105
N O N E X P E C T E D - U T I L I T Y T H E O R Y
In what follows, my aim will be to set out what I take to have been key theoret-
ical developments in the area, to review the related evidence, and to draw conclu-
sions about the current state of play and the prospects for the future. In doing so,
rather than simply present an exhaustive list of models, my aim will be to identify
and discuss different modeling
strategies
, picking specific models as illustrations.
I also intend to narrow my sights in two significant respects. First, my focus will
be on
descriptive
as opposed to
normative
issues. Second, I will concentrate on
the problem of modeling choices under
risk
as opposed to the more general cate-
gory of
uncertainty
(the distinction is explained in the next section). Clearing the
ground in this way will, I hope, sharpen the focus on one central research problem
that continues to motivate much of the research in this arena: the endeavor to de-
velop a “satisfactory” account of
actual
decision behavior in situations of
risk
. It
will be a personal view, but one which I hope will help the interested nonspecial-
ist find a trail through this expansive and quite detailed literature.
The chapter is organized as follows. Sections 2 and 3 set the scene with discus-
sions of the standard theory and the evidence that prompted theorists to look for
alternatives. Section 4 provides the core overview of nonexpected utility theories.
Section 5 seeks to evaluate what has been achieved so far, and in three subsec-
tions I discuss (1) how new theories have fared in a second phase of experimental
testing, (2) how new theories may help us to explain a range of phenomena “in
the field,” and (3) whether nonexpected utility theory offers a viable alternative to
EU for everyday theoretical use.
2
. Where It Began
Although the primary purpose of this chapter is to review alternatives to EU, that
theory provides the natural point of departure since most of the theories I will be
discussing can be understood as generalizations of this base theory.
1
EU was first
proposed by Daniel Bernoulli (1738) in response to an apparent puzzle surround-
ing what price a reasonable person should be prepared to pay to enter a gamble. It
was the conventional wisdom at the time that it would be reasonable to pay any-
thing up to the expected value of a gamble, but Bernoulli presents this counterex-
ample: A coin is flipped repeatedly until a head is produced; if you enter the
game, you receive a payoff of, say, $2
n
where
n
is the number of the throw pro-
ducing the first head. This is the so-called St. Petersburg game. It is easy to see
that its expected monetary payoff is infinite, yet Bernoulli believed that most peo-
ple would be prepared to pay only a relatively small amount to enter it, and he
took this intuition as evidence that the “value” of a gamble to an individual is not,
in general, equal to its expected monetary value. He proposed a theory in which
individuals place subjective values, or “utilities,” on monetary outcomes and the
value of a gamble is the expectation of these utilities. While Bernoulli’s theory—the
1
I shall not dwell on this account of EU. For those interested in further discussion, an excellent
starting place is Paul Schoemaker’s (1982) review.
Do'stlaringiz bilan baham: |