In a new volume chapter, Shai Agmon and Samuel Bagg argue that academic and policy references to “competition” often fail to distinguish between competition’s many forms. Their disaggregation of competition into two complementary processes—parallel and friction competition—helps to clarify the neo-Brandeisian approach to competition policy and its advantages over the traditional consumer welfare approach.
The design and regulation of market competition was once a central political struggle. In the late nineteenth and early twentieth centuries, battles over monopolies and antitrust stood at the forefront of electoral campaigns, fueled fiery rhetoric, and at times, mobilized broad social movements. Competition policy was not a technical subfield. It was a live question about economic power and democratic control.
Over the course of the twentieth century, that struggle was gradually displaced. Decisions about how markets should be structured—like so many other questions of economic governance—migrated into the domain of technocratic management. Antitrust became the province of lawyers and economists. The political contest over market design receded.
This trend has recently begun to reverse. Drawing inspiration from one of the original architects of twentieth-century antitrust, Louis Brandeis, prominent neo-Brandeisians—among them former Federal Trade Commission Chair Lina Khan—have revived a long-dormant strain of antimonopoly thinking. Their ambition is to make competition policy political again. Pointing to rising market concentration and unfair competitive practices across the economy—especially by dominant firms in the increasingly indispensable technology sector—they have criticized the legal standards and analytical assumptions that helped stabilize and depoliticize competition policy in the twentieth century.
The legal, political, and economic merits of a neo-Brandeisian approach remain hotly contested—including on the pages of ProMarket. One thing, however, commands broad agreement: the goal is to promote competition.
What is striking is how rarely the term itself is examined. What exactly is the phenomenon that “competition” policy is meant to promote? And how, precisely, does it generate the social benefits so often attributed to it?
In our chapter for a new volume on Democracy and Competition, we argue that much turns on the answers to these questions. Clarifying the underlying concept will not resolve the dispute between neo-Brandeisians and their critics on its own. But it does take us closer. Once we identify which kind of competition we are trying to promote, familiar policy alignments begin to shift. Measures typically described as “pro-competition” turn out to undermine the very process they claim to protect, once we have the right kind of competition in view. Conversely, regulations often criticized as anticompetitive may, in a different light, appear as necessary preconditions for sustaining competition.
Two concepts of competition: is the market like sprinting or boxing?
The first step is to recognize that nearly all of the social benefits commonly attributed to competition, across an enormous range of contexts, can ultimately be attributed to one of two distinct processes. In some cases, the point is for competitors to maximize their performance entirely in parallel: i.e., without any interference (or “friction”) between them. Well-designed “parallel” competition thus carves out zones of non-interference for each competitor; protected lanes in which they have maximal freedom from one another to attempt their task. Think of the physical barriers that separate swimmers or sprinters.
“Friction” competition, by contrast, relies on some mutual interference. Sticking to sports analogies, we might think of basketball, or boxing, since their logic depends fundamentally on mutual interference between the competitors. A more instructive example is the adversarial legal system employed by the United States and many other nations with British legal heritage. Each party in such a system is incentivized to fight zealously for their own side—and not simply by developing good legal arguments of their own, but also by directly challenging or refuting the other side’s claims. Rather than securing a space of non-interference for each party, adversarial legal systems are designed to encourage certain socially beneficial forms of meddling: namely, those which protect individual rights (by minimizing false convictions) and help to uncover the truth (through procedures like discovery and cross-examination).
So, what about the market? On standard accounts of market competition, the incentive to maximize profits leads firms to produce valuable goods and services more efficiently, which in turn stimulates innovation in technology and organization. As such, the broader social benefits of markets are thought to arise from the aggregated parallel efforts of many independent firms to develop the most efficient ways to produce the goods and services that people value most highly. It is natural to presume that market regulators should assume the same posture as referees in other parallel competitions—like pure contests of athletic or artistic skill—whose social benefits arise entirely from their ability to elicit and reward certain forms of human excellence. In such cases, referees have a fairly simple task: prevent all friction between competitors, while otherwise interfering as little as possible, so that each competitor has maximal freedom to pursue greatness within their own lane.
This minimalist approach to regulation makes perfect sense in cases of pure parallel competition. Its influence has spread far beyond that context, however, shaping how competition (and competition policy) is understood across widely varying spheres of economic competition, political contestation, legal procedure, and even competition over ideas. When we call for “more competition” or seek to make competition more effective, we tend to imagine expanding zones of non-interference: increasing competitors’ “freedom,” securing their separate lanes, and insulating them from one another.
We contend that this dominance is mistaken. Frictions, particularly those introduced by regulators, deserve a more central place in how we think about markets and other forms of institutional competition. Many of the social benefits of market competition do arise from parallel mechanisms. But the central place of markets in our economic and political life makes a minimalist regulatory posture deeply inappropriate—not only within traditional competition policy, but well beyond it.
In praise of friction, part 1: a minimalist approach does not protect parallel competition
One reason market competition differs is that some degree of friction is simply inevitable. In a pure parallel contest, where all interference can feasibly be prevented, the only way for competitors to maximize their chances of victory is to improve their own performance. The contest is structured so that any strategy pursued by purely self-interested competitors will end up serving the socially beneficial ends of the competition.
Not so in markets. While firms can pursue profits by engaging in socially beneficial forms of innovation, they may also choose socially harmful strategies, such as externalizing costs, evading labor standards, or over-hyping their products to generate short-term windfalls. More to the point, they may also employ predatory pricing, exploit platform power or information asymmetries, lobby for favorable regulation, and pursue a variety of other tactics designed to undermine their competitors. Regulators can and should try to make these socially harmful strategies unprofitable, but in the extraordinarily complex context of the modern global marketplace, they cannot feasibly eliminate these temptations entirely.
What this means is that even if we assume the benefits of markets result entirely from parallel competition, a minimalist approach to market regulation—one focused on carving out zones of non-interference—will fail to protect those benefits. Rather than attempting to achieve the impossible, regulators aiming to protect parallel competition may do so more effectively by encouraging certain forms of friction.
How might this work in practice? Think back to contemporary struggles over the aims of antitrust or “competition” policy. According to the still-reigning “consumer welfare” approach, market concentration—and the behavior firms may have used to attain it—merits punishment only if it can be shown to result in inflated consumer prices. After all, large and even monopolistic firms can often be quite efficient, and if the point of market competition is to generate growth and innovation, it makes sense to evaluate firm behavior based on these outcomes, rather than assuming all market concentration is inherently suspect.
By contrast, neo-Brandeisians insist that the concentrated economic power wielded by extremely large firms is inherently dangerous, regardless of immediate price impacts, because it increases their ability to pursue negative-sum profit-seeking strategies, including those that involve meddling with their competitors. Instead of the minimalist approach to competition policy implied by the consumer welfare standard, neo-Brandeisians argue that antitrust should be guided by the much broader goal of antimonopoly: i.e., contesting highly concentrated economic power as such.
We can understand the neo-Brandeisians’ criticisms as internal to the model of parallel competition, whereby the relevant benefits of markets arise entirely from firms’ parallel efforts. Under this interpretation, neo-Brandeisians simply claim that the consumer welfare approach to antitrust fails to sufficiently protect parallel competition from frictional interference. Instead of assuming that we can prevent such interference while remaining oblivious to the relative strength of the various competitors—as we might when refereeing a track-and-field event—the alternative approach they suggest assumes that some friction between competitors will occur no matter what we do, and thus adopts the aim of preventing any participant from having dangerous or pervasive advantages when it does.
Firms that do not enjoy outsized power, vis-à-vis their rivals, are less likely to switch from productive innovation to unproductive interference in the first place. And when they do, it is less likely to threaten the larger benefit-generating process of parallel competition—at least in part because their rivals can respond in kind, thereby nullifying the competitive advantage such interference might otherwise bring. As such, the signature proposal of neo-Brandeisians is an aggressive approach to antitrust, which aims to prevent any competitor from gaining the kind of market power that would allow it to interfere effectively, and in socially unproductive ways, with their rivals’ ability to compete. Ideally, this will keep all competitors in their own lanes, and if not, it will keep the negative impacts of their interference to a minimum. Neo-Brandeisians seek to prevent the most socially harmful frictions—those employed by dominant firms to shut out competitors—by encouraging certain power-dispersing frictions instead.
In reality, aggressive antitrust is just one tool in the neo-Brandeisian arsenal. Their more general aim is to sustain a roughly egalitarian balance of power. In their early days, for instance, antitrust laws were often used—against the wishes of their architects—to criminalize the organization of labor unions as monopolies of workers. And even though a “labor exemption” to rules about marketplace collusion has since been carved out, the imperatives of labor and antitrust are still widely understood to be in tension with one another. In emphasizing the danger of concentrated economic power as such, by contrast, a neo-Brandeisian approach makes it perfectly clear why both aggressive antitrust action and aggressive labor organizing ultimately serve the same goal. More subtly, antitrust laws are still employed to criminalize various forms of coordination among small businesses that are perfectly legal when internalized within a single corporate structure. As enforced, therefore, antitrust laws function to privilege hierarchical organization and the consolidation of economic power over more horizontal and cooperative forms of organization. A neo-Brandeisian approach focused on the dangers of concentrated economic power gives us strong reasons to eliminate or reverse these advantages.
