Lockdown changed the grocery cart, but it didn’t last

Photo of author

| Published:

American households moved toward less processed food during pandemic lockdowns. Once restrictions were lifted, ultra-processed food spending climbed back and kept climbing, settling above where it stood before 2020.

Shelves in a grocery store display green baskets filled with assorted fresh vegetables, including yellow bell peppers, broccoli, and other produce.
Lockdown pushed American shoppers toward less processed groceries. New basket data shows how fast that reversed. Photo credit: Depositphotos.

That reversal shows up in 9,367,550 weekly grocery baskets. The record covers more than 30,000 U.S. households across six years, which makes it one of the longest continuous views of what Americans actually bought rather than what they said they ate.

What the baskets showed

A team led by Guanlan Hu of the University of Toronto and Columbia University published the analysis in Nature Human Behaviour. The researchers combined household purchase records from 2017 through 2022 with county-specific pandemic policies, then scored every product two separate ways: how heavily it had been processed, and how it rated on nutrients such as sugar, sodium and fiber. Those two scores do not measure the same thing, and keeping them separate matters.

The short-term picture matched the sourdough-era headlines. During lockdowns, households shifted money toward less processed food. The long-term picture did not. By the post-pandemic period, spending on ultra-processed products had passed its pre-pandemic baseline and stayed there.

What a less-processed cart looks like

A processing score moves when packaged items are replaced with parts they are made of. That makes it a measure of how much assembly happens at home rather than in a factory. Those swaps tend to be small and unglamorous. Carton broth becomes a slow-cooker vegetable broth made with vegetable scraps, seasoning packets become jars of individual spices and bottled dressing becomes oil and vinegar in the bottom of a bowl.

None of that changes dinner much. A household making all three swaps eats roughly the same meals, with roughly the same nutrition, from a very differently processed cart.

2 rulers give 2 answers

The gap is exactly what the study found at scale. The reversal showed up sharply on the processing measure, a system called NOVA that researchers and regulators use to sort food by degree of industrial processing. It showed up far less when the same baskets were scored on nutrients, through the Healthy Eating Index and the Food Standards Australia New Zealand profile.

The authors treat the two approaches as complementary rather than interchangeable. That distinction matters for anyone writing rules, because the two systems would flag different products on the same shelf.

The categories that stayed

Prepared and frozen foods held their gains after reopening. So did carbonated beverages. Both showed persistent effects, meaning the change outlasted the policy that set it off.

Some of the categories with staying power were convenience oriented, a pattern consistent with a lasting preference for food that arrives mostly finished. The study does not establish why those categories persisted. What did not persist was the scratch-cooking habit that briefly displaced them. Much of that habit lives in small, boring items rather than showstopper dinners, which is why the standard advice for households trying to hold the line is to start by making pantry staples at home instead of rebuilding the entire week at once.

Where the eating went instead

Grocery carts are only part of the picture. Per capita spending on food away from home grew 12% in 2023 to $4,485, while food-at-home spending rose 1.8% to $3,187, according to the USDA Economic Research Service.

The pattern held. By 2025, food away from home accounted for 56.3% of total U.S. food expenditures, with 43.7% going to groceries. The agency notes that the food-at-home share rose temporarily in 2020 and then resumed its earlier downward trend once restaurant spending recovered.

That does not mean Americans ate more restaurant meals than home-cooked ones, since restaurant food costs more per unit. It does mean the money moved, and it moved back fast.

Prices did not reset

Grocery prices rose 2.7% in the 12 months through July 2026, and restaurant prices rose 3.4%, according to the Bureau of Labor Statistics. Grocery inflation has cooled sharply, with the BLS reporting that food-at-home prices rose 11.4% in 2022 but just 2.2% over the 12 months ending in August 2026. The price level has not come down.

The return to packaged and prepared food happened inside a considerably more expensive food environment than the one households left in early 2020. That is the practical case for the dishes that sit between a frozen entree and a project: sweet potato black bean tacos built from canned beans and a roasting sheet, or a pan of shakshuka that turns eggs and canned tomatoes into dinner in half an hour. Neither asks for the empty calendar that 2020 briefly handed out.

A federal definition is still pending

The study lands in the middle of an unresolved regulatory question. On Aug. 10, 2026, the Department of Health and Human Services and the U.S. Department of Agriculture submitted the federal government’s first proposed definition of ultra-processed foods for final review at the Office of Management and Budget. No standardized federal definition has existed before now.

When the agencies opened the process in 2025, they warned that a definition drawn too broadly could sweep in whole-grain products and yogurt alongside soda and packaged snacks. The Nature findings sharpen that problem rather than solving it, since the processing-based measure and the nutrient-based measures disagreed about the same baskets.

Consumption data gives the scale of what is being defined. Between August 2021 and August 2023, Americans age 1 and older consumed an average of 55% of their calories from ultra-processed foods, the National Center for Health Statistics reported. Adults account for 53%. For children and teenagers ages 1 to 18, the figure is 61.9%.

Shruthi Baskaran-Makanju is a food and travel writer and a global food systems expert based in Seattle. She has lived in or traveled extensively to over 60 countries and shares stories and recipes inspired by those travels on Urban Farmie. 

Leave a Comment

This site uses Akismet to reduce spam. Learn how your comment data is processed.