The empty promise of behavioral antitrust.

• Author: Devlin, Alan
• Position: II. The Empty Promise of Behavioral Antitrust C. Behavioral Analysis of Specific Business Practices through V. Conclusion, with footnotes, p. 1038-1063

3. Behavioral Analysis of Specific Business Practices

   The problems of indeterminacy outlined above are not limited to substitution and entry. Because behavioral antitrust operates only in hindsight, it has no value for policymakers charged with predicting whether markets can fully and promptly correct short-term imperfections or whether intervention is required to restore competition. Consider three business practices that neoclassical economics and game theory have determined generally to be efficient but occasionally harmful to competition: predatory pricing, unilateral refusals to deal, and product tying and bundling.

   Predatory pricing involves a dominant firm setting its price below cost to entrench its monopoly, principally by denying fringe rivals the scale necessary to achieve economies in production and by fostering an aggressive reputation so as to deter entry. (126) Concerned with the risk of mistakenly deterring price-cutting, and informed by the game-theoretic literature showing that the strategy is generally an irrational and ineffective means by which to exclude equally or more efficient competitors, the law takes a skeptical view of predatory-pricing claims. (127) To prevail, a plaintiff must establish not only that a dominant firm set price below cost, but also that a dangerous probability exists that in the future the predator will recoup its losses. (128) This law, which demands much of a plaintiff, reflects the premise that below-cost pricing will not usually harm competition, as future entry will promptly remedy any subsequent attempt by the would-be predator to exercise market power and recoup its losses. (129)

   Unilateral refusals to deal occupy a controversial position within competition law. When a dominant firm controls a physical or technological infrastructure necessary for viable operation in a market, competitors denied access to the infrastructure may seek a court order compelling the dominant firm to share its infrastructure on fair and reasonable terms. (130) U.S. law recognizes a duty to deal only in narrow circumstances--specifically, where the parties formerly engaged in a mutually profitable course of dealing, and when the dominant party terminated that dealing suddenly and without a valid business justification. (131) The law takes this view out of concern that mandatory sharing would reduce incentives to invest ex ante. (132)

   Product tying arises when a dominant seller requires its customers to purchase a second good (the tied product) as a condition of buying the tying good. (133) These restraints often yield overriding efficiencies, particularly where the items sold together are economic complements. (134) The practices reduce search and negotiation costs, entail production-side savings, facilitate interoperability, and typically result in lower combined prices and higher output because they eliminate the double marginalization caused by divided ownership. (135) They can sometimes harm competition, though, by denying rivals scale in the tied product markets or by bolstering the dominant firm's position in the tying market. (136) Although price theory suggests that even a monopolist in the tying-goods market cannot profitably charge a double markup (137) by using tie-ins to achieve monopoly in an otherwise-competitive tied market, this outcome necessarily holds true only for fixed-proportions tying. (138) It also ignores the possibility that a dominant share of a tied market may enable the tying firm to realize future profit opportunities in that market. (139)

   Under today's antitrust rules, a firm can violate the law by product tying only if it has monopoly power in the tying product market and its actions substantially foreclose its rivals in the market for the tied product. (140) Many economists have argued that the Supreme Court should relax these rules. (141) The strength of their argument hinges in part on whether consumers will react rationally to contractual restraints that require them to purchase the tied products only from the tying firm: in particular, whether they will add the price of the tied product to the purchase price of the tying good. If buyers rationally combine the present tying-product price and the later tied-good price, sellers will be unable to charge a second monopoly price. If buyers discount the future, or otherwise focus only on the tying-good price that they have to pay today, sellers may be able to extract a second monopoly price later when lock-in has occurred. (142) The strength of their arguments also depends on the assumption that new entry is likely to occur in the tying and tied product markets, should the tie-in prove inefficient.

