An essay on labor market dynamics : theory and evidence

Abstract

A stochastic model of the labor market is developed in which job search takes place within a framework which describes participation decisions as a Markov process. Prospective labor force participants, faced with a known distribution of wage offers, must invest resources to obtain information about actual wage rates paid by particular firms. Decisions concerning entry, the continuation of search once begun, or withdrawal from the labor force are made each period on the basis of a comparison of the value of a non-market alternative with the expected gains from additional periods of job-seeking activity. Through aggregation of individual decisions, the basic units of observation become various "states" in the process. These are (1) not in the labor force with no previous employment experience (the source of new entrants); (2) not in the labor force, but having held at least one previous job (the source of reentrants); (3) employed; (4) unemployed and actively searching; (5) new hires waiting to report to work; and (6 ) on layoff and waiting to be recalled. By solution of a system of differential equations, the steady-state proportions of the total population expected to be found in each of the six states are determined in a probabalistic sense. Further analysis establishes the direction of parametric changes on the steady-state probabilities. A stability requirement of the model indicates that there must be, on average, a positive probability that unfilled job vacancies exist. Further, unemployment is positive even in the long run..