Using a Field Experiment to Understand Skill Formation in the Adolescent Years

We combine a field experiment with structural estimation to study skill formation in the adolescent years. By randomizing nearly 1,000 adolescents from low-income communities to distinct interventions that lasted 10-11 weeks, we explore novel aspects of the education production function. First, we document evidence of strong self-productivity in cognitive skill formation and synergy in executive functioning skill formation, whereby future cognitive and executive functioning skills depend heavily on past cognitive skills. We find evidence that low-cost investments during adolescence can improve executive functioning skills, and that the technology of skill formation itself is impacted by our treatments. We estimate production functions for each treatment group to uncover the underlying mechanisms: the treatments increased the self-productivity of executive functioning skills and significantly reduced synergy in skill formation. This means that students with low cognitive skills could better improve executive functioning skills without being as disadvantaged by their low prior cognitive skills. We predict that the program increases high school (college) graduation by at least 2 (1) percentage points. Our results suggest that the policymaker not paying attention to the adolescent years will miss out on a key sensitive period