Files
Abstract
The three chapters of this dissertation are independent of one another.
Gender, Risk, and Signaling: Preferences and Beliefs
In many signaling environments, female senders take fewer risks than male senders. To what extent are these different behaviors a result of different preferences versus different beliefs? In this chapter I present an experiment which parses out the effect of preferences and beliefs on risk taking by male and female senders in a signaling environment. In particular, I compare risk preferences to three beliefs: confidence in one's own abilities, general beliefs about risk, and anticipation of discrimination. Consistent with previous findings, I find that women take fewer risks than men in this signaling environment. However, there is no evidence for differences in risk preferences between men and women. Instead, there is strong evidence for differences in beliefs. In particular, women are less confident in their own abilities than men. There is some evidence for more pessimistic beliefs about risk among women than men and no evidence of anticipated discrimination.
Sense of Control and Willingness to Pay For Non-Instrumental Control
Control can be desirable for two reasons: its instrumental value and its intrinsic value. Separating the instrumental value of control from its intrinsic value can be difficult since most control comes with both. But, it is important to distinguish between them since they play very different roles in economic decision-making. To manage this problem, I focus on control with no instrumental value, which I call non-instrumental control. In particular, in this chapter I present an experiment in which I measure the effect of changes in participant's sense of control on their willingness to pay for non-instrumental control. The results are inconclusive. The sense of control manipulation did not significantly affect participants' sense of control or their willingness to pay for non-instrumental control. However, participants' willingness to pay was significantly positive with notable and predictable heterogeneity.
The ``Blue Wall of Silence'' and Police Careers
Many police departments in the United States have an unwritten code ``which prohibits disclosing perjury or other misconduct by fellow officers, or even testifying truthfully if the facts would implicate the conduct of a fellow officer'' (Chin and Wells 1997). In this chapter, I use administrative data on Chicago police officers and the investigation of complaints against them to create a proxy for instances of officers testifying for or against other officers. I validate the proxy using a subset of data for which validation is possible, and test for correlations between the proxy and important career outcomes: leaving the police force, being promoted to higher pay grades, being accused in future complaints, and having those complaints sustained in investigations. I find that officers who do not testify against other officers are less likely to leave the force and more likely to be accused in future complaints, and officers who do testify against other officers are more likely to have complaints sustained against them.