11ts49p01g&hl=en (last visited 12 Jan. 2024) and,%2Fg%2F11ts49p01g&hl=en (last visited 12 Jan. 2024).

[54] See David F. Carr, As ChatGPT Growth Flattened in May, Google Bard Rose 187%, Similarweb Blog (5 Jun. 2023),

[55] See Press Release, Introducing New AI Experiences Across Our Family of Apps and Devices, Meta (27 Sep. 2023),; Sundar Pichai, An Important Next Step on Our AI Journey, Google Keyword Blog (Feb. 6, 2023),

[56] See Ion Prodan, 14 Million Users: Midjourney’s Statistical Success, Yon (Aug. 19, 2023),; see also Andrew Wilson, Midjourney Statistics: Users, Polls, & Growth [Oct 2023], ApproachableAI (13 Oct. 2023),

[57] See Hema Budaraju, New Ways to Get Inspired with Generative AI in Search, Google Keyword Blog (12 Oct. 2023),; Imagine with Meta AI, Meta (last visited Jan. 12, 2024),

[58] Catherine Tucker, Digital Data, Platforms and the Usual [Antitrust] Suspects: Network Effects, Switching Costs, Essential Facility, 54 Rev. Indus. Org. 683, 686 (2019).

[59] Manne & Auer, supra note 31, at 1345.

[60] See, e.g., Stefanie Koperniak, Artificial Data Give the Same Results as Real Data—Without Compromising Privacy, MIT News (3 Mar. 2017), (“[Authors] describe a machine learning system that automatically creates synthetic data—with the goal of enabling data science efforts that, due to a lack of access to real data, may have otherwise not left the ground. While the use of authentic data can cause significant privacy concerns, this synthetic data is completely different from that produced by real users—but can still be used to develop and test data science algorithms and models”.).

[61] See, e.g., Rachel Gordon, Synthetic Imagery Sets New Bar in AI Training Efficiency, MIT News (20 Nov. 2023), (“By using synthetic images to train machine learning models, a team of scientists recently surpassed results obtained from traditional ‘real-image’ training methods.).

[62] Thibault Schrepel & Alex ‘Sandy’ Pentland, Competition Between AI Foundation Models: Dynamics and Policy Recommendations, MIT Connection Science Working Paper (Jun. 2023), at 8.

[63] Igor Susmelj, Optimizing Generative AI: The Role of Data Curation, Lightly (last visited 15 Jan. 2024),

[64] See, e.g., Xiaoliang Dai, et al., Emu: Enhancing Image Generation Models Using Photogenic Needles in a Haystack, ArXiv (27 Sep. 2023) at 1, (“[S]upervised fine-tuning with a set of surprisingly small but extremely visually appealing images can significantly improve the generation quality”.); see also, Hu Xu, et al., Demystifying CLIP Data, ArXiv (28 Sep. 2023),

[65] Lauren Leffer, New Training Method Helps AI Generalize like People Do, Sci. Am. (26 Oct. 2023), (discussing Brendan M. Lake & Marco Baroni, Human-Like Systematic Generalization Through a Meta-Learning Neural Network, 623 Nature 115 (2023)).

[66] Timothy B. Lee, The Real Research Behind the Wild Rumors about OpenAI’s Q* Project, Ars Technica (Dec. 8, 2023),

[67] Id.; see also GSM8K, Papers with Code (last visited 18 Jan. 2023), available at; MATH Dataset, GitHub (last visited 18 Jan. 2024), available at

[68] Lee, supra note 66.

[69] Geoffrey Manne & Ben Sperry, Debunking the Myth of a Data Barrier to Entry for Online Services, Truth on the Market (26 Mar. 2015), (citing Andres V. Lerner, The Role of ‘Big Data’ in Online Platform Competition (26 Aug. 2014),

[70] See Catherine Tucker, Digital Data as an Essential Facility: Control, CPI Antitrust Chron. (Feb. 2020), at 11 (“[U]ltimately the value of data is not the raw manifestation of the data itself, but the ability of a firm to use this data as an input to insight”.).

[71] Or, as John Yun put it, data is only a small component of digital firms’ production function. See Yun, supra note 28, at 235 (“Second, while no one would seriously dispute that having more data is better than having less, the idea of a data-driven network effect is focused too narrowly on a single factor improving quality. As mentioned in supra Section I.A, there are a variety of factors that enter a firm’s production function to improve quality”.).

[72] Luxia Le, The Real Reason Windows Phone Failed Spectacularly, History–Computer (8 Aug. 2023),

[73] Introducing the GPT Store, Open AI (Jan. 10, 2024),

[74] See Michael Schade, How ChatGPT and Our Language Models are Developed, OpenAI,; Sreejani Bhattacharyya, Interesting Innovations from OpenAI in 2021, AIM (1 Jan. 2022),; Danny Hernadez & Tom B. Brown, Measuring the Algorithmic Efficiency of Neural Networks, ArXiv (8 May 2020),

[75] See Yun, supra note 28 at 235 (“Even if data is primarily responsible for a platform’s quality improvements, these improvements do not simply materialize with the presence of more data—which differentiates the idea of data-driven network effects from direct network effects. A firm needs to intentionally transform raw, collected data into something that provides analytical insights. This transformation involves costs including those associated with data storage, organization, and analytics, which moves the idea of collecting more data away from a strict network effect to more of a ‘data opportunity.’”).

[76] Lerner, supra note 69, at 4-5 (emphasis added).

[77] See Clayton M. Christensen, The Innovator’s Dilemma: When New Technologies Cause Great Firms to Fail (2013).

[78] See David J. Teece, Dynamic Capabilities and Strategic Management: Organizing for Innovation and Growth (2009).

[79] Antitrust merger enforcement has long assumed that horizontal mergers are more likely to cause problems for consumers than the latter. See: Geoffrey A. Manne, Dirk Auer, Brian Albrecht, Eric Fruits, Daniel J. Gilman, & Lazar Radic, Comments of the International Center for Law and Economics on the FTC & DOJ Draft Merger Guidelines, (18 Sep. 2023),

[80] See Hagiu & Wright, supra note 32, at 32 (“We use our dynamic framework to explore how data sharing works: we find that it in-creases consumer surplus when one firm is sufficiently far ahead of the other by making the laggard more competitive, but it decreases consumer surplus when the firms are sufficiently evenly matched by making firms compete less aggressively, which in our model means subsidizing consumers less”.); see also Lerner, supra note 69.

[81] See, e.g., Hagiu & Wright, id. (“We also use our model to highlight an unintended consequence of privacy policies. If such policies reduce the rate at which firms can extract useful data from consumers, they will tend to increase the incumbent’s competitive advantage, reflecting that the entrant has more scope for new learning and so is affected more by such a policy”.); Jian Jia, Ginger Zhe Jin, & Liad Wagman, The Short-Run Effects of the General Data Protection Regulation on Technology Venture Investment, 40 Marketing Sci. 593 (2021) (finding GDPR reduced investment in new and emerging technology firms, particularly in data-related ventures); James Campbell, Avi Goldfarb, & Catherine Tucker, Privacy Regulation and Market Structure, 24 J. Econ. & Mgmt. Strat. 47 (2015) (“Consequently, rather than increasing competition, the nature of transaction costs implied by privacy regulation suggests that privacy regulation may be anti-competitive”.).

Continue reading
Antitrust & Consumer Protection

Against the ‘Europeanization’ of California’s Antitrust Law

Regulatory Comments We are grateful for the opportunity to respond to the California Law Revision Commission’s Study of Antitrust Law with these comments on the Single-Firm Conduct . . .

We are grateful for the opportunity to respond to the California Law Revision Commission’s Study of Antitrust Law with these comments on the Single-Firm Conduct Working Group’s report (the “Expert Report”).[1]

The International Center for Law & Economics (ICLE) is a nonprofit, nonpartisan global research and policy center based in Portland, Oregon. ICLE was founded with the goal of building the intellectual foundations for sensible, economically grounded policy. ICLE promotes the use of law & economics methodologies to inform public-policy debates, and has longstanding expertise in the evaluation of competition law and policy. ICLE’s interest is to ensure that competition law remains grounded in clear rules, established precedents, a record of evidence, and sound economic analysis.[2]

I. Introduction

The urge to treat antitrust as a legal Swiss Army knife—capable of correcting all manner of economic and social ills—is difficult to resist. Conflating size with market power, and market power with political power, recent calls for regulation of large businesses are often framed in antitrust terms, although they rarely are rooted in cognizable legal claims or sound economic analysis.

But precisely because antitrust is such a powerful regulatory tool, we should be cautious about its scope, process, and economics, as well as its politicization. For the last 50 or so years, U.S. law has maintained a position of relative restraint in the face of novel, ambiguous conduct, while many other jurisdictions (particularly the European Union) have tended to read uncertainty as the outward expression of a lurking threat. This has led to a sharp policy divergence in the area of competition policy, with the EU passing the Digital Markets Act,[3] while the United States has, to date, continued to rely on tried-and-tested principles crafted by courts over years on a case-by-case basis.

Despite—or perhaps because of—this divergence, many advocates of more aggressive antitrust intervention assert that the United States or individual states should emulate the EU’s approach. This disposition underpins much of the California Law Review Commission’s Report on Single Firm Conduct.[4] Despite some reassuring conclusions—such as the recognition that “protecting competing businesses, even at the expense of consumers and workers” would not “provide a good model for California”[5]—the policies that the report proposes would significantly broaden California antitrust law, bringing it much closer to the European model of competition enforcement than the U.S. one.

Unfortunately, this European-inspired approach to competition policy is unlikely to serve the interests of California consumers. As explained below, the European model of competition enforcement has at least three features that tend to chill efficient business conduct, with few competitive benefits in return (relative to the U.S. approach).

A. ‘Precautionary Principle’ vs Error-Cost Framework

Differentiating pro- from anticompetitive conduct has always been the central challenge of antitrust. When the very same conduct can either benefit or harm consumers, depending on complex and often unknowable circumstances, the potential cost of overenforcement is at least as substantial as the cost of underenforcement.

The U.S. Supreme Court has repeatedly recognized that the cost of “false positive” errors might be greater than those attributable to “false negatives” because, in the words of Judge Frank Easterbrook, “the economic system corrects monopoly more readily than it corrects judicial errors.”[6] The EU’s “precautionary principle” approach is the antithesis of this. It is rooted in a belief that markets are generally unlikely to function well, and certainly are not better at mitigating harm than technocratic regulatory intervention.

The key question is whether, given the limits of knowledge and the errors that such limits may engender, consumers are better off with a more discretionary regime or one in which enforcement is limited to causes of action that policymakers are fairly certain will serve consumer interests. This is a question about changes at the margin, but it is far from marginal in its significance. As we explain below, the U.S. approach to antitrust law performs better in this respect. Departing from it would not benefit California consumers.

B. Presumptions vs Effects-Based Analysis

EU antitrust rests heavily on presumptions of harm, while U.S. courts require plaintiffs to demonstrate that the conduct at-issue actually has anticompetitive effects.

Crucially, the U.S. approach is more consistent with learnings from modern economics, which almost universally counsel against presuming competitive harm on the basis of industry structure and, in particular, in favor of presuming benefit from vertical conduct. Indeed, the EU approach often disregards these findings and presumes the contrary. As evidenced by its recent Intel decision, even the EU’s highest court has finally recognized the paucity of the European Commission’s analysis in this area. But because judicial review of antitrust decisions in the EU is so attenuated, it is not clear if the high court’s admonition will actually affect the Commission’s approach in any substantial way.

California policymakers would be wrong to emulate the European model by introducing more presumptions to California antitrust law.

C. Extraction of Rents vs Extension of Monopoly

U.S. monopolization law prohibits only predatory or exclusionary conduct that results in harm to consumers. The EU, by contrast, also regularly punishes the mere possession of monopoly power, even where lawfully obtained. Indeed, the EU goes so far as to target companies that may lack monopoly power, but merely possess an innovative and successful business model. For example, in actions involving companies ranging from soda manufacturers to digital platforms, the EU repeatedly has required essential-facilities-style access to companies’ private property for less-successful rivals.

As we explain below, the Expert Report essentially calls on California lawmakers to replicate the European model by seeking to protect even those competitors that are less efficient, thus challenging the very existence of legitimately earned monopolies. Unfortunately, this approach would diminish the incentives to create successful businesses in the first place. Such an outcome would be particularly unfortunate for California, which is host to arguably the most vibrant startup ecosystem in the world.

D. The Danger of the European Approach

In endorsing the European approach to antitrust in order to justify high-profile cases against large firms, California would effectively be prioritizing political expediency over the rule of law and consumer well-being.

The risk of an EU-like approach in California is that it would thwart technological progress and enshrine mediocrity. This is particularly true in the digital economy, where innovative practices with positive welfare effects—such as building efficient networks or improving products and services as technologies and consumer preferences evolve—are often the subject of demagoguery, especially from inefficient firms looking for a regulatory leg up.

While advocates for a more European approach to antitrust assert that their proposals would improve economic conditions in California (and the United States, more generally), economic logic and the available evidence suggest otherwise, especially in technology markets.

Once antitrust is expanded beyond its economic constraints, it ceases to be a uniquely valuable tool to address real economic harms to consumers, and becomes instead a tool for evading legislative and judicial constraints. This is hardly the promotion of democratic ideals that proponents of a more EU-like regime claim to desire.

In the following sections, we expand upon these distinctions between EU and U.S. law and explain how elements of the Expert Report’s analysis and proposed statutory language would shift California’s antitrust law toward the EU model in problematic ways. We urge the California Law Revision Commission to consider not just whether emulating the EU approach would permit the state to reach a preconceived outcome—i.e., placing large firms under increased antitrust scrutiny—but whether doing so would ultimately benefit California and its consumers.

II. The EU ‘Precautionary Principle’ Approach vs the US Error-Cost Framework

The U.S. Supreme Court has repeatedly recognized the limitations that courts face in distinguishing between pro- and anticompetitive conduct in antitrust cases, and particularly the risk this creates of reaching costly false-positive (Type I) decisions in monopolization cases.[7] As the Court has noted with respect to the expansion of liability for single-firm conduct, in particular:

Against the slight benefits of antitrust intervention here, we must weigh a realistic assessment of its costs…. Mistaken inferences and the resulting false condemnations “are especially costly because they chill the very conduct the antitrust laws are designed to protect.” The cost of false positives counsels against an undue expansion of § 2 liability.[8]

The Court has also expressed the view—originally laid out in Judge Frank Easterbrook’s seminal article “The Limits of Antitrust”—that the costs to consumers arising from Type I errors are likely greater than those attributable to Type II errors, because “the economic system corrects monopoly more readily than it corrects judicial errors.”[9]

The EU’s more “precautionary” approach to antitrust policy is the antithesis of this.[10] It is rooted in a belief that markets do not—or, more charitably, are unlikely to—function well in general, and certainly not sufficiently to self-correct in the face of monopolization.

While the precautionary principle may generally prevent certain fat-tailed negative events,[11] these potential benefits come, almost by definition, at the expense of short-term growth.[12] Adopting a precautionary approach is thus a costly policy stance in those circumstances where it is not clearly warranted by underlying risk and uncertainty. This is an essential issue for a state like California, whose economy is so reliant on the continued growth and innovation of its vibrant startup ecosystem.

While it is impossible to connect broad macroeconomic trends conclusively to specific policy decisions, it does seem clear that Europe’s overarching precautionary approach to economic regulation has not served it well.[13] In that environment, the EU’s economic performance has fallen significantly behind that of the United States.[14] “[I]n 2010 US GDP per capita was 47 percent larger than the EU while in 2021 this gap increased to 82 percent. If the current trend of GDP per capita carries forward, in 2035, the average GDP per capita in the US will be $96,000 while the average EU GDP per capita will be $60,000.”[15]

Of course, no one believes that markets are perfect, or that antitrust enforcement can never be appropriate. The question is the marginal, comparative one: Given the realities of politics, economics, the limits of knowledge, and the errors to which they can lead, which imperfect response is preferable at the margin? Or, phrased slightly differently, should we give California antitrust enforcers and private plaintiffs more room to operate, or should we continue to cabin their operation in careful, economically grounded ways, aimed squarely at optimizing—not minimizing—the extent of antitrust enforcement?

This may be a question about changes at the margin, but it is far from marginal. It goes to the heart of the market’s role in the modern economy.

While there are many views on this subject, arguments that markets have failed us in ways that more antitrust would correct are poorly supported.[16] We should certainly continue to look for conditions where market failures of one kind or another may justify intervention, but we should not make policy on the basis of mere speculation. And we should certainly not do so without considering the likelihood and costs of regulatory failure, as well. In order to reliably adopt a sound antitrust policy that might improve upon the status quo (which has evolved over a century of judicial decisions, generally alongside the field’s copious advances in economic understanding), we need much better information about the functioning of markets and the consequences of regulatory changes than is currently available.

To achieve this, antitrust law and enforcement policy should, above all, continue to adhere to the error-cost framework, which informs antitrust decision making by considering the relative costs of mistaken intervention compared with mistaken nonintervention.[17] Specific cases should be addressed as they come, with an implicit understanding that, especially in digital markets, precious few generalizable presumptions can be inferred from the previous case. The overall stance should be one of restraint, reflecting the state of our knowledge.[18] We may well be able to identify anticompetitive harms in certain cases, and when we do, we should enforce the current laws. But we should not overestimate our ability to finetune market outcomes without causing more harm than benefit.

Allegations that the modern antitrust regime is insufficient take as a given that there is something wrong with antitrust doctrine or its enforcement, and cast about for policy “corrections.” The common flaw with these arguments is that they are not grounded in robust empirical or theoretical support. Indeed, as one of the influential papers that (ironically) is sometimes cited to support claims for more antitrust puts it:

An alternative perspective on the rise of [large firms and increased concentration] is that they reflect a diminution of competition, due to weaker U.S. antitrust enforcement. Our findings on the similarity of trends in the United States and Europe, where antitrust authorities have acted more aggressively on large firms, combined with the fact that the concentrating sectors appear to be growing more productive and innovative, suggests that this is unlikely to be the primary explanation, although it may be important in some industries.[19]

Rather, such claims are little more than hunches that something must be wrong, conscripted to serve a presumptively interventionist agenda. Because they are merely hypotheses about things that could go wrong, they do not determine—and rarely even ask—if heightened antitrust scrutiny and increased antitrust enforcement are actually called for in the first place. The evidence strongly contradicts the basis for these hunches.

Critics of U.S. competition policy sometimes contend that markets have become more concentrated and thus less competitive.[20] But there are good reasons to be skeptical of the national-concentration and market-power data.[21] Even more importantly, the narrative that purports to find a causal relationship between these data and reduced competition is almost certainly incorrect.

Competition rarely takes place in national markets; it takes place in local markets. Recent empirical work demonstrates that national measures of concentration do not reflect market structures at the local level.[22] To the extent that national-level firm concentration may be growing, these trends are actually driving increased competition and decreased concentration at the local level, which is typically what matters for consumers:

Put another way, large firms have materially contributed to the observed decline in local concentration. Among industries with diverging trends, large firms have become bigger but the associated geographic expansion of these firms, through the opening of more plants in new local markets, has lowered local concentration thus suggesting increased local competition.[23]

The rise in national concentration is predominantly a function of more efficient firms competing in more—and more localized—markets. Thus, rising national concentration, where it is observed, is a result of increased productivity and competition that weed out less-efficient producers. Indeed, as one influential paper notes:

[C]oncentration increases do not correlate to price hikes and correspond to increased output. This implies that oligopolies are related to an offsetting and positive force—these oligopolies are likely due to technical innovation or scale economies. My data suggest that increases in market concentration are strongly correlated with innovations in productivity.[24]

Another important paper finds that this dynamic is driven by top firms bringing productivity increases to smaller markets, to the substantial (and previously unmeasured) benefit of consumers:

US firms in service industries increasingly operate in more local markets. Employment, sales, and spending on fixed costs have increased rapidly in these industries. These changes have favored top firms, leading to increasing national concentration. Top firms in service industries have grown by expanding into new local markets, predominantly small and mid-sized US cities. Market concentration at the local level has decreased in all US cities, particularly in cities that were initially small. These facts are consistent with the availability of new fixed-cost-intensive technologies that yield lower marginal costs in service sectors. The entry of top service firms into new local markets has led to substantial unmeasured productivity growth, particularly in small markets.[25]

Similar results hold for labor-market effects. According to one recent study, while the labor-market power of firms appears to have increased:

labor market power has not contributed to the declining labor share. Despite the backdrop of stable national concentration, we… find that [local labor-market concentration] has declined over the last 35 years. Most local labor markets are more competitive than they were in the 1970s.[26]

In short, it is inappropriate to draw conclusions about the strength of competition and the efficacy of antitrust laws from national-concentration measures. This is a view shared by many economists from across the political spectrum. Indeed, one of the Expert Report’s authors, Carl Shapiro, has raised these concerns regarding the national-concentration data:

[S]imply as a matter of measurement, the Economic Census data that are being used to measure trends in concentration do not allow one to measure concentration in relevant antitrust markets, i.e., for the products and locations over which competition actually occurs. As a result, it is far from clear that the reported changes in concentration over time are informative regarding changes in competition over time.[27]

It appears that overall competition is increasing, not decreasing, whether it is accompanied by an increase in national concentration or not.

A. The Expert Report’s Treatment of Error Costs

Implicitly shunning the evidence that demonstrates markets have become more, not less, competitive, the Expert Report proposes that California adopt a firm stance in favor of false positives over false negatives—in other words, that it tolerate erroneously condemning procompetitive behavior in exchange for avoiding the risk of erroneously accepting anticompetitive conduct:

Whereas the policy of California is that the public is best served by competition and the goal of the California antitrust laws is to promote and protect competition throughout the State, in interpreting this Section courts should bear in mind that the policy of California is that the risk of under-enforcement of the antitrust laws is greater than the risk of over-enforcement.[28]

Of course, it is possible that, in some markets, there are harms being missed and for which enforcers should be better equipped. But advocates of reform have yet to adequately explain much of what we need to know to make such a determination, let alone craft the right approach to it if we did. Antitrust law should be refined based on an empirical demonstration of harms, as well as a careful weighing of those harms against the losses to social welfare that would arise if procompetitive conduct were deterred alongside anticompetitive conduct.

Dramatic new statutes to undo decades of antitrust jurisprudence or reallocate burdens of proof with the stroke of a pen are unjustified. Suggesting, as the Expert Report does, that antitrust law should simply “err on the side of enforcement when the effect of the conduct at issue on competition is uncertain”[29] is an unsupported statement of a political preference, not one rooted in sound economics or evidence.

The primary evidence adduced to support the claim that underenforcement (and thus, the risk of Type II errors) is more significant than overenforcement (and thus, the risk of Type I errors) is that there are not enough cases brought and won. But even if superficially true, this is, on its own, just as consistent with a belief that the regime is functioning well as it is with a belief that it is functioning poorly. Indeed, as one of the Expert Report’s authors has pointed out:

Antitrust law [] has a widespread effect on business conduct throughout the economy. Its principal value is found, not in the big litigated cases, but in the multitude of anticompetitive actions that do not occur because they are deterred by the antitrust laws, and in the multitude of efficiency-enhancing actions that are not deterred by an overbroad or ambiguous antitrust.[30]

At the same time, some critics (including another of the Expert Report’s authors) contend that a heightened concern for Type I errors stems from a faulty concern that “type two errors… are not really problematic because the market itself will correct the situation,” instead asserting that “it is economically naïve to assume that markets will naturally tend toward competition.”[31]

Judge Easterbrook’s famous argument for enforcement restraint is not based on the assertion that markets are perfectly self-correcting. Rather, his claim is that the (undeniable) incentive of new entrants to compete for excess profits in monopolized markets operates to limit the social costs of Type II errors more effectively than the legal system’s ability to correct or ameliorate the costs of Type I errors. The logic is quite simple, and not dependent on the strawman notion that markets are perfect:

If the court errs by condemning a beneficial practice, the benefits may be lost for good. Any other firm that uses the condemned practice faces sanctions in the name of stare decisis, no matter the benefits. If the court errs by permitting a deleterious practice, though, the welfare loss decreases over time. Monopoly is self-destructive. Monopoly prices eventually attract entry. True, this long run may be a long time coming, with loss to society in the interim. The central purpose of antitrust is to speed up the arrival of the long run. But this should not obscure the point: judicial errors that tolerate baleful practices are self-correcting while erroneous condemnations are not.[32]

Moreover, anticompetitive conduct that is erroneously excused may be subsequently corrected, either by another enforcer, a private litigant, or another jurisdiction. Ongoing anticompetitive behavior will tend to arouse someone’s ire: competitors, potential competitors, customers, input suppliers. That means such behavior will be noticed and potentially brought to the attention of enforcers. And for the same reason—identifiable harm—it may also be actionable.

By contrast, procompetitive conduct that does not occur because it is prohibited or deterred by legal action has no constituency and no visible evidence on which to base a case for revision. Nor does a firm improperly deterred from procompetitive conduct have any standing to sue the government for erroneous antitrust enforcement, or the courts for adopting an improper standard. Of course, overenforcement can sometimes be corrected, but the institutional impediments to doing so are formidable.

The claim that concern for Type I errors is overblown further rests on the assertion that “more up-to-date economic analysis” has undermined that position.[33] But that learning is, for the most part, entirely theoretical—constrained to “possibility theorems” divorced from realistic complications and the real institutional settings of decision making. Indeed, the proliferation of these theories may actually increase, rather than decrease, uncertainty by further complicating the analysis and asking generalist judges to choose from among competing theories, without any realistic means to do so.[34]

Unsurprisingly, “[f]or over thirty years, the economics profession has produced numerous models of rational predation. Despite these models and some case evidence consistent with episodes of predation, little of this Post-Chicago School learning has been incorporated into antitrust law.”[35] Nor is it likely that the courts are making an erroneous calculation in the abstract. Evidence of Type I errors is hard to come by, but for a wide swath of conduct called into question by “Post-Chicago School” and other theories, the evidence of systematic problems is virtually nonexistent.[36]

Moreover, contrary to the Expert Report’s implications,[37] U.S. antitrust law has not ignored potentially anticompetitive harm, and courts are hardly blindly deferential to conduct undertaken by large firms. It is impossible to infer from the general “state of the world” or from perceived “wrong” judicial decisions that the current antitrust regime has failed or that California, in particular, would benefit from a wholesale shifting of its antitrust error-cost presumptions.[38]

III. The Reliance on Presumptions vs the Demonstration of Anticompetitive Effects

While U.S. antitrust law generally requires a full-blown, effects-based analysis of challenged behavior—particularly in the context of unilateral conduct (monopolization or abuse of dominance) and vertical restraints—the EU continues to rely heavily on presumptions of harm or extremely truncated analysis. Even the EU’s highest court has finally recognized the paucity of the European Commission’s analysis in this area in its recent Intel decision.[39]

The degree to which the United States and EU differ with respect to their reliance on presumptions in antitrust cases is emblematic of a broader tendency of the U.S. regime to adhere to economic principles, while the EU tends to hold such principles in relative disregard. The U.S. approach is consistent with learnings from modern economics, which almost universally counsel against presuming competitive harm on the basis of industry structure—particularly from the extent of concentration in a market. Indeed, as one of the Expert Report’s own authors has argued, “there is no well-defined ‘causal effect of concentration on price,’ but rather a set of hypotheses that can explain observed correlations of the joint outcomes of price, measured markups, market share, and concentration.”[40]

Concerns about excessive concentration are at the forefront of current efforts to expand antitrust enforcement, including through the use of presumptions. There is no reliable empirical support for claims either that concentration has been increasing, or that it necessarily leads to, or has led to, increased market power and the economic harms associated with it.[41] There is even less support for claims that concentration leads to the range of social ills ascribed to it by advocates of “populist” antitrust. Similarly, there is little evidence that the application of antitrust or related regulation to more vigorously prohibit, shrink, or break up large companies will correct these asserted problems.

