Association for Public Policy Analysis & Management

For most of the twentieth century, the policies governing children’s living arrangements after divorce were to set custody based on the “tender years” doctrine or the “best interests of the child,” both of which resulted in the vast majority of children living with their mother, with occasional or regular visitation by their father. Policy began to change in the 1970s to allow children to spend substantial periods with both parents (shared custody). This increase in nonresident father involvement could be quite important for child well-being, but little is known about the level of shared custody, or the types of cases in which it is awarded.

A substantial portion of the reason for our lack of knowledge comes from measurement and data limitations. One issue has been an inability to separate physical custody (where children live) from legal custody (which parent makes major decisions). Second, accurately ascertaining living arrangements, when these can vary by the week, by holidays and vacations, and by summer versus the school year, has been difficult. Third, ex-partners can give substantially different reports of their children’s arrangements. Finally, the lack of comparably-collected data over a substantial period limits studies of trends.

In this paper we use unique court records from 21 Wisconsin counties to examine custody arrangements upon divorce. Our data enable us to meet the prior research limitations. The court record differentiates physical and legal custody and shows that these are often different. The court record provides substantial detail about the child’s living arrangements, enabling us to count the number of overnights with each parent over the course of a year. This means we can differentiate between equal-shared custody and unequal shared custody (one parent has the child 25-49% of the overnights). Using administrative records mitigates problems caused by differential reporting. Finally, data have been collected in a similar way over a long period, 1988-2010.

Updating and extending our earlier analyses that showed significant increases in shared custody, we estimate that shared custody (combining equal-shared and unequal-shared) has now replaced sole-mother custody as the most common post-divorce parenting arrangement – accounting for just over half (50.3%) of all cases in the most recent cohort. In contrast to the early period, equal-shared custody is now twice as likely as unequal-shared custody, reflecting the growing involvement of fathers in their children’s lives after separation. Our empirical analyses predict custody by considering changes in policy and changes in the characteristics of families entering the court system. We do not find substantial changes at the point of policy change. We do find that as shared custody has become more common, it has become less distinctive, so that it is now a common arrangement for any size family, for any number of children, and for any gender combinations of children. Higher-income families are still more likely to share custody, but this relationship is now weaker than it was early in the study period. We discuss the implications for policy and practice.