Mixed feelings: Theories of and evidence on giving

This paper examines possible motives and institutional factors that impact giving. Specifically, I consider alternative theories parallel to dictator experiments that generate evidence on both allocation decisions and their effect on feelings. A number of new empirical findings as well as new interpretations for previously reported findings result. A novel test distinguishes warm glow from impure altruism and rules out the former as the sole motive for giving. Very generous donations to charities that aid the needy (with modal gifts of the entire dictator's stakes) cannot be attributed to familiarity with the charities. A charity that offers a matching grant increases its revenues by drawing donors and donations away from one that does not, although aggregate charitable donations do not rise. Additional results on emotions paint a picture of "mixed feelings:" generosity creates good feelings when the recipients are charities and bad feelings when they are fellow students. No group of dictators, however, feels better, on average, than a control group that is given no opportunity to donate. I propose a simple model that accounts for these results on allocation behavior and feelings by incorporating elements of two approaches, unconditional altruism and social preference theories, that to date have mostly evolved independently. A critical feature of this model is the social norm, and the results of the experiments corroborate the theory in the context of two norms of distributive justice that are important to real world giving: equity and need.