Adapting the universal signals of traffic lights to food packaging
David Leonhardt’s “Budgets Behaving Badly” column from yesterday’s New York Times is a must-read for Nudge enthusiasts interested in how behavioral economics can be applied to public policy. One of the ideas in the column, also discussed in Nudge, is the redesign of fuel economy stickers. Another potential area for redesigned labels is food packaging, which one Nudge reader thought could benefit from a traffic light theme.
The Cancer Council of New South Wales in Australia is currently pushing the traffic light idea to help people make better nutritional decision making, following some field tests and surveys involving breakfast cereal, crispbread, and lasagna. (Hat tip to Stephen Laniel for pointing us to the study, which is here.)
Researchers created a simple traffic light system ranking for levels of total fat, saturated fat, sugar and sodium. They compared these labels against other more detailed and complicated labels, which provided specific information on the percentage daily intake of bad items like fat and sodium, and good items like protein and fiber found in each food item. Some of the daily intake labels used a traffic light color coding as well. The various labels are shown below.
Each label was attached to two versions of each food product, one healthy and one unhealthy. (People weren’t told which.) After observing the food package, consumers were asked to identify the healthy product. In the end, simpler turned out to be better.
Consumers using the Traffic Light system were five times more likely to correctly identify healthier food products compared to the Monochrome %DI system, and three times more likely to correctly identify the healthier products compared to the Colour-Coded %DI system. There were no significant differences in consumers’ ability to use the Traffic Light and the Traffic Light + Overall Rating systems, therefore including this additional overall rating information on front-of-pack labels may not be necessary.
Consumers reported that they could compare the healthiness of food products the fastest when using the Traffic Light and the Traffic Light + Overall Rating labelling systems, with significantly more consumers reporting they could make product comparisons using these systems at a glance, compared to consumers using the Monochrome %DI and Colour-Coded %DI systems.
What is most interesting about the survey is that consumers initially thought the traffic light labeling would be less useful than the daily intake percentage labels. Before seeing the products, just 14 percent thought the traffic light label would be the most helpful, compared to 41 percent for the color-coded daily intake percentage.
In other words, consumers assumed they would benefit from an overload of information and graphics. But in the end, as behavioral economists know all too well, the exact opposite was the case. A preference for the complicated label may be another example of overconfidence, in which people think they can process more information than they actually can.
The traffic light labels were not superior on all fronts, however. They helped consumers choose the healthy lasagna, but not the healthy breakfast cereal, leading the study’s authors to suggest that “preconceived ideas and prior association(s)” with food can play a role in decision making.
Thus, more consumers rated the breakfast cereal as healthy using the Monochrome %DI and Colour-Coded %DI labels, as consumers might generally consider breakfast cereals to be relatively healthy. In contrast, lasagna might be generally thought of as unhealthy.