Do child-welfare agencies discriminate against people with disabilities? That’s the implication of an investigation that the Justice Department launched last month into the Department of Human Services in Allegheny County, Pennsylvania. Relying on a few confidential sources, the Associated Press broke the news of the investigation and noted that it focused on “how algorithmic tools could potentially exacerbate disparities, including for people with disabilities.”
Someone unfamiliar with how child-welfare agencies make their decisions might wonder what this could possibly mean. Are officials investigating or removing children from homes because a parent is in a wheelchair or is hearing-impaired? Of course not.
Rather, the algorithms that Allegheny County has used for the past few years to determine which kids are most at risk (once a report is made to the local child-abuse hotline) include factors such as whether a caregiver has a serious mental-health disorder or a history of substance abuse. So the question that the Justice Department investigation must answer is this: If a caregiver has been reported for child abuse or neglect, why wouldn’t you consider a diagnosis of, say, bipolar disorder or a history of overdoses in determining whether that child is at risk of imminent harm?
Perhaps not coincidentally, the American Civil Liberties Union is passing around its own paper documenting problems with the Allegheny Family Screening Tool. The paper, titled “The Devil is in the Details: Interrogating Values Embedded in the AFST” and to be published later this spring, explains that the tool “recorded diagnoses of various behavioral and mental health disorders that have been considered disabilities under the Americans with Disabilities Act (ADA), including schizophrenia, major depressive disorder, and broad categories of other mood disorders.” The paper notes: “Several versions of the tool have also included features related to public benefits—such as eligibility for Supplemental Security Income (SSI) benefits—that may be related to or potentially proxies for an individual’s disability status.”
Does the Americans with Disabilities Act shield parents with histories of mental-health problems or substance abuse from any additional scrutiny when incidents of abuse or neglect get reported? No. To begin with, the ADA protects a person with a history of substance abuse only when he is in recovery and not currently using illegal substances. According to the ADA National Network, “in recovery” means a person is “no longer engaged in the illegal use of drugs” or “is participating in a supervised rehabilitation program and is no longer using drugs illegally.” It also clarifies that “current” illegal use of drugs could mean not only that the person has failed a drug test but also that “the use of the drugs occurred recently enough” to justify a reasonable belief that “a person’s drug use is a real and ongoing problem.”
For child-welfare investigators, this means that if they see evidence of drug paraphernalia or hear reports of neglect, a history of substance abuse can inform their decisions about what services to offer the family or whether to remove a child from the home. Information about substance abuse and serious mental-health disorders is vital for any job that involves life-or-death decisions. Pilots must pass health exams every six months, including tests for mental-health disorders. Doctors are required to disclose psychiatric disorders. Parents may not be responsible for entire planes full of people, but their decisions, particularly regarding young or medically fragile children, can be the difference between life and death.
Indeed, information about past or present substance abuse or mental-health disorders isn’t just used by algorithms. Every child-welfare agency and family court in the country must consider these histories to assess a child’s risk for maltreatment.
Unfortunately, a movement is growing in the world of child welfare to ignore the past when assessing risk. In 2021, the California legislature considered a bill promoting “blind removals,” which would have prevented child-welfare decisionmakers not only from knowing where a family lives or the race or sex of the child involved in an allegation of abuse but also whether the family has any past unsubstantiated allegations. “Unsubstantiated” in this case doesn’t mean that the allegation was false or frivolous, merely that it could not be proven. In fact, unsubstantiated allegations are strongly correlated with future substantiated allegations.
America prides itself on giving everyone the opportunity to start fresh, rather than being punished indefinitely for crimes committed long ago, so it’s easy to see why people might think ignoring a person’s past is the best policy decision. But in the case of child welfare, the past is the best predictor of a parent’s future behavior. Using that information to assess whether a child is at risk of imminent harm is not a violation of a parent’s rights; it is a vital means of protecting the child’s.