Meanwhile, economic theory, empirical evidence, and experience all teach that vertical restraints—several of which would be treated more harshly under the Expert Report’s recommendations[42]—rarely harm competition. Indeed, they often benefit consumers by reducing costs, better distributing risk, better informing and optimizing R&D activities and innovation, better aligning manufacturer and distributor incentives, lowering price, increasing demand through the inducement of more promotional services, and/or creating more efficient distribution channels.

As the former Federal Trade Commission (FTC) Bureau of Economics Director Francine Lafontaine explained in summarizing the body of economic evidence analyzing vertical restraints: “it appears that when manufacturers choose to impose [vertical] restraints, not only do they make themselves better off but they also typically allow consumers to benefit from higher quality products and better service provision.”[43] A host of other studies corroborate this assessment.[44] As one of these notes, while “some studies find evidence consistent with both pro- and anticompetitive effects… virtually no studies can claim to have identified instances where vertical practices were likely to have harmed competition.”[45] Similarly, “in most of the empirical studies reviewed, vertical practices are found to have significant pro-competitive effects.”[46]

At the very least, we remain profoundly uncertain of the effects of vertical conduct (particularly in the context of modern high-tech and platform industries), with the proviso that most of what we know suggests that this conduct is good for consumers. But even that worst-case version of our state of knowledge is inconsistent with the presumptions-based approach taken by the EU.

Adopting a presumptions-based approach without a firm economic basis is far more hostile to novel business conduct, especially in the innovative markets that distinguish California’s economy. EU competition policy errs on the side of condemning novel conduct, deterring beneficial business activities where consumers would be better served if authorities instead tried to better understand them. This is not something California should emulate.

A. The Expert Report’s Quantification of Anticompetitive Harm and Causation

European competition law imposes a much less strenuous burden on authorities to quantify anticompetitive harm and establish causation than does U.S. law. This makes European competition law much more prone to false positives that condemn efficiency-generating or innovative firm behavior. The main cause of these false positives is the failure of the EU’s “competitive process” standard to separate competitive from anticompetitive exclusionary conduct.

While the Expert Report rightly recognizes that adopting an abuse-of-dominance standard (similar to that which exists in Europe) would be misguided, its proposed focus on “competitive constraints,” rather than consumer welfare, would effectively bring California antitrust enforcement much closer to the EU model.[47]

At the same time, the Expert Report counsels adopting a “material-risk-of-harm” standard, which is foreign to U.S. antitrust law:

(e) Anticompetitive exclusionary conduct includes conduct that has or had a material risk of harming trading partners due to increased market power, even if those harms have not yet arisen and may not materialize.[48]

While such a standard exists in U.S. standing jurisprudence,[49] antitrust plaintiffs (and private plaintiffs, in particular) must typically meet a higher bar to prove actual antitrust injury.[50] Moreover, the focus is generally on output restriction, rather than the risk of “harm” to a trading partner:

The government must show conduct that reasonably seems capable of causing reduced output and increased prices by excluding a rival. The private plaintiff must additionally show an actual effect producing an injury in order to support a damages action or individually threatened harm to support an injunction. The required private effect could be either a higher price which it paid, or lost profits from market exclusion.[51]

Again, this is a fairly concrete application of the error-cost framework: Lowering the standard of proof required to establish liability increases the risk of false positives and decreases the risk of false negatives. But particularly in California—where so much of the state’s economic success is built on industries characterized by large companies with substantial procompetitive economies of scale and network effects, novel business models, and immense technological innovation—the risk of erroneous condemnation is substantial, and the potential costs significant.

Further, defining antitrust harm in terms of “conduct [that] tends to… diminish or create a meaningful risk of diminishing the competitive constraints imposed by the defendant’s rivals”[52] opens the door substantially to the risk that procompetitive conduct could be enjoined. For example, such an approach would seem at odds with the concept of antitrust injury for private plaintiffs established by the Supreme Court’s Brunswick case.[53] “Competitive constraints” may “tend” to be reduced, as in Brunswick, by perfectly procompetitive conduct; enshrining such a standard would not serve California’s economic interests.

Similarly, the Expert Report’s proposed statutory language includes a provision that would infer not only causation but also the existence of harm from ambiguous conduct:

5) In cases where the trading partners are customers…, it is not necessary for the plaintiff to specify the precise nature of the harm that might be experienced in the future or to quantify with specificity any particular past harm. It is sufficient for the plaintiff to establish a significant weakening of the competitive constraints facing the defendant, from which such harms to direct or indirect customers can be presumed.[54]

The Microsoft case similarly held that plaintiffs need not quantify injury with specificity because “neither plaintiffs nor the court can confidently reconstruct a product’s hypothetical technological development in a world absent the defendant’s exclusionary conduct.”[55] But Microsoft permits the inference only of causation in such circumstances, not the existence of anticompetitive conduct. Most of the decision was directed toward identifying and assessing the anticompetitiveness of the alleged conduct. Inference is permitted only with respect to causation—to the determination that such conduct was reasonably likely to lead to harm by excluding specific (potential) competitors. Establishing merely a “weakening of the competitive constraints facing the defendant,” by contrast, does not permit an inference of anticompetitiveness.

Such an approach is much closer to the European standard of maintaining a system of “undistorted competition.” European authorities generally operate under the assumption that “competitive” market structures ultimately lead to better outcomes for consumers.[56] This contrasts with American antitrust enforcement which, by pursuing a strict consumer-welfare goal, systematically looks at the actual impact of a practice on economic parameters, such as prices and output.

In other words, European competition enforcement assumes that concentrated market structures likely lead to poor outcomes and thus sanctions them, whereas U.S. antitrust law looks systematically into the actual effects of a practice. The main consequence of this distinction is that, compared to the United states, European competition law has established a wider set of per se prohibitions (which are not discussed in the Expert Report) and sets a lower bar for plaintiffs to establish the existence of anticompetitive conduct (which the Expert Report recommends California policymakers emulate).[57] Because of this lower evidentiary threshold, EU competition decisions are also subject to less-stringent judicial review.

The EU’s competitive-process standard is similar to the structuralist analysis that was popular in the United States through the middle of the 20th century. This view of antitrust led U.S. enforcers frequently to condemn firms merely for growing larger than some arbitrary threshold, even when those firms engaged in conduct that, on net, benefited consumers. While EU enforcers often claim to be pursuing a consumer-welfare standard, and to adhere to rigorous economic analysis in their antitrust cases,[58] much of their actual practice tends to engage in little more than a window-dressed version of the outmoded structuralist analysis that U.S. scholars, courts, and enforcers roundly rejected in the latter half of the 20th century.

To take one important example, a fairly uncontroversial requirement for antitrust intervention is that a condemned practice should actually—or be substantially likely to—foster anticompetitive harm. Even in Europe, whatever other goals competition law is presumed to further, it is nominally aimed at protecting competition rather than competitors.[59] Accordingly, the mere exit of competitors from the market should be insufficient to support liability under European competition law in the absence of certain accompanying factors.[60] And yet, by pursuing a competitive-process goal, European competition authorities regularly conflate desirable and undesirable forms of exclusion precisely on the basis of their effect on competitors.

As a result, the Commission routinely sanctions exclusion that stems from an incumbent’s superior efficiency rather than from welfare-reducing strategic behavior,[61] and routinely protects inefficient competitors that would otherwise rightly be excluded from a market. As Pablo Ibanez Colomo puts it:

It is arguably more convincing to question whether the principle whereby dominant firms are under a general duty not to discriminate is in line with the logic and purpose of competition rules. The corollary to the idea that it is prima facie abusive to place rivals at a disadvantage is that competition must take place, as a rule, on a level playing field. It cannot be disputed that remedial action under EU competition law will in some instances lead to such an outcome.[62]

Unfortunately, the Expert Report’s repeated focus on diminished “competitive constraints” as the touchstone for harm may (perhaps unintentionally) even enable courts to impose liability for harm to competitors caused by procompetitive conduct. For example, the Expert Report would permit a determination that:

[C]onduct tends to… diminish or create a meaningful risk of diminishing the competitive constraints… [if it] tends to (i) increase barriers to entry or expansion by those rivals, (ii) cause rivals to lower their quality-adjusted output or raise their quality-adjusted price, or (iii) reduce rivals’ incentives to compete against the defendant.[63]

But market exit is surely an example of a reduced incentive to compete, even if it results from a rival’s intense (and consumer-welfare-enhancing) competition. Depending on how “barrier to entry” is defined, innovation, product improvement, and vertical integration by a defendant—even when they are procompetitive—all could constitute a barrier to entry by forcing rivals to incur greater costs or compete in multiple markets. Similarly, increased productivity resulting in less demand for labor or other inputs or lower wages could enable a “defendant [to] profitably make a less attractive offer to that supplier or worker… than the defendant could absent that conduct,”[64] even though the increase in market power in that case would be beneficial.[65]

It is true that the Expert Report elsewhere notes that “it is sometimes difficult for courts to distinguish between anticompetitive exclusionary conduct, which is illegal, from competition on the merits, which is legal even if it weakens rivals or drives them out of business altogether.”[66] Thus, it is perhaps unintentional that the report’s proposed language could nevertheless support liability in such circumstances. At the very least, California should not adopt the Expert Report’s proposed language without a clear disclaimer that liability will never be based on “diminished competitive constraints” resulting from consumer-welfare-enhancing conduct or vigorous competition by the defendant.

IV. Penalizing the Existence of Monopolies vs Prohibiting Only the Extension of Monopoly Power

While U.S. monopolization law prohibits only predatory or exclusionary conduct that results in both the unlawful acquisition or maintenance of monopoly power and the creation of net harm to consumers, the EU also punishes the mere exercise of monopoly power—that is, the charging of allegedly “excessive” prices by dominant firms (or the use of “exploitative” business terms). Thus, the EU is willing to punish the mere extraction of rents by a lawfully obtained dominant firm, while the United States punishes only the unlawful extension of market power.

There may be multiple reasons for this difference, including the EU’s particular history with state-sponsored monopolies and its unique efforts to integrate its internal market. Whatever the reason, the U.S. approach, unlike the EU’s, is grounded in a concern for minimizing error costs—not in order to protect monopolists or large companies, but to protect the consumers who benefit from more dynamic markets, more investment, and more innovation:

The mere possession of monopoly power, and the concomitant charging of monopoly prices, is not only not unlawful; it is an important element of the free-market system. The opportunity to charge monopoly prices—at least for a short period is what attracts “business acumen” in the first place; it induces risk taking that produces innovation and economic growth. To safeguard the incentive to innovate, the possession of monopoly power will not be found unlawful unless it is accompanied by an element of anticompetitive conduct.[67]

At the same time, the U.S. approach mitigates the serious risk of simply getting it wrong. This is incredibly likely where, for example, “excessive” prices are in the eye of the beholder and are extremely difficult to ascertain econometrically.

This unfortunate feature of EU competition enforcement would likely be, at least in part, replicated under the reforms proposed by the Expert Report. Indeed, the report’s focus on the welfare of “trading partners”—and particularly its focus on trading-partner welfare, regardless of whether perceived harm is passed on to consumers—comes dangerously close to the EU’s preoccupation with reducing the rents captured by monopolists.[68] While the Expert Report does not recommend an “excessive pricing” theory of harm—like the one that exists in the EU—it does echo the EU’s fixation on the immediate fortunes of trading partners (other than consumers) in ways that may ultimately lead to qualitatively equivalent results.

V. The Emulation of European Competition Law in the Expert Report’s Treatment of Specific Practices and Theories of Harm

Beyond the high-level differences discussed above, European and U.S. antitrust authorities also diverge significantly on numerous specific issues. These dissimilarities often result from the different policy goals that animate these two bodies of law. As noted, where U.S. case law is guided by an overarching goal of maximizing consumer welfare (notably, a practice’s effect on output), European competition law tends to favor structural presumptions and places a much heavier emphasis on distributional considerations. In addition, where the U.S. approach to many of these specific issues is deeply influenced by its overwhelming concern with the potentially chilling effects of intervention, this apprehension is very much foreign to European competition law. The result is often widely divergent approaches to complex economic matters in which the United States hews far more closely than does the EU to the humility and restraint suggested by economic learning.

Unfortunately, the recommendations put forward in the Expert Report would largely bring California antitrust law in line with the European approach for many theories of harm. Indeed, the Expert Report rejects the traditional U.S. antitrust-law concern with chilling procompetitive behavior, even proposing statutory language that would hold that “courts should bear in mind that the policy of California is that the risk of under-enforcement of the antitrust laws is greater than the risk of over-enforcement.”[69] Not only is this position unsupported, but it also entails an explicit rejection of a century of U.S. antitrust jurisprudence:

[U]sing language that mimics the Sherman Act would come with a potentially severe disadvantage: California state courts might then believe that they should apply 130 years of federal jurisprudence to cases brought under California state law. In recent decades, that jurisprudence has substantially narrowed the scope of the Sherman Act, as described above, so relying on it could well rob California law of the power it needs to protect competition.[70]

The evidence suggesting that competition has been poorly protected under Sherman Act jurisprudence is generally weak and unconvincing,[71] however, and the same is true for the specific theories of harm that the Expert Report would expand.

A. Predatory Pricing

Predatory pricing is one area where the Expert Report urges policymakers to copy specific rules in force in the EU. In its model statutory language, the Expert Report proposes that California establish that:

liability [for anticompetitive exclusionary conduct] does not require finding… that any price of the defendant for a product or service was below any measure of the costs to the defendant for providing the product or service…, [or] that in a claim of predatory pricing, the defendant is likely to recoup the losses it sustains from below-cost pricing of the products or services at issue[.][72]

U.S. antitrust law subjects allegations of predatory pricing to two strict conditions: 1) monopolists must charge prices that are below some measure of their incremental costs; and 2) there must be a realistic prospect that they will be able to recoup these first-period losses.[73] In laying out its approach to predatory pricing, the Supreme Court identified the risk of false positives and the clear cost of such errors to consumers. It therefore particularly stressed the importance of the recoupment requirement because, without recoupment, “predatory pricing produces lower aggregate prices in the market, and consumer welfare is enhanced.”[74]

Accordingly, in the United States, authorities must prove that there are constraints that prevent rival firms from entering the market after the predation scheme or that the scheme itself would effectively foreclose rivals from entering in the first place.[75] Otherwise, competitors would undercut the predator as soon as it attempts to charge supracompetitive prices to recoup its losses. In such a situation—without, that is, the strong likelihood of recouping the lost revenue from underpricing—the overwhelming weight of economic learning (to say nothing of simple logic) makes clear that predatory pricing is not a rational business strategy.[76] Thus, apparent cases of predatory pricing in the absence of the likelihood of recoupment are most likely not, in fact, predatory, and deterring or punishing them would likely actually harm consumers.

In contrast, the legal standard applied to predatory pricing in the EU is much laxer and almost certain, as a result, to risk injuring consumers. Authorities must prove only that a company has charged a price below its average variable cost, in which case its behavior is presumed to be predatory.[77] Even when a firm imposes prices that are between average variable and average total cost, it can be found guilty of predatory pricing if authorities show that its behavior was part of “a plan to eliminate competition.”[78] Most significantly, in neither case is it necessary for authorities to show that the scheme would allow the monopolist to recoup its losses.[79]

[I]t does not follow from the case-law of the Court that proof of the possibility of recoupment of losses suffered by the application, by an undertaking in a dominant position, of prices lower than a certain level of costs constitutes a necessary precondition to establishing that such a pricing policy is abusive.[80]

By affirmatively dispensing with each of these limitations, the Expert Report effectively recommends that California legislators shift California predatory-pricing law toward the European model. Unfortunately, such a standard has no basis in economic theory or evidence—not even in the “strategic” economic theory that arguably challenges the dominant, “Chicago School” understanding of predatory pricing.[81] Indeed, strategic predatory pricing still requires some form of recoupment and the refutation of any convincing business justification offered in response.[82] As Bruce Kobayashi and Tim Muris emphasize, the introduction of new possibility theorems, particularly uncorroborated by rigorous empirical reinforcement, does not necessarily alter the implementation of the error-cost analysis:

While the Post-Chicago School literature on predatory pricing may suggest that rational predatory pricing is theoretically possible, such theories do not show that predatory pricing is a more compelling explanation than the alternative hypothesis of competition on the merits. Because of this literature’s focus on theoretical possibility theorems, little evidence exists regarding the empirical relevance of these theories. Absent specific evidence regarding the plausibility of these theories, the courts… properly ignore such theories.[83]

The case of predatory pricing illustrates a crucial distinction between European and American competition law. The recoupment requirement embodied in U.S. antitrust law essentially differentiates aggressive pricing behavior that improves consumer welfare by leading to overall price decreases from predatory pricing that reduces welfare due to ultimately higher prices. In other words, it is entirely focused on consumer welfare.

The European approach, by contrast, reflects structuralist considerations that are far removed from a concern for consumer welfare. Its underlying fear is that dominant companies could, through aggressive pricing—even to the benefit of consumers—by their very success, engender more concentrated market structures. It is simply presumed that these less-atomistic markets are invariably detrimental to consumers. Both the Tetra Pak and France Télécom cases offer clear illustrations of the European Court of Justice’s reasoning on this point:

[I]t would not be appropriate, in the circumstances of the present case, to require in addition proof that Tetra Pak had a realistic chance of recouping its losses. It must be possible to penalize predatory pricing whenever there is a risk that competitors will be eliminated… The aim pursued, which is to maintain undistorted competition, rules out waiting until such a strategy leads to the actual elimination of competitors.[84]


[T]he lack of any possibility of recoupment of losses is not sufficient to prevent the undertaking concerned reinforcing its dominant position, in particular, following the withdrawal from the market of one or a number of its competitors, so that the degree of competition existing on the market, already weakened precisely because of the presence of the undertaking concerned, is further reduced and customers suffer loss as a result of the limitation of the choices available to them.[85]

In short, the European approach leaves much less room for analysis of a pricing scheme’s concrete effects, making it much more prone to false positives than the Brooke Group standard in the United States. It ignores not only the benefits that consumers may derive from lower prices, but also the chilling effect that broad predatory-pricing standards may exert on firms that attempt to attract consumers with aggressive pricing schemes. There is no basis for enshrining such an approach in California law.

B. Refusals to Deal

Refusals to deal are another area where the Expert Report’s recommendations would bring California antitrust rules more in line with the EU model. The Expert Report proposes in its example statutory language that:

[L]iability… does not require finding (i) that the unilateral conduct of the defendant altered or terminated a prior course of dealing between the defendant and a person subject to the exclusionary conduct; [or] (ii) that the defendant treated persons subject to the exclusionary conduct differently than the defendant treated other persons[.][86]

The Expert Report further highlights “Discrimination Against Rivals, for example by refusing to provide rivals of the defendant access to a platform or product or service that the defendant provides to other third-parties” as a particular area of concern.[87]

U.S. and EU antitrust laws are hugely different when it comes to refusals to deal. While the United States has imposed strenuous limits on enforcement authorities or rivals seeking to bring such cases, EU competition law sets a far lower threshold for liability. The U.S. approach is firmly rooted in the error-cost framework and, in particular, the conclusion that avoiding Type I (false-positive) errors is more important than avoiding Type II (false-negative) errors. As the Supreme Court held in Trinko:

[Enforced sharing] may lessen the incentive for the monopolist, the rival, or both to invest in those economically beneficial facilities. Enforced sharing also requires antitrust courts to act as central planners, identifying the proper price, quantity, and other terms of dealing—a role for which they are ill suited.[88]

In that case, the Court was unwilling to extend the reach of Section 2, cabining it to a very narrow set of circumstances:

Aspen Skiing is at or near the outer boundary of §2 liability. The Court there found significance in the defendant’s decision to cease participation in a cooperative venture. The unilateral termination of a voluntary (and thus presumably profitable) course of dealing suggested a willingness to forsake short-term profits to achieve an anticompetitive end.[89]

This highlights two key features of American antitrust law concerning refusals to deal. To start, U.S. antitrust law generally does not apply the “essential facilities” doctrine—indeed, as the Court held in Trinko, “we have never recognized such a doctrine.”[90] Accordingly, in the absence of exceptional facts, upstream monopolists are rarely required to supply their product to downstream rivals, even if that supply is “essential” for effective competition in the downstream market.

Moreover, as the Court observed in Trinko, the Aspen Skiing case appears to concern only those limited instances where a firm’s refusal to deal stems from the termination of a preexisting and profitable business relationship.[91] While even this is not likely to be the economically appropriate limitation on liability,[92] its impetus—ensuring that liability is found only in situations where procompetitive explanations for the challenged conduct are extremely unlikely—is appropriate for a regime concerned with minimizing the cost to consumers of erroneous enforcement decisions.

As in most areas of antitrust policy, EU competition law is much more interventionist. Refusals to deal are a central theme of EU enforcement efforts, and there is a relatively low threshold for liability.[93] In theory, for a refusal to deal to infringe EU competition law, it must meet a set of fairly stringent conditions: the input must be indispensable, the refusal must eliminate all competition in the downstream market, and there must not be objective reasons that justify the refusal.[94] Moreover, if the refusal to deal involves intellectual property, it must also prevent the appearance of a new good.[95] In practice, however, all of these conditions have been significantly relaxed by EU courts and the Commission’s decisional practice. This is best evidenced by the lower court’s Microsoft ruling. As John Vickers notes:

[T]he Court found easily in favor of the Commission on the IMS Health criteria, which it interpreted surprisingly elastically, and without relying on the special factors emphasized by the Commission. For example, to meet the “new product” condition it was unnecessary to identify a particular new product… thwarted by the refusal to supply but sufficient merely to show limitation of technical development in terms of less incentive for competitors to innovate.[96]

Thus, EU competition law is far less concerned about its potential chilling effect on firms’ investments than is U.S. antitrust law.

The Expert Report’s wording suggests that its authors would like to see California’s antitrust rules in this area move towards the European model. This seems particularly misguided for a state that so heavily relies on continued investments in innovation.

In discussing its concerns with the state of refusal-to-deal law in the United States, the Expert Report notes that:

[E]ven a monopolist can normally choose the parties with which it will deal and [] a monopolist’s selective refusal to deal with another firm, even a competitor, violates antitrust law only in unusual circumstances…. [The Court] explained that courts are ill-equipped to determine the terms on which one firm should be required to deal with another, so a bright line is necessary to preserve the incentives of both the monopolist and the competitor to compete aggressively in the marketplace. Such a rule may have been reasonable in a setting where “dealing” often meant incurring a large fixed cost to coordinate with the other firm. In an economy containing digital “ecosystems” that connect many businesses to one another, and digital markets with standardized terms of interconnection, such as established application program interfaces (APIs), that rule may immunize much conduct that could be anticompetitive.[97]

This approach is unduly focused on the welfare of specific competitors, rather than the effects on competition and consumers. Indeed, in the Aspen Skiing case (which did find a duty to deal on the defendant’s part), the Supreme Court is clear that the assessment of harm to competitors would be insufficient to establish that a refusal to deal was anticompetitive: “The question whether Ski Co.’s conduct may properly be characterized as exclusionary cannot be answered by simply considering its effect on Highlands. In addition, it is relevant to consider its impact on consumers and whether it has impaired competition in an unnecessarily restrictive way.”[98]

The Expert Report’s additional proposal that liability should not turn on whether the defendant treated particular parties differently in exercising exclusionary conduct (including refusal to deal)[99] is a further move away from effects-based analysis and toward the European model. As Einer Elhauge has noted, there is an important distinction between unconditional and discriminatory exclusionary conduct:

Efforts to simply improve a firm’s own efficiency and win sales by selling a better or cheaper product at above-cost prices should enjoy per se legality without any general requirement to share that greater efficiency with rivals. But exclusionary conditions that discriminate on the basis of rivalry by selectively denying property or products to rivals (or buyers who deal with rivals) are not necessary to further ex ante incentives to enhance the monopolist’s efficiency, and should be illegal when they create a marketwide foreclosure that impairs rival efficiency.[100]

By arguing to impose liability regardless of whether conduct is exercised in a discriminatory fashion, the Expert Report would remove the general protection under U.S. antitrust law for unconditional refusals to deal, and would instead apply the conditional standard to all exclusionary conduct.

It seems quite likely, in fact, that this provision is proposed as a rebuke to the 9th U.S. Circuit Court of Appeals’ holding in FTC v. Qualcomm, which found no duty to deal, in part, because the challenged conduct was applied to all rivals equally.[101] At least three of the Expert Report’s authors are on record as vigorously opposing the holding in Qualcomm.[102] But far from supporting a challenge to Qualcomm’s conduct on the grounds that it harmed competition by targeting threatening rivals, the Expert Report authors’ apparent preferred approach to Qualcomm’s alleged refusal to deal was to attempt to force a wholesale change in Qualcomm’s vertically integrated business model.

In other words, the authors would find liability regardless of how Qualcomm enforces its license terms, and would prefer a legal standard that does not condition that finding on exclusionary conduct against only certain rivals. In essence, they see operating at all in the relevant market as a harm.[103] Whatever the merits of this argument in the Qualcomm case, it should not be generalized to undermine the sensible limits that U.S. antitrust has imposed on the refusal-to-deal theory of harm.

C. Vertical and Platform Restraints

Finally, the Expert Report would take a leaf out of the European book when it comes to vertical restraints, including rebates, exclusive dealing, “most favored nation” (MFN) clauses, and platform conduct. Here, again, the Expert Report singles these practices out for attention:

Loyalty Rebates, which penalize a customer that conducts more business with the defendant’s rivals, as opposed to volume discounts, which are generally procompetitive;

Exclusive Dealing Provisions, which disrupt the ability of counterparties to deal with the defendant’s rivals, especially if such provisions are widely used by the defendant;

Most-Favored Nation Clauses, which prohibit counterparties from dealing with the defendant’s rivals on more favorable terms and conditions than those on which they deal with the defendant, especially if such clauses are widely used by the defendant.[104]

There are vast differences between U.S. and EU competition law with respect to vertical restraints. On the one hand, since the Supreme Court’s Leegin ruling, even price-related vertical restraints (such as resale price maintenance, or “RPM”) are assessed under the rule of reason in the United States.[105] Some commentators have gone so far as to say that, in practice, U.S. case law almost amounts to per se legality.[106] Conversely, EU competition law treats RPM as severely as it treats cartels. Both RPM and cartels are considered restrictions of competition “by object”—the EU’s equivalent of a per se prohibition.[107] This severe treatment also applies to nonprice vertical restraints that tend to partition the European internal market.[108] Furthermore, in the Consten and Grundig ruling, the ECJ rejected the consequentialist (and economically grounded) principle that inter-brand competition is the appropriate touchstone to assess vertical restraints:

Although competition between producers is generally more noticeable than that between distributors of products of the same make, it does not thereby follow that an agreement tending to restrict the latter kind of competition should escape the prohibition of Article 85(1) merely because it might increase the former.[109]

This especially stringent stance toward vertical restrictions flies in the face of the longstanding mainstream-economics literature addressing the subject. As Patrick Rey and Jean Tirole (hardly the most free-market of economists) saw it as long ago as 1986: “Another major contribution of the earlier literature on vertical restraints is to have shown that per se illegality of such restraints has no economic foundations.”[110]

While there is theoretical literature (rooted in so-called “possibility theorems”) that suggests firms can engage in anticompetitive vertical conduct, the empirical evidence strongly suggests that, even though firms do impose vertical restraints, it is exceedingly rare that they have net anticompetitive effects. Nor is the relative absence of such evidence for a lack of looking: countless empirical papers have investigated the competitive effects of vertical integration and vertical contractual arrangements and found predominantly procompetitive benefits or, at worst, neutral effects.[111]

Unlike in the EU, the U.S. Supreme Court in Leegin took account of the weight of the economic literature and changed its approach to RPM to ensure that the law no longer simply precluded its arguable consumer benefits: “Though each side of the debate can find sources to support its position, it suffices to say here that economics literature is replete with procompetitive justifications for a manufacturer’s use of resale price maintenance.”[112] Further, “[the prior approach to resale price maintenance restraints] hinders competition and consumer welfare because manufacturers are forced to engage in second-best alternatives and because consumers are required to shoulder the increased expense of the inferior practices.”[113]

By contrast, the EU’s continued per se treatment of RPM strongly reflects its precautionary-principle approach to antitrust, under which European regulators and courts readily condemn conduct that could conceivably injure consumers, even where such injury is, according to the best economic understanding, unlikely (at best).[114] The U.S. approach to such vertical restraints, which rests on likelihood rather than mere possibility,[115] is far less likely to erroneously condemn beneficial conduct.

