Research Insights on Front of Package Labeling
By Laura Stortz
The current federal government has proposed mandatory front-of-package nutrition warning labels to deter purchases of prepackaged foods ‘high in’ sugar, sodium, and/or saturated fat. The labels are part of the federal government’s Healthy Eating Strategy to combat high obesity levels in the population (67% of adults have weight categorized as overweight or obese). Health Canada estimates that the annual economic burden of unhealthy eating amounts to $6.6 billion dollars annually.
Many lobby groups have been opposing the labels, as they are projected to cost upwards of $2 billion to adopt and many producer groups believe that the labels send the wrong message. The most vocal to resist the labels include the dairy industry, who believe that all dairy products should be exempt not just those without added sugars. While there is opposition to how strict the labels are, some believe that they could be harsher. The federal government has completed a consultation process which highlighted the two lobbying group extremes of the issue: nutrition groups who want harsher, more explicit labels and food manufacturers who want less focus on warnings and more on information.
Despite the visceral reaction of both sides, Health Canada has decided to move forward with labels. My Masters thesis was part of a project funded by the Ontario Ministry of Agriculture, Food, and Rural Affairs to look at the impact of these labels. No other study has been able to test the effectiveness of the labels in a grocery store setting with a full range of products. We had 457 products in the store and about 33% of them qualified for the label—which varied based on the category.
When the products were labeled, both the number of ‘high in’ items (-12%) and the proportion of 'high in' items (-6%) per basket were reduced. The treatment group had a reduced likelihood of purchasing 'high in' items overall.
Previous studies have highlighted that warning labels are more likely to effect choice in categories where nutritional content is ambiguous. Our findings confirm this pattern. The label decreased likelihood of purchasing ‘high in’ juice (-43.5%), tomato sauce (-12%), and chocolate milk (-50%). People may perceive juice to be healthy but it seems the warning labels correct these assumptions and decrease the likelihood of purchasing juice with a high in label. Importantly, the label especially affected participants with children.
The dairy industry is very worried about the effect on the labels on demand for dairy products. We did not find any negative effect on likelihood of purchasing ‘high in’ yogurt in the treatment group, we even found that reading nutrition facts tables increased the likelihood of purchasing high-in yogurt in the treatment group. The industry’s fears about the effect of the labels on the likelihood of purchasing chocolate milk were confirmed, we saw a marked 50% decrease in the purchasing likelihood.
One feature of the study is that we specifically wanted representation from marginalized groups. We did not find that the label made those who are most at risk of being obese (those with lower socio-economic status) less likely to purchase ‘high in’ items. Contrary to expectations, we did not find that those who had less education or low income had increased likelihood of purchasing high-in items. However, the average grams per serving-size of sugar (-36%) and sodium (-27%) per basket were reduced for low-income participants in the treatment group. It seems that participants seem to think of saturated fat differently than other nutrients.
The warning labels are supposed to decrease barriers to accessing nutrition information by attracting attention and providing clear, simple information. We found that the labels were more effective for those who reported that they always or mostly look at nutrition facts tables, which we used as a proxy for interest in nutrition information. However, we did not find that they seemed to make a difference for those who reported never using NFTs, which may show that they do not reduce information costs for these groups.
Overall, warning labels seem to do what they are intended to do but this is not without caveats. For potato chips, time spent observing warning labels actually increased likelihood of purchasing labelled items. While this is certainly an unintended consequence of the labels, it makes sense on some level. Other studies have found that people may use warning labels as a signal of how tasty an item will be. I like to use the example of ice cream—if I am buying ice cream, I am certainly not going to search for ice milk, in fact, I may search for ice cream touting ‘100% Canadian Cream’.
This type of labelling may significantly reduce healthcare costs if Canadians respond to the labels over time. The ambiguous categories showed the greatest change in result to the label. Warning labels were proposed to first, decrease consumption of ‘high in’ foods and induce food manufacturers to reformulate. Reformulation may be achieved through manufacturers wanting to avoid costs of the adopting the label or decreased demand. If consumers react to the labels by switching to unlabelled items or dropping out of the category, food manufacturers may reformulate… but if the labels increase attractiveness then reformulation would be against their best interests.
Recommended citation format: Stortz, Laura. "Research Insights on Front of Package Labeling". Food Focus Guelph (90), Department of Food, Agricultural and Resource Economics, University of Guelph, June 15, 2020.