In both of these cases—by encouraging labor unions and allowing other forms of coordination among weaker market actors—the implication of a neo-Brandeisian approach is that regulators should limit the dangers of the objectionable, but inevitable friction that will take place between market actors by ensuring that none of these actors have power that is lopsided enough to be able to protect their position through socially harmful forms of interference. On the one hand, this means intervening directly in the marketplace to ensure a degree of balance among economic actors, using tools such as antitrust enforcement. On the other hand, it also means structuring markets in ways that indirectly foster certain forms of friction—for instance, by facilitating coordination among workers and smaller businesses. Rather than seeking in vain to prevent friction between economic actors altogether, a neo-Brandeisian approach suggests that the best way to protect parallel competition is to encourage and facilitate power-dispersing frictions—on reasonably egalitarian terms—among large firms, between large firms and coalitions of smaller firms, and between firms and unions.
In praise of friction, part 2: there is more to life than parallel competition
The second point to highlight here is that market competition has many other significant consequences, beyond those associated with parallel competition. Once we add these considerations into the mix, we have further reasons to reject minimalist approaches to market regulation and embrace certain forms of friction in the marketplace.
Consider the many downstream social consequences of what political economist Joseph Schumpeter called “creative destruction”—economic progress that inevitably destroys old industries to make room for new innovative ones. On the one hand, this process redistributes economic rents in unpredictable ways, thus preventing any particular political-economic elite from growing too secure in its position and providing potential challengers with resources. Over the long term, it promotes pluralistic and inclusive forms of political order, as opposed to the closed and extractive institutions characteristic of societies that rely on more predictable sources of economic rents, such as land ownership or resource extraction. In preventing the entrenchment of any particular coalition of elites, creative destruction unleashed by market competition supports liberal democracy. On the other hand, of course, creative destruction also has very serious costs. Even where it generates economic growth in the aggregate, it will often harm the interests of many particular groups in the short and medium term. Indeed, that is precisely the tension to which the term “creative destruction” alludes.
Both the benefits and the costs of creative destruction give us further reasons to engage in both kinds of regulation demanded by neo-Brandeisians. On the one hand, direct interference in markets to ensure a greater balance of power—through antitrust policy, for instance—can facilitate the churn that is crucial for creative destruction. Rather than allowing successful firms to entrench their position and prevent competitors from entering the market, actively supporting a more egalitarian balance among rival firms may help to support inclusive institutions over the long term. On the other hand, facilitating coordination between weaker market actors can not only help to prevent giant firms from accumulating dangerous levels of power. It can empower countervailing organizations to mitigate the negative impacts of creative destruction on otherwise vulnerable groups. A regulatory approach that incorporates a comprehensive accounting of the normatively relevant social consequences of market economies thus has additional reasons to abandon a minimalist approach to market regulation in favor of one that aims to maintain parity among actors and foster certain beneficial types of competitive friction.
Conclusion
Many of the most important social benefits of competitive markets emerge from a classic form of parallel competition: profit-seeking firms striving independently to produce goods and services more efficiently. We do not deny that.
What we question is the regulatory inference often drawn from it. Approaches that seek to protect parallel competition by minimizing friction—most notably the standard consumer-welfare framework—fall short in two respects.
First, in complex market environments, protecting parallel competition may require fostering certain forms of friction that disperse market power and ensure that they occur on reasonably egalitarian terms. Preventing interference is not always the same as preserving competition.
Second, markets have normatively significant consequences beyond efficiency. Creative destruction can support democratic pluralism by unsettling entrenched power, but it also imposes real costs on particular groups. These broader effects give regulators additional reasons to structure market competition in ways that go beyond minimalist lane-keeping.
If competition policy is to take competition seriously, it must take both of these dimensions seriously as well.
This post was adapted from Shai Agmon and Samuel Bagg. “In Praise of Friction: Mutual Interference, Parity, and the Benefits of Competition.” In Democracy and Competition: Rethinking the Forms, Purposes, and Values of Competition in Democracy, edited by Samuel Bagg and Alfred Moore. 282:219–43. The British Academy, 2026.
Author Disclosure: The author reports no conflicts of interest. You can read our disclosure policy here.
Articles represent the opinions of their writers, not necessarily those of the University of Chicago, the Booth School of Business, or its faculty.
Subscribe here for ProMarket’s weekly newsletter, Special Interest, to stay up to date on ProMarket’s coverage of the political economy and other content from the Stigler Center.
