State Contexts and Food Bank Use By Low-Income Households with Children

Background

The food security infrastructure in the United States includes public programs, such as the Supplemental Nutrition Assistance Program (SNAP; formerly Food Stamps) and private service providers such as food banks. Cash transfer programs like the Earned Income Tax Credit (EITC) and cash assistance under Temporary Assistance for Needy Families (TANF) also facilitate the purchase of necessities such as food. Public food and cash aid programs are a common focus of studies examining food security in low-income households (Shaefer & Gutierrez, 2013; Bartfeld & Dunifon, 2006).

The use of private food providers by households in need has been less well examined (Bhattarai, Duffy, & Raymond, 2005). Additionally, the public and private food safety nets could interact. A household may be more likely to turn to private food providers when public supports are less accessible. Importantly, states are afforded varying degrees of discretion over the implementation of programs such as TANF and SNAP, and state policies affect program take-up. In this paper, I consider the relationship between state safety net policy and the use of private emergency food resources by food insecure households with children, a group for whom food security has consequences in areas such as academic performance and emotional wellbeing.

Methods

I draw data from the Current Population Survey Food Security Supplement (CPS-FSS). The CPS is a large, nationally representative survey of non-institutionalized U.S. households conducted monthly. The FSS is a supplement administered once per year to households below 185% of poverty to gauge food hardship. It also includes questions on use of food resources. I use the CPS-FSS to create a binary dependent variable indicating if the household used a food bank in the past year.

I estimate multilevel logit models, with households nested in states, of the probability of a household reporting food bank use. I restrict the sample to households with children and estimate models for a year prior to the “Great Recession” (2005; n=1,316), during the recession (2010; n=1,627), and well into the economic recovery (2015; n=1,153). Key predictors are variables representing various aspects of state policy, including TANF cash assistance rules, SNAP/food stamps accessibility, and presence of a state EITC. I control for relevant household (e.g., number of children) and state (e.g., unemployment rate) characteristics.

Results and Implications

Models indicate that the probability of a household using a food bank varies by state. Among state policies, a TANF cash assistance time limit of less than the federal guideline of 60 months is associated with an increased probability of food bank use in all three years, though it is only statistically significant (at the 0.05 level) in 2005. There are few other consistent state-level predictors of food bank use. Interestingly, the patterns of sign, magnitude, and statistical significance of several household-level predictors noticeably vary across years. I consider possible reasons for and consequences of these differences. Findings are relevant to current discussions about further restricting access to programs such as SNAP, which may in turn burden the private food safety net.