There are also significant differences between the U.S. and EU stances on the issue of rebates. This reflects the EU’s relative willingness to disregard complex economics in favor of noneconomic, formalist presumptions (at least, prior to the ECJ’s Intel ruling). Whereas U.S. antitrust has predominantly moved to an effects-based assessment of rebates,[116] this is only starting to happen in the EU. Prior to the ECJ’s Intel ruling, the EU implemented an overly simplistic approach to assessing rebates by dominant firms, where so-called “fidelity” rebates were almost per se illegal.[117] Likely recognizing the problems inherent in this formalistic assessment of rebates, the ECJ’s Intel ruling moved the European case law on rebates to a more evidence-based approach, holding that:

[T]he Commission is not only required to analyse, first, the extent of the undertaking’s dominant position on the relevant market and, secondly, the share of the market covered by the challenged practice, as well as the conditions and arrangements for granting the rebates in question, their duration and their amount; it is also required to assess the possible existence of a strategy aiming to exclude competitors that are at least as efficient as the dominant undertaking from the market.[118]

As Advocate General Nils Wahl noted in his opinion in the case, only such an evidence-based approach could ensure that the challenged conduct was actually harmful:

In this section, I shall explain why an abuse of dominance is never established in the abstract: even in the case of presumptively unlawful practices, the Court has consistently examined the legal and economic context of the impugned conduct. In that sense, the assessment of the context of the conduct scrutinised constitutes a necessary corollary to determining whether an abuse of dominance has taken place. That is not surprising. The conduct scrutinised must, at the very least, be able to foreclose competitors from the market in order to fall under the prohibition laid down in Article 102 TFEU.”[119]

The Expert Report, however, contains a direct refutation of Intel, thus “out-Europing” even Europe itself in its treatment of vertical restraints:

7) Plaintiffs need not show that the rivals whose ability to compete has been reduced are as efficient, or nearly as efficient, as the defendant. Harm to competition can arise when the competitive constraints on the defendant are weakened even when those competitive constraints come from less efficient rivals. Indeed, harm to competition can be especially great when a firm that faces limited competition further weakens its rivals.[120]

If adopted, this language would significantly limit the need for California courts to show actual anticompetitive harm arising from challenged vertical conduct. Similarly, the Expert Report’s rejection of the “no-economic-sense” test—“liability…does not require finding… that the conduct of the defendant makes no economic sense apart from its tendency to harm competition”[121]—removes another mechanism to ensure that vertical restraints lead to actual consumer harm, rather than simply injury to a competitor.

As Thom Lambert persuasively demonstrates, there are imperfections with both the “as efficient competitor” test and the “no economic sense” test. But these commonly applied tools do at least help to ensure that courts undertake to find actual anticompetitive harm.[122] The rejection of both simultaneously is decidedly problematic, suggesting a preference for no serious economic constraints on courts’ discretion to condemn practices solely on the ground of structural harm—i.e., harm to certain competitors.

By contrast, the alternative definition that Lambert proposes “would deem conduct to be unreasonably exclusionary if it would exclude from the defendant’s market a ‘competitive rival,’ defined as a rival that is both as determined as the defendant and capable, at minimum efficient scale, of matching the defendant’s efficiency.”[123] While this test may appear to have some traits in common with the Expert Report’s “diminishing competitive constraints” approach, it incorporates a much more robust set of principles and limitations, designed to more clearly distinguish conduct that merely excludes from exclusions that actually cause anticompetitive harm, while minimizing administrative costs.[124] The Expert Report, by contrast, explicitly removes such limitations.

A related problem concerns the Expert Report’s proposal that “when a defendant operates a multi-sided platform business, [liability does not turn on whether] the conduct of the defendant presents harm to competition on more than one side of the multi-sided platform[.]”[125] This provision is meant to reverse the Supreme Court’s holding on platform vertical restraints in Ohio v. American Express that:

Due to indirect network effects, two-sided platforms cannot raise prices on one side without risking a feedback loop of declining demand. And the fact that two-sided platforms charge one side a price that is below or above cost reflects differences in the two sides’ demand elasticity, not market power or anticompetitive pricing. Price increases on one side of the platform likewise do not suggest anticompetitive effects without some evidence that they have increased the overall cost of the platform’s services. Thus, courts must include both sides of the platform—merchants and cardholders—when defining the credit-card market….

…For all these reasons, “[i]n two-sided transaction markets, only one market should be defined.” Any other analysis would lead to “mistaken inferences” of the kind that could “chill the very conduct the antitrust laws are designed to protect.”[126]

As Greg Werden notes, “[a]lleging the relevant market in an antitrust case does not merely identify the portion of the economy most directly affected by the challenged conduct; it identifies the competitive process alleged to be harmed.”[127] Particularly where novel conduct or novel markets are involved, and thus the relevant economic relationships are poorly understood, market definition is crucial to determine “what the nature of [the relevant] products is, how they are priced and on what terms they are sold, what levers [a firm] can use to increase its profits, and what competitive constraints affect its ability to do so.”[128] This is the approach the Supreme Court employed in Amex.

The Expert Report’s proposal to overrule Amex in California is deeply misguided. The economics of two-sided markets are such that “there is no meaningful economic relationship between benefits and costs on each side of the market considered alone…. [A]ny analysis of social welfare must account for the pricing level, the pricing structure, and the feasible alternatives for getting all sides on board.”[129] Assessing anticompetitive harm with respect to only one side of a two-sided market will arbitrarily include and exclude various sets of users and transactions, and incorrectly assess the extent and consequences of market power.[130]

Indeed, evidence of a price effect on only one side of a two-sided platform can be consistent with either neutral, anticompetitive, or procompetitive conduct.[131] Only when output is defined to incorporate the two-sidedness of the product, and where price and quality are assessed on both sides of a sufficiently interrelated two-sided platform, is it even possible to distinguish between procompetitive and anticompetitive effects. In fact, “[s]eparating the two markets allows legitimate competitive activities in the market for general purposes to be penalized no matter how output-enhancing such activities may be.”[132]

Notably, while some scholars have opposed the Amex holding that both sides of a two-sided market must be included in the relevant market in order to assess anticompetitive harm, some of these critics appear to note that the problem is not that both sides should not be taken into account at all, but only that they should not be included in the same relevant market (thus, permitting a plaintiff to make out a prima facie case by showing harm to just one side).[133] The language proposed in the Expert Report, however, would go even further, seemingly permitting a finding of liability based solely on harm to one side of a multi-sided market, regardless of countervailing effects on the other side. As in the Amex case itself, such an approach would confer benefits on certain platform business users (in Amex, retailers) at the direct expense of consumers (in Amex, literal consumers of retail goods purchased by credit card).

Adopting such an approach in California—whose economy is significantly dependent on multisided digital-platform firms, including both incumbents and startups[134]—would imperil the state’s economic prospects[135] and exacerbate the incentives for such firms to take jobs, investments, and tax dollars elsewhere.[136]

[1] Antitrust Law — Study B-750, California Law Revision Commission (last revised Apr. 26, 2024), available at

[2] We welcome the opportunity to comment further or to respond to questions about our comments. Please contact us at [email protected].

[3] Regulation (EU) 2022/1925 of the European Parliament and of the Council of 14 September 2022 on Contestable and Fair Markets in the Digital Sector and amending Directives (EU) 2019/1937 and (EU) 2020/1828, 2022 O.J. (L 265) 1.

[4] See Aaron Edlin, Doug Melamed, Sam Miller, Fiona Scott Morton, & Carl Shapiro, Expert Report on Single Firm Conduct, 2024 Cal. L. Rev. Comm’n (hereinafter “Expert Report”), available at ExRpt-B750-Grp1.pdf.

[5] Id. at 14.

[6] Frank H. Easterbrook, The Limits of Antitrust, 63 Tex. L. Rev. 1, 15 (1984).

[7] See, especially, Pac. Bell Tel. Co. v. linkLine Commc’ns, Inc., 555 U.S. 438 (2009); Credit Suisse Sec. (U.S.A) LLC v. Billing, 551 U.S. 264, 265 (2007); Verizon Comm. v. Law Offices of Trinko, 540 U.S. 398 (2004).

[8] Trinko, 540 U.S. at 414 (quoting Matsushita Elec. Indus. Co. v. Zenith Radio Corp., 475 U.S. 574, 594 (1986)).

[9] Easterbrook, supra note 6, at 7.

[10] See, e.g., Aurelien Portuese, The Rise of Precautionary Antitrust: An Illustration with the EU Google Android Decision, CPI EU News November 2019 (2019) at 4 (“The absence of demonstrated consumer harm in order to find antitrust injury is not fortuitous, but represents a fundamental alteration of antitrust enforcement, predominantly when it comes to big tech companies. Coupled with the lack of clear knowledge, a shift in the burden of proof, and the lack of a consumer harm requirement in order to find abuse of dominance all reveal the precautionary approach that the European Commission has now embraced.”).

[11] See Nassim Nicholas Taleb, Rupert Read, Raphael Douady, Joseph Norman, & Yaneer Bar-Yam, The Precautionary Principle (With Application to the Genetic Modification of Organisms), arXiv preprint arXiv:1410.5787, 2 (2014). (“The purpose of the PP is to avoid a certain class of what, in probability and insurance, is called “ruin” problems. A ruin problem is one where outcomes of risks have a non-zero probability of resulting in unrecoverable losses.”).

[12] The precautionary principles implies that policymakers should bar certain mutually advantageous transactions due to the social costs that they might impose further down the line. Moreover, the precautionary principle has historically been associated with anti-growth positions. See, e.g., Jaap C Hanekamp, Guillaume Vera?Navas, & SW Verstegen, The Historical Roots of Precautionary Thinking: The Cultural Ecological Critique and ‘The Limits to Growth’, 8 J. Risk Res. 295, 299 (2005) (“The first inklings of today’s precautionary thinking as a means of creating a sustainable society can be traced historically to ‘The Limits to Growth’…”).

[13] See, e.g., Greg Ip, Europe Regulates Its Way to Last Place, Wall St. J. (Jan. 31, 2024), (“Of course, Europe’s economy underperforms for lots of reasons, from demographics to energy costs, not just regulation. And U.S. regulators aren’t exactly hands-off. Still, they tend to act on evidence of harm, whereas Europe’s will act on the mere possibility. This precautionary principle can throttle innovation in its cradle.”) (emphasis added).

[14] See, e.g., id.; Eric Albert, Europe Trails Behind the United States in Economic Growth, Le Monde (Nov. 1, 2023), (“For the past fifteen years, Europe has been falling further and further behind…. Since 2007, per capita growth on the other side of the Atlantic has been 19.2%, compared with 7.6% in the eurozone. A gap of almost twelve points.”).

[15] Fredrik Erixon, Oscar Guinea, & Oscar du Roy, If the EU Was a State in the United States: Comparing Economic Growth Between EU and US States, ECIPE Policy Brief No. 07/2023 (2023), available at

[16] Among other things, the Expert Report argues that antitrust should be used to address alleged policy concerns broader than protecting competition, and should accept reductions in competition to do so. See Expert Report, supra note 1, at 2 (“Nonetheless, these important values [‘broader social and political goals’] can influence the evidentiary standards that the Legislature instructs the courts to apply when handling individual antitrust cases. For example, the California Legislature could instruct the courts to err on the side of enforcement when the effect of the conduct at issue on competition is uncertain.”). But as one of the authors of the Expert Report has himself noted elsewhere: “while antitrust enforcement has a vital role to play in keeping markets competitive, antitrust law and antitrust institutions are ill suited to directly address concerns associated with the political power of large corporations or other public policy goals such as income inequality or job creation.” Carl Shapiro, Antitrust in a Time of Populism, 61 Int’l J. Indus. Org. 714, 714 (2018) (emphasis added).

[17] See generally Easterbrook, supra note 6, at 14-15. See also Geoffrey A. Manne & Joshua D. Wright, Innovation and the Limits of Antitrust, 6 J. Comp. L. & Econ. 153 (2010).

[18] See Robert W. Crandall & Clifford Winston, Does Antitrust Policy Improve Consumer Welfare? Assessing the Evidence, 17 J. Econ. Persp. 3, 4 (2003) (“[T]he economics profession should conclude that until it can provide some hard evidence that identi?es where the antitrust authorities are signi?cantly improving consumer welfare and can explain why some enforcement actions and remedies are helpful and others are not, those authorities would be well advised to prosecute only the most egregious anticompetitive violations.”).

[19] David Autor, David Dorn, Lawrence F. Katz, Christina Patterson & John Van Reenen, The Fall of the Labor Share and the Rise of Superstar Firms, 135 Q.J. Econ. 645, 651 (2020) (citations omitted) (emphasis added).

[20] See, e.g., Thomas Philippon, The Great Reversal: How America Gave Up on Free Markets (2019); Jan De Loecker, Jan Eeckhout, & Gabriel Unger, The Rise of Market Power and the Macroeconomic Implications, 135 Q. J. Econ. 561 (2020); David Wessel, Is Lack of Competition Strangling the U.S. Economy?, Harv. Bus. Rev. (Apr. 2018),; Adil Abdela & Marshall Steinbaum, The United States Has a Market Concentration Problem, Roosevelt Institute Issue Brief (2018), available at

[21] A number of papers simply do not find that the accepted story—built in significant part around the famous De Loecker, Eeckhout, & Unger study, id.—regarding the vast size of markups and market power is accurate. The claimed markups due to increased concentration are likely not nearly as substantial as commonly assumed. See, e.g., James Traina, Is Aggregate Market Power Increasing? Production Trends Using Financial Statements, Stigler Center Working Paper (Feb. 2018), available at; see also World Economic Outlook, April 2019 Growth Slowdown, Precarious Recovery, International Monetary Fund (Apr. 2019), available at Another study finds that profits have increased, but are still within their historical range. See Loukas Karabarbounis & Brent Neiman, Accounting for Factorless Income, 33 NBER Macro. Annual 167 (2019). And still another shows decreased wages in concentrated markets, but also that local concentration has been decreasing over the relevant time period, suggesting that lack of enforcement is not a problem. See Kevin Rinz, Labor Market Concentration, Earnings, and Inequality, 57 J. Hum. Resources S251 (2022).

[22] See Esteban Rossi-Hansberg, Pierre-Daniel Sarte, & Nicholas Trachter, Diverging Trends in National and Local Concentration, 35 NBER Macro. Annual 115, 116 (2020) (“[T]he observed positive trend in market concentration at the national level has been accompanied by a corresponding negative trend in average local market concentration…. The narrower the geographic definition, the faster is the decline in local concentration. This is meaningful because the relevant definition of concentration from which to infer changes in competition is, in most sectors, local and not national.”).

[23] Id. at 117 (emphasis added).

[24] Sharat Ganapati, Growing Oligopolies, Prices, Output, and Productivity, 13 Am. Econ. J. Micro. 309, 323-24 (2021) (emphasis added).

[25] Chang-Tai Hsieh & Esteban Rossi-Hansberg, The Industrial Revolution in Services, 1 J. Pol. Econ. Macro. 3, 3 (2023) (emphasis added). See also id. at 39 (“Over the past 4 decades, the US economy has experienced a new industrial revolution that has enabled ?rms to scale up production over a large number of establishments dispersed across space. The adoption of these technologies has particularly favored productive ?rms in nontraded-service industries. The industrial revolution in services has had its largest effect in smaller and mid-sized local markets…. The gain to local consumers from access to more, better, and novel varieties of local services from the entry of top ?rms into local markets is not captured by the BLS. We estimate that such ‘missing growth’ is as large as 1.6% in the smallest markets and averages 0.5% per year from 1977 to 2013 across all US cities.”) (emphasis added).

[26] David Berger, Kyle Herkenhoff & Simon Mongey, Labor Market Power, 112 Am. Econ. Rev. 1147, 1148-49 (2022).

[27] Shapiro, Antitrust in a Time of Populism, supra note 16, at 727-28.

[28] Expert Report, supra note 1, at 15 (emphasis added).

[29] Id. at 2.

[30] A. Douglas Melamed, Antitrust Law and Its Critics, 83 Antitrust L.J. 269, 285 (2020).

[31] Herbert J. Hovenkamp & Fiona Scott Morton, Framing the Chicago School of Antitrust Analysis, 168 U. Penn. L. Rev. 1843, 1870-71 (2020).

[32] Easterbrook, supra note 6, at 2-3.

[33] Hovenkamp & Scott Morton, supra note 31, at 1849.

[34] See generally Geoffrey A. Manne, Error Costs in Digital Markets, in Global Antitrust Institute Report on the Digital Economy (Joshua D. Wright & Douglas H. Ginsburg eds., 2020), available at

[35] Bruce H. Kobayashi & Timothy J. Muris, Chicago, Post-Chicago, and Beyond: Time to Let Go of the 20th Century, 78 Antitrust L.J. 147, 166 (2012).

[36] See id. at 166 (“[T]here is very little empirical evidence based on in-depth industry studies that RRC is a significant antitrust problem.”); id. at 148 (“Because of [the Post-Chicago School] literature’s focus on theoretical possibility theorems, little evidence exists regarding the empirical relevance of these theories.”).

[37] See Expert Report, supra note 1, at 7 (“The history of federal antitrust enforcement of single-firm conduct illustrates that when courts are uncertain about how to assess conduct, they often find in favor of defendants even if the conduct harms competition simply because the plaintiff bears the burden of proof.”).

[38] See supra notes 19-27, and accompanying text.

[39] See Case C-413/14 P Intel v Commission, ECLI:EU:C:2017:788.

[40] See Steven Berry, Martin Gaynor, & Fiona Scott Morton, Do Increasing Markups Matter? Lessons from Empirical Industrial Organization, 33 J. Econ. Persp. 48 (2019). See also Jonathan Baker & Timothy F. Bresnahan, Economic Evidence in Antitrust: Defining Markets and Measuring Market Power in Handbook of Antitrust Economics 1 (Paolo Buccirossi ed., 2008) (“The Chicago identification argument has carried the day, and structure-conduct-performance empirical methods have largely been discarded in economics.”).

[41] See, e.g., Gregory J. Werden & Luke Froeb, Don’t Panic: A Guide to Claims of Increasing Concentration 33 Antitrust 74 (2018),, and papers cited therein. As Werden & Froeb conclude: No evidence we have uncovered substantiates a broad upward trend in the market concentration in the United States, but market concentration undoubtedly has increased significantly in some sectors, such as wireless telephony. Such increases in concentration, however, do not warrant alarm or imply a failure of antitrust. Increases in market concentration are not a concern of competition policy when concentration remains low, yet low levels of concentration are being cited by those alarmed about increasing concentration…. Id. at 78. See also Joshua D. Wright, Elyse Dorsey, Jonathan Klick, & Jan M. Rybnicek, Requiem for a Paradox: The Dubious Rise and Inevitable Fall of Hipster Antitrust, 51 Ariz. St. L.J. 293 (2019).

[42] See, e.g., Expert Report, supra note 1, at 15.

[43] Francine Lafontaine & Margaret Slade, Exclusive Contracts and Vertical Restraints: Empirical Evidence and Public Policy, in Handbook of Antitrust Economics 391 (Paolo Buccirossi ed., 2008).

[44] See, e.g., Daniel P. O’Brien, The Antitrust Treatment of Vertical Restraints: Beyond the Possibility Theorems, in The Pros and Cons of Vertical Restraints 40, 72-76 (Swedish Competition Authority, 2008) (“[Vertical restraints] are unlikely to be anticompetitive in most cases.”); James C. Cooper, et al., Vertical Antitrust Policy as a Problem of Inference, 23 Int’l J. Indus. Org. 639 (2005) (surveying the empirical literature, concluding that although “some studies find evidence consistent with both pro- and anticompetitive effects… virtually no studies can claim to have identified instances where vertical practices were likely to have harmed competition”); Benjamin Klein, Competitive Resale Price Maintenance in the Absence of Free-Riding, 76 Antitrust L.J. 431 (2009); Bruce H. Kobayashi, Does Economics Provide a Reliable Guide to Regulating Commodity Bundling by Firms? A Survey of the Economic Literature, 1 J. Comp. L. & Econ. 707 (2005).

[45] James Cooper, Luke Froeb, Daniel O’Brien, & Michael Vita, Vertical Restrictions and Antitrust Policy: What About the Evidence?, Comp. Pol’y Int’l 45 (2005).

[46] Id.

[47] Expert Report, supra note 1, at 16: (b) Conduct, whether by one or multiple actors, is deemed to be anticompetitive exclusionary conduct, if the conduct tends to (1) diminish or create a meaningful risk of diminishing the competitive constraints imposed by the defendant’s rivals and thereby increase or create a meaningful risk of increasing the defendant’s market power, and (2) does not provide sufficient benefits to prevent the defendant’s trading partners from being harmed by that increased market power.

[48] Id.

[49] See TransUnion LLC v. Ramirez, 141 S. Ct. 2190, 2210-11 (2021) (“The plaintiffs rely on language from Spokeo where the Court said that ‘the risk of real harm’ (or as the Court otherwise stated, a ‘material risk of harm’) can sometimes ‘satisfy the requirement of concreteness…. [but] in a suit for damages, the mere risk of future harm, standing alone, cannot qualify as a concrete harm—at least unless the exposure to the risk of future harm itself causes a separate concrete harm.”) (citations omitted).

[50] In essence, for uncertain future effects, U.S. antitrust law applies something like a “reasonableness” standard. See U.S. v. Microsoft Corp., 253 F.3d 34, 79 (D.C. Cir. 2001) (enjoining “conduct that is reasonably capable of contributing significantly to a defendant’s continued monopoly power”) (emphasis added). Of course, “material risk” is undefined, so perhaps it is meant to accord with this standard. If so, it should use the same language.

[51] Herbert Hovenkamp, Antitrust Harm and Causation, 99 Wash. U. L. Rev. 787, 841 (2021). See also id. at 788 (“While a showing of actual harm can be important evidence, in most cases the public authorities need not show that harm has actually occurred, but only that the challenged conduct poses an unreasonable danger that it will occur.”) (emphasis added).

[52] Expert Report, supra note 1, at 16.

[53] See Brunswick Corp. v. Pueblo Bowl-O-Mat, Inc., 429 U.S. 477, 487-88 (1977) (“If the acquisitions here were unlawful, it is because they brought a ‘deep pocket’ parent into a market of ‘pygmies.’ Yet respondents’ injury—the loss of income that would have accrued had the acquired centers gone bankrupt—bears no relationship to the size of either the acquiring company or its competitors. Respondents would have suffered the identical ‘loss’—but no compensable injury—had the acquired centers instead obtained refinancing or been purchased by ‘shallow pocket’ parents, as the Court of Appeals itself acknowledged. Thus, respondents’ injury was not of ‘the type that the statute was intended to forestall[.]’”) (citations omitted).

[54] Expert Report, supra note 1, at 17.

[55] Microsoft, 253 F.3d at 79.

[56] Treaty on European Union, Protocol (No27) on the internal market and competition, Official Journal 115.

[57] See especially Expert Report supra note 1, at 17, §§ (f)(8) & (g) through (i).

[58] See, e.g., Joaquín Almunia, Competition and Consumers: The Future of EU Competition Policy, Speech at European Competition Day, Madrid (May 12, 2010), available at (“All of us here today know very well what our ultimate objective is: Competition policy is a tool at the service of consumers. Consumer welfare is at the heart of our policy and its achievement drives our priorities and guides our decisions.”). Even then, however, it must be noted that Almunia elaborated that “[o]ur objective is to ensure that consumers enjoy the benefits of competition, a wider choice of goods, of better quality and at lower prices.” Id. (emphasis added). In fact, expanded consumer choice is not necessarily the same thing as consumer welfare, and may at times be at odds with it. See Joshua D. Wright & Douglas H. Ginsburg, The Goals of Antitrust: Welfare Trumps Choice, 81 Fordham L. Rev. 2405 (2013).

[59] See Commission Guidance on the Commission’s Enforcement Priorities in Applying Article 82 of the EC Treaty to Abusive Exclusionary Conduct by Dominant Undertakings, 2009 O. J.(C 45)7 at n. 5, §6 (“[T]he Commission is mindful that what really matters is protecting an effective competitive process and not simply protecting competitors.”).

[60] See Case C-209/10, Post Danmark A/S v Konkurrencerådet, ECLI:EU:C:2012:172, §22 (“Competition on the merits may, by definition, lead to the departure from the market or the marginalisation of competitors that are less efficient and so less attractive to consumers….”).

[61] See Pablo Ibáñez Colomo, Exclusionary Discrimination Under Article 102 TFEU, 51 Common Market L. Rev. 153 (2014).

[62] Id.

[63] Expert Report, supra note 1, at 16.

[64] Id.

[65] See Brian Albrecht, Dirk Auer, & Geoffrey A. Manne, Labor Monopsony and Antitrust Enforcement: A Cautionary Tale, ICLE White Paper No. 2024-05-01 (2024) at 21, available at (“[Conduct] that creates monopsony power will necessarily reduce the prices and quantity purchased of inputs like labor and materials. But this same effect (reduced prices and quantities for inputs) would also be observed if the [conduct] is efficiency enhancing. If there are efficiency gains, the [] entity may purchase fewer of one or more inputs than [it would otherwise]. For example, if the efficiency gain arises from the elimination of redundancies in a hospital…, the hospital will buy fewer inputs, hire fewer technicians, or purchase fewer medical supplies.”). See also Ivan Kirov & James Traina, Labor Market Power and Technological Change in US Manufacturing, conference paper for Institute for Labor Economics (Oct. 2022), at 42, available at (“The labor [markdown] therefore increases because ‘productivity’ rises, and not because pay falls. This suggests that technological change plays a large role in the rise of the labor [markdown].”).

[66] Expert Report, supra note 1, at 15 (emphasis added).

[67] Trinko, 540 U.S. at 407.

[68] See Expert Report, supra note 1, at 16 (“‘Trading partners’ are parties with which the defendant deals, either as a customer or as a supplier. In [assessing anticompetitive exclusionary conduct], a trading partner is deemed to be harmed or benefited even if that trading partner passes some or all of that harm or benefit on to other parties.”).

