Two Challenges for Neo-Brandeisian Antitrust

Abstract

Several scholars and policy-makers have proposed a “Neo-Brandeisian” reform of U.S. antitrust law, aimed at reviving “republican” antitrust. Republicanism conceives of domination as inherently detrimental to freedom. Republican antitrust considers antitrust law as an “institution of antipower,” aimed at dispersing economic power. This paper sets out two key challenges to the Neo-Brandeisian reform agenda and argues for legal formalism to address them. First, republicanism would alter the normative justification, but not necessarily the content of antitrust law. Neoclassical antitrust law does not broadly reflect a Schumpeterian endorsement of dominance. Rather, its epistemological priors and methodology entail skepticism about the mere presence of economic power. Thus, mainstream antitrust law and policy remain unfazed by the Neo-Brandeisian claim that antitrust should target domination instead of consumer welfare. Second, Neo-Brandeisian reform proposals are inherently polycentric. How Neo-Brandeisians aim to balance distinct values including the competitive process, the harm of concentrated power, and the protection of democracy and egalitarianism has remained unclear. This paper argues that both challenges demand for a formalistic approach to Neo-Brandeisian antitrust. Compared to a case-by-case approach, adopting general rules through legislative or administrative decision-making may legitimately overturn current precedent, incorporate alternate methods of measuring power and competitive harm, and pursue a variety of republican goals. Neo-Brandeisian formalism would essentially reinvigorate the Harvard school’s insight that multiple purposes—including both efficiency and republican liberty—can be attained by formalistic rules.

I. Introduction

Several scholars and policy-makers, many of whom self-identify as “Neo-Brandeisians,” ¹ have proposed a “republican” reform of U.S. antitrust law. ² According to Neo-Brandeisians, notably including FTC Chair Lina Khan and law professor and former presidential advisor Tim Wu, republican principles can guide U.S. antitrust law from its current focus on efficiency and consumer welfare toward a renewed focus on combating economic and political domination, as it had had before the rise of the Chicago school of antitrust. ³

Republicanism in this sense refers to a theory of liberty that conceives of liberty as the absence of domination. ⁴ Republicanism is built on the premise that domination as such is inherently detrimental to freedom, even if there is no actual or likely interference with the negative or positive freedom of others. ⁵ Just as slaves are unfree even if their master is benevolent—simply because the master has the capacity to change his mind and interfere with the slaves’ freedom at will ⁶ —the fact that economically dominant undertakings have the mere capacity to interfere with the economic freedom of smaller competitors implies a lack of freedom. ⁷ As economic power translates into political power, ⁸ moreover, curtailing economic power—among others through antitrust law—has as one of its distinct functions the protection of republicanism. ⁹ Thus, a republican theory of antitrust conceives of antitrust law as an “institution of antipower,” ¹⁰ aimed at protecting a republican system of government and society by dispersing economic power. ¹¹

This republican understanding of U.S. antitrust law is clearly visible, for example, in Justice Black’s Opinion of the Court in Northern Pacific R. Co. v. United States:

The Sherman Act was designed to be a comprehensive charter of economic liberty aimed at preserving free and unfettered competition as the rule of trade. It rests on the premise that the unrestrained interaction of competitive forces will yield the best allocation of our economic resources, the lowest prices, the highest quality, and the greatest material progress, while at the same time providing an environment conductive to the preservation of our democratic political and social institutions. ¹²

The Neo-Brandeisian antitrust agenda aims to reinvigorate the republican conception of antitrust both at the level of abstract theory and at the level of concrete legal doctrine. At the goal level, Neo-Brandeisians reject the focus on “efficiency” or “consumer welfare” as the supreme goal of U.S. antitrust law. ¹³ At the doctrinal level, Neo-Brandeisian proposals include a number of specific interventions ranging from a renewal of the “big case” tradition, ¹⁴ to stricter tests for predatory pricing and a prophylactic ban on vertical integration by already dominant platform companies. ¹⁵ Their overall purpose is a radical overhaul of current, neoclassical antitrust doctrine. ¹⁶

This essay sets out two key challenges to the reinvigoration of a republican conception of antitrust and provides an argument in favor of legal formalism as an important tool to address these two challenges. My aim is not to argue in favor or against Neo-Brandeisian antitrust as such. Rather, I aim to show that, if one wants to implement a Neo-Brandeisian antitrust agenda, these two challenges are the primary obstacles to overcome.

The first challenge centers on the Neo-Brandeisians’ argument that antitrust law should target “domination” instead of advancing “consumer welfare” or “efficiency.” Most of neoclassical antitrust doctrine remains unfazed by this claim. Adopting republicanism instead of consumer welfare as one of antitrust’s central values would, as such, not substantially alter the technical content of antitrust law and its resulting laissez-faire attitude toward markets, because neoclassical antitrust analysis precisely denies the prevalence of economic power and domination. ¹⁷

The second challenge pertains to the complexity and polycentricity of the move from neoclassical antitrust toward republican antitrust. Neoclassical economics has provided antitrust law with a highly formalized and workable decision-making procedure. ¹⁸ Insofar as the Neo-Brandeisian antitrust agenda takes into account republicanism as a central value of antitrust law, it puts complex balancing exercises back on the table.

