This paper examines the effects of a comprehensive performance pay program for teachers implemented in high-need schools on students' longer-run educational, criminal justice, and economic self-sufficiency outcomes. Using linked administrative data from a Southern state, we leverage the quasi-randomness of the timing of program adoption across schools to identify causal effects of the school reform. The program improved educational attainment and reduced both criminal activity and dependence on government assistance in early adulthood. We find little scope for student sorting or changes in the composition of teacher workforce, and that program benefits far exceeded its costs. We propose mechanisms for observed long-run effects and provide evidence consistent with these explanations. Several robustness checks and placebo tests support our findings.