The Intergenerational Effects of Head Start on Infant Health

• In Event: Enrollment and Policy Effects of Assistance Programs Including SNAP, TANF, and Head Start

Abstract

This paper evaluates the intergenerational effects of maternal access to the largest federal preschool program in the United States, Head Start, on infant health at birth. We use Vital Statistics Natality records covering births from 1970 to 2019 to mothers born between 1955 and 1975. To identify causal effects, we leverage variation in the year in which Head Start was introduced in each county from 1965 to 1980, constructed from three administrative sources: the Poverty Program Information reports, the Community Action Program files from the National Archives and Records Administration, and the Federal Outlay System Files. Our empirical strategy uses a difference-in-differences design combined with inverse propensity-score weighting to address non-random rollout timing, and we employ the heterogeneity-robust estimator of Sun and Abraham (2021) to accommodate the staggered introduction of the program. 

We find that a mother's access to Head Start during childhood improves infant health in the next generation. Treatment-on-the-treated estimates show that maternal exposure to Head Start increases birth weight by 44 grams (a 1.3% gain) and reduces the likelihood of low birth weight by 11.8%. These improvements extend across multiple dimensions of infant health: we also document reductions in the incidence of small-for-gestational-age births and preterm births. The effects are concentrated around the low-birth-weight threshold rather than in the extreme left tail of the distribution, and are substantially larger for Black mothers and for mothers residing in counties with above-median poverty rates in 1960, consistent with the program disproportionately benefiting children from the disadvantaged backgrounds it was designed to reach. The results are robust to alternative fixed-effects specifications, sample restrictions, and recent estimators for staggered adoption designs. 

As a potential pathway, we show that Head Start availability reduces the likelihood of drinking during pregnancy, while having no detectable effect on smoking, the timing of prenatal care, or reported medical risk factors. We also find a small increase in the probability of being married at the time of the first birth, suggesting that marriage-market channels may contribute to the improvements in birth outcomes. In contrast, we find no evidence that Head Start changed the age distribution of first births or overall fertility. 

These findings indicate that the benefits of early childhood programs have enduring effects that carry over to the next generation, implying that the returns on these types of public investments may be significantly underestimated by evaluations focused on a single generation.

Author

• Esra Kose, University of California, Merced
  Presenting Author