Food systems represent a major contribution to environmental pressures, such as greenhouse gas emissions and biodiversity loss. As a result, large-scale dietary changes are needed to avoid exceeding the safe limits of the planet. Although consumers generally report favorable attitudes toward sustainable food consumption, these attitudes often do not translate into actual behavior. One factor contributing to this attitude-behavior gap is that many consumers have inaccurate or incomplete knowledge on the environmental impact of food products.

Understanding how people perceive the environmental impact of food is key to supporting dietary changes. A recent study in which British participants (n = 168) organized various supermarket products into categories based on perceived impact (using a card sorting method) revealed a novel structure to these perceptions.

Two dimensions of sustainability in the consumer's mind

Results analyses Multidimensional scaling (MDS) showed that participants create two-dimensional mental models food sustainability. They primarily distinguished products based on two main criteria:

1. Origin of animals vs. plants: This criterion was represented approximately by a diagonal from the upper left corner (plant origin) to the lower right corner (animal origin).
2. Processing level: This criterion was represented by a diagonal from the lower left corner (low processing) to the upper right corner (high processing).

According to these mental models, they were meat/dairy and highly processed products perceived as worse for the environment. Using an open-ended classification method (where participants created and named their own categories) confirmed that this dimensionality reflects deeper psychological structures.

Key areas of misconception

When comparing participants' perceptions with scientifically estimated environmental impacts (based on LCA data from Clark et al., 2022) showed that there are significant differences. Specifically, the study identified key areas where consumers are making mistakes:

• Overvaluing highly processed foods: The largest overestimation of impact was for highly processed products such as potato chips, baked chips and orange syrup. Although processing contributes to the overall impact (e.g. in the production of potato chips, processing can account for up to three quarters of GHG emissions), their overall impact is still relatively low compared to other products with which consumers categorize them (e.g. cheese, butter).
• Underestimating water load: The largest underestimations of the impact mainly concerned foods with high water consumption (so-called scarcity-weighted water use), such as various nuts, rice and almond milk. This suggests that consumers are unaware that the production of these foods contributes significantly to water stress.
• Non-differentiation of meat: Participants did not distinguish beef from other meat products when categorizing them, despite the vast differences in scientific estimates of their environmental impacts.

The impact of surprise on future purchase intentions

A positive finding was that exposure to scientific data can influence future behavior. Participants who were surprised by how high the scientifically estimated impact is of the given product, they showed intention to reduce consumption of that product in the future. Conversely, surprise at low impact led to an intention to increase consumption. This relationship was significant only for products that participants were currently purchasing.

These findings have practical implications for information strategies such as eco-labelling. As consumers perceive the impacts of processed foods and meat as incomparable, eco-labels can help facilitate comparisons across categories by providing a common metric (e.g. A–E scale). However, to maintain long-term public trust, it is necessary that additional awareness campaigns better aligned consumers' mental models with factors such as water stress and highlighted the differential impacts of red and white meat. JRi

The study was published in the journal Science