Smaller class sizes improve student achievement.

Parents and teachers overwhelmingly favour smaller classes, but until Tennessee's Project STAR, rigorous causal evidence was scarce. STAR randomly assigned students and teachers to small or regular classes in the early grades.

Analysing STAR, Krueger (1999) found that students in small classes scored meaningfully higher on tests, with larger benefits for minority and low-income children, and some effects persisted for years. That is genuine support. But the picture is mixed: the gains, while real, were modest; effects were clearest in the earliest grades; and because reducing class size everywhere is very expensive and requires hiring many teachers (potentially lowering average teacher quality), other reforms can be more cost-effective. So smaller classes help, especially early and for disadvantaged students, but they are not a cheap or universal fix.