Head Start was a War on Poverty improvisation. In 1965, a panel of psychologists, pediatricians, and educators convened by Sargent Shriver designed, in about three months, a summer program meant to give poor children a "head start" before kindergarten. The program enrolled half a million children in its first summer. Edward Zigler, the Yale psychologist who became its intellectual godfather, later said the design was rushed, the ambitions were larger than the evidence base, and the program survived less because it worked perfectly than because it had a constituency and a clear moral story. Sixty years later, Head Start still serves about a million children a year, still operates under federal performance standards, and still occupies a contested position in the evidence base.
The honest accounting goes something like this. Head Start, as a category, has produced large benefits for the children and families it has served — measurable improvements in health, nutrition, parental engagement, and access to services. Long-term studies, including Eliana Garces, Duncan Thomas, and Janet Currie's sibling-comparison analyses, find effects on educational attainment, employment, and reduced criminal justice involvement that persist into adulthood. These effects are real and have been replicated. They are also smaller than the effects of the demonstration programs — Perry, Abecedarian — that informed the Head Start design. The reason is not mysterious: Head Start at scale is, on average, a lower-quality program than the demonstrations. Workforce wages are lower. Teacher qualifications are mixed. Curriculum quality varies by site. Child-staff ratios are higher than in the demonstrations. Head Start, in other words, demonstrates what early childhood education at the actual scale of American public investment produces, rather than what early childhood education optimally configured produces.
The 2010 Head Start Impact Study, a federally funded randomized controlled trial, produced findings that were both important and widely misread. Children randomly assigned to Head Start showed small to moderate positive effects on cognitive and social-emotional outcomes at the end of the program year. By third grade, most of the cognitive effects had faded relative to the control group, though some effects on parenting and on the highest-risk subgroups persisted. The headline coverage emphasized "fade out." The more careful interpretation is that the comparison group was not "no early childhood education" — it was "other early childhood arrangements," some of which were comparably good. The contrast effect was therefore attenuated. Head Start was being compared to its own diluted alternatives, not to nothing.
This is a Law 5 problem — revision — at the scale of national policy. The original 1965 design was a compromise, intended to be improved. Some improvements have happened: performance standards have been tightened, two-generation models have been added, Early Head Start was created in 1995, teacher qualification requirements have been raised. But the central structural problems — under-paid workforce, fragmented funding, variable quality, partial-day and partial-year service models that do not match working parents' schedules, integration failures with public schools — have not been substantially solved. Head Start has been revised at the margin and left structurally intact. The structural intactness is the problem.
The Law 1 dimension — Unity — is the program's founding insight: that a society should not allow poverty to determine the developmental trajectory of children, and that public action is required to prevent it. The Law 2 dimension — Think — is the discipline required to honestly evaluate what sixty years of data show, which is that the principle is right and the execution has been chronically under-resourced. The Law 4 dimension — Plan — is the original planning failure: launching at scale with a design that the evidence base of 1965 could not actually support, and then never returning to redesign with the evidence base that accumulated subsequently.
What follows is a careful tour of Head Start's history, evidence, design choices, and limitations. The article is neither hagiography nor takedown. It is an attempt to understand a sixty-year-old social program that, at this point, is one of the longest-running public investments in early childhood in the world, and to extract the lessons — both about what works and about what the political economy of American social policy permits.