Wulczyn, F., Gibbons, R., Snowden, L., & Lery, B. (2013). Poverty, social disadvantage, and the black/white placement gap. Children and Youth Services Review, 35, 65-74.
Much of the child welfare literature indicates that black children are placed in out-of-home care at a higher rate than white children. Wulczyn and colleagues examine whether county-level measures of poverty and social disadvantage are correlated with county-level variation in the Black/White foster care placement gap. Specifically, the study aims to identify whether certain counties with an above average poverty rate have an above average black/white placement gap.
The sample consisted of children from 822 counties from 16 states and only included children that were being placed in child welfare care for the first time below the age of 18 in the year 2000. The focus of the study was whether the gap varied with respect to race, specific macro structural variations in poverty, and social disadvantage.
The dependent variables in the study include: black child placements per 1000 black children, white child placements per 1000 white children, and the gap between these two rates (i.e., the black/white placement gap). The principal independent variable is poverty at the macro structural level. The first measure is the proportion of children living in families with incomes below the poverty line. The second measure considers poverty through the lens of social context and macro structural measures of social disadvantage (e.g., family structure, education level, unemployment, and housing).
Twenty one million children were sampled (the total number of children living in the identified counties in 2000); 21% were black. There were 45,617 admissions to child welfare care and 20,959 were black children. Overall there were 4.56 placements per 1000 black children and 1.47 placements per 1000 white children; therefore the difference between the two rates yields a black/white placement gap of 3.09 placements per 1000 children. The gap is substantially greater in urban areas; however, placement rates are higher in urban counties for black children and higher in rural areas for white children.
Higher child poverty rates were associated with lower black child placement rates. For example, at the county level, black child placement rates declined as white child poverty rates increased. The effect of white child poverty was more pronounced, in part because the wide variation in white child placement rates is set against the relatively narrow range of white poverty rates.
The results suggest that placement rates, when black and white children are compared, are more similar in counties with higher white poverty rates in large part because black placement rates do not vary with measures of black poverty and social disadvantage.
A multilevel Poisson event count model was used for placement rate disparity. Authors note that although multilevel modeling was utilized in the analysis, individual child level data was not available and thus not included in the study. Another limitation was the inability to differentiate between counties with various indicators of social disadvantage. Moreover, county-specific measures of maltreatment or substantiation were intentionally not included in the analysis, given the theoretical assumption that there is differential treatment between black and white families. The data were limited to a single point in time; therefore the model does not capture changes in the placement gap or social context.