A multilevel analysis of the connection between female labour force participation and divorce in Canada 1931-1991.

Abstract

This study investigates the causal order between women's employment and divorce in Canada 1931-1991. An econometric model-the Granger-Hsiao test-is applied to time series data to identify the form and direction of the relationship between these two events and predict their pattern of change over time. Autoregressive estimates show that increased entry of women into the work force was a causal factor in the rise in divorce rates from 1931-1969; after 1969 the direction of causality switched. Relative risk estimates obtained from individual-level survey data using Cox's proportional hazard models confirm the time series results but, in addition, show that labour force participation was a significant predictor of the risk of marital dissolution only among women who married between 1950 and 1969 and who worked without interruption. Logistic regression results show that compared to women who did not experience marital disruption, divorced and separated women had higher odds of being employed while those who remarried were less likely to be employed. This study thus demonstrates that the causal connection between these two events have changed over time among Canadian women, and that analyses at different levels of data measurement can be used to support, refine or refute results obtained at either level.