Market Governance as a Balance of Power

This essay conceptualizes market governance as a balance of power and discusses the implications for current debates over antitrust policy. This framework offers a way to interpret and evaluate the “neo-Brandeisian” school that views concentrated market power as a threat to democracy as well as to economic goals, such as productivity and innovation. It suggests that the government can deploy antitrust policy to alter the balance of power to promote the public welfare without necessarily impeding competition or otherwise distorting markets. And antitrust policies that constrain market power can have the double benefit of making both markets and politics more competitive.

The proponents of a "neo-Brandeisian" approach to antitrust took on leadership roles in the Joe Biden administration, including Lina Khan, chair of the Federal Trade Commission (FTC); Jonathan Kanter, head of the Antitrust Division at the Department of Justice (DOJ); Tim Wu, special assistant for technology and competition policy at the National Economic Council; and Rohit Chopra, director of the Consumer Finance Protection Bureau (CFPB). The Biden administration issued a sweeping executive order to bolster competition in the American economy. 1 The FTC and the DOJ launched a public inquiry into strengthening merger guidelines in January 2022, and the FTC issued a statement declaring that it would use its full statutory authority to combat anticompetitive behavior in November 2022. 2 Democrats in the US Congress sponsored a series of high-profile hearings on the power of big tech firms and proposed significant new antitrust legislation. Republicans called for more robust antitrust policy as well, although they tended to be more concerned about the power of the big tech firms to control social media and other sources of information than about their market power per se. The House of Representatives prepared legislation to limit the ability of dominant platform firms to favor their own products and services, and to require them to open their app stores to rival marketplaces. Yet the big tech firms and other incumbents fought back hard.
The "neo-Brandeisian" school refers to a broad range of scholars and practitioners who critique the Chicago school of antitrust that focuses on a narrow "consumer welfare" standard to render judgments on merger cases and to evaluate dominant firm behavior, and favors a cautious approach toward antitrust enforcement. 3 The neo-Brandeisians call for a more robust antitrust policy to combat concentrated market and political power. 4 They have garnered this label because they share the concern of the legendary Supreme Court justice Louis Brandeis (who served from 1916 to 1939) with market concentration as a threat to democratic politics. 5 As discussed in more detail below, the neo-Brandeisians tend to advocate much stronger policies and more proactive enforcement to combat market power, to highlight the distinctive anticompetitive features of the digital platform firms, and to argue that antitrust should address the concentration of political as well as market power.
The critique of the Chicago school begins with an appreciation for the central role of power in markets. Markets require rules, including the rule of law and the protection of private property. Even the most ardent advocates of free markets would concede that. 6 Hence there is no "free market" in the literal sense: markets are always governed. 7 Governments do not govern markets alone, but share that role with firms, societal organizations, and individuals. Hence market governance spans from government regulation to business practices to social norms. Governments play the central role because they have the unique ability to create and enforce laws and regulations over a given territory. I refer to this role as "marketcraft" because it constitutes a core function of government roughly comparable to statecraft. 8 The proposition that markets are governed may seem fairly obvious, perhaps even banal. In this essay, however, I contend that this simple proposition logically begets some less obvious implications. Specifically, market governance is not only inevitable-but it is not neutral. That is, any regulation or practice or norm will favor some market participants over others, even if that was not the primary intent. Thus market governance not only defines a market, but it shapes the balance of power among market participants. If we imagine a spectrum of market governance options from those most favorable to incumbents to those most advantageous to challengers, we cannot identify a specific point along that spectrum that represents the free market or even a neutral position. That in turn suggests that it may be possible to alter that balance of power to promote public welfare without impeding competition or otherwise distorting markets. Many standard economic models, especially those of the Chicago school, bestow an undeserved positive valence upon the status quo by beginning from an assumption of perfect markets and characterizing any government action as an "intervention" that requires justification by reference to a specific market failure. 9 Antitrust drives home this point because it demonstrates how competitive markets are not the product of an absence of government regulation, but rather the outcome of specific types of government action. After all, markets require more than property rights and the enforcement of contracts to function properly. Businesspeople may prefer to collude rather than compete if left to their own devices. They may boost rents (returns not justified by value added) for themselves rather than add value for their customers. Or they may deceive customers or investors rather than disclose key information. So the government has to compel them to compete, to provide value to consumers, and to provide the information that fosters dynamic markets. A market liberal would counter that governments can also constrain competition, foster corruption, or mislead consumers-and they would be right. But governments remain the critical force that can foster and sustain competitive markets, and press firms from rent extraction toward value creation. The alternative to government regulation (including antitrust) is not the free market but real-world markets thoroughly sullied with collusion, fraud, and imbalances of power. The choice is not whether to regulate but how: how effectively, toward what goals, and in whose interests.