[69] Id. at 15 (emphasis added).

[70] Id. at 13.

[71] See supra Section II.

[72] Expert Report, supra note 1, at 17. As the Expert Report acknowledges elsewhere, recoupment is a “requirement for a predatory pricing claim under federal antitrust law.” Id. at 15.

[73] See Brooke Group Ltd. v. Brown & Williamson Tobacco Corp., 509 U.S. 209, 222-27 (1993).

[74] Id. at 224.

[75] On entry deterrence, see Steven C. Salop, Strategic Entry Deterrence, 69 Am. Econ. Rev. 335 (1979).

[76] See generally John S. McGee, Predatory Pricing Revisited, 23 J.L. Econ 289 (1980). Some economists have more recently posed a “strategic” theory of predatory pricing that purports to expand substantially (and redirect) the scope of circumstances in which predatory pricing could be rational. See, e.g., Patrick Bolton, Joseph F. Brodley, & Michael H. Riordan, Predatory Pricing: Strategic Theory and Legal Policy, 88 Geo. L. J. 2239 (2000). While this and related theories have, indeed, likely expanded the theoretical scope of circumstances conducive to predatory pricing, they have not established that these conditions are remotely likely to occur. See Bruce H. Kobayashi, The Law and Economics of Predatory Pricing, in 4 Encyclopedia of Law and Economics (De Geest, ed. 2017) (“The models showing rational predation can exist and the evidence consistent with episodes of predation do not demonstrate that predation is either ubiquitous or frequent. Moreover, many of these models do not consider the welfare effects of predation, and those that do generally find the welfare effects ambiguous.”). From a legal perspective, particularly given the risk of error in discerning the difference between predatory pricing and legitimate price cutting, it is far more important to limit cases to situations likely to cause consumer harm rather than those in which harm is a remote possibility. The cost of error, of course, is the legal imposition of artificially inflated prices for consumers.

[77] Case C-62/86, AKZO v Comm’n, EU:C:1991:286, ¶¶ 71-72.

[78] Id. at ¶ 72 (“[P]rices below average total costs, that is to say, fixed costs plus variable costs, but above average variable costs, must be regarded as abusive if they are determined as part of a plan for eliminating a competitor.”).

[79] Case C-333/94 P, Tetra Pak v Comm’n, EU:C:1996:436, ¶ 44. See also, Case C-202/07 P, France Télécom v Comm’n, EU:C:2009:214, ¶ 110.

[80] Id. at ¶ 107.

[81] See, e.g., Bolton, Brodley, & Riordan, supra note 76.

[82] See id. at 2267 (“[A]nticipated recoupment is intrinsic in [strategic] theories, because without such an expectation predatory pricing is not sensible economic behavior.”). See also Kenneth G. Elzinga & David E. Mills, Predatory Pricing and Strategic Theory, 89 Geo. L.J. 2475, 2483 (2001) (“Of course, no proposed scheme of predation is credible unless it embodies a plausible means of recoupment, but this does not justify taking shortcuts in analysis. In particular, it is unwise to presume that a plausible means of recoupment exists just because facts supporting other features of a strategic theory, such as asymmetric information, are evident. Facts conducive to probable recoupment ought to be established independently.”).

[83] Kobayashi & Muris, supra note 35, at 166.

[84] Tetra Pak, supra note 79, at ¶ 44.

[85] France Télécom, supra note 79, at ¶ 112.

[86] Expert Report, supra note 1, at 17.

[87] Expert Report, supra note 1, at 15.

[88] Trinko, 540 U.S. at 408.

[89] Trinko, 540 U.S. at 409.

[90] Trinko, 540 U.S. at 411. See also Phillip Areeda, Essential Facilities: An Epithet in Need of Limiting Principles, 58 Antitrust L.J. 841 (1989).

[91] Aspen Skiing Co. v. Aspen Highlands Skiing Corp., 472 U.S. 585, 610-11 (1985).

[92] See Alan J. Meese, Property, Aspen, and Refusals to Deal, 73 Antitrust L. J. 81, 112-13 (2005).

[93] See Joined Cases 6/73 & 7/73, Instituto Chemioterapico Italiano S.p.A. and Commercial Solvents Corporation v. Comm’n, 1974 E.C.R. 223, [1974] 1 C.M.L.R. 309.

[94] See Case C-7/97, Oscar Bronner GmbH & Co. KG v Mediaprint Zeitungs, EU:C:1998:569, §41.

[95] See Case C-241/91 P, RTE and ITP v Comm’n, EU:C:1995:98, §54. See also, Case C-418/01, IMS Health, EU:C:2004:257, §37.

[96] John Vickers, Competition Policy and Property Rights, 120 Econ. J. 390 (2010).

[97] Expert Report, supra note 1, at 7.

[98] Aspen Skiing Co. v. Aspen Highlands Skiing Corp., 472 U.S. 585, 605 (1985).

[99] Expert Report, supra note 1, at 17.

[100] Einer Elhauge, Defining Better Monopolization Standards, 56 Stan. L. Rev. 253, 343 (2003).

[101] See Fed. Trade Comm’n v. Qualcomm Inc., 969 F.3d 974, 995 (9th Cir. 2020) (“Finally, unlike in Aspen Skiing, the district court found no evidence that Qualcomm singles out any specific chip supplier for anticompetitive treatment in its SEP-licensing. In Aspen Skiing, the defendant refused to sell its lift tickets to a smaller, rival ski resort even as it sold the same lift tickets to any other willing buyer (including any other ski resort)…. Qualcomm applies its OEM-level licensing policy equally with respect to all competitors in the modem chip markets and declines to enforce its patents against these rivals…. Instead, Qualcomm provides these rivals indemnifications…—the Aspen Skiing equivalent of refusing to sell a skier a lift ticket but letting them ride the chairlift anyway. Thus, while Qualcomm’s policy toward OEMs is ‘no license, no chips,’ its policy toward rival chipmakers could be characterized as ‘no license, no problem.’ Because Qualcomm applies the latter policy neutrally with respect to all competing modem chip manufacturers, the third Aspen Skiing requirement does not apply.”)

[102] Carl Shapiro was an economic expert for the FTC in the case, and Fiona Scott Morton was an economic expert for Apple in related litigation against Qualcomm. Doug Melamed was co-author of an amicus brief supporting the FTC in the 9th U.S. Circuit Court of Appeals. (In the interests of full disclosure, we authored an amicus brief, joined by 12 scholars of law & economics, supporting Qualcomm in the 9th Circuit. See Brief of Amici Curiae International Center for Law & Economics and Scholars of Law and Economics in Support of Appellant and Reversal, FTC v. Qualcomm, No. 19-16122 (9th Cir., Aug. 30, 2019), available at

[103] For a discussion of the frailties of these arguments, see Geoffrey A. Manne & Dirk Auer, Exclusionary Pricing Without the Exclusion: Unpacking Qualcomm’s No License, No Chips Policy, Truth on the Market (Jan. 17, 2020), (“The amici are thus left with the argument that Qualcomm could structure its prices differently, so as to maximize the profits of its rivals. Why it would choose to do so, or should indeed be forced to, is a whole other matter.”). For a response by one of the Expert Report authors, see Mark A. Lemley, A. Douglas Melamed, & Steve Salop, Manne and Auer’s Defense of Qualcomm’s Licensing Policy Is Deeply Flawed, Truth on the Market (Jan. 21, 2020),

[104] Expert Report, supra note 1, at 15.

[105] See Leegin Creative Leather Prods., Inc. v. PSKS, Inc., 551 U.S. 877 (2007).

[106] See, e.g., D. Daniel Sokol, The Transformation of Vertical Restraints: Per Se Illegality, The Rule of Reason, and Per Se Legality, 79 Antitrust L.J. 1003, 1004 (2014) (“[T]he shift in the antitrust rules applied to [vertical restraints] has not been from per se illegality to the rule of reason, but has been a more dramatic shift from per se illegality to presumptive legality under the rule of reason.”).

[107] See Commission Regulation (EU) No 330/2010 of 20 April 2010 on the Application of Article 101(3) of the Treaty on the Functioning of the European Union to Categories of Vertical Agreements and Concerted Practices, 2010 O.J. (L 102) art.4 (a).

[108] See, e.g., Case C-403/08, Football Association Premier League and Others, ECLI:EU:C:2011:631, §139. (“[A]greements which are aimed at partitioning national markets according to national borders or make the interpenetration of national markets more difficult must be regarded, in principle, as agreements whose object is to restrict competition within the meaning of Article 101(1) TFEU.”).

[109] Joined Cases-56/64 and 58/64, Consten SARL & Grundig-Verkaufs-GMBH v. Commission of the European Economic Community, ECLI:EU:C:1966:41, at 343.

[110] Patrick Rey & Jean Tirole, The Logic of Vertical Restraints, 76 Am. Econ. Rev. 921, 937 (1986) (emphasis added).

[111] These papers are collected and assessed in several literature reviews, including Lafontaine & Slade, supra note 43; O’Brien, supra note 44; Cooper et al., supra note 44; Global Antitrust Institute, Comment Letter on Federal Trade Commission’s Hearings on Competition and Consumer Protection in the 21st Century, Vertical Mergers (George Mason Law & Econ. Research Paper No. 18-27, Sep. 6, 2018). Even the reviews of such conduct that purport to be critical are only tepidly so. See, e.g., Marissa Beck & Fiona Scott Morton, Evaluating the Evidence on Vertical Mergers 59 Rev. Indus. Org. 273 (2021) (“[M]any vertical mergers are harmless or procompetitive, but that is a far weaker statement than presuming every or even most vertical mergers benefit competition regardless of market structure.”).

[112] Leegin, 551 U.S. at 889.

[113] Id. at 902.

[114] See, e.g., Lafontaine & Slade, supra note 43.

[115] See Leegin, 551 U.S. at 886-87 (holding that the per se rule should be applied “only after courts have had considerable experience with the type of restraint at issue” and “only if courts can predict with confidence that [the restraint] would be invalidated in all or almost all instances under the rule of reason” because it “‘lack[s]… any redeeming virtue’”) (citations omitted).

[116] See Bruce Kobayashi, The Economics of Loyalty Rebates and Antitrust Law in the United States, 1 Comp. Pol’y Int’l 115, 147 (2005).

[117] See, e.g., Case C-85/76, Hoffmann-La Roche & Co. AG v Commission of the European Communities, EU:C:1979:36, at 7.

[118] See Intel, supra note 39, at ¶ 139 (emphasis added).

[119] Opinion of AG Wahl in Case C-413/14 P Intel v Commission, ECLI:EU:C:2016:788, para 73.

[120] Expert Report, supra note 1, at 17.

[121] Id.

[122] See, e.g., Thomas A. Lambert, Defining Unreasonably Exclusionary Conduct: The Exclusion of a Competitive Rival Approach, 92 N.C. L. Rev. 1175, 1175 (2014) (“This Article examines the proposed definitions or tests for identifying unreasonably exclusionary conduct (including the non-universalist approach) and, finding each lacking, suggests an alternative definition.”).

[123] Id.

[124] Id. at 1244 (“Drawing lessons from past, unsuccessful attempts to define unreasonably exclusionary conduct, this Article has set forth a definition that identifies a common thread tying together all instances of unreasonable exclusion, comports with widely accepted intuitions about what constitutes improper competitive conduct, and generates specific safe harbors and liability rules that would collectively minimize the sum of antitrust’s decision and error costs.”).

[125] Expert Report, supra note 1, at 17.

[126] Ohio v. Am. Express Co., 138 S. Ct. 2274, 2286-87 (2018).

[127] Gregory J. Werden, Why (Ever) Define Markets? An Answer to Professor Kaplow, 78 Antitrust L.J. 729, 741 (2013).

[128] Geoffrey A. Manne, In Defence of the Supreme Court’s ‘Single Market’ Definition in Ohio v. American Express, 7 J. Antitrust Enforcement 104, 106 (2019).

[129] David S. Evans, The Antitrust Economics of Multi-Sided Platform Markets, 20 Yale J. Reg. 325, 355-56 (2003). See also Jean-Charles Rochet & Jean Tirole, Platform Competition in Two-Sided Markets, 1 J. Eur. Econ. Ass’n 990, 1018 (2003).

[130] See, e.g., Michal S. Gal & Daniel L. Rubinfeld, The Hidden Cost of Free Goods, 80 Antitrust L.J. 521, 557 (2016) (discussing the problematic French Competition Tribunal decision in Bottin Cartographes v. Google Inc., where “[d]isregarding the product’s two-sided market, and its cross-network effects, the court possibly prevented a welfare-increasing business strategy”).

[131] See, e.g., Brief of Amici Curiae Prof. David S Evans and Prof. Richard Schmalensee in Support of Respondents in Ohio, et al. v. American Express Co., No. 16-1454 (Sup. Ct. Jan. 23, 2018) at 21, available at (“The first stage of the rule of reason analysis involves determining whether the conduct is anticompetitive. The economic literature on two-sided platforms shows that there is no basis for presuming one could, as a general matter, know the answer to that question without considering both sides of the platform.”).

[132] United States, et al. v. Am. Express Co., et al., 838 F.3d 179, 198 (2nd Cir. 2016).

[133] See, e.g., Michael Katz & Jonathan Sallet, Multisided Platforms and Antitrust Enforcement, 127 Yale L.J. 2142, 2161 (2018) (“[I]t is essential to account for any significant feedback effects and possible changes in prices on both sides of a platform when assessing whether a particular firm has substantial market power.”).

[134] California earned 10% of its statewide GDP from the tech industry in 2021, and just over 9% in 2022. See SAGDP2N Gross Domestic Product (GDP) by State, Bureau of Economic Analysis (last visited May 1, 2024),

[135] See Joseph Politano, California Is Losing Tech Jobs, Apricitas Economics (Apr. 14, 2024), (“[California’s] GDP fell 2.1% through 2022, the second-biggest drop of any state over that period, driven by a massive deceleration across the information sector. That allowed states like Texas to overtake California in the post-pandemic GDP recovery, creating a gap that California still hasn’t been able to close despite its economic rebound in 2023.”).

[136] See id. (“[T]he Golden State has been bleeding tech jobs over the last year and a half—since August 2022, California has lost 21k jobs in computer systems design & related, 15k in streaming & social networks, 11k in software publishing, and 7k in web search & related—while gaining less than 1k in computing infrastructure & data processing. Since the beginning of COVID, California has added a sum total of only 6k jobs in the tech industry—compared to roughly 570k across the rest of the United States.”).

Continue reading
Antitrust & Consumer Protection

Labor Monopsony and Antitrust Enforcement: A Cautionary Tale

ICLE White Paper Executive Summary In recent years, there has been growing interest among economists, lawyers, and policymakers in the concept of monopsony power, particularly in labor markets. . . .

Executive Summary

In recent years, there has been growing interest among economists, lawyers, and policymakers in the concept of monopsony power, particularly in labor markets. This interest has been spurred partially by academic research suggesting that labor-market concentration may be more prevalent than previously thought, as well as policy developments signaling a more aggressive approach by antitrust authorities to labor-monopsony issues. Despite this momentum, however, significant empirical and conceptual challenges remain in the use of antitrust law to address labor monopsony.

A. Economics Challenges

On the empirical front, the evidence on the extent and impact of labor monopsony is mixed. While some studies have found evidence of labor-market concentration and its effects on wages, these studies often rely on indirect measures that have limited applicability to antitrust cases. More direct estimates of monopsony power are rare, and often rely on stylized economic models that may not capture the complexities of real-world labor markets. Moreover, the economics literature has not reached a clear consensus on the appropriate framework to assess labor-market power in antitrust contexts.

Conceptually, there are important differences between monopoly and monopsony that complicate the application of traditional antitrust tools and standards to labor markets. One key difference is that monopsony and monopoly markets do not sit at the same place in the supply chain. This matters because all supply chains end with final consumers, and antitrust policy must grapple with how to balance effects at different levels of the distribution chain. In evaluating monopsony, authorities must consider the “pass through” to final product markets, a complication that does not arise in the mirror-image case of monopoly.

Another conceptual challenge is how to handle merger efficiencies in labor-market cases. In input markets, traditional efficiencies and increased buyer power are often two sides of the same coin, presenting difficult tradeoffs for authorities. Additionally, market definition—a cornerstone of modern antitrust policy—becomes more complex in labor markets, where the boundaries between different occupations, industries, and geographic areas can be blurry.

B. Policymakers’ Response

Despite these challenges, antitrust authorities have recently signaled a more aggressive approach to labor-monopsony issues. The Federal Trade Commission’s (FTC) noncompete ban, challenge to the Kroger/Albertsons merger, and the 2023 Merger Guidelines’ discussion of labor-market effects are all prominent examples of this trend. But these enforcement actions and policy statements often gloss over the unsettled state of the economics literature and the legal difficulties of proving labor-market harms under existing antitrust standards.

For example, the 2023 Merger Guidelines assert that labor markets have unique features that may exacerbate the competitive effects of mergers, but do not fully grapple with the limitations of the economic models and empirical evidence underlying these claims. Similarly, while the FTC’s Kroger/Albertsons complaint advances a novel “union grocery labor” market definition, it is unclear whether this approach aligns with economic realities or legal precedent.

C. Legal Difficulties

More broadly, it remains uncertain whether demonstrating and remedying monopsony power is feasible under existing legal standards. While harms to workers can theoretically be cognizable under the antitrust laws, proving such harms is challenging, especially under the prevailing consumer-welfare standard. Recent criminal cases targeting wage fixing and no-poach agreements have faced difficulties, and civil cases require showing harm to downstream consumers, not just workers.

Addressing these issues may require rethinking the goals and methods of antitrust enforcement. The consumer-welfare standard becomes difficult to apply when a merger may harm workers but benefit consumers downstream. Weighing these cross-market effects raises unresolved questions about the proper balance between consumer and producer surplus. While the 2023 Merger Guidelines assert that harms to upstream competition cannot be offset by benefits to downstream consumers, the basis for this stance in case law is questionable.

There are also important differences between monopoly and monopsony that complicate the mirror-image application of antitrust tools to labor markets. Most fundamentally, authorities must grapple with how to balance effects at different levels of the supply chain—an issue that does not arise in the standard monopoly context.

Moreover, the unique features of labor markets—such as the importance of firm-specific investments in human capital—pose challenges for market definition and the assessment of competitive effects. Traditional concentration measures and econometric tools used in product markets may not readily translate to the labor context. And the potential for countervailing effects on workers and consumers creates difficult tradeoffs in merger review.

Given these complexities, this paper urges caution and further study before radically expanding labor-antitrust enforcement. Advocates of reform should engage seriously with the empirical and conceptual issues highlighted here, rather than assuming that current law and economics support their policy prescriptions. Courts and enforcers should carefully consider the limitations of existing approaches and develop more robust analytical frameworks suited to the realities of labor markets.

D. The Road to Antitrust Enforcement in Labor Markets

This does not mean that antitrust has no role to play in addressing labor-market power. But it does counsel against a rush to condemn mergers and practices based on simplistic models or tenuous evidence. A more gradual, case-by-case approach focused on building legal precedent and economic consensus may be warranted. In the meantime, further dialogue between labor economists, antitrust experts, and policymakers is essential to aligning theory, evidence, and doctrine.

Such an agenda might include:

  • Developing more direct, antitrust-relevant measures of labor-market power beyond concentration ratios.
  • Studying the effects of specific mergers and practices on labor-market outcomes, rather than simply correlating concentration with wages.
  • Refining models of dynamic competition and firm-specific investments in labor markets and considering their implications for antitrust enforcement.
  • Clarifying the goals of antitrust in labor markets and how to weigh effects on different stakeholders under the consumer-welfare standard (or alternative frameworks).

The paper concludes by noting that, while the road ahead is challenging, the growing interest in labor antitrust presents an opportunity for interdisciplinary research and policy innovation. By carefully building on existing knowledge and legal frameworks, academics and practitioners can help craft an antitrust regime that promotes competition and welfare in labor markets without unduly chilling procompetitive conduct. The key is to remain grounded in sound economics and committed to empirical rigor, while adapting to the unique features of labor markets. With such an approach, antitrust can play a valuable role in ensuring that workers share in the benefits of a well-functioning economy.

I. Introduction

Market power—traditionally discussed in terms of monopoly power on the sell side—has faced increasing scrutiny from the buy-side perspective. This is especially true regarding labor monopsony, where employers may exert undue control over employees, thereby influencing wages and working conditions. This shift in focus reflects a growing concern among economists, lawyers, and policymakers about the implications of such power dynamics in the labor market. The growing discourse around monopsony power in labor markets has been further marked by a keen interest in applying antitrust laws to combat these concerns.

Recent policy initiatives and enforcement decisions indicate a burgeoning will to leverage antitrust law against perceived labor-market power abuses. In the first half of 2024 alone, the Federal Trade Commission (FTC) has enacted a rule banning noncompete agreements for nearly all workers in the United States, justified on grounds that such agreements amount to “unfair methods of competition.”[1] The FTC has also brought an enforcement action challenging the proposed Kroger/Albertsons merger, in part predicated on concerns about the combination’s potential to diminish labor competition and exacerbate monopsony power in local labor markets.[2] At year-end 2023, meanwhile, the FTC and the U.S. Justice Department (DOJ) Antitrust Division published updated merger guidelines that, for the first time, included an expanded discussion of monopsony issues.[3] While the noncompete ban, the Kroger/Albertsons merger challenge, and the 2023 Merger Guidelines are the most prominent examples, they are far from the only ones.[4]

This paper argues that, despite growing interest in the use of antitrust law to address labor monopsony, such efforts are not supported by empirical and theoretical foundations sufficient to bear the weight of these galvanized efforts. While policy proceeds apace, the debate is far from settled on the economic evidence, analytical tools, and legal standards appropriate for understanding and addressing monopsony power in labor markets as an antitrust concern. In fact, the current state of economic research and antitrust jurisprudence raises more questions than answers about the appropriate framework for assessing labor-market power.

Examples of this disconnect are legion. Empirical data concerning the magnitude and impact of labor monopsonies is inconsistent. Evidence on the extent of labor-market power is mixed, with studies reaching divergent conclusions depending on the data, methodology, and markets analyzed. While the Biden administration has been quick to cite economic research on labor-market concentration and earnings as motivating factors,[5] the referenced studies provide only indirect evidence of monopsony power and have limited applicability to antitrust cases, while direct estimates of monopsony power are rare and often rely on economic models that have not yet been accepted within antitrust. A more complete analysis of the literature on concentration in labor markets, meanwhile, does not support the narrative that labor markets are extremely concentrated across wide swathes of the economy. From a theoretical standpoint, the economics literature has not reached a clear consensus on the appropriate antitrust framework for labor markets. Moreover, the distinct economics of monopsony contrast with those of monopoly, introducing unresolved complexities into customary modes of antitrust analysis, such as market definition, assessment of efficiencies, and the consumer-welfare standard.

The antitrust authorities have ignored these complications in their recent actions. For example, Guideline 10 of the 2023 Merger Guidelines states that labor markets frequently have unique characteristics that may exacerbate the competitive effects of mergers:

[L]abor markets often exhibit high switching costs and search frictions due to the process of finding, applying, interviewing for, and acclimating to a new job. Switching costs can also arise from investments specific to a type of job or a particular geographic location. Moreover, the individual needs of workers may limit the geographical and work scope of the jobs that are competitive substitutes.[6]

This implies that market attributes like switching costs, search costs, and transportation costs are unique to labor markets. Of course, this is not true. Nor is there any reason to think labor markets are even relatively more susceptible to such costs. At the same time, the guidelines’ statement implies that these labor-market costs are borne only by workers, rather than employers. But there is no reason why that should be the case. Indeed, switching costs do not always make markets less competitive.[7]

The guidelines further assert that relevant labor markets “can be relatively narrow,” and that “the level of concentration at which competition concerns arise may be lower in labor markets than in product markets, given the unique features of certain labor markets.”[8] Because these are the merger guidelines and are meant to cover a wide variety of situations, one could read “may” as implying something more than a possibility. Indeed, the guidelines clearly appear to indicate that, following mergers, anticompetitive effects are more of a concern in labor markets than in product markets.

Unfortunately, the models commonly employed in labor economics to support these claims rely on assumptions about worker mobility, employer conduct, and market structure that likely oversimplify real-world dynamics. All models are simplifications, but how important are those simplifications for antitrust? The economic models commonly used to study labor markets have not been subjected to the same level of antitrust scrutiny as those employed in industrial-organization (IO) economics to analyze product markets. Over the past several decades, IO models of imperfect competition have been rigorously adapted and applied to assess the competitive effects of mergers, collusive agreements, and exclusionary practices in antitrust matters. Empirical IO research has frequently focused on questions of direct relevance to antitrust enforcement, and IO economists have often played an active role in developing the analytical tools used by agencies and courts.

In contrast, most labor-economics research has been conducted without an explicit focus on antitrust policy and, until recently, labor economists were rarely involved in antitrust matters. As a result, the key assumptions and implications of labor-economics models have not been fully stress tested against the evidentiary burdens and legal standards of antitrust cases—at least, not in the same ways as their IO counterparts. This disconnect poses challenges to the effective application of labor economics to antitrust enforcement, as the models and empirical techniques most familiar to labor economists may not align well with the demands of antitrust law.

Moreover, it’s not just the economics that is more unsettled than the current administration would like to claim; the law is unsettled, too. It is unclear whether demonstrating and remedying monopsony power is feasible under existing legal standards, for example. It is true that harms to labor can be cognizable under the antitrust laws, which prohibit certain exercises of monopsony power, and not just monopoly power. There are, however, ambiguities in accurately defining the boundaries of relevant labor markets. And establishing tangible anticompetitive effects on workers as “consumers” of jobs also poses challenges.

Wage-fixing agreements are per se illegal, but the decisions in recent criminal no-poach and wage-fixing cases suggest difficulties in proving that such agreements amount to meaningful market allocation, rather than insignificant job-posting-policy changes, that would be inconsistent with a per se rule. For example, in United States v. DaVita Inc., the judge ruled that no-poach agreements could be an illegal market-allocation agreement.[9] But the jury acquitted the defendants of criminal no-poach charges, finding that the DOJ had failed to prove that the agreements at-issue were made with the purpose of allocating the market and ending meaningful competition for employees. The government has faced similar difficulties in other cases.[10]

Outside of per se cases, antitrust becomes even more complicated. Addressing labor-market power requires tradeoffs under established antitrust standards, raising unresolved questions about the goals of antitrust enforcement. As Herbert Hovenkamp notes, “it has been explicit from the start that antitrust’s concern is protection from reduced market output and, concurrently, higher prices.”[11] This focus on output and price effects in downstream product markets sits uneasily with concerns about labor market harms, which may not always manifest in higher consumer prices or reduced output in the downstream product market.

For example, the consumer-welfare standard becomes difficult to apply when a merger may harm workers, but benefit consumers downstream, as when wage reductions for workers accompany consumer benefits (such as lower prices) in downstream product and service markets. Do all mergers that reduce wages for one market of workers “substantially lessen competition” in a “line of commerce”?[12] In practice, weighing these cross-market effects raises unresolved questions about the goals of antitrust enforcement. Is the sole focus on final-product consumers, or should producer surplus also be considered? If so, how should we value and compare producer versus consumer harms?

