Who benefits from compulsory education? Evidence from the average and heterogeneous effects on migrant welfare in China

Abstract

This paper examines the heterogeneous treatment effects of the 1986 compulsory education law on migrant welfare in China. We develop a novel instrument by exploiting variation in exposure intensity across cohorts and local enforcement strength across counties of birth. Based on the marginal treatment effect (MTE) framework, we find that migrants with higher gains on wage are more likely to receive compulsory education than migrants with lower gains. In contrast, other welfare outcomes have their own heterogeneous patterns. Counterfactual analysis shows that as an equalizer, compulsory education almost removes the intergroup difference in welfare outcomes. The findings above are robust to various checks. Policy simulation reveals that targeted education policies could be more cost-benefit efficient than universal education programs.