In Michigan, 140,000 cases of suspected child abuse or neglect are investigated each year — with many of these youths ending up in foster care, where they can be especially vulnerable to long-term difficulties.

In an effort to reduce child maltreatment recidivism and foster care placement, the University of Michigan Youth Policy Lab will evaluate a new Michigan Department of Health and Human Services in-home, parent-focused program called SafeCare.

The Youth Policy Lab, which helps community and government agencies make better decisions by measuring what really works, received a $765,000 grant from Arnold Ventures to conduct a randomized controlled trial until 2028.

Co-principal investigators for the evaluation are lab founder and faculty co-director Brian Jacob, Walter H. Annenberg Professor of Education Policy and professor of public policy in the Gerald R. Ford School of Public Policy, professor of economics in LSA, and professor of education in the Marsal Family School of Education; and Joseph Ryan, professor of social work in the School of Social Work.

SafeCare targets risk factors for child neglect and physical abuse. It teaches caregivers how to interact positively with their children, recognize hazards in the home to improve the home environment, and recognize and respond to symptoms of illness and injury.

While several studies have evaluated SafeCare against comparison groups, including randomized controlled trials and quasi-experimental studies, prior research does not provide definitive causal evidence.

U-M’s evaluation will be relevant to bolster the justification for MDHHS as it allocates more than $6 million in funding for SafeCare over the course of its first three-year pilot, and also to be able to draw on federal funding under the 2018 Family First Prevention Services Act.

If the study finds evidence for SafeCare’s effectiveness and is well-rated by the Prevention Services Clearinghouse, this will pave the way for states nationwide to use Family First funds for SafeCare, influencing program choices in child welfare throughout the country.

U-M’s randomized controlled trial will evaluate the effectiveness of SafeCare relative to a “services as usual” counterfactual in which caseworkers will provide services from the existing suite of programs available to families.

Randomization will be achievable because in the 17 counties where the program is being piloted, only 996 families can be served per year, against a population of more than 10,000 eligible families.

Drawing on child welfare administrative data from MDHHS for background characteristics such as demographics and past maltreatment allegations, the study will seek to measure long-term recurrence of child maltreatment and out-of-home placement.