Market Veridiction and the Monopoly Paradox

At some point in the middle of the eighteenth century, classical political economy replaced law and moral reasoning as the “mode of rationality” that liberal government “fostered and upon which it depended” (Best Reference Best2018, 93; Brown Reference Brown2015, 57–58, 67–68). With political economy, truth replaced moral and juridical claims about the market. “The market no longer appeared as a site of jurisdiction” that marked off a realm of individual autonomy from the sovereign, lectured Foucault (Reference Foucault2010, 32–33). It was no longer a site of “justice” to assess whether the Prince’s actions were noble or corrupt. Instead, the market appeared as “something that obeyed” “natural” and “spontaneous mechanisms.” Attempts to modify those mechanisms could only “impair and distort them.” Nowhere was this more evident than in market prices, which expressed “true” relations between the “cost of production and the extent of demand.” As such, prices became a “standard of truth that enabled us to discern which government policies are correct and which are erroneous.” The market, in short, became a “site of veridiction,” which told truths that governments “must obey.”

It is important to understand that market veridiction is an historical concept meant to situate modern liberalism in light of what came before it. For all classical political economy’s claims to reveal market truths that governments must obey, Foucault explains how it also admitted the limits of its own logic when it observed that competition could lead to monopoly as easily as it could to natural prices. Liberal political economy, then, could not serve either as a schema that fully determined state conduct, a “policy paradigm” that structured official cognition (Hall Reference Hall1993), an “ideational blueprint” (Blyth Reference Blyth2003; Reference Blyth2013), a “totalizing paradigm,” or what Bevir (Reference Bevir, Bevir and Rhodes2010, 42; Reference Bevir and Bevir2016) calls a “quasi-structuralist episteme.”Footnote ⁴ As Rabinow (Reference Rabinow2011) writes, truths for Foucault are tools: pragmatic instrumentalities to help subjects find their way about in problematic situations or what Foucault calls “foyers d’experience” (liminal spaces). Subjects cannot help but draw on historical modes of veridiction—truths formed amid prior problems and situations—even as they innovate to address new circumstances. Uncovering the history of truths, the problems they were meant to solve, their historical contingencies, fissures, paradoxes, and pathologies is Foucault’s unique method of genealogy (see also Koopman Reference Koopman2009; Reference Koopman2017; Lemke Reference Lemke2019; Novkov Reference Novkov2011; Vucetic Reference Vucetic2011).

Nowhere was the fissured nature of market truths more evident than in the problem of monopoly, writes Foucault. For no sooner did classical political economy identify the price mechanism as a truth machine than it also acknowledged that monopolies undermined the truth-telling capacity of markets. Foucault (Reference Foucault2010, 64, 134–35, emphasis added) put it this way: “In the classical conception…of the economy, monopoly is seen as the semi-natural, semi-necessary consequence of competition… . Monopoly is an actual part of the logic of competition.” If governments want to “save competition from its own effects,” it “must act on…[those] economic mechanisms. This is a paradox.” The “monopoly paradox” ensured that market veridiction rested on a circular logic: governments were bound to respect the truth of natural prices only insofar as they also acted to prevent the monopolies that undermined natural prices.

The monopoly paradox, Foucault (Reference Foucault2010: 64–65) noted, ensured that neither autonomous markets nor political economy applied to state action could, in and of themselves, produce liberty: “Freedom in the regime of liberalism is a not a given, it is not a ready-made region which has to be respected…. Freedom is something which is constantly produced. Liberalism is not acceptance of freedom; it proposes to manufacture it constantly, to arouse it and produce it.” Nor does liberty derive from an ideal theory of markets. Indexed by a paradox wherein competition produces natural prices and monopolies—where markets are self-regulating and self-destructive—liberal government is an “art” that “entails at its heart a productive/destructive relationship with freedom.” The same might be said for markets as for liberty; that is, the market is not a “ready-made region to be respected.” It must be “constantly” nurtured by the “art of liberal government”: that is, an art that acknowledges both natural prices and natural monopolies.

A Genealogy of Neoliberalism

At the end of World War II, some German intellectuals began to worry that the monopoly paradox had undermined liberalism. Once an agreed-on exception to the rule of natural prices, now socialists saw monopoly everywhere. Monopoly was not a paradox for socialists; it was a fatal contradiction of capitalism. As such, it became a justification for nationalization, collective bargaining, redistributive taxation, and social welfare. If German neoliberals were to neutralize these threats to liberty, denaturalizing monopoly was a necessary first step.

As Foucault explains, if Hayek, Von Mises, and Ropke were to curb socialist threats to liberalism, they needed to show how competition did not result in monopoly. Hayek led the way. More often than markets, he charged, it was the state that granted monopoly privileges and politics that entrenched them against competitive challenge. To be sure, Hayek and his colleagues admitted that sometimes monopoly had technical causes (such as scale economies). But these were always temporary, insisted Von Mises, because they were inevitably challenged by technical progress and by new rivals that were encouraged by monopoly profits. In Foucault’s words (Reference Foucault2010, 136–37), the German neoliberals aimed to free liberalism “from…the handicap of monopoly,” remove the monopoly paradox from market veridiction, and set liberalism on a fresh and “utopian” foundation. Resolving the monopoly paradox once and for all was foundational to the neoliberal project.

