Preference Reversals

Definition

Preference reversals refer to the observation that there are systematic changes in people's preference order between options. Preference order refers to an abstract [Page 693]relation between two options. It is assumed that when an individual is presented with options A and B, he or she either prefers A to B or prefers B to A (or is indifferent between A and B). Systematic changes refer to the observation that people exhibit different or even reverse preferences for the same options in normatively equivalent evaluation conditions (i.e., conditions that differ at first sight but in which the options that people are presented with have essentially remained the same).

History and Background

The preference reversal phenomenon was first observed in the late 1960s and the early 1970s by Sarah Lichtenstein and Paul Slovic in a gambling context. They observed that if people are asked to choose between a relatively safe bet with a low payoff and a relatively risky bet with a high payoff, and if they are asked to indicate their selling prices if they were to sell these very lotteries, people's choice ordering is systematically different from their price ordering. More specifically, people tend to state a preference for the safer bet but tend to state a higher selling price for the riskier one. Very soon, this finding was replicated several times.

Although the theoretical concept of preferences as an abstract relation between two options seems very clear and natural, this abstract relation is a psychological construct that must be operationalized or measured by some observable behavior. Researchers have introduced multiple elicitation methods or methods that enabled them to observe decision makers' preferences. Besides asking individuals to choose among different options or to indicate how much they are willing to accept to forego or sell an option (i.e., willingness to accept), people have been asked to indicate how much they are willing to pay to obtain an option (i.e., willingness to pay), to state a price that is considered to be equivalent to an option (i.e., a certainty equivalent), or to give the probability of winning an option that is considered equivalent to another option (i.e., a probability equivalent). Researchers then use this information to rank order people's preferences. Consistently, the rank order of preferences produced by one measurement method did not correspond with the rank order produced by a second measurement method. In other words, systematic preference reversals were found repeatedly.

In addition, the preference reversal phenomenon has not stopped at lotteries. Rather than being a peculiar characteristic of a choice between bets, it has been found to be an example of a general pattern. Research has also shown preference reversals when options offering a certain but delayed outcome are used. When faced with a choice between delayed payments, decision makers often select the short-term option but assign a higher certainty equivalent to the long-term option. Different descriptions of the same problem also cause individuals to exhibit different preferences.

Nowadays, preference reversals are firmly established as robust phenomena. Contrary to what researchers assumed originally, preferences do depend on the method of elicitation (i.e., there is no procedure invariance) and they do depend on how the options are described (i.e., there is no description invariance). Rather than trying to eliminate the preference reversal phenomenon as they tend to have done in the past, researchers are now trying to explain it.

...