At the end of 2001, the Indian Supreme Court issued a directive ordering states to institute school lunches known locally as midday meals in government primary schools. This paper provides a large-scale assessment of the enrollment effects of India's midday meal scheme, which offers warm lunches, free of cost, to 120 million primary school children across India and is the largest school feeding program in the world. To isolate the causal effect of the policy, we make use of staggered implementation across Indian states in government but not private schools. Using a panel data set of almost 500,000 schools observed annually from 2002 to 2004, we find that midday meals result in substantial increases in primary school enrollment, driven by early primary school responses to the program. Our results are robust to a wide range of specification tests.
primary school enrollment school lunches natural experiment ITT