This paper develops and estimates an empirical framework that evaluates the impact of charter school choice on education quality in the aggregate. We estimate the model using student-level data from North Carolina. We find that North Carolina's lifting of its statewide charter school cap raised the average public school's value-added by around 0.01 standard deviations (on the student test score distribution). We calculate the total human capital returns of the expansion at above $100,000 per charter school enrollee. We further show that competition drives the aggregate gains; test score impacts on students induced into charter schools by the policy are negative.