Alan B. Krueger analyzes data from the Project STAR experiment, a large-scale randomized study of class size effects on student achievement in Tennessee. The study involved 11,600 kindergarten students and teachers randomly assigned to small classes (13-17 students), regular-size classes (22-25 students), or regular-size classes with a teacher's aide. Students remained in their initial class type through third grade, though some were re-assigned later. Standardized tests were administered annually, and the results showed that students in small classes performed better than those in regular-size classes, with performance improvements persisting over time. Teacher aides had little effect on student achievement. The study found that class size had a larger effect on test scores for minority students and those on free lunch. The beneficial effect of smaller classes was not due to Hawthorne effects. The study used statistical methods to address potential biases from attrition and re-randomization. The results suggest that smaller classes improve student achievement, particularly for lower-achieving students, and that the effects are consistent across different student backgrounds. The study also highlights the importance of controlling for school-level effects and the limitations of observational data in estimating education production functions. The findings contribute to the understanding of how class size impacts student achievement and provide insights for policy decisions in education.Alan B. Krueger analyzes data from the Project STAR experiment, a large-scale randomized study of class size effects on student achievement in Tennessee. The study involved 11,600 kindergarten students and teachers randomly assigned to small classes (13-17 students), regular-size classes (22-25 students), or regular-size classes with a teacher's aide. Students remained in their initial class type through third grade, though some were re-assigned later. Standardized tests were administered annually, and the results showed that students in small classes performed better than those in regular-size classes, with performance improvements persisting over time. Teacher aides had little effect on student achievement. The study found that class size had a larger effect on test scores for minority students and those on free lunch. The beneficial effect of smaller classes was not due to Hawthorne effects. The study used statistical methods to address potential biases from attrition and re-randomization. The results suggest that smaller classes improve student achievement, particularly for lower-achieving students, and that the effects are consistent across different student backgrounds. The study also highlights the importance of controlling for school-level effects and the limitations of observational data in estimating education production functions. The findings contribute to the understanding of how class size impacts student achievement and provide insights for policy decisions in education.