In several of our recent program evaluations, we have had the opportunity to talk with child welfare workers in Illinois, Oregon, and Wisconsin and gain their insight on the nature of their work and what makes their jobs easier or harder. One of the things that is most difficult for child welfare workers, and something that is completely out of their control, is change in child welfare leadership. Changes in leadership often mean changes in programming, as the new leadership seeks to address the most pressing concerns facing the agency in the ways that they deem best.
In the face of all that, here's what amazes me: the steady, inspiring dedication of those frontline workers who, day in and day out, dedicate themselves to the children and families in their community. When children are safe and families able to thrive on their own, the credit goes to the families. When children are unsafe and families need help, the blame goes to the workers. And yet our child welfare workers continue their tireless work on behalf of children and families.
So let me just take a moment in this opening letter to say what can never be said enough: THANK YOU to everyone working on the frontlines of child welfare. Through leadership and policy changes, your steadfast dedication to children and families is an inspiration to us all. Thank you for everything you do!
We hope you enjoy our August newsletter!
This year's annual monitoring report of the
B.H. consent decree includes an additional chapter focused on racial disproportionality in Illinois. To understand more about this important topic, we talk to CFRC research specialist Dr. Yu-Ling Chiu, who led the analysis.
Q: What is racial disproportionality?
Racial disproportionality is essentially any over- or under-representation of a racial group in the child welfare system compared to that group's proportion of the general population.
Q: Why are states concerned about it?
We want child welfare systems across the country to respond to reports of child maltreatment and not let possible racial biases affect how they handle reports and investigations. For example, we want children to be taken into care because they are unsafe at home; we don't want some children staying in the home because of biases that some racial groups have closer knit families and thus are less likely to re-abuse. Here in Illinois, DCFS wants to track racial equity and disparity at critical decision points to help inform their planning and decision making.
Q: How does CFRC measure disproportionality?
We use a standard measure called a racial disproportionality index (RDI). That involves dividing the proportion of children in a given racial group at a decision point in the child welfare system by their proportion at a meaningful comparison point. We have two types of RDIs that we use. Absolute RDIs use proportions in the general population as the denominator. Relative RDIs, on the other hand, use the most recent previous decision point as the denominator. For example, the relative RDI for entries into substitute care uses a denominator of the proportion of a given racial group among indicated reports.
Q: What does the RDI tell us?
If the RDI is near 1, it means the two proportions are the same. So, for example, it means that that a given group makes up the same proportion of indicated reports as they do among entries into substitute care. An RDI above 1 means that group is overrepresented; below 1 means they are underrepresented. The tricky part is that over- or underrepresentation isn't necessarily a bad thing. That's because we don't know the "true" rate of maltreatment in the population. If one racial group does experience a lower rate of maltreatment compared to another racial group, then we should want that group to be "underrepresented" in the child welfare system. So instead of telling us that there is or isn't a problem, RDIs act as a caution flag. They point out places where racial bias may exist.
Q: What were the findings reported in this year's B.H. report?
In general, we found disproportionality in Illinois that is similar to states across the nation, specifically that African American children are overrepresented. We also found that Hispanic children are underrepresented, and White children proportionally represented, compared to their makeups in the general population. Additionally, relative RDIs showed entries into protective custody and substitute care were more likely for African American children and less likely for Hispanic children, when compared against the rates at which maltreatment allegations are investigated and maltreatment indicated.
Q: What can we do with these findings to make improvements?
We see disproportionality right from the first decision point, with African American children overrepresented in investigations and Hispanic children underrepresented. Child protection agencies can only act on what is reported to them, so right away, they are beginning with a disproportionate population. To me, this suggests the importance of addressing disproportionality both at the CPS level but also--and more importantly--at the societal level. We cannot solve societal problems like poverty, racism, and so on, through the child welfare system.
Our use of relative RDI does help shed light on two decision points of concern. It appears that African American children are more likely to be taken into care compared to the rates at which their maltreatment is investigated and indicated. This suggests a possible source of racial bias that should be examined with more research. There are also some regional differences, including lower disproportionality in Southern region, that should be explored more.
We look forward to performing that additional research and making disproportionality a central focus of our
B.H. work going forward. In the meantime, we hope everyone will read the disproportionality chapter in this year's
B.H. report, and all the other chapters too!
There is so much great research published each month, we can hardly keep up! Here are the articles that we have read as part of our monthly Journal Club.
Child welfare worker turnover is a constant and vexing problem in the child welfare system, and Frank Vandervort and colleagues at the University of Michigan explored one possible reason for this: the difficulty and stress caused by interacting with the legal system. In their 2008
, they describe the conflicts between legal ethics and social work ethics as increasing stress among workers. It's a thought-provoking perspective on the long-standing problem of worker turnover.
As part of our expanding work in the field of predictive analytics, one article that caught our attention was a 2017
by Kelly Capatosto from the Kirwan Institute for the Study of Race and Ethnicity. In the paper, Ms. Capatosto describes how predictive analytics models can further implicit and institutional biases. She raises questions that everyone who uses predictive analytics must grapple with, and we will use her thoughts to guide our own work in this area.
And finally, as we expanded our knowledge on disproportionality, we read Michelle Johnson-Motoyama and colleague's 2017
titled "Using Administrative Data to Monitor Racial/Ethnic Disparities and Disproportionality Within Child Welfare Agencies." It's wonderful to compare and contrast the approaches that other researchers take, and this article was full of new ideas for us to consider.
What have you been reading lately? Let us know via the contact information below.
Connect with Us