Based on these two challenges, this essay argues in favor of a “new” legal formalism to implement a Neo-Brandeisian antitrust agenda. A formalistic approach—in which political determinations constrain administrative and judicial decision-makers ¹⁹ —directly remedies the second challenge, but also allows for a solution to the first challenge by legislatively altering the metrics of power and domination. In turn, however, this article also recognizes the new political and constitutional challenges that such legal formalism invites.

The remainder of this article is divided in four parts. Section II analyzes the first challenge, focusing on the epistemic and methodological relevance of power and domination in republicanism and neoclassical antitrust, respectively. Section III analyzes the second challenge, focusing on the complexity and polycentricity of the Neo-Brandeisian antitrust agenda, and argues in favor of legislative or administrative rather than judicial intervention. Section IV makes an overarching, normative claim in favor of implementing a Neo-Brandeisian antitrust agenda through legal formalism. Section V concludes.

II. Power and Competition in Republican and Neoclassical Antitrust

The first challenge for Neo-Brandeisian antitrust is that the doctrinal content of neoclassical antitrust law is unfazed by whether or not republicanism underlies antitrust. Neo-Brandeisians have emphasized the Chicago school’s—and current doctrine’s—endorsement of consumer welfare as the sole goal of antitrust law. ²⁰ By focusing on consumer welfare, however, Neo-Brandeisians only highlight the claim that whenever something does not harm consumers, it cannot be an antitrust violation. Neoclassical antitrust has followed the Chicago school’s more fundamental belief, however, that power concentration and domination are much rarer than pre-Chicago legal doctrine had presumed. And while antitrust law to some extent has moved beyond the Chicago school’s specific doctrinal prescriptions, ²¹ lax enforcement of both section 1 and section 2 of the Sherman Act 1890 shows that the belief that power and domination are rare phenomena in the economy continues to permeate the current neoclassical paradigm in U.S. antitrust law. ²²

Thus, the Chicago school primarily casted epistemological and methodological doubt on the implementation of republican antitrust in the 1940s to 1960s. The epistemic priors and methodology of the Chicago school and neoclassical antitrust have allowed them to take over the low evidentiary standard characteristic of republican antitrust—focusing on the mere “capability” to affect competition—and incorporate it in a laissez-faire approach. Neo-Brandeisian antitrust, in turn, will also need to emphasize methodology, rather than normative theory.

A. The Implementation of Republican Antitrust: Domination and Capability

As Neo-Brandeisians rightly emphasize, pre-Chicago antitrust built on the republican premise that “monopoly power, whether lawfully or unlawfully acquired, may itself constitute an evil and stand condemned under section 2 even though it remains unexercised.” ²³ A key challenge for republican antitrust, however, was to reconcile this republican skepticism of domination with the belief that legally acquired private power, for instance through superior quality or efficiency, should remain unpunished.

In this regard, Learned Hand’s famous Opinion in the Alcoa case came very close to a “no fault” approach to section 2 of the Sherman Act. On this view, limiting the prohibition of monopolization to “manoeuvres not honestly industrial, but actuated solely by a desire to prevent competition” would “emasculate the Act; would permit just such consolidations as it was designed to prevent.” ²⁴ Thus, the only exception to the prohibition in section 2 was unavoidable monopolization, so as to exempt “those who do not seek, but cannot avoid, the control of a market.” ²⁵

Alcoa, however, did not represent a clear legal consensus. ²⁶ Section 3 of the Clayton Act 1914, which was adopted in response to the introduction of the rule of reason in Standard Oil, ²⁷ prohibited specific types of conduct—tying, exclusive dealing and price discounts—insofar as they substantially lessened competition. ²⁸ The Robinson–Patman Act 1936 added a ban on various forms of price discrimination. ²⁹ Other judicial decisions, moreover, rejected the prohibition of monopoly as such, and emphasized the need to prove monopolization through certain abusive practices, such as those prohibited by the Clayton Act and the Robinson–Patman Act. ³⁰

While there were diverging views as to whether size or dominance as such sufficed to violate section 2 of the Sherman Act, it is clear that pre-Chicago antitrust was based on a republican concern about dominance and market power. ³¹ According to Elias Deutscher, a low threshold for section 2 liability, which emphasized the capability to interfere with other firms’ freedom, characterized this republican approach. ³² Carl Kaysen and Donald Turner’s seminal statement of the Harvard antitrust school went even further and proposed structural remedies against sustained “unreasonable market power”—even without evidence of actual anticompetitive conduct—combined with an efficiency defense for the dominant company concerned. ³³

The Neo-Brandeisian agenda aims to reinvigorate this republican conception of antitrust. Irrespective of whether this is done through “no fault” prohibitions, or through conduct-based prohibitions focusing on the mere “capability” to affect competition, Neo-Brandeisians presuppose that economic domination is a major problem in the current economy ³⁴ As it turns out, this is precisely what Chicago school analysis—and to a large extent neoclassical antitrust doctrine—denies.

