Heterogeneous Effects of SNAP Participation on Food Insecurity across Income Levels

Background / Purpose: The Supplemental Nutrition Assistance Program (SNAP) is a critical policy aimed at improving food access and reducing food insecurity among low-income households. Nutritional intake is linked to long-term health outcomes, and early-life health plays a significant role in shaping later socioeconomic outcomes, including income. In this context, food insecurity is not only a short-term hardship but also a key factor associated with broader health and socioeconomic inequalities. Prior research shows that SNAP reduces household food expenditure burdens and helps alleviate food insecurity. A large body of literature has examined the relationship between SNAP participation and food insecurity, with income as a key variable. However, relatively little attention has been paid to whether the effect of SNAP on food insecurity varies across income levels. This is an important limitation, as SNAP’s impact is likely heterogeneous. Understanding this heterogeneity is essential for evaluating SNAP effectiveness and informing policy design.

Methods: This study used 2024 Survey of Income and Program Participation (SIPP) data from 382 households in Virginia, a single-state sample chosen to ensure administrative and policy consistency across households. The unit of analysis was the household, and a household was considered a SNAP participant if at least one member reported participating in SNAP. Food insecurity was measured using the USDA 6-item scale and constructed as a continuous variable based on the USDA scoring guidelines. The key independent variables were SNAP participation and annual household poverty ratio. An interaction term between SNAP and poverty ratio was included to test whether the association between SNAP participation and food insecurity varied by income. Control variables included age, gender, race, number of children aged 15 and older, and employment status. Hierarchical regression models were estimated by sequentially adding income, controls, and the interaction term.

Results: In Model 1, SNAP participation was positively associated with food insecurity (B= 2.90, p <.001), likely reflecting selection into the program among more vulnerable households. In Model 2, SNAP remained significant (B= 2.67, p <.001), while higher income was associated with lower food insecurity (B= -0.04, p= .006). In Model 3, the interaction between SNAP participation and income was negative and statistically significant (B= -0.92, p <.001), indicating that the negative association between income and food insecurity was stronger among SNAP participants. Specifically, while SNAP participants exhibited higher levels of food insecurity at lower income levels, increases in income were associated with a more pronounced reduction in food insecurity compared to non-participants.

Discussion and implication: The findings are consistent with prior research showing that higher income is associated with lower food insecurity. The interaction results indicate that the association between SNAP participation and food insecurity varies by income level, with stronger associations observed among lower-income households and weaker associations as income increases. These findings suggest that SNAP participation is more strongly associated with outcomes among economically disadvantaged households. From a policy perspective, the results highlight the importance of considering heterogeneity within low-income populations when evaluating SNAP effectiveness.