An analysis of 8,251 homeless children in New York City found that 18% of them received child welfare services over the five-year period following their first shelter admission, and an additional 6% had a history of having received such services before their first shelter admission. Recurrent use of public shelters, exposure to domestic violence, older age at first episode of homelessness, and larger number of children in a household were associated with an increased risk of child welfare involvement. The high rate of crossover between homelessness and the child welfare system suggests the need for service coordination for children in homeless families.
Although researchers have found high rates of childhood experiences in out-of-home care among homeless adults (Bassuk et al, 1997; Burt et al., 1999; Herman, Susser, Struening, & Link, 1997; Piliavin, Sosin, Westerfelt, & Matsueda, 1993), the dynamics between child welfare services and homelessness among children are largely unexplored. Researchers have noted that economic hardships, housing instability, and psychological distress related to homelessness lead parents to place their children with extended family members or friends, or lead child welfare services to intervene (Nelson, 1992; Steinbock, 1995; Williams, 1991). McChesney (1995) noted that homeless mothers were more likely than housed mothers to have an open child neglect or abuse case or to have a child protective services caseworker assigned to them.
Although children in homeless families are considered at great risk of child welfare involvement, only a few empirical studies have examined the prevalence of child welfare involvement among this population. A study of 178 sheltered homeless adults in Maryland found that among parents of minor children, approximately 6% had at least one child in foster care (DiBlasio & Belcher, 1992). A study of 179 homeless women in California showed that 15% of homeless mothers had children in the formal foster care system (Zlotnick, Robertson, & Wright, 1999). A study of 543 homeless and housed low-income families in New York City found that homelessness was an important predictor of mother-child separations among those families and that 11% of children from homeless families were in foster care (Cowal, Shinn, Weitzman, Stojanovic, & Labay, 2002).
These studies, however, were based on small samples and retrospective self-reports by homeless parents. Using an alternate approach, J. E Culhane, Webb, Grim, Metraux, and Culhane (2003) followed an entire one-year cohort of children born to women in Philadelphia and found that 37% of children of ever-homeless mothers received in-home services or were placed in foster care. This study is limited in that it did not look into the sequencing of homelessness and child welfare involvement.
The present study prospectively follows a cohort of children in sheltered homeless families to examine the extent of their subsequent involvement in the child welfare system. This entails examining child welfare service use among the children before and after their initial episode of homelessness. In addition, this study examines the associations among experiences in the homeless shelter system, children's demographic and familial characteristics, and children's entry into the child welfare system, either through a foster care placement or nonplacement preventive services.
To examine child welfare involvement among homeless children, this study used administrative data from two New York City service systems: the Department of Homeless Services, which maintains comprehensive records of shelter use and basic demographic information on family shelter users, and the Administration for Children's Services, which maintains the Child Care Review Service database on children who enter the child welfare system. The data on family shelter users were available back to 1986, and information on children in child welfare was available from 1981 through 2001. …