Experiments in Altruism and Charitable Giving

This dissertation introduces three sections on giving, employing a variety of methods in different environments to investigate giving. The first section, utilizes a simple incentivized game to assess differences in risk-sharing norms across communities. In the game, subjects decide whether to share resources with anonymous group members who have lost a lottery where everyone in the group has the same potential for a positive or negative outcome. To gauge the impact of formal institutions on informal risk sharing, subjects make a second sharing decision; but this time they have the opportunity to purchase simple insurance, which guarantees a positive outcome. I found that insurance crowds out informal risk sharing; the amount people share decreases significantly when self-insurance is offered, no matter whether the decision maker chooses to insure. The second section, studies the effects of successful fundraising campaigns on individuals’ philanthropic behavior. We investigate the source of funds that are raised in successful campaigns. Using a controlled lab experiment, we ask whether new funds are raised, or if individuals merely redirect funds from contributions to other organizations following a successful campaign. The results show that a successful campaign increases funds raised by the charity. However, the increase in giving to the target charity comes entirely at the expense of the other charities. This provides strong evidence for a ‘crowding-out’ effect for targeted campaigns. My third and final section, investigates cooperation, giving, and the effect of punishment in two simple games, the trust game and the public goods game. Notably, the paper examines the puzzle presented by past experimental results of punishment effects in the games. Punishment increases contributions in the public good game, while decreasing cooperation in the trust game. To test the effect of punishment on various sets of game parameters, subjects play modified versions of the two games. By observing the differences in contributions with and without punishment, I find design features are not related to the different effects of punishment