Full bibliography

Panel data obtained from Canada's Labour Market Activity Survey are used to estimate union effects on wages, wage dispersion, and pension coverage. The selectivity corrected estimate of the union-nonunion wage differential is in the 13.1%-15.5% range, which is consistent with other selectivity corrected estimates for Canada but most notably those by Grant, Swidinsky, and Vanderkamp (1987). The estimated union effect on wage dispersion is less pronounced: The standard deviation of 1n wages falls by .02 (or by roughly 5%) when workers become unionized. However, the union effect on the probability of pension coverage is considerably greater. Selectivity corrected estimates indicate that a worker employed in a union job is 22%-25% more likely to have pension coverage than if the individual is employed in a nonunion job.