B. Domination and Neoclassical Antitrust

Whether the Neo-Brandeisian critique of neoclassical antitrust law follows from republican theory depends on whether Neo-Brandeisians are right to claim that it is not concerned with domination in absence of concrete evidence of consumer harm. Republicanism as such, however, does not imply a rejection of current doctrine. The laissez-faire attitude of neoclassical antitrust doctrine is not the result of limiting the goal of antitrust to protecting consumers, nor of neglecting republicanism. Instead, the content of neoclassical antitrust is primarily the result of a particular epistemology and methodology for measuring power.

The Neo-Brandeisian concern about domination primarily targets a Schumpeterian perspective on antitrust law. According to this view, temporary domination is a central and necessary part of innovation-generating waves of creative destruction. ³⁵ Consequently, a Schumpeterian perspective on antitrust entails a more cautious attitude toward antitrust intervention against temporary economic dominance. ³⁶ By emphasizing the inherent harm of domination, the Neo-Brandeisian antitrust reform agenda is the antithesis of this Schumpeterian perspective.

Some antitrust doctrines indeed seem to take a Schumpeterian view. The U.S. Supreme Court’s judgment in Trinko, in particular, clearly rejected a republican conception of antitrust law:

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. ³⁷

These Schumpeterian principles led the Supreme Court to effectively nullify the exploitative prices doctrine, ³⁸ substantially limit the refusal to deal doctrine, ³⁹ and allow certain types of regulation to pre-empt antitrust intervention. ⁴⁰ Contrary to republican concerns about monopoly, the Opinion of Justice Scalia applauds dominance as “an important element of the free-market system.” ⁴¹

However, Trinko is not representative of the key principles and rules of neoclassical antitrust law. At the core of the Chicago school’s analysis was a deep skepticism about the prevalence of economic dominance and the capability of allegedly dominant companies to affect competition. ⁴² Although neoclassical antitrust doctrine has far from incorporated all of the Chicago school’s prescriptions, its legal epistemology similarly centers on a narrow conception of power that does not expressly reject, but is rather blind to Neo-Brandeisian concerns about dominance.

Skepticism about the prevalence of market power and its threat to competitiveness can be linked to the four below-mentioned epistemic priors and methodological commitments. The first three points characterize major contributions to Chicago school scholarship, while at least the first point has become a commonplace in antitrust legal doctrine more generally. In response to alleged power concentration in digital markets, some scholars wary of antitrust intervention have increasingly emphasized the fourth point. Importantly, all four points refer to either epistemological priors or methodological commitments that are independent of whether or not antitrust law should aim to preserve republican values; for there is no need to combat domination if it is (thought to be) absent in the economy.

1. A narrow conception of power based on neoclassical price theory;

2. A presumption of efficiency over domination and exploitation;

3. A presumption that markets are contestable and contested; and in some cases,

4. A shift from market-oriented analysis to a focus on alleged competition between markets offsetting power within markets.

As to the first point, antitrust law has embraced neoclassical price theory to analyze market power, understood as the power of a firm over price or output. ⁴³ The difficulty of applying this conception of market power to contemporary digital markets, rife with zero-pricing business models, is well known. ⁴⁴ More generally, power over price or output on a particular market neglects other forms of economic power including but not limited to “gatekeeping” power and “vertical” power that firms may exercise vis-à-vis their trading partners, ⁴⁵ which precisely has been a main concern for Neo-Brandeisian scholarship. ⁴⁶

Regarding the second point, Frank Easterbrook famously argued that economists and lawyers simply do not understand the rationale of many business practices and—as a result of their ignorance—wrongly regard them with suspicion. ⁴⁷ For Easterbrook, practices such as exclusive dealing, tying and resale price maintenance just “cannot be anticompetitive.” ⁴⁸ Similarly, Richard Posner rejected the theory of anticompetitive leveraging because “the [. . .] monopolist will gain nothing from the second monopoly.” ⁴⁹ The argument is not that domination is unproblematic; it is that many business practices are incapable of affecting competition. But if such practices are allegedly incapable of affecting competition, it is immaterial whether their legality is assessed under a neoclassical or a republican standard. ⁵⁰

Similarly, in Brooke Group, the Supreme Court held that predatory pricing requires evidence that the defendant has “a reasonable prospect of recouping its investment in below-cost prices.” ⁵¹ This recoupment requirement reflects the view that “below-cost pricing must be capable, as a threshold matter, of producing the intended effects on the firm’s rivals[.]” ⁵² While a capability threshold is usually associated with republican antitrust, ⁵³ Brooke Group shows that it equally conforms to a neoclassical approach.