Neoliberalism Comes to America

The Germans exported their efforts to denaturalize monopoly to the United States where they assumed a different form. Whereas the German response was institutional, the American response was epistemological. If governments, not markets, created and sustained monopolies, law and policy could be altered to reinvigorate liberalism, argued the Germans. An “ordoliberal” state would discourage monopoly and encourage a flourishing competitive economy of small and medium-sized enterprises. Carefully designed markets could prevent an entrenched proletariat, and demands for unionization, nationalization, and social provision would fade (Foucault Reference Foucault2010, 120–21, 176–79).Footnote ⁵

American neoliberalism took a different turn. To Foucault’s surprise, Americans looked to Viennese Positivism for a philosophically grounded defense of markets. Led by intellectuals at the University of Chicago, neoliberals sought to replace a putatively haphazard combination of ideologies, legal doctrines, and ad hoc economic principles with a general theory of economic behavior. In doing so, Foucault (Reference Foucault2010, 222–33) noted, the Americans shifted the domain of economic analysis from production, exchange, and distribution under various conditions to a more “general field of reference of economic analysis,” wherein the “object of study” attended to “the analysis of the internal programming of individuals’ activity…. For the [American] neo-liberals, economic analysis should [explain] the nature and consequences of…substitutable choices, that is to say, the study and analysis of the way in which scarce means are allocated to competing ends.” They aimed to build a general theory, founded on the psychology of human choice, that could be applied to any realm of life where scarcity reigned—from the market and politics to crime and the family (Bevir Reference Bevir and Bevir2016; Brown Reference Brown2015; Cooper Reference Cooper, Plehwe, Slobodian and Mirowski2020).

From the Monopoly Paradox to the Antitrust Paradox

German neoliberals exported their attack on the monopoly paradox to the United States. Formed in 1946 under Hayek’s guidance at the University of Chicago Law School, the Free Market Project turned its attention to denaturalizing monopoly; it was renamed the Antitrust Project in 1952. Under Robert Bork’s guidance, the Antitrust Project built a theory of markets that transformed a durable truth in classical political economy—the monopoly paradox—into a thinking error, a logical contradiction to be solved by a foundational theory of markets (Eisner Reference Eisner2010; Khan Reference Khan2020b; Sawyer Reference Sawyer2024; Shaviro Reference Shaviro2012; Teles Reference Teles2010; Van Horne Reference Van Horne2011).

The result of 25 years of scholarship, Bork’s Antitrust Paradox detailed confusions in twentieth-century antitrust and outlined a model for rationalizing law and policy. It is easy to read Bork, as do the Polanyians, as a champion of deregulation, free markets, or disembedded liberalism (Britton-Purdy et al. Reference Britton-Purdy, Grewal, Kapcynski and Rahman2020; Fligstein and Vogel Reference Fligstein and Vogel2020; Somers Reference Somers, Mackert, Wolf and Turner2021; Vogel Reference Vogel2024). That is true enough, but Bork was more ambitious. He sought nothing less than to save the liberal rule of law from caprice, bias, and incoherence and, by doing so, to save democracy from interest group capture and ideological prejudice. Ridding antitrust of the monopoly paradox and setting market veridiction on empirically verifiable logical foundations were the bleeding edges of his grand project.

Consider Bork’s brief for deregulation. Two problems plagued mid-century antitrust: ideology and interests (Bork and Bowman Reference Bork and Bowman1965). Ideologically, antitrust had always harbored the “loosely Jeffersonian” sentiment that “social and political health” was “engendered by a sturdy, independent yeomanry in the business world.” The patron saint of modern antitrust, Louis Brandeis, Bork (Reference Bork1978, 47) charged, “was not so much a believer in competition as a believer in safety and smallness in the economic world.” Interests also degraded antitrust. Like his colleague George Stigler (Reference Stigler1971), Bork argued that regulation was inevitably designed by those with the greatest stake in redistributing rents to themselves. In antitrust this meant that the most “politically potent…groups” used their influence to suppress, rather than encourage competition (Bork Reference Bork1978, 55). Add yeoman ideology to interest group domination, and the outcome was toxic. Antitrust “protected competitors not competition” (56). Bork’s solution was deregulation, to disembed markets from a democracy corrupted by interest groups and ideology by realigning antitrust from redistribution to efficiency. Armed with a theory of economic efficiency, antitrust would enable markets, rather than regulators or judges, to select the most efficient forms of economic organization.

Bork’s project was more ambitious than insulating the market from ideology and the “serfdom of interest group liberalism” (Lowi Reference Lowi1986–87). Armed with an economic theory predicated on the “internal rationality…of human behavior” (Foucault Reference Foucault2010, 223), Bork sought to reinvigorate liberalism by saving the rule of law from subjectivity (judge-made law), capriciousness (treating unlike cases alike), and imprecision (meaningless concepts). Antitrust became the cutting edge of that project. This required setting political economy on a solid epistemological foundation by showing how the monopoly paradox was not a market truth but instead was a logical error or, in Foucault’s idiom, a false mode of market veridiction. Reconceptualized on logical micro-foundations, or what Dewey in 1924 called “antecedent principles,” Bork would show how the monopoly paradox evaporated from antitrust doctrine. In its stead, he revealed an antitrust paradox, a policy at war with itself because it was so riddled with logical contradictions that it undermined the law’s promises to bring about prosperity, liberty, and democracy.

In Bork’s deductive critique, two associated principles of midcentury jurisprudence led judges and regulators astray: the doctrines of multiple goals and market fragmentation. On the first, antitrust statutes voiced social, political, and economic goals. Congress meant not only to ensure that monopoly did not undermine efficiency and low prices but also that it did not injure liberty and democracy. In the absence of a theory that explained how monopoly undermined these goals or how to arbitrate conflicts between them, jurists invented a proxy measure: market fragmentation.

