Essays on Corruption and Enforcement

Thesis (Ph.D.)--University of Washington, 2016-08

This dissertation examines the role and limits of enforcement in environments where the enforcing party faces issues related to corruption and misreporting. In the first chapter, we construct a principal – agent – auditor model with adverse selection in which the principal optimally allows bribery to occur due to the potential for extortion. This result mirrors the moral hazard model of Khalil, Lawarrée and Yun (2010). I introduce a probability that the agent is “honest” in that she cannot lie or engage in corruption. Because the principal cannot distinguish who is honest and who is not a priori, the existence of honest agents creates an additional dimension of adverse selection. Honest agents cannot reduce their expected penalties through bribery, and strategic agents can pretend to be honest, so the principal must either allow additional rent for all dishonest agents or shut down honest, low-income agents. This would avoid the new adverse selection issue, at the cost of revenue. In this way, honesty hurts the principal. In the second chapter, I examine the optimality of commitment to an auditing scheme. Some auditors forgo committing to audit when they seemingly have the capability of doing so, such as the IRS. One explanation is that auditors tend to receive subjective information that may be too late to affect production and transfers but could affect auditing decisions, and contracting upon this information creates additional incentive issues which prevent the auditors from effectively committing to future actions. I construct a principal – agent model with costly auditing occurring after production and transfers, and I find that the principal may optimally forgo commitment to auditing when he observes a strong private signal of the agent’s type prior to auditing. In the third chapter, we examine how the death penalty affects the incidence of reported child rape for different types of offenders. In 2008, the Supreme Court ruled capital punishment for child rape unconstitutional. We use this natural experiment to find evidence that the death penalty reduces reported child rapes for strangers and acquaintances. This evidence suggests that capital punishment deters potential offenders from committing child rape.