Playing it Safe: Actions Attractive to the Risk Averse

We introduce a way to compare actions in decision problems. One action is safer than another if the set of beliefs at which the decision-maker prefers the safer action expands as the decision-maker becomes more risk averse. We provide a full characterization of this relation, show that it is equivalent to robust conceptions of single-crossing and second-order stochastic dominance, and reveal that in monotone decision problems it totally orders the decision-maker’s set of actions. We discuss applications to games, insurance, investment hedging, and security design.