This article assesses causal effects of formal microcredit on children's educational outcomes by using household panel data (2000 and 2004) in north-west rural China. The unobservables between borrowers and non-borrowers are controlled in static and dynamic regression-discontinuity designs. The...