LAUSR

The health impact of attending early childhood development programs in developing countries remains largely unknown. In this study, we focus on the health consequences of attending preschool programs in India. Using a unique longitudinal dataset, we allow for heterogeneity in the impact of preschool across the distribution of health outcomes while controlling for time-invariant unobservables. We detect unique temporal variation in the effect of preschool attendance - growth of preschool attendees is slower than non-attendees in various parts of the distributions of several anthropometric measures when evaluated in the early years between ages 1 and 5. This effect is likely to reverse in the longer term at age 8. The early years' adverse effect can be explained in part by over-attendance in the form of long daily hours, excessive attendance days, and early entry. The findings are insensitive to nutritional incentives like free meals provided in public schools. The growth-retarding effect remains robust for weight-for-age z-scores, implying that the impact of preschool attendance is not only heterogeneous, but differs across dimensions of health status. Our study highlights the need for strengthening the delivery of childhood programs in developing countries in order to prevent adverse health effects in the critical years. Copyright © 2016 John Wiley & Sons, Ltd.