If you Give a Kid a Preschool, She’ll Probably Want to Learn: Insights into Head Start

“Study after study shows that the sooner a child begins learning, the better he or she does down the road. But today, fewer than three in ten four-year-olds are enrolled in a high-quality preschool program.” In the 2013 State of the Union address, President Obama clearly laid out his plan to improve early education for American children through federal programming and legislation. Federal involvement in early education is not new: The federal government has been operating Head Start for the last fifty years, a program which now serves 1.13 million children and pregnant women.

The president is right, however, to raise concerns about quality. Recent results from the Head Start Impact Study (HSIS), a randomized trial of programs, show that children in Head Start have smaller, less persistent gains than those who attended “high-quality preschool programs” such as Perry Preschool and Abecedarian, two programs designed to research the potential effects of early education. This discrepancy led Christopher Walters to research variation in Head Start quality as well as its causes and effects on student achievement. His findings are outlined in the National Bureau of Economic Research working paper “Inputs in the Production of Early Childhood Human Capital: Evidence from Head Start.”

Walters uses the Head Start Impact Study (HSIS) data to categorize programs based on factors associated with success in high-quality programs including using the HighScope curriculum (a curriculum of “active participatory learning”), teacher characteristics, class size, instructional time, frequency of home visits, instructor experience, student demographics, and availability of alternative preschool programs. He then tested the impact of each factor on student outcomes—measuring both cognitive effects (changes to student performance on tests of academic intelligence) and non-cognitive effects (changes to behavior and socio-emotional development).

Walters found that full day programs had greater cognitive effects and frequent home visits had greater non-cognitive effects, while High/Scope curriculum, class size, teacher education, and teacher experience had no effect. Additionally, children with less educated mothers had greater cognitive benefits from the program, but no additional benefits were associated with any other background characteristics. Also, programs that drew students from home-based care had greater cognitive gains than those who drew students from other center-based preschools.

Ultimately, Walters notes that all these factors combined explained one-third of the variation in cognitive and non-cognitive effects. There are two important limitations of this data: 1) that high-quality characteristics were not randomly assigned to programs, and 2) that non-compliance with treatment assignment was common. It is also worth noting that no single Head Start center has all the characteristics thought of as being markers of high quality.

Using the HSIS data along with data produced by Raj Chetty on lifetime earnings, Walters also finds that programs in the 84th percentile of quality lead to a $3,400 increase in earnings per child, which is roughly half of the Head Start per student expenditure.

Interestingly, Walters compares some of his findings to a forthcoming article by Cohodes and Goodman, which shows that a state scholarship offered in Massachusetts induces students who would otherwise attend a private college to attend a state college, leading to poorer academic outcomes and degree attainment. Similarly, Walters hypothesizes that Head Start programs can induce families to send their children to the free program rather than a private community-based program, leading to poorer outcomes.

This research can help policymakers identify high-performing programs and improve those that are not. It also demonstrates that not all of the factors identified in previous research on early education lead to effective programs in practice and that, contrary to popular belief, the High/Scope curriculum is not a guarantee of quality.

Additionally, these findings offer important policy insights for program expansion by showing that the children who gain the most from Head Start are those with no alternatives. Policymakers can begin to determine what factors contribute to creating high-quality programs with constrained budgets. It is unlikely that any government-sponsored early childhood program will be able to replicate the success of Perry Preschool, which spent over $11,000 annually per student. An increase to those spending levels would cost the federal government an additional $12.3 billion per year. Nonetheless, Walters’ research helps us understand the necessary steps to create a high-quality program with limited resources.

Article Source: Walters, Christopher. (2014). Inputs in the production of early childhood human capital: Evidence from Head Start. National Bureau of Economic Research, working paper 20639

Feature Photo: cc/(Franklin Park Library)