The Determinants of First and Second Marital Dissolution. Evidence from Britain

This paper investigates whether the determinants of first marriage dissolution differ from determinants of second marriage dissolution. Using data from British Household Panel Survey, I estimate a simultaneous hazard model for the risk of ending a first and a second marriage, taking into account unobserved heterogeneity. The model is estimated for men and women, separately. Cross-equations unobserved heterogeneity is not statistically different from zero. The most interesting findings is related to cohabitation before marriage: it is found to affect first and second marriage dissolution risk differently, namely it is found to increase the hazard of first marriage dissolution, but to reduce the hazard of second one.