A School-based Intervention to Prevent Expulsions for Violence: A Targeted Minimum Loss-Based Estimation Analysis

Abstract Childhood and adolescence are critical periods for preventing behavioral problems, and intervening during these stages can produce lifelong benefits. If basic needs remain unaddressed, however, behavioral problems can escalate to include violence. Schools may expel students in severe cases, further ostracizing them and often setting them on a course of continued aggression. We evaluated a school-based intervention, Pathways to Potential (P2P), designed to address students’ basic needs. By addressing social determinants of behavioral problems, we hypothesized that P2P also would be associated with fewer violence-related expulsions. This study used targeted minimum loss-based estimation to estimate average treatment effects (ATE) of P2P on rates of expelling students in Michigan K-12 schools from 2013–14 through 2016–17. Outcomes included expulsions for (1) overall violence, (2) weapons, and (3) firearms. Treated schools received P2P from 2013–14 through 2016–17 ( n = 98). Comparison schools were untreated but in districts where some schools received P2P ( n = 346). Mean pre-treatment rates for overall violence-related, weapon-related, and firearm-related expulsions were 0.85 (SD = 2.32), 0.30 (SD = 1.07), and 0.04 (SD = 0.33) per 1,000 student-years, with no differences between treatment and comparison schools. Analyses did not find that P2P was associated with changes in expulsion rates for overall violence (ATE: − 0.01 [ − 0.10, 0.08]), but it was associated with decreases for weapons (ATE: − 0.06 [ − 0.10, − 0.02]), including firearms (ATE: − 0.03 [ − 0.06, − 0.01]). Findings were robust to sensitivity checks. School-based interventions that address fundamental causes of health and well-being may offer a promising strategy for preventing severe aggression at a critical stage in children’s lives.