Union-nonunion wage differential: a human capital approach
Description
The objective of this study is to empirically examine union-nonunion wage differential in the context of human capital theory As a starting point, previous theoretical and empirical studies explaining the union-nonunion wage differential are surveyed. Most recent empirical studies surveyed since Lewis' work (1963) have shown that there exists a quite large wage differential of about 15 to 30 percent. In examining the underlying causes which bring it about, most conventional studies stand on the view of the wage differential as monopoly rent. This approach, however, does not fit in explaining several aspects of real phenomena. Recently, several theories to interpret the wage differential from different points of view have been developed This study attempts to analyze the role of unions in the creation of wage differentials via their effect on investment in specific human capital. From the theoretical argument one testable hypothesis follows: the worker-financed stock of specific human capital would be increased under unionism, and thereby some portion of the allegedly higher wage of union workers would be explained by the return to increased union worker-owned specific human capital. The other implication of the thoretical analysis for an empirical testing is that the stock of specific human capital in a particular industry might be proxied by the rehire rate of the industry The basic wage equation to be employed for the empirical test is the expanded human capital model in which the specific human capital variable and other control variables are included. The empirical content of the hypothesis is tested by the introduction of the interaction term between union dummy variable and rehire rate. The primary data for the study were taken from the National Longitudinal Surveys of Labor Market Experience (NLS) for young men. Cross-sectional results show that one-third or one-quarter of union wage premium might be credited to the specific human capital possessed by union members. Additionally, the empirical estimates from the wage change equation, which are utilized to take fuller advantage of the longitudinal nature of the data, provide some indirect evidence for the support of the hypothesis