Certification outcomes and returns to shareholders in Canada

Event study methods are used to estimate the effects of union certification applications on the returns to shareholders in Canada. Two methods of inference are employed: a classical method, and a resampling method which makes no assumptions about the distributions of share returns. Certifications granted without a representation vote have virtually no effect on returns, whereas certifications granted after a vote have a negative impact. Dismissed applications that required a representation vote also have a negative effect. These results suggest that the certification process itself is important; providing another explanation for the difference in the experiences of the Canadian and U.S. labour movements.