Use Data / Analytics to Prevent Child Welfare Involvement

About this recommendation

When resources, including staff time and funds, are limited, data can help to target communities, criteria, and specific services that can have the greatest positive impact for particular populations.

The use of data to predict child welfare involvement can be very challenging and ultimately cause more harm to families, if not used carefully. Any data analysis needs to recognize the racial and socioeconomic biases baked into historical data and decision making. Predictive data or artificial intelligence should never, ever, ever be used for automatic decision making.

In the examples below, we share examples of using data to identify communities where the provision of thoughtful, culturally responsive services can help to safely stem or solve problems before they ever rise to the level of child welfare involvement.

Who's doing this

• Child Opportunity Index - The Child Opportunity Index is a composite index of children's neighborhood opportunity that contains data for every neighborhood (census tract) in the United States from every year for 2012 through 2021. It comprises 44 indicators in three domains (education, health and environment, and social and economic) and 14 subdomains.

  One jurisdiction combines COI data with its child neglect reporting , to prioritize engagement. It also uses this information to drive other supports, such as prioritizing grant allocations.

  Look up your own service area on their website.

• Hello Baby - Allegheny County, Pennsylvania created Hello Baby, a voluntary program for any family in the county with a newborn or young child (see their methodology report).

• Predict Align Prevent - Through geospatial risk analysis, strategic alignment of community initiatives, and implementation of accountable prevention programs, we discover practical solutions to the fundamental problems of child maltreatment, preventing the suffering and death of little children due to abuse and neglect.

• By cross-referencing aggregate racial and demographic data with screening, testing, and CPS referral numbers, Washington State DCYF can direct anti-bias training to targeted referral sources and support policy that standardizes screening, testing, and referral practices.

  Washington State DCYF maintains a public data dashboard to support the agency’s efforts to prevent child maltreatment.

• Washington’s Prenatal to Three Coalition uses Targeted Universalism to develop key focus areas and identify sub-populations of focus.

Do you know of other data-driven prevention programs? Please let us know.