Disparités interindustrielles dans les taux de départs volontaires : une étude empirique

This article refers to a LISREL specification to study empirically the determinants of job mobility in Quebec. A LISREL approach is particularly useful for a system of simultaneous equations with latent variables. The industry level data used in this study are for 1982 and combine various statistical sources. The job mobility model includes three endogenous latent variables: job mobility, economic opportunity and risk factors. There are ten exogenous variables associated with labor personal characteristics, human capital, labor environment and labor relations. In general, the results confirm our anticipations. A significative result suggests that risk factors reduce job mobility. This is contrary to previous results, but can be explained from theoretical considerations. The use of LISREL helps to explain the complex issue of job mobility. L’objet de cet article est d’utiliser le modèle d’estimation LISREL dans l’étude empirique des départs volontaires au Québec. Il s’agit d’un modèle d’estimation de coefficients inconnus d’un ensemble d’équations structurelles linéaires avec variables latentes. Les données utilisées sont au niveau de l’industrie. Elles portent sur l’année 1982 et proviennent de diverses sources.