A closely related set of anomalies involves sequences of outcomes. Until re-
cently, most experimental research on intertemporal choice involved single out-
comes received at a single point in time. The focus was on measuring the correct
form of the discount function and it was assumed that once this was determined,
the value of a sequence of outcomes could be arrived at by simply adding up the
present values of its component parts. The sign and
magnitude effects and the
delay / speedup asymmetry focused attention on the form of the utility function
that applies to intertemporal choice, but retained the assumption of additivity
across periods. Because they
involved only single outcomes, these phenomena
shed no light on the validity of the various independence assumptions that involve
multiple time periods.
Research conducted during the past decade, however,
has begun to examine
preferences toward sequences of outcomes and has found quite consistently that
they do not follow in a simple fashion from preferences for their component parts
(Loewenstein and Prelec 1993). People care about the “gestalt,” or overall pattern
of a sequence, in a way that violates independence.
A number of recent studies have shown that people generally favor sequences
that improve over time. Loewenstein and Sicherman (1991) and Frank and
Hutchens (1993 and this volume), for example, found that a majority of subjects
prefer an increasing wage profile to a declining or flat one, for an otherwise iden-
tical job. Preference for improvement appears to be driven in part by savoring and
dread (Loewenstein 1987), and in part by adaptation and loss-aversion. Savoring
and dread contribute to preference for improvement because, for gains, improving
sequences allows decision makers to savor the best outcome until the end of the
sequence. With losses, getting undesirable outcomes over with quickly eliminates
dread. Adaptation leads to a preference for improving sequences because people
tend to adapt to ongoing stimuli over time and to evaluate new stimuli relative to
their adaptation level (Helson, 1964), which means that people are sensitive to
change
. Adaptation favors increasing sequences, which provide a series of posi-
tive changes—i.e.,
gains
—over decreasing sequences, which provide a series of
negative changes—i.e.,
losses
. Loss-aversion intensifies the preference for im-
provement over deterioration (Kahneman and Tversky 1979).
The idea that adaptation and loss-aversion contribute to the preference for se-
quences, over and above the effects of savoring and dread, was suggested by a
study conducted by Loewenstein and Prelec (1993). They asked subjects first to
state a preference between a fancy French restaurant dinner for two either on Sat-
urday in one month or Saturday in two months. Eighty percent preferred the more
immediate dinner. Later, the same respondents were asked whether they would
prefer the sequence fancy French this month and mediocre Greek next month, or
mediocre Greek this month and fancy French next month. When the choice was
expressed as one between sequences, a majority of respondents shifted in favor of
preferring the improving sequence—which delayed
the French dinner for two
months. The same pattern was observed when the mediocre Greek restaurant was
replaced by “eat at home,” making it even more transparent that the sequence
frame was truly changing people’s preferences. The conclusion of this research is
Do'stlaringiz bilan baham: