Nearly 5,000 Children, a Decade of Data: Head Start Gains Gone by Third Grade
November 3, 2026 · Frisian News
A congressionally mandated study of nearly 5,000 children found that Head Start produced real preschool gains, but the government's own OPRE Report 2012-45 found very few impacts across cognitive, social-emotional, health, and parenting outcomes by the end of third grade.
By third grade, the early cognitive gains had shrunk to a single statistically significant measure for each age group. That is what the US Department of Health and Human Services found when it published OPRE Report 2012-45, the congressionally mandated follow-up to its own Head Start Impact Study.
The study was not a small survey. Nearly 5,000 newly entering three- and four-year-olds across 84 grantee agencies were randomly assigned to either a Head Start group or a control group whose parents could choose other early childhood programs. Data collection ran from fall 2002 through 2008. The What Works Clearinghouse later rated the design as meeting its randomized-trial evidence standards without reservations.
The program produced real gains at the start. During the Head Start years, three-year-old participants scored higher in every area tested. Four-year-olds showed higher school readiness in language and literacy. Congress had ordered the study in the 1998 reauthorization of Head Start specifically to measure these effects at national scale, and the initial numbers were encouraging.
The gains then faded. The Michigan State University Institute for Public Policy and Social Research, summarizing the 2010 first-grade findings, noted that by the end of first grade most of the original benefits had disappeared. The 2012 third-grade follow-up confirmed that direction. Across all four measured domains, cognitive development, social-emotional development, health, and parenting practices, the study found 'very few impacts' for either age cohort. In the cognitive domain, the report states, these early effects 'rapidly dissipated in elementary school,' leaving a single statistically significant measure for the three-year-old cohort and a single for the four-year-old cohort at the end of third grade.
Some subgroups held onto more. Among children who entered at age three, those from high-risk households and those whose parents reported no depressive symptoms showed more sustained cognitive effects. Among four-year-old entrants, children of parents reporting mild or severe depressive symptoms and Black children retained some benefits through third grade. The report describes these as subgroup findings, distinct from the program's overall result.
Data collection closed in 2008. The third-grade follow-up appeared in 2012. HHS both funds and administers Head Start, which makes a finding of near-total fade-out from its own contracted research team unusually candid. The report was authored by Mike Puma, Stephen Bell, Ronna Cook, Camilla Heid, Pam Broene, Frank Jenkins, Andrew Mashburn, and Jason Downer.
The study has genuine limits. OPRE Report 2012-45 measured cognitive scores, behavioral ratings, health status, and parenting practices through third grade only. It did not follow children into adulthood.
What the data do not support is the routine claim that Head Start boosts third-grade academic outcomes. The program's own funder ran the study, nearly 5,000 children were tracked from fall 2002 through 2008, and the verdict was 'very few impacts' by third grade. That verdict deserves a cleaner headline.
Published November 3, 2026 · Frisian News