Fragmentation, Bork (Reference Bork1978, 60–61) complained, was a “word without content.” Is the “atomization of society” a good in and of itself? Does it provide a uniform standard for assessing multiple goals when they conflict? Is fragmentation a means to protect rivalry or its end? Reasoning deductively from antecedent principles, Bork reconceptualized market structure as a result of competitive processes among rational rivals, instead of an a priori normative standard for markets (fragmentation). When driven by competition to allocate resources to their most efficient use, Bork reasoned that the most resourceful rivals will always win. The predictable result of competition, then, is the “elimination of rivalry”; that is, market concentration. Fragmentation was precisely the opposite of a correct marker for effective competition.

Still worse, the fragmentation standard provided litigants with perverse legal incentives, because it granted losers entitlements to attack successful rivals. In the absence of a deductive theory of markets, rooted in sound micro-foundations, Bork wrote, antitrust doctrine became “paradoxical.” A legal standard meant to “protect competition”—fragmentation—“protected competitors” from the discipline of market rivalry. Antitrust became a policy “at war with itself” (Bork Reference Bork1978, 60; Bork and Bowman Reference Bork and Bowman1965).Footnote ⁶ In Foucault’s language, Bork deduced a theory of market competition that transformed a durable feature of liberal market veridiction—the monopoly paradox—into a contradiction in formal logic, an antitrust paradox. A law meant to serve economic prosperity through competition undermined competition. In Bork’s analysis, the monopoly paradox was no longer a challenging problem to be ameliorated by the “art” of liberal government (Foucault Reference Foucault2010): it was an “antitrust paradox” that could be settled by a formal theory of law and thereby excised from antitrust doctrine.

To be sure, Bork conceded that jurists occasionally recognized that fragmentation was an inappropriate proxy for multiple goals in statutory law. In those cases, they proposed to adjudicate competing goals on a “case-by-case” basis. Only facts could arbitrate whether a market structure or behavior was consistent with statutory intent. In the absence of a theory that derived rules from general principles and facts from rules, Bork objected, this was a recipe for caprice or “judge made law.” Bork asked jurists instead to conceptualize the problem of monopoly from the perspective of the calculating monopolist, who was weighing the costs and benefits of market control. Consider, for example, the aspiring monopolists who engage in pricing below cost to drive rivals from the market. This strategy is never costless: they must consider whether they can recoup losses once they gain control. If they raise prices too high, they will draw new rivals into the market. The real question, then, is whether aspiring monopolists can recoup the cost of predatory behavior. Thus, instead of asking whether pricing below costs leads to monopoly (the absence of fragmentation), Bork asks jurists to deduce a determinant rule from general principles: Do aspiring monopolists raise prices to recoup the cost of pricing below costs once they gain market power? Logic, Bork concludes, replaces the fragmentation standard with the recoupment rule.Footnote ⁷ (As we shall see, it is precisely Bork’s account of the relationship between general principle, rule, and fact selection that neoBrandeisians criticize when they explain why neoliberal regulators were unable to perceive how Amazon monopolized digital commerce.)

If fragmentation is an incoherent proxy measure for multiple ends, then another way to evaluate trade-offs is necessary. Appealing to diverse values is useless, Bork objected, because they are incommensurate. Antitrust needs an objective “common denominator”: a uniform category to arbitrate competing interests or assess value trade-offs. In a commercial democracy, there is only one interest or value universal enough to serve that purpose: “consumer well-being.” Take the trade-off between the gains of breaking up a monopoly and potential losses to efficiency. If a judge were to measure the gains to smaller-scale rivals against the efficiency losses of an incumbent enterprise, the result would be arbitrary or political. Held to the universal standard of consumer welfare, however, jurists can compare losses and gains on a uniform metric. In the absence of a consumer welfare standard, calculating trade-offs is capricious, unpredictable, and indeterminant. Antitrust risked becoming nothing more than a battle for private advantage (Bork Reference Bork1978, 22–26, 83–115).

Bork’s Antitrust Paradox succeeded in excising the monopoly paradox from antitrust doctrine for a generation (Lynn Reference Lynn2009; Reference Lynn2017; Sawyer Reference Sawyer2024; Vogel Reference Vogel2024; Wu Reference Wu2018). Markets were no longer understood as ambiguous sites of self-regulation (natural prices) and self-destruction (monopolies). Conceptualized instead as sites of rivalry between rationally calculating firms, the problem of monopoly evaporated. The following section will show how neoliberal theory succeeded in transforming antitrust rules concerning predatory pricing, vertical integration, and the timing of antitrust prosecution, and why the neoBrandeisians turn to Legal Pragmatism to explain why those rules fail.

NeoBrandeisians on Markets

New Brandeisians conceptualize markets pragmatically, not deductively. Like the legal pragmatists before them, they are antidualists, who reject the metaphysical distinction between artificial and natural markets. Markets, in their view, are both legal constructions and innate sites of power and coercion. On the one hand, this means that markets reflect the organization of entitlements to property and exchange in common, statutory, and constitutional law. On the other hand, it means that neoBrandeisians are antifoundationalists, who—like Dewey, Holmes, and Brandeis—think that law is never unassailable. It is always possible for an enterprising monopolist to capture, entrench, or abuse market power by finding ways to manipulate, end-run, or invent new interpretations of the law. Antitrust, argue the New Brandeisians, is an appropriate response to this dilemma because it acknowledges both the possibility and fallibility of the law to protect markets from paralyzing concentrations of power. In William James’s (Reference James1907) idiom, antitrust is a meliorist orientation to an irresolvable dilemma: the monopoly paradox. It is easy to misinterpret the New Brandeisians as Marxists or populists if you see only their attention to power, or as Polyanians if you see only their attention to the embeddedness of markets in law. Once you see their perspective on markets in all its complexity and on the imperfect nature of antitrust, their pragmatism comes into focus.

