Preference Reversals

Grether and Plott (1979)

A common way to spot preference reversals is to ask people to make a choice between two options as well as to state their valuation of those same options. For example, when exposed to the question below, 69.39% of participants chose the Left Bet.

…yet when asked to state their valuation of these lotteries, 71.43% of participants priced the Right Bet higher!

In fact, 67.65% of those that chose the Left Bet valued the Right Bet more.

Across all examples from Grether and Plott (1979) in our replication, we find a preference reversal rate of 45.58%.

Slovic, Griffin, and Tversky (1990)

Slovic, Griffin, and Tversky documented preference reversals even in riskless lotteries.

Replicating their study with stakes five times larger, we observe a preference reversal rate of 55.08%!

People choose an option with a higher probability of winning more than 75% of the time, but assign higher prices to the lottery with the larger reward (but lower probability) 56.78% of the time.

Tversky, Slovic, and Kahneman (1990)

Preference reversals also occur in intertemporal choices. In the example below, 83.33% of subjects chose the first option.

But 64.81% value the delayed payoff more.

Mean response: $1510

Mean response: $1906