SNAP Purchasing Power and Household Time Allocation: Evidence from Geographic Variation in Food Prices

• In Event: SNAP Program Design: Implications for Participants and Policy

Abstract

The Supplemental Nutrition Assistance Program (SNAP) provides uniform benefits across the contiguous United States, yet geographic variation in food prices means that two otherwise identical households may experience very different levels of real support depending on where they live. While research has extensively examined SNAP's effects on food security and nutrition, we know relatively little about how the program's real value shapes household time allocation. This paper investigates how differences in SNAP purchasing power affect how households spend their time.

I construct local measures of purchasing power by combining the USDA's Quarterly Food-at-Home Price Database with maximum SNAP benefit levels across 32 market areas from 2003 to 2010. Purchasing power is defined as the ratio of the SNAP maximum allotment to the local cost of the Thrifty Food Plan, the USDA's reference food basket for an adequate diet. I merge this information with detailed time diaries from the American Time Use Survey to measure time spent on six activities: market work, home production, childcare, sleep, leisure, and purchasing time.

The empirical strategy exploits within-market, overtime variation in purchasing power. Because local food prices can affect time use for all low-income households, I implement a triple-difference design that compares time-use responses of SNAP-eligible households to near-eligible households within the same market and year. The identifying assumption is that both groups face similar price environments, but only eligible households can access benefits.

The results show that SNAP purchasing power affects time allocation along two margins. A 10-percent increase in purchasing power raises leisure time among eligible households by approximately 11 minutes per day – 3.9 percent of the baseline mean – and reduces time spent purchasing by approximately 3 minutes per day, both relative to near-eligible households facing the same price changes. Effects on market work, home production, childcare, and sleep are small and statistically indistinguishable from zero in the full sample. These results are robust across alternative specifications, placebo tests, and permutation inference. Heterogeneity analysis reveals meaningful variation across groups. Non-White households exhibit substantially larger responses, with a 14 minutes per day increase in leisure, a 12 minutes per day reduction in childcare time, and 6 minutes per day reduction in purchasing time. Women show a significant increase in market work (9 minutes per day), suggesting that relief from food budget management may free time and cognitive capacity for employment.

Author

• Georgina Cisneros, American University
  Presenting Author