This paper uses a quasi-experimental framework to analyse the impact of India's largest public works programme, the National Rural Employment Guarantee Scheme (NREGS), on schooling enrolment, grade progression, reading comprehension test scores, writing test scores, math test scores and Peabody Picture Vocabulary Test (PPVT) scores. The availability of pre- and two rounds of post-intervention initiation data from the three rounds of the Young Lives Panel Study allow us to measure both the short- and medium-run intent-to-treat effects of the programme.

We find that the programme has no effect on enrolment but has strong positive effects on grade progression, reading comprehension test scores, math test scores and PPVT scores. The average effect size computed over several outcomes is similar to the effects of conditional cash transfer programmes implemented in Latin America. These short-run impact estimates all increased in the medium run, that is, there is no decaying of impact but instead medium-run augmentation of the estimated short-run effects. The findings reported here are robust to attrition bias, endogenous programme placement, type I errors and type II errors.