Multi-generational Effects of Prenatal and Early Life Access to SNAP

Year: 2017

Research Center: Tufts University/University of Connecticut (UConn) RIDGE Program

Investigator: Page, Marianne and Chloe East

Institution: University of California, Davis

Project Contact:
Marianne Page
University of California, Davis
Department of Economics
1 Shields Avenue
Davis, CA 95616


There is substantial evidence that health and socioeconomic inequalities persist across generations. Poor children begin life with significant health and academic disadvantages compared to non-poor children, and these gaps widen as children age. By the time the children enter adulthood, individuals who grow up in low-income families are more likely to exhibit a number of chronic health conditions, including obesity and diabetes, and are more likely to grow up to experience poor labor market outcomes. Recent research finds that early life access to the Supplemental Nutrition Assistance Program (SNAP) may help alter these trajectories, however: In-utero and early childhood access to SNAP has been shown to reduce the incidence of metabolic syndrome in adulthood and increase self-sufficiency.

This study builds on these recent findings to consider whether the benefits of SNAP also persist to later generations. Literature in the biological sciences, economics and child development all predict that the causal impacts of effective nutrition interventions should spill over onto offspring. First, the effects of improved prenatal nutrition on exposed children’s health may transmit to later generations’ health through direct biological processes. Second, improvements in later life economic outcomes that result from early life nutrition programs will facilitate individuals’ ability to invest in their offspring’s health. SNAP may also indirectly affect the later generation’s outcomes by changing treated individuals’ fertility.

In spite of these predictions, to date, little is known about the extent to which public policies like SNAP alter multigenerational linkages. As SNAP currently reaches approximately 20 percent of U.S. children, the program has significant potential to reduce heath and socioeconomic disparities that persist across generations. Evidence of substantive, generationally persistent effects, would also suggest that existing cost-benefit analyses underestimate the SNAP’s true value.

This study uses sharp differences in the county-by-county timing of SNAP (formerly called Food Stamp) program roll-out between 1961 and 1975 to create “treatment” and “control” groups.

These treatment and control groups are then used to identify the effects of early life food assistance on individuals’ later life health, and the health of their eventual offspring. The identification strategy builds on previous work documenting that the timing of counties’ program adoption was quasi-random. The hypothesis is that infants born in counties that had already adopted Food Stamps experienced a healthier in-utero environment than infants born before the program was implemented, and that their improved health lead to better health in the next generation.

California Vital Statistics Natality files were used to match information provided on the birth certificates of women born in California during the 1960s and 1970s, when the Food Stamp program was in the process of being adopted, to health information provided on the (California) birth certificates of their eventual offspring. The study examines first and second-generation birthweight, which is known to be a strong predictor of later life health and labor market success. The study also examines whether early childhood access to Food Stamps affected later life fertility. This is important, as changes in fertility could be driven by changes in infant survival: Increases in survival will generally be driven by infants on the margin, who will be on the low end of the birth-weight distribution, which may lead to a decline in the second generation’s average birth weight. The study focuses on event study research designs that control for time invariant county characteristics, and national “shocks” as well as county specific-time varying measures of welfare spending and economic conditions.

The results suggest that in utero and early life exposure to the Food Stamp program during its initial rollout period led to improvements in the “first generation’s” health at birth. This result is consistent with evidence from previous studies. Among later offspring (the “second generation”) average birth weight was lower for treated cohorts than for untreated cohorts, but fertility was higher. These results are suggestive of “selection effects,” whereby access to SNAP during the prenatal period improves females’ later reproductive health. Future research will investigate this potential mechanism.