The antitrust realm is distinctive because it explicitly confronts the issue of power in the sense of the market power that dominant firms exercise in a given market. It also addresses the relationships between assembly firms and their suppliers, retailers and their suppliers, and employers and their workers. It grapples with the thorny questions of how to determine when that power is problematic and what to do about it. And that in turn highlights broader questions of how to conceptualize power in the study of political economy and in the practice of policy. 10 In this sense, antitrust is an important element of what has been called the "predistribution" policy agenda, as opposed to the redistribution agenda of progressive taxation and welfare spending. 11 The predistribution policy agenda deploys market governance and public investment to affect how the market distributes returns in the first place rather than allow markets to operate and then redistribute after the fact. The predistribution agenda is more foundational because it seeks to alter the power relationships that generate economic inequalities in the market itself. Formulating predistribution policies requires a sophisticated understanding of how power is embedded in markets and how it is deployed in the political arena.

The Balance of Power
I propose to conceive of market governance as a balance of power ( Figure 1). To illustrate what that means, let us begin with a simple example from telecommunications regulation. When government officials first introduced competition into the telecommunications sector in the United States and Britain, they had to increase regulation, not decrease it. 12 They introduced "asymmetrical" regulation, meaning regulation that favored challengers over incumbents. The incumbents had such an overwhelming advantage because of their control of the network that this was the only way to induce new entrants. 13 The authorities not only had to force the incumbents to lease their lines to their competitors, but they also had to do so on terms favorable to the competitors. The incumbents and the challengers naturally advocated different ways to calculate the interconnection charges. The incumbents preferred historical cost accounting, whereby the charges would compensate them for their long-term investment in building the infrastructure. The challengers preferred marginal cost accounting, whereby the charges would only reflect the additional costs for the incumbents to provide access to the challengers. At its core, however, this was not a strictly technical debate. The government had to decide how much to favor the challengers over the incumbents. Economic theory could not give them a precise interconnection charge that would reflect the free-market price, or even a neutral rate. They were left with a regulatory dial, which they could turn in one direction or the other.
A sports analogy offers another way to illustrate the same basic point. In baseball, the batter is out with three strikes and walks to first base with four balls. Those numbers strongly affect the competitive balance between the pitcher and the batter. If the rules gave the batter only two strikes for an out, that would favor the pitcher. If the batter could walk with three balls, then that would favor the batter. The architects of the game cannot opt out of these determinations. They cannot just say, "Well, let's just let each pitcher and batter decide the rules for themselves." And they cannot rely on some theory to tell them what would be a fair or neutral set of rules. They have to set the bars somewhere.
The telecommunications interconnection example represents a simple case of market governance as a balance of power in the sense that the government can manipulate one variable to alter the competitive balance between incumbents and challengers in a fairly predictable manner. In most realms, however, the relationship between government action and the balance of power is more complex. For example, corporate governance shapes the balance of power between market actors, but the effects on market outcomes such as prices, profits, and wages are less straightforward. If corporate law requires a certain number of independent directors (those not otherwise employed by or affiliated with the company) in large companies, those directors would be more likely to focus on financial returns to shareholders and less likely to consider benefits for stakeholders such as workers-if all else were equal. Yet the relationship between corporate law and the employer-worker power balance varies considerably depending on the broader legal and societal environment. 14 The corporate governance example illustrates the core argument of this essay even better than the telecommunications one because it incorporates power in a deeper way. In the telecommunications case, the regulatory authority shifts the advantage in one direction or another in a fairly mechanical manner, thereby affecting the prices, profits, and market shares for incumbents and challengers. In the corporate governance case, however, the causal chain from market governance (laws, practices, and norms) to market outcomes operates via power relationships as key intervening variables. That is, market and social and political power reinforce each other. This underscores the neo-Brandeisian case that we should not artificially separate out the analysis of market power from that of political power.
