Efficacy of United States' federally-funded interventions in increasing school capacities to improve student mental health and education outcomes in Tennessee

About two in five children and adolescents will meet the criteria for a mental disorder by age 18, and more than half of youth who are accessing mental health services receive them in an educational setting. Yet there is limited evidence on the effectiveness of school-based interventions on children's mental health and education outcomes. We examine the effectiveness of two key United States' federally-funded interventions for expanding school-based capacities to improve children's health and education outcomes-School-Based Health Centers (SBHCs) and Advancing Wellness and Resiliency in Education (AWARE) grants-in a mixed method, longitudinal study of low-income, Tennessee children. We linked health insurance claims data for children enrolled in Tennessee's Medicaid program with administrative education records for students attending Tennessee public schools between 2006 and 2019. We also implemented a census of Tennessee school districts to determine which had SBHCs and AWARE grants and their start years, and we conducted semi-structured interviews with each treated district to assess their infrastructure, programs, staffing, partnerships, health services offered, and more. We estimated effects of SBHCs and AWARE grants on school-level rates of mental health conditions, behavioral health conditions, preventive health care visits, absences, chronic absences, and disciplinary incidents using a staggered adoption, difference-indifferences (DiD) approach. We found a statistically significant reduction in diagnosed mental health conditions among treated schools of 6 percent relative to their baseline prevalence, which our qualitative findings suggest might be related to increased health staffing in schools, earlier detection of mental health needs, and greater use of prevention strategies. We saw larger effects in some school districts with more extensive mental health infrastructure. We did not satisfy model assumptions for estimating causal effects on preventive health care visits, absences, chronic absences, and disciplinary incidents, although associations were in the expected direction.