The idea of removing race from child removal decisions is growing, despite some skepticism.
In an all-too-common occurrence in the nation’s largest local child welfare system, a 37-year-old mother of five from Los Angeles County dialed 911 about two years ago, seeking protection from an abusive partner. That call brought not only the police but the Department of Children and Family Services to her home.
Under the watch of social workers, Kenia Charles said, she moved into a shelter for domestic violence survivors, but still, the child welfare agency argued in court it had concerns about her mental health and ability to parent.
“I don’t do any drugs. My kids have never been in question of me harming them. I never hit them. I never did anything to hurt them,” Charles said in a series of interviews over several months, describing the torment of losing her two youngest children to foster care.
From her vantage point, social workers wrenched the family apart, in part because of her race. “They just looked at me as a Black woman, and they didn’t like what they saw,” Charles said.
Since the death of George Floyd nine months ago prompted America to re-examine entrenched racism in all its institutions – from police departments to corporations and colleges – the child welfare system too, has had to reckon with its troubled past and deeply flawed present. Driven by evidence that child welfare decision-makers judge parents of color more harshly and are more likely to remove their children, calls for systemic change have grown more urgent among parent advocates, scholars and even agency leaders.
Against that backdrop, there is growing interest in a program that proposes to weed out racial bias when social workers weighing allegations of abuse and neglect decide whether to remove a child from their parents.
The method, known as “blind removals,” was pioneered a decade ago in Nassau County, New York, where Black children are the largest non-white group and represent a disproportionate share of who gets taken into foster care. The race-blind process works off a simple premise: If the race of parents and children is not known to the decision-makers, the bias of social workers won’t determine their fate.
Drawn from experiments such as those masking the identity of musicians auditioning for orchestras, blind removals involve a committee of child welfare workers that must reach conclusions without knowing the family members’ names, race, ethnicity or any other identifying demographic information.
The results of this novel approach spread three years ago, after a team led by researcher Jessica Pryce studied the color-blind program, with Pryce sharing state data on a Ted Talk viewed by more than 1.3 million people. Between 2011 and 2016, after a “color-blind removal” process was deployed, Pryce said, the percentage of Black children removed from their homes in Nassau County went from 57% to 21%.
Impressed with the success, New York officials last year expanded a version of the color-blind model statewide that removes race and other demographic information from a case file, joined by counties in Michigan and Minnesota. A bill before the California Legislature would adopt the program in select counties “to promote racial equity and advance practices of equity and inclusion in the child welfare system.”
Now, agencies in Los Angeles County; Austin, Texas; Chicago; New Jersey; Baltimore; Washington state and Oklahoma have also expressed some interest, according to proponents. Last month, the influential philanthropic group Casey Family Programs highlighted the practice on its website.
Yet as enthusiasm has grown, so has skepticism. The color-blind model has a range of critics, including a data scientist who questions its research base and some social workers who fear it would hinder their work. Activists call it too small a tweak to a fundamentally flawed system in need of a more radical overhaul to stop penalizing poverty and tearing Black and Native American families apart.
One prominent critic is University of North Carolina researcher Emily Putnam-Hornstein, whose work centers on providing child welfare decision-makers with extensive data on families and predictive analytics to calculate their risks of harming children.
“It drives me crazy when people run wild and try to give things this sheen of fact and kind of scientific credibility when there is none,” she said.
Recent data on Nassau County that is now available show a muted record on the success of color-blind removals at reducing disproportionality and one that is less pronounced and more nuanced than initial findings suggested. Before the program began in 2010, of all children removed from their homes, 54% were Black. In the pandemic year of 2020 – a decade’s worth of color-blind removals later – that number had declined to 42%.
But new data from New York’s Office of Children and Family Services shows the decline was not steady or consistent. The share of Black children that were removed from home in Nassau County has bounced around – from 62% in 2014, down to 35.5% in 2016 and rising again in 2019 to 52.5%.
Officials there say their program works to reduce implicit bias and that the process has coincided with a sharp drop statewide in the number of children of all races removed from their homes, and that too is a goal of color-blind removals. From 2010 to 2019, the number of all children in foster care in Nassau County has plunged 60%, more than twice the decrease statewide.
“While Nassau County has seen fluctuations in the rate of Black children entering care over the past 10 years, overall, the trend has been very positive,” promoting a “more just and equitable” process, a spokesperson for New York State’s Office of Children and Family Services, John Craig, told The Imprint. “OCFS remains unwavering in its commitment to implement the blind removal process statewide and believes it is a contributing factor in the decreased number of Black children entering foster care.”
