Experiments on bivariate risk preferences

We study cross-risk preferences over wealth and two other attributes to obtain theory-free evidence on correlation aversion as well as higher-order cross-traits like cross-prudence and cross-temperance. Two experiments elicit the dependence structure of risk preferences between wealth and, respectively, waiting time and someone else's wealth. The latter experiment, which could be called ``risk preferences for you, me, and us'', connects bivariate risk preferences with (higher-order) inequality aversion. This first systematic exploration of bivariate risk attitudes helps assess the appropriateness of assumptions in economic models on health prevention, inequality, saving, or insurance.