What if we could just snap our fingers and get rid of racial disparities in the child-welfare system? Boosters of a recent experiment in “race blind” decision-making, launched first in Nassau County, say we can do just that. Sound too good to be true? It is.
In 2010, Nassau County initiated a policy of “blind removals,” which means that a committee of child-welfare professionals determines whether to take a child into foster care based on an assessment that does not include race or other demographic information. The advent of this policy coincided with a five-year-decline in the percentage of black children going into foster care. As a result, public agencies and private philanthropies have been clamoring to adopt blind removals around the country.
New York State has already instructed local agencies to develop a plan for blind removals. California’s state assembly is considering a bill to fund blind-removal programs across the state; the legislation would not only ensure that the committee doesn’t see a child’s race but also that it redacts information about gender and any evidence of prior unsubstantiated reports of maltreatment. Casey Family Programs, a private foundation focused on foster care, has also provided support for the development of these programs and notes on its website that “there is strong evidence that ‘blinding,’ . . . along with continued training and coaching for staff, is a technical strategy that can have a significant impact on racial inequities and disproportionality.”
Looking at the evidence from Nassau County, though, one would be hard-pressed to reach the same conclusion. Emily Putnam-Hornstein, a professor at the University of North Carolina, Chapel Hill and co-director of the Children’s Data Network, has made some important observations about these miraculous findings.
First, while it’s true that the share of black children entering foster care from 2011 to 2016 dropped from 55.5 percent to 29.8 percent, the trend wasn’t exactly linear. The number stood at 29.2 percent in 2012, then jumped to 39.6 percent in 2013, 48.1 percent in 2014, and 51.3 percent in 2015. What explains all this variation if the program was supposedly producing such great results? One possibility is that the numbers of children involved were so small that only a few cases produced a big percentage jump, meaning the evidence isn’t statistically significant in the first place. Also, we don’t know what happened after 2016.
The way race is coded in these statistics is also a problem. In 2011, “multi-race” was an option, and under this system 2.9 percent of children entering foster care were listed as “other/missing” in the race category. But by 2016, the system had changed to use single-race only, and 17.5 percent of children entering foster care were listed as “other/missing” in the race category.
A final reason to be skeptical: while the number of black children removed declined by 77 percent, the number of black children considered for removal also declined by 74 percent. In other words, whatever is happening, it isn’t occurring at the stage where the committee has any input. Likelier than not, caseworkers are making a concerted effort to divert black children into kinship care, or they’re simply deciding that, no matter the risk, these black children shouldn’t be removed.
None of this disproves the idea that blind removals are helping to combat whatever bias exists in the child-welfare system. But before we head farther down this road, we should weigh the problems with blind removals. The most obvious one is that caseworkers—who actually meet the families and are most familiar with their situations—are not involved in making decisions about removal. They may be biased, but they also have important observations (and yes, gut feelings) about the risks to a child.
California’s lawmakers, meanwhile, seem inclined to remove all sorts of relevant information from the decision-making process. Why shouldn’t decision-makers know the gender of a child? Gender affects the dynamics of a family in many ways. What if there are allegations of sexual abuse? Is such information relevant then?
If, as California legislators wish, the move toward “blindness” ensures that the committee making these decisions will not know about past unsubstantiated allegations, that will also result in poorer outcomes. It may seem fair to us that allegations that cannot be “proved” should be dismissed, but in child welfare that’s not the case. Caseworkers are not simply assessing guilt; they are assessing risk. And the risks associated with unsubstantiated allegations are just as high as they are from substantiated ones.
A study published in Child Maltreatment using data from more than 4.3 million children born in California between 1999 and 2006, for instance, found that “children with a prior allegation of maltreatment died from intentional injuries 5.9 times as often as unreported children, and they died from unintentional injuries twice as often as unreported children.” In fact, “a prior allegation to CPS proved to be the strongest independent risk factor for injury mortality before the age of five.”
No magic spell exists for eliminating racial disparities in child welfare. High-risk family structures, particularly the presence of nonrelative males in a home with children, are distributed unevenly across racial and ethnic populations. As long as this remains the case, there will be differences in rates of child abuse and neglect. It’s easy to blame racism for the disparities in child removals, but it’s just not true. And making decisions blindfolded won’t help anyone.
Photo: AHMET YARALI/iStock