Mixed Feelings: Theories and Evidence of Warm Glow and Altruism

This paper presents theoretical and empirical analyses of experiments that test competing theories of altruism, including pure altruism (a preference for the well-being of others), warm glow (a good feeling from giving) and impure altruism (a combination of pure altruism and warm glow). These theories produce different predictions regarding crowding out, i.e., the reduction in private donations due to public spending. Variations on dictator experiments involving both students and charities examine the incidence of crowding out and provide a new direct measure of the effect of giving on feelings. The results indicate that crowding out is incomplete, i.e., less than dollar for dollar. The evidence on warm glow suggests mixed feelings: giving may be associated with good or bad feelings, depending on the context. As a way to resolve apparent inconsistencies and reconcile the evidence on crowding out and feelings, this paper proposes a theory of conditional altruism, which extends previous models to incorporate social norms that arise in the workplace, marketplace and laboratory.