Moral Behavior and Happiness

A growing empirical literature uniformly claims that prosocial (or moral) behavior increases happiness. Previous causal evidence, however, focuses on short-term effects and non-choice data. We test the link between prosocial behavior and happiness in a novel high-stakes paradigm that provides a comprehensive, dynamic perspective on the pattern of derived happiness. In our longitudinal study, each subject either saved one (statistical) human life or personally received 100 euros. Exploiting choice data at the same time as random variation in the actual outcome, our results tentatively confirm the consensus from the literature only for the very short run. Once we assume a longer-term perspective and account for the impact on happiness after a delay of 4 weeks, we strikingly find a large negative causal effect of saving a life relative to receiving money. This reversal over time suggests that positive effects on happiness from policies promoting prosocial behavior would likely be short-lived. Our findings provide a clear rationale for why people are not universally altruistic in practice. Moreover, we report direct evidence on the role of self-image for moral behavior. In line with theoretical models, self-signaling has powerful effects in the short term, but fades over time as those signals are forgotten or imperfectly accessible from memory. JEL classification: D03, I31, D64