The effect of custody arrangement on re-partnering after divorce. Evidence from a policy reform promoting joint physical custody

Jan Van Bavel, Katholieke Universiteit Leuven
Christine Schnor, Vrije Universiteit Brussel
Inge Pasteels, Universiteit Antwerpen
Lindsay Theunis, Vrije Universiteit Brussel

We investigate the effect of the physical child custody arrangement following divorce on re-partnering. Previous studies showed that full-time custody lowers the chances of re-partnering for divorced parents. This paper uses a policy reform from 1995 promoting joint custody as an instrumental variable to identify the causal effect of the presence of children in the household on the transition to a new residential partnership after divorce. Data comes from the Divorce in Flanders (DiF) study and includes 1767 divorced parents. As methodological approach we use a recursive bivariate probit model that estimates the effect of a binary endogenous variable (full-time custody) on a binary outcome (post-marital household formation with new partner within five years after divorce). We find that the negative correlation between full-time custody and re-partnering is causal. Divorced parents with full-time custody have a 32 percent lower probability of re-partnering than parents in other custody arrangements. Failing to consider the endogeneity of the custody arrangement choice leads to underestimating the negative effect of full-time physical custody on the probability to move together with a new partner in the first five years after divorce.