Imagine a product that is described as “natural”. It’s plant-based, locally grown in America. It’s also organic and 100 percent additive-free. It certainly sounds like a healthier choice than other brands in its category and that is, in fact, what 64 percent of its buyers believe. In reality, the natural, plant-based, locally grown and 100 percent additive-free product is…tobacco, and cigarettes made using this particular tobacco are just as dangerous as any other cigarettes.

Perception matters, and interest in healthy eating is increasing, as I have written in previous Knowledge articles based on my research. But what people and brands mean by “healthy” is changing. It’s less about checking nutrition information on products and more about selecting “real food”, “clean eating” or following some downright unhealthy ideas peddled on Instagram. 

In a recent article, I studied the categorisation of front-of-package claims and their impact on consumers' choices and their expectations of taste, healthiness and weight loss. “Healthy Through Presence or Absence, Nature or Science? A Framework for Understanding Front-of-Package Food Claims” is published by the Journal of Public Policy & Marketing, with Quentin André (INSEAD PhD ’18) of the Rotterdam School of Management and Kelly Haws of Vanderbilt University.

Healthy through presence or absence, nature or science

Earlier research had considered food claims at either the macro level (using broad descriptions such as “healthy” or “tasty”) or the micro level (studying single claims such as “low fat”). We found a middle ground solution of four clusters, based on two underlying dimensions:

1. The first dimension is claims of health based on the presence of something good or the absence of something bad.
2. The second dimension is claims of health based on either improving the food (science) or on preserving the food (nature).

The outcome is four distinct ways to be considered healthy, as shown in Figure 1. A carton of milk, for example, could be “high in vitamins” (adding positive); “low fat” (removing negatives); “no artificial growth hormone” (not adding negatives); or “all natural” (not removing positives).

Figure 1

We asked 443 people to rate 37 common claims (distilled from a list of 107) on these two dimensions (to prevent survey fatigue, each person rated eight claims). As shown in Figure 2, there was no middle ground and claims clustered well around the four types: adding (or not removing) positives and removing (or not adding) negatives.

Figure 2

Perceptions and reality

We examined the consequences of these claims, both perceived and real, in three studies of breakfast cereals, a popular food category in which food claims are common and that often has a “health halo” despite large differences in actual nutrition quality. To explore the perceived benefits of different types of food claims, we asked 363 Americans to evaluate breakfast cereal boxes with a variety of health messages.

We found that consumers had a more positive attitude towards claims that are based on positivity, the presence of something good, compared to claims that are about the absence of something bad. People expected breakfast cereals with both “adding positives” (“high protein”, “high antioxidants”, “high fiber” and “high calcium) and “not removing positives" (“all natural”, “made with whole grains”, “wholesome” or “organic”) to be healthier than brands with claims about “removing negatives” or “not adding negatives”, even if the messages claimed the absence of something considered to be harmful.

We also found that when consumers see nature-focused claims of health, like “homemade” or even “no preservatives”, they are inspired to believe the food would be tasty. Finally, if the claim is about removing negatives (“light”, “low fat”, “low calorie” or “low sugar”), consumers believe that the breakfast cereal helps with weight loss or weight maintenance.

The idea that these claims influence perceptions isn’t that surprising. After all, it’s the goal. But are consumers correct in making inferences, particularly about the healthiness of a food, based on the type of claim that is on the box? To find out, we used data from Open Food Facts, a collaborative, free and open database of food products from around the world, which has food claim and nutrition information for 633 breakfast cereals. This database also provides the Ofcom nutrient profiling score developed for the British Food Standards Agency, which is a validated measure of nutritional quality. To our surprise, the correlation between the type of “healthy" claim made and the actual nutritional quality of the breakfast cereal was almost zero (0.04, to be precise).

Predicting food choices, depending on people’s goals

Does understanding the four ways brands claim to be healthy help predict what consumers will choose? In our fourth study, we introduced three goals – healthy eating, hedonic eating or weight loss – to see how the type of claim helps predict consumers’ choices between different foods with or without food claims. With the help of PRS IN VIVO, a leader in shopper and nudge-related research, we randomly assigned 611 breakfast buyers to one of the three shopping goals and asked them to choose between foods with different claims on the label, as seen in Figure 3.

Figure 3

Compared with the breakfast cereal and milk carton without claims, all those with any claim at all had a higher probability of being chosen. More importantly, we found that goals influenced the effectiveness of the claims. When choosing a product for taste or health reasons, study participants went for the brands with the nature-based claims. When the goal was losing weight however, they selected the foods with the “removing negative” claims, like “low sugar” and “low fat”. Overall, what consumers expect from food had an impact on their choices.

What’s next for policymakers and marketers

Although there was no link between the type of claims and overall nutrition quality in our particular product category, consumers expected the type of claim to be a strong predictor of the healthiness, taste and dieting properties of the products. None of the claims we surveyed explicitly said that the product would make people healthier (or help them lose weight or stay thin), yet consumers interpreted these claims as such. And their perceptions influenced their choices.

As people are moving away from nutrition-based to nature-based approaches to “healthy” eating, so are food claims. The danger is that claims are receding from nutrition, based on science, towards beliefs based on folk theories or, worse, the opinions of celebrities. This should motivate legislators to consider stricter regulation of food claims.

Now, the first principle in the regulation of marketing claims is to ensure that they are not incorrect. But this is not enough. After all, the tobacco product mentioned at the beginning of this article really is organic and 100 percent additive-free. The problem is the misleading interpretation that this makes it healthier. And this is why in 2010 the Federal Trade Commission forced the company to include the disclaimer that “no additives in tobacco does NOT mean a safer cigarette” and why, in 2015, the FDA took action to force it to remove the "natural” and “additive-free” claims altogether.

There is no reason why food should be treated differently. Obviously, claims must be accurate, but they should also not be misleading in such a way that consumers might expect benefits that the food cannot deliver. Rather than basing their regulations on what the mythical “rational consumer” would think, policymakers should take into account what real consumers do expect from products with health claims.

The association we have between nature and health is often true. Michael Pollan’s advice of “Eat food. Not too much. Mostly plants.” is sound. Ironically however, the only foods that are universally accepted as healthy are fruits, vegetables and some nuts – that is, the few foods without a claim on the label. Because they don’t have a label. Nature, as most consumers understand, knows best.