The 2023 Merger Guidelines acknowledge these issues, but sidestep them, by asserting that:

If the merger may substantially lessen competition or tend to create a monopoly in upstream markets, that loss of competition is not offset by purported benefits in a separate downstream product market. Because the Clayton Act prohibits mergers that may substantially lessen competition or tend to create a monopoly in any line of commerce and in any section of the country, a merger’s harm to competition among buyers is not saved by benefits to competition among sellers.[13]

As we explain below, however, the issue is not so simple, and its resolution cannot be assumed simply by quoting the Clayton Act.[14]

While the guidelines propose treating labor markets similarly to product markets for analytical purposes, the Kroger/Albertsons complaint suggests that, in practice, the agency believes that labor markets should be defined more narrowly—for example, unionized workers in very narrow geographic areas.[15] This approach raises further conceptual issues in market definition, as labor markets may transcend traditional industry and geographic boundaries in complex ways. More work is needed to align labor economics with the realities of antitrust enforcement. Answering these questions may require revisiting foundational assumptions that currently guide antitrust policy. Caution is thus warranted before concluding that antitrust can or should seek to remedy monopsony, absent harm to consumers of final goods.

Therefore, while monopsony concerns are becoming more prevalent in academic and policy discussions, the agencies should be extremely hesitant as they move forward. Some have argued that “[m]ergers affecting the labor market require some rethinking of merger policy, although not any altering of its fundamentals.”[16] As we discuss below, however, while the economic “fundamentals” undergirding merger policy may not change for labor-market mergers, the “rethinking” required to properly assess such mergers entails fundamental changes that have not yet been adequately studied or addressed. As many have pointed out, there is only a scant history of merger enforcement in input markets in general, and even less in labor markets.[17] It is premature to offer guidelines or impose nationwide bans on labor practices, while purporting to synthesize past practice and the state of knowledge, when neither is well-established.

The following sections illustrate some of the significant disconnects between labor economics and antitrust enforcement, highlighting the need for further research and dialogue between the two fields. In short, while interest is growing, labor economics cannot yet be readily plugged into antitrust enforcement in the same way that IO theory and empirics have been.

II. The Contemporary Relationship Between Labor and Antitrust

As discussed in the previous section, the 2023 Merger Guidelines, Kroger/Albertsons complaint, and the FTC’s noncompete rule evidence an invigorated policy effort to address competition concerns in labor markets. The merger guidelines discuss the potential labor-market implications of mergers in multiple sections, and adopt a guideline specifically related to labor-market considerations that calls out the purportedly unique features of labor-monopsony markets “that can exacerbate the competitive effects of a merger.”[18] While the noncompete ban contains an extensive discussion of the labor-economics literature on noncompetes,[19] the sweeping nature of the ban suggests that policymakers view monopsony power as a pervasive issue affecting most workers, despite the nuances and ambiguity of the literature.[20] And the FTC’s complaint in the Kroger/Albertsons case argues that the merger would eliminate labor-market competition between Kroger and Albertsons and would increase their leverage in negotiations with local unions over wages, benefits, and working conditions in an asserted “union grocery labor” market—introducing a novel and remarkably narrow market definition and an untested, contentious theory of harm (reduction in bargaining leverage) particular to labor markets.[21]

While these efforts may signal a newly heightened attention to labor-market concerns, the antitrust focus on labor monopsony did not originate with them. In recent years, there has been growing interest in using the tools of antitrust to address labor issues, with both academic literature and enforcement actions paving the way for a more labor-centric approach to antitrust. This section provides an overview of some of the key developments in this area, illustrating the growing attention given to labor-market power by antitrust authorities and scholars.

Conceptually, the relationship between labor economics and antitrust law has also been a subject of growing academic attention in recent years. A number of law-review articles have highlighted the historical disconnect between the two fields, noting that labor markets have often been overlooked in antitrust analysis.[22] They also point, however, to some areas where labor economics has begun to make inroads into antitrust enforcement.

On the policy front, President Joe Biden explicitly called for greater scrutiny of “monopsony power” in labor markets in his 2021 executive order on competition.[23] The U.S. antitrust agencies have similarly been ramping up enforcement and other policy work at the intersection of labor and competition policy. For instance, the DOJ sued to block Penguin Random House’s acquisition of Simon & Schuster, in part based on monopsony concerns regarding the market for top-selling book authors.[24] Under the current leadership, the FTC has brought and settled several enforcement actions alleging that certain noncompete agreements violated the FTC Act’s prohibition on “unfair methods of competition.”[25] The day after announcing the first three of those settlements, the FTC first proposed a nationwide ban on the use of noncompetes via a notice of proposed rulemaking.[26]

As noted above, the DOJ has brought several recent wage-fixing cases, albeit with limited success.[27] Previously, during the Obama administration, the DOJ and FTC jointly issued antitrust guidance for human-resource professionals that warned that agreements among competing employers to fix terms of employment may violate the antitrust laws.[28] The DOJ also brought suits against major Silicon Valley employers for entering into anticompetitive “no-poach” agreements to restrict hiring of engineers and programmers from competitor firms.[29] The department alleged in those suits that the agreements amounted to unlawful allocation of the relevant labor market among horizontal competitors. The DOJ also challenged a hospital association’s members agreement to set uniform billing rates for certain nurses as an improper exertion of buyer power.[30] Although both the “no-poach” and nurse wage-setting actions ultimately settled, these cases demonstrated an increasing willingness to extend antitrust scrutiny to labor-market effects and to discipline allegedly monopsonistic practices by dominant buyers of labor.

Finally, in 2022, the FTC signed a memorandum of understanding with the National Labor Relations Board (NLRB) “regarding information sharing, cross-agency training, and outreach in areas of common regulatory interest.”[31] In 2023, the FTC signed a similar memorandum of understanding with the U.S. Labor Department.[32]

While these recent developments reflect growing interest in the application of antitrust law to labor-monopsony concerns, the linkage between labor economics and antitrust is not yet as developed as the one between antitrust law and IO and antitrust economics for output markets. Over the 20th century, the fields of IO economics and antitrust law evolved considerably. While the two fields are not co-extensive, the mutual influence has been considerable and ongoing, as strong connections have developed between economic theory, empirical study, and legal doctrine. Models of imperfect competition were incorporated into analyses of mergers, collusion, and exclusionary practices.[33] Notably, even the Chicago School, despite some scholars’ claims to the contrary,[34] made extensive use of models beyond perfect competition as a central part of its approach to antitrust.[35] Empirical IO research also frequently studied topics directly relevant to antitrust inquiries.[36] This close, co-evolutionary relationship does not yet exist—at least, not to the same extent—between labor economics and antitrust.[37]

While some scholars have worked to integrate labor and antitrust economics more closely, most empirical research remains focused on indirect concentration measures, rather than pricing conduct directly relevant to antitrust enforcement. Labor economics does not yet have IO’s established track record of successful application to assessing the competitive impact of mergers, restraints, or exclusionary practices. Before that sort of track record can be built, certain limitations must be overcome—not least that labor research has largely developed without a focus on, or involvement in, antitrust policy.

III. The Newly Developing Economic Literature on Labor-Market Power

Labor markets have become an increasingly popular topic in antitrust-policy debates. These debates have, at least in part, been spurred by academic research that purports to find widespread market power in labor markets, thus warranting the need for antitrust scrutiny.[38] For example, the U.S. Treasury Department’s report on “The State of Labor Market Competition” connects the economics research to “a description of Biden Administration actions to improve competition.”[39] Unfortunately, conclusions that the labor-market-power literature supports tougher antitrust enforcement often rely on indirect measures of market power, such as concentration figures, that are sometimes far-removed from the needs of antitrust enforcement, which usually requires more direct measures and more antitrust-relevant markets.[40]

Against this backdrop, this section reviews the scholarly evidence on labor-market power. Subsection A reviews economic papers that attempt to measure firms’ labor-market power directly, while Subsection B reviews papers that rely on such proxies as industry-concentration measures (i.e., indirect evidence of labor-market power). Ultimately, we find that these bodies of research say little about the need for tougher antitrust enforcement, largely because their measures of market power fail to indicate that there is an antitrust-relevant problem that is currently unaddressed in labor markets.

A. Direct Evidence: Do Employers Have Significant Labor-Market Power?

How do we measure labor-market power? While the bulk of the evidence on labor markets is only indirectly related to market power (if related at all), there have been a few explicit attempts to quantify the extent of labor-market power within U.S. markets.

The most popular way to directly estimate labor-market power is through the residual labor-supply elasticity that a firm faces. A labor-supply elasticity measures how responsive the supply of labor is to a change in wages. In the simplest model, a more elastic labor supply means workers have more outside options and employers have less wage-setting power. In the extreme, a perfectly competitive firm faces a perfectly elastic residual supply curve; in the baseline (two-firm) model, if one firm pays $0.01 less than the other employer, all the employees will leave for the other employer.

Outside of the perfectly competitive case, a firm may have some degree of labor-market power, which can be measured by the difference between the wage and the marginal revenue product, known as the wage “markdown.”[41] In the case of perfect competition (i.e., no market power), the firm is unable to pay wages below the marginal product of labor (the revenue generated for the firm by an additional worker), and thus the labor markdown of wages is zero. By contrast, the presence of a larger wage markdown (because of a lower labor elasticity) indicates greater labor-market power.[42]

Naidu, Posner, and Weyl summarize estimates of labor-supply elasticity from several studies, finding evidence of substantial market power in some labor markets, but by no means all.[43] Indeed, the underlying papers find residual labor elasticities ranging from 0.1 to 4.2, which would mean that workers are receiving between 9% and 81% of their marginal product, depending on the particular paper’s estimate.[44] While the list of papers estimating labor elasticity is too lengthy to detail in this paper, the upshot for antitrust policy is that low elasticity (and thus large labor-market power) is not universal (nor should we expect it to be; even if average market power is large, not every market is average).[45]

But even if the empirical labor-economics literature unanimously identified a large degree of labor-market power, which it does not, it would remain unclear what the implications are for antitrust policy. The crux of the problem is that the literature’s estimates of labor elasticities generally rely on assumptions that may not mirror those typically used in antitrust analysis. Applying these estimates to a simple antitrust model of monopsony generates implications that go against the data. For example, a labor-supply elasticity of 0.1 would imply a labor share of income of just 8% in the model described in Naidu, Posner, & Weyl.[46] That is far lower than the actual labor share observed in most countries, which has fallen, but is still closer to 60%, not 8%.[47] This suggests that the connection between the estimate and the model may not be appropriate. Thus, while labor-supply elasticities can provide valuable information about the degree of labor-market competition, antitrust practitioners should be wary of applying them mechanically to standard models of product-market competition without considering the unique features and dynamics of labor markets.

There can also be discrepancies between the tools employed to estimate labor-supply elasticities, on the one hand, and the needs of antitrust enforcement, on the other. For instance, a study by Ransom and Sims employs a search model—a standard tool in labor economics, but not a model generally seen in antitrust. The model is based on the idea of “search frictions,” which refers to the time and effort required for workers to find jobs and for employers to fill vacancies.[48] Because of these frictions, workers may accept lower-paying jobs while continuing to search for better opportunities.

This model assumes that, in the long run, the number of workers leaving a job is equal to the number of workers taking a new job. While this “steady state” assumption may hold in many contexts, it is not one typically seen in antitrust analysis of product markets. If the assumption is violated, estimates of labor-market power derived from the model could be biased in either direction, depending on the specific imbalance of worker flows. In the realm of antitrust enforcement, this could lead to both false positives and false negatives. It remains to be seen what courts would do when confronted with these new models.

Conversely, other papers attempt to apply the standard Cournot model from antitrust product-market analysis to labor markets.[49] In this approach, the authors take the median Herfindahl-Hirschman Index (HHI), a common measure of market concentration, and divide it by the aggregate labor-supply elasticity to estimate labor-market power. But there may be a mismatch here, as well. Indeed, it is unclear whether the Cournot model, where firms commit to hiring a certain number of workers each period, is a realistic representation of labor markets for antitrust purposes, because it relies on critical assumptions that may not be present in real-world markets, such as simple wage-posting, monopsony models. In fact, this may explain why search models, despite their flaws, remain the most common approach to assessing labor markets.

Recognizing these limitations, a burgeoning literature attempts to design labor-market competition models that better align with the needs and realities of antitrust analysis. But as yet, there is no silver bullet. Azar, Berry, and Marinescu, for example, combine elements of a static model of imperfect competition (commonly used in IO economics) with a labor-market model.[50] This approach aims to capture the dynamics of labor-market competition more accurately by considering the differentiation among jobs and workers’ preferences.

The authors use data on job vacancies from (a popular online job board) to estimate a model of differentiated jobs and workers’ preferences for those jobs. Because of data limitations, however, they only have information on the elasticity of vacancy demand—i.e., the intensity of responses to posted job vacancies—not on actual wages. To overcome this, they assume a simple model where employers post wages and workers choose whether to accept those offers, similar to how firms post prices in the Cournot model of product-market competition. Using this approach, the authors estimate that workers are paid 21% less than their marginal product, suggesting significant labor-market power.[51] But their model relies on the same long-run-equilibrium assumption discussed earlier, where the number of workers leaving a job equals the number of workers taking a new job.

One final approach uses wage markdowns to estimate labor-market power, but this, too, is far from perfect. Yeh, Macaluso, and Hershbein, for example, use data from the U.S. Census Bureau to estimate markdowns in the manufacturing sector.[52] They find that, on average, workers earn about 65 cents for every dollar of value they generate for their employer.[53] This would imply a significant degree of labor-market power. The researchers also find that markdowns tend to be larger for bigger companies, suggesting that these firms have more power to set wages.[54] Interestingly, they find that markdowns decreased from the late 1970s to the early 2000s, but have increased sharply over the past 20 years.[55] This recent increase in markdowns could indicate a growing problem of labor-market power.

Unfortunately, interpreting markdowns as a clear sign of labor-market power is not always straightforward, and there are reasons to be skeptical of these results. To see why, imagine two hair salons: Salon A is a basic salon that charges $20 for a haircut, while Salon B is a luxury salon that charges $40 for a haircut that the econometrician believes is the same quality. If both salons hire hairdressers who can do one haircut per hour, Salon B might pay only slightly more than Salon A—say $21 per hour—to attract hairdressers. This means that the hairdressers at Salon B are receiving a wage that is far less than the $40 value of their marginal product. Superficially, this might look like a sign of labor-market power.

But where the price difference is attributable to non-labor factors—such as the salon’s luxury branding, posh environment, and free drinks—the apparent markdown might, in fact, reflect the salon owner’s return on investment, rather than its power to set wages. This is why some economists view markdowns as a “residual”—the leftover value after accounting for other factors.[56] In the real world, we do not know whether an apparent markdown comes from labor-market power due to weak competition, or whether it is a return to something the owner contributes that the economist does not see.

In fact, some evidence suggests that a significant portion of markdowns may be just that: a return to some technology the firm has rather than labor-market power. Kirov and Traina look at markdowns in U.S. manufacturing over time and find that workers received the full value of their output in 1972, but only about half in 2014.[57] They argue that this increase in markdowns was driven largely by rapid productivity growth due to technological advancements, not by slower wage growth. The authors find that markdowns were strongly correlated with measures of information technology, management practices, and automation. This suggests that the growing gap between worker pay and productivity might be more about technological change than about employers’ bargaining power—a very different issue than the monopsony problem that antitrust law could (potentially) address.

All of this is not to say that labor-economics tools are unsuitable for antitrust policy or enforcement. Rather, it highlights the need for further research and legal precedent to establish how these tools can be effectively adapted to meet the evidentiary standards and analytical frameworks of antitrust law. While proponents of increased labor-antitrust enforcement may be eager to apply insights from labor economics to antitrust cases, it is crucial to recognize that this translation is not always straightforward and may require careful consideration of the underlying assumptions and their implications for antitrust analysis.

In short, there is a gap between existing direct evidence on labor-market power and the needs of antitrust policy and enforcement. Labor economics generally relies on models that are not germane to antitrust enforcers, while the models that are common in antitrust enforcement might not fully capture the dynamics of labor markets. Further research and dialogue between labor economists and antitrust experts is needed to develop a consistent and reliable framework to analyze labor-market power in antitrust cases. Until then, the inapt assumptions and limitations of the models presented to antitrust authorities and courts call their predictive value into question.

Ultimately, the direct evidence from labor-elasticity estimates and other measures of labor-market power remains limited in scope and varies widely across studies. While these studies provide valuable insights, they are far from conclusive, and do not yet approach the level of evidence and analysis typically relied upon in the IO literature to assess product-market competition. Courts and policymakers are likely to expect a more robust and consistent body of evidence before making significant changes to antitrust enforcement in labor markets. The disputes over direct evidence on labor-market power underscore the need for further research and highlight the challenges of applying antitrust tools to labor markets based on the current state of knowledge. Antitrust enforcers should take policy insights gleaned from labor-economics studies with a grain of salt, as they may be of limited use when informing antitrust policy decisions.

B. Indirect Evidence: Are Labor Markets ‘Relatively Narrow’?

The 2023 Merger Guidelines assert that labor markets can be “relatively narrow” and that “the level of concentration at which competition concerns arise may be lower in labor markets than in product markets, given the unique features of certain labor markets.”[58] The academic literature, however, presents a more nuanced picture that casts doubt on some of these claims. This section provides an abbreviated review of that literature. A more thorough explanation is provided in the Appendix.[59]

Given the limited direct evidence discussed in the previous section, as well as the difficulties entailed in collecting and applying it, it is not surprising that many scholars have turned to indirect measures of market power to fill the evidentiary gap. There are, however, significant issues with these indirect measures, as they often rely on concentration metrics, such as the Herfindahl-Hirschman Index (HHI), which are more readily available, but considerably less reliable than direct estimates of market power.[60]

While all indirect data sources have limitations, some are more comprehensive and reliable than others. The most comprehensive data is administrative data. While these differ on the levels of concentration, depending on how narrowly the market is defined, they consistently document falling concentration levels in local labor markets, where most job search and hiring occurs. [61] These studies have the advantage of comprehensive coverage of employers and workers, but often define labor markets based on industry codes, rather than occupations, which may not fully capture the relevant competitors for specific types of labor.

On the other hand, the administrative data concern all employer establishments.[62] The administrative data directly measure employment levels and shares, instead of being restricted to online vacancies as a proxy for employment.[63] This distinction matters, because employment shares are the natural counterpart of market shares—a cornerstone of antitrust enforcement. Concentration measures based on vacancies will be systematically higher than those based on employment, because not all firms will hire in any particular period (in addition to any other issues with the data sample). Using the most direct comparison available, the governmental microdata finds an average HHI roughly one-tenth as large as that found using vacancy data.[64]

Unfortunately, no dataset is perfect, even the administrative data. For example, many rely on employment data organized by North American Industry Classification System (NAICS) codes for market definition, which are organized by establishment, not by occupation. For example, all Wal-Mart employees at a store are labeled as NAICS 4521 (Department Stores), instead of being broken out by different occupations (Standard Occupational Classifications or “SOC”) for different vacancies.[65] That makes their results better interpreted as local industrial-concentration measures, instead of true labor-market concentration measures.

For pure concentration measures, this may not matter too much. Berger, Herkenhoff, and Mongey argue that “there is little practical difference in defining a market at the occupation-city level rather than the industry-city level as these two measures are highly correlated.”[66] But at the more granular level of antitrust enforcement, the difference between measures may be significant. In particular, many workers may be able to easily substitute between employers located in different industries. An accountant, for instance, might be just as qualified to work for a bank as for a hotel or a tech company. This cross-industry substitution is obscured by market definition undertaken at the NAICS level.

With these caveats about market definition, what does the administrative data show about concentration? Rinz uses the Longitudinal Business Database, covering nearly all private-sector employers, to estimate labor-market concentration from 1976 to 2015.[67] At the beginning and end of the time period studied, unsurprisingly, Rinz finds rural labor markets to be more concentrated than urban markets.[68] He finds that the average local HHI, defined by commuting zones and four-digit NAICS industries, decreased from 0.16 in 1976 to 0.12 in 2015, indicating a shift toward less-concentrated local markets. Local concentration fell in all population quintiles.[69]

By contrast, national HHI increased modestly over the same period, driven by large firms entering more local markets.[70] Similarly, Lipsius documents falling local concentration from 1976 to 2015, using alternative market definitions based on five-digit NAICS codes and urban areas, rather than commuting zones.[71] Despite these definitional differences, the average local HHI remains consistently low, ranging from 0.14 to 0.17 depending on the year and market definition. Berger, Herkenhoff, & Mongey further corroborate these findings with a different way of averaging HHI measures across markets.[72] They estimate an average local HHI of 0.17 for the year 2014, with even lower concentration levels when analyzing individual sectors like manufacturing and services. The average local HHI levels documented in these studies are below the 1,800 (or 0.18) threshold associated with highly concentrated markets in the 2023 Merger Guidelines.[73]

Studies using job vacancies, rather than employment data, tend to find higher market concentration, but this may partly be driven by their omission of job openings that are not published online (or at all). Indeed, the most well-cited papers on labor-market concentration use online job postings to measure concentration.[74] These studies can define labor markets more granularly, but they may not capture all employers and job openings, particularly those that are not advertised online. This focus on vacancies rather than employment may not always reflect the actual options available to workers, as not all job vacancies are advertised (online).[75]

While the 2023 Merger Guidelines suggest that labor markets warrant a lower concentration threshold for competition concerns, they do not provide a clear basis for this assertion or specify what that threshold should be. The indirect evidence from local labor-market concentration metrics does not support the notion that labor markets are inherently more problematic than product markets, from a concentration perspective. Instead, these low and falling concentration levels suggest that many local labor markets are relatively competitive and do not necessarily require a lower concentration threshold for merger analysis. While the guidelines’ recognition of labor markets’ unique features is important, this acknowledgment should be coupled with a more precise and empirically grounded approach to defining concentration thresholds.

More fundamentally, regardless of the data source used, market-definition issues remain. The variety of concentration estimates stemming from different geographic units and shifting occupational groupings demonstrates the lack of clarity around reasonable market boundaries. Worker mobility also introduces questions about appropriate geographic scope. While some labor markets may be highly concentrated, it does not follow that relevant antitrust labor markets are often relatively narrow. Establishing narrowness, in the antitrust sense, requires specific proof that additional employer options do not provide meaningful competitive discipline against potential wage reductions—something these papers do not do.

The upshot is that antitrust enforcers will need to rely on case-specific evidence, rather than broad claims of high concentration levels and narrow labor markets. Concentration measures have long been considered imperfect indicators of market power in antitrust policy and IO debates.[76] While high concentration may be suggestive of market power, it is not conclusive evidence. Many factors other than concentration can affect wages, such as differences in firm productivity, local labor-market conditions (e.g., urban vs. rural), and institutional factors like unionization rates.

Moreover, there is good evidence that employer concentration does not lead to depressed wages.[77] For example, Kirov and Traina find that rising markdowns (the gap between worker productivity and wages) are more strongly associated with technology-related factors, such as automation and managerial practices, than with employer concentration.[78] Moreover, they caution that:

These results suggest the workhorse assumptions behind some of the labor-market power literature might need reevaluation, particularly work that uses cross-sectional variation to infer trends in labor-market power. Concentration is likely an inappropriate measure of labor-market power in this case.[79]

Their critique underscores the limitations of relying heavily on concentration metrics to assess labor-market competition, especially when making claims about trends over time. As Berry, Gaynor, and Scott Morton write:

A main difficulty in [the monopsony power literature] is that most of the existing studies of monopsony and wages follow the structure-conduct-performance paradigm; that is, they argue that greater concentration of employers can be applied to labor markets and then proceed to estimate regressions of wages on measures of concentration. For the same reasons we discussed above, studies like this may provide some interesting descriptions of concentration and wages but are not ultimately informative about whether monopsony power has grown and is depressing wages.[80]

This is not to say that indirect evidence of market power is entirely without value. These studies can provide useful background information to guide antitrust policy. Moreover, antitrust law itself often relies on indirect measures of market power, such as concentration ratios and HHIs. In the case of antitrust enforcement, however, these measures are typically derived from carefully defined relevant markets. Defining the relevant market for labor is a complex task that requires considering such factors as job characteristics, worker skills, worker mobility, and geographic scope. There is currently little consensus among labor economists about the best way to define labor markets for antitrust purposes.

Ultimately, the indirect evidence from concentration metrics does not support the merger guidelines’ strong claims about ubiquitous labor-market narrowness or the need for a lower concentration threshold in merger analysis. While concentration trends are not uniform across all markets and data sources, the weight of the evidence points toward falling local concentration and increasing labor-market competition over time (if concentration is a proxy for competition). Antitrust authorities should engage with this evidence and provide a stronger empirical basis for their policy recommendations, rather than relying on unsubstantiated assumptions about the inherent narrowness of labor markets.

IV. The Problems of Addressing Labor-Market Power Under Antitrust Law

The empirical literature that attempts to measure labor-market power remains unsettled and limited, and provides, at best, only indirect evidence of economy-wide monopsony power. But even if robust measures of labor monopsony were available, applying antitrust laws to remedy monopsony power would still face conceptual hurdles. Economic theory indicates important differences between monopoly and monopsony power that complicate simple policy translation.

While antitrust statutes technically apply equally both upstream and downstream,[81] the economics of monopoly versus monopsony raise thorny theoretical issues regarding dynamic efficiency, merger efficiencies, market definition, and more that may differ between the two. Just as the empirical questions remain far from settled, the theory provides little straightforward guidance on how to address these concerns.

U.S. antitrust agencies have nevertheless long sought to reinvigorate anti-monopsony enforcement. Before concluding that labor-monopsony enforcement should be a priority for antitrust enforcers, both the evidentiary limitations and conceptual challenges warrant careful consideration by enforcers, scholars, and the courts.

On the surface, it may appear that monopsony is simply the “mirror image” of monopoly.”[82] There are, however, several important differences between monopoly and monopsony, as well as several complications that monopsony analysis raises that significantly distinguishes it from monopoly analysis. Most fundamental among these, monopsony and monopoly markets do not sit at the same place in the supply chain.[83] This matters, because all supply chains end with final consumers. Accordingly, from a policy standpoint, it is essential to decide whether antitrust ultimately seeks to maximize output and welfare at that (final) level of the distribution chain (albeit indirectly); whether intermediate levels of the distribution chain (e.g., an input market) should be analyzed in isolation; or whether effects in both must be somehow aggregated and balanced.

This has important ramifications for antitrust enforcement against monopsonies. As we explain below, competitive conditions of input markets have salient impacts on prices and output in product markets. Given this, any evaluation of monopsony must consider the “pass through” to the final product market. There is, however, no such “mirror image” complication in the consideration of final-product monopoly markets. Along similar lines, treating the assessment of mergers in input markets as the simple mirror image of product-market mergers presents important problems for how authorities address merger efficiencies, as traditional efficiencies and increased buyer power are often two sides of the same coin. Finally, it is unclear how authorities should think about market definition—a cornerstone of modern antitrust policy—in labor markets, in particular.

The upshot is that, while monopsony concerns have become more prevalent in academic and policy discussions, the agencies should be extremely hesitant as they move forward. Some have argued that “[m]ergers affecting the labor market require some rethinking of merger policy, although not any altering of its fundamentals.”[84] As we discuss below, however, while the economic “fundamentals” undergirding merger policy may not change for labor-market mergers, the “rethinking” required to properly assess such mergers does entail fundamental changes that have not yet been adequately studied or addressed. As many have pointed out, there is only a scant history of merger enforcement in input markets in general, and even less in labor markets.[85] It is premature to offer guidelines that purport to synthesize past practice and the state of knowledge, when neither is well-established.

