There is a growing debate over police accountability in the United States, and, in particular, a growing interest from the public in police use of force and incidents of officer misconduct. Surprisingly, there is limited empirical evidence on whether or not law enforcement mis- conduct affects the level of crime, citizens’ perceptions of the police, and citizens’ criminal records (and subsequently their education and labor market outcomes).,The first two chapters of this dissertation focus on issues related to police accountability while the final chapter delves into understanding the methodologies behind the identification of treatment effects using duration models.,In Chapter 1, “Going the Extra Mile: the Cost of Complaint Filing, Accountability, and Law Enforcement Outcomes in Chicago,” I study civilian willingness to pay to file a complaint against a police officer. As officers fulfill their duties to protect and serve the community, they often receive complaints from civilians with whom they have interacted. This setting makes policing fraught with agency problems. I use new, detailed administrative data to study the costs and benefits associated with filing a complaint against the police in Chicago. I exploit the fact that allegations without affidavits signed by complainants are considered null. I use variation in people’s distance to the oversight agency where they sign affidavits as a way to study the effect of civilian oversight on policing. An administrative change of location of the reporting center provides a quasi-experimental setup for the identification strategy. A difference-in-difference analysis suggests that a one-standard deviation increase in traveling distance to the reporting center decreases the likelihood of a signed complaint by 6.2 percent for allegations of constitutional violations and 16.3 percent for complaints alleging the police’s failure to provide service. In non-white residential areas, higher injury rates due to use of force and a higher level of force used per arrest were observed as distance from the reporting center increased. Individuals who benefit most from oversight are those with lowest valuation of complaining. I simulate counterfactual scenarios under a policy that would reduce the cost,of signing the complaint. This policy would largely increase the number of investigations and the sustained rates for failure to provide service complaints in the most violent police districts. On the other hand, for allegations of constitutional violations, this policy would reduce sustained rates overall and marginally increase the number of investigations. This research sheds light on the tradeoffs that arise when increasing the cost of reporting police misconduct.,While this paper documents the role of civilians’ oversight on police performance and their willingness to complain based on their incentives, it is also important to understand police officers’ behaviors. As society delegates to police the authority to enforce laws, including the right to use force when needed, the question raised is whether or not police use force more than necessary.,Chapter 2 addresses this question in joint work with J. Grogger, “The Introduction of Tasers and Police Use of Force: Evidence from the Chicago Police Department,” where we study police use of force. In March 2010, the Chicago Police Department changed its Taser policy, issuing the weapons to patrol officers instead of largely restricting their use to sergeants. We used that policy change to obtain difference-in-difference estimates of how the availability of Tasers affected the types of force employed by police, the total number of use-of-force incidents, injury rates per incident, the total number of injuries, and the race distribution of civilians involved in use-of-force incidents. The policy change initially led to a large increase in the use of Tasers, with limited substitution from other types of force. After a period of retraining, substitution between Tasers and other types of force, both greater and lesser, increased. Police injuries fell, but neither injury rates nor the number of injuries to civilians were affected. Contrary to popular arguments in support of Tasers, we find there is no evidence that Tasers led to a reduction in police use of firearms.,Chapter 3 focuses on understanding the methodologies behind the identification of treatment effects using duration models. In joint work with J. Ham, R. Lalonde, and X. Li, “On the Identification, Estimation, and Use of Dynamic Treatment Effects,” we estimate the effect of endogenous training participation on transitions in and out of employment for disadvantaged women in the Job Training Partnership Act (JTPA) study. Decomposing the effect of training on employment into its effects on transitions in and out of employment has the potential to develop more effective programs. We also consider a potentially serious identification problem that arises when individuals do not undertake training immediately, and we propose a test to shed light on this problem. We find that this problem is not important in our context. JTPA classroom training substantially reduced unemployment durations, and thus it complements programs that increase employment durations.