Score contribution per author:
α: calibrated so average coauthorship-adjusted count equals average raw count
We estimate the nonlinear impact of class size on student achievement by exploiting regulations that cap class size at 20 students per class in kindergarten. Based on student-level information from a previously unexploited and unique large-scale census survey of kindergarten students, this study provides clear evidence of the nonlinearity of class size effects on development measures. While the effects are largest on cognitive development, class size reductions also improve noncognitive skills for children living in disadvantaged areas. These findings suggest that sizeable class size reductions targeted at disadvantaged areas would achieve better results than a marginal reduction across the board.