Displaced Worker Earnings: New Theory and Observation.

This dissertation applies theory and empirical analysis to help explain the earnings experience of involuntary job losers. Chapter II shows that sheer misfortune can account for the entirety of earnings losses experienced by displaced workers. Coming out of unemployment, ex-ante identical workers start in employment relationships that are, on average, less profitable than their jobs prior to separation. Since workers have to look for their ideal job while employed, it takes time for newly hired workers to move up the job ladder and to find a good quality match. This induces a slow recovery in earnings after displacement. The model successfully matches observed wage dispersion, and this in large part explains the success of the model. As a result of serial correlation in displacements, the framework also delivers the empirical decomposition of earnings losses into lower wages and lower employment. Chapter III investigates the implications of job displacement for the aggregate labor market. The framework features large movements of unemployment during recessions. Due to low match-qualities among new hires, many employment relationships exist near the reservation frontier, which results in a large burst of layoffs at the onset of a recession. Since poor quality employment relationships are destroyed at the beginning of a recession, the average match-quality rises initially (cleansing effect). Due to fewer posted vacancies and lower job-finding rates, the average match-quality begins to fall (sullying effect). The model also delivers substantial propagation of aggregate productivity shocks due to the slow movement of workers up the job ladder. Chapter IV uses data from the Panel Study of Income Dynamics to present results from two popular empirical approaches and describes the difference between them. The specification that uses the never displaced as a control group implies much larger earnings losses than a specification that uses those not displaced in a given year as the control group. The analysis shows that despite substantial differences in the incidence of displacement, the earnings losses conditional on displacement are similar. Finally, this chapter presents new evidence on earnings losses by worker characteristics, including heterogeneity by pre-displacement wealth.PhDEconomicsUniversity of Michigan, Horace H. Rackham School of Graduate Studieshttp://deepblue.lib.umich.edu/bitstream/2027.42/100014/1/pawelk_1.pd