Scholar proposes framework for the use of empirical evidence in child welfare policies.
Each year, reports of alleged abuse and neglect of nearly 7.5 million children reach local child welfare agencies across the United States. With so many reports, agencies need to determine which require an urgent response. Many child welfare agencies are turning to empirical evidence to help triage these cases.
In a recent paper, Clare Huntington, a law professor at Fordham University, acknowledges that empirical evidence—especially the use of predictive analytics—can improve child welfare policies and practices. But she argues that empirical data must be used with great caution.
Huntington is especially wary of the drawbacks that an emphasis on empiricism can have on local agencies’ policy choices. She proposes a framework to mitigate these concerns, and she argues that empirical evidence will be most useful when decision-makers agree about “how to balance competing values,” such as children’s well-being and parental liberty interests.
Local agencies increasingly use predictive analytics to determine which cases to prioritize. “Predictive analytics” refers to the use of statistical analysis to “develop models that predict future events or behavior.” These models can assist child welfare agencies when deciding whether to respond to allegations of abuse and how to allocate limited resources.
For example, experts in Hillsborough County, Florida developed a model to evaluate child welfare data after a surge in child fatalities. The model reviews data from past cases to function as a “second set of eyes” to assist with and supervise caseworkers’ decision-making. Agencies in San Diego and New York have developed similar tools.
Although these tools can identify potentially high-risk cases, predictive analytics yields many false positives, which inappropriately bring innocent families and safe children under state surveillance.
Judicial review, according to Huntington, may help alleviate concerns about the reliability of empirical evidence, as courts may invalidate flawed analytic tools. These tools, however, may not be subject to the same oversight as federal regulations. Furthermore, Huntington suggests local agencies may not “have the capacity to handle empirical evidence in a careful and sophisticated manner” because they have fewer resources.
Huntington’s primary concern is that using empirical evidence distracts from the values underlying policy choices by focusing instead on measurable outcomes. This emphasis on outcomes, Huntington argues, “discourages a forthright debate about these competing values and the tradeoffs inherent in any legal regulation.”
Huntington emphasizes the need to balance the prevention of abuse and neglect with other concerns, such as “family autonomy, family integrity, and disproportionate intervention in families of color.” She encourages policymakers to consider potential value tradeoffs—such as between providing services to children in need while inappropriately involving others in harmful, unnecessary state intervention—and to be critically aware of how value judgments manifest in these decisions.
The Indian Child Welfare Act (ICWA) demonstrates how competing values inform child welfare policy, Huntington explains. Enacted in 1978, the ICWA sought to “protect the best interests of Indian children and to promote the stability and security of Indian tribes and families.” Congress passed this law in response to the frequent removal of Native American children by nontribal officials, which occurred in large part because states failed to appreciate the social and cultural importance of tribal relationships. The ICWA thus prioritizes two potentially competing values: first, it seeks to promote the best interests of Native American children, and, second, it aims to preserve tribal autonomy and children’s relationships with their tribes.
Although the ICWA is often considered the “gold standard” of child welfare policy, Huntington identifies problems with child welfare agencies’ use of empirical evidence in this context. She writes that empirical evidence will overemphasize measurable values, which are “defined along traditional metrics, such as physical health and high school graduation rates.” Because it is harder to measure the benefits of preserving children’s relationships to their tribe, it is easier for policymakers to ignore these relationships when making decisions about a child’s best interests, she claims.
When policymakers do not agree about the underlying values of policies or how to balance them, Huntington’s framework discourages reliance on empirical evidence and encourages more debate about competing values. Recognizing that empirical evidence is an important tool, she stresses that it should not dominate decision-making and should not be used to distract from the values underlying these decisions.
Huntington advocates for practical tools, including more deliberate gatekeeping about what kind of evidence policymakers consume, and she suggests that local agencies can partner with universities to access the resources required for responsible reliance on empirical evidence.
Furthermore, Huntington argues that policymakers must examine whether the use of empirical evidence will compound discrimination and inequality. Moving forward, she challenges policymakers to develop a critical awareness of their ideological commitments and to pay careful attention to their values when interpreting empirical evidence.