No cohort left behind?

Much of the debate over the allocation of education resources focuses on the alleged benefits of smallness—of classroom or school—and is based on evidence from small-scale studies. This paper reframes the question in terms of cohort size. Using national data, we find that a 10% increase in kindergarten enrollment yields a 0.5% increase in cohort shrinkage across early grade transitions, which implies that larger cohorts feature higher rates of retention. Consistent with previous work on class and school size in more restricted settings, this cohort-tracking exercise provides robust evidence at the national level that smallness confers benefits.