Why Grocery Prices Don't Come Back Down: The One-Way Ratchet, Explained

There's a moment most grocery shoppers have had in the past few years: the news says inflation is "cooling," "easing," "back to normal" — and then the receipt at the register looks exactly as bad as it did last month. Both things are true at once, and the reason they can both be true is the single most misunderstood mechanic in American economic life.

This explainer walks through that mechanic: why grocery prices ratchet up during an inflationary burst, click into place, and then — with rare, specific exceptions — never come back down.

"Inflation fell" doesn't mean what it sounds like

Inflation is a rate of change, not a level. When the Bureau of Labor Statistics reports that food-at-home prices (that's the government's term for groceries, as opposed to restaurant meals) rose 1.2% in 2024, that means prices ended the year 1.2% higher than they started it — on top of every previous increase.

Here's the recent sequence, from USDA Economic Research Service and BLS data: grocery prices rose 3.5% in 2020, another 3.5% in 2021, a remarkable 11.4% in 2022, 5.8% in 2023, and 1.2% in 2024. In 2025, per BLS, they rose another 2.4%. USDA researchers calculate that between January 2020 and January 2023 alone, food-at-home prices climbed 24.0% — against a long-run average pace of about 2.6% a year over 2005–2024.

Notice what's not in that chart: a negative number. When 11.4% gives way to 1.2%, headlines say prices are cooling — and they are, in the rate sense. But the level, the thing your receipt reflects, is the sum of every bar stacked on top of the last. A slowdown in inflation (economists call it disinflation) still means the ratchet clicked forward. It just clicked forward more slowly.

Disinflation means the ratchet clicks slower. Deflation — the ratchet turning backward — is the thing that almost never happens.

The ratchet, part one: most of your food dollar isn't food

To see why the ratchet rarely reverses, follow the money. USDA's Food Dollar Series decomposes each dollar Americans spend on domestically produced food. As of the 2024 data, just 11.8 cents of that dollar went to farms. The other 88.2 cents — the "marketing share" — paid for everything that happens after the farm gate: processing, packaging, transportation, energy, wholesaling, retailing, advertising, and, biggest of all, food-service labor.

This split explains a puzzle: farm commodity prices do fall all the time. Wheat, corn, milk, and cattle trade on volatile markets that swing down as easily as up. But since farm value is barely a dime of your retail dollar, even a steep drop in commodity prices moves the shelf price only slightly. The other 88 cents ride on costs with a very different personality.

The ratchet, part two: sticky costs

Economists call prices and wages that resist falling "sticky," and stickiness is concentrated exactly where the food dollar lives.

Wages are the stickiest cost of all. Trucking, warehousing, meatpacking, stocking, and checkout are labor-intensive, and wages in America essentially never fall in nominal terms. Employers who cut pay outright tend to lose their best workers first, so instead they freeze pay or slow raises. Once the wage increases of 2021–2023 were baked into the cost of moving a box of cereal from a plant in Iowa to a shelf in Ohio, they were permanent. (The same dynamic shows up in the restaurant world — we've traced how labor costs and customs interact in our explainer on why tipping took over America.)

Menu costs and contracts lock prices in. Food manufacturers negotiate long-term supply contracts, and retailers change shelf prices deliberately, not continuously. When costs surge, firms raise prices in visible jumps; when costs ease, the path of least resistance is to hold prices and let improved margins quietly rebuild profits — or spend them on promotions rather than permanent cuts.

Shrinkflation is a one-way valve, too. A common response to cost spikes is shrinking the package — 59 ounces of orange juice where 64 used to be. Packages very rarely grow back when costs ease. The BLS does adjust the Consumer Price Index for package size, so shrinkflation shows up in the data as a price increase, but on the shelf it hides the ratchet from casual view.

Expectations do the rest. Once shoppers, suppliers, and competitors all expect prices to stay at the new level, the new level becomes the baseline everyone plans around. Contracts, wage negotiations, and next year's budgets are written against it. Unwinding it would require a coordinated retreat nobody has an incentive to lead.

