Preference Reversals and Scale Compatibility

intermediate3 min readpsychology · economics · behavioral_science

The compatibility hypothesis explains why preferences between delayed payoffs reverse depending on the evaluation task: in pricing, the larger but longer-delayed option is often overvalued compared to choice, leading to systematic reversals that stem from scale compatibility between payoffs and pricing.

THE IDEA

Study 4: Time Preferences

The compatibility hypothesis entails that preference reversals should not be restricted to risky prospects and that they should also be found in riskless choice. The present study investigates this hypothesis using delayed payoffs that differ in size and length of delay (see Tversky, Slovic, & Kahneman, 1990). Consider a delayed payoff of the form (X, T) that offers a payment of X dollars T years from now.

One hundred sixty-nine students from the University of Oregon participated in a study of choice between delayed payoffs. Half the subjects first chose between S and L in each pair and later priced all eight options by stating “the smallest immediate cash payment for which you would be willing to exchange the delayed payment.” The other subjects performed the choice and pricing tasks in the opposite order. There were no systematic differences between the groups, so their data were combined.

Because both the given payoffs and the stated prices are expressed in dollars, the compatibility hypothesis implies that the payoffs will be weighted more heavily in pricing than in choice. As a consequence, the preference for the short-term option (S) over the long-term option (L) should be greater in choice than in pricing. Overall, S was chosen over L 74% of the time, but S was priced higher than L only 25% of the time, yielding 52% preference reversals, as compared with 3% reversals in the opposite direction. The application of the diagnostic analysis described in Tversky, Slovic, and Kahneman (1990) revealed that, as in the case of choice between simple bets, the major determinant of PR was overpricing of the long-term option, as suggested by compatibility.

In the pricing task, each option is evaluated singly, whereas choice involves a direct comparison between options. The standard demonstrations of PR, therefore, are consistent with the alternative hypothesis that payoffs are weighted more heavily in a singular than in a comparative evaluation. To test this hypothesis against compatibility, we replicated the study presented previously with a new group of 184 students from the University of Oregon, with one change. Instead of pricing the options, the subjects were asked to rate the attractiveness of each option on a scale from 0 (not at all attractive) to 20 (extremely attractive). If PR is controlled, in part at least, by the nature of the task (singular versus comparative), we should expect L to be more popular in rating than in choice. However, if PR is produced by scale compatibility, there is no obvious reason why rating should differ from choice. Indeed, no discrepancy between choice and rating was observed. Overall, S was chosen over L 75% of the time (as in the original study), and the rating of S exceeded the rating of L in 76% of the cases. Only 11% of the patterns exhibited PR between choice and rating, as compared to 52% between choice and pricing.

Study 3 showed that the use of nonmonetary prizes greatly reduced the amount of PR, whereas Study 4 demonstrated substantial PR in the absence of risk. Evidently, preference reversals are controlled primarily by the compatibility between the price and the payoffs, regardless of the presence or absence of risk.

When to use it

Choosing between a smaller immediate bonus and a larger deferred bonus in a work contract—pricing the options might lead to overvaluing the deferred bonus.
Investment decisions between short-term and long-term financial instruments; when pricing, investors might overweight the long-term return.
Evaluating credit card offers: an upfront cashback vs. a larger rewards point accumulation over time—choice may favor immediate cashback while pricing might favor points.
Negotiating a settlement: immediate lump sum vs. future installments—pricing the installments could lead to a preference reversal.

FROM THEORY TO EVERYDAY LIFE

In practice

Original Study: Choice vs. Pricing

169 students chose between short-term (S) and long-term (L) delayed payoffs and priced all eight options. Half did choice first; the others did pricing first. S was chosen 74% of the time but priced higher than L only 25% of the time. The authors report 52% preference reversals favoring L in pricing.

Replication with Rating Task

184 students repeated the same choice pairs but instead of pricing, they rated each option's attractiveness on a 0-20 scale. S was chosen 75% of the time and rated higher 76% of the time, with only 11% preference reversals between choice and rating, supporting the scale-compatibility explanation over the proposed singular-versus-comparative explanation.

✳

A MOMENT TO REFLECT

Put it to work.

Think of a time you chose a smaller reward now over a larger reward later. If you had to put a price on the larger reward, would that change your preference? Why?
How might scale compatibility affect your evaluation of job offers that differ in immediate salary versus long-term benefits?
When comparing options directly, do you think you weigh outcomes differently than when you evaluate them one at a time? Can you recall an example?