Eliciting risk preferences using choice lists

Abstract

We study the effect of embedding pairwise choices between lotteries within a choice list on measured risk attitude. Using an experiment with online workers, we find that subjects choose the risky lottery rather than a sure payment significantly more often when responding to a choice list. This behavior can be rationalized by the interaction between nonexpected utility and the random incentive system, as suggested by Karni and Safra (1987).

Framing and repetition effects on risky choices: A behavioural approach

“…However, the aforementioned bias was not found when they analysed the certainty equivalent (CE) elicitation method. Contrary to these results, Freeman et al (2019) found that embedding a pairwise choice in a choice list increased the fraction of subjects choosing the riskier lottery when the safer alternative was certain, but it did not significantly affect choices when the safer alternative is risky. Erev et al (2008) and Blavatskyy and Köhler (2009) analysed the robustness of the CE mechanism to elicit risk preferences and found that elicited payoffs were systematically affected by the range of certain payoffs to which the lottery was compared.…”

The empirical relationship between nonstandard economic behaviors

“…decisions may not see such choices as providing certainty (53). This could lead to noisy or inaccurate measurement of these variables and an underestimation of any resulting correlations (see SI Appendix, section S1 for a discussion).…”

Loss Attitudes in the U.S. Population: Evidence from Dynamically Optimized Sequential Experimentation (DOSE)

“…Paying for randomly selected questions is incentive compatible under Expected Utility, but not necessarily under more general risk preferences, where it is known that no such mechanism may exist(Karni and Safra, 1987;Azrieli et al, 2018). An old and still growing literature suggests this theoretical concern may not be empirically important(Beattie and Loomes, 1997;Cubitt et al, 1998;Hey and Lee, 2005;Kurata et al, 2009), but there are some exceptions(Freeman et al, 2015). Dynamic designs are generally not incentive compatible, however in practice this is of little concern-see Appendix A.19 The conversion from points to awards can only be done at specific point values, which leads to a slightly convex payoff schedule.…”