Search Results

In Chapter 2, I develop a dynamic (search-matching-bargaining) model of employer discrimination with on-the-job search by workers. Workers are either black or white, and employers differ by whether they discriminate against black workers or not. I analyze the impact of taste-based employer discrimination on wages and job-to-job turnover for both races. I use data from National Longitudinal Survey of Youth 79 (NLSY79) to find model parameters that matches some carefully chosen moments from the data. Two variables of particular interest are the extent of disutility suffered by prejudiced employers from employing minority workers, and the proportion of such employers in the market. I find that the extent of discrimination is significantly lower than what has been found elsewhere, although they vary slightly across specifications used. Discrimination accounts for between 6% to 8% of differences in mean wages vi between races, depending on the controls used. Also, differences in exogenous job-termination rates play an important role in generating wage differentials. In Chapter 3, the assumption of an exogenous job-separation rate to unemployment is relaxed. In each period, workers experience productivity shocks that changes the match quality in the current, match. Consequently, some matches are rendered unpr ofitable, and hence the worker chooses to be unemployed. While discrimination is still a minor contributor to wage differences, it's impact is more profound on job-to-unemployment transition rates, whereas productivity differences are a much bigger contributor to wage differences as well as significant contributor to job-to-unemployment movements. The magnitude of the productivity shocks needed to generate differences observed in the data are not much different between blacks and whites.