How do Suggested Donations Affect Charitable Gifts? Evidence from a Field Experiment in Public Broadcasting

David Reiley, University of Arizona

Abstract

Direct-mail charitable fundraising campaigns frequently provide response cards with checkboxes and suggested donation amounts, along with an additional checkbox with a fill-in-the-blank amount. Previous economic theory on charitable giving does not predict any effect of such suggested donation amounts, because people can always choose to write in their preferred amount if it doesn't correspond to one of the suggested donation amounts. However, the prevalence of this fundraising strategy suggests that it might have an effect. This raises two questions: (1) do these amounts actually matter, and (2) what can we say about optimal policy for the charity on suggested donation amounts? To address these questions, I conduct an experiment in cooperation with three local public-broadcasting stations (one TV, two radio) to assess the effects of varying these suggested donation amounts. We assign each of the 25,000 solicitations to be in one of four different experimental treatments for a direct-mail campaign beginning in November 2003, in order to observe the effects of the different suggested-donation strategies on overall giving behavior. Preliminary results indicate that when suggested-donation amounts increase by 20% across the board, the number of gifts decreases by 10%, as if there were a "downward-sloping demand curve" for the charity. None of our four experimental treatments seems to provide significantly more revenue than the others, though the data suggest further experimental treatments that might come closer to the optimal fundraising strategy. I discuss the need for improved theory in this area, and point out some directions the theory might go.