Other supporters say a fuller picture of the color-blind program could come with more-rigorous study.
Trina Greene Brown of the Los Angeles-based nonprofit group Parenting for Liberation said she is convinced the model shows promise and supports California’s Assembly Bill 656 as a first step toward piloting it in the state.
“This bill is really to give Black parents a chance to actually raise our children without systems utilizing our race, our income, our ZIP codes and all these other things that have become reasons to take our children away,” Greene Brown said.
By virtually all accounts, the problem of racial disproportionality is endemic to the child welfare field and has been poorly addressed in the past. In an opinion piece published last month by The Imprint, even the Trump administration’s top child welfare official expressed the weight of the times.
“The new administration must fully acknowledge and address social and racial injustice in child welfare,” wrote Jerry Milner, who recently stepped down as associate commissioner of the U.S. Children’s Bureau. “The glaring inequities in child welfare are not a matter of opinion – they are a matter of fact.” Nationally, Black children are nearly 1.7 times more likely to be in foster care, and that data is even more skewed for Native American children, who are roughly 2.7 times more likely to be separated from their families.
In a typical child welfare case, investigating social workers and their supervisors weigh information that includes the current allegations of abuse or neglect; interviews with family members, mandated reporters and medical personnel; and details on any prior history with the child welfare system. The professionals considering the case will have access to the family’s race and ethnicity, as well as other information that could serve as a proxy for race, like surnames and where the family lives.
Federal figures show that nearly two-thirds of decisions to remove a child involve allegations of neglect – conditions closely linked to poverty, substance abuse, mental health concerns and domestic violence.
Under the blind removal process developed in New York, participating social workers are trained to recognize implicit bias in their decision-making and to stay focused on immediate safety risks, rather than being influenced by things like whether a child lives in a public housing project.
“We want to build an entirely new system that actually knows how to help families that are poor, not investigate them and make them feel bad about being poor,” said Pryce, the director of the Florida Institute for Child Welfare who evaluated the Nassau County program. “Really changing the system is going to take time, but a blind removal is something we can do right now.”
Kent County in Michigan – which includes the city of Grand Rapids – began piloting the program in August 2019 after several months of training and discussion.
Bob Wheaton, a spokesman for the Michigan Department of Health and Human Services, said the county adopted the program to reduce the overrepresentation of children of color in foster care as well as to develop a culturally competent workforce. In Kent County, a child of color is 3.4 times more likely to be the subject of a child abuse or neglect investigation than a white child, he said.
To date, the state has only preliminary data but a few trends have emerged: There has been a decline in the number of children of color taken from their homes but a “slight increase” in the overall percentage of those children in foster care. In the year before the blind removal program was implemented, 110 Black children in Kent County were removed by the agency; in the first year of the new initiative, that number dropped to 78.
Still, he said the color-blind practice of “eliminating racial, socio-economic descriptors, and other demographics that contribute to judgment” have improved casework, with social workers focusing on the facts of the case, the imminent risk of harm, as well as the strengths of the extended family and services that could be provided. While the pandemic has made measuring progress more complicated, Wheaton said with additional scrutiny happening before children are removed from home, there has been about a 50% reduction in the number of children of all races being taken into foster care.
Ramsey County, Minnesota began training child welfare workers on blind removals last year. Officials there said they are still developing the process and collecting input from community members, with no firm launch date.
But the need for some intervention is apparent: In Minnesota, although white children make up the largest group in the child welfare system, Native American children are 18 times more likely to be placed into foster care, and Black children are three times more likely, according to a 2019 report from the Minnesota Department of Human Services.
Kelis Houston, a cultural consultant working with Ramsey County to reduce disparities for Black children, supports trying out the new approach.
“Oftentimes, we are blinded, unfortunately, by both our explicit and implicit biases within child protection,” Houston said.
California – which has the most disproportionate share of Black children in foster care of any state in the nation, according to the National Center on Juvenile Justice – is considering a bill that would create blind removal pilot projects in several counties.
Going further than other color-blind programs, the bill would require that during meetings to discuss removing a child, case file redactions include not only the race, ethnicity and address of parents under investigation but also their religious beliefs, education level, sexual orientation, number of children, marital status and income. Case reviewers also would not be provided information on prior investigations that did not result in a child abuse or neglect finding.
