Preferences for skewness: evidence from a binary choice experiment

Abstract

In this paper, we experimentally test skewness preferences at the individual level. Several prospects that can be ordered with respect to the third-degree stochastic dominance criterion are ranked by the participants of the experiment. We find that the skewness of a distribution has a significant impact on the decisions. Yet, while skewness has an impact, its direction differs substantially across subjects: 39% of our subjects demonstrate a statistically significant preference for skewness and 10% seem to avoid skewness (at 5% level). On the level of individual decisions we find that the variances of the prospects and subjects’ experience increase the probability of choosing the lottery with greater skewness.