Ever since the Food Stamp Act of 1964 created the U.S. Supplemental Nutrition Assistance Program (SNAP), policymakers have debated whether benefits should be permitted to be spent on sugary drinks like soda. From an economics perspective, the impact of such a restriction is unclear. Standard theory suggests that because the majority of recipients spend more on groceries than their SNAP allotment, they will simply switch to paying for soda with non-SNAP funds. But recent research suggests that because of mental accounting—where recipients treat SNAP benefits as a separate budget for eligible groceries—removing sugary drinks from the eligible list could reduce consumption altogether.
In this paper, the authors test this question in the context of a recent policy change. Throughout the first half of 2026, ten states implemented the first-ever policies disallowing the use of SNAP benefits on sugary drinks. The authors use data on grocery shopping behavior to compare purchases in these states with those in control states before and after the policy took effect. They also survey recipients to observe changes in public opinion and perceptions of stigma, and combine their results with an economic model to estimate the overall welfare effects: changes in economic welfare caused by changes in economic variables, like economic shocks, income, employment, taxes, and so on .
They find the following:
- The restrictions reduced SNAP recipients’ retail purchases of excluded drinks by 12.4%. This translates into a reduction in sweetened drink consumption of about 404 fluid ounces per covered person per year, or about 34 12-ounce cans.
- In states that excluded only some sugary drink categories—for example, soda but not energy drinks or fruit drinks—SNAP recipients diverted up to 39% of the excluded drink calorie reduction to increases in retail purchases of non-excluded sugary drinks and fruit juices. This suggests that continuing to include some sugary drinks in SNAP will naturally undo some of the potential consumption reductions.
- The authors find much stronger effects on retail excluded drink purchases among households that pay for a larger share of their groceries with EBT cards, consistent with the restrictions operating through mental accounting. They also find suggestive evidence of stronger effects among households whose survey respondents reported perceiving little or no health risk from consuming soft drinks.
- Support for SNAP sugary drink restrictions was much larger among non-SNAP recipients than SNAP recipients.
- The restrictions caused statistically significant 0.10 to 0.17 standard deviation increases in three measures of SNAP recipients’ perceptions of stigma: feeling judged when paying with SNAP benefits and perceiving that the current SNAP policies in their state were disrespectful or took away personal freedom.
- Excluding all sugary drinks from SNAP nationwide would provide benefits of about $1.1 billion per year. About 70% of these benefits come from the predicted health care cost reduction, stemming from predictions that the restrictions would cause the average adult SNAP recipient to lose 0.27 pounds and would reduce the risk of developing type 2 diabetes over the next 10 years by 2.5%. The remaining 30% comes from consumer welfare gains from reduced sugary drink consumption in the presence of an over-consumption internality. This estimate of consumer welfare gains may be overstated as it does not account for potential losses from increased stigma.
The SNAP sugary drink restrictions are effective at reducing consumption, and welfare modeling suggests that extending these restrictions to all sugary drinks nationwide would generate about $1.1 billion in annual benefits, driven primarily by reductions in public health care spending. Two findings, however, complicate a straightforward policy endorsement. First, in states that excluded only some sugary drink categories, recipients substituted toward the drinks that remained eligible, partially offsetting the consumption reductions and the associated health and welfare gains. Second, the restrictions increased SNAP recipients’ perceptions of stigma, costs which the authors’ welfare estimates do not account for. Whether stigma effects persist over time, and how to value them, remains an open question.