Why the Married Live Longer: Evidence on the Selection versus Protection Effect of Marriage Using Fixed-Effect Survival Models

Hans-Peter Kohler, University of Pennsylvania
Axel Skytthe, University of Southern Denmark

A robust finding in the demographic research is that married people have a lower risk of death as compared to divorced or single individuals. Two main hypotheses have been proposed to explain this relationship: marital selection versus marital protection. In this paper we investigate this interdependence between marriage and mortality proposing new statistical methods and data. We develop fixed-effect survival model that to our knowledge has not been applied to the analysis of mortality and twin data. This fixed-effect survival model allows the estimation of the parameter of interest, which measures the impact of individual characteristics on the level of mortality, without imposing the assumption of independence between the distribution of unobserved heterogeneity and individual characteristics. Our analyses are based on the Danish twin register. Our results suggest that the effect of marriage on mortality may be primarily traced back to the effect of selection rather than to marital protection.