Small Class Sizes and Health: Causality, Mechanisms, and Lessons for Policy

This project provides a follow-up to the Tennessee STAR class size experiment conducted in the mid-1980s. Project STAR included over 11,601 students who were randomly assigned to the treatment group (class size 13–17 students) or to larger class sizes (22–26 students). Project STAR showed that smaller class sizes in kindergarten through 3rd grade led to decreases in special education needs and grade repetition; improved academic performance and general well-being; and increased high school graduation and college entrance exam completion rates. Reduced class sizes also appeared to lower educational disparities based on income and race, even though the intervention was not specifically targeted to at-risk populations. Now that these original subjects have grown older, investigators will combine the Project STAR data with federal health data to ascertain whether children who experienced reduced class sizes within Project STAR have subsequently experienced higher earnings, and lower disability and mortality rates than their peers who attended regular size classes. Addressing these questions will also help identify which aspects of the educational system might require enhancements.