Three Essays on Microeconomic Theory and Experiment

My dissertation uses game theoretical and experimental approaches to study how individual's behavior in different informational environments affects economic outcomes and motivations for charitable giving. Chapter 2 "Bargaining with Uncertain Value Distributions" studies a bargaining model in which the seller is uncertain not only about the buyer's value but also about which distribution the buyer's values are drawn from. Different from the classical models, the distribution of the buyer's values is fixed across periods, while the buyer's values are drawn independently from the distribution each period. I find that adding this additional layer of uncertainty improves the seller's profit when her ex ante beliefs are sufficiently optimistic. Chapter 3, "Social Norms, Information, and Trust among Strangers: An Experimental Study" (with John Duffy and Yong-Ju Lee), investigates whether norms of trust and reciprocity arise in response to different reputational mechanisms. We conduct an experiment where anonymous subjects are randomly matched each period and play a series of indefinitely repeated trust games. We find that the social norm of trust and reciprocity is difficult to sustain without reputational information, although it is supported as an equilibrium by the parameters. The provision of information on players' past decisions significantly increases trust and reciprocity. Furthermore, making such information available at a small cost also leads to a significant improvement, despite that most subjects do not choose to purchase this information. Finally, Chapter 4 "Motives for Charitable Giving" (with Lise Vesterlund and Mark Wilhelm) reports an experiment which tests the pure altruistic and the impure altruistic explanations for charitable giving. We focus on the comparative static predictions of both models and quantify the relative weight attached to the warm-glow component of giving in the impure altruism model. A methodological innovation is to create the equivalent of a series of real-world charities. Each participant is paired with a child who has suffered a severe fire and informed that they single-handedly determine the size of a gift given to the child. Our results show that participants behave as predicted by the impure altruism model. However the relative weight attached to the warm-glow of giving is very small.