Service Referral Patterns Among Black and White Families Involved with Child Protective Services

Background and Purpose: Sources of racial disproportionality in the child welfare system continue to be an issue of concern. However, most research on this topic has focused on overrepresentation in the system, or on negative case outcomes.  Evidence is limited as to whether and in what ways race factors into caseworkers’ service referral decisions. This study examines race differences in several aspects of service referral: whether services are referred, how many services, and which types of services.  Specifically, two questions are addressed: (1) Does the probability of service referral, and do the number and type of services, differ depending on the race of the child and caseworker?; (2) Are differences explained by variation in household, caseworker, case, risk, and county characteristics?

Data and Methods: A sample of 1,511 cases from 81 counties from second cohort of the National Survey of Child and Adolescent Well-Being is used to examine differences in the probability of service referral, types of services, and number of services for black and white families. We first use multiple regression models to assess whether the interaction between the race of the caseworker and the race of the child predicts differences in caseworkers’ service referral decisions. Using Blinder-Oaxaca decompositions, we then estimate the extent to which service disparities are explained by racial differences in families’ characteristics (risk factors) versus differences in associations between characteristics and service referral decisions (differential treatment). Control variables include sociodemographic characteristics of the family, case and risk factors, and caseworker demographics. County-level fixed effects are used to address differences in the availability of services across counties.

Results: Regression results suggest that the combination of caseworker and child race are significantly associated with differences in the probability of referral to different types of services, even after controlling for county, case factors, and household and caseworker demographics. Decomposition models suggest that most differences in service referrals between black and white families are explained by differences in risk factors rather than differential treatment by workers.  Yet, we find that a higher proportion of black families were referred to parenting services by both black and white caseworkers, and that this disparity was not explained by differences in risk factors.

Implications and Conclusions: Ensuring unbiased case decisions in child welfare is an important aim, and this study adds to our current knowledge of racial disproportionality by exploring service referral decisions. Though efforts to address the black-white gap in child welfare system involvement have tended to focus on the capacity of caseworkers to make unbiased substantiation and placement decisions, this study highlights a potential, though complex, role of race in service referral decisions.  Overall, this study suggests that caseworkers’ service referral decisions often reflect families’ risk factors; yet, in some instances, racial differences in service referral decisions cannot be explained. More research is needed to further explore how caseworkers decide whether to refer families to services and which types of services are needed.