Can't We Just Live Together?: New Evidence on the Effect of Relationship Status on Health

Abstract

An extensive literature suggests that marriage confers health benefits to men and women. However, several econometric issues raise doubts about this prevalent conclusion. First, the evolution of health is a strongly persistent dynamic process; yet extant research relies almost exclusively on static mean-differenced fixed effects estimators. Second, measures of morbidity vary across studies making comparisons difficult, and the often used categorical measure of self-assessed health raises difficult interpretation and econometric issues associated with non-linear models. Finally, existing research has not confirmed whether these benefits accrue to couples who cohabit. We employ the Blundell and Bond dynamic panel data estimator to control for selection into marriage and a multiple correspondence analysis to identify weights for a continuous health index that reflects the multiple facets of morbidity and alleviates the econometric issues associated with a discrete dependent variable. Contrary to extant literature, our analysis using a balanced panel of nearly 2,430 individuals over 17 waves of the British Household Panel Survey finds that the effect of cohabitation on health is not statistically different from the effect of marriage on health after controlling for both health dynamics and selection.

Keywords: selection effect, dynamic panel data, health and marriage

JEL Classification: I1, J12