It should be emphasized that “capability” might have a different meaning in neoclassical antitrust than it has in republican antitrust. Arguably, Brooke Group understands “capability” probabilistically, its meaning being close to that of “likelihood,” ⁵⁴ while republican antitrust understands the capability to detrimentally affect competition non-probabilistically, in line with the belief that the mere existence of economic power is detrimental to freedom. ⁵⁵ Notwithstanding this semantic ambiguity, Brooke Group provides a illustrative example of how neoclassical antitrust, rather than taking a Schumpeterian perspective, is mostly concerned with the question of whether particular types of business conduct generate the degree of power necessary to harm competition—and in practice usually answers this question in the negative. ⁵⁶

Priors about market contestability have been complementary to the presumption that business conduct is generally pro-competitive. The Chicago school embraced George Stigler’s narrow definition of a barrier to market entry. ⁵⁷ Consequently, Chicago school scholarship denied the prevalence of most barriers to entry other than government regulations. ⁵⁸ This narrow conception of entry barriers meshed with newly developed economic theories, according to which markets can be competitive even if there is only one competitor, as long as potential market entrants are taken into account. ⁵⁹ The belief in perennial market contestability and the resulting unlikelihood that incumbents are ever capable of excluding competition has been bolstered by today’s digital economy, in which supposedly “competition is just one click away.” ⁶⁰

Finally, economic theories about two-sided markets and competition between markets or “ecosystems” have further fueled the view that domination is rare in today’s economy. In two-sided digital markets, (supracompetitive) profits from one side of the market are often used to cross-subsidize the other side of the market. ⁶¹ Such business models have led to concerns about the dominant positions of companies like Google vis-à-vis advertisers and application developers. ⁶² However, neoclassical antitrust doctrine insists that measuring market power requires taking into account both sides of the market. ⁶³ Thus, in Ohio v. American Express, the Supreme Court held that the plaintiffs had failed to demonstrate market power and lost the case, even though they had plausibly demonstrated evidence of anticompetitive effects. ⁶⁴ Along the same lines, some scholars have gone further by questioning the relevance of power within a specific market. ⁶⁵ Nicolas Petit for instance argued that even orthodox antitrust analysis inflates the degree of dominance of large technology firms because—even though they may be monopolists on specific markets—actual competition takes place between monopolists on distinct markets. ⁶⁶

To be clear, I do not want to engage substantively in the debate as to the merits of Ohio v. American Express or as to whether competition between ecosystems is more important than competition within specific relevant markets. My point is that neoclassical antitrust doctrine can immunize itself from principles-based—rather than empirical—challenges based on republican theory by denying the prevalence of power and domination in the economy.

In summary, the legal epistemology of neoclassical antitrust law is blind to many of the issues that are of concern to the Neo-Brandeisian program. The laissez-faire approach of neoclassical antitrust is not the result of the consumer welfare standard or the values and objectives of antitrust law. ⁶⁷ Rather, its central characteristic is the belief that domination is unlikely in the first place, which is caused not by its values and goals, but its epistemic priors and methodological commitments. ⁶⁸ Although, as noted above, neoclassical antitrust doctrine has not incorporated many of the Chicago School’s doctrinal prescriptions such as formalistic per se legality rules, ⁶⁹ and post-Chicago economic and econometric analysis has found its way into enforcement practice and (lower) federal courts’ case law, ⁷⁰ the lack of substantial antitrust enforcement in the past decades clearly shows that antitrust law and practice—rightly or wrongly—has been deeply skeptical of the presence of economic power and domination. ⁷¹

Consequently, the first challenge for Neo-Brandeisian antitrust is to respond to the following objection: Neo-Brandeisians have mainly attacked the consumer welfare standard, but their argument misses the key premise of neoclassical antitrust, namely that market power and domination are simply not as prevalent as Neo-Brandeisian scholarship asserts. The Neo-Brandeisian mistake that has made this counterargument possible is their focus on the consumer welfare standard and the values of antitrust, instead of the epistemology and methodology of measuring domination. ⁷² Re-emphasizing methodology, then, can address this challenge.

C. Republican Antitrust: From Theory to Methodology

In order to make their case, Neo-Brandeisians should target the specific epistemological priors and methodological commitments that neoclassical antitrust employs to measure power and dominance, and provide concrete alternatives.

Neo-Brandeisians, to be sure, have not only addressed underlying theories. Lina Khan has offered several proposals for specific doctrinal reforms, and a concrete list of Neo-Brandeisian doctrinal interventions has been provided by Tim Wu with the support of several colleagues. ⁷³ However, both Khan and Wu have not provided a coherent set of priors and methodologies that connect a republican conception of antitrust to these doctrinal interventions. They have rather confused their audiences by arguing for a departure from current doctrine, while at the same time using the technical language of neoclassical antitrust including the term “market power,” without making clear that they mean something else than power over price or output. ⁷⁴

The rejection of the consumer welfare standard and Wu’s and Khan’s proposed alternative of “protecting the competitive process” fail, in this regard, to do the work. ⁷⁵ It remains unclear how “protecting the competitive process” will distinguish between pro-competitive and anticompetitive practices, and how it translates into a different threshold for intervention. ⁷⁶ Furthermore, Chicago-inspired antitrust analysis will have no difficulty re-engineering the current doctrines based on the goal of protecting the “competitive process,” since nothing in the concept of protecting the competitive process prevents including potential competition, or an epistemic focus on market power as power over price or output.

