risk can be likewise mirrored across the two domains. For instance, a given indi-
vidual who displays risk-aversion in a choice among
particular prospects with
nonnegative outcomes may display risk-seeking if all outcomes are changed to
losses of the same absolute magnitude. Kahneman and Tversky report evidence
for this kind of effect from an experiment involving choices among prospects of
the form
s
5
5
(
x
,
p
; 0, 1
2
p
) and
r
5
5
(
y
, •
p
; 0, 1
2
•
p
). For given absolute val-
ues of
x
and
y
the majority of subjects revealed
s
5
s
r
5
when
y
.
x
.
0 and
r
5
s
s
5
when
y
,
x
,
0.
The “Asian disease” example discussed at the end of section 3 is consistent
with the reflection effect. In that example, the choice between prospects was af-
fected by the description of options. When outcomes were framed as lives saved,
the majority of choosers were attracted to a sure gain of 200 out of 600 lives;
when framed as losses the majority rejected the sure loss of 400 out of 600 deaths,
preferring instead to take the risk. The effect observed there can be interpreted as
a reflection effect with risk aversion in relation to
gains and risk-seeking for
losses. Before we could think this an explanation of the Asian disease problem,
however, we need an account of how consequences are
interpreted
. From an ob-
jective standpoint, two hundred lives saved out of six hundred
is the same thing as
four hundred lives lost, hence a full explanation would require a theory of how
framing affects whether an outcome is interpreted as a gain or a loss. Kahneman
and Tversky go some way toward this in their discussion of editing.
Prospect theory assumes that prior to the second stage of evaluation, individu-
als will edit prospects using a variety of heuristics. One of the major editing oper-
ations involves the
coding
of outcomes as gains and losses relative to a reference
point. Kahneman and Tversky argue that the reference point will typically be the
current asset position, but they allow the possibility that “the location of the ref-
erence point, and the consequent coding of outcomes as gains or losses, can be af-
fected by the formulation of the offered prospects, and by the expectations of the
decision maker” (p. 274). Notice that this possibility of differential coding under
the two problem descriptions is a necessary step in explaining responses to the
Asian disease problem. While some economists might be tempted to think that
questions about how reference points are determined sound more like psycholog-
ical than economic issues, recent research is showing that understanding the role
of references points may be an important step in explaining real economic behav-
ior in the field (see, for example, Heath, Huddart, and Lang 1999).
Several of the other editing routines in prospect theory are essentially rules for
simplifying prospects and transforming them into a form that can be more easily
handled in the second phase. One such operation is the rule of
combination
,
which simplifies prospects by combining the probabilities associated with identi-
cal outcomes. For example, a prospect
described
as (
x
1
,
p
1
;
x
1
,
p
2
;
x
3
,
p
3
; . . .) may
be
evaluated
as the simplified prospect (
x
1
, (
p
1
1
p
2
);
x
3
,
p
3
; . . .).
Notice that
these two prospects are not, in general, equivalent if • (
?
) is nonlinear. Decision
makers may also simplify prospects by rounding probabilities and/or outcomes.
Further operations apply to
sets
of prospects. The operation of
cancellation
in-
volves the elimination of elements common to the prospects under consideration.
Do'stlaringiz bilan baham: