Percentage of fraud in SNAP.
Executive summary
A simple percentage for “SNAP fraud” does not exist; data show that intentional fraud is relatively rare while broader measures of improper payments— which mix errors, administrative mistakes and some fraud—can be materially larger. USDA/FNS reports emphasize very high eligibility and payment accuracy (over 98% eligible; 95.64% payment accuracy), while GAO/USDA estimates for FY2023 put improper payments at 11.7% (about $10.5 billion) of benefits—an apples-to-oranges contrast that must be unpacked [1] [2].
1. What the headline numbers actually measure
The USDA/FNS emphasizes that “over 98% of those receiving SNAP benefits are eligible” and highlights a 95.64% payment accuracy rate, framing most problems as mistakes and showing long-term improvement in payment accuracy [1]. By contrast, the U.S. Government Accountability Office reported USDA’s FY2023 estimate that 11.7% of SNAP benefits were improper—about $10.5 billion—making clear that “improper payments” are broader than intentional fraud and include wrong amounts, eligibility errors and administrative mistakes [2]. Congress’s research office likewise notes there is no single metric that captures all forms of SNAP fraud and that some acts (like trafficking) are always fraud while others (like duplicate enrollment) can be error or fraud depending on circumstances [3].
2. How much of that improper spending is actually intentional fraud?
Multiple sources caution that intentional fraud by recipients or retailers is much smaller than the headline improper‑payment figures imply: policy summaries and advocacy research repeatedly describe SNAP fraud as “rare” and emphasize that many improper payments are errors rather than deliberate theft [3] [4]. Industry and legal analyses point to historic estimates that explicit SNAP fraud—measured narrowly—has been below about 1% of total benefits in some assessments, even as absolute dollar losses can be large when benefits disbursed are themselves sizable [5].
3. Where the biggest criminal losses are occurring and why percentages hide reality
Recent reporting and private-sector studies highlight thefts from Electronic Benefit Transfer (EBT) cards—skimming, cloning and organized crime rings—which can produce big spikes in incidents or dollars stolen in short windows without changing long‑run percentage rates dramatically [6] [7]. News outlets and vendors show a sharp increase in reported fraudulent EBT transactions in particular quarters, but those incident counts do not automatically translate into a stable, program‑wide fraud percentage because USDA’s formal measures don’t always capture every trafficking or EBT‑theft vector in the same metric [8] [6] [3].
4. Accountability, measurement gaps and competing agendas
GAO has urged USDA to improve how it measures retailer trafficking and to act on recommendations—points that expose measurement gaps that allow differing narratives to flourish [2]. Private vendors and fraud‑prevention firms push multi‑layered tools and studies claiming large costs and higher fraud rates, which can reflect business agendas for selling solutions [9] [10]. At the same time, USDA and state agencies emphasize program integrity and high eligibility rates to defend the program’s core purpose and political support [1].
5. Bottom line: answer to “Percentage of fraud in SNAP”
If the question means “what share of SNAP benefits are improper,” USDA’s FY2023 estimate is 11.7% (about $10.5 billion) but that figure includes errors and administrative mispayments as well as some fraud [2]. If the question means “what share is intentional fraud (trafficking, recipient or retailer fraud),” authoritative summaries and CRS/advocacy reporting characterize that kind of fraud as rare and, in many analyses, materially below the headline improper‑payment rate—often cited in single digits and in some estimates under 1% of benefits—though exact national percentages for all fraud subtypes are not available in a single consistent measure [3] [5] [4]. The public debate thus conflates different measures: improper payments (11.7% in FY2023) versus intentional fraud (a much smaller, harder‑to‑pinpoint share) [2] [3].