Crowd-out in school-based health interventions: Evidence from India’s midday meals program

•We study the implementation of school-based nutrition programs in Odisha, India, using a randomized controlled trial.•We find that a newly implemented nutrition program crowded out the implementation of a pre-existing nutrition program, resulting in no net effect on child health.•We find that high intensity monitoring improved implementation of the pre-existing nutrition program, resulting in a reduction in anemia rates among school children.•Both crowd-out and responses to monitoring intensity were predominantly found among schools that were capacity constrained.