This paper illustrates the nexus between Social Choice Theory, the Social Preferences literature (the study of other--regarding preferences in behavioral game theory) and decision making under risk, all of which study preferences over income distributions. Motivated by this connection we introduce a new functional for capturing preferences for fairness based on rank-dependent utility, which we call General Fairness Preferences, that includes most other proposals as special cases. We structurally estimate this functional using
an incentivized laboratory experiment in which individuals decide over income distributions in three different treatments. In the first treatment subjects decide over risky prospects whose outcomes are only pay-off relevant for themselves. In the second treatment subjects decide over income distributions under the veil of ignorance, that is over an income distribution for the whole population without knowing their own position beforehand. Lastly, individuals make decisions while knowing their own position in advance. We introduce another novelty to the experimental economics literature as we generally allow for ten distinct lottery
outcomes using a novel intuitively appealing lottery design. Our results allow us to disentangle risk aversion and social preferences as well as to test whether decision makers follow certain principles of fairness typically assumed in Social Choice theory.