The empty promise of behavioral antitrust.

• Author: Devlin, Alan
• Position: Abstract through II. The Empty Promise of Behavioral Antitrust B. Entry as a Constraint on Market Power, p. 1009-1037

ABSTRACT

Microeconomic theory has long guided competition law. Using price and game-theoretic models, antitrust has settled on rules that have endured because they are more coherent, easier to understand, and simpler to apply than those of any other methodology. In application, these rules predict the market consequences of business transactions innumerable in form and uncertain in outcome. But this coherent framework is now under attack. Entranced by the larger "behavioral law and economics" movement, certain academics have questioned the pillars of doctrine built upon the foundation of rational-choice theory. In their view, bounded rationality, willpower, and self-interest afflict firms' and consumers' decisionmaking, inducing systemic departures from the predictions of neoclassical economics and game theory. Current antitrust laws, they argue, fail to account for those departures from rationality. Because these rules and standards produce what they regard as unduly permissive treatment, behavioral antitrust scholars urge more interventionist policy.

This Article contends that, whatever its virtues for the larger field of law and economics, behavioral economics can play no useful role in contemporary antitrust policy. It is hopelessly vague, untethered to a theory, and reliant on biases that routinely operate in opposing directions. Although it can sometimes describe the past, it is incapable of predicting the future--a fatal shortcoming for any method of antitrust analysis. We test whether behavioral antitrust can produce a coherent theory for predicting the market effects of impugned restraints on trade and exclusionary conduct. In doing so, we show that the biases prove either too much--all results are possible--or too little, canceling themselves out and reverting to the (rational) mean. The suggested utility of behavioral antitrust depends entirely upon which biases are thought to explain the conduct in question. But because the movement lacks any method for determining the explanatory power of a particular bias ex ante, choosing between conflicting biases is either a random act or a political one.

We also show what the behavioralists must be loath to mention. Despite the claimed empirical superiority of behavioral antitrust, on one key point the psychological literature has failed to supply it with the evidence or theory critical for its application: the frequency with which one set of biases dominates others. Without this key bit of inductive fact or deductive logic, all that remains of the approach is conjecture or cynical manipulation. We conclude that the behavioral movement is a non-event for antitrust policy today.

ABSTRACT INTRODUCTION I. A BRIEF OVERVIEW OF BEHAVIORAL ANTITRUST A. The Core Principles of Behavioral Antitrust B. Cognitive Biases and Prospect Theory II. THE EMPTY PROMISE OF BEHAVIORAL ANTITRUST A. Price-Induced Substitution as a Constraint on Market Power 1. Biases Suggesting a Lack of Substitutability 2. Biases Indicative of Rapid Substitution B. Entry as a Constraint on Market Power 1. Biases Suggesting that Entry Is More Likely than Predicted in the Conventional Account 2. Other Biases Suggest that Less Entry Will Occur than Under the Conventional Account C. Behavioral Analysis of Specific Business Practices III. DEBUNKING THE BEHAVIORAL ANTITRUST LITERATURE A. Behavioral Scholars Err in Criticizing the "Realism" of Neoclassical Antitrust Economics B. In Focusing on Empiricism, Behavioral Antitrust Reveals Its Emptiness C. Behavioral Economics Is Anti-Theoretical D. Behavioral Antitrust Cannot Yield More Accurate Predictions E. Behavioral Antitrust's Substantive Recommendations Are Unsound F. The Behavioral Antitrust Literature Lacks Specificity G. Behavioral Antitrust as Modest Adjustment? H. Debunking the Literature: Conclusion IV. THE STAYING POWER OF NEOCLASSICAL ANTITRUST ECONOMICS A. Neoclassical Assumptions Are Not Meant to Be Realistic B. Price Theory as an Organizing Principle C. Neoclassical Antitrust Is Not Synonymous with the Chicago School V. CONCLUSION INTRODUCTION

