A's from Zzzz's? The Causal Effect of School Start Time on the Academic Achievement of Adolescents

US Air Force Academy freshman cohorts across two schedule policy changes

US Air Force Academy freshmen (college students, ages 18+)

Combines two policy changes in the Academy's daily schedule with the Academy's randomized placement of freshman students to courses and instructors. The randomization is genuine and the design is strong — it is arguably the cleanest identification anywhere in the start-time literature — but the population is college freshmen at a military service academy, outside this archive's ages 4-18 scope, and the setting (mandatory military routine, controlled dormitory lights-out) does not generalize to secondary schooling.

Starting the school day 50 minutes later has a significant positive effect on student achievement, roughly equivalent to raising teacher quality by one standard deviation. Frequently cited as the strongest causal evidence for later start times; the citation usually omits that the subjects are adults in a service academy, not high school students.

Low by design — random assignment of students to courses and instructors, combined with institution-wide schedule shocks. Genotype cannot differ across the treated and untreated schedule regimes.