Much of the education literature explores the relationships between the effort, academic self-concept, and achievement of students in isolated, static models. However, students experience these mechanisms jointly and dynamically within a broader ecological system of coevolving peers. In the interest of using systems thinking to reduce systemic disadvantage, I formulate and estimate a dynamic model that reveals how the interplay between these entities can lead to phase transitions, tipping points, and other counterintuitive properties of complex systems. Through simulations of early childhood, growth mindset, double-dose algebra, and self-affirmation interventions, the model shows how reinforcement processes in human development allow for appropriately timed and targeted policy interventions to have persistent impacts on disadvantaged groups. The results contain important implications for the achievement gap, and shed new light on multiple policy debates on persistence and fadeout.