Poster #28 - Does Head Start Create Lasting Benefits to Children? Analysis with Additional CNLSY Data

• In Event: Saturday Poster Luncheon

Abstract

Background: Head Start was launched in 1965 as part of the “War on Poverty” and, by 2019, served over onemillion children annually with a budget of $10 billion (Bailey et al., 2021). Although a substantial body of researchdocuments the program’s short-term benefits, questions remain about its long-term impact. Deming (2009) evaluated long-term outcomes for Head Start participants across several domains—including academic achievement, high schoolgraduation, criminal behavior, employment, and health—using data through 2004. By then, the youngest cohort in thesample had reached age 19, an age sufficient for assessing adolescent and early adult outcomes such as high schoolgraduation, college attempt, criminal involvement, and health status. However, outcomes like college completion and adultcriminal behavior were less fully observed, as disadvantaged youth may delay postsecondary enrollment, and criminalrecords often emerge later in adulthood. These data limitations underscore the need to revisit Deming’s analysis with morerecent data—a gap this study seeks to address.
Research Question: Does Head Start generate lasting benefits for disadvantaged children that persist into later adulthood?

Methods: Building on Deming’s 2009 paper, Early Childhood Intervention and Life-Cycle Skill Development: Evidence fromHead Start, this study extends the analysis using an additional 16 years of data from the National Longitudinal Mother-ChildSupplement (CNLSY). CNLSY follows all children born to female respondents from the NLSY79 cohort, with biannual surveysconducted between 1986 and 2014. This panel structure enables sibling comparisons using a family fixed-effects approach.We focus on families with children over age four by 1990, comparing siblings who differed in Head Start participation toestimate effects on long-term outcomes, particularly focusing on higher education enrollment, criminal behavior, and healthstatus.

Results: This study finds that Head Start participants are 11 percentage points more likely to enroll in college than theirnon-participating siblings—nearly doubling the 6 percentage point gain observed in Deming’s original study for collegeattempt. Gains in health status also persist in the extended follow-up, reinforcing prior findings. In contrast, effects oncriminal behavior remain statistically insignificant.
Implications: These findings suggest that the long-term benefits of Head Start not only persist but may grow stronger overtime. By documenting sustained gains in education and health, this study adds to the body of evidence supporting HeadStart as an effective and worthwhile investment for improving life outcomes among disadvantaged children.

Authors

• Hye-Min Jung, Columbia University
  Presenting Author

• Jiawen Cui, Columbia University
  Non-Presenting Co-Author