Markets, write the New Brandeisians, are inexorably sites of power that tell truths that liberal democracies ignore at their peril. Far from a solution to the nineteenth-century legal formalist and twentieth-century neoliberal problem of coercion, writes Rahman (Reference Rahman2017, 10), markets are “domains of power and conflict, riven by inequalities of bargaining power, welfare, and position.” Teachout, Khan, and Rahman outline a framework to conceptualize market power and explain how it threatens prosperity, liberty, and democracy. Drawing on Progressive-era legal pragmatists Louis Brandeis, John Dewey, Robert Hale, and others, neoBrandeisians theorize markets as sites of “coercion, power, and social harm” (Rahman Reference Rahman2017, 54–77). Rahman distinguishes “dyadic” from “structural” power. Brandeis, he writes, diagnosed the former as domination by a “discrete, cognizable actor who intentionally exerts arbitrary influence on another.” Institutionalized in a “particular group of private actors,” dyadic power is manifested as the “capacity to dominate others…whether directly in the economy or indirectly in the policy process.” But it also manifests in an institutionalized “deficit” of power in “citizens.” In Brandeis’s time, citizen agency went unrealized in corporate hierarchies and the policy process because there were insufficiently “powerful channels” through which they could “contest the…power [of] economic elite[s]” (Rahman Reference Rahman2017, 82–83).

Today, write Teachout and Khan (Reference Teachout and Khan2014, 58), dyadic power is exercised in concentrated market structures through the power to regulate and tax. These powers manifest by governing “what is in a market” and by “self-regulat[ion]” (53–55). “Market power enables a company [or a cartel] to raise what it charges consumers and lower what it pays suppliers” (55–58). Such wealth extraction is better understood as a tax than a market transaction (see also Khan and Vaheesen 2017). Unlike public taxation, this form of coercion is unaccountable. Self-regulation occurs when one company or a small group of companies have sufficient market power to “set standards for the rest of the industry,” which affect a broad array of citizens. Here again, market power takes on the “form” of coercive “political governance” without accountability (Teachout and Khan Reference Teachout and Khan2014, 55).

Rahman (Reference Rahman2017, 83–84) theorizes a second form of market power—structural domination—in which “no single actor intends to create…an unequal social structure.” Rather, this power is the aggregate effect of each party seeking its own advantage against the background of established laws. Dewey and the legal pragmatists explained how putatively free markets could become “inequitable systems” of structural domination by exercising “instabilities, externalities, and risks that effected those far beyond the binary transactional relationships of private exchange” (Rahman Reference Rahman2017, 83–84). Teachout and Khan (Reference Teachout and Khan2014, 588) quote Brandeis: domination may exist, “although it is not manifested in any overt act, and even though there is no intent…. Words of advice, seemingly innocent and perhaps benevolent, may restrain [others], when uttered under circumstances that make advice equivalent to command. For the essence of restraint is power; and power may arise merely out of position.”

Although neoBrandeisians naturalize market power, unlike Marxists they do not think it is a fatal flaw of capitalist democracy. They are antidualists, who also think that markets are artificial creations of law and politics, and meliorists, who think antitrust is necessary and possible. “The New Brandeisians,” writes Khan (Reference Khan2018b, 132), “believe the political economy is structured only through law and policy.” The “economic order is fundamentally ‘embedded in social relations,’ and shaped by and subordinated to the social order,” adds Rahman (Reference Rahman2011, 277). “‘Natural’ market forces are the product of human agency.” Nowhere is the artificial nature of markets more evident than in the long history of antimonopoly law. Built on eighteenth-century common law, modern antitrust is the quintessential “predistributive” policy, aimed at structuring markets to check private power before it threatens prosperity, liberty, and democracy (Lynn Reference Lynn2020; Teachout Reference Teachout2020; Wu Reference Wu2018).Footnote ⁸

In short, like classical liberals and the legal pragmatists before them, neoBrandeisians think markets have two paradoxical qualities: they are natural sites of power and artificial creations of law, and their primary governing mechanism—competition—can be predatory or productive. Far from being fatal contradictions for liberal market democracies, these qualities are sources of meliorist vitality for neoBrandeisians. They signal the pressing need for antitrust as an imperfect but workable solution to the monopoly paradox. As Jonathan Kanter (Reference Kanter2023c) puts it, when open and competitive, markets “promote material progress,” “choice and opportunity,” and “democracy.” When predatory, monopolistic, and closed, “real people suffer real harm,” which threatens “the individual liberty of…citizens” and “undermin[es] the democrac[y] they hold dear.” The purpose of antitrust, Kanter (Reference Kanter2023a, emphasis added) has always been to ameliorate economic and political power. The Founders grasped this lesson; antimonopoly planks were written into state constitutions; and America’s “greatest political leaders—Frederick Douglass, Senator Sherman, Franklin Roosevelt, Senator Kefauver—understood this constitutional political economy. They knew, as FDR put it, that ‘freedom is no half and half affair.’ Freedom in…the marketplace and the polling place go hand-in-hand. These leaders leaned on the antitrust laws and a competitive economy as a solution, as we must today.”

