Job loss from ill-health, smoking and obesity : concurrent evidence for direct and indirect selection

Background and objectives: Health selection into unemployment may be either direct or operate by reference to health-related behaviours rather than health per se (indirect selection). Panel data are desirable to investigate selection effects, and the two types of selection processes may be concurrent. We examine jointly the roles of health and health-related behaviours as precursors of unemployment, in order to disentangle direct from indirect selection processes. Design: The data of a multi-round nationally representative health survey in France were analysed longitudinally, based on three data collection rounds: 1992–5, 1996–8 and 2000–2. Following employees salaried in the private sector and aged 30–54 years at baseline, we explored through logistic regression the influence of non-optimal self-rated health, smoking and obesity on the risk of being found unemployed 4 years later. Results: After adjustment for self-rated health, obesity was found to be a significant precursor of unemployment in women, and heavy smoking had that role in men. After adjustment for smoking and obesity, poor health at baseline was found to be a significant precursor of unemployment in both genders. Conclusion: Those findings confirm the intrinsic role of poor health and of health-related behaviours as precursors of unemployment, with gender-specific patterns for the latter. Public policy prescriptions regarding employees’ protection from job insecurities should integrate appropriate accommodations of health limitations, and the personal factors underlying unfavourable work and health behaviours should be investigated, in order to thwart indirect selection phenomena.