Jobless durations of displaced workers: A comparison of Canada and the United States

This paper deals with one facet of the unemployment rate gap between Canada and the United States that started in the early 1980s. We seek to analyse discrepancies in the search behaviour and environment of displaced workers which give rise to a higher observed average jobless duration for Canadian workers. A common hazard function is estimated from a data set which combines comparable information from the American and the Canadian Displaced Worker Surveys for 1986. A descriptive analysis of the characteristics and the distribution of jobless spells of displaced workers in the two countries reveals some relevant differences across countries. The results from the regression model are roughly similar for the two countries, with the exception of significant differences in the impact of a few variables, such as the cause of displacement (plant closure versus production cutback); there is also a higher degree of negative duration dependence in the US. An empirical decomposition exercise suggests that the differences in the characteristics of displaced workers and their labour markets are relatively more important than differences in the estimated regression coefficients of the corresponding variables in generating a longer average duration in Canada.