Focusing on methodologies of measuring power and domination, by contrast, directly targets the prescriptions of neoclassical antitrust, whose main vulnerability is its narrow reliance on neoclassical price theory. For example, neoclassical antitrust analyses of the alleged dominance of companies such as Facebook or Google are often always based on the so-called “revealed preferences” of consumers, ⁷⁷ ignoring decades of cognitive and psychological research showing that consumer preferences are endogenous. ⁷⁸

Neo-Brandeisian proposals for a more interventionist antitrust policy could use this methodological weakness by appealing to different and arguably more sophisticated methods to measure power and to determine the capability of business practices to affect competition. ⁷⁹ Such methods may include behavioral economics, ⁸⁰ cognitive and psychological research, ⁸¹ empirical economics, ⁸² and complexity studies. ⁸³ Even mainstream economics, moreover, has offered concrete empirical support for the Neo-Brandeisian view that American markets are increasingly concentrated, and that domination is a pressing concern. ⁸⁴

Furthermore, Neo-Brandeisians can coherently maintain their claims without proposing theoretical changes regarding the goals or underlying values of antitrust. The consumer welfare standard is sufficiently elastic to capture Neo-Brandeisian proposals for doctrinal reform, as long as this standard is interpreted and implemented in view of a wider range of methods of measuring power. ⁸⁵ It should also be emphasized that “consumer welfare” can be indirectly protected through a structuralist approach, as recent case law from the Court of Justice of the European Union has illustrated. ⁸⁶

While emphasizing methodology over theory makes the Neo-Brandeisian project more viable, this strategy would require extensive use of interdisciplinary methods so as to challenge the mainstream, formalistic framework of antitrust analysis. However, if the point is not whether antitrust should be based on a “neoliberal” or a “republican” theory of liberty, but rather which epistemological priors and methodologies antitrust should employ to measure market power and anticompetitive effects, the turn from theory to methodology is necessary to make a plausible case for a Neo-Brandeisian antitrust agenda.

III. Polycentricity and Formalism

The second challenge for the Neo-Brandeisian agenda is to operationalize its key principles and objectives into workable rules that can be applied by courts, in a manner that both courts themselves and their interlocutors consider legitimate.

Neoclassical antitrust law is perceived to be a “technocratic” discipline, ⁸⁷ even though a longstanding tradition of scholarship has shown that antitrust doctrine is neither value-neutral nor a-political. ⁸⁸ Reliance on microeconomic language, however, has given antitrust an aura of objectivity, allowing courts and other legal decision-makers to avoid the impression of making political judgments. While the foundations of antitrust can never be “a-political,” neoclassical antitrust law internally functions in a relatively value-neutral manner. ⁸⁹

Neo-Brandeisians aim to disrupt current antitrust’s formalism by fundamentally challenging the core hypotheses of the neoclassical research program and its implications for antitrust. Their proposed shift toward a broader concern for domination and power includes a considerably larger number of relevant values and metrics. ⁹⁰ The formalistic world of neoclassical price theory is substituted with an empirical perspective on economic power. ⁹¹

The key question is how this more inclusive conception of antitrust law—which is not just preoccupied with “the best allocation of our economic resources, the lowest prices, the highest quality, and the greatest material progress” but equally with “the preservation of our democratic political and social institutions” ⁹² —can and should be operationalized. The exact manner in which Neo-Brandeisians aim to strike a balance between consumer welfare, the competitive process, the inherent harm of concentrated power, and republican ideals of democracy and egalitarianism, has remained unclear. ⁹³ Two distinct strategies can however be deduced from the work of, respectively, Tim Wu and Lina Khan.

A. The “Protection of Competition” Standard

Wu argues that the Neo-Brandeisian antitrust agenda can be implemented within the current antitrust framework. He proposes a “protection of competition as a process” standard that should replace the consumer welfare standard, so as to allow for a higher degree of antitrust intervention. ⁹⁴ Wu is remarkably optimistic about the determinacy of this new standard, arguing that it is equally, if not more precise than the consumer welfare standard. ⁹⁵

More specifically, Wu proposes a list of five questions (including several sub-questions) that an enforcer should ask in order to determine whether competition has been harmed. Multiple of these questions are notoriously open-ended. Whether “the firm appear[s] to have sufficient market power to actually affect the process of competition?” ⁹⁶ for example, is a question that Neo-Brandeisians and Chicagoans would answer radically differently, as shown in Section II.

