Much of the law and policy encouraging the use of kinship foster families as a placement resource rests on the implicit assumption that kinship foster families constitute a homogeneous population—that is, that they are very much alike. However, findings of studies of different child welfare jurisdictions suggest that the kinship family population is not, in fact, homogeneous. Nevertheless, few—if any—published studies have attempted to differentiate kinship foster families based on family characteristics or foster child outcomes.

In order to address this gap, a recent Chapin Hall study undertook to characterize the differences among kinship families along two related dimensions: family structure and household composition, both of which are related to a range of child outcomes. Using data from a web-based survey of public and private agency caseworkers in Illinois (N = 453), we used latent class analysis to develop a typology of kinship foster families based on indicators of family structure and household composition. Latent class analysis uses the statistical associations between observed indicators to identify similar groups, types, or latent classes. The resulting family types were then compared based on other family characteristics.

The findings reveal four types of kinship foster families, distinguished from one another based on the caregiver’s relationship to the foster child and by whether the family is caring for other related, non-foster children. For example, the largest group of families (41.4%)—labeled empty-nest grandparents—were generally headed by children’s grandparents and had few or no non-foster children living in their household (Table 1). In contrast, almost all of the families in the smallest group (15.5%)—labeled parenting grandparents—had one or more of their own children living with them. The remaining two kinship family types were almost exclusively headed by non-grandparent relatives (e.g., aunts, uncles, and cousins). Like the two grandparent types, however, these family types—non-grandparents with some children (24.7%) and parenting non-grandparents (18.4%)—could be distinguished from one another by whether they had non-foster children currently living in the household.

These kinship family types differ from one another on a number of other family characteristics, including caregiver age, perceived fostering competence, and geographic proximity to foster children’s biological families (Table 2). These findings suggest interesting questions about how these family types may differ from one another in more fundamental ways. For example, are parenting grandparents the group of families most closely allied with the parents of the children for whom they are caring? Most of these families are headed by caregivers who are the parents of foster children’s own mother or father and, on average, parenting grandparent families are located closer to the homes of children’s biological parents than other kinship family types. Also, parenting grandparents were described by caseworkers as being less likely to provide children with a safe living environment than the other four family types, which may reflect attributes or circumstances that these caregivers share with the children’s parents.

Perhaps the most important implication of this study is that, although kinship families may be different in important ways from non-kinship foster families, kinship family populations are not all alike. Differences in family attributes and circumstances may exist, and these differences may lead to very different outcomes for children. In light of the growing number of state and federal policies that draw distinctions between kinship foster family and other forms of substitute care, it is more important than ever that scholars and policymakers test the implicit assumption that the differences between kinship and nonkinship families are greater than those among kinship families themselves.