A. Theoretical Differences Between Monopoly and Monopsony

Before getting to the practical differences of a monopoly case versus a monopsony case, consider the theoretical differences between identifying monopsony power and monopoly power.[86] Suppose, for now, that a merger either generates efficiency gains or market power, but not both. In a monopoly case, if there are efficiency gains from a merger, the quantity sold in the output market will increase. With sufficient data, the agencies will be able to see (or estimate) the efficiencies directly in the output market. Efficiency gains result in either greater output at lower unit cost, or else product-quality improvements that increase consumer demand. In contrast, if the merger simply enhances monopoly power without efficiency gains, the quantity sold will decrease, either because the merging parties raise prices or quality declines. The empirical implication of the merger is seen directly in the market in question.

The monopsony case is, however, rather more complicated. Ultimately, we can be certain of the effects of monopsony only by looking at the output market, not the input market where the monopsony power is claimed. To see this, consider again a merger that generates either efficiency gains or market (now monopsony) power. A merger that creates monopsony power will necessarily reduce the prices and quantity purchased of inputs like labor and materials. But this same effect (reduced prices and quantities for inputs) would also be observed if the merger is efficiency enhancing. If there are efficiency gains, the merged entity may purchase fewer of one or more inputs than the parties did pre-merger. For example, if the efficiency gain arises from the elimination of redundancies in a hospital merger, the hospital will buy fewer inputs, hire fewer technicians, or purchase fewer medical supplies.

We have seen there are scale efficiencies associated with a hospital merger. As work from the FTC’s Bureau of Economics explains, there can be scale efficiencies associated with “surgical procedures that exhibit a volume-outcome relationship.”[87] Typically, these are high-risk, complex procedures. “By consolidating such procedures at fewer hospitals, or by sending experienced personnel from one hospital to another, a system potentially can reap the benefits of increased scale.”[88] That is, reassignment of personnel and/or consolidation of procedures (and attendant personnel) at fewer hospitals can facilitate more efficient, and higher quality, provision of services, even as it may decrease labor demand in certain geographic markets. This may even reduce the wages of technicians or the price of medical supplies, even if the newly merged hospitals do not exercise any market power to suppress wages.[89]

Decisionmakers 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. The court can only differentiate a merger that generates monopsony power from a merger that increases productive efficiencies by looking at the output market. Once we look at the output market, as in a monopoly case, if the merger is efficiency-enhancing, there will be an increase in output-market quantity. If the merger increases monopsony power, by contrast, the firm perceives its marginal cost as higher than before the merger and will reduce output.[90]

In short, the assumption that monopsony analysis is simply the mirror image of monopoly analysis does not hold.[91] In both types of mergers—those that possibly generate monopoly and those that possibly generate monopsony—agencies and courts cannot look solely at the input market to differentiate them from efficiency-enhancing mergers; they must also look at the output market. Therefore, it is impossible to discuss monopsony power coherently without considering the output market.

This crucial conceptual difference in the theoretical understanding of monopsony versus monopoly has important implications for antitrust enforcement in labor markets. The need to look at output markets to distinguish efficiency-enhancing mergers from monopsonistic ones complicates the analysis and may require a different approach than traditional monopoly cases. Antitrust authorities and courts must carefully consider how a merger affects both output and input markets, and weigh potential efficiencies against anticompetitive effects.

This is particularly challenging under the consumer-welfare standard, which focuses on output-market effects. The potential for countervailing effects on output and input markets creates difficult tradeoffs for enforcers and courts, who must balance the interests of consumers, workers, and overall economic efficiency.

B. Monopsony and Merger Efficiencies

In real-world cases, mergers will not necessarily be either solely efficiency-enhancing or solely monopsony-generating, but a blend of the two. Any rigorous consideration of merger effects must account for both and make some tradeoff between them. It’s true that, in some cases, there will be output increases alongside labor-market increases and, in such scenarios, we can look simply at output.[92] In the standard monopsony models in economics, there is no offsetting effect; harm to sellers of inputs (workers) hurts consumers, as well.[93] This was the case in the recent successful action to block Penguin-Random House from merging with Simon & Schuster.[94] The parties agreed that, if there was harm to the authors, there would be fewer books, thereby harming consumers.[95] There was no need to think about offsetting harms. That’s the easy case.

But what about other cases where the effects are not so clearcut? The question of how guidelines should address monopsony power is inextricably tied to consideration of merger efficiencies—particularly given the point above that identifying and evaluating monopsony power will often depend on its effects in downstream markets.

This reality raises some thorny problems for monopsony-merger review that have not been well-studied to date:

Admitting the existence of efficiencies gives rise to a subsequent set of difficult questions central to which is “what counts as an efficiency?.” A good example of why the economics of this is difficult is considering the case in which a horizontal merger leads to increased bargaining power with upstream suppliers. The merger may lead to the merging parties being able to extract necessary inputs at a lower price than they otherwise would be able to. If so, does this merger enhance competition in a possible upstream market? Perhaps not. However, to the extent that the ability to obtain inputs at a lower price leads to an increase in the total output of the industry, then downstream consumers may in fact benefit. Whether the possible increase in the total surplus created by such a scenario should be regarded as off-setting any perceived loss in competition in a more narrowly defined upstream market is a question that warrants more attention than it has attracted to.[96]

With monopoly mergers, plaintiffs must show that a transaction will reduce competition, leading to an output reduction and increased consumer prices. This finding can be rebutted by demonstrating cost-saving or quality-improving efficiencies that lead to lower prices or other forms of increased consumer welfare. In evaluating such mergers, agencies and courts must weigh the upward pricing pressure from reduced competition against the downward pricing pressure associated with increased efficiencies and the potential for improved quality.

As we have explained above, this analysis becomes more complicated when a merger raises monopsony concerns. In a simple model, the monopsony merger would increase market power in the input market (e.g., labor), leading to a lower price paid for the input and a smaller quantity used of the input relative to pre-merger levels. Assuming no change in market power in the final product market, these cost savings would result in lower prices paid by consumers. Should such efficiency effects “count” in evaluating mergers alleged to lessen competition in input markets? It is surely too facile a response to assert that such efficiency effects would be “out of market” and thus irrelevant. If that were the case, the legality of a merger would turn arbitrarily on the choice of input or output market, while flatly ignoring evident and quantifiable effects in an equally affected market. No sensible approach to antitrust would countenance this arbitrariness.[97]

Some would argue these are the types of efficiencies that merger policy is meant to encourage. Others may counter that policy should encourage technological efficiencies, while discouraging efficiencies stemming from the exercise of monopsony power.

But this raises another complication: How do agencies and courts distinguish “good” efficiencies from “bad?” Is reducing the number of executives pro- or anticompetitive? Is shutting down a factory or health-care facility made redundant post-merger pro- or anticompetitive? Trying to answer these questions places agencies and courts in the position of second guessing not just the effects of business decisions, but also the intent of those decisions (to a first approximation, the observed outcomes are identical). But intent is far from dispositive in determining the competitive effects of business conduct, and it may be misleading.[98] Even worse, it can create a Catch-22 where an efficiencies defense in the product market is turned into an efficiencies offense in the input market—e.g., a hyper-efficient merged entity may outcompete rivals in the product market, possibly leading to monopsony in the input market. In ambiguous cases, this means the outcome may depend on whether it is challenged on the input or output side of the market. It even implies that overcoming a challenge by successfully identifying efficiencies in one case creates the predicate for a challenge based on effects on the other side of the market.

Hemphill and Rose argue that “harm to input markets suffices to establish an antitrust violation.”[99] But surely, this cannot be a general principle, at least not if markdowns are seen as a form of anticompetitive harm. To see why, consider a merger that has no effect on either monopoly or monopoly power; it solely improves the merging parties’ technology by removing redundancies. For example, suppose the merged firms require fewer janitors. By assumption, this merger lowers consumer prices and increases consumer and total welfare. But proponents of the Hemphill and Rose view would likely call it an antitrust violation, because it harms the input market for janitors. Fewer janitors will be hired, and janitors’ wages may fall (even though, by assumption, there is no monopsony power pushing down wages).

This likely explains why Marinescu and Hovenkamp recognize that assessing a monopsony claim requires looking at both input and output markets:

To have a chance of succeeding, an efficiency case for a merger affecting a labor market must show that post-merger reorganization will decrease the need for workers and will not lower total production. Both of these requirements are essential. A merger that decreases the need for workers may represent nothing more than an exercise of monopsony power, but in that case, ceteris paribus, it will also reduce production. By contrast, a merger that eliminates duplication can also reduce the need for workers, but production will not go down. Indeed, it should go up to the extent that the post-merger firm has lower costs.[100]

The complications only multiply once we move beyond a classical, wage-posting monopsony. For example, many labor-market models include some form of wage bargaining.[101] Labor economists believe this captures important aspects of labor markets that are not purely about wage-posting.[102] With bargaining—as compared to classical monopsony—when firms achieve more product-market power, they generate higher profits and, therefore, more potential surplus to be split between employers and employees.[103] Workers (at least those who keep their jobs), may welcome greater monopoly power, as they are able to extract higher wage rents, which would not be the case for a firm earning thin or no margins in an extremely competitive product market. Consequently, this generates the opposite implication at the firm level: more product market power puts upward, not downward, pressure on wages. Yet, presumably, no one would argue that courts should allow mergers simply because they raises wages. But then the reverse should also be true: courts should not block mergers simply because they lower wages.

Far from being a theoretical curiosity, bargaining is of first-order importance when we are thinking about unions and labor markets. In its Kroger/Albertsons complaint, for example, the FTC defines the relevant labor market as “union grocery labor” and alleges that the merger would harm competition specifically for these workers.[104] But 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.[105] If there is no increase in monopsony power, but there is an increase in monopoly power, the union will bargain to split that profit and increase wages.

How likely is this outcome? One local union endorsed the merger and divestiture package, arguing that “[e]mployees of Kroger and C&S will be better off than employees of other potential buyers.”[106] Of course, it is possible that most unions do not believe wages will increase; after all, delegates of the UFCW unanimously voted to oppose the merger.[107] And yet, rather than citing concern over monopsony power or lower wages, the union delegates’ stated reason for their opposition was lack of transparency.[108] The point is not to draw a conclusion about this particular merger’s likely effects on wages; it is to point out the complex tradeoffs inherent in applying antitrust to labor markets.

And there are further complications. When dynamic effects are taken into account, for example, even apparent harms confined to the seller side of an input market may turn into benefits:

[T]he presence of larger buyers can make it more profitable for a supplier to reduce marginal cost (or, likewise, to increase quality). This result stands in stark contrast to an often expressed view whereby the exercise of buyer power would stifle suppliers’ investment incentives. In a model with bilateral negotiations, a supplier can extract more of the profits from an investment if it faces more powerful buyers, though the supplier’s total profits decline. Furthermore, the presence of more powerful buyers creates additional incentives to lower marginal cost as this reduces the value of buyers’ alternative supply options.[109]

Of course, none of this is to say that creation of monopsony power should categorically be excluded from the scope of antitrust enforcement. But it is quite apparent that this sort of enforcement raises complicated tradeoffs that are elided or underappreciated in the current discourse, and manifestly underexplored in the law.[110]

C. Determining the Relevant Market for Labor

Even in the most basic monopoly cases, agencies and courts face enormous challenges in accurately identifying relevant markets. These challenges are multiplied in input markets—especially labor markets—in which monopsony is alleged. Many inputs are highly substitutable across a wide range of industries, firms, and geographies. For example, changes in technology—such as the development of PEX tubing and quick-connect fittings—allow laborers and carpenters to perform work previously done exclusively by plumbers. Technological changes have also expanded the relevant market in skilled labor. Remote work during the COVID-19 pandemic, for example, demonstrates that many skilled workers are not bound by geography and compete in national—if not international—labor markets.

When Whole Foods attempted to acquire Wild Oats, the FTC defined (and the court accepted) the relevant market as “premium natural and organic supermarkets,” as a way to exclude larger firms, such as Walmart and Kroger, from the relevant product market.[111] But even if one were to accept the FTC’s product-market definition, it is unlikely that anyone would consider employment at a “premium natural and organic supermarket” as a distinct input market.[112] Even the narrowest industries considered in the economics literature would never be defined that narrowly. This is because the skillset required to work at Whole Foods overlaps considerably with the skillset demanded by myriad other retailers and other employers, and virtually completely overlaps with the skillset needed to work at Kroger or another grocer.

As noted above, the FTC’s complaint in Kroger/Albertsons defines the relevant labor market as “union grocery labor” in “local CBA areas” (i.e., the geographic areas covered by each collective-bargaining agreement’s jurisdiction).[113] While the alleged product-market definition aligns with the FTC’s approach in past supermarket mergers, the labor-market definition is novel and does not appear to have a direct precedent in prior cases.[114] By focusing on unionized workers in specific localized areas, the FTC is implicitly arguing that the merger’s potential anticompetitive effects on labor are limited to these narrow categories of workers.

This approach to labor-market definition diverges from much of the economic literature on labor monopsony, which often defines markets based on industry or occupation codes that may not capture the full scope of competition for workers.[115] The FTC’s narrow market definition may reflect the practical challenges of bringing a labor-monopsony case under existing antitrust frameworks. But it also risks overlooking the fluid and dynamic nature of labor markets, where workers may have employment options across different industries, occupations, and geographies.[116]

We can see the difficulty with pursuing a labor-monopsony case by recognizing 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,[117] the labor market for most employees is much larger than the two merging companies. This fact alone likely renders the DOJ’s successful challenge in that case more of an aberration than a model for future labor-market enforcement actions, as is sometimes claimed.[118]

Indeed, the relevant market often cannot be narrowed down to even a handful of readily identifiable companies. For the vast majority of workers, a great number of potential employers would remain following a merger. This “potential competition”—the range of feasible employers that present an outside option to the merged companies’ present employees—limits the merged firm’s ability to exercise monopsony power in its labor negotiations. While we are not aware of publicly available data that would more comprehensively illustrate worker flows among different companies (and industries), such flows of retail workers into and out of roughly adjacent labor markets make 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 five 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.[119]

In any particular merger—such as between Kroger and Albertsons, for example—an overwhelming majority of Kroger workers’ next best option (i.e., what they would do if a store closed) will not be at an Albertsons store, but something completely outside of the market for grocery-store labor (or even outside the retail-food industry more broadly). Where that is the case, the merger would not take away those workers’ next best option, and the merger cannot be said to increase labor-monopsony power to the extent necessary to justify blocking it.[120]

Fundamentally, the labor-economics literature has offered little guidance to date on how to define markets in labor cases. As explained above, concentration varies greatly, depending on the exact definition of the relevant market, especially the geographic market.[121] It is virtually impossible to know what outside options to include in the relevant market, and it may not always be possible to identify even where such potential employers are located (e.g., are commuting zones, for example, better proxies for the relevant geographic labor market than metropolitan areas?). These market-definition issues are far more acute in monopsony cases than in traditional monopoly cases, both because the intrinsic question of substitutes is more complicated and because there is far less precedent to guide parties and enforcers.

D. Labor Markets Are Not Spot Markets

The merger guidelines stress that labor markets are not simple spot markets where each side calls out a price and the two make an exchange when bid/ask prices align. As the guidelines state, “labor markets often exhibit high switching costs and search frictions due to the process of finding, applying, interviewing for, and acclimating to a new job.”[122] Moreover, “finding a job requires the worker and the employer to agree to the match. Even within a given salary and skill range, employers often have specific demands for the experience, skills, availability, and other attributes they desire in their employees.”[123]

The typical employment contract is often more complicated than the typical end-user purchase agreement. Employment contracts are, indeed, not spot contracts, and thus contain a temporal dimension often absent from the product markets at-issue in monopoly cases. The terms of employment contracts are also rarely purely monetary, and the value of any given employment contract (and especially of aggregated “employment data”) may not be reflected in the nominal “price” (i.e., wage) of the agreement. Various benefits, deferred compensation, location, start date, moving costs and the like can dramatically complicate identifying the value of employment contracts. Complicating matters further is that the value of these terms to any given employee may vary widely, as people’s preferences for employment terms are significantly idiosyncratic. All of which makes the analysis of observable employment terms inordinately complicated and assessments of market power fraught with error.

There are, however, additional relevant aspects of labor markets that distinguish them from spot markets and that warrant consideration in antitrust analysis. One crucial factor is that employment relationships frequently involve mutual investments by both parties that develop over time. Employers often make substantial investments to build workers’ firm-specific skills through training, knowledge sharing, and opportunities to form client relationships.[124] Some of these skills are general and portable across firms, while others are firm specific and have limited value to other employers.

Firm-specific investments can increase workers’ productivity at their current firms, but also make it more costly for them to switch jobs, potentially giving employers some labor-market power. This “lock in” effect exists because the worker’s current role is more valuable due to firm-specific investments and, in some cases, this increased value cannot be ported to a new employer.

In other cases, however, employers can and do invest in training that provides workers with general—and thus transferable—skills.[125] In such examples, there is a risk that those workers will leave for a competitor before the employer can fully recoup its investment. A higher wage may be justified for a subsequent employer, as the employee comes with the added value provided by the former employer (e.g., training, knowledge of competitively valuable information, relationships with potential customers). This “holdup” problem can lead firms to underinvest in worker training, even when such training would be socially beneficial.

To mitigate this risk, firms and workers may seek contractual solutions that incentivize workers to stay long enough for the firm to earn a return on its investment. These arrangements could include promises of future wage increases, promotions, or other benefits that are contingent on the worker remaining with the firm. In turn, these contractual mechanisms create a new problem: once the investment is made and the worker has acquired valuable skills, they may be “locked in” to their current employer through the promise (implicit or explicit) of future wage gains or other benefits.

Of course, to the extent these arrangements give firms some ex-post market power, they are accompanied by terms implicitly or explicitly sharing the benefits with employees. But if a merger enhances employers’ ability to make such productivity-enhancing investments, it could simultaneously increase labor-market power while generating efficiencies, which may be shared with employees in ways that are difficult to identify or to value. Assessing the competitive effects of such a merger requires identifying and weighing these competing effects, which may be extremely difficult.

The FTC’s complaint against the proposed Kroger/Albertsons merger provides a concrete example of how antitrust enforcers must grapple with these issues in practice.[126] In defining the relevant labor markets, the FTC focuses on “union grocery labor” in “local CBA areas” (i.e., the geographic areas covered by each collective-bargaining agreement’s jurisdiction).[127] By narrowing the market to unionized workers covered by specific CBAs, the FTC appears to be making a form of lock-in argument. The complaint alleges that “[u]nion grocery workers can move between grocery employers covered by their union while retaining their pension and healthcare benefits, as well as other valuable workplace benefits and protections provided by the CBAs. If a union grocery worker leaves for a non-union employer, however, the worker will lose any non-vested CBA benefits and protections.”[128] In other words, the CBA-specific benefits function similarly to firm-specific investments in tying workers to a particular set of employers, or a contractual solution to the holdup problem involving promised future benefits, potentially giving those employers monopsony power.

From an antitrust perspective, assessing such a merger’s effect on firm-specific investments is complex. Will the merger increase or decrease employers’ incentive to provide worker training? How should antitrust balance potential productivity gains against increased labor-market power over workers? Efficiency arguments by merging parties should be met with appropriate skepticism, but such investments may be more than a rounding error in calculating overall effects. Indeed, the concept of firms investing in building worker skills is more than just a theoretical curiosity; there is clear empirical evidence that these investments occur, affect human capital, and have effects on wages.[129] These dynamic investment effects are first-order factors in labor markets, but are not easily captured in a static monopsony framework. Further study on these tradeoffs within merger analysis is essential.

The complications caused by the importance of investment in workers show up in antitrust contexts beyond merger enforcement, such as the FTC’s noncompete rulemaking.[130] The FTC recognized as much, noting that “[t]here is some empirical evidence that non-competes increase investment in human capital of workers, capital investment, and R&D investment,”[131] and citing numerous studies indicating such effects.[132] Of course, the commission nevertheless adopted a rule banning all noncompete agreements outright, despite this recognition.

All of this makes the simple monopsony model difficult to apply and map to the actual competition that occurs in the market. For example, to estimate labor-supply elasticities, many papers take a traditional monopsony model that assumes a spot market where the buyer sets a price and lets as many people buy as are willing.[133] Such analysis can be informative, but it may say little about the competitive effects of various practices in real-world antitrust markets.

The point is not to establish the proper model of human-capital formation. Instead, it is simply to point out that human-capital development is of first-order importance in labor markets. How should antitrust treat it? Contrary to the impression from the merger guidelines (and the short shrift given this point in the proposed NCA rules), not every feature of the labor market simply points toward a need for more enforcement.

V. Monopsony and the Consumer-Welfare Standard

As discussed in the previous sections, using antitrust enforcement to thwart potential monopsony harms is a task full of evidentiary difficulties and complex, poorly understood tradeoffs. Perhaps more problematically, it is also unclear whether (and, if so, how) such an endeavor is consistent with the consumer-welfare standard—the lodestar of antitrust enforcement, at least as it is currently understood and implemented by courts.[134]

Marinescu & Hovenkamp assert that:

Properly defined, the consumer welfare standard applies in exactly the same way to monopsony. Its goal is high output, which comes from the elimination of monopoly power in the purchasing market.… [W]hen consumer welfare is properly defined as targeting monopolistic restrictions on output, it is well suited to address anticompetitive consequences on both the selling and the buying side of markets, and those that affect labor as well as the ones that affect products. In cases where output does not decrease, the anticompetitive harm to trading partners can also be invoked.[135]

And Hemphill & Rose state that:

Overall, then, a trading partner welfare approach accords well with the case law and economic reasoning, and under this approach, a merger that results in increased classical monopsony power may be condemned on account of harm to the input market.[136]

But this is far from self-evident. There are at least two problems with this reasoning.

To start, the assertion that harm to input providers that does not result in reduced product output is actionable is based on a tenuous assertion that a mere pecuniary transfer is sufficient to establish anticompetitive harm.[137] This is problematic, because such “harms” actually benefit consumers in the baseline model. In the extreme example, all of the benefits of a better negotiating position are passed on to consumers, and the firm is more of a direct intermediary trading on behalf of consumers, rather than a monopolistic reseller.[138]

The main justification for ignoring these cross-market effects (as with all market-definition exercises) is primarily a pragmatic one (although it is rather weakened in light of modern analytical methods).[139] But particularly in the context of inputs to a specific output market, these cross-market effects are inextricably linked and hardly beyond calculation. As the enforcement agencies have previously recognized, “[i]nextricably linked out-of-market efficiencies, however, can cause the Agencies, in their discretion, not to challenge mergers that would be challenged absent the efficiencies.”[140]

The assertion that pecuniary transfers of bargaining power are actionable is also inconsistent with the fundamental basis for antitrust enforcement, which seeks to mitigate deadweight loss, but not mere pecuniary transfers that do not result in anticompetitive effects.[141]

Second, it is unclear whether the consumer-welfare standard applies to input markets. At its heart, the consumer-welfare standard focuses on the effects that a(n) (incipient) monopolist’s behavior may have on consumers. And courts have extended this welfare calculation to all direct purchasers affected by anticompetitive behavior. Less clear is whether courts have consistently extended (or would extend) this notion of anticompetitive harm to all “trading partners” in input markets.[142] This goes to the very heart of the consumer-welfare standard:

[I]f only consumers matter, then a buying cartel should be perfectly legal and indeed should be encouraged. Monopsony power would not matter in antitrust cases, because the fact that sellers are harmed is irrelevant under a consumer surplus standard. I know of no proponent of the consumer surplus standard who endorses buyer cartels, or who believes that monopsony is not harmful. Instead, proponents of a consumer surplus rule tend to argue that buyer cartels and monopsony are exceptions to the otherwise sensible rule of maximizing consumer surplus. However, the need for these exceptions illustrates the lack of a coherent logic for the consumer surplus standard.[143]

Other scholars appear too ready to accept that there is a “coherent logic” of the consumer-welfare standard that unquestionably contemplates upstream trading-partner welfare because their interests align with those of consumers:

A useful definition of “consumer welfare” is that antitrust should be driven by concerns for trading partners, including intermediate and final purchasers, and also sellers, including sellers of their labor. These all benefit from high output, high quality, competitive prices, and unrestrained innovation. Higher output and lower prices are good indicators of competitive benefit, and there is little practical difference between the way courts talk about antitrust harm and the idea of “consumer welfare.”[144]

As we explain above, however, this coincidence of interest is far from complete, and lower wages could be consistent with both efficiency and monopsony.[145] As the FTC summarized in closing the investigation of a merger between two pharmacy benefit managers, “[a]s a general matter, transactions that allow firms to reduce the costs of input products have a high likelihood of benefitting consumers, since lower costs create incentives to lower prices.”[146] “Higher output and lower prices [may be] good indicators of competitive benefit,” but it seems problematic to assume they reflect a clear benefit to workers if they result from lower wages. Indeed:

Larger buyers may also be able to reduce their purchasing costs at the expense of suppliers…. The concept of buyer power as an efficiency defence rests squarely on such a presumption. What is more, the argument also posits that the exercise of buyer power will not only have distributional consequences, but also increase welfare and consumer surplus by reducing deadweight loss. As we spell out in detail below, welfare gains may arise both at the upstream level, i.e., in the transactions between the more powerful merged firm and its suppliers, as well as at the downstream level, where the creation of buyer power may translate into increased rivalry and lower prices. The extent to which final consumers ultimately benefit is of particular importance if antitrust authorities rely more on a consumer standard when assessing mergers. If total welfare is the standard, however, distributional issues are not directly relevant and any pass-on to consumers is thus only relevant in as much as it contributes to total welfare.[147]

This raises an obvious question: can the consumer-welfare standard (and thus antitrust authorities and courts) reach a finding of anticompetitive harm if consumers (at least, in the narrow market under investigation) are ultimately charged lower prices?

Consider Judge Breyer’s Kartell opinion. As Steve Salop explains:

The famous Kartell opinion written by Judge (now Justice) Stephen Breyer provides an analysis of a buyer-side “cartel” (comprised of final consumers and their “agent” insurance provider, Blue Cross) that also is consistent with the true consumer welfare standard.… Buyer-side cartels generally are inefficient and reduce aggregate economic welfare because they reduce output below the competitive level…. However, a buyer-side cartel. comprised of final consumers generally would raise true consumer welfare (i.e., consumer surplus) because gains accrued from the lower prices would outweigh the losses from the associated output reduction, even though the conduct inherently reduces total welfare (i.e., total surplus).…

… Judge Breyer treated Blue Cross essentially as an agent for the customers it insured, rather than as an intermediary firm that purchased inputs and sold outputs as a monopolistic reseller. The court apparently assumed (perhaps wrongfully) that Blue Cross would pass on its lower input costs to its customers in the form of lower insurance premiums….

… In permitting Blue Cross to achieve and exercise monopsony power by aggregating the underlying consumer demands for medical care—i.e., permitting Blue Cross to act as the agent for final consumers—the Kartell court implicitly opted for the true consumer welfare standard. Blue Cross’s assumed monopsony conduct on behalf of its subscribers would thus lead to higher welfare for its subscribers despite reduced efficiency and lower aggregate economic welfare. Thus, this result represents a clear (if only implicit) judicial preference for the true consumer welfare standard rather than the aggregate economic welfare standard.[148]

By this logic, it seems, the relevant “consumer” welfare in antitrust analysis—including that of mergers that increase either monopoly or monopsony power—is that of the literal consumer: the final product’s end-user. But this contrasts quite sharply with the standard mode of analysis in monopsony cases as the mirror image of monopoly, in which the merging parties’ trading partner (whether upstream or downstream) is the relevant locus of welfare analysis.