Similarly, whether the conduct “is competition on the merits (i.e., a better or cheaper product) or a potentially illegitimate methods (sabotage, exclusionary deals, tying, predation, manipulation of a standards process, and so on)” ⁹⁷ is hardly helpful. The question assumes the illegitimacy of business conduct such as tying and exclusionary dealing in order to answer the question of whether this conduct is legitimate. ⁹⁸ Wu’s questions suggest determinacy, but expose widespread disagreement. This is even clearer with the question whether “the complained-of conduct or merger tend to implicate important non-economic values, particularly political values?” ⁹⁹ which brings antitrust enforcement squarely within the realm of political philosophy.

While Wu’s proposal for a “protection of competition” standard may indeed “capture far more of the dynamics of the competitive process then [sic] does existing analyses,” ¹⁰⁰ the flipside is that the competitive process is complex, polycentric, and as such irreducible to a simple decision-making procedure. Wu concludes by advocating for Brandeis’s formulation of the rule of reason. ¹⁰¹ That formulation, however, has been rightly criticized for requiring judges to decide each case on the basis of all circumstances, thus making nothing dispositive. ¹⁰²

Wu has also emphatically argued for a reinvigoration of the “big case tradition” of the breakups of Standard Oil and AT&T, and the cases against IBM and Microsoft. ¹⁰³ The fact that there have been few successful “big case” breakups in more than a century of antitrust history, however, indicates the stakes of structural remedies. ¹⁰⁴ Apart from whether Wu’s optimism as to the government’s ability to meet the required standard of proof is warranted, this strategy arguably is not normatively desirable. I agree, in this regard, with economist Paul Romer that in the absence of “clearly defined notions about what’s allowed and what’s not allowed and then consequences that follow from that,” courts should not decide on the suitability and proportionality of major behavioral and structural remedies, because “judges are just not in the position to make decisions like that.” ¹⁰⁵

B. Overturning Precedent

Khan advances a slightly different approach. While agreeing with Wu on a “protection of competition as a process” standard, ¹⁰⁶ Khan’s proposals are more formalistic. In contrast to Wu’s range of open-ended questions, she proposes simple rules and presumptions of illegality, which err on the side of intervention. ¹⁰⁷

Simple rules and presumptions are formalistic because they exclude a range of considerations from the decision-making process. ¹⁰⁸ Decision-makers should not decide on the basis of all considerations of the concrete case, as they typically would do in applying a “standard” instead of a “rule,” ¹⁰⁹ but rather based on the literal meaning of the rule and whether its conditions obtain. ¹¹⁰ Khan allows, however, for a possibility to rebut presumptions of illegality based on legitimate business justifications. ¹¹¹ This approach of a strict rule that errs on the side of over-inclusiveness, with the possibility for the relevant firm to justify its conduct, mirrors both the traditional approach to abuse of dominance in European Union (EU) competition law, ¹¹² and Kaysen and Turner’s classic statement of the Harvard antitrust school. ¹¹³

Introducing such formalistic rules requires overruling Chicago school-inspired precedents that had themselves overturned older case law. ¹¹⁴ Khan, for example, advocates a reinvigoration of the Philadelphia National Bank standard, ¹¹⁵ according to which mergers leading to a post-merger market share of more than 30 percent are presumptive unlawful, ¹¹⁶ and contends that the recoupment criterion for predatory pricing should be overruled. ¹¹⁷

While I sympathize with Khan’s proposals for specific, deliberately formalistic rules, ¹¹⁸ it seems unrealistic to expect the federal judiciary to overturn 40 years of Chicago school-inspired precedent by re-introducing formalistic rules and presumptions. Although precedent is certainly no insurmountable barrier, ¹¹⁹ arguably it was considerably easier for the Supreme Court to overturn the formalistic per se prohibitions of 1950s and 1960s case law than it would be to overturn current precedent today. Although it is widely recognized that the rule of reason—and similar doctrines in the context of section 2—are “little more than a euphemism for nonliability,” ¹²⁰ in theory current doctrine provides for inclusive, all-things-considered tests. ¹²¹ This means that the Supreme Court can only reinvigorate a pre-Chicago school approach by openly and deliberately substituting more formalistic rules for open-ended standards, which would itself be contrary to precedent. ¹²²

Since formalistic rules and presumptions are unlikely to be (re-)introduced through judicial law-making, Neo-Brandeisian antitrust likely requires legislative or administrative rule-making, even though this would in turn entail additional challenges of a political and constitutional nature.

IV. An Argument for Legal Formalism

Adding to the practical necessity of implementing Neo-Brandeisian antitrust by other means than through judicial law-making, adopting formalistic rules through legislative or administrative rule-making is also the best way to implement Neo-Brandeisian antitrust from a normative viewpoint, as this section aims to show.

It is important to note that while this section provides an argument for legal formalism, the argument itself is thoroughly functionalist and pragmatic: it builds, among others, on the pragmatic claim that the federal courts are unlikely to overturn decades of precedent, and the normative claim that especially congressional legislation would be the most legitimate way of implementing a Neo-Brandeisian antitrust agenda. While the political and constitutional challenges to a new and politically legitimated legal formalism in antitrust law should be recognized, such challenges do not undermine the normative and pragmatic reasons for taking this course.

