Economic shocks produced by the Great Recession have contributed to rising food insecurity, with 14.7 percent of U.S. households being food insecure in 2009, compared to 11.1 percent in 2007. Using data from the first two waves of the Michigan Recession and Recovery Survey (MRRS), a unique panel survey of a representative sample of working-age adults in the Detroit Metropolitan Area merged with a novel data set capturing local food assistance programs, this paper explores three research questions related to food security among low-income households: How are adverse economic shocks, such as job loss, associated with increased risk of low or very low food security among households with children? How is access to local food resources (e.g., grocery stores, SNAP offices, food pantries) related to food security? To what extent does the local food resource infrastructure moderate the impact of economic shocks?

To answer these questions, we combine data from a telephone survey of food pantry programs in metropolitan Detroit, location information about SNAP eligibility offices, and data about the location of food retailers with the MRRS survey data which contains the geo-locations of respondents. We then place these data in a GIS and calculate distance- and commute-mode weighted measures of accessibility to different types of neighborhood food resources (e.g., supermarkets, SNAP authorized retailers, SNAP offices, charitable food assistance programs). These access measures allow us to describe the local food resource environment around poor and near-poor households with greater precision than most other studies to date and to carefully examine the association between local food resource infrastructure and food security.

Consistent with other studies, we find that roughly one-third of households with children at or below 300% of the federal poverty line in each wave reported food insecurity in the previous year. Contrary to expectations from the literature on food deserts, we find that low-income residents of the City of Detroit have greater access to food retailers than low-income suburban residents. Low-income urban households also have much greater access to food assistance programs than low-income suburban households. When controlling for household characteristics, we find longer spells of unemployment and detachment from the labor market due to health limitations are associated with food insecurity. Despite differences in levels of access to food retailers and food assistance, we do not find the level of food resource access to be associated with household food security even after accounting for selection.

Our findings provide important insights for future research, underscoring the importance of finding data that can link household food security, earnings and work, and food resource access in space.  Improved understanding of the spatial antecedents of food assistance and food insecurity also provides insights that may translate into more effective programming and more efficient allocation of public program dollars, private capital, entrepreneurial activity, and philanthropic resources to programs intended to reduce food insecurity.