Session Recap: New Evidence on Head Start
November 11, 2014 10:30 AM
By Alix Gould-Werth, University of Michigan
Head Start has been heralded as an intervention with both short- and long-term effects on participant human capital, and also decried as an inefficient waste of taxpayer dollars. The papers in this session contributed to the large body of work identifying impacts of early childhood education on a wide range of outcomes. Two papers examined the multi-generational impacts of program participation, and two others focused on variation in response to treatment, by classroom quality and by neighborhood context.
Passing the Baton: Does a Parent's “Head Start' Affect Their Children's Outcomes? By Cassandra Hart, University of California, Davis
Hart’s paper used PSID data to examine how long-run Head Start outcomes affect parenting behavior and thus child outcomes. While the results were characterized by null findings and large standard errors in many domains, some findings were interesting and counter-intuitive. For example, parental Head Start participation was associated with lower levels of warmth toward children. The paper also identified effects for child cognitive ability which faded out over time—as the second Head Start generation ages, it will be possible to see if there is a resurgence of effect later in the life course with G2, as has been documented in G1.
The Longer Long-Term Impact of Head Start: Intergenerational Transmission of Program Effects by Andrew Barr and Chloe Gibbs, University of Virginia
This paper used data from NLSY79 and C-NLSY, family fixed effects, and Head Start availability as an instrument to examine multi-generational impact of parental Head Start attendance. Despite the small sample size (n=318), these analyses uncovered “improved longer-term outcomes for the second generation, in the form of increased educational attainment and decreased engagement in risky behaviors,” which were particularly concentrated among African Americans.
Program Level Variation in Head Start Impacts: The Moderating Role of Classroom Quality by Allison H. Friedman-Krauss, Maia Connors, and Pamela Morris, New York University; Avi Feller and Dana Charles McCoy, Harvard University
The authors in this study acknowledge that Head Start participation is not a uniform treatment, asking the question “to what extent does variation in the quality of the Head Start center account for variation in cognitive and social-emotional outcomes for participants?” This study used random assignment and HLM to show that Head Start Center quality does affect early math achievement, vocabulary, and display of externalizing behaviors. The center qualities that were associated with improved outcomes were home visits, student-teacher ratio, and class-size, though these outcomes could serve as proxies reflecting overall levels of fidelity to the Head Start model and best practices.
Head Start and Children's School Readiness: Variations by Family and Neighborhood Poverty by Taryn Morrissey, U.S. Department of Health and Human Services, and Katie Vinopal, American University
This paper is theoretically situated in the debate between the compensatory hypothesis (the idea that the most disadvantaged children have the most to gain from intervention) and the idea of accumulated advantage (the idea that children who already have resources and support are best poised to take advantage of the benefits offered by interventions). Using ECLS-B data linked to neighborhood characteristics from the ACS, this paper investigated how school readiness varies by poverty level and if neighborhood poverty mediates school readiness. Results presented suggested that neighborhood poverty does not act as a moderator.
Comments from discussants and audience members pushed the panelists in two major directions. First, they encouraged panelists looking at multi-generational effects to examine a wider range of mechanisms through which Head Start participation could affect parent behavior and thus child outcomes. Particularly germane was the link between parent participation Head Start as a child and employment status as an adult, which could affect child outcomes in both positive and negative directions. Second, presenters were asked to imagine data sources that could allow them to more precisely measure variation in context, both classroom quality and neighborhood characteristics.