One frequent consequence of homelessness among families is separation of children from their parents. In a national survey, 60 percent of homeless women and 41 percent of homeless men had at least one minor child, but only 39 percent of women and 3 percent of men lived with any children (Burt et al., 1999). Some separations occur in shelter systems that do not allow men or older boys to be housed with women and younger children. But separations are common even in cities where families can be housed together. Some separations are associated with formal placements in foster care; probably more are informal arrangements where children stay with relatives or friends. For example, Park and colleagues (2004) found that 16 percent of 8,251 children under 16 who entered shelters with their families in New York City for the first time in 1996 spent some time in out-of-home placements in the child welfare system (before shelter or in the next five years). In the same city, Cowal and colleagues (2002) used interview data that tracked informal as well as formal placements in a smaller sample of 543 homeless and low-income housed families sampled in 1988. They found that five years after entering shelter in New York City in 1988, 44 percent of mothers experiencing homelessness had become separated from one or more of their children, compared to only 8 percent of continuously housed mothers. Consistent with Park and colleagues (2004), fewer than half of these separations were the result of actions by child welfare authorities or the courts, and many children were in informal placements. Similarly, Bassuk and colleagues (1997) in a study of 167 homeless and low-income families found that 19 percent of preschool-age children in homeless families had been placed in foster care, as compared to 8 percent of the low-income children.
There is some literature on the circumstances of family separations. Cowal and colleagues (2002) found that the mother’s drug dependence, institutional placement (most often for drug treatment), and experience of domestic violence each made independent contributions to the prediction of separation. Homelessness did not augment the effects of other risk factors but was itself by far the most powerful risk factor for separation. A homeless mother with no other risk factor had about the same risk of becoming separated from a child as a housed mother who experienced domestic violence and was drug dependent. Barrow and Lawinski (2009), using qualitative interviews with homeless mothers with psychiatric or substance use disorders who had been separated from one or more children, confirmed these three factors and added partner abuse of children; substance use by others in household, building, or neighborhood; and children’s needs. They also documented mothers’ problem-solving efforts in an attempt to find the best choices for their children, often among a set of undesirable alternatives.
Barrow and Laborde (2008) focused on homeless mothers who were separated from their children and recruited in single shelters for women with psychiatric problems or substance abuse problems. All 20 of the women they interviewed were actively involved in parenting their children, and most hoped to be reunited, although none were currently living with them. Barrow and Laborde describe women who were caught between demands and conflicting expectations of multiple systems, including shelters, child welfare, foster care, and family courts. Although the researchers did not interview children, they documented multiple cases in which children were shifted repeatedly among relatives or foster care placements and where both shelters and foster care agencies canceled or changed plans for visitation.
Model child welfare guidelines state that homelessness is not by itself a reason to remove children from homes (Williams, 1991), but empirically, it played a large role in these studies both as a cause of initial separations and as a barrier to reunification. A handbook published by the New York State Society for the Prevention of Cruelty to Children (1990) (during the time of the study by Cowal et al.) listed poverty and homelessness as risk factors that “may predispose an individual to abuse or neglect a child” (p. 21). Research suggests that observers rate the same parental behaviors as more abusive in low-income rather than middle-income families (McLoyd, 1990), and it is possible that homeless families were judged more harshly than comparable families that were not homeless. Freidman (2000) and Park and colleagues (2004) note that homeless parents must do their parenting “in a fishbowl” under the watchful eye of shelter and social service staff who may not forgive lapses that go unobserved for housed families. Further, homeless shelters are difficult places to parent due to financial stress, crowding, lack of privacy, and the fact that staff often usurp parental functions, such as providing meals or setting curfews (e.g., Boxhill & Beaty, 1990; Hausman & Hammen, 1993; Lindsey, 1998).
Studies have not examined the effects of separations of children from homeless families on the children themselves. A hint of the effects of separation more generally may be found in a study that compared two groups of 4th to 6th grade children from highly stressed families: those deemed stress-resilient on the basis of “wholesome adjustment in the face of profound life stress” and those deemed stress-affected by both teacher and parent report. An important predictor of resilience was lack of separation of the child and the primary caregiver during the first two years of the child’s life (Cowen, Wyman, Work, & Parker, 1990). A number of studies have documented that childhood separation is a predictor of future homelessness in adults (cf. Rog & Buckner, 2007). However, it is not clear to what extent separation serves as a marker of other factors that may lead families to become homeless or is itself causal. Childhood separations ceased to predict housing instability in adulthood in New York after access to subsidized housing was controlled (Shinn et al., 1998), suggesting that separations may reduce access or indicate the lack of access to familial financial and housing resources.