(Withdrawn) Racial/Ethnic and Gender Disparities in Child Protection Decision-Making: What Role Do Stereotypes Play?

Background/purpose: Child protection decisions, such as providing services or placing a child out of home, seem impacted by disparities. This means that marginalized groups, including minoritized racial/ethnic groups or women, seem negatively impacted by child protection decisions. Research has shown that children from disadvantaged racial/ethnic groups are more likely to receive child protection interventions and maltreatment committed by mothers seems associated with an increased likelihood of out-of-home placements. Research on the underlying causes and biases of these disparities is scarce. In this study, we investigate whether stereotypes mediate disparities in child protection decision-making, by applying the Stereotype Content Model (SCM) which states that stereotypes are formed by perceptions of sociability, morality, and competence.

Methods: We conducted an experimental 3x2 vignette study to investigate disparities in decision-making. We altered family race/ethnicity (white, Black, or Latinx) and the gender of the parent who seemed responsible for child maltreatment (mother or father), while all other case facts remained constant. 258 child protection professionals from Colorado participated in the study (125 trainees and 133 current welfare staff). After randomly presenting them with one of the vignettes, they rated how they perceived the sociability, morality, and competence of the parent responsible for child maltreatment. Then, they assessed the risk for future harm and to what extent they believed that a supervision order should be provided. Using a structural equation model, we examined whether the dimensions of the SCM (i.e., perceived sociability, morality, and competence) mediated potential ethnic and gender disparities in decision-making.

Results: We obtained a SEM model (cfi = .95, rmsea = .04) with different model coefficients for trainees and current staff. For both trainees and current staff, there was no substantial evidence that ethnicity and gender in the vignette impacted stereotypes for sociability, morality. However, evidence appeared for disparities in decision-making. Trainees agreed to a higher extent that a supervision order should be provided for the white father (β = 0.20), Black mother (β = 0.22), Latinx mother (β = 0.27), and Latinx father (β = 0.18) compared to white mother condition. Current staff rated the risk for the child’s future harm higher in cases with Black mothers (β = 0.20), Latinx mothers (β = 0.21), and Latinx fathers (β = 0.23). Then, higher risk scores increased the score for providing a supervision order (β = 0.26). Overall, it appeared that perceived morality had a significant total effect on decision-making, as increased perceptions of morality decreased the score to decide for a supervision order (trainees: β = -0.55, current staff: β = -0.37).

Conclusions/Implications: We were unable to find evidence that stereotypes mediate disparities in decision-making. However, we found evidence for disparities in decision-making. Further, we also found that stereotypes, in particular perceived morality of the parent responsible for maltreatment, seems to play an important role in the decision-making process. Future research is needed to investigate which factors impact professional judgement concerning parents’ morality and how the impact of stereotypes on decision-making can be eliminated.