If market governance is a balance of power, that suggests that we should incorporate political power into market analysis and market power into political analysis. In the language of social science, we have to recognize the endogeneity of power in market relations. We risk obscuring the essential nature of markets if we do not recognize that supply and demand are themselves shaped by politics and society from the outset. Integrating power into the supply and demand equation opens up the possibility of a more interdisciplinary political economy, one in which political and social variables are incorporated not as external to the workings of the market but as integral to it. Social scientists outside of economics have much to offer by showing how political and social variables affect the market outcomes that concern economists the most, such as prices, profits, and wages. With some notable exceptions, however, they have tended to focus more on political outcomes, such as policies, or social outcomes, such as interpersonal networks or personal well-being. 15 Moreover, we cannot really understand market power without recognizing how it is connected to political power: how market power begets political power, and how political power begets market power. 16 Incumbent firms can achieve profits and stability via two strategies. 17 They can seek (1) to raise productivity (create wealth) and/or (2) to secure rents (extract wealth). Incumbent firms may prefer the latter because they may be vulnerable to challengers that develop a product or process that renders their competitive advantage obsolete. They may seek to insulate themselves from this threat rather than to rely on their ability to outrun the competition. They may prefer collusion to competition. To put this differently, firms can focus on building a better widget or manipulating market governance to maximize stability and profits with the widget they already have. In practice, of course, they do both. But they can gain security by ensuring that they will survive even if they cannot always produce the best widget at the best price. And they are best able to maximize profits if they are able to supplement any price or quality advantage they might have with the exercise of market power to deter, absorb, or impede competitors.
Firms may pursue rent extraction via two interrelated mechanisms. First, they deploy political strategies, lobbying government to create laws and regulations to protect them from competition, stabilize prices, and secure profits. This might include market governance, trade protection, subsidies, or tax credits. Firms with greater market power naturally have an advantage in this political competition. And they might deploy that market power to promote legislation not aligned with the interests of the electorate. 18 They rely heavily on their "instrumental power": their advantage in financial and organizational resources. But they also benefit from "structural power" because incumbent political leaders depend on business to support the performance of the economy to improve their chances for reelection. 19 Second, incumbent firms employ business strategies to increase their rents. For example, they can structure their business relationships to moderate competition via long-term alliances, interlocking shareholding, or exclusive supply chains. They can exploit their advantages in bargaining power to squeeze their workers, suppliers, borrowers, other business partners, and/or customers. They can manipulate their boards of directors to increase executive pay. They can engage in anticompetitive practices, from exclusive dealing to outright collusion; they can use intellectual property strategically to stifle challengers; or they can pre-empt competition by buying out potential challengers. Antitrust policies are essential tools to press firms from rent extraction to value creation.

The Antitrust Debate
Market governance may reflect a balance of power, but this does not necessarily mean that these are zero-sum relationships. The telecommunications example above may at first appear to present a fairly straightforward zero-sum situation: higher interconnection charges benefit the incumbent while lower rates benefit the challengers. Yet interconnection rates also raise public-interest concerns: How should we balance between the public welfare benefits from the economies of scale of the incumbent versus the benefits from the competitive challenge from new entrants? Which will produce more innovation: the continued dominance of the incumbent or the rise of the challengers? And should the authorities be concerned that market dominance could translate into disproportionate political power as well?
Antitrust debates have centered on precisely these sorts of questions. In the progressive era, the proponents of antitrust remedies were concerned about the concentration of political power as well as market power more narrowly defined. They felt that the big trusts had too much of both, and they relied on antitrust policy as a mechanism to rein in this power. 20 In the postwar era, the debate centered more on the economic costs and benefits of market power. It juxtaposed two different market visions: a procompetitive stance that favored activist government to sustain and promote competition, and a laissez-faire approach that advocated government restraint even if this means less competition. 21 This debate took the form of a rivalry between the "Harvard school" and the "Chicago school." Harvard school scholars focused more on the structural attributes of a market, such as firm size and product differentiation. They tended to view monopoly and oligopoly as more durable than their Chicago school counterparts, and to support more government action to remedy these market structures and the practices created and perpetuated by them. They would be inclined to rule against mergers at lower market shares and in less concentrated markets. The Harvard school shifted partway toward the Chicago school in the 1970s, however, relaxing its concern with entry barriers and vertical integration, and favoring a detailed examination of firm conduct prior to endorsing government action. 22 Scholars in the Chicago school contended that monopolies tend to be fragile and competition robust because firms that attempt to charge monopoly prices are likely to be challenged by new competitors over time. They asserted that most practices that might be deemed "exclusionary" are in fact procompetitive because they reflect firms' efforts to lower prices or to innovate. In any case, the government may be incapable of devising an appropriate remedy, or it may be captured by political interests, such as small businesses. Hence these scholars were reluctant to prescribe government action even where they found that a firm dominates a market or engages in anticompetitive practices. 23 With the ascendance of the Chicago school in the 1980s, US antitrust policy shifted from defining competition by the number of competitors to economic analysis centered on consumer welfare: low prices, high output, and potential for innovation.
The rise of the Chicago school represents one element of the broader neoliberal turn in ideology and policy since the late 1970s. The term "neoliberal" can refer to anything from a stage of capitalism to a market fundamentalist doctrine, and from the "free-market" policies of Thatcher and Reagan to the center-left agenda of Bill Clinton and Tony Blair. 24 Here I restrict the term to a market-liberal ideology shared by theorists such as Hayek and Friedman and political leaders such as Thatcher and Reagan. By this definition, neoliberals advocate less government intervention and more market freedom-as if the two are necessarily associated. They regard the market as an arena of freedom (voluntary exchange) and government regulation as a constraint on that freedom. And they view markets as spontaneously evolving rather than actively constructed, and market behavior as natural rather than learned. The irony is that in practice the neoliberal reformers deployed the language of free markets-such as deregulation, privatization, and liberalization-as smokescreen for a program that rigged market governance in favor of the wealthy and the powerful (upward redistribution). 25 The Chicago school approach to antitrust provided an intellectual rationale for policies that allowed market power and political power to reinforce each other over time.

The Neo-Brandeisians
Critics of the Chicago school have gained prominence in recent years in both scholarly debates and policy circles. They have responded to two developments in particular: the slow rise in market concentration in the United States over several decades and the more recent and rapid emergence of dominant digital platform firms such as Amazon and Google. Mergers boomed in the 1980s, including strategic acquisitions to strengthen core competencies and hostile takeovers designed to boost shareholder returns. 26 Thomas Philippon argues that the decrease in competition in the US economy since the 1990s has led to higher price markups, higher after-tax profits, lower investment, lower productivity, a lower labor share of income, and higher inequality. He stresses, moreover, that the stakes are huge. He estimates that if the economy were as competitive as it was twenty years ago, GDP would rise by 5 percent and income for American workers would increase by $1.5 trillion. Hence strengthening antitrust would deliver higher economic benefits than most any other proposed policy reform. 27 The Obama administration (2009-17) turned some attention to this issue late in its second term. A study by the Council of Economic Advisers found that market concentration increased in many core sectors from 1997 through 2012, with the largest increases in transportation, retail trade, finance and insurance, wholesale trade, real estate, and utilities. 28 Meanwhile, the digital platform firms have posed particular challenges for antitrust because they enjoy powerful network effects and winner-takes-all market dynamics; they operate their own market infrastructures; they do not compete on consumer prices; they benefit from distinctive new forms of patient capital; and they control massive amounts of data.
The neo-Brandeisians vary among themselves, but collectively they present at least six core arguments. First, they emphasize market power over economic efficiency. 29 They challenge the Chicago school's focus on a consumer welfare standard defined by low prices and high output. They argue that antitrust policy should consider the interests of workers, suppliers, entrepreneurs, and citizens as well as consumers. 30 They offer several alternatives to the consumer welfare standard. Lina Khan, for example, calls for a return to a more structural approach that centers more on a firm's market share and power in the marketplace than on prices and output. 31 Tim Wu argues that the primary goal should be the protection of competition. 32 And Sandeep Vaheesan suggests that the core purpose should be to disperse power and promote fair competition. 33 Second, the neo-Brandeisians charge that the Chicago school is too lenient on monopoly. Whereas Chicago school scholars believe that dominant firms tend to grow by providing better goods and services and lower prices and contribute to market efficiency via economies of scale, the neo-Brandeisians emphasize that these firms stifle competition and extract rents. They are particularly concerned with dominant firm (anticompetitive) behavior, not only size. And whereas Chicago school adherents stress that dominant firms drive innovation, the neo-Brandeisians emphasize that these firms impede innovation by blocking the entry of new challengers. In practice, the neo-Brandeisians advocate more proactive action to contain and even break up dominant firms, more zealous enforcement against anticompetitive practices, and more rigorous merger review.
