Canadian Union-Nonunion Wage Differentials

Abstract

Using extensive Canadian longitudinal data from the years 1969–71, the authors estimate union-nonunion wage differentials of 12–14 percent for 1969 and 13–16 percent for 1970. These estimates are not adjusted for selectivity because three different tests to identify selectivity yield no evidence of selectivity bias. The authors argue that although testing for selectivity is often essential, selectivity adjustments have resulted in greatly inflated estimates of union-nonunion wage differentials in some studies and should therefore be used with caution.