Teachout and Khan agree. When efforts to regulate the political influence of private power fail, antimonopoly law has always served as a backstop for democracy. When the Supreme Court struck down campaign finance regulation in Citizens United v FEC, it unintentionally opened the constitutional door to antitrust. By declaring corporations immune from regulation because they are indispensable to the “architecture” of American democracy, argue Teachout and Khan (Reference Teachout and Khan2014, 71), the Court made it possible to pose the predistributive question about the way we craft markets: Does “excess corporate size, scope, and industry concentration undercut democratic self-governance in ways that lobbying, and campaign finance reform do not address?” This is the central political and economic problematic for antitrust.

Unlike neoliberals, then, neoBrandeisians think that there is no theoretical solution to the problem of private power that can produce definitive legal judgments. Power will always threaten markets, liberty, and democracy. But neoBrandeisians are legal pragmatists, for whom private power can be lessened sufficiently to make markets workable and compatible with liberty and democracy. Antitrust authorities must be vigilant, and the law must adapt to businesses’ endless inventiveness in finding ways to escape scrutiny by end-running the law. As Dewey (Reference Dewey, Fisher, Horwitz and Reed1924, 193) wrote, the law must be a site of inquiry and innovation, not mechanical jurisprudence (Pound Reference Pound1908; 1910), especially at moments of economic flux like the present. Antitrust, Brandeis said, must be a “living law,” in which empirical inquiry serves to clarify rules as much as rules guide inquiry. Nowhere is this lesson more pressing today, insist neoBrandeisians, than in the digital economy.

The Dynamics of Monopoly and the Limits of Neoliberal Antitrust

In 2017, Lina Khan published “Amazon’s Antitrust Paradox.” Arguably the most influential text in the neoBrandeisian canon (Horowitz-Ghazi Reference Horowitz-Ghazi2023; Streitfeld Reference Streitfeld2018), it demonstrated the limits of Bork’s promises to objectivity, consistency, and coherence in antitrust. It is easy to read “Amazon’s Antitrust Paradox” as a root-and-branch critique of neoliberal deregulation: the old antitrust rules would have prevented a digital monopoly. But Khan concedes too much to neoliberalism for this to be a satisfactory interpretation. Amazon, she admits, realized neoliberal metrics for consumer welfare. It built a mass market, kept visible prices low, and reduced distribution costs. But Amazon also invented technologies and business practices that created and concealed monopoly power from neoliberal scrutiny. Like Neil Fligstein’s seminal (Reference Fligstein1993) study of the historical relationship between corporate strategy and antitrust, which shows how managers repeatedly innovate to escape antitrust rules, Khan shows how Amazon gamed neoliberal rules and, by doing so, revealed the limits of Bork’s promise to legal determinacy. Like the legal pragmatists, Khan shows how, in a moment of economic flux, the formalist promise to deduce a determinant relationship between general principles, rules, and facts failed.

Amazon, writes Khan, enacted neoliberal principles with predictably positive results for consumer welfare. Focusing on building a mass consumer following, it grew online commerce at an extraordinary pace, integrated forward into a distribution system optimized by inventory and transportation logistics, and diversified into mass-market services. “Customers universally seem to love the company. Close to half of all online buyers go directly to Amazon first to search for products, and in 2016, the Reputation Institute named the firm the “most reputable company in America for the third year running.” At the same time, there is a “gnawing sense of that its dominance … may pose hazards,” as well as confusion about how a company that provides so many benefits threatens markets and democracy (Khan Reference Khan2017): “It is as if [Jeff] Bezos charted the company’s growth by first drawing a map of antitrust laws, then devising routes to smoothly pass them…. With its missionary zeal for consumers, Amazon marched toward monopoly by singing the tune of contemporary antitrust” (710).

Khan asks us to consider first Amazon’s practice of predatory pricing, or pricing below cost to drive rivals from the industry. Once presumptively unlawful, neoliberals argue that predatory pricing must be analyzed from first principles, which guide analysis to the subjective perspective of an aspiring monopolist who weighs costs against benefits or, in Foucault’s idiom, calculates “substitutable choices.” As Bork theorized, pricing below costs is only rational if a monopolist can recoup losses once rivals are driven from the market. The longer recoupment takes, the more costly it is, and the higher the recoupment price, the more likely competitors will enter and drive down prices. In truth, predatory pricing is irrational and rarely occurs (Khan Reference Khan2017, 722–31). Yet, neoliberals derive an empirically verifiable standard for antitrust action from this analysis: the recoupment rule. If firms price below cost and do not raise prices once they have attained market power to recoup their losses, then the threat of entry is in force. Prices and costs, in this foundational analysis, are precise, measurable, and definitive.

Look closely at Amazon’s dominance in the e-book market, urges Khan. Initially, pricing evidence demonstrates that, despite its market standing, Amazon did not set prices to recoup losses from pricing below costs. Instead, Amazon priced its e-book reader, the Kindle, and electronic bestsellers below cost to gain market dominance. Once it succeeded in dominating the e-book market, price increases were negligible. Both the DOJ and the FTC considered the charge of predatory pricing and dismissed it because it did not meet the recoupment standard (Khan Reference Khan2017, 711–17).

