Estimating Private School Effects for School Children in Peru: Evidence from Individual‐level Panel Data