School’s Out: The Impact of MGNREGA on School Dropout Rates in Rural India

Abstract

This paper examines whether participation in India’s Mahatma Gandhi National Rural Employment Guarantee Act (MGNREGA) affected school dropout rates among children in rural households, and whether these effects differed by gender. Using nationally representative data from the National Sample Survey’s 66th (2009-10) and 68th (2011-12) Employment and Unemployment rounds, covering over 70,000 rural Indian households, I estimate a two-stage least squares model that instruments for household-level MGNREGA enrollment using a yearly leave-one-out district-level participation rate to address self-selection and reverse causality. I find no statistically significant evidence that MGNREGA enrollment affected household school dropout rates overall or disaggregated by gender, with point estimates suggesting a 9 percentage point increase in dropout that is indistinguishable from zero. A complementary random forest analysis reveals that household head age and education are the dominant predictors of school dropout, while MGNREGA enrollment ranks sixth out of ten covariates, suggesting the program was a secondary rather than primary determinant of household schooling decisions.