Answers are grounded in this paper and cite their sources.
We survey a representative sample of US households to evaluate the impact of uncertainty on their willingness to insure against a potential loss. We document three main patterns: individuals switch from risk averse to risk loving as risks become more likely; individuals exhibit uncertainty aversion; and risk and uncertainty aversion are negatively correlated. We theoretically show the need for probability weighing to account for the switch from risk aversion to risk loving. We propose a generalization of recursive anticipated utility, which models uncertainty as a two-stage lottery and applies probability weighting to both stages, that can explain our data. We estimate the distribution of preferences in the population and find that most individuals overweigh small probabilities and underweigh large probabilities in both stages. We also suggest how to account for unobserved uncertainty when estimating preferences from insurance data.
No takes yet. Share an insight, caveat, or question.
Gandhi et al. (2025) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: