112
S T A R M E R
The two pairs of options are stochastically equivalent. The only difference is that
the group I description presents the information in terms of
lives saved
while the
information presented to group II is in terms of lives lost. Tversky and Kahneman
found a very striking difference in responses to these two presentations: 72% of
subjects preferred option A to option B while only 22% of subjects preferred C to
D. Similar patterns of response were found among groups of undergraduate stu-
dents, university faculty, and practicing physicians.
Failures of procedure invariance and description invariance appear, on the face
of it, to challenge the very idea that choices can, in general, be represented by
any
well-behaved preference function. If that is right, they lie outside the explanatory
scope of the conventional strategy. Some might even be tempted to say that
choices lie outside the scope of economic theory altogether. That stronger claim,
however, is controversial, and I will not be content to put away such challenging
evidence so swiftly. For present purposes, let it suffice to make two observations.
First, whether or not we have adequate economic theories of such phenomenon,
the “Asian disease” example is clearly suggestive that framing effects have a bear-
ing on issues of genuine economic relevance. Second, there are at least some theo-
ries of choice that predict phenomena like preference reversal and framing effects,
and some of these models have been widely discussed in the economics literature.
Although most of these theories—or at least the ones I will discuss—draw on
ideas about preference to explain choices, they do so in unorthodox ways, and
many draw on concepts more familiar to psychologists than economists. The one
feature common to this otherwise heterodox bunch of theories is that none of
them can be reduced to or expressed purely in terms of a single preference func-
tion
V
(
?
) defined over individual prospects. I will call such models
nonconven-
tional theories
. These theories step into what has been relatively uncharted water
for the economics profession. One of the aims of this chapter will be to reflect on
the relative merits of the conventional and nonconventional approaches.
4
. Nonexpected Utility Theories
4.1. The Conventional Strategy
One way to approach this literature is to ask a question that motivated a number of
theories: what properties would a conventional theory of preference need to ex-
plain the known violations of independence? To pursue that question, it will be
helpful to introduce an expositional device known as the probability triangle dia-
gram,
11
this will also prove useful as a vehicle for comparing the predictions of
alternative theories.
11
Although the probability triangle had appeared in the literature many years before (see Marschak
1950), Mark Machina’s use of it in the 1980s (see further on) popularized it to the extent that some
have called this diagram the “Machina triangle.”
Do'stlaringiz bilan baham: |