Khan, like Dewey, asks us to consider the recoupment rule as a hypothesis for inquiry, not as a decision rule. From this perspective, the neoliberal promise to distinguish salient from irrelevant facts fails. Recoupment measures obscure as much as reveal evidence. Although Amazon kept visible prices low, it designed technologies, legal instruments, and managerial practices to build and conceal power in the e-book market. First, Amazon established a property right called Digital Rights Management (DRM), which restricted certain e-book formats to the Kindle. DRM meant that Amazon could compel readers to purchase a Kindle by offering bestsellers below cost and then locking them into future purchases. Second, Amazon designed software to gather data on reader habits, which it used to tailor recommendations and product placement. Together, these innovations locked readers into the Kindle, even as competitors offered better products at lower prices. Moreover, Amazon invented data-harvesting software to customize prices, making it impossible to locate a uniform baseline from which to distinguish predatory from competitive prices. By selling in multiple markets, Amazon recouped losses from e-book bestsellers and on e-books facing little competition. Finally, Amazon used its vendors’ sales data to go into publishing itself and drive them out of business by privileging its own content (Khan Reference Khan2017, 756–68). In Dewey’s (Reference Dewey, Fisher, Horwitz and Reed1924) language, the relationship between general principles, rules, and facts is indeterminate, because situations are never exactly the same and because salient facts change during moments of economic change.

In short, deployed as a hypothesis for inquiry, instead of a “mechanically” (Pound Reference Pound1908) applied rule, the recoupment standard reveals how Amazon invented methods for “rolling up” (Lynn Reference Lynn2009) the e-book market. Like Brandeis, Khan poses three questions: (1) Was Amazon’s e-book strategy predatory or productive: What is the competitive process?; (2) Are better products driven from the market: What are the results for economic performance?; and (3) Can rivals compete, once Amazon achieved dominance: What is the resulting distribution of power in the industry? By answering these questions, Khan shows how Amazon gamed the recoupment standard by holding visible prices low and concealing predatory pricing from neoliberal scrutiny though data harvesting, cross-subsidization, vertical exclusion, and competing with platform vendors. Her antifoundational analysis, then, reveals the limits of a veridiction regime that promised determinant methods for evaluating predatory pricing.

Consider next the problem of forward integration, or the methods by which Amazon formed its own distribution system. Under pre-neoliberal doctrine, antitrust officials worried that vertical integration would lock out rivals or limit their access to necessary supply or distribution markets. When combined with market power in commerce or production, vertical integration could be used as leverage to bar entry at another level. From a neoliberal perspective, however, it is a mistake to see integration into distribution as a method to foreclose competition or leverage power. Analyzed from the perspective of an aspiring monopolist weighing substitutable choices, two questions emerge. Is it cheaper to contract for distribution or to do it oneself? And if a firm does integrate to undermine rivals, can it maintain control if it does not introduce efficiencies to ward off new competitors? Bork Reference Bork1978, 231, 234) answers that “predation through vertical merger is extremely unlikely” because “a manufacturer would not favor its retail subsidiary…unless it was cheaper to do so.” Then it would “pass those savings on to consumers” or face new entrants. “The law against vertical mergers is merely a law against the creation of efficiency.”

The trouble with Bork’s analysis, writes Khan, is that it is impossible to falsify. If vertical integration occurs in this theory, it must have been motivated by and achieve cost reductions. Suppose, instead, writes Khan, echoing Brandeis’s principles of pragmatist inquiry, we ask these questions: What was the distribution of power as Amazon built a distribution network, what competitive tactics did Amazon deploy (process), and what were the consequences for economic performance? From that perspective, writes Khan, the evidence is mixed. As Bork predicts, Amazon constructed an efficient delivery service through forward integration. It “established an extensive network of physical infrastructure: hundreds of warehouses, sorting centers, delivery stations…, and Prime Now hubs. [Amazon] is within 20 miles of 31% of the population and 20 miles of 60% of its core same-day base…. It expanded into trucking, container ships, planes, and drones, through which it speeds up how quickly it can transport goods to customers” (Khan Reference Khan2017, 777–78).

Neoliberal theory verified, however, does not eliminate the relationship between theory, rule, and fact of indeterminacy, writes Khan, because Amazon learned how to game neoliberalism’s veridiction (fact) regime. Amazon joined its efficient distribution infrastructure to three innovations, which entrenched its power over potential rivals and the vendors who use its platform: Amazon Prime, Fulfillment-by-Amazon (FBA), and Amazon Marketplace. Marketplace allows sellers to list their products on the platform, and Amazon then collects fees on their sales (Khan Reference Khan2017, 781). Third-party sellers’ share of total items sold on Amazon increased from 36% in 2011 to 50% in 2015. Amazon Prime asks consumers to pay an annual fee for free expedited shipping. Amazon bundled other deals with Prime, such as music streaming and faster deliveries. Fulfillment-by-Amazon (FBA) allowed merchants to store their products in Amazon warehouses and then provided vendors with packing, shipping, and customer services. Products sold through FBA were eligible for other advantages, such as faster shipping (776).

Coupled with Marketplace, Prime, FBA, and Amazon’s capacity to harvest data, Amazon’s efficient distribution system became an instrument of monopoly power at the same time as it served consumers by providing cheaper goods and faster service. Take Amazon’s power over independent vendors. Tying Amazon’s suite of services to its distribution system entrenches consumers and locks independent vendors into its platform. It also makes it harder for a rival platform to compete. Amazon’s capacity to cross-subsidize losses with revenues from less competitive products meant that it could keep vendors from using independent shippers by charging less. And by favoring its own access to distribution services, Amazon’s products outperformed those of independent merchants on distribution speed and quality (Khan Reference Khan2017, 779).

