Research That Matters (January 17 - 20, 2008)

Background and Purpose. Numerous studies on racial disparities have had mixed results with respect to identifying the effect of race on case outcomes (Hill, 2006). However, few of these have been designed explicitly to examine potential bias in child protective services (CPS) casework practice. The hypothesis was: characteristics of the child, family, or report interact with the child's race and worker practice resulting in differential casework decisions by race. Methods. This was a longitudinal study of matched pairs of African American and white children substantiated as neglected in four counties. From an original random sample of 103 pairs, 206 cases of children ages 0 through 11 were read. The case was followed from the time of the report through case closing or a minimum of 18 months of service. After a thorough review, 23 cases were excluded because at least one parent was an immigrant or of Latino/a ethnicity or because the child was mixed race. Of the remaining 180 cases, 81 matched pairs were identified. A logistic regression analysis was conducted to identify predictors of ongoing placement (>72 hours). A qualitative analysis was then conducted to identify issues such as parent cooperation, worker attitudes, and details of the case not ordinarily available. All 162 cases were coded line-by-line and analyzed to identify types of cases or problems associated with case decisions. Cases were initially coded blind, without knowledge of race. Analysis was conducted by age group: 1-4, 5-9 and 10-11. Children <1 were not coded because they had similar dispositions regardless of race. After the major themes were identified, race was introduced to determine whether any of the observed patterns were explained by these variables. Results. White children 1-4 were more likely to go to ongoing placement. For children ages 5 -9, the reverse was true. Early police involvement in the case played a striking role in determining eventual ongoing placement with notable variation by race for those 5 and over. As a result of parental arrests, police were more likely to place white children in informal placements and African American children in formal placements. In another finding, the role of drug and alcohol abuse in interaction with race had differential effects on the placement decision for different age groups. Selected additional qualitative findings regarding factors interacting with race included the influence of mother's financial crises, worker attitude and level of parental cooperation. Conclusions and Implications. With the caveat that these findings result from four counties examined at one point in time, the need for worker and supervisor training regarding the potential interaction of case factors and race was identified. In policy, it is important to level access for all races to alternative services, e.g., mental health, prior to CPS involvement. In addition, the need for more active collaboration with law enforcement regarding the handling of arrests and raids where children are likely to be involved is imperative. Reference: Hill, R. B. (2006). Synthesis of research on disproportionality in child welfare: An update. Washington, DC: Westat.