Loading field map
Market competition theory is the branch of industrial organization that studies how firms’ strategic behavior is shaped by market structure and how that behavior, in turn, determines market outcomes such as prices, quantities, product quality, and innovation. At its core, the field asks a deceptively simple question: when firms compete, what does competition actually accomplish? The answers matter because they inform antitrust policy, regulation, and any assessment of whether markets serve consumers well.
The field is not a single unified doctrine. It is best understood as a set of overlapping research traditions that differ in their assumptions about how firms think, what information they possess, and what “competition” means in practice. These traditions have developed partly in succession and partly in parallel, and contemporary work routinely combines elements from several of them.
Three enduring questions organize the field. First, how does market structure affect performance? Classic structure–conduct–performance analysis asks whether industries with fewer firms tend to charge higher prices or earn higher profits, and whether that relationship is causal or merely coincidental. Second, how do firms strategically interact? When a firm’s optimal action depends on what rivals do—whether to cut price, build capacity, launch a product, or enter a market—the outcome depends on the rules of the game and what each player knows. Third, when does competition fail? Markets can fail to produce efficient outcomes for reasons that have nothing to do with monopoly in the textbook sense: firms may collude tacitly, predatory pricing may deter entry, or network effects may lock in an inferior standard.
The stakes are practical. Competition policy—antitrust law in the United States, competition law in the European Union and elsewhere—relies on theoretical predictions about which business practices harm consumers. The field’s disputes are therefore not merely academic. Whether a merger is presumed harmful, whether exclusive contracts are suspect, and whether below-cost pricing is evidence of predation all depend on which theory of competition is treated as authoritative.
The oldest systematic approach, developed in the mid-twentieth century, is the structure–conduct–performance (SCP) paradigm. Its intellectual home was Harvard, and its organizing assumption was that market structure—the number and size distribution of firms, barriers to entry, product differentiation—determines conduct (pricing, advertising, investment) which then determines performance (efficiency, profitability, innovation). The paradigm’s empirical workhorse was cross-industry regression: industries with higher concentration ratios, it was claimed, exhibited higher profits.
The SCP approach has been heavily criticized and is no longer accepted as an adequate theory, but it remains historically important. Its critics, particularly from the University of Chicago, pointed out a fundamental identification problem: high profits in a concentrated industry might indicate collusion, but they might equally indicate superior efficiency. A firm that is more efficient than its rivals will both gain market share and earn high profits; concentration is then the result of competitive success, not its absence. This critique, associated with Harold Demsetz and others, did not merely challenge a statistical method—it challenged the presumption that concentrated structures are socially costly.
The SCP paradigm also suffered from a theoretical weakness: it lacked an account of how structure leads to conduct. Why would a concentrated industry produce collusive pricing? The paradigm treated the connection as almost mechanical. The need for a rigorous account of firm interaction led to the field’s most transformative development.
Beginning in the 1970s and accelerating through the 1980s, game theory transformed market competition theory from a largely empirical and descriptive enterprise into a discipline built on formal models of strategic interaction. The central innovation was to treat firms as rational players making optimal choices in full awareness that their rivals are doing the same. The solution concept most often used is the Nash equilibrium: a set of strategies, one per firm, such that no firm can improve its payoff by unilaterally changing its own strategy.
The game-theoretic approach addressed the SCP paradigm’s explanatory gap. It showed precisely how structure—the number of firms, the degree of product differentiation, the cost structure—generates conduct. Two canonical models anchor the tradition.
The Cournot model, named for the nineteenth-century French mathematician Augustin Cournot, assumes firms choose quantities simultaneously, and the market price adjusts to clear supply and demand. The model predicts that equilibrium price declines as the number of firms increases, approaching the competitive price only in the limit of many firms. The Bertrand model, named for Joseph Bertrand, assumes firms choose prices simultaneously. With homogeneous products and constant marginal costs, even two firms are driven to price at marginal cost—the fully competitive outcome.
