The Long-Term Effects of Universal Preschool in Boston

~4,215 randomised applicants (working-paper figure; verify against published QJE version)

Boston Public Schools preschool lottery applicants, cohorts 1997-2003.

Admissions-lottery instrument — grade A identification, modern, at scale, in a public program.

The most important modern pre-K result and the cleanest demonstration that ATTAINMENT can move while TEST SCORES do not: a precisely-estimated zero on achievement across grades 3-10, alongside real gains in high-school graduation and college enrolment. Crucially, its counterfactual was mostly OTHER preschool, which should have shrunk effects — so the substitution story cannot explain why other scaled programs fail. Read the attainment result with care: on-time college enrolment is solid, but ever-enrolling is marginal and GRADUATION is null.

NOT 'no preschool': roughly 33% of new BPS preschool students were drawn from Head Start and ~29% from private preschool; only ~38% from no preschool.

Lottery randomisation defeats family selection — its central strength.

Weiland et al. (2020, n=3,182) independently confirms the grade-3 test-score null for the same program. Effects are larger for BOYS. Note that college graduation — arguably the outcome that matters most — is not significant.