Amazon augmented its power over potential entrants and Marketplace vendors by designing software to monitor consumer behavior on sales, failed searches, and more. It then used that data to go into business for itself and outcompete independent merchants by exploiting its power over product placement, access to distribution, and Prime services. “The current antitrust regime,” Khan (Reference Khan2017, 783) concludes, “has yet to reckon with the fact that firms with concentrated control over data can systematically tilt the market in their favor.” Under such circumstances, firms seeking dominance are always tempted to game existing rules and facts to their advantage. As Dewey put it in 1924, precisely “because circumstances are really novel and not covered by old rules…shrewd and enterprising men are encouraged to sail close to the wind and trust to ingenious lawyers to find [the] rule under which they can get off scot free.”

In short, although Khan criticizes neoliberalism’s deference to vertical integration, she concedes that its predictions of distributive efficiency and some measures of consumer welfare are accurate. Nonetheless, neoliberalism mistakes partial evidence for the whole because it insists on a foundational theory, which selects facts that blind it to business’s ongoing incentive to innovate in ways that do an end-run around the prevailing rules (Fligstein Reference Fligstein1993).

This is not to say that neoBrandeisians think economic analysis is useless or ideological. Echoing legal pragmatist Roscoe Pound (Reference Pound1908, 609–10; 1910), who wrote that science should serve law rather than subordinate it to formal logic parading as science, Jonathan Kanter (Reference Kanter2022a) notes, “We must be mindful that economics—and expertise, more broadly—are merely tools to understand facts relevant to a particular case. Courts should use economics…to address questions of fact, not to resolve questions of law. And where there are natural experiments and direct evidence of competitive harm, economic theory must give way to market realities.” As Kanter (Reference Kanter2022c) explained to the New York Bar Association, “law enforcers are not central planners. At a certain point—in antitrust, and in every area of the law—we must exercise judgment.” It is an “unavoidable part of law.” The key is to make sure we “do it in a way that is reliable…and consistent with the statutes we enforce.” Reliability means fidelity to precedent. Consistency means that the “right debate” will always address Congress’s plain directive to “protect competition,” not to achieve consumer welfare (see also Kanter Reference Kanter2022b).

Timing, Rules, and Truth

Perhaps nowhere is the debate between neoliberals and neoBrandeisians over market veridiction and the monopoly paradox starker than in a dispute over the timing of antitrust action. Neoliberals caution humility: authorities should wait until markets reveal truths about alleged legal infractions before they act, because markets are less likely to make costly mistakes than judges and regulators. In contrast, neoBrandeisians advocate prompt action before market power becomes so entrenched it is hard to dislodge. They apply pragmatist critical methods to reveal analytical and empirical flaws in neoliberal reasoning and offer a meliorist justification for prompt action: although decisive action may sacrifice short-run efficiencies, it is necessary to protect liberty, prosperity, and democracy from concentrated private power.

In an influential 1984 article, which advocated waiting, neoliberal judge Frank Esterbrook wrote, “The entire corpus of antitrust doctrine is based on the belief that markets do better than judges or regulators in rewarding practices that create economic benefit and penalizing others. The common belief that if markets are imperfect then something else must be better is a logical fallacy. One need not pretend that markets work perfectly to see that they are better than judges at penalizing inappropriate conduct” (24). Antitrust authorities, Esterbrook went on, face a “fundamental difficulty” in assessing the effects of conduct. If they “err by condemning a beneficial practice, the benefits may be lost for good.” If they err “by permitting a deleterious practice, though, the welfare loss decreases over time,” because “monopoly is self-destructive” (1, 2). As Hayek and Bork explained, monopoly prices, restricted output, or sluggish innovation will always provide incentives for new rivals to enter the market. The monopoly paradox is an illusion. Thus, the cost of a false-positive decision, Esterbrook concludes, is undoubtedly higher than the cost of a false negative. Although this might look like a recipe for deregulation, Esterbrook does not argue that government should never act but only that it should wait until markets have revealed truth first. His prescription is humility in the face of epistemological fallibility.

New Brandeisians criticize Esterbrook’s argument for its conceptual fallacy, empirical error, and excising the monopoly paradox from antitrust. Like Dewey and Brandeis, they ask us to consider waiting as a hypothesis for inquiry, instead of a decisive rule. Three flaws of waiting become evident. First, historical evidence demonstrates that it can cause irremediable damage. Witness the Microsoft case, in which the court moved so slowly that Microsoft drove its rival Netscape from the market before the court found the defendant guilty (Reback Reference Reback2009). Second, the more entrenched that market power becomes, the more likely it will be translated into political influence over judges and regulators. Put differently, history confirms the monopoly paradox: it is impossible to sustain the rules necessary for open markets in the face of private power (Teachout and Khan Reference Teachout and Khan2014). Third, like classical liberals and legal pragmatists, neoBrandeisians argue that the more entrenched monopoly power becomes, the less likely are markets to provide useful information to guide state action. In Foucault’s idiom, the monopoly paradox belies Esterbrooke’s waiting prescription, because markets can only produce truths (e.g., natural prices) that governments must obey if governments act to prevent the monopolies that undermine the capacity of markets to produce those truths.