These two models give starkly different predictions for duopoly, and the contrast has shaped decades of research. The difference lies in the strategic variable: quantity competition softens rivalry because a firm’s best response to a rival’s expansion is to contract its own output; price competition is fierce because undercutting a rival grabs the entire market. The lesson, which has become a central insight of the field, is that competition is not a single phenomenon. Its intensity depends on the strategic instruments firms have at their disposal.
Game theory also provided tools for analyzing dynamic competition. The folk theorem of repeated games shows that in infinitely repeated interactions, collusion is sustainable if firms care enough about future profits. This result explained why tacit collusion is possible without any explicit agreement, and it highlighted the role of factors that make future punishment more or less effective: the interest rate, the frequency of interaction, the observability of rivals’ actions, and the ease of detecting deviations.
The game-theoretic revolution brought with it a new style of theorizing. Models became precise, results became conditional, and the field became increasingly comfortable with the idea that equally plausible assumptions could yield opposite predictions. This was both a strength and a liability. The strength was rigor: no conclusion could survive without explicit assumptions about information, timing, and strategy space. The liability was indeterminacy: with enough freedom to choose assumptions, a theorist could produce a model supporting almost any conclusion.
Running alongside the game-theoretic revolution was the intellectual movement known as the Chicago School of antitrust. Although not a formal model-building tradition, Chicago’s influence on market competition theory is profound. Its defining stance was skepticism toward government intervention in markets and a corresponding confidence that most business practices that appear anticompetitive have efficiency explanations.
Chicago scholars argued that many practices the SCP tradition viewed with suspicion—vertical restraints such as resale price maintenance, exclusive dealing, tie-in sales—are better understood as ways of solving coordination problems or aligning incentives in a distribution chain. A manufacturer, for example, might set a minimum resale price to induce retailers to provide pre-sale services that consumers free-ride on when they buy elsewhere at a discount. The practice can increase demand and lower the manufacturer’s price, benefiting consumers.
The Chicago approach was less a formal theory than a set of presumptions: that markets are generally self-correcting, that entry erodes monopoly power, and that firms rarely succeed in monopolizing markets without being more efficient than their rivals. Its policy conclusions—narrow the scope of antitrust, treat vertical restraints as presumptively legal, evaluate mergers mainly by their efficiency effects—have been enormously influential, particularly in U.S. courts.
The Chicago School’s limits became apparent over time. Its confidence in entry as a check on market power assumed that entry is easy, which is not the case in industries with large sunk costs, network effects, or intellectual property protection. Its claim that predatory pricing is rarely rational rested on specific assumptions about information and capital markets that do not always hold. And its presumption that firms’ practices reflect efficiency is a presumption, not a demonstrated fact; in many cases, the same practice can be explained either way.
By the 1990s, game-theoretic modeling had been turned against Chicago’s conclusions, producing what is loosely called post-Chicago economics. This approach does not reject Chicago’s emphasis on economic reasoning; it rejects Chicago’s presumption that markets are efficient absent clear proof of harm. Instead, it uses formal models to show that apparently benign practices can be anticompetitive under plausible conditions, and that the Chicago conclusions depend on assumptions that are not always realistic.
Key post-Chicago results concern topics such as exclusive dealing, raising rivals’ costs, and predatory pricing. Chicago scholars argued that exclusive dealing contracts cannot be anticompetitive because a buyer would not accept a contract that forecloses cheaper alternatives unless compensated. Post-Chicago models show that a dominant firm can, however, offer exclusive contracts to buyers that are individually rational but collectively harmful: each buyer accepts because the payments are generous, but the cumulative effect is to deny entry to a more efficient rival. The model’s insight is that the externality across buyers—each buyer’s acceptance makes entry less likely for all—means that the market outcome is not efficient even though no individual was coerced.
The post-Chicago approach did not replace Chicago as a unified doctrine. It is better described as a collection of models that share a methodology and a temper: rigorous, game-theoretic, policy-relevant, and willing to question both market efficiency and government intervention. The field today is largely post-Chicago in the sense that serious arguments on both sides are conducted in its formal language.
