When children are involved in divorce, at least four sets of agents are involved: the children themselves, the two parents, and institutional actors such as lawyers, judges, and social workers. Regarding the determination of divorce settlements, one school of thought advocates that such settlements be defined with respect to the """"""""best interests of the child."""""""" In Del Boca and Flinn (1990), we construct a model in which the behavior of divorced parents and institutional actors can be analyzed. The institutional actors in our approach are assumed to act so as to maximize child welfare, with no direct concern for parental welfare levels. However, since the parents are responsible for the day to day decisions affecting the child, the institutional actor must consider the parental reactions to settlements in arriving at optimal custody arrangements and child support orders. Thus, parental behavior and welfare is properly viewed as a constraint in such a formulation of institutional behavior. We plan to extend our earlier work in a number of directions, all of which deal with the issue of compliance. In the first project proposed below, institutional preferences of the type discussed above will be explicitly estimated using data on settlements and estimates of parental preferences. In the second project, we will use direct observations on the expenditure patterns of custodial parents to determine whether households which receive a large portion of their income from child support incomes have higher expenditure levels on child-specific consumption goods. Finally, we intend to estimate dynamic models of compliance with child support orders. This project involves the use of both structural and non-structural approaches to the problem. The structural approach considered extends the static compliance framework to a dynamic setting in a simple manner by precluding strategic considerations. Some attempts at including strategic behavior in such a model will be attempted at some latter point in the project.
