To understand how clients use food assistance programs, and how policy changes in public food assistance impact private food pantries, we constructed a novel data set that matched administrative client use records from the Food Bank of the Southern Tier, which services the food pantries in a six-county region of Upstate New York, with similar administrative client data from SNAP in the same region. The resulting data set is, to the best of our knowledge, the first and largest such matched private and public food assistance data set in existence. To study longer-term food security dynamics, we developed a new measure that can be applied to existing, long-standing household survey data: the Probability of Food Security (PFS), which represents the estimated probability that a household’s food expenditure equals or exceeds the minimal cost of a healthful diet, based on the USDA’s Thrifty Food Plan cost. PFS is calculated using simple household food expenditure data, rather than the more complex household food security survey module used by USDA (Food Security Scale Score – FSSS). We find that PFS and FSSS are strongly and positively correlated and that their associations with household attributes are consistent, indicating that our simpler PFS model is a useful complement to the FSSS.