Third, the neo-Brandeisians characterize the rise of digital platform firms as a novel phenomenon that requires distinctive policy and enforcement remedies. This contrasts with more established antitrust scholars who contend that these firms exhibit network effects that have been common in earlier periods and that can be addressed with a traditional antitrust toolbox. 34 The digital revolution poses particular challenges for the antitrust regime because of the powerful network effects in the information technology sector. Consumers want goods and services that interoperate with what they already have, or with what others have, and they will pay more for those goods and services. So market leaders can eliminate competitors by leveraging these effects. In fact, the neo-Brandeisians stress that the digital platform firms achieve their dominant position not simply by offering better products and services but by engaging in anticompetitive practices. In an influential article on Amazon, Lina Khan stressed several distinctive features of the platform firms. 35 The digital platform economy is particularly conducive to monopoly because the platform operators set the terms of competition on their platforms. Amazon, for example, operates a retail marketplace and sells on that marketplace at the same time. That gives it the ability to structure the market rules in its own favor. The big tech firms are also distinctive in that they do not compete on prices for consumers. Facebook, for example, does not charge most customers for its service, but garners the bulk of its revenue from advertising. So the antitrust authorities cannot use consumer prices to measure anticompetitive behavior. The big tech firms also enjoy access to patient capital, because investors do not insist on profits in the short term, and that enables firms like Amazon to operate without profits for years while they outlast the competition. And they use their control over data to boost their market power and consolidate their dominance. Some neo-Brandeisians conclude that the antitrust authorities should impose structural rules, such as prohibiting platform operators from competing on their open platform. 36 Fourth, some neo-Brandeisians are concerned not only about limiting the power of dominant firms but also about protecting the countervailing power of association among weaker actors. They not only believe that antitrust laws should not apply to labor unions, but they also contend that antitrust should support the ability of small businesses or groups of professionals to engage in economic coordination. Sanjukta Paul argues that antitrust policy privileges the coordination rights of business firms while unduly restricting other coordination mechanisms such as labor organizations and the public coordination of markets. 37 Likewise, Sandeep Vaheesan argues that antitrust policy should be less hostile to horizontal coordination among less powerful actors-such as workers, professionals, and small firms-because collusion among these actors can represent socially desirable cooperation. Reconstructing antitrust in this way would transfer power from corporate executives to these other actors. 38 Fifth, some argue that the authorities should consider the impact of antitrust on a much broader range of policy outcomes, such as inequality, supply chain resistance, and climate change. Khan and Vaheesan, for example, contend that a revived antitrust movement could help to reverse the rise in inequality in the United States. 39 And scholars have increasingly focused on the impact of antitrust policy on labor markets, the power balance between capital and labor, and the impact of this balance on wages, benefits, and working conditions. 40 Sixth, as stressed throughout this essay, the neo-Brandeisians contend that antitrust policy must concern itself with the concentration of political power as well as market power. Dominant firms can leverage their market power into outsized political influence, and this can undermine democracy. K. Sabeel Rahman argues that market governance, including antitrust, should counteract domination, whether that be in the form of corporate power or inequitable markets. 41 And Barry Lynn, one of the architects of the neo-Brandeisian movement, claims that concentrated market power has created a new autocracy that undermines democracy in the United States. 42

Antitrust and Democracy