While theory was being transformed, empirical work also underwent major changes. The old SCP regressions—industry-level correlations between concentration and profitability—were replaced by two more sophisticated approaches.
Structural estimation uses formal economic models as the lens through which data are interpreted. A researcher specifies a model of demand and supply, estimates its parameters using market data, and then uses the estimated model to simulate counterfactuals: What would prices be if two firms merged? What would profits be if a firm entered? This approach, associated with the New Empirical Industrial Organization, is powerful because it allows for direct evaluation of policy questions. Its weakness is that the results depend on the model’s assumptions being correct; if the demand system is misspecified, the merger simulation is unreliable.
Reduced-form empirical work, by contrast, uses natural experiments, instrumental variables, or quasi-experimental variation to estimate the causal effect of competition on outcomes without specifying a full model. The goal is to find settings where the researcher can compare otherwise similar markets that differ in some dimension of competition—perhaps due to a policy change, an entry event, or institutional variation—and measure the effect. This approach is more robust than structural estimation in the sense that it does not depend on a full model, but it typically provides less guidance for predicting the effects of actions not observed in the data.
Both approaches coexist in contemporary research, and cutting-edge work increasingly combines them: reduced-form estimates can validate structural models, and structural models can extrapolate beyond the observed range. The empirical turn has made market competition theory more directly tied to policy, as merger review in the United States and the European Commission routinely relies on these methods.
Several active research fronts define the current landscape. Contestability theory, associated with William Baumol, holds that what matters for performance is not the number of firms in a market but the threat of entry. A monopoly that faces no entry barriers is forced to price at the competitive level to avoid inviting entry. The theory identifies sunk costs rather than fixed costs as the real entry barriers: if entrants can recoup their investments upon exit, the market is contestable. Its influence has waned as empirical research has found that many markets have substantial sunk costs, but it remains a useful benchmark.
Platform markets—two-sided markets such as credit cards, search engines, app stores, and ride-hailing networks—pose new questions that do not fit the traditional mold. A platform serves two distinct groups whose participation is interdependent: more merchants attract more consumers, and more consumers attract more merchants. Pricing decisions are correspondingly complex; platforms may subsidize one side while charging the other. Competition between platforms raises questions about multihoming (using several platforms at once), tipping (the market collapsing to one dominant platform), and network effects (the value of the platform increasing with the number of users). These markets have become a central focus both because of their economic importance and because antitrust authorities struggle to apply traditional tools—market definition itself is contested when the product is free to users.
A quieter development is the incorporation of behavioral considerations. Traditional theory assumes firms maximize expected profits and consumers maximize expected utility. Behavioral industrial organization relaxes these assumptions, asking what happens when consumers are inattentive, overestimate the benefits of complex products, or fail to switch providers even when cheaper alternatives exist. The findings complicate both Chicago and post-Chicago conclusions: behavioral consumers can make some practices more harmful than the rational model suggests, but they can also make other practices benign. The field is young, and its policy implications are still being worked out.
The approaches described here are not mutually exclusive stages in a linear progression. The SCP paradigm is largely a historical precursor, but its concern with the link between structure and performance remains the field’s animating question. Game theory is not a school with a doctrine but a toolbox; both Chicago and post-Chicago economists use it. The Chicago School is best understood as a set of presumptions and policy preferences rather than a formal theory; its influence persists in the burden of proof assigned in antitrust decisions. Post-Chicago work is not a settled doctrine but an ongoing research program. Empirical work does not replace theory; it tests, calibrates, and informs it.
The field’s current state is thus one of productive tension. Market competition theory possesses rigorous methods but no settled conclusions about which of those methods’ insights apply in any given market. That uncertainty is not a failure of the field; it reflects the genuine complexity of its subject matter. Competition is a process, not a state, and the field’s enduring contribution is to have made precise the conditions under which that process serves consumers well—and the conditions under which it does not.