ECIN Replication Package for "Universal Free Meals and School Suspensions"

We evaluate whether providing free meals to all students causes changes in out-of-school suspensions for schools adopting those policies in the United States. The study builds on earlier work that has shown null effects of universal free meals on suspensions. Specifically, this study analyzes an updated dataset and incorporates recent methods designed to estimate program effects when program adoption is staggered over time. We adopt a difference-in-differences research design and exploit the time variation in adopting provisions that permit free meals to be provided to all enrolled students (i.e., Provisions 2 and 3 of the National School Lunch Act, and the Community Eligibility Provision of the Healthy, Hunger-Free Kids Act). Adopting universal meals decreases suspensions by approximately ten percent for elementary students and six percent for middle and high school students. These impacts are more pronounced in schools with fewer students eligible for free and reduced-price meals in preadoption periods. We show that differences in our conclusions from those reported previously arise primarily out of methods that better account for staggered program adoption as opposed to the updated dataset. This replication package provides data and code accompanying the article.