Rationality Gone Awry? Decision Making Inconsistent with Economic and Financial Theory - Review

Business Economics, July, 1999 by Herbert A. Simon

By Hugh Schwartz, Westport, CT: Praeger Publishers, 1998, 209 pp., hardcover, $65.00.

In the past few years, it has become quite acceptable (some would even say "fashionable") in economics to challenge the heroic portrait of human rationality that is displayed by the theory of maximization of expected utility. One erstwhile devotee (and coinventor) of rational expectations, Thomas Sargent, has even titled a recent book "Bounded Rationality." So perhaps we can say that one important lesson has been learned by the discipline: Any veridical account of economic decisionmaking must take into account the psychological capabilities of decisionmakers, and in particular their limited abilities to envision "all possible" decision alternatives, to deal with the uncertain consequences of taking one decision or another (including especially the reactions of other actors), and even to mediate among their own competing preferences.

But before neoclassical images can be replaced by something that bears a closer resemblance to reality, a theory must come into existence that is based firmly on knowledge about the actual decisionmaking process of real human beings. The lesson that economics is just beginning to learn is that such a theory cannot be constructed in the comfort of an armchair but has to have strong empirical foundations. These foundations must come, on the one hand, from psychology, where a substantial empirically grounded theory of human thinking, learning, problem solving and decisionmaking has been constructed in the past half century. They must come, on the other hand, from economics itself, through direct empirical study of human decisionmaking processes in markets and within business organizations. This process, especially at the microlevel rather than through econometric massaging of rough aggregate data, is just beginning to get under way.

In Rationality Gone Awry? Professor Schwartz makes an important contribution to acquainting economists with the existing foundations of a behavioral theory of economics, particularly, but not exclusively, those originating in psychology. In so doing, he surveys a vast number of studies that provide evidence of specific kinds of departures of the decisions of individuals and firms from the neoclassical prescriptions. He devotes a chapter (Chapter 3) to the objectives that individuals and enterprises actually seek, and the frequency and nature of their departures from profit maximization. Chapter 4 focuses on the processing of information: the enormous gap between selecting goals and searching for and analyzing information to arrive at the best choices for implementing these goals. Chapter 5 summarizes the evidence for major anomalies in choice under uncertainty in situations of the sorts that have been explored by Kahneman and Tversky and others who interrogate subjects about decisions involving risk.

Chapter 6 gives a brief account of expectations (including the "rational" kind). Chapter 7 examines the evidence for departures from rationality in individual and corporate investment decisions, including the behavior of financial markets (e.g., their volatility, choices between dividends and capital gains, short-run and long-run profits, corporate takeovers, and portfolio diversification). Although a number of references are made to the work of Roth and Smith, it would have been helpful to have had a more explicit account of the findings from research in experimental economics, which has now become one of the most important sources of new empirical evidence about the working of markets.

In these five chapters we are confronted with an enormous array of specific empirical studies, nearly all of which reveal one or more substantial departures of economic behavior from the behavior that neoclassical theory predicts. In a brief final chapter, Professor Schwartz sums up with some "guidelines for decisionmaking," which consist largely in an enumeration of the dimensions along which "boundedness" may enter into the decision process. But we should not complain if the author does not do what others have not done: lay out systematically the new theory that is to substitute for expected utility maximization. To fill in the rough framework that is already in place will be a task for the next half century.

Those of us who want to work at this task will find a valuable resource in the final forty pages of Rationality Gone Awry? that provides an extensive bibliography, much of it annotated, of the books anti papers that the author has examined, which constitutes an enormous mine of facts (and some theories) about how economic decisions are actually made. I found lots of items with which I had not been familiar, and missed only a few that I would have hoped to find there.

If we hold a Popperian view that the function of evidence is to disprove incorrect theories, then against neoclassical economics we can bring in, without delay, a verdict of "failure proven." But then we are faced with shaping the alternative, the new theory that is supposed to replace the old. The evidence amounts to a vast collection of particulars, but these do not painlessly assemble themselves into a new theory. From what we have already learned, what can we say about the shape of that new theory?