Marital dissolution among young Americans: the role of income variables and past family formation behaviour

This paper presents an econometric model of marital dissolution using data on young American women. A distinguishing feature of this paper is that the initial conditions (the family formation process leading up to divorce) are modelled explicitly, and thus included as covariates in the divorce transitions. A considerable advantage of this approach is that marriage (and divorce) and fertility are modelled in a joint econometric framework. The paper examines how economic variables affect marital dissolution. In addition to the respondent's predicted wages, welfare benefits, and family income, we introduce a new income variable which represents the respondent's earnings capacity (or quality) which is unlikely to be fully anticipated by the partner at the time of marriage. The model is estimated using data on young American women from the NLSY. The results indicate that the respondent's earnings quality tend to be positively associated with divorce, whereas positive values of the spouse's earnings quality is negatively associated with divorce. Quality has relatively little explanatory power when there is no marital specific capital present in the household. The other income variables are also found to be important determinants of divorce. Furthermore, the results indicate that past family formation transitions are crucial in determining marital dissolution.