The highest-return years.

High-quality intensive preschool programs produce large, durable gains in adult outcomes — but lower-intensity programs often do not.

The Perry Preschool Project randomized 123 Black children in Ypsilanti, Michigan to receive 2.5 hours of daily preschool plus weekly home visits, or no program. At age 40, participants had higher employment (76% vs. 62%), higher earnings, and dramatically lower arrest rates (36% vs. 55% convicted of a felony). The Abecedarian Project — a full-day year-round program from birth to age 5 — found similarly large effects on IQ, education, and health at age 35. Both programs were expensive and intensive. Head Start, which is lower in cost and intensity, produced much smaller effects in the national randomized evaluation — primarily cognitive gains that faded by third grade.

Parenting programs that improve caregiver quality produce lasting gains for children — even starting in the prenatal period.

The Nurse-Family Partnership, which pairs low-income first-time mothers with nurse home visitors from pregnancy through age 2, reduced child abuse and neglect by 48% and improved children's academic achievement at age 15. The Jamaica stimulation study found that 24 months of psychosocial stimulation from community health workers for stunted children produced earnings 25% higher than controls at age 22 — one of the largest and longest-run effects ever documented for a low-income country intervention. Both programs work through improving caregiver responsiveness, not direct child instruction.

Structured pre-K curricula with teacher coaching produce larger effects than typical public programs.

The Boston Public Pre-K program — using structured curriculum and intensive teacher coaching — produced language gains of 0.32 SD and executive function gains of 0.45 SD, among the largest effects documented in a US public program. The Tulsa pre-K program showed similar large effects. In contrast, the national Head Start randomized evaluation found modest and fading effects. The critical difference appears to be curriculum structure, teacher quality, and coaching intensity — not the existence of pre-K per se.

Head Start Impact Study

United States (nationally representative) · 2002

Head Start's national randomized evaluation found that most academic gains faded by the end of first grade. The program exists in over 1,600 communities with highly variable quality. The average-quality Head Start program does not replicate the Perry or Abecedarian results — which were unusually intensive and high-quality for their era.