On the use of honesty priming task to mitigate hypothetical bias in choice experiments

Abstract

We test whether the use of an honesty priming task from the social psychology literature can help mitigate hypothetical bias in stated preference choice experiments (CE). Using a between-sample design, we conducted experiments with five treatments: (1) hypothetical CE without cognitive task, (2) hypothetical CE with cheap talk script, (3) hypothetical CE with neutral priming task, (4) hypothetical CE with honesty priming task, and (5) non-hypothetical CE. Results generally suggest that marginal willingness to pay estimates from treatment 4 where subjects are given honesty priming task before the choice experiment are not statistically different from marginal valuations from treatment 5 where subjects are in a non-hypothetical choice experiment. Values from both these treatments are significantly lower than those from other three hypothetical treatments (treatments 1-3). Using hold out tasks, our results also suggest that one could get higher correct predictions of participants’ choices in treatments 4 and 5 than in treatments 1-3 and that there is no significant difference in percentage of correct predictions between treatments 4 and 5