| dc.description.abstract | This dissertation addresses three policy-relevant issues using experimental methodology. These studies illustrate the value of using experimental methods to study policy questions.
First is the blood donation game, which is related to health policy. In this essay I design an experiment to investigate two incentive treatments intended to increase donations, a waiver of blood transfusion fees and priority access to blood supplies. I find that both the waiver treatment and the priority treatment significantly raise donations, and combining the two incentives has the most impact. Second, I study private provision of public goods and examine how introducing the possibility that group members can punish each other affects provision, varying group sizes. Results indicate that introducing a punishment institution has no effect when group size is small. However, for large groups, introducing the punishment institution dramatically increases provision. Finally, I use a popular contest game, the non-constant-sum Colonel Blotto game, to study economic policy. This game mimics the R&D investment decisions of companies that compete with each other. I investigate two factors that might make it easier for firms to engage in a kind of tacit collusion, collectively lowering their investment in innovation: the stability of the relationship (pairs are either stable over time or re-matched each round), and the number of prizes (which proxy for inventions). I conclude that subjects are more successful in tacitly colluding when groups are stable, regardless of the number of prizes. However, when randomly re-matched every period, subjects only collude when there are more prizes: that is, having more prizes facilitates tacit collusion. Both players are worse off when the number of prizes is small because they increase their bids to compete more aggressively against each other. | en |
