Tendencies versus Boundaries: Levels of Generality in Behavioral Law and Economics

Article excerpt

In this reply to Professor Prentice's article, Professor Mitchell offers some additional thoughts in favor of a modest approach to revising the law's assumption of rationality, as compared to the bolder approach argued for by Professor Prentice. After discussing how much of the evidence on human rationality can be used both to attack and defend the rational actor assumption due to ambiguities in this evidence, Professor Mitchell turns to the larger question of whether legal decision theorists describe behavior at too general a level to be useful in the formulation of legal policy. Professor Mitchell argues that legal decision theorists have placed too great an emphasis on finding and describing behavioral tendencies toward irrationality, without due regard for the boundary conditions on these supposed tendencies. As a result, much of the interesting and important information about the constraints on rational versus irrational behavior is consigned to ceteris paribus clauses and treated as "noise" that should be controlled and ignored rather than elucidated and understood.

When evidence on the truth or falsity of a proposition is ambiguous and open to multiple interpretations, psychologists warn about "biased assimilation" of the evidence to support pre-existing theories, beliefs, and attitudes.1 Therefore, when a skeptic about the public policy implications of psychological research examines the complex mix of evidence on human rationality, he may find much to support his skepticism about the use of psychology to reform the law. Likewise, an optimist about the public policy contributions of psychology may find within this same body of evidence much to bolster his optimistic view that psychological research can be used to refashion the law to better predict and regulate human behavior.

Two of my previous articles on the subject of behavioral law and economics,2 or, as I prefer to call the field, legal decision theory,3 and Robert Prentice's excellent discussion of my articles and of legal decision theory in general may illustrate the biased assimilation phenomenon at work.4 Indeed, Professor Prentice and I often cite the very same works to support our different perspectives on legal decision theory-with Prentice's article emphasizing how much we know about the quasi-rationality of human judgment and decision making and my articles emphasizing how little we know in light of the complexity of the evidence.5 Notwithstanding the possible influence of our predispositions, both of our readings of the evidence bring out valid points about the possibilities and limitations of legal decision theory.

Not yet content to call our debate a draw or to concede defeat, however, in this reply to Professor Prentice's article I offer some additional thoughts in favor of a modest approach to revising the law's assumption of rationality, as compared to the bolder approach supported by Professor Prentice. I begin by discussing how the ambiguous nature of much of the evidence on human rationality may explain some of the differences between Professor Prentice and me. Next, I turn to the larger question of whether legal decision theorists describe behavior at too general a level to be useful in the formulation of legal policy. In particular, I discuss the tendency of legal decision theorists to speak in terms of "behavioral tendencies" and how this approach to behavioral description may impede the progress of behavioral law and economics. I contend that legal decision theorists have placed too great an emphasis on finding and describing behavioral tendencies toward irrationality, without due regard for the boundary conditions on these supposed tendencies. As a result, much of the interesting and important information about the constraints on rational versus irrational behavior is consigned to ceteris paribus clauses and treated as "noise" that should be ignored and controlled rather than elucidated and understood.

I. IS THE RATIONALITY GLASS HALF FULL OR HALF EMPTY? …