The Effects of Unions on Wages: Findings from Pooled Data

Cross-section and time-series data from the Panel Study of Income Dynamics from 1967 to 1981 are pooled generating a data set large enough to allow calculation of union-nonunion wage differentials for as many as 260 subgroups in the population. No previous study has examined as many groups. The diversity of groups considered within the framework of a common model and identical variable definitions allows a more precise intergroup comparison of union wage effects than has been possible hitherto. Many of the earlier results in the literature are supported, while some fresh ones are found.