The purpose of this paper is to study the key determinants of the union status of workers in Canada and to evaluate the relative significance of labour market segmentation by gender, in explaining the lower incidence of unionization among Canadian women. Using a unique micro data set, this study assesses the respective roles of demographic/human capital factors and the industry-occupation of employment in explaining gender differences in union membership in Canada. First, a union status probability equation is estimated on a pooled sample using explanatory variables such as age, sex, marital status, education, job tenure, province of residence, part-time/full-time status of the worker, and industry and occupation of employment, on the assumption that only intercept coefficients differ between men and women. Following this, separate equations are estimated for males and females allowing for differences in slope coefficients. Next, we estimate three separate union membership status equations based on the pooled sample of individuals of both sexes. The first equation contains only demographic/human capital factors as explanatory variables. The second equation adds controls for occupation of the individual to the demographic/human capital vector of variables. Finally, the equation is augmented with controls for industry of employment. The estimated coefficient on the sex variable in the three equations is compared to evaluate the role of industry and occupation of employment in explaining the differential in the extent of unionization between men and women. As well, the likelihood ratio tests are performed to test the joint significance of industry, occupation and human capital/demographic variables as predictors of union membership.
Gender Differences in Union Membership Status: The Role of Labour Market Segmentation
David Cowan, Pradeep Kumar
March 1, 1969