   Because a similar analysis applies to these practices as to the overarching questions of substitution and entry, we provide a table highlighting the offsetting insights of a non-exhaustive list of behavioral biases on the ultimate effect of the restrictions:

   Restraint Biases Protecting Biases Harming Competition Competition Predatory Irrational escalation Loss aversion Pricing Bandwagon effect Hyperbolic Endowment effect discounting Optimism bias Pessimism bias Status quo bias Anchoring Aversion to sunk Overconfidence bias costs Overconfidence bias Refusal to Optimism bias Loss aversion Deal Anchoring bias Sunk-cost aversion Bandwagon effect Pessimism bias Reputational goals Exaggerate low Incentive to grow probability share Status quo bias Availability bias Tying and Loss aversion Anchoring bias Bundling Availability bias Availability bias Framing effect Endowment effect Anchoring bias Sunk-cost aversion Bandwagon effect Post-purchase Overconfidence bias rationalization Selective perception Thus, behavioral antitrust is a disorganized amalgam of context-dependent biases that operate in varying directions and to varying degrees. No behavioral theory of antitrust exists, nor does one appear to be possible. Indeed, the preceding analysis illustrates the fatal problems that result from the absence of an organizational principle in complex environments with many explanatory variables. Next, we critique the scholarship that has promoted behavioral antitrust. We find it to be deficient and unconvincing.

   3. Debunking the Behavioral Antitrust Literature

   Although still in its infancy, the behavioral antitrust literature has reached the point where an overarching thesis has emerged, comprised of three related propositions. According to the first tenet, the conventional economic account of antitrust fails because its analysis rests on unrealistic assumptions of rationality and market efficiency. (143) Second, antitrust's attachment to neoclassical economics blinds policymakers to evidence at variance with the conceptually attractive, but ultimately erroneous, notion that irrational behavior either does not exist or cannot endure. (144) As a corollary, behavioral antitrust scholars almost uniformly take issue with what they perceive to be the laissez faire nature of the modern antitrust enterprise. (145) The third proposition argues that, because firms and consumers are systemically irrational, society cannot rely on private contract to fashion efficient outcomes. (146) For certain scholars, this means that antitrust's relatively passive role in modern times should give way to a policy founded on "promoting" competition. (147)

   In assailing the rationality foundation of modern antitrust theory, however, behavioral scholars reveal a serious misapprehension of their target. For the reasons that follow, the academic literature promoting behavioral economics' application to antitrust is flawed.

1. Behavioral Scholars Err in Criticizing the "Realism" of Neoclassical Antitrust Economics

   Above all, the behavioral antitrust literature criticizes the economic assumptions of rationality, profit maximization, and efficiency that underlie modern competition policy. (148) This critique, though, targets a straw man, characterizing the standard economic account in inaccurate and easily caricatured terms. One commentator derides "the suffocating straitjacket of neoclassical economics and its unrealistically static models," (149) while others urge that "one cannot assume that markets operate as efficiently as the Chicago School predicts." (150) Yet another maintains "the assumption that humans behave as perfectly rational, profit-maximizing actors has taken center stage in modern antitrust law." (151) In contrast, "behavioral economics ... has questioned the assumption that humans always behave perfectly rationally." (152) Former FTC Commissioner J. Thomas Rosch, a prominent advocate of behavioral antitrust, has suggested that "the orthodox and unvarnished Chicago School of economic theory is on life support, if it is not dead." (153)

   If this criticism were true, it would be hard not to embrace the realistic behavioral enterprise. But this critique exists only in the writings of behavioral scholars, who conjure an image sharply at odds with the reality of price theory. The perfect-competition model is an abstraction used to facilitate analysis, not a description of every bit of economic life. (154) Used as a benchmark, the model shows that market inefficiencies are ubiquitous. (155) The notion that the neoclassical model of perfect competition reflects the neoclassical view of actual competition is seriously mistaken. (156)

   Behavioralists' criticism of rational choice theory's assumed rationality is equally wrong. Neoclassical models assume profit and utility maximization by firms and consumers to generate predictions. They make no claim that individual, real-world actors are invariably rational. Rather, they maintain that the impulse to maximize profit and utility--along with opportunities for learning, competitive pressures, and canceling-out effects--is sufficiently strong that the explanatory variables price theory highlights correlate in statistically significant fashion with actual market outcomes. In short, the...