How many calories are in a medium-size muffin? I’ll give you a minute to think.
The answer is about 520 — probably more than many people think. In contrast, a handful of almonds has 160 calories. To put this in perspective: replacing a morning muffin snack with a handful of nuts every day for a year would cut 131,400 calories, or about 35 pounds.
Despite these large differences — and despite the facts that a third of U.S. adults are obese and an estimated 63 percent of obese individuals are trying to lose weight — most people do not know much about how many calories they are eating. In one study of consumers in New York City and Newark, New Jersey, researchers found that consumers at fast-food chains underestimated the number of calories in the meal they purchased by about 200 calories. Only 15 percent of people surveyed stated an estimate within 100 calories of the content of their meal.
Notably, however, people who do pay attention to calories tend to buy fewer of them. (It’s possible these people do not actually eat less; all the studies I’ll discuss here focus on how many calories people buy, not how many they eat.) A highly cited study from 2008 in the American Journal of Public Health compared customers of the sandwich chain Subway who said they paid attention to calorie labels to those who said they did not. The study found that those who paid attention bought 52 fewer calories on average; that’s a little more than a third of a 12-ounce soda. It’s not a huge number, but if someone made this change every day, they could lose up to 5 pounds in a year.
With this backdrop, it’s not surprising that policymakers have focused on calorie labeling at restaurants as a major policy initiative. Last week, the Food and Drug Administration announced sweeping rules that will require extensive calorie labeling nationwide. This includes on menus at chain restaurants, movie theaters, bakeries, coffee shops and vending machines. This change has been hailed by legislators and food-policy experts as a landmark in fighting obesity.
But will it work? The evidence I pulled together above isn’t conclusive. The fact that people do not know the calorie contents of their food could just be because they don’t care. And that people who do pay attention to calories eat fewer of them is hardly evidence that forcing this information on everyone will change average behavior.
The best evidence to answer this question comes not from a compilation of circumstantial facts, but from evaluations of food-labeling programs. The nationwide FDA rules are new, but New York City and a few other cities and counties have had labeling rules in place in chain restaurants for several years. This enables researchers to look at the simple question of whether telling people the calorie content of meals reduces their caloric intake.
In most studies, the answer is no. In some that focused on low-income individuals and chain restaurants, there was no evidence of behavior change in response to calorie labeling. However, there are a few exceptions, and those seem to highlight differences among particular demographic groups. One study, run in 2008, surveyed 1,156 individuals eating at chain restaurants. The study compared the changes in calories purchased in New York City after calorie labeling was introduced to changes in Newark, where no such labeling was introduced. Thirty percent of people surveyed in New York reported that labeling influenced their food choices. This might seem encouraging, but there was no change in calories purchased.
A similar result shows up in data from King County, Washington, which instituted a calorie-labeling program in 2009. Researchers compared the change in calories purchased at Taco Time restaurants in King County before and after the labeling program to changes over a similar period in surrounding areas without the labeling. The Taco Time calorie labels had no impact on calories purchased.
This evidence is supported by randomized laboratory experiments. In one study, participants who were randomly given a menu with calorie counts displayed did not order fewer calories than those given a menu without such information.
A meta-analysis from 2011 concluded: “The current evidence suggests that calorie labeling does not have the intended effect of decreasing calorie purchasing or consumption.”
There are, however, at least two recent studies that contradict this. One, by three economists, looked at the effect of calorie labeling in Starbucks stores. The authors found that after the labeling, calories per transaction fell by 6 percent. This change was entirely due to changes in food purchases, not beverage purchases. In other words, knowing that a 16-ounce, whole milk Caramel Brulee Frappucino has 430 calories didn’t change whether people wanted one, but learning that there were 420 calories in a blueberry scone did decrease demand.
This study is encouraging, although it’s subject to the concern that people may be more willing to limit their food purchases at Starbucks because food is not the major reason to visit the coffee chain. This tendency could limit the ability to generalize these results to restaurants, fast food or otherwise.
A second study, from Philadelphia, looked at purchasing behavior in a set of full-service chain restaurants. Researchers compared outlets with calorie labeling to those without, and found that consumers in the calorie-labeling outlets purchased 155 fewer calories than those at the non-labeling outlets. A notable feature of this study is that when the researchers asked consumers whether the nutrient information affected their orders, those most likely to say yes were people with higher incomes and more education. Forty-nine percent of consumers with a graduate degree said they were affected, compared to just 9 percent with less than a high school degree.
This last result may give some clue as to why this study differs from the previous ones. Fast-food chains attract, on average, lower-income consumers. If that group is less likely to respond to calorie labels, it could explain why we see no impact of calorie labeling in those restaurants but do see some impact in chains typically frequented by higher-income individuals, such as full-service restaurants and Starbucks.
What does this mean for the success of the FDA labeling effort? It seems plausible that some subset of individuals at some restaurant locations will decrease their calorie intake. In fact, this seems almost certain. Every time I bring up calorie labeling with people I work with, someone will describe how the practice changed their behavior. One colleague realized that her habit of purchasing muffins rather than croissants to save calories was a waste — croissants have fewer calories than muffins.
On the other hand, if the idea is that these changes will drastically affect obesity in low-income populations eating at fast-food chains, the evidence doesn’t support this hope.
A supporter of calorie labeling will ask, “What is the downside?” Even if only a small fraction of people respond by decreasing their calories, there is no loss to those who did not change. So, why worry about it? This ignores the obvious cost to restaurants of calculating the calories in their products and of changing their menus. A grocery-store industry trade group estimated the cost of compliance with calorie-labeling rules could initially amount to $1 billion. The change also ignores the possible psychic costs to people.
Even if I am still going to eat a muffin after I learn that there are 520 calories in it, I may feel worse about doing so. If the actual impact of calorie labeling is to encourage only a few people to eat fewer calories but to make many more people feel worse about themselves, it seems less than obvious that it is a welfare-improving idea.