The Role of the Supplemental Nutrition Assistance Program in Food Insecurity and Food Acquisitions

Given the size of SNAP in both households served and public dollars spent, it is of great importance to understand the impact of SNAP dollars and how they affect households. Food insecurity and diet quality have potential impact on our health care system and economy. This dissertation looks at some of the most vulnerable SNAP households – those near the $0 net income threshold for maximum SNAP benefits – using the Household Food Acquisition and Purchase Survey (FoodAPS), collected in 2012 to early 2013, in a fuzzy regression kink framework using net income as an instrument, as well as descriptive analyses. First stage findings indicate that net income based on household-reported survey data does not perfectly fit expectations for SNAP benefit allotment, implying that current household economic status does not necessarily match status at time of application. This is as expected since the measure of net income used here is not collected at the time of SNAP benefit application. However, households with positive net income do receive lower benefit amounts with higher net income, though the slope is less steep than predicted by the benefit allotment calculation. I examine the effect of SNAP benefit amount on three outcomes: food insecurity, food spending, and nutritional quality. I am unable to reject no effect of SNAP benefit amount on any outcome in this subpopulation. However, in these households food insecurity remains high and nutritional quality does not meet national recommendations. Additionally, households with negative net income are, as we might expect, worse off – they are more likely to struggle to pay important bills on time. However, they are also less likely to turn to places of worship for food support. I also find suggestive evidence that nutritional quality is improved by SNAP benefit amount particularly in week 3 after receiving SNAP benefits. This research suggests that the SNAP benefit formula is not harmful, but further that if SNAP goals are to be met, policies may need to increase benefits for all households. Further work should focus on the effects of increased benefit amounts among a larger portion of the SNAP population.