Indeed, extended to other current potential cases, this mode of analysis raises a distinct problem for the agencies. Consider, for example, a hypothetical case against Kroger-Albertsons that did not mention the product market and in which the merger was alleged to increase monopsony power, but not monopoly power. Should such a challenge fail regardless of the effect on input providers because Kroger can be considered “an agent for the customers it [sells to]”? There is, as Salop seems to suggest,[149] some merit in such an approach, but it is certainly not how similar cases have been evaluated in the past.

Indeed, the rule of reason arguably contemplates some sort of balancing of effects across markets.

Critically, the balancing required by the rule of reason is neither quantitative nor precise. In California Dental Association, the Supreme Court described a court’s task as reaching a “conclusion about the principal tendency of a restriction” on competition. If a restraint suppresses competition in one market and promotes competition in a related market, the Chicago Board of Trade and Sylvania statements of the rule of reason can be read to hold that legality turns on which effect predominates in a qualitative sense.[150]

The U.S. Supreme Court’s Alston case highlights this dynamic, and in a case involving labor-market monopsony, no less. Despite the NCAA’s undisputed monopsony power in the “market for athletic services” (an upstream labor market), the Court considered its proferred procompetitive justification of preserving amateurism in college sports—an effect avowedly in the downstream, output market.[151] As the Court described the proceedings below:

The NCAA’s only remaining defense was that its rules preserve amateurism, which in turn widens consumer choice by providing a unique product—amateur college sports as distinct from professional sports. Admittedly, this asserted benefit accrues to consumers in the NCAA’s seller-side consumer market rather than to student-athletes whose compensation the NCAA fixes in its buyer-side labor market. But, the NCAA argued, the district court needed to assess its restraints in the labor market in light of their procompetitive benefits in the consumer market—and the district court agreed to do so.[152]

Tellingly, the district court’s rejection of the NCAA’s procompetitive justification turned on the lack of connection between it and the challenged conduct in the input market. “As the court put it, the evidence failed ‘to establish that the challenged compensation rules, in and of themselves, have any direct connection to consumer demand.’”[153] The plain implication is that, where restraints in one market are sufficiently connected to benefits in another market, those benefits will be considered—and may turn out to justify—the challenged restraints.[154]

There is perhaps no easy answer to the difficulty of assessing harm in upstream markets when downstream markets benefit. At first blush, excluding deadweight losses that stem from monopsony power (or, at least, forcing plaintiffs to show that downstream purchasers are also harmed) seems like legalistic reasoning largely incompatible with the welfarist ancestry of the consumer-welfare standard.[155] Indeed, the consumer-welfare standard is largely premised on the assumption that increased output is desirable, and deadweight losses are harmful to society, regardless of their second-order effects.

There is no tension here when output and labor both benefit from an action; sometimes, output reduction goes directly with labor harms.[156] But what about the cases that are not so neat? It seems odd to depart from this focus on output as the lodestar of antitrust just because a supplier, rather than a consumer, is being harmed.

Faced with what may potentially be intractable economic questions, antitrust courts have, for the sake of administrability, often decided to limit antitrust analysis to what economics generally refer to as partial-equilibrium analysis.[157] This largely explains, e.g., why only direct purchasers can claim antitrust damages.[158] Perhaps it also explains why the Court in Ohio v. American Express chose to simply ignore potential harm to cash purchasers in limiting the market in that case to the “market for credit-card transactions,” even though the district court found that Amex’s conduct would increase retail prices for cash consumers [159]

But much to some commentators’ chagrin,[160] the Court in Amex did take account of cross-market effects—in that case, by combining both sides of a two-sided market into a single market—and noted that failing to do so would lead to error.[161] While the Court limited its holding to two-sided, “simultaneous transaction” markets,[162] it is difficult to escape the realization that the logic of the holding—and the arbitrariness of considering effects on one side in isolation—would apply as well to the analysis of upstream and downstream trading partners:

Absent consideration of both sides of a platform, the analysis will arbitrarily include and exclude various sets of users and transactions, and incorrectly assess the extent and consequences of market power. Indeed, evidence of a price effect on only one side of a two-sided platform can be consistent with either neutral, anticompetitive, or procompetitive conduct. Only when output is defined to incorporate the two-sidedness of the product, and where price and quality are assessed on both sides of a sufficiently interrelated two-sided platform, is it even possible to distinguish between procompetitive and anticompetitive effects.[163]

The upshot is that, with some notable exceptions (such as the case of two-sided markets in Amex), antitrust courts have been reluctant to analyze competitive effects in adjacent markets. Alas, it is unclear where that line is appropriately drawn, or whether it has been drawn somewhat arbitrarily in the past.

What might seem like an arbitrary decision appears more reasonable, of course, when one considers the sheer complexity of the task at-hand. Economic behavior will often have second-order effects that run in an opposite direction to its first-order or “partial equilibrium” ones. A coal monopoly may cause buyers to opt for cleaner energy sources; a conservation cartel may maximize the long-term value of scarce resources.[164] Yet surely there are cases where out-of-market effects are “inextricably linked” to in-market effects, and where extending the analysis would not create insurmountable burdens. A practical approach—and one consistent with the broad scope of the rule of reason—would at least consider out-of-market effects when they are a direct and identifiable consequence of conduct challenged in a separate market.

The question is further complicated in merger cases where the Clayton Act’s “any line of commerce” language seems to limit merger analysis to a single market, and where the Court’s holding in Philadelphia National Bank clearly reiterates this apparent constraint.[165] But those legal rules do not address the economic propriety of so limiting merger analysis, and neither is predicated on the complexity of undertaking the requisite economic analysis. Indeed, whatever the merits of such an approach at the time Philadelphia National Bank was decided, both the law and the economics have moved past them:

Despite the incorporation of efficiencies analysis into modern merger evaluation, and the advances in economics that allow efficiencies to be identified and calculated more accurately than at the time of Philadelphia National Bank, antitrust doctrine in the United States still supports a regime that fails to take into account efficiencies arising outside of the relevant market. Only a handful of federal court cases since Philadelphia National Bank raise the issue of out-of-market efficiencies, and those that address the merits quickly dispatch such efficiencies as being precluded by the Supreme Court precedent. In light of the advances in the ability to identify and measure efficiency benefits, the federal courts should update antitrust doctrine to support a serious and committed treatment of out-of-market efficiencies in merger analysis.[166]

In part reflecting this change in approach, the Court in Baker Hughes held that “[t]he Supreme Court has adopted a totality-of-the-circumstances approach to the statute [Section 7], weighing a variety of factors to determine the effects of particular transactions on competition.”[167] And lower courts have been increasingly willing to consider efficiencies in evaluating the application of Section 7 to proposed mergers.[168] It is even arguable that the district court in New York v. Deutsche Telekom (reviewing the T-Mobile/Spring merger) credited out-of-market efficiencies in approving the merger.[169]

Moreover, as with virtually all legislative language, the Clayton Act’s language is not as clear as some make it out to be. The phrase “in any line of commerce” need not be interpreted to constrain the permissible zone of analysis, or to condemn effects in a single “line of commerce” regardless of its effects in another. Rather, the phrase’s most obvious meaning is to indicate that no area of commercial activity is exempted from the Clayton Act. Indeed, the use of the word “line” to refer to the indicated area rather than “market” seems clearly to indicate general categories of business that are to be included in the law’s prescriptions, rather than specific markets for identifying effects.

In other words, “it is plain that Section 7 does not limit the range of ‘lines of commerce’ that can trigger a merger’s prohibition.”[170] But it is by no means clear that Section 7 proscribes liability when a merger “lessen[s] competition” in a single market, regardless of whether it may enhance competition elsewhere in the same “line of commerce.”[171] As the Court suggested in Amex, the relevant “line of commerce” may incorporate distinct markets that need not exist on the same side of a given transaction. Indeed, modern “business ecosystem” theories suggest that conglomerate businesses with widely different “markets,” interrelated by an overarching business model that “inextricably links” them, may constitute something like a single “line of commerce,” despite the superficial distinctions between the components that comprise them.[172]

The question remains whether antitrust law has a comparative advantage in dealing with more “systemic” issues (like worker welfare, environmental effects, or even the “amateurism” offered by the NCAA in Alston), or whether other legal frameworks are better adapted. Put differently, antitrust law’s main strength might be that it is mostly a consumer-oriented body of law that focuses on a single tractable problem: the prices consumers and other direct purchasers pay for goods. If that is true, then other bodies of law (such as, e.g., labor and environmental laws) may be better suited to deal with broader harms. Indeed, in the case of each of these fields, there exists a massive regulatory apparatus specifically designed to implement government standards. Under the law as it stands, where antitrust law and a regulatory regime conflict, antitrust must give way.[173]

We do not purport to have a satisfactory answer to this complicated question. In fact, it is probably fair to say that one does not exist. Antitrust law can either depart from its welfarist underpinnings—a large loss for its economic consistency—or it can follow those principles toward difficult problems that may ultimately impair its administrability. At this juncture, it is not clear there is a compromise that would enable enforcers to thread the needle to solve this complex conundrum. And if such as solution exists, it has yet to be articulated in a convincing manner that may lead to actionable insights for enforcers or courts. But it is crucial to note that some cross-market analysis may be unavoidable under a welfarist approach if antitrust is going to continue to attempt to address potential harms in upstream markets, including labor markets.

Given all of this, the FTC and DOJ’s update of their merger guidelines to address monopsony harms, while clearly important, also appears to be premature, compared to the state of the economic literature, and potentially unactionable (or, at least, incoherent as stated) under the consumer-welfare standard. This is not to say the antitrust-policy world should simply ignore monopsony harms, but rather that more research, discussion, and case law are needed before definitive guidelines can be written. Ultimately, it may well be that legislative change is needed before any such guidelines will be enforceable before the courts.

VI. A Path Forward: An Agenda for Antitrust and Labor Markets

The previous sections have highlighted the empirical and conceptual challenges that complicate the application of antitrust law to labor monopsony. While the growing interest in this area presents opportunities for research and policy innovation, it is important to approach these issues with a mix of enthusiasm and skepticism. The current state of economic knowledge and antitrust doctrine suggests that we are not yet ready for a major expansion of enforcement in labor markets. This, however, does not mean that antitrust has no role to play or that the status quo is optimal. Rather, it suggests the need for a thoughtful and incremental approach that prioritizes the development of better analytical tools, evidence-based policymaking, and inter-disciplinary collaboration.

The recent FTC complaint against the proposed Kroger/Albertsons merger underscores the importance of the issues raised in this paper, as well as the ongoing challenges that antitrust authorities face when assessing labor-market effects in merger cases.[174] While the complaint reflects an increased focus on labor issues in merger enforcement, it also highlights the complexities of defining markets, assessing competitive effects, and weighing efficiency claims in this context. The Kroger/Albertsons case provides a real-world example of how the FTC is grappling with these issues in practice, but also raises questions about the rigor of its proposed market definitions, the sufficiency of evidence required, and the theories of harm proposed.

Perhaps most notably, although the complaint proposes two distinct markets, one on either side of the supermarket business (“union grocery labor” on the one hand, and “the retail sale of food and other grocery products,” on the other), it fails to note that both are simultaneously intrinsic to the operation of supermarkets. It also fails to offer any suggestion for how a court should respond if, for example, harm is found in one market but not the other. Of course, as noted, the complaint does not even contemplate the possibility that its alleged theory of harm in the labor market could result in procompetitive effects in the retail market.[175]

As labor-market concerns continue to arise in antitrust cases, it will be critical for the FTC and other enforcers to develop more robust analytical frameworks and evidentiary standards to support their claims, and for courts and policymakers to provide clearer guidance on how labor-market harms should be assessed under existing legal standards. While the FTC’s increased focus on labor issues is noteworthy, the Kroger/Albertsons complaint also demonstrates that the agency’s approach needs to be further refined and clarified.

One key priority should be to develop more direct, antitrust-relevant measures of labor-market power. While some recent studies have proposed measures such as labor-supply elasticity[176] and wage markdowns,[177] these tools have not been widely validated in antitrust contexts. Moreover, as discussed earlier, these measures may be sensitive to assumptions about the nature of competition.[178] Further refinement and testing of these measures, with a focus on their robustness and applicability to antitrust cases, is needed.

In addition, scholars should continue to study the effects of specific mergers and practices on labor-market outcomes, using more sophisticated research designs that can isolate causal impacts. While some recent studies have taken steps in this direction,[179] much more work is needed to build a body of evidence that can inform antitrust enforcement. In particular, studies that can disentangle the effects of labor-market concentration from other factors, such as firm-specific investments and productivity differences, would be valuable.

Scholars and policymakers should also continue to refine models of dynamic competition and firm-specific investments in labor markets, with an eye toward their implications for antitrust enforcement. As discussed earlier, standard static models of monopsony may not fully capture the complexities of labor-market competition, such as the role of search frictions, bargaining, and human-capital investments. Some recent papers have started to incorporate these features,[180] but more work is needed to develop tractable models that can guide enforcement decisions. It remains to be seen to what extent the FTC’s lock-in argument in the Kroger/Albertsons complaint will be supported with such models.[181]

Another key priority should be to clarify the goals and legal standards for antitrust enforcement in labor markets. The consumer-welfare standard, which has long guided antitrust policy, becomes difficult to apply when a merger or practice may harm workers but benefit consumers.[182] While some have argued for a “worker-welfare standard” that would prioritize the interests of workers,[183] it is not clear whether this would be consistent with the goals of antitrust law, nor how it would be reconciled with simultaneous findings of countervailing consumer effects.[184] Policymakers, courts, and scholars should continue to grapple with these normative questions and work toward developing a coherent and administrable framework for weighing labor-market effects in antitrust cases.

Finally, it is important to foster dialogue and collaboration between antitrust and labor experts to develop a shared understanding of the issues at-stake. Economists, lawyers, and policymakers approaching these issues from different perspectives must find common ground and a common language to assess concerns about labor-market power.

While these challenges are significant, there are reasons for cautious optimism. The increased attention to labor-market power from scholars, policymakers, and the public has created a unique opportunity to reexamine long-held assumptions and explore new approaches. By pursuing an agenda that emphasizes empirical rigor, legal clarity, and interdisciplinary collaboration, we can make progress toward more competitive labor markets. This will not happen overnight, just as the development of the consumer-welfare standard and the integration of antitrust with economic theory did not happen overnight. By staying focused on the ultimate goal of promoting the welfare of both workers and consumers, and being willing to adapt to new evidence and insights, we can move closer to an antitrust regime that is suited to the realities of the modern labor market.

Given that these complex tradeoffs still lack anything approaching definitive resolution in research or precedent, antitrust authorities would best serve the integrity of enforcement standards by exercising restraint. The disregard of difficult tradeoffs and the premature or overzealous application of questionable theories both risk distorting competition and innovation incentives more than protecting them. This is not an argument against addressing labor-market power entirely through uncertain means, as further co-evolution of economic and legal understanding may resolve some quandaries. It is, however, an argument that threading the needle to expand prohibitions into input markets requires a cautious, studious approach—especially when they conflict with the consumer interests that antitrust ultimately aims to safeguard.

Appendix: Detailed Discussion of Labor-Market Concentration Research and Its Implications for Antitrust

The 2023 Merger Guidelines assert that labor markets can be “relatively narrow” and that “the level of concentration at which competition concerns arise may be lower in labor markets than in product markets, given the unique features of certain labor markets.”[185] The academic literature presents a more nuanced picture, however, and casts doubt on these claims. This section provides a more thorough review of the literature discussed in Section III.B, infra.

By examining the strengths and limitations of each approach, we aim to provide a balanced assessment of what the current evidence can (and cannot) tell us about the extent of labor-market power in the U.S. economy. Our review suggests that, while some labor markets may indeed be highly concentrated, the evidence does not support a blanket characterization of labor markets as “narrow.” Antitrust authorities should carefully consider the specific contours of the relevant labor market in each case, drawing on multiple data sources and methodologies. The broad pattern does not support general presumptions that mergers systematically make already-narrow labor markets dramatically more concentrated over time. If anything, concentration data indicate that labor markets are growing more competitive.

I. Administrative Data

The narrative of rising employer dominance and increasing labor-market concentration has been challenged by recent research using comprehensive administrative data. These studies generally find that, while national labor-market concentration has been rising, local concentration levels have declined or remained stable over recent decades.

Papers leveraging datasets like the Longitudinal Business Database, which covers nearly all private-sector employers, point to falling concentration within local labor markets, such as commuting zones and urban areas. Rinz[186] and Lipsius[187] both used this data and estimated decreasing local concentration from 1976-2015, even as national measures increased. Their explanation is the entry of large firms into more local markets over time.

Autor, Patterson, and Van Reenen reinforce these findings using Economic Census data across major sectors. They estimated local-employment concentration fell from 0.35 in 1992 to 0.30 in 2017, contrary to rising national concentration.[188] This divergence was partly driven by employment shifts away from the highly concentrated manufacturing sector toward more competitive services sectors.

Focusing on just manufacturing, Benmelech, Bergman, and Kim found relatively stable average local concentration from 1978-2016 in the Longitudinal Business Database.[189] Importantly, their wage data allowed them to examine concentration’s direct earnings impact, suggesting a 3% wage decrease when moving from a low to high concentration market, or 9-14% using mergers as an instrument. This correlation, even with an instrument, should be interpreted with caution.

Modeling by Berger, Herkenhoff, and Mongey highlighted weighting concentration by payroll, rather than employment.[190] Though producing lower estimates, their approach still showed the diverging national/local trends.

While mixed, this literature consistently finds declining or stable local-labor market concentration when leveraging government-collected microdata. This casts doubt on claims of pervasive local-monopsony power and suggests national trends may be more relevant for assessing competitiveness. These findings have antitrust-policy implications regarding employer concentration and merger effects.

The papers that use administrative data find a trend that contradicts the popular narrative. They generally find a decline in local-labor market concentration, alongside a rise at the national level. Such findings suggest that employer dominance in the labor market may not be as pervasive or detrimental at the local level as it is nationally, complicating the narrative of widespread monopsony power in labor markets.

A. Rinz (2022) and Lipsius (2018)

First, let us consider papers that use administrative data, generally considered to be the best when available. Rinz uses administrative data from the Longitudinal Business Data and finds that local labor-market concentration has been declining, while national concentration has been increasing.[191] Lipsius uses the same dataset and finds the same result, but focuses on connecting labor-market concentration to changes in labor share of income.[192] Both papers have data on employment at the firm level for the years 1976-2015, so they are able to study the evolution over time. The data cover the near universe of non-farm, private establishments with employees.

The two papers use different levels of aggregation. Rinz uses four-digit NAICS for the job description and commuting zones for the location. Lipsius used 5-digit NAICS codes and urban areas, which are smaller than commuting zones but based on economic integration instead of political lines, such as counties.

Rinz assesses concentration using HHI measures. He finds that, at the national level, HHI declined roughly 40 percent from 1976 to 1983, stayed flat through the 1980s and has risen since. When divided into commuting zones, however, he finds a falling trend in concentration. The difference in trends has various explanations, but the simplest is that large firms are entering more and more labor markets. For example, when Wal-Mart enters a small town with one retail store, national concentration may rise, even though the town’s concentration falls.

Source: Rinz (2022)[193]

B. Autor, Patterson, & Van Reenen (2023)

Recent work by Autor, Patterson, and Van Reenen provides additional evidence on trends in local labor-market concentration using establishment-level data from the Economic Census.[194] Autor, et al. analyze six broad sectors—manufacturing, retail trade, wholesale trade, services, utilities/transportation, and finance—that comprise roughly 80% of U.S. employment and GDP. The authors have data covering the period from 1982-2017 for manufacturing, retail, wholesale, and services, and going back to 1992 for the others. They define markets by county and by six-digit NAICS industry, and find that employment-based HHI fell from 0.35 in 1992 to 0.30 in 2017.[195] Similar results hold for three- and four-digit NAICS.[196] This contrasts with the rise in national employment concentration over the same period, which rose by 1.7 points for employment (from 0.025 in 1992 to 0.042 in 2017).[197] The authors also show substantial divergence between national and local concentration trends over the longer 1982 to 2017 period for the four sectors with available data. Moreover, the local-employment HHI exhibits a consistent downward trend over most five-year intervals between 1992 and 2017. Overall, the results point to a robust fall in local employment concentration that runs counter to the rise in national concentration.

Some of this trend is structural. A key element of Autor et al.’s analysis is distinguishing between changes occurring within industries, versus those across industries. The divergence between national and local employment-concentration trends is largely attributable to the reallocation of economic activity from more-concentrated manufacturing industries to less-concentrated service industries. In fact, the authors show that, holding industry structure fixed at 1992 levels, local employment concentration would have risen by about 9%, rather than falling by 5%.[198] This between-industry reallocation had a smaller dampening effect on sales concentration, since the shift from manufacturing to services was greater for employment than sales. At the same time, Autor et al. find that concentration has risen within detailed industries and localities for both employment and sales.

C. Benmelech, Bergman, & Kim (2022)

Diving into manufacturing, specifically. Benmelech, Bergman, and Kim uses administrative, micro-level data on manufacturing establishments (“plants”), covering the period 1978-2016.[199] To calculate concentration measures, they use the Longitudinal Business Database (as did Rinz and Lipsius).[200] They use four-digit standard industry-classification codes (the predecessor of NAICS codes). For concentration measures, their data shares all the costs and benefits of the Longitudinal Business Database discussed above.

For manufacturing, they find the average levels of concentration have remained relatively stable, with employment-weighted HHI being 0.569 for the period 1978-1987 and 0.587 for 2008-2016.[201] One should be careful when extrapolating from manufacturing to the whole U.S. economy, given that manufacturing has been declining and the forces changing manufacturing may not apply to the rest of the economy. According to the U.S. Bureau of Labor Statistics, the percentage of employment in manufacturing sector dropped from roughly 22% in 1980 to slightly more than 10% in 2012 (Lipsius 2018, p. 4).

They supplement the concentration measures with two data sets: the Census of Manufacturers, which covers all plants in years ending in 2 and 7, and the Annual Survey of Manufacturers, which covers about 50,000 plants with a threshold of 250-1000 employees for the non-Census years. Other smaller firms are sampled randomly. The Annual Survey of Manufacturers is mandatory reporting, subject to fines for misreporting. They collected data on many things, such as value of shipments. For our discussion, the important thing is that they collect data on actual wages and labor hours, compared to simply posted wages. Moreover, since they are looking at manufacturing, they have better estimates of productivity of firms, as they have better data on inputs and outputs at the plant level. In their baseline regression, moving from a market that is one standard deviation below the median to one standard deviation above is associated with a 3% decline in wages.

Moreover, they are able to use mergers and acquisitions to instrument for concentration to potentially estimate a causal effect of concentration on wages. Using their instrumental-variable approach, they estimate that moving from a market that is one standard deviation below the median to one standard deviation above is associated with a decline in wages of between 9% and 14%.

D.  Berger, Herkenhoff, & Mongey (2022)

Berger, Herkenhoff, and Mongey estimate a general-equilibrium model to measure labor-market power.[202] In the process, their model suggests a certain way to average HHI across markets. They start with LBD at the 3-digit industry level within commuting zones, but they are still left with the problem of how to weight different markets. Instead of weighting by employment level or vacancies level, they weight by market-level payroll, which lowers concentrations slightly, although the trend remains the same.

They find that local concentration is declining over the full period, while national-concentration measures are more complicated. For tradeable sectors, national concentration is falling. For non-tradeable sectors, after falling in the early 1980s, it has slowly risen. But non-tradeables are larger, so the overall national concentration measure has also been rising since the mid 1980s.

In the data (model) weighted average concentration measured in terms of employment is 0.15 (0.16) and in terms of payroll is 0.17 (0.17). In the data (model) unweighted average concentration measured in terms of employment is 0.45 (0.32) and in terms of payroll is 0.48 (0.33).

Source: Berger, Herkenhoff, & Mongey (2022)[203]

E. Handwerker & Dey (2023)

Handwerker and Dey use microdata from the Occupational Employment and Wage Statistics, mapped to the Quarterly Census of Employment and Wages, which records quarterly employment levels for each establishment in the United States that reports to state-level Unemployment Insurance departments.[204] They define markets by 6-digit SOC by metropolitan area. They also look by industry, instead of occupation. They focus on the case where they weight markets by payroll shares, following the theory of Berger, Herkenhoff, and Mongey.[205]

They find an average HHI that is relatively stable and low. They also only look at the private sector and weight by employment, so their results are more directly comparable to some other papers. For example, they directly compare the concentration measures in their data to the 26 occupations of Azar, Marinescu, and Steinbaum.[206] Handwerker and Dey find an HHI in the private sector of one-tenth that found in Azar, Marinescu, and Steinbaum (0.0383 vs. 0.3157).[207] This is the clearest example of how the different data sources matter for concentration numbers.

Source: Handwerker & Dey (2023)[208]

II. Online Job Vacancies

While the above papers use administrative data, other papers on labor-market concentration use online job vacancies (postings) to measure concentration.

A. Azar, Marinescu, Steinbaum, & Taska (2020)

Azar, Marinescu, Steinbaum, and Taska use data on job openings from Burning Glass Technologies (BGT), which collects online job-posting data from 40,000 websites.[209] They restrict their analysis to calendar year 2016, which was the most recent year with available data when the paper was first written. They claim the years 2007-2015 show similar concentration measures (footnote 4).

The papers that use job openings, compared to measures of employment levels, claim openings are a better way to gauge how easy it is for searching workers to find a new job.[210] The nearest government-data product to BGT’s is the Job Opening and Labor Turnover Survey (JOLTS), which is a nationally representative sample of employers. When comparing BGT’s collected job postings to the job postings in JOLTS, the authors estimate that they captured roughly 85% of the job openings in the United States during 2016.

BGT cleans the data to remove double postings and consolidate different spellings for the same employer; i.e., “Bausch and Lomb”, “Bausch Lomb”, and “Bausch & Lomb” are marked as the same employer. After cleaning, 35.9% of employer names are missing, especially if staffing companies do not want to disclose the employer. They assume that all of these with missing employer names are different employers. This means that they have a lower bound on market-concentration measures.

