King, B., Fallon, B., Boyd, R., Black, T., Antwi-Boasiako, K., & O’Connor, C. (2017). Factors associated with racial differences in child welfare investigative decision-making in Ontario, Canada. Child Abuse & Neglect, 73, 89-105.
This is a report of decisions made by workers about Black and White families investigated for child protection concerns in Ontario, Canada. Data were from the Ontario Incidence Study of Reported Child Abuse and Neglect (OIS-2013), which had been designed to provide as representative a sample as possible of Ontario agencies, and of cases within agencies.
It was found that Black children were more likely to be investigated than White children, but this was not due to race alone; Black and White children differed significantly with respect to child characteristics, caregiver risk factors and socioeconomic risk factors. However, even adjusting for these factors, Blacks had 33% greater odds of being transferred to ongoing services. For Black families, the strongest predictors of the decision to provide child welfare services were the assessed quality of the parent-child relationship and severe economic hardship.
In their discussion of the implications of this overrepresentation, the authors point out that poverty is an important problem, and stress possible corrective actions with this in mind: “additional financial, social and mental health resources for caregivers are critical,” and child welfare workers need to “more effectively integrate anti-poverty work into their practice.”
The unweighted sample comprised all cases, 14 years or younger, opened for investigation at the selected agencies in the quarter Oct.1/13 to Dec.31/13 (N = 5265). To provide a weighted annual estimate this number was multiplied by four; as the nature and number of child and family problems could vary with season and school term, this step might affect data validity. Regional weights were also applied to reflect the relative size of each site, giving a full weighted sample of 83,400 investigations. Population racial estimates were based on census data (2011 National Household Survey).
The elaborate data analysis was based on child and maltreatment characteristics, caregiver and socioeconomic risk factors, and allegation/referral source. Agency variables such as size, location, funding, or worker qualifications, were not included, though the concentration of Black families in urban/suburban regions was mentioned.
Statistical procedures included Black and White disparity calculations at each decision point: investigation, substantiation, transfer, and placement into out--of–home care. Both bivariate (crosstabs & Χ2) and multivariate (logistic regression) procedures were used. A potential limitation mentioned by the authors was the use of the full weights in these analyses, which may “overstate the significance of these estimates.”