During the past decade a substantial effort has been devoted to estimating labor market models of job search. These are important models in that their estimation, especially using panel data, can yield relevant predictions about the effects of various government policies on the unemployment rate and on wages. Several approaches have been taken to model the search for work. Most of the models set the job searcher in the framework of a price taker, that is the distributions of wages across jobs is given, and the person searches for a position which meets at least the minimum wage rate required to entice him into the labor force. This type of model has been criticized on the grounds that the wage distribution might not be stable over even very short periods of time, and more recent analyses have tried to incorporate a changing wage distribution by using models of partial equilibrium in the labor market. This project uses a partial equilibrium framework in which individuals search in a market where productivity varies from firm to firm. The workers themselves are heterogeneous with respect to their desires for employment versus leisure time. The wage distribution is not assumed to be stable independent of the workers, but to change in response to differences among workers' desire for leisure and in response to productivity differences among firms. The model is estimated using data from the National Longitudinal Surveys of Labor Market Experience of Youth in which the data follow several cohorts of high school graduates from the time they graduate until the time they accept their first full-time job. The estimates can be used to test the effects of minimum wages and unemployment compensation on the unemployment rate, and also for testing for labor market segmentation and discrimination, and to give these sometimes loosely defined concepts more rigor. This project is done in collaboration with Kenneth Wolpin of Ohio State University.