A. The Desirability and Feasibility of Rule-Making

Chopra and Khan have argued that the FTC could use its rule-making power under section 5 of the FTC Act to adopt general rules banning, for instance, exclusive dealing and tying. ¹²³ Chopra and Khan argue that section 5 rule-making could specifically pertain to “unfair methods of competition” instead of business conduct subject to sections 1 and 2 of the Sherman Act, since FTC rules arguably may not run counter to the federal courts’ interpretation of the Sherman Act. ¹²⁴ This distinction could shield such rule-making from judicial review in light of Chevron deference. ¹²⁵ On January 5, 2023, the FTC indeed proposed, following Khan’s earlier work, a per se rule against noncompete clauses on the basis of section 5. ¹²⁶

Whether the federal courts will actually apply Chevron deference in this scenario, however, is far from certain. ¹²⁷ The kind of administrative rules that Chopra and Khan envision, especially as they run counter to the current interpretation of the antitrust laws, may be considered questions of deep “economic and political significance” running foul to the major question doctrine in its strong form. ¹²⁸

Alternatively, formalistic rules and presumptions could be adopted by congressional legislation. This is one of the pillars of the House Antitrust Subcommittee’s report on digital markets, which proposes to legislatively overturn numerous Supreme Court precedents. ¹²⁹

In view of the inherently political nature of several of the Neo-Brandeisians proposals, a legislative approach offers two clear advantages over administrative rule-making. First, from a pragmatic perspective, federal legislation can more expressly depart from the normative choices of Supreme Court precedent. Since administrative regulation cannot overturn precedents interpreting the Sherman Act, the exercise of section 5 rule-making will entail incongruencies between the Sherman Act and administrative rules on unfair competition, while these latter rules effectively aim to circumvent traditional antitrust rules.

Second, the question of how to interpret and apply the antitrust laws is controversial and polycentric. Various Neo-Brandeisian proposals make profound value choices by balancing, among others, efficiency and democracy, the benefits of economic size and its perils, and consumer convenience and a fair economy. ¹³⁰ They often require overturning decades of case law either explicitly or implicitly. Such reforms are—also from a legitimacy perspective—better implemented at the highest political level, mirroring the adoption of the Clayton Act 1914 and the Robinson–Patman Act 1936.

In sum, while the Neo-Brandeisian antitrust agenda is bold and stimulating, its holistic and polycentric nature poses major substantive and procedural challenges. Substantively, Neo-Brandeisian antitrust would introduce value-laden balancing exercises, which cannot be answered by a value-neutral decision-making procedure, at the heart of antitrust. Consequently, they also invite the difficult procedural question of which institution can legitimately implement and operationalize these proposals. ¹³¹ I agree with Khan that a Neo-Brandeisian antitrust agenda should center on simple rules aimed at operationalizing a complex and multifaceted conception of republican antitrust, which ideally would be adopted by Congress. Whether Congress will be able to pass antitrust legislation appears increasingly doubtful. ¹³² While this political reality cannot be ignored as a substantial challenge to antitrust reform, however, it does not detract from the substantive claim that new, formalistic antitrust rules are best adopted through congressional legislation, administrative rule-making being the second-best option.

B. Formalizing Methodology

Leaving aside political realities, the previous sub-section showed how legal formalism offers a solution for the complex and polycentric nature of Neo-Brandeisian antitrust. Legal formalism, however, can also be part of the solution to the indeterminacy of republicanism in regard to the technical content of antitrust doctrine, discussed in Section II above.

Insofar as current antitrust doctrine is indeed formalistic to a significant degree, ¹³³ a Neo-Brandeisian focus on actual power dynamics and the interrelationship between economic and political domination aims to rely more heavily on empirical analysis. While such a change in the methods of measuring power may seem primarily suited for a case-by-case approach, methodological change can also take place at the rule-creating level.

In the EU, for example, the EU legislature recently adopted the Digital Markets Act (DMA). ¹³⁴ The DMA represents both a substantive and a methodological departure from the existing European antitrust law framework. Substantively, the DMA centers on formalistic prohibitions: it neither requires a case-by-case assessment of anticompetitive effects, nor contains an efficiency defense. ¹³⁵ The DMA is also based on a methodological change in respect of measuring power. ¹³⁶ It does not require either market definition or evidence of dominance, in contrast to the prohibition of abuse of a domination position in Article 102 of the Treaty on the Functioning of the European Union (TFEU). ¹³⁷ Instead, providers of core platform services are designated as “gatekeepers” and become subject to the DMA primarily based on quantitative thresholds related to turnover and the number of users. ¹³⁸

Like substantive changes to the antitrust rules, ¹³⁹ a methodological change of the metrics of power and anticompetitive effect is better done at the rule-creating level—rather than at the level of applying rules to individual cases—for two reasons.

