For better and for worse: The relationship between union dissolution and self-assessed health in European Panel Data

We investigate the association between union dissolution and self-assessed health in European panel data. Previous studies suggest that this association might be negative, yet it is unclear to what extent this reflects causation (an effect of union dissolution of health) and/or selection (an effect of health on union dissolution). We analyzed the relationship between self-assessed health and 3,894 union dissolutions in about 60,000 respondents aged 18–59 across eight waves of the European Community Household Panel (ECHP). Fixed-effect panel analyses show that the negative association between dissolution and self-assessed health is not due to a general negative effect of divorce on health. Rather, after union dissolution self-assessed health improves among some individuals, while for others it declines. Compared to being in a union continuously, union dissolution seems to cause increases and decreases in health. We also find evidence for a negative effect of self-assessed health on dissolution risks. This selection effect appears to be associated with being in poor self-assessed health for a prolonged period of time rather than by the immediate effect of a health decline. Our results support the idea that the association between union dissolution and self-assessed health is an outcome of both causation and selection and that the effect of union dissolution on self-assessed health is heterogeneous.
Keywords: Union dissolution, Divorce, Self-assessed health, Panel data, Fixed-effects models, Event-history analysis, Europe