This paper was prepared for John B. Taylor and Michael Woodford, Editors

---

*

This paper was prepared for John B. Taylor and Michael Woodford, Editors, Handbook of

Macroeconomics.  An earlier version was presented at a conference “Recent Developments in Macro-

economics” at the Federal Reserve Bank of New York, February 27–8, 1997.  The author is indebted

to Ricky Lam for research assistance, and to Michael Krause, Virginia Shiller, Andrei Shleifer, David

Wilcox, and the editors for helpful comments.  This research was supported by the National Science

Foundation.

1

The Roberts (1967) paper has never been published; the fame of his paper apparently owes to

the discussion of it in Fama (1970).

1

Human Behavior and the Efficiency

of the Financial System

by

Robert J. Shiller

*

Abstract

Recent literature in empirical finance is surveyed in its relation to

underlying behavioral principles, principles which come primarily from

psychology, sociology and anthropology.  The behavioral principles

discussed are: prospect theory, regret and cognitive dissonance, anchoring,

mental compartments, overconfidence, over- and underreaction, repre-

sentativeness heuristic, the disjunction effect, gambling behavior and

speculation, perceived irrelevance of history, magical thinking, quasi-

magical thinking, attention anomalies, the availability heuristic, culture and

social contagion, and global culture.

Theories of human behavior from psychology, sociology, and anthropology have helped

motivate much recent empirical research on the behavior of financial markets.  In this paper

I will survey both some of the most significant theories (for empirical finance) in these other

social sciences and the empirical finance literature itself.

Particular attention will be paid to the implications of these theories for the efficient

markets hypothesis in finance.  This is the hypothesis that financial prices efficiently

incorporate all public information and that prices can be regarded as optimal estimates of

true investment value at all times.  The efficient markets hypothesis in turn is based on more

primitive notions that people behave rationally, or accurately maximize expected utility, and

are able to process all available information.  The idea behind the term “efficient markets

hypothesis,” a term coined by Harry Roberts (1967),

1

 has a long history in financial

research, a far longer history than the term itself has.  The hypothesis (without the words

---

2

efficient markets) was given a clear statement in Gibson (1889), and has apparently been

widely known at least since then, if not long before.  All this time there has also been

tension over the hypothesis, a feeling among many that there is something egregiously

wrong with it; for an early example, see MacKay (1841).  In the past couple of decades the

finance literature, has amassed a substantial number of observations of apparent anomalies

(from the standpoint of the efficient markets hypothesis) in financial markets.  These

anomalies suggest that the underlying principles of rational behavior underlying the efficient

markets hypothesis are not entirely correct and that we need to look as well at other models

of human behavior, as have been studied in the other social sciences.

The organization of this paper is different from that of other accounts of the literature

on behavioral finance (for example, De Bondt and Thaler, 1996 or Fama, 1997):  this paper

is organized around a list of theories from the other social sciences that are used by

researchers in finance, rather than around a list of anomalies.  I organized the paper this way

because, in reality, most of the fundamental principles that we want to stress here really do

seem to be imported from the other social sciences.  No surprise here:  researchers in these

other social sciences have done most of the work over the last century on understanding

less-than-perfectly-rational human behavior.  Moreover, each anomaly in finance typically

has more than one possible explanation in terms of these theories from the other social

sciences.  The anomalies are observed in complex real world settings, where many possible

factors are at work, not in the experimental psychologist’s laboratory.  Each of their theories

contributes a little to our understanding of the anomalies, and there is typically no way to

quantify or prove the relevance of any one theory.  It is better to set forth the theories from

the other social sciences themselves, describing when possible the controlled experiments

that demonstrate their validity, and give for each a few illustrations of applications in

finance.

Before beginning, it should be noted that theories of human behavior from these other

social sciences often have underlying motivation that is different from that of economic

theories.  Their theories are often intended to be robust to application in a variety of

everyday, unstructured experiences, while the economic theories are often intended to be

robust in the different sense that, even if the problems the economic agents face become

very clearly defined, their behavior will not change after they learn how to solve the

problems.  Many of the underlying behavioral principles from psychology and other social

sciences that are discussed below are unstable and the hypothesized behavioral phenomena

may disappear when the situation becomes better structured and people have had a lot of

opportunity to learn about it.  Indeed, there are papers in the psychology literature claiming

that many of the cognitive biases in human judgment under uncertainty uncovered by

experimental psychologists will disappear when the experiment is changed so that the

probabilities and issues that the experiment raises are explained clearly enough to subjects

(see, for example, Gigerenzer, 1991).  Experimental subjects can in many cases be con-

vinced, if given proper instruction, that their initial behavior in the experimental situation

was irrational, and they will then correct their ways.

To economists, such evidence is taken to be more damning to the theories than it would

be by the social scientists in these other disciplines.  Apparently economists at large have

not fully appreciated the extent to which enduring patterns can be found in this ‘unstable’

---