First, the current methodological status quo in antitrust law—in summary, microeconomic analysis and industrial organization economics ¹⁴⁰ —has been legitimized by more than forty years of federal courts precedent. Methodological change, therefore, requires a significant departure from administrative and judicial practice. ¹⁴¹

Recent events in both the United States and Europe indicate that courts are not necessarily receptive to such radical change. In the United States, two FTC cases against Facebook were initially dismissed because “[t]he FTC ha[d] failed to plead enough facts to plausibly establish [. . .] that Facebook has monopoly power.” ¹⁴² While the Court allowed the FTC to repair its defects, and subsequently allowed the case to move forward, it emphasized that the FTC “stumbled out of the starting blocks” and “may well face a tall task down the road in proving its allegations.” ¹⁴³ Meanwhile in the EU, the EU Courts appear to be moving against the Neo-Brandeisian tide toward substantially stricter scrutiny of administrative decisions based on mainstream neoclassical analysis. ¹⁴⁴

Second, the use of alternative, including possibly heterodox, methodologies and methods in individual antitrust cases is complex and resource-intensive. Such case-by-case assessments will likely involve value judgments in light of significant uncertainty. It is by no means guaranteed that, for example, experimental economics and cognitive psychology will provide complementary—rather than conflicting—perspectives in concrete cases. This practical difficulty adds to the legitimacy problem of allocating such complex decision-making power to administrative authorities, as discussed in Section III.

However, new methods for measuring power and domination can also inform rule-making, which will not only obviate the need to incorporate such methods on a case-by-case basis, but which will also politically legitimize such methods. The manner in which the DMA introduced a new way of measuring the power of providers of core platform services shows how legal formalism can effectively bring about methodological changes to regulation, ¹⁴⁵ albeit at the inevitable cost of over- and under-inclusiveness. ¹⁴⁶

This is why, in my view, legal formalism not only offers a legal solution to the polycentric nature of Neo-Brandeisian antitrust reform proposals, but also addresses the need to focus on epistemic priors and methodology rather than abstract theory and goals. Legislative and administrative law-makers are free to take into account a considerably wider array of possible informants and considerations, without being constrained by particular (allegedly) value-neutral rules and methods. ¹⁴⁷

Consequently, a Neo-Brandeisian antitrust reform agenda, if such is deemed desirable, is best implemented by using the legislative and/or the administrative process to adopt—in a democratically legitimated manner—a set of new legal formalisms that would (1) settle polycentric disputes and (2) incorporate new modalities of measuring power, domination, and anticompetitive effects through sufficiently clear and largely self-executing rules. These rules almost certainly will not be perfectly calibrated to the protection of either competition or consumer welfare, in the sense that any formalistic rule is inherently both over-inclusive and under-inclusive. ¹⁴⁸ But avoiding false positives and false negatives in individual cases is obviously not the point, for the polycentric objectives of Neo-Brandeisian antitrust may well be better attained through the aggregate application of formalistic rules. Arguably, this point is true for antitrust law in general. ¹⁴⁹

V. Conclusion

Neo-Brandeisian antitrust proposals have forcefully argued for a revolutionary reform of mainstream U.S. antitrust law and policy. Relying on empirical research showing that the American economy is increasingly concentrated and underperforming, ¹⁵⁰ Neo-Brandeisians have argued in favor of bolstering antitrust intervention and abandoning the consumer welfare standard in order to revive the republican tradition in pre-Chicago school antitrust.

Neo-Brandeisian antitrust proposals face two key challenges. First, Neo-Brandeisians have insufficiently realized that the core cause of the laissez-faire attitude of neoclassical antitrust doctrine is not the consumer welfare standard itself. This attitude instead results from the epistemic priors and methodological commitments of the neoclassical program, which entail a general skepticism about the prevalence of economic domination. In order to meet this challenge, Neo-Brandeisians should focus on epistemic priors and methodology instead of abstract normative justification.

The second challenge is that Neo-Brandeisian antitrust reform proposals appeal to a variety of mostly incommensurate values and objectives. Wu’s suggestion that a “protection of competition as a process” standard could be easily implemented by enforcers and courts underestimates the indeterminacy of his proposals, and the reasonable disagreements that they invite. By contrast, Khan’s proposals focus heavily on the introduction of new formalistic rules and presumptions. This strategy, which would ideally take the shape of congressional legislation, is more desirable, notwithstanding constitutional and political obstacles.

Current antitrust doctrine is in many ways highly formalistic and detached from empirical realities. This conclusion, however, does not necessarily warrant more case-by-case enforcement. New legal formalisms could reinvigorate antitrust law and policy, thereby returning to the Harvard school’s key insight that a plethora of purposes—including both efficiency and republican liberty—can be attained by simple, formalistic rules combined with an efficiency defense or legitimate business justifications. ¹⁵¹ Such formalism could, moreover, reflect different methods of measuring power relations and domination in the real-life world. Finally, formalistic rules, especially if adopted by Congress, are capable of effectively and legitimately overturning decades of Chicago school-inspired precedents, which otherwise likely prevent an overhaul of the current neoclassical paradigm in U.S. antitrust law and policy.