While the popularity of blind removals has surged in recent months, so has criticism of the much-heralded program. Putnam-Hornstein said the initial enthusiasm about Nassau County's results – widely reported as a 26 percentage point drop in the share of Black children entering foster care between 2011 and 2016 – has led to conclusions that aren't supported by empirical research. When Putnam-Hornstein reviewed the Nassau County findings, she said she found it disturbing that multiracial children were counted in some years and not others, and that the findings had been accepted without the scrutiny of a peer-review process.
“The whole field should be upset that people ran this far with a single statistic,” she said. “People got out a bit over their skis on the conclusions that were drawn, which happens when you have something that's potentially exciting.”
Nassau County declined to comment on Putnam-Hornstein's concerns and the recent revision of its numbers. Pryce said her research relied on Nassau County’s data, and her evaluation of the impact of the blind removals program there has revealed changes in the way social workers discuss cases and a greater acknowledgment of how personal values and biases can influence decisions.
Pryce added that the blind removal process does not guarantee a continuous drop in the number of Black children entering foster care. “There's no one thing that's magically going to fix the issue,” she said.
The idea of color blind removals has also received pushback from some social workers, who say the process overlooks their expertise and experience. Pryce has said some workers object to assumptions that they take different actions depending on a child’s race, or say there’s no time in emergency situations to follow the coloblind removal process. And leaders of SEIU California, a group that lobbies on behalf of child welfare workers up and down the state, say members have significant concerns about the California legislation, though they declined to provide a comment.
Corey Best, a child welfare consultant in Broward County, Florida, said he believes the process is useful in training settings but should not be used in active cases.
“Not seeing families is probably one of the most dehumanizing things that we can do,” Best said, “especially when there are a multitude of adverse experiences.”
Miranda Sheffield, 35, who grew up in Los Angeles County’s foster care system, agrees it's a limited approach. She said intervening at the removal stage is too late because millions of Black families are the subject of calls to child abuse hotlines each year, and that’s where government intervention begins. Federal research from 2017 shows that more than half of all Black children will be the subject of a child abuse or neglect investigation by age 18, nearly double the rate of white children. “All these solutions that we’re coming up with right now are just Band-Aids – they're not really getting to the core of these issues,” Sheffield said. “The blind removals idea is losing track of the root causes of why Black folks are being so overly represented in the first place.”
Often, it can feel like being watched and mistrusted.
Since late last year, Charles has had her two youngest children back home, but she was not yet free of the county’s supervision of her family. In late March, she faced a virtual hearing that would close her case, ending sometimes-daily scrutiny of her home and court appearances.
In interviews before the hearing, she described fears that her case could be extended another six months or a year because she said social workers thought she needed more therapy. During the pandemic, Charles was asked to walk around her apartment holding her phone up to show them her living conditions, including the inside of her refrigerator and whether her bed is made.
She also felt pressure to present herself well, by dressing up and putting on makeup.
“If you're a white person you can get up and put on some sweats and flip-flops and sandals,” Charles said. “But me being Black, I feel like when I dress like that, oh my God, it's a problem.”
Due to confidentiality laws established to protect families receiving child welfare services, the Department of Children and Family Services declined to provide a comment on Charles' case.
But Louise Godbold knows her story well. Godbold, executive director of the Los Angeles-based nonprofit Echo, met Charles last year when the mother took a court-ordered parenting class provided by her organization. Godbold said she completed all classes and therapy with “grace and with dignity.”
But she watched her client pay a steep price when a violent man came to her door and she chose to call 911 to protect herself and her children.
“She made the call and because of that, she's been punished over and over again, and not been given the support she needs,” Godbold said.
Charles said the county helped her with rent since her younger children returned home and a lack of child care during the pandemic forced her to leave her retail job. But still, she worried that social workers paid too much attention to the state of her house and her daughter’s progress with potty training, and not enough on what matters most to her family’s success. As she tried to earn the approval of the Department of Children and Family Services, she remained terrified that her kids would be taken back into foster care.
“I’m left feeling like I’m on a desert island,” Charles said. “I just want a chance to breathe and to be able to take care of my kids just like everybody else.”
That opportunity arrived two weeks ago, when she said the county terminated her case.
After two years of government supervision of her family, Charles said she is “still looking over my shoulder,” checking her phone for the texts and calls that used to arrive daily from social workers.
Though she still struggles to understand why her children were taken away, Charles is thankful the family can move forward, together.
“I'm stronger than I think I am,” she said. “I am a good mom. And even if it's just me that thinks that, I'm proud of myself.”
Freelance writer Colleen Connolly and senior reporter Michael Fitzgerald contributed to this report.