Modelling social inequalities in health in contemporary Switzerland

The relationship between income and health has been widely established. However, there are some issues which need to be addressed in order to study accurately the socio-economic gradient in health. Firstly, there is a specific issue linked to the dynamic nature of health. The current health status is closely linked to past health circumstances and is a good predictor of future health status. Hence we should implement models which are able to take into account the dynamic nature of health. Secondly, the study of health can be affected by a more general problem that often plagues sociological models aimed at estimating causal effects, that of non-observed heterogeneity leading to endogeneity. Both genetic factors and psychological predispositions that are important predictors of health may easily be correlated with one’s socio-economic status, thus creating endogeneity. If analyzed with the appropriate methods, panel data help us tackle these issues. In this work we study the effect of income on self-assessed health in Switzerland using 13 waves (from 1999 to 2012) of the Swiss Household Panel. We apply a fixed effects model, a random effect model and a dynamic panel model (more precisely, the Mundlak–Chamberlain model). Whereas in both the fixed and the random effects models income appears to be significantly related to health, in the Mundlak–Chamberlain model this effect disappears. Results show that models based on different assumptions and including a different set of controls give quite different results.