Antitrust law has evolved incrementally over the past several decades into a coherent body of doctrine. Policymakers have drawn on the price and game-theoretic insights of modern industrial organization to formulate rules and standards with distinct substantive and administrative advantages. The law that has emerged is logical, simple to understand, usefully predictive, and--for those reasons--relatively easy to apply. Illustratively, ease of entry defeats monopolization claims; (1) mergers effecting modest increases in concentration pass muster under the agencies' review standards; (2) complaints alleging behaviors that make no economic sense fail to state a claim; (3) a manufacturer can require its retailers not to sell below a minimum price; (4) firms with less than fifty percent of a market cannot monopolize that market as a matter of law; (5) and firms have no duty to deal with their competitors unless they have unilaterally terminated a profitable course of prior dealing. (6) Many such rules now form the basis for the administration of competition law not just in the United States but in most of the world as well.

Critical assumptions about the rationality of individuals and firms lie at the heart of current analysis. Consumers respond rationally to the price and quality signals that sellers of desired goods provide. They will buy products and services that provide them with the most relative value, compare competing products sensibly, and switch to rival products should the quality-price mix of their initial choice rise unacceptably. And they will continue to remain informed of changes in the marketplace that might lead them to switch.

By the same token, firms are assumed to be rational profit-maximizers, endeavoring to make as much profit as they can. Because their profits come from sales, they must be attuned to shifts in consumer demand, the need for new products, the actions of their rivals, and the possibility of making greater profits in some new market. And they must respond correctly to all of these forces if they are to become and remain profitable. Economists have derived these rationality assumptions and deployed them in the service of a straightforward, coherent, and predictive methodology that has dominated antitrust law and policy for the past thirty-five years. That analytic method uses models based on constrained optimization and noncooperative game theory. Pursuant to this inductive analysis, economists have empirically tested their models based on observed economic behavior. Since the 1970s, price and game theories have dominated the structuralist approach to industrial economics that previously held sway under the Harvard School of Economics.

Now, however, competition policy is approaching a crossroads. Some academics and regulators have begun to challenge antitrust's theoretical and evidentiary underpinnings, questioning what they perceive to be the laissez faire policy encouraged by the conventional account. (7) They argue that the pricetheoretic view that the Chicago School first propounded--a methodology that has long rationalized, structured, and guided antitrust enforcement--is deeply flawed, its core assumptions unrealistic, its "goals" indistinct and unattainable, its key terminology indeterminate, and its implementation a form of politics masquerading as neutral decisionmaking. (8)

This critique seeks to import into antitrust law lessons drawn from behavioral economics, which purports to enrich the realism of economic thinking by incorporating insights from cognitive psychology. (9) Decrying the rational-choice assumptions underlying conventional antitrust analysis, behavioral scholars contend that the psychological literature yields revolutionary insights for competition policy. Drawing on experimentally observed departures from rational choice, they have identified biases that lead firms and consumers to act contrary to their self-interest and hence to the predictions of neoclassical and game theories. (10) Using this literature, which focuses on reference points, heuristics, and mental biases in lieu of expected-utility maximization, behavioral antitrust scholars claim an ability to supplant, or more modestly to complement, the conventional approach to antitrust analysis. (11)

This Article makes two claims. First, behavioral antitrust rests on a series of observations that necessarily look backward and offer neither a model against which to assess business behavior nor a theory by which to predict its future effects. We do not argue that neoclassical economic analysis is infallible or that its predictions are invariably accurate. Nor do we steadfastly defend a non-interventionist antitrust policy. We argue instead that, because it lacks a predictive component, behavioral economics adds nothing to competition policy beyond what empiricism has long contributed. In that respect, no one can doubt that empirical observation has long been at the core of industrial organization and that it has shaped the development and refinement of neoclassical theory. But this fact makes behavioral antitrust redundant as an advocate for empiricism and useless when the relevant effects of impugned behavior will arise in the future, and must therefore be predicted.

Second, we show that behavioral antitrust is malleable to the point of being meaningless. In...