Association for Public Policy Analysis & Management

The Supplemental Nutrition Assistance Program (SNAP) is the largest of the federal food programs aimed at addressing food insecurity and serves a significant share of food insecure households. By effectively addressing food insecurity, SNAP has the potential to also significantly impact the many adverse child and adult outcomes related to food insecurity, including malnutrition, developmental deficits, and mental health problems. Although SNAP can alleviate food insecurity, there is also compelling evidence that SNAP recipients’ caloric intake, fruit and vegetable consumption, and food expenditures decline over the monthly SNAP cycle. Such declines also point to possible variability in food insecurity over the month, with food insecurity increasing as these other food consumption and purchasing behaviors decline. As yet, no studies have captured daily or weekly variation in food insecurity to assess the change in food insecurity over the SNAP cycle. This paper will address this gap in the literature using a new survey instrument to determine whether food insecurity increases as a function of the SNAP cycle for a sample of North Carolina SNAP recipients.

Data and Methods

Using an adaptation of the Household Food Security Module, a common measure of food insecurity, this study measures perceived food security in a convenience sample of SNAP participants in Durham, North Carolina. Study participants were approached at public housing communities and at the local social services administrative offices. The target sample size is 200, with recruitment scheduled to be completed by the end of the month. The instrument also collects information about the number of days passed since the participant’s most recent SNAP benefit receipt to identify whether food security is a function of the recency of receipt. Because, in North Carolina, the last digit of the head of household’s social security number determines the day of the month a household receives SNAP benefits, this recency measure is effectively random and exogenous. This serves to control for any unobserved individual differences between participants. It is predicted that perceived food insecurity will be positively associated with the number of days passed since the most recent SNAP benefit disbursement. To rule out two important alternative explanations, this study also collects information about consumption smoothing habits and alternative food sources. Structural regression models will be utilized and will also include demographic covariates.

Results

Currently, the sample includes 61 households, 60% of which are households with children. On average, all households are experiencing moderate food insecurity (Mean = 5.2 on a scale of 0 to 10, with 10 being most food insecure). However, preliminary analyses find that the relationship between days since the most recent SNAP benefit disbursement and food insecurity is qualitatively different for households with and without children. For households with children this relationship is curvilinear, with food insecurity peaking toward the middle of the SNAP month. For households without children this relationship is linear and increasing over the SNAP month. Results from this study will have implications for the design of SNAP policies that may reduce the variability in food insecurity of SNAP households.