The exceptions that prove the mechanism

Some grocery prices genuinely do fall, and the exceptions are instructive because they're all cases where the farm share of the price is unusually large and the supply shock is temporary.

Eggs are the classic example. Egg prices have repeatedly spiked during outbreaks of highly pathogenic avian influenza — when millions of laying hens are culled, supply collapses — and then fallen substantially as flocks are rebuilt, a cycle USDA has documented across multiple outbreaks since 2015. Produce behaves similarly around droughts and freezes. In these categories, the commodity is most of the product: there's little processing, packaging, or brand markup to hold the price up once supply recovers.

Now compare a box of cornflakes. The corn in it costs pennies; the processing, box, advertising, logistics, and retail margin cost the rest. Corn prices can crater and the cornflakes won't care. The more transformed a product is between farm and shelf, the more of its price is made of sticky, never-retreating costs — and the more perfectly it ratchets.

Why falling prices aren't actually the goal

Here's the part that surprises people: the institution in charge of U.S. price stability doesn't want grocery prices to come back down — not across the board, anyway.

The Federal Reserve explicitly targets inflation of about 2% per year, on average, and has since it formalized the target in 2012. Why not 0%, or a return to 2019 prices? Because broad deflation — a falling overall price level — has a nasty feedback loop attached. If prices are falling, consumers delay purchases (why buy today what's cheaper next month?), which cuts demand, which pushes prices down further, which triggers layoffs and pay cuts, which cuts demand again. Debt makes it worse: your grocery bill would shrink, but your mortgage payment wouldn't, so debts become heavier in real terms. That spiral is a big part of the story of the Great Depression and of Japan's long stagnation after 1990, and central bankers design policy specifically to avoid it.

So when inflation surged in 2022, the Fed's goal was never to push the price level back down. It was to slow the rate of increase back toward 2% — which is what "victory over inflation" means in practice: your groceries getting more expensive at a couple percent a year, forever, on top of whatever level was reached during the surge. The Fed's main tool for engineering that slowdown is interest rates, and the transmission chain from a Fed decision to your cart is its own piece of machinery — we've mapped it start to finish in how the Fed actually moves your interest rates.

What actually offsets the ratchet

If prices don't come back down, how does the sting fade? Through the other side of the ledger: incomes. Over time, nominal wages tend to rise, and the standard measure of pain — the share of income spent on food — drifts back down as pay catches up to the new price level. USDA's long-running data on food spending shows Americans spend a historically small share of disposable income on food compared with the mid-20th century, even after the 2020s surge, though that surge did push the share up from its lows (as of the most recent USDA data, released in 2025).

That's also why "are groceries expensive?" and "is inflation high?" are genuinely different questions. Inflation can be fully back to normal while the memory of the old price level is fresh — and research on consumer sentiment consistently finds that shoppers benchmark against prices from a few years back, not last month. The gap between those benchmarks closes slowly, through raises rather than rollbacks.

The short mechanical summary

Run the ratchet one more time, end to end. A shock hits — a pandemic snarls supply chains, a war spikes energy and fertilizer costs, an outbreak culls hens. Costs surge through the 88-cent marketing chain, and retailers pass them on in visible price jumps. Then the shock fades. But by now, wages along the chain have permanently stepped up; contracts have repriced; packages have shrunk; expectations have reset. Competition doesn't force prices back down because every competitor's cost floor rose together, and the central bank is engineering a gentle 2% upward drift on purpose, because the alternative — deflation — is worse. The rate falls. The level stays. Click.

It's the same logic that governs plenty of other American prices, from airline seats (where, unusually, decades of falling real prices show what genuine competition plus falling unit costs can do) to housing, where a shortage keeps its own ratchet turning. Once you see prices as a level built from layers of sticky costs — rather than a dial that turns both ways — grocery receipts stop being mysterious. Not cheaper. Just explicable.

This explainer is written to stay accurate over time. Facts and figures were verified against the primary sources listed above as of August 22, 2026. If you spot an error, our corrections policy explains how we fix it.