Contrary to Esterbrooke’s advice, neoBrandeisians ask regulators to act quickly before monopoly becomes entrenched. In a striking reversal of twentieth-century regulation, which was governed largely by general rules—or what antitrust doctrine called “rules of reason”—Rahman and Khan (Reference Rahman, Khan, Abernathy, Konczal and Milani2016, 20–21) call for a “policy in favor of simple rules and presumptions over ‘rule-of-reason’ analyses.” Antitrust needs more “per se” standards, wherein “certain practices are presumed to be illegal without requiring plaintiffs to make significant showings. This would involve blocking problematic deals, rather than seeking to mitigate harms through conduct remedies or divestitures” (see also Kanter Reference Kanter2022c). Legislation should “change the ‘unreasonable restraint’ standard to an ‘abuse of dominance’ standard…, allowing for greater scrutiny of the business practices of market-dominant firms” (Rahman and Khan Reference Rahman, Khan, Abernathy, Konczal and Milani2016, 23). Likewise, Khan and former FTC commissioner Rohit Chopra, called on the commission to use its rulemaking power to set “clear rules” of conduct to “help deliver consistent enforcement and predictable results. Reducing ambiguity about what the law is will enable market participants to channel their resources and behavior more productively” (Chopra and Khan Reference Chopra and Khan2020, 357).

NeoBrandeisians justify legal certainty by pragmatic (practical), not theoretical, necessity. They want to impose per se rules—with bright-line empirical standards (such as quantitative limits on market share)—to ameliorate predatory behavior and market power, not to solve the monopoly problem once and for all. NeoBrandeisians draw on a long record of per se rules in public law as a precedent for antitrust. Teachout’s (Reference Teachout2014) history of corruption law explains how the framers of the Constitution used per se rules to preclude activities perceived as so threatening to liberty and democracy that they warranted absolute prohibition: “Structural or prophylactic” rules “do not require proof of intent to convict. They work through changing incentives before the fact instead of punishing activity after the fact” (4–5). “Clearly defined,” “bright line” rules are “made up of objective elements, with little room for different interpretation” (think “speed limits”). At the very least, they make definitive government action possible. At best, they take on “moral weight” that protects the “social fabric” (4–5, 184).

In antitrust, per se rules are meant to be applied before predatory conduct does irreparable damage. The risks associated with the timing of governmental action are precisely the reverse of neoliberal humility. Where Esterbrook calculates the risks of a false-positive decision as undeniably higher than those of a false negative, neoBrandeisians assess the risk of a false negative in some instances as categorically unacceptable. Structural rules, writes Teachout, are imperfect: they are a meliorist trade-off. They inevitably “cover innocent activity as well as insidious transactions.” The Founders included a per se residency requirement for political office in the Constitution, because they thought “adventurism” so threatened republican government that it was necessary to preclude “potentially brilliant and important candidates.” “Corruption,” they said, “cannot be made to vanish”; it can only “be subdued” (Teachout Reference Teachout2014, 4–5, 47). Likewise, predatory conduct and market power cannot be made to vanish—they can only be ameliorated.

The historical record, write the New Brandeisians, belies neoliberal theory: more often than not, power begets power, not destabilizing competition. Structural rules warrant prophylactic injunction, even if that may mean setting aside short-term efficiency gains. Markets cannot realize gains from trade, generate innovation, cultivate liberty, or disperse power in service to democracy unless threats to competition are prevented (Kanter Reference Kanter2023b; Reference Kanter2023c). Nor can markets reveal usable truths about prices, costs, output, or innovation when monopolies become entrenched.Footnote ⁹ NeoBrandeisians, in short, demand legal certainty by fiat, not, as neoliberal formalists do, by claims to logical necessity.

To be sure, the New Brandeisians are legal pragmatists who observe that all rules are fallible. They concede that some predatory behaviors or monopolistic market structures will escape scrutiny and some regulatory or judicial actions will fail. Sometimes, action will be necessary after the fact: antitrust can always dismantle concentrated power through divestiture. But here too Rahman and Teachout (Reference Rahman and Teachout2020) argue that per se rules will dismantle power more decisively than the rule of reason. Finally, even where neoBrandeisians call for professional discretion, as in the implementation of their new merger guidelines, they also advocate extensive rulemaking, with clear distinctions of legality and illegality (US Department of Justice 2023).

The prevailing interpretations of the New Brandeisians cannot make sense of their demand for per se rules because they cast the movement as a return to twentieth-century regulation. However, that era was marked by general statutory rules, which granted judges and regulators enormous discretion in applying them to facts (Berk Reference Berk2009; James Reference James2000; Lowi Reference Lowi1979, 92–126; Orren and Skowronek Reference Orren and Skowronek2017; Peritz Reference Peritz1996; Sanders Reference Sanders1999; Sawyer Reference Sawyer2018). Indeed, it was precisely this sort of discretion that neoliberals like Bork opposed. Responding to the neoliberal demand for legal accountability, neoBrandeisians use pragmatist methods to reach different conclusions than their Progressive mentors. Rules should be precise and discretion limited.

Seen through Foucault’s genealogy, the debate between neoliberals and neoBrandeisians over the timing of state action is better understood as a contest over market veridiction and the monopoly paradox than an ideological dispute over regulation. According to neoliberal theory, in time competition will always destabilize monopoly. Antitrust authorities who act prematurely ignore this truth at prosperity’s peril. As legal pragmatists, neoBrandeisians reconceptualize waiting from a decision rule to a hypothesis. Inquiry, they conclude, demonstrates that power begets power, and only prompt action, enabled by bright-line rules, can protect markets from domination and stagnation.