Explainer

This Pint of Ben & Jerry’s Is Four Servings?

How do food companies determine “serving size”?

Is serving size a useful metric?

The food at many chain restaurants and supermarkets contains more calories than indicated on nutrition labels, according to a study published in this month’s Journal of the American Dietetic Association. This discrepancy is due in part to the fact that actual portions are often bigger than the serving size indicated on the label—for example, a McDonald’s employee might put more mayo on a burger than he’s supposed to. How do food companies determine “serving size,” anyway?

With old data. The Food and Drug Administration provides an exhaustively detailed set of guidelines to help manufacturers set portion size based on the amount the average person is likely to eat at once—or, in jargon, the “reference amount customarily consumed.” The agency arrives at this figure—always expressed in grams—by factoring in survey data, recommendations from food and nutrition organizations, and customs in other countries. One reason official serving sizes seem so small is that the survey data comes largely from studies conducted by the U.S. Department of Agriculture in the 1970s and ‘80s, when Americans consumed less food than they do now. In 2005, the FDA began the slow process of updating the rules to reflect how we eat now, but the agency is far from finished.

The simplest way to arrive at the correct serving size of a food is to convert the RACC from grams into a household measurement like teaspoons, tablespoons, or cups. A serving of sugar, for example, is the teaspoon equivalent of 4 grams. For slices of bread, cookies, and other foods that come in discrete units, the serving size is expressed as the number that comes closest to the recommended amount. Cookies, for example, have an RACC of 30 grams. If each cookie weighs 18 grams, the serving size would be two cookies, or 36 grams. (Companies aren’t expected to change the size of their cookies to fit the RACC exactly.) If a product comes in a “single-serving container”—individually sold products that contain less than 200 percent of the RACC—companies can simply list a serving size as “1 container” or “1 package” or “1 can.” Things get more complicated with food that requires adding water (like condensed milk) or food that gets mixed with other ingredients (like cake mix), since the label will often list both the nutrition facts “as packaged” and “as prepared.” There are also rules for rounding (if a package contains between two and five servings, round to the nearest half serving, but if it contains more than five servings, round to the nearest serving) as well as rules for measuring different types of pasta, like macaroni versus lasagna. (See here for a good serving size FAQ.)

There’s still room for fudging. For simplicity’s sake, the FDA encourages manufacturers to label a container as a single-serving “if the entire contents of the package can reasonably be consumed at a single-eating occasion”—even if it’s more than twice the RACC. For example, someone who buys a 20 oz. soda bottle is probably going to drink it all at once, even though the RACC is 8 fluid ounces. It’s more important for the food label to reflect actual consumption habits, the thinking goes, than to adhere strictly to the technical guidelines. Nevertheless, manufacturers insist on labeling soda bottles “2.5 servings.”

Food manufacturers undergo regular inspection by the FDA, which flags a product if its serving size is mislabeled. For example, the agency recently sent a warning letter to a manufacturer of pierogi for, among other violations, calling its serving size “5 blintzes” without defining blintz. Companies that don’t comply with warning letters can be subject to fines, seizure of their product, or criminal prosecution.

Got a question about today’s news? Ask the Explainer.

Explainer thanks Siobhan DeLancey of the Food and Drug Administration and Marion Nestle of New York University.

Become a fan of the Explainer on Facebook.