SNAP benefits surged ahead of the 2021-22 grocery-price spike, then fell back below it once pandemic boosts expired

Summary

Food assistance more than kept pace with the grocery-price spike while pandemic-era emergency allotments and the 2021 Thrifty Food Plan (TFP) update were both in effect (fiscal 2021-2022), then fell behind once those temporary boosts expired, even though grocery prices kept rising. The average SNAP benefit per person (nominal) rose about 49% from FY2020 to its FY2022 peak, versus a 19% rise in the food-at-home CPI over the same span — real purchasing power roughly +25%. But from FY2022 to FY2024, the average per-person benefit fell about 19% in nominal terms (emergency allotments ended February/March 2023) while food-at-home CPI kept climbing another 3%, so by FY2024 the average benefit's real purchasing power was back to, and by FY2025 essentially at, its pre-spike FY2020 level (index ~96-98 vs. 100).

What the numbers show (AskAmerica data)

Computed from fiscal.snap_benefits_by_geography (USDA FNS, sum of total_benefits_usd / sum of persons_participating across all states, annual fiscal-year rows) and econ.food_cpi (BLS CUUR0000SAF11, December level each year):

SQL: SELECT "year", SUM(total_benefits_usd) tot_b, SUM(persons_participating) tot_p FROM fiscal.snap_benefits_by_geography WHERE "month" IS NULL AND "year" BETWEEN 2018 AND 2025 GROUP BY "year", divided by 12 and by tot_p; and SELECT "year","value" FROM econ.food_cpi WHERE series='CUUR0000SAF11' AND "period"='M12'. Note: rows for FY2018-19 in the SNAP table carry an internal scale inconsistency (values orders of magnitude smaller than surrounding years) and were excluded as a likely data-quality defect, not used in any of the figures above.

Why benefits initially outran prices, then fell behind

• Emergency Allotments (EAs). From spring 2020 through February/March 2023, COVID-era emergency allotments topped every household up to the maximum benefit for its size (or added a flat $95/month minimum), pushing the per-person average well above the pre-pandemic formula amount — this is most of the FY2021-22 surge in the chart.
• The 2021 Thrifty Food Plan re-evaluation. In October 2021, USDA re-costed the Thrifty Food Plan — the basket that sets base SNAP benefit levels — for the first time without the 45-year-old "cost-neutral" constraint, permanently raising benefits about 21% above where simple CPI-indexing would have put them (National Agricultural Law Center; CBPP).
• The October 2022 cost-of-living adjustment (COLA) ran 12.5% for FY2023 — the largest since 1975 — but food-at-home prices were up roughly 13.1% year over year at that point, so even that unusually large increase came in slightly under the inflation it was meant to offset (NYC Food Policy Center; GAO).
• EAs ended in Feb/March 2023 (all remaining EA states wound down by the FY2023-FY2024 transition), which is the single largest driver of the nominal per-person benefit falling from about $230 (FY2022) to about $187 (FY2024) — an 18.8% nominal drop — while grocery prices continued rising, if more slowly, over the same window. USDA's own Economic Research Service documented this same pattern nationally: "SNAP Spending Rose and Fell With Pandemic-Era Changes to Benefit Amounts" (USDA ERS, June 2024).
• The October 2023 COLA was only about 3.5%, roughly 1.4 points below the trailing year-over-year food inflation rate at that point (per Yahoo Finance/various coverage summarizing the FY2024 adjustment), meaning the annual COLA mechanism itself was, by then, no longer even matching current food inflation, let alone making up the post-EA drop.

Independent research confirms the same pattern

The Urban Institute and the Congressional-Farm-Bill-mandated Urban Institute study of the reevaluated TFP and Emergency Allotments both find that (a) the combination of EAs and the re-costed TFP pushed SNAP purchasing power well above pre-pandemic and pre-inflation levels during 2021-2022, and (b) the subsequent unwind — EAs ending nationally by March 2023, combined with continued (if decelerating) grocery inflation — left a "persistent gap between benefits and food costs" that by 2024 had returned households closer to, or below, their 2019-2020 real purchasing power. The Cato Institute reaches the same factual conclusion about the TFP's scale (21% above the cost-neutral trajectory) from the opposite policy angle, arguing it broke the program's traditional cost discipline rather than that it was needed.

Caveats

• This uses the U.S. average benefit and the national food-at-home CPI; state-level SNAP maximum benefits and state-level grocery CPI both vary, and geographic mismatch (SNAP benefits are set nationally with only an Alaska/Hawaii adjustment, while grocery prices vary regionally) is not captured here.
• Average benefit per person mixes household-size and income composition effects with the policy/COLA effects being described; it is a reasonable proxy but not a pure "same household, same size, before and after" comparison.
• FY2018-2019 rows in the SNAP source table show an internal unit-scale inconsistency versus FY2020+ and were excluded rather than risk a spurious pre-pandemic comparison; this does not affect the FY2020-2025 comparison presented.
• Fiscal-year (Oct-Sep) SNAP totals are compared against December CPI as an approximation of the year's price level; a fiscal-year-average CPI would shift point estimates slightly but not the direction of the finding.