Behavioral Economics in Antitrust

Mon Apr 25 16:24:38 EDT 2011
By Dr. Elizabeth M. Bailey

Imagine this: When it comes to making decisions, people are like Mr. Spock. They dispassionately weigh pros and cons. They don’t let their emotions cloud their decisions. If mistakes are made - maybe buying an overvalued Internet stock or a Las Vegas condo - they learn from them and don’t repeat them.

None of this is true, of course. But conventional economic theory assumes it is. And these assumptions helped to shape the economics of antitrust. Conventional economic models used in antitrust assume that consumers and firms are rational.

US courts, government agencies and practitioners rely on those models and economic techniques derived from them to predict the economic effect of alleged anti-competitive conduct and to make decisions about challenging proposed mergers and acquisitions. However, a growing body of research shows that people and firms do not always behave like Spock.

An entire field of research - behavioral economics - has emerged that is devoted to understanding how consumers and firms depart from the standard assumptions. A primary goal of the “behavioralists” is to understand consumer and firm decision-making in order to make economic models more realistic.

Behavioral researchers have identified many real-world examples of irrationality. Consumers may not take simple steps to maximize their welfare, for example, failing to enroll in 401(k) plans or accepting the employer-provided default investment when another option would serve them better.

This body of economic research has sparked a debate about whether the conventional economic models used in antitrust analyses adequately account for real-world behavior. One assumption that has been questioned is whether firms maximize profits as the models used in antitrust analyses assume they do.

The standard model of firm decision-making assumes that a firm makes choices about price, quality, innovation and output to maximize its profits. And for many, if not most, firms, this is likely to be a reasonable assumption.

But for some firms, it may not be. Some enterprises by their nature may not maximize profits. It seems sensible to question whether a not-for-profit university or a not-for-profit hospital should be modeled as maximizing profit. And even for-profit companies may depart from strict profit maximization over the short or medium term. In the short-run, they might choose instead to increase revenues or market share. Or a firm’s managers may simply grow complacent and pay insufficient attention to the firm’s bottom line.

If the models used to evaluate potential antitrust concerns rest on assumptions that are flawed, then the resulting analyses may not provide useful predictions.

As an example, conventional economic models of firm pricing yield a relationship between a firm’s gross profit margin and its own-price elasticity of demand. In merger analysis, this relationship, coupled with an estimate of the acquiring firm’s gross profit margin, is sometimes used to draw inferences about the elasticity of demand faced by the acquirer.

However, if the firm’s short-run goal is something other than profit maximization, then an analysis based on a flawed inference about the firm’s own-price elasticity of demand drawn from the firm’s gross profit margin may not be useful for making predictions about post-merger market power and post-merger pricing.

As another example, conventional economic models predict that variable cost savings, such as lower input costs, are more likely to lead to lower prices post-merger than fixed cost savings, such as lower overhead costs. These economic models provide the basis for the U.S. merger enforcement agencies putting more weight on efficiencies generated from variable cost savings than those generated from fixed cost savings.

However, too little credit may be given to fixed cost savings if the conventional economic models do not describe adequately how firms set prices. If an acquiring firm sets prices taking explicit account of fixed costs, then a reduction in fixed costs post-merger may well lead to lower prices post-merger.

There is precedent for antitrust agencies entertaining the notion that a firm may have goals besides profit maximization. The Federal Trade Commission investigated Genzyme Corp.’s acquisition of Novazyme Pharmaceutical in 2001, focusing on the extent to which the transaction might lessen the pace of research and development for a treatment for Pompe disease. Given that both firms were developing treatments for the disease, a reduction in head-to-head competition could have slowed the pace of R&D post-merger.

The FTC, however, took account of facts that conventional models did not capture. In closing the investigation without taking any enforcement action, then-FTC Chairman Timothy J. Muris noted that the structure of the Genzyme/Novazyme transaction would not dampen incentives to develop a treatment because the manager slated to be in charge of the Pompe disease research program had two children afflicted with the disease.

In making that decision, the FTC was, in effect, tweaking its models so that they better accorded with the facts about how individuals and firms made decisions in that specific situation.

How closely models’ assumptions track actual behavior determines, in part, how useful these models are in predicting potential anti-competitive effects. The rational “behavior” embedded in neoclassical economics is likely to be a reasonable assumption a lot of the time. It is important, however, to test whether modeling assumptions accord with the facts.

It is also important to assess whether behavior that deviates from the conventional assumptions is systematic and persistent. If the relevant facts suggest that a firm might behave in ways that depart from conventional assumptions, then private parties, government agencies and the courts should be willing to consider alternate economic models that account appropriately for the observed behavior.