If antitrust policy should address political goals as well as economic ones, then how should we define those goals? In this special issue, Katharine Jackson and Samuel Bagg provide distinctive answers. They both contest any sharp distinction between the political and the economic realms and stress that antitrust policy has an important role in ensuring the quality of democracy. They also emphasize the importance of the balance of power among actors in the market and in politics, consistent with the argument of this essay. Jackson maintains that all trusts (corporations) are enabled by law, so the government has a legitimate claim to regulate them. But on what principles? She proposes that we should evaluate antitrust policies in terms of their contribution to equal liberty: the egalitarian distribution of economic associational rights. On the one hand, individuals should be able to exercise associational freedom in their economic activities, including the formation of corporations. On the other hand, the rights of corporations should not preclude the rights of others to associate, to secure social and economic resources, or to preserve their autonomy from the corporation. This can be problematic because corporations may have a power advantage relative to other actors, such as workers: the ability to coerce, to exclude, or to leverage greater resources. So equal liberty in practice requires a balance of power-and antitrust law offers an important tool to achieve that balance. It can contribute to a balanced distribution of rights, including the ability for challengers to compete, consumers to choose products, and workers to choose employers. And consistent with one of the threads in the neo-Brandeisian school, Jackson suggests that it can also protect associational rights for small businesses, farmers, retailers, workers, and consumers. Jackson recognizes that to design antitrust policy to promote equal liberty requires complex judgments about how policies and enforcement affect balances of power in particular contexts. Yet she stresses that these difficulties should not be allowed to undermine the constructive potential of an equal liberty approach to antitrust. Samuel Bagg reaches similar conclusions by different means, framing antitrust policy, and the balance of power it helps to bring about, as demands of democracy. While neo-Brandeisians presume that limits on market concentration enhance democracy, Bagg notes that the most prominent theories of democracy focus more on processes of collective will-formation and decision making and remain hostile or indifferent to an expansive conception of antitrust. Yet Bagg stresses that a broader egalitarian balance among social forces is even more fundamental to democracy than any particular procedures. This perspective clarifies the democratic value of antitrust policy, and enables a response to two common concerns about neo-Brandeisian approaches. First, those on the market liberal or libertarian side might worry that expanding the purview of antitrust beyond the consumer welfare standard to more "political" concerns would only invite more capture, allowing antitrust policy to be manipulated for partisan gain. Bagg recognizes this risk but stresses that antitrust is not any more vulnerable to capture than other types of regulation and that antitrust is itself an important tool to prevent capture. Second, radicals and socialists might fear that antitrust could undermine their goal of centralizing democratic control over economic activity in the long term and undercut the power of labor unions in the short term. Yet Bagg argues that antitrust policy is fundamentally allied with labor organization and other projects of building countervailing power, and he stresses that large firms are not necessarily easier to unionize.
Brian Callaci takes up the second of the two concerns discussed in Bagg's essay in further detail. Specifically, he notes that many union leaders and scholars sympathetic to labor believe that large companies and oligopolistic market structure were integral to the high-wage political economy in the postwar era. These companies achieved economics of scale, invested in human and physical capital for the long term, and shared their oligopoly rents with their workers. Callaci disputes this account, stressing that some unions-such as those in the garment, coal, and trucking industries-stabilized competition without oligopolistic industry structures. In the first part of the twentieth century, the garment and coal workers unions played a central role in market governance, moderating competition while setting standards for wages and working conditions. Then, in the "Golden Age" (1945-74), workers at large firms were able to bargain for higher wages, benefits, and working conditions-but they failed to attain a prominent role in firm management or market governance. The Teamsters Union, however, followed the garment and coal unions' playbook and positioned itself as the coordinating force in an atomized industry. Meanwhile, giant corporations in the manufacturing sector increasingly "fissured" the workplace by outsourcing labor to third-party contractors. 43 While antitrust enforcement was relatively assertive in the 1950s and 1960s, it was unable to rebalance the relationship between capital and labor because it was geared more toward regulating corporate behavior than restructuring markets. Over the long run, Callaci concludes, high market concentration undermined labor power by establishing corporations as the sole governors of their markets, dampening competition and thereby undermining productivity, and denying workers influence over corporate and market governance. Thus, he maintains that trade unionism and robust antitrust policy complement rather than conflict with each other as components of a more egalitarian political economy.