For job description, the BGT dataset uses the Standard Occupational Code (SOC). In the baseline, they consider 200 occupations, which capture 90% of the vacancies in their dataset.[211] For occupations, the authors use six-digit SOC codes for their baseline, but argue that is likely too broad.[212] For location, they use commuting zones, which are geographic definitions based on groups of counties and were developed by the U.S. Department of Agriculture (USDA) to capture local economies and labor markets.[213]

In the SOC-6 occupation by commuting zone by quarter, they find an average HHI of 0.44. For reference, the 2010 Horizontal Merger Guidelines defined markets with post-merger HHIs exceeding 2,500 or 0.25 as “highly concentrated,” and held that mergers in such markets that also increase the HHI level by at least 100 points “raise significant competitive concerns and often warrant scrutiny.”[214] Using the 2010 thresholds, they find that 60% of markets were considered “highly concentrated.”[215] They calculate many other measures of concentration, including at different percentiles and how they vary across the country.[216]

B. Schubert, Stansbury, & Taska (2024)

Schubert, Stansbury, and Taska also use BGT data on vacancies, but with data from 2011 through 2019.[217] They define markets by SOC-6, but use metropolitan area as the location. They do not focus on trends in concentration but on the distribution of concentration and its relationship to wages through outside options to other markets. While the median market has an HHI of 0.0882, the 75th percentile market has an HHI of 0.2143 and the 95th percentile market has an HHI of over 0.8025.[218]

If, however, you weight by level of employment—since many markets have low levels of employment but high levels of concentration—the 50th percentile worker works in a market with an HHI of 0.0137; the 75th percentile worker in a market with an HHI of 0.0404; and the 95th percentile worker in a market with an HHI of 0.1845.[219] That means that under their data and definition of markets, around 5% of workers are in markets that cross the merger-guidelines threshold for a structural presumption (an HHI greater than 1,800 or 0.18, along with an increase of HHI of 100 or 0.01).[220]

When weighting each labor market equally, instead of by size, they find around 25% of markets are over the new threshold.[221] In contrast, using the same data source (BGT) but defining markets differently, Azar, Marinescu, Steinbaum, and Taska find 60% of markets were above the 2,500 threshold.[222]

Source: Schubert, Stansbury, & Taska (2024)[223]

C. Azar, Marinescu, & Steinbaum (2022)

Azar, Marinescu, and Steinbaum use data from, which is a large online job board.[224] The total number of vacancies on represents 35% of the total vacancies in the US in January 2011, as counted by JOLTS. They consider the SOC-5 definition and pick the 13 most frequent occupations over the 2009 to 2012 window, plus the three most frequent occupations in manufacturing and construction. They then consider the SOC-6 definition, which further splits the SOC-5, and end up with 26 occupations in total.[225]

Like Azar, Marinescu, Steinbaum, and Taska,[226] they use commuting zones. They also have data on the number of applicants, which allows measures of “tightness” as (number of vacancies)/(number of applications). They calculate an average HHI for vacancies of 0.3157. When they look at the average based on applications, they find a higher HHI of 0.3480.[227] Again, this is significantly higher than the HHI measure found for the same occupations but using the administrative microdata.[228]

[1] Non-Compete Clause Rule, Final Rule (RIN 3084-AB74, adopted Apr. 23, 2024) (to be codified at 16 C.F.R. Part 910), available at

[2] Complaint, In the Matter of the Kroger Company and Albertsons Companies, Inc., FTC Docket No. D-9428 (Feb. 26, 2024),

[3] U.S. Dep’t. of Just. & Fed. Trade Comm’n, Merger Guidelines 27 (2023), available at

[4] See infra Part II.

[5] See generally U.S. Dep’t of the Treas., The State of Labor Market Competition (Mar. 7, 2022), available at

[6] Merger Guidelines, supra note 3, at 27.

[7] See Jean-Pierre Dubé, Günter J. Hitsch, &Peter E. Rossi, Do Switching Costs Make Markets Less Competitive?, 46 J. Marketing Rsrch. 435, 435 (2009) (“In the simulations, prices are as much as 18% lower with than without switching costs. More important, equilibrium prices do not increase even in the presence of switching costs that are of the same order of magnitude as product price.”).

[8] Merger Guidelines, supra note 3, at 27.

[9] United States v. DaVita Inc., et al., Case No. 21-cr-00229 (D. Colo. 2021).

[10] See, e.g., United States v. Patel, et al., Case No. 21-cr-00220 (D. Conn. 2021) (acquitting all defendants and holding that the evidence did not permit a reasonable jury to conclude there was an agreement to meaningfully allocate the labor market for engineers); United States v. Manahe, et al., Case No. 22-cr-00013 (D. Me. 2022) (acquitting all defendants of charges of a wage-fixing conspiracy among home-healthcare agencies); United States v. Surgical Care Affiliates LLC, et al., Case No. 21-cr-00011 (N.D. Tex. 2021) (DOJ voluntarily dismissed its indictment of a no-poach conspiracy of senior-level surgical facility employees).

[11] Herbert Hovenkamp, The Slogans and Goals of Antitrust Law, 25 Leg. & Pub. Pol’y 705, 705 (2023).

[12] See 15 U.S.C. § 18 (2018) (“No person… shall acquire… the whole or any part of the stock… of another person…, where in any line of commerce…, the effect of such acquisition may be substantially to lessen competition….”).

[13] Merger Guidelines, supra note 3, at 27 (bold/italics emphasis added; italics-only emphasis in original).

[14] See infra Sections IV.B and V.

[15] Complaint, In the Matter of Kroger/Albertsons, supra note 2.

[16] Ioana Marinescu & Herbert J. Hovenkamp, Anticompetitive Mergers in Labor Markets, 94 Indiana L.J. 1031, 1034 (2019).

[17] See, e.g., id. (“While the use of section 7 to pursue mergers among buyers is well established, there is relatively little case law.”).

[18] Merger Guidelines, supra note 3, at 26-27.

[19] See Non-Compete Clause Rule, Final Rule, supra note 1.

[20] For an extensive review of the noncompete literature relied upon by the FTC and a discussion of the nuances and limitations of that literature, see Alden Abbott, et al., Comments of Scholars of Law & Economics and ICLE in the Matter of Non-Compete Clause Rulemaking, FTC Matter No. P201200 (Apr. 19, 2023),

[21] Complaint, In the Matter of Kroger/Albertsons, supra note 2, at ¶¶ 63 & 70.

[22] See, e.g., Suresh Naidu, Eric A. Posner, & Glen Weyl, Antitrust Remedies for Labor Market Power, 132 Harv. L. Rev. 536 (2018), (“As far as we know, the DOJ and FTC have never challenged a merger because of its possible anticompetitive effects on labor markets, or even rigorously analyzed the labor market effects of mergers as they do for product market effects. Nor have we found a reported case in which a court found that a merger resulted in illegal labor market concentration.”). Ioana Marinescu & Eric A. Posner, Why Has Antitrust Law Failed Workers?, 105 CORNELL L. REV. 1343 (2020)

[23] Exec. Order No.14036, 86 FR 36987 (2021).

[24] See United States v. Bertelsmann SE & Co. KGaA, et al., 646 F. Supp. 3d 1 (D.D.C. 2022).

[25] See Press Release, FTC Cracks Down on Companies That Impose Harmful Noncompete Restrictions on Thousands of Workers, Fed. Trade. Comm’n, (Jan. 4, 2023), See also, e.g., Complaint and Decision and Order, In the Matter of Anchor Glass Container Corp., et al., Fed. Trade. Comm’n (Jun. 2, 2023),

[26] See Non-Compete Clause Rule, Notice of Proposed Rulemaking, 88 Fed. Reg. 3482 (RIN 3084, proposed Jan. 19, 2023) (to be codified at 16 C.F.R. Part 910).

[27] See cases referenced supra note 10.

[28] Dept of Just. & Fed. Trade Comm’n, Antitrust Guidance For Human Resource Professionals (2016),

[29] Press Release, Justice Department Requires Six High Tech Companies to Stop Entering into Anticompetitive Employee Solicitation Agreements, U.S Dept. of Just. (Sep. 24, 2010),

[30] United States v. Arizona Hosp & Healthcare Ass’n & AzHHA Service Corp., No. CV07-1030-PHX (D. Az. May 22, 2007).

[31] Memorandum of Understanding Between the Fed. Trade Comm’n and the Nat’l Labor Relations Bd. Regarding Information Sharing, Cross-Agency Training, and Outreach in Areas of Common Regulatory Interest (Jul. 19, 2022), available at

[32] Memorandum of Understanding Between the U.S. Dep’t of Labor and the Fed. Trade Comm’n (Aug. 30, 2023), available at

[33] See generally Herbert Hovenkamp, The Antitrust Enterprise: Principle and Execution 38-9 (2005) (citing examples and noting that “post-Chicago theory typically models strategic behavior by use of game theory, with alternatives that reach far beyond the conventional Cournot oligopoly analysis”). See also, e.g., Edward J. Green & Robert H. Porter, Noncooperative Collusion under Imperfect Price Information, 52 Econometrica 87 (1984); Thomas G. Krattenmaker & Steven C. Salop, Anticompetitive Exclusion: Raising Rivals’ Costs to Achieve Power over Price, 96 Yale L.J. 209 (1986).

[34] Herbert J. Hovenkamp & Fiona Scott Morton, Framing the Chicago School of Antitrust Analysis, 168 U. Pa. L. Rev. 1843, 1847 (2020) (“Built into Chicago School doctrine was a strong presumption that markets work themselves pure without any assistance from government. By contrast, imperfect competition models gave more equal weight to competitive and noncompetitive explanations for economic behavior….”).

[35] See, e.g., Lester G. Telser, Why Should Manufacturers Want Fair Trade?, 3 J.L. & Econ. 86 (1960); George J. Stigler, A Theory of Oligopoly, 72 J. Pol. Econ. 44 (1964); Howard Marvel, Exclusive Dealing, 25 J. L. Econ. 1 (1982).

[36] See, e.g., Gregory J. Werden & Luke M. Froeb, The Effects of Mergers in Differentiated Products Industries: Logit Demand and Merger Policy, 10 J.L. Econ. & Org. 407 (1994); Jonathan B. Baker & Timothy F. Bresnahan, The Gains from Merger or Collusion in Product-Differentiated Industries, 33 J. Indus. Econ. 427 (1985).

[37] To be clear, this is merely a descriptive claim about the present state of the relationship between labor economics and antitrust, not a normative claim that the two fields should not develop stronger connections.

[38] See, e.g., Jose Azar, Iona Marinescu, & Marshall Steinbaum, Labor Market Concentration, 57 J. Hum. Res. S167, S197 (Supp. 2022) (“The type of analysis we provide could be used to incorporate labor market concentration concerns as a factor in antitrust analysis.”).

[39] See U.S. Dep’t of the Treas., supra note 5.

[40] See, e.g., Azar, Marinescu, & Steinbaum, supra note 38, at S174 (“Our baseline measure of market power in a labor market is the Herfindahl–Hirschman index (HHI)….”); Carl Shapiro, Protecting Competition in the American Economy: Merger Control, Tech Titans, Labor Markets, 33 J. Econ. Persp. 69, 75-76 (2019). (“Measures of industry concentration based on data from the US Economic Census are simply not very informative for merger analysis because these data are available only at an aggregated level. The modest increases in concentration observed when using these data confirm that the largest firms are responsible for a greater portion of economic activity in many industries, but they tell us very little about concentration in properly defined relevant antitrust markets… Furthermore, it is important to remember that an increase in concentration in a properly defined relevant market does not prove that competition in that market has declined.”).

[41] This is effectively the labor-market equivalent of markups that measure whether firms enjoy market power in the market for goods or services. See, e.g., Naidu, Posner, & Weyl, supra note 22, at 556 (“The firm’s absolute markup is the gap between this price and the firm’s cost. The markup equals the difference between the monopoly price and the competitive price, and thus serves as a natural gauge of market power… As in the monopoly case, a monopsonist will not internalize this effect on workers and will choose an “absolute markdown” of wages below the marginal revenue product.”).

[42] As we will discuss later, this connection between labor-supply elasticities, marginal products, and wages is more complicated. For example, the markdown could be a mismeasured return to technology, not traditional market power. See, e.g., Ivan Kirov & James Traina, Labor Market Power and Technological Change in US Manufacturing, conference paper for Institute for Labor Economics (Oct 2022), at 42, available at (“The labor [markdown] therefore increases because “productivity” rises, and not because pay falls. This suggests that technological change plays a large role in the rise of the labor [markdown].”).

[43] See Naidu, Posner, & Weyl, supra note 22.

[44] Id. at 567. See also Douglas O. Staiger, Joanne Spetz, & Ciaran S. Phibbs, Is There Monopsony in the Labor Market? Evidence from a Natural Experiment, 28 J. LAB. ECON. 211 (2010); Arindrajit Dube, Laura Giuliano, & Jonathan Leonard, Fairness and Frictions: The Impact of Unequal Raises on Quit Behavior, 109 AM. ECON. REV. 620 (2019).

[45] For one example, Matsudaira uses a natural experiment around the introduction of state minimum-nurse-staffing laws and evidence consistent with perfect competition and zero market power for nurse-aides. High and low market power can exist at the same time. See Jordan D. Matsudaira, Monopsony in the Low-Wage Labor Market? Evidence from Minimum Nurse Staffing Regulations, 96 Rev. Econ. & Stat. 92 (2014).

[46] See Naidu, Posner, & Weyl, supra note 22, at 564-566.

[47] Loukas Karabarbounis & Brent Neiman, The Global Decline of the Labor Share, 129 Quarterly J. of Econ. 61, 71 (2013).

[48] Michael R. Ransom & David P. Sims, Estimating the Firm’s Labor Supply Curve in a “New Monopsony” Framework: Schoolteachers in Missouri, 28 J. LAB. ECON. 331 (2010).

[49] See, e.g., Efraim Benmelech, Nittai K. Bergman, & Hyunseob Kim, Strong Employers and Weak Employees, How Does Employer Concentration Affect Wages?, 57 J. Hum. Res. S200 (Supp. 2022). See also David Berger, Kyle Herkenhoff, & Simon Mongey, Labor Market Power, 112 Am. Econ. Rev. 1147 (2022).

[50] José A. Azar, Steven T. Berry, & Ioana Marinescu, Estimating Labor Market Power (Nat’l Bureau of Econ. Rsch., Working Paper No. 30365, 2022).

[51] Id. at 35.

[52] Chen Yeh, Claudia Macaluso, & Brad Hershbein, Monopsony in the US Labor Market, 112 Am. Econ Rev. 2099 (2022).

[53] Id. at 2099.

[54] Id. at 2114.

[55] Id. at 2099.

[56] See Steven Berry, Market Structure and Competition Redux, Presentation at Fed. Trade. Comm’n Micro Conference (Nov. 2017), available at; See also Brian Albrecht, Markups as Residuals, Economic Forces (Nov. 17, 2022),

[57] See Kirov & Traina, supra note 42.

[58] 2023 Merger Guidelines, supra note 3.

[59] See infra Appendix.

[60] In order to evaluate concentration, the relevant market must be defined. For labor markets, the relevant market is usually defined as both the job description (e.g., nurse) and the location of the job (e.g., Portland area). Using this, one can calculate some measure of concentration, such as the HHI. Economics papers tend to report HHI as a percentage, instead of as a cardinal number out of 10,000, as used in the merger guidelines. For example, an HHI of 1,800 would be written as “0.18.”

[61] See, e.g., Kevin Rinz, Labor Market Concentration, Earnings, and Inequality, 57 J. Hum. Res. S251 (Supp. 2022); David Autor, Christina Patterson, & John Van Reenen, Local and National Concentration Trends in Jobs and Sales: The Role of Structural Transformation, 5 (Nat’l Bureau of Econ. Rsch., Working Paper No. 31130, 2023) at 7 (“The employment-based HHI fell by 2.3 points, from 33.3 in 1992 to 31.0 in 2017, which stands in contrast to the 3.4 point rise in the sales HHI. Our estimates for local employment concentration echo those of Rinz (2022), who uses the LBD.”) (emphasis in original).

[62] Rinz, id. at S256.

[63] See Azar, Marinescu, & Steinbaum, supra note 38.

[64] Handwerker & Dey directly compare the concentration measures in their data to the 26 occupations studied by Azar, Marinescu, & Steinbaum. They find an HHI in the private sector of 0.0383, compared to 0.3157 in Azar, Marinescu, & Steinbaum. See Elizabeth Weber Handwerker & Matthew Dey, Some Facts About Concentrated Labor Markets in the United States, 63 Indus. Rel. 132, 135 (2023); Azar, Marinescu, & Steinbaum, supra note 38.

[65] A firm may have multiple establishments, and the data allow different NAICS codes for each establishment, so, in some cases and to some extent, different types of workers can be separated out if they work in different locations.

[66] Berger, Herkenhoff, & Mongey, supra note 49, at 1169 (citing Elizabeth Handwerker & Matthew Dey, Megafirms and Monopsonists: Not the Same Employers, Not the Same Workers (Unpublished)).

[67] Rinz, supra note 61.

[68] Id. at S264 (“In both years, the areas that are most concentrated tend to be rural. In particular, the Great Plains region has a relatively large number of highly concentrated commuting zones in both 1976 and 2015. The least concentrated markets tend to be in urban areas.”).

[69] Kevin Rinz, Labor Market Concentration, Earnings Inequality, and Earnings Mobility, National Bureau of Economic Research Summer Institute (Jul. 23, 2019) (slides obtained from author).

[70] Rinz, supra note 61 at S253.

[71] See Ben Lipsius, Labor Market Concentration Does Not Explain the Falling Labor Share, Working Paper (2018), available at

[72] See Berger, Herkenhoff, & Mongey, supra note 49.

[73] See 2023 Merger Guidelines, supra note 3.

[74] See, e.g., Azar, Marinescu, and Steinbaum, supra note 38; Jose Azar, Iona Marinescu, Marshall Steinbaum, & Bledi Taska, Concentration in US Labor Markets: Evidence from Online Vacancy Data, 66 Labor Econ. 101886 (2020).

[75] For a more detailed discussion of these papers and their limitations, see Appendix Section II, infra.

[76] See Harold Demsetz, Industry Structure, Market Rivalry, and Public Policy, 16 J. L. & Econ. 1 (1973). See also, e.g., Richard Schmalensee, Inter-Industry Studies of Structure and Performance, in 2 Handbook of Industrial Organization 951 (Richard Schmalensee & Robert Willig, eds., 1989); William N. Evans, Luke M. Froeb, & Gregory J. Werden, Endogeneity in the Concentration-Price Relationship: Causes, Consequences, and Cures, 41 J. Indus. Econ. 431 (1993); Berry, supra note 56; Nathan Miller, et al., On the Misuse of Regressions of Price on the HHI in Merger Review, 10 J. Antitrust Enf. 248 (2022).

[77] Some papers find lower wages in markets with higher employer concentration, but do not differentiate rural from urban labor markets. Rural and urban labor markets can differ significantly in terms of their economic structures, job opportunities, and wage levels. Any regression of wages on concentration is likely picking up something unrelated to concentration directly. See Benmelech, Bergman, & Kim, supra note 49.

[78] Kirov & Traina, supra note 42.

[79] Id. at 46 (emphasis added).

[80] Steven Berry, Martin Gaynor, & Fiona Scott Morton, Do Increasing Markups Matter? Lessons from Empirical Industrial Organization, 33 J. Econ. Persp. 44, 57 (2019) (emphasis added).

[81] The antitrust statutes do not distinguish buy-side and sell-side behavior, besides the partial exception in Section 6 of the Clayton Act, which provides that workers do not violate antitrust laws when they organize unions. See 15 U.S.C. § 17 (“The labor of a human being is not a commodity or article of commerce. Nothing contained in the antitrust laws shall be construed to forbid the existence and operation of labor… organizations, instituted for the purposes of mutual help…, or to forbid or restrain individual members of such organizations from lawfully carrying out the legitimate objects thereof….”). In practice, however, it seems the agencies have historically treated labor markets differently. See, e.g., Naidu, Posner, & Weyl, supra note 22.

[82] See, e.g., Roger G. Noll, Buyer Power and Economic Policy, 72 Antitrust L.J. 589, 589 (2005) (“[B]uyer power arises from monopsony (one buyer) or oligopsony (a few buyers), and is the mirror image of monopoly or oligopoly.”); id. at 591 (“Asymmetric treatment of monopoly and monopsony has no basis in economic analysis.”).

[83] Of course, monopoly markets in intermediate products (i.e., products sold not to end users, but to manufacturers who use them as inputs for products that are, in turn, sold to end users) may indeed sit in the same place in the supply chain as the typical monopsony market. Some, but not all, of the complications associated with monopsony analysis are relevant to these monopoly situations, as well.

[86] For purposes of this discussion, “monopoly” refers to any merger (or other conduct) that would increase market power by a seller in a product market, and “monopsony” refers to any merger (or other conduct) that would increase market power by a buyer in an input market (including a labor market).

[87] Keith Brand, Martin Gaynor, Patrick McAlvanah, David Schmidt, & Elizabeth Schneirov, Economics at the FTC: Office Supply Retailers Redux, Health Care Quality Efficiencies Analysis, and Litigation of an Alleged Get Rich Quick Scheme, 45 Rev. Indus. Org. 325 (2014).

[88] Id.

[89] Some efficiency-enhancing mergers will be identifiable, of course. For example, if the merger raises quantities and prices for all inputs, that must be efficiency enhancing. The problem, as always, is with the hard cases.

[90] See C. Scott Hemphill & Nancy L. Rose, Mergers that Harm Sellers, 127 Yale L.J. 2078 (2018).

[91] In theory, one could force a monopsony model to be identical to monopoly. The key difference is about the standard economic form of these models that economists use. The standard monopoly model looks at one output good at a time, while the standard factor-demand model uses two inputs, which introduces a tradeoff between, say, capital and labor. See Sonia Jaffe, Robert Minton, Casey B. Mulligan, and Kevin M. Murphy, Chicago Price Theory (2019) at Ch. 10. One could generate harm from an efficiency for monopoly (as we show for monopsony) by assuming the merging parties each produce two different outputs, apples and bananas. An efficiency gain could favor apple production and hurt banana consumers. While this sort of substitution among outputs is often realistic, it is not the standard economic way of modeling an output market.

[92] Herbert Hovenkamp, Worker Welfare and Antitrust, 90 U. CHI. L. REV. 511, 529 (2023) (“To the extent that such actions lead to higher prices or reduced product output, labor as well as consumers suffer.”).

[93] Marinescu & Hovenkamp, supra note 16 at 1042 (“The key message from economic theory is that as one moves away from the competitive equilibrium towards a situation of monopsony in the labor market, wages and production both generally tend to decrease.”).

[94] See United States v. Bertelsmann SE & Co. KGaA, et al., supra note 24.

[95] Id. at 23 (“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.”)

[96] John Asker & Volker Nocke, Collusion, Mergers, and Related Antitrust Issues, in 5 HANDBOOK oF INDUSTRIAL ORGANISATION 177, 221-22 (Kate Ho, Ali Hortasçu & Alessandro Lisseri eds., 2021).

[97] But see United States v. Bertelsmann SE & Co. KGaA, et al., supra note 24, at 28 (“Thus, even if alternative submarkets exist at other advance levels, or if there are broader markets that might be analyzed, the viability of such additional markets does not render the one identified by the government unusable.”). Of course, in that case, the parties (and the court) did identify downstream harms. See id. at 23.

[98] See generally, Geoffrey A. Manne & E. Marcellus Williamson, Hot Docs vs. Cold Economics: The Use and Misuse of Business Documents in Antitrust Enforcement and Adjudication, 47 Ariz. L. Rev. 609, 619 (2005).

[99] Hemphill & Rose, supra note 90. The authors make a useful distinction between mergers that generate classical monopsony and those that increase buyer leverage. As explained below, however, increased buyer bargaining leverage is just a transfer from sellers to buyers. If it truly has no effect on output, as supposed for Hemphill & Rose, it is not anticompetitive. If antitrust is to weigh in on splitting the surplus and conclude that a merger that leads to more of the surplus going to the buyer is anticompetitive, the courts would be implicitly saying that either the division before the merger was optimal or that more surplus going to sellers is always better. While people may have an intuition that more surplus going to sellers of labor (i.e., workers) is better, do we have the same intuition for all types of sellers? Moreover, would we be willing to apply the same logic to mergers to monopoly? If so, and mergers that increase buyer leverage are bad and mergers that increase seller leverage are bad (again with no effect on output), are we concluding all mergers are bad, full stop?

[100] Marinescu & Hovenkamp, supra note 16, at 1040 (emphasis added).

[101] Such bargaining models have been awarded Nobel prizes. See Peter Diamond, Wage Determination and Efficiency in Search Equilibrium, 49 Rev. Econ. Stud. 217 (1982); Christopher A. Pissarides, Equilibrium Unemployment Theory (2017).

[102] See, e.g., Richard Rogerson, Robert Shimer, & Randall Wright, Search-Theoretic Models of the Labor Market: A Survey, XLIII J. ECON. LIT. 959,961 (2005) (“Bargaining is one of the more popular approaches to wage determination in the literature…”).

[103] See, e.g., John Van Reenan, Labor Market Power, Product Market Power and the Wage Structure: A Note 224 (Program on Innovation and Diffusion, Working Paper No. 085, 2023),, (“Here, when firms achieve more product market power there are higher profits and therefore more of a potential surplus to be split between employers and employees. Workers (at least those who keep their jobs), may welcome greater monopoly power as they are able to extract higher wage rents, which would not be the case for a firm earning thin or no margins in an extremely competitive product market. Consequently, this generates the opposite implication at the firm level – more product market power generates higher, not lower, wages.”).

[104] Complaint, In the Matter of Kroger/Albertsons, supra note 2, at ¶ 63 (“Union grocery labor is a relevant market in which to analyze the probable effects of the proposed acquisition.”).

[105] Indeed, increased bargaining power is the purpose of a union. Whether the coordination leads to equivalent, lesser, or greater bargaining power than that of employers in a given case depends on many specifics. But the whole point of both the union and the labor antitrust exemption is to facilitate the exercise of this increased bargaining power on the labor side.

[106] Lynn Petrak, Local Union Supports Kroger-Albertsons Merger, Progressive Grocer (Feb. 21, 2024),

[107] Press Release, America’s Largest Union of Essential Grocery Workers Announces Opposition to Kroger and Albertsons Merger, United Food and Commercial Workers (May 5, 2023),

[108] See Petrak, supra note 106.

[109] Roman Inderst & Christian Wey, Countervailing Power and Dynamic Efficiency, 9 J. Eur. Econ. Ass’n 702, 715 (2011).

[110] For further discussion of the problems of reconciling upstream and downstream market effects when labor markets are taken into account, see Section V, infra.

[111] FTC v. Whole Foods Mkt., Inc., 548 F.3d 1028, 1063 (D.C. Cir. 2008). See also Geoffrey Manne, Premium, Natural, and Organic Bullsh**t, Truth on the Market (Jun. 6, 2007), (“In other words, there is a serious risk of conflating a ‘market’ for business purposes with an actual antitrust-relevant market.”).

[112] Unsurprisingly, there is no SOC code that corresponds to such a market definition, and the FTC did not allege it. See Occupational Employment and Wage Statistics, May 2023 Occupation Profiles, Bureau of Labor Statistics (last visited Apr. 23, 2024),

[113] Complaint, In the Matter of Kroger/Albertsons, supra note 2, at ¶ 63.

[114] See Brian Albrecht, Dirk Auer, Eric Fruits, & Geoffrey A. Manne, Food-Retail Competition, Antitrust Law, and the Kroger/Albertsons Merger, Int’l. Ctr. for Law & Econ. White Paper 2023-10-17 (2023),

[115] See generally Section A, infra.

[116] See, e.g., Amos Golan, Julia Lane, & Erika McEntarfer, The Dynamics of Worker Reallocation within and across Industries, 74 Economica. 1 (2007). (“About 27% of workers who had previously exhibited a substantial degree of attachment to their employer reallocate in a given year. About two-thirds of this reallocation is job-to-job reallocation, split roughly evenly between, within and across broadly defined industries.)

[117] See United States v. Bertelsmann SE & Co. KGaA, et al., supra note 24.

[118] See, e.g., Press Release, Justice Department Obtains Permanent Injunction Blocking Penguin Random House’s Proposed Acquisition of Simon & Schuster, US Dep’t of Justice (Oct. 31, 2022), (“‘The decision is also a victory for workers more broadly,’ said AAG Kanter. ‘It reaffirms t