alternative jobs that are like their present job in all respects except for the
amount of social contact and the daily commuting time. The relevant informa-
tion is summarized in the accompanying table. Subjects are divided into two
groups: one group is told that they presently hold job A, the second group is
told they presently hold job B. Both groups are then asked to choose between
job X and job Y. Because their current jobs are said to be ending, maintaining
the status quo is not an option.
Social Contact Daily Travel TimePresent Job A isolated for long stretches 10 min.
Job X limited contact with others 20 min.
Job Y moderately sociable 60 min.
Present Job B much pleasant social interaction 80 min.
Notice that both X and Y are better than A and worse than B with respect to
social contact, and both are worse than A and better than B in terms of com-
muting time. According to standard economic analysis, the choice between X
and Y should not depend on the decision maker’s current reference point. On
the other hand, if subjects treat their present job as a reference point and if dis-
advantages relative to this reference point loom larger than corresponding
advantages, then subjects are more likely to choose the job with the smaller
disadvantage relative to their current job. Thus, subjects who currently hold job
A are expected to favor job X, whereas subjects who currently hold job B are
expected to favor job Y. The data confirm this expectation: more than two-
thirds of subjects in each group chose the predicted option.
Loss aversion, or the asymmetry between the evaluation of gains and losses,
emerges as an important empirical generalization that has implications for a
wide range of decisions. It promotes stability rather than change by inducing
people to maintain their current position. A loss-averse individual at position X
would be reluctant to switch to position Y, even though, were she at position Y,
she would be reluctant to switch to X. Along these lines, the reluctance to
change induced by loss aversion can hinder the negotiated resolution of dis-
putes. If each side to a dispute evaluates the opponent’s concessions as gains
and its own concessions as losses, then agreement will be hard to reach because
each side will perceive itself as relinquishing more than it stands to gain. A
skillful mediator may facilitate agreement by framing concessions as bargain-
ing chips rather than as losses.
26.5 Eliciting Preference
Preferences can be elicited by different methods. People can be asked to indi-
cate which option they prefer; alternatively, they can be asked to price each
option by stating the amount of money that is as valuable to them as that op-
tion. A standard assumption, known asprocedure invariance, demands that log-
ically equivalent elicitation procedures should give rise to the same preference
order. Thus, if one option is chosen over another, it is also expected to be priced
higher. Procedure invariance is essential for interpreting both psychological
and physical measurement. For example, the ordering of physical objects with
610 Eldar Shafir and Amos Tversky