The Digital Platform Economy
The essays by Gerald Berk and AnnaLee Saxenian and by Zephyr Teachout build on the neo-Brandeisian analysis of the digital platform firms, but advance the debate in new directions. Berk and Saxenian argue that promoting innovation in the era of cloud computing requires a reconceptualization of antitrust from combating market power to fostering a market infrastructure conducive to open innovation. They depart from both the neo-Brandeisian view of Tim Wu, who advocates aggressive antitrust measures to rein in the big tech firms, and the market liberal view of Tyler Cowen, who argues that these firms face plenty of competition and they should not be punished for providing better services for lower prices (often for free). 44 Berk and Saxenian focus on cloud services, juxtaposing two contrasting trajectories for the cloud ecosystem. In the autarkic model, exemplified by Amazon Web Services, the platform firm forces the independent firms that use its services to accept its terms, competes directly with those firms, and appropriates the open-source database innovations for its own purposes. In the collaborative model, exemplified by Google Cloud, the platform firm collaborates with start-ups, contributes to open-source projects, and partners with open-source institutions such as the Linux Foundation. The collaborative model fosters an architecture of participation that enables thousands of individual contributions to software code to foster faster, more reliable, and more flexible interoperable innovation. Viewed in this way, the goal of antitrust policy should be to shift the platform firms from the autarkic model toward the more collaborative one. Antitrust litigation can play a role in shifting power from autarkic platforms like Amazon to independent database companies. The FTC should also foster a more collaborative ecosystem via rulemaking to promote interoperability. The antitrust authorities should shift the focus of pre-and post-merger reviews from market concentration, start-up rates, or the protection of independent software developers in cloud computing to promoting a cloud ecosystem that will spur open-source contributions and increase innovation velocity. The government can also promote a more favorable ecosystem by investing in open-source institutions, thereby encouraging cloud providers to partner with open-source companies and to collaborate with the open-source foundations.
Zephyr Teachout envisions a different sort of antitrust challenge in the digital era. She is concerned about the ability of employers to leverage digital technology to monitor and control workers, including their workplace behavior and their political voice. She fears that monopoly power and data collection will combine to undermine working conditions and democracy via what she calls "surveillance wages." By this she means that employers will leverage the data they collect to individualize compensation and benefits for workers in ways unrelated to tasks or responsibilities. For example, employers can monitor worker behavior such as speech, social connections, or online interactions. And they can run experiments on their workers to refine labor incentives, lower wage costs, and maximize returns. Teachout sees early signs of this phenomenon in the labor practices of dominant platform firms such as Uber and Amazon, but she predicts that these strategies will proliferate rapidly in employment practices for low-wage workers. This will lead to downward pressure on wages and benefits, an erosion in worker solidarity, and a chilling of political speech and organization. She views antitrust policy as a promising mechanism to combat this trend. Like others in the neo-Brandeisian camp, she does not view the core goal of antitrust policy to be to approximate perfect markets but rather to decentralize economic and political power. Specifically, she proposes that government authorities should give greater weight to labor market monopsony in antitrust policy and enforcement. And she recommends that they ban first-degree wage discrimination (differential pay not based on differential work), either for all firms or for those with a dominant position in a given labor market.

Rethinking Antitrust Policy
The neo-Brandeisians offer a powerful critique of the Chicago school focus on the consumer welfare standard, but they do not provide a single coherent alternative for how to assess the harm from market power. This leaves us with three critical questions regarding the substantive scope, the measurement of costs and benefits, and the political independence of antitrust policy. First, if we are going to expand the analysis beyond consumer welfare, then what do we expand it to? Second, if we are going to assess the public interest in a way that transcends consumer welfare, then how do we evaluate (measure) those public-interest costs and benefits? And third, how do we ensure that antitrust policy and enforcement operate in the public interest and not in the private interest of the incumbent firms?
With respect to the scope of antitrust policy, Carl Shapiro challenges the neo-Brandeisian case, arguing that antitrust policy is ill suited to address other public policy goals such as income inequality or job creation. 45 Antitrust authorities should stick to their core task. They are better able to focus on antitrust concerns because they bracket out broader issues and leave those to other agencies. Yet the government cannot opt out of many judgments that affect the balances of power among market actors. So the withdrawal (or abdication) of the government is not a solution. Government authorities have a responsibility to consider the full range of the consequences of their policies, and antitrust is no exception. And antitrust policy has implications for a broader array of public-interest goals. Market concentration has a major impact on local communities and quality of life. For example, the consolidation of the retail sector has decimated local business districts, including bookstores and other small retail establishments. In recent years, economists have focused more on labor market monopsony, suggesting that antitrust policy should consider the impact on workers as well as consumers. They have found that labor market power has contributed substantially to wage inequality and economic stagnation. 46 Likewise, one could argue that antitrust authorities should also consider some of the most pressing challenges of our time, such as the coronavirus pandemic. Critics of the US government response to the pandemic have emphasized how weak antitrust enforcement allowed market concentration that compromised resiliency in the face of the pandemic. The United States suffered supply chain challenges in part because it relied on a small number of suppliers for protective equipment and ventilators.
With regard to measuring costs and benefits, moving away from the consumer welfare standard implies a shift from a relatively narrow economic calculus to a more "structural" approach to antitrust policy. But that does not mean that a neo-Brandeisian antitrust policy cannot benefit from economic analysis. We must recognize, however, that the conception of market governance as a balance of power challenges the way that economists use the concept of "rents" to measure the distance from competitive market outcomes. This practice is both useful for understanding real-world markets and potentially misleading. It depicts markets as contests among market actors (such as capital and labor) for returns; it shows how firms and workers can generate returns above value added; and it offers methods for estimating the scale and distribution of those rents. Yet the concept of rents also builds on perfect markets as a reference point. Then how do we make sense of this concept if there are no perfect markets? If there are no perfect markets, that implies that there is no absolute scale for what constitutes a rent. There is only a spectrum of power balances from most favorable to capital to most favorable to labor. This still leaves room for a productive analysis of rents in a relative sense. We could compare one mode of market governance to another (situated on points along the lines in Figure 1, for example), or to the status quo, rather than to a perfect market. We could compare two different power balances rather than comparing a state of imbalance relative to an imaginary equilibrium devoid of power relations. In fact, economists' measures of rents, including market concentration, retail markups, and Tobin's Q, would still be quite useful in this world of rent relativity. 47 They could not meaningfully compare a given situation to a state of nature, since that does not exist. But they could compare how incumbents have more market power in one situation than in another, for example, and they could use those measures to estimate the difference. Likewise, the recognition of the relativity of rents does not undermine the concept of rent-seeking more broadly, which then simply signifies the effort to shift the rules in one's favor, such as toward incumbents or employers or shareholders.
And with regard to political independence, the skeptics of the neo-Brandeisian position might argue that antitrust policy would be more vulnerable to political capture if it expanded in scope, and especially if it addresses the concentration of political power. Yet there is no shortcut to the kinds of political and administrative reforms required to combat political capture, and in fact a more rigorous antitrust policy would provide one mechanism in that effort. As Jackson (this issue) argues, antitrust policy requires the careful balancing of complex trade-offs, but that does mean that the authorities should forego this challenge and focus on narrower goals. And as Bagg (this issue) notes, antitrust is not uniquely vulnerable to political capture, so we should deploy the same vigilance to impede capture in this realm as in many others. More fundamentally, vigorous antitrust policy is a counterbalance against political capture.
Fortunately, these concerns about substantive scope, the measurement of costs and benefits, and the political independence of antitrust policy may be less daunting in practice than in theory-especially in light of the balance of power framework introduced here. For in all of the policy arenas discussed above, the base policy recommendation would be to shift the market governance power balance in the direction of challengers relative to incumbents. So making some headway would not require a precise economic analysis of every relevant market but only administrative reforms and policy changes to facilitate this shift. At the administrative level, for example, Congress could increase funding for the antitrust agencies to enhance their ability to conduct the kind of in-depth research to identify anticompetitive practices by digital platform firms. Or it could expand the authority of the agencies, or shift the burden of proof in antitrust cases from plaintiffs to defendants. 48 On the policy side, the antitrust authorities could tighten merger enforcement. In particular, they could apply tougher standards to mergers that could lessen competition in the future, even if they do not do so right away. 49 Sandeep Vaheesan proposes some specific measures to shift the balance of power: prohibit exclusive dealing and other exclusionary contracting by dominant firms; ban below-cost pricing by near-dominant firms; prevent incumbents from deploying financial power as a competitive weapon; and deem violations of other laws, such as environmental and labor laws, as unfair competition. 50 Likewise, the authorities may want to favor more structural rules to address the power of the platform firms, such as prohibiting them from providing a market infrastructure and competing on that infrastructure at the same time.
If market governance represents a balance of power, and there is no balance that represents a free market equilibrium or a neutral position, that suggests it may be perfectly reasonable to alter market governance to shift the power balance between incumbents and challengers, or among other market actors, to pursue public welfare goals. And antitrust policies constitute one of the core mechanisms to accomplish that. The markets-as-power-relations perspective also has profound implications for the study of political economy more broadly. It suggests that economists should bring power more to the center of their models of supply and demand and their understanding of market outcomes such as prices, profits, and wages. Political scientists should focus more on market power and how it distorts the balance of political power. And both should probe further into the ways in which market power and political power reinforce each other, and how a more rigorous antitrust policy might moderate that cycle.