Conditions for extrapolating differences in consumption to differences in welfare
研究了在什么条件下，更好的消费分布意味着更高的福利，通过反例和不同设定（如固定依赖或秩不变假设）来阐明这一关系，并结合现实例子讨论。
Abstract We characterize conditions under which a better consumption distribution implies higher welfare. Specifically, here “better consumption” means first‐order stochastic dominance, and “higher welfare” means higher expected utility for every subpopulation of individuals with the same utility function. Although this implication seems natural, we first provide a counterexample wherein better consumption risk allocation outweighs lower consumption. We then show that higher expected utility results from higher consumption in different settings, including fixed dependence (fixed copula) between consumption and individual risk preferences, or alternatively using the rank invariance assumption from the treatment effects literature. These are discussed in several real‐world examples.