| The literature on the effectiveness of the United States unemployment insurance (UI) system does not simultaneously address the heterogeneity of the U.S. population and its workplace and consequently may not accurately portray the UI system. In this thesis, I provide the first analysis of the U.S. UI system that simultaneously incorporates the three channels through which UI enhances agents' welfare: consumption smoothing across employment states, improved job matching during the reemployment process, and redistribution of wealth across agents.; In Chapter 1, I calibrate an equilibrium model of agent saving, effort, and employment decisions, in the U.S. economy and UI system that incorporates moral hazard. Comparing labor and capital market measures in this model to those from an identical economy without UI, I find that mean agent welfare would decrease by 3.1% if the current UI system were abolished, but could increase by over 2% if several statutory changes were made.; In Chapter 2, I use dynamic contracting methods to find a social planner's solutions to my model. For two lower bounds on promised expected utility, I find that mean welfare could be increased by over 10% if there were no restrictions on transfers between agents, and by up to 8.5% in the complete absence of transfers across educational groups. Additionally, I find that a fully experience-rated system would have a negligible effect on the unemployment rate.; In Chapter 3, I analyze an optimal UI contract in an economy with duration dependence in the employment process. Numerical analyses of the optimal contract indicate that: (1) while the wage tax rate after reemployment is constant, it should decrease with the duration of unemployment; and (2) there is an endogenous lower bound on the utility promised to agents in the presence of negative duration dependence. Further numerical analyses assess the role of active labor market policies in an optimal unemployment insurance program. |
