Picture yourself walking into a coffee shop on a busy morning. The menu board lists your usual order, and beside the price sits a small number telling you how many calories it contains. Does seeing that figure change what you buy? Governments are betting that it does. In 2022, the United Kingdom required large cafes and restaurants to print calorie counts on their menus, hoping the information would nudge people toward lighter choices.
A study published in Appetite set out to test how that nudge actually works, and whether it changes behavior at all. The short version: calorie labels shifted what people believed about menu items and improved how accurately they could guess calorie content, but they did not significantly change what people chose to order.
The question behind the policy
Calorie labeling has been studied many times, with mixed results. Some reviews report modest reductions in calories ordered, somewhere between 11 and 50 calories per purchase. Others find little or no effect. What has been missing, according to the researchers, is an explanation of why labels might work when they do. Knowing whether a policy is effective is one thing. Understanding the steps in between, from reading a label to placing an order, is another.
Khaleda Ahmadyar of City St George’s, University of London, along with colleagues Katy Tapper from the same institution and Eric Robinson of the University of Liverpool, focused on coffee shops specifically. That setting has received less attention than full restaurants, even though many hot drinks carry a surprising number of calories. Liquid calories also tend to leave people feeling less full than calories from food, which means a daily latte habit can quietly add up.
The team drew on a framework from psychology called the expectancy-value approach. The basic idea is that people change their behavior when two things happen together: their expectations about an outcome shift, and they actually care about that outcome. Applied here, a calorie label might make someone see a drink as less healthy, and if that person values eating healthily, they might then choose something lighter. The study was built to test each link in that chain.
How the study worked
The researchers recruited adults through Prolific, an online research platform, using quotas for age, sex, and ethnicity so the sample broadly matched the UK population. After exclusions for failed attention checks, 577 participants remained, split almost evenly between two groups.
Everyone saw a menu board listing ten items: seven drinks (three coffees, tea, two hot chocolates, and water) and three food items (a chocolate brownie, a blueberry muffin, and a banana), with prices based on a well-known coffee chain. One group saw calorie counts printed next to the prices. The other did not. Participants were asked to imagine walking into a coffee shop and to click on the items they would buy for themselves.
Before and after viewing the menu, people rated each item on five expectations: whether it would taste good, leave them full, offer good value for money, make them feel like they were eating healthily, and help them avoid gaining weight. Participants also reported how much they personally cared about health, weight control, price, and taste in their everyday food choices, along with their current hunger level. Afterward, they estimated the calorie content of each item and explained, in their own words, why they chose what they did.
The main measure was simple: the total calories in the items each person selected.
What the analysis revealed
On the central question, the labels fell short of statistical significance. People who saw calorie counts selected an average of 371 calories, compared with 392 calories for those who did not, a difference of about 5 percent. The direction matched the prediction, and the size of the gap was in line with what earlier reviews have reported, but the result did not reach the threshold that would let the researchers rule out chance (p = 0.18).
The labels did, however, change what people believed. Seeing calorie counts produced larger shifts in expectations about weight control, health, and value for money. The pattern was specific and revealing. For low-calorie drinks like tea and americano, where people tended to overestimate the calories, beliefs about healthiness improved once the true numbers appeared. For high-calorie items like lattes and hot chocolates, where people tended to underestimate the calories, beliefs grew more negative. This shift did not happen in the group without labels.
Here is where the expected chain broke. Even though labels changed these beliefs, the changed beliefs did not translate into different orders. People who valued health and weight control did not select lower-calorie items as a result of their shifted expectations. The study, in other words, did not find evidence that calorie labels work by changing beliefs and then changing behavior, at least not in this single decision.
The authors suggest a few reasons. One is the well-documented gap between intentions and actions. Another is that ordering a coffee is a quick, low-stakes, habitual decision, the kind people often make on autopilot rather than through deliberate reasoning. As they note, behavior may still have been “primarily driven by more automatic, habitual processes” despite the change in beliefs.
The role of personal motivation
The team also asked whether labels work better for people who are already trying to manage their weight. The evidence here was mixed and should be read with caution. In one version of the statistical analysis, calorie labels were linked to lower-calorie choices among people with high weight control motivation, with no effect for those less motivated. But this result did not hold up across the other analysis models, and it disappeared when the researchers removed participants who had guessed the study’s purpose.
Some support came from participants’ own explanations. When describing their choices, 12 percent of people in the labeling group mentioned weight or calories, compared with just 2 percent in the no-label group. Still, taste was by far the most common reason for any choice, cited by roughly 70 percent of participants in both groups.
Where labels clearly made a difference
The clearest effect appeared in knowledge. People who saw the calorie counts were significantly more accurate when later estimating how many calories each item contained. Across both groups, people tended to overestimate low-calorie items and underestimate high-calorie ones, but the labels narrowed that error. Younger participants, those dieting to lose weight, women, and people with higher health motivation also tended to estimate more accurately.
The researchers point to this as a meaningful result. Even if labels do not change a single coffee order in the moment, they argue, better knowledge of calorie content might influence eating decisions later in the day or in other settings over time.
Caveats worth keeping in mind
This was a hypothetical exercise, not a real purchase. People clicked on a screen rather than spending money or drinking anything, which limits how far the findings stretch to actual coffee shop behavior. The researchers acknowledge this and suggest that field experiments in real cafes, or tests built into mobile ordering apps, would offer stronger evidence.
The setup itself may also have shaped the results. Participants rated items on multiple dimensions before choosing, which could have nudged them to think about considerations they might otherwise have ignored. And roughly 19 percent of participants correctly guessed what the study was investigating, which can color responses. The measure of personal motivation captured what mattered to people “on a typical day,” which may not match what was driving them during the specific task.
Taken together, the study suggests that calorie labels on coffee shop menus may do a better job of informing people than of changing their immediate choices. As the authors put it, “while calorie labelling may influence beliefs and knowledge, its acute impact on population level behaviour may be minimal.” Whether that improved knowledge eventually shows up as different habits remains an open question for future research.
