Examining the potential for racial disparity in out-of-home placement decisions: A quantitative matched-pair study

Abstract

Despite some reductions in recent years, African American children in the United States continue to be disproportionately placed into out-of-home care. This paper describes a secondary analysis of data collected by Minnesota's African American Disparities Committee to identify whether factors in the casework process were contributing to Minnesota's extreme rates of disproportionality. The study, first described in a 2005 Minnesota Department of Human Services report, identified over 200 non-Hispanic, non-immigrant, white and African American child neglect cases substantiated in 2001 in four counties. Cases were then followed until case closure or for at least 18months following the time of report to detail the casework process. This analysis used recursive partitioning to more accurately identify primary factors and potential interactions related to worker decision-making. African American and white children were matched by type of neglect, gender, age group, and county using administrative data. After rigorous data cleaning, 81 matched pairs were identified from a final sample of 180 cases. The progress of the children through the system was evaluated for potential race-based differences in the decision to refer a child for ongoing placement. Number of children, age, history of maltreatment, and report source were associated with placement decisions. Logistical regression follow-ups did not identify race-based interaction effects on placement decisions. However, African American cases statistically differed from white counterparts on presence of biological father and number of children in the household, identity of the perpetrator, parental and household risk factors, and whether an informal placement was arranged by police, some of which were associated with placement. Results are informed by a more detailed, qualitative follow-up analysis in a separate article.