The Determinants of Workers' Union Status: Evidence from the National Longitudinal Surveys

Using data from the 1980 wave of the young men NLS sample, this paper provides estimates of the queue model developed by Farber to explain the process of union status determination. The use of NLS data largely avoids the censoring problem encountered by Farber in estimating his model and permits direct estimation of crucial probabilities associated with the model. Despite important differences between econometric methodologies and the data sets analyzed, the NLS results support Farber's conclusion that queueing for vacancies in existing union jobs is an important labor market phenomenon. In particular, supply constraints on the availability of union jobs are found to dominate worker preferences in explaining differences in extent of unionization by occupation and region but not race.