Family First Evidence Standards Risk Worsening Racial Inequities in Child Welfare
The Family First Prevention Services Act has the potential to move significant child welfare resources upstream—to fund services that help families provide safe, loving homes for their children. While this funding could help decrease the need for foster care and is worthy of celebrating, the act also has the potential to further racial inequities within the child welfare system.
To draw on the funding available through Family First, states must use programs that meet a set of rigorous evidence standards, which are detailed in the Title IV-E Prevention Services Clearinghouse Handbook of Standards and Procedures. These standards were created to ensure that federal dollars are used for high-quality, effective programming; yet they can rule out programs that could be especially helpful in some communities.
The types of programs that meet the standards tend to have benefitted from large randomized controlled trials or quasi-experimental designs, which are often hugely resource intensive and privilege Western-centric evaluation methods. Excluded are qualitative methods and pre-post designs, which are often more feasible and sometimes preferred for telling the story of a program and centering the voices of those served.
Within large quantitative studies, evidence often shows whether the intervention was effective for marginalized families, including Black, multiracial, Native American, Asian, Pacific Islander, and Latinx families. However, the Family First standard explicitly prohibits results of these sub-analyses from use in the evidence rating. Many studies of family services have a majority white sample; therefore, their experiences determine the final evidence rating. A program that was especially effective for Black parents but not effective for white parents, for example, may not qualify if white parents made up a large proportion of the study participants. Conversely, a program that was effective for white parents but not at all effective for Black parents can gain a high evidence rating.
In addition, community-based programs are often too small and under resourced to be considered for inclusion in Family First prevention plans—or to conduct the type of evaluations needed. Even if more evaluation resources were available, these programs may be too small to achieve the statistical power required by Family First standards. Furthermore, a small but impactful provision of the standards requires evaluations to include multiple administrative units (e.g., clinics, agencies). This provision is called the “n of 1 administrative unit confound,” and it is not possible for small, one-site community-based programs to meet this requirement.
States must submit state-level plans to receive Family First funding. In county-run states especially, the incentive is strong to select large, statewide scalable programs that have already been approved as meeting Family First standards, even if they are not well-suited for individual communities and particularly communities of color.
In one example of the recognition of the past and ongoing harms of child welfare agencies and the importance of expanding the definition of a successful program, Family First allows Native American tribes flexibility in the evidence standards. This approach could be broadened.
Recommendations for addressing the potential for increased racial inequities through Family First include the following.
1. Short Term
Remove the “n of 1 administrative unit confound” standard.
Remove exclusion of subgroup analyses from the standards.
NOTE: These changes could be implemented quickly, with significant impact. If these standards are removed, smaller programs will stand a better chance of being reviewed and rated as well-supported, supported, or promising.
2. Medium Term
Provide resources for organizations and programs known for effectively serving families of color in their communities to pursue evaluations that both satisfy the current evidence standards and tell the stories of their successes and challenges in a way that is relevant and meaningful to the community. Even the playing field and resist the homogenization and perpetuation of the status quo of services.
3. Longer Term
Reconsider the evidence standards in their entirety. In a field with historical and ongoing oppression of marginalized groups, “what works” is a conversation that needs to be more inclusive and open to a broad range of definitions and types of evidence. Examples might be large changes, such as the admissibility of qualitative evidence, or smaller ones like increasing the threshold of statistical significance (e.g., p value higher than .05) to accept smaller pilot studies.