Identifying the Effects of the Supplemental Nutrition Assistance Program on Child Health

Background: The primary objectives of the Supplemental Nutrition Assistance Program (SNAP), formerly known as the Food Stamp Program, are to improve diets and reduce food insecurity among low-income households, especially those with children. Childhood obesity has more than doubled in the last 30 years (CDC, 2015), and 15.8 million children are living in food insecurity in 2014 (USDA, 2015). There is no consensus in the literature on the effect of food assistance programs on food insecurity, the nutritional quality of diets, and overall health of participants, especially among children. Given that participants in food assistance programs are not randomly selected, and that unobserved factors may determine program participation (Kaushal, 2007; Mykerezi & Mills, 2010; Meyerhoefer & Yang, 2011), identification strategies are crucial to understanding the programs’ impacts on child health outcomes. Thus, I focus on identifying the effects of SNAP by addressing the endogeniety of program participation.

Methods: This study uses data from the Fragile Families and Child Wellbeing Study (FFCWS), which contains rich measures of child health and cognitive development in low-income households in the US. I use four follow-up waves: 1-year, 3-year, 5-year, and 9-year for general health outcomes reported by mothers; I use three follow-up waves: 3-year, 5-year, and 9-year for obesity outcomes. The sample consists of 3,860 children who were identified in all four waves. Participations in food programs are measured as binary variables and the amounts of benefits are measured as continuous variables. I use individual fixed-effects models to capture how child health status changes as program participation changes, although these models are subject to major endogeneity concerns. To address this, I expand the model by using instrumental variables for panel-data models to control for unobservable state-benefit heterogeneity, which may influence the program participation. The key instrumental variables are state payment error rates, broad-based categorical eligibility, and the implementation of fingerprinting.

Results: Results from the simple fixed effects models show that program participation negatively influences general health (b=-.055, p<.05) and BMI (b =-.074, p<.05). These results are consistent with previous studies reporting that children with parents’ receiving welfare are more likely to have health problems and lower educational attainment (Chen el al., 2005; Danzinger et al., 1999; Ku & Plotnick, 2003; Pylypchuck, 2008). However, the model does not take into account the endogeniety problems of program participation. Using the IV’s described above, I find that these negative relationships disappear for both outcome variables. These findings are consistent with current research that addresses endogeneity problems (Kaushal, 2007; Kreider, Peper, Gundersen, & Jolliffe, 2012), suggesting that participation in nutrition assistance programs does not negatively affect child health.

Implications: This study found that receiving SNAP benefits does not negatively influence child health. The findings indicate that food assistance programs do not negatively impact children’s health and development in vulnerable populations.