How do we improve population diets? There is now general agreement that education alone can't achieve this goal. Instead, we need to look at more structural interventions that support healthier food choices by creating a healthier food environment—i.e. making the healthy choice the easy choice. Food pricing and food labeling are two structural interventions that are quickly gaining popularity.
The Question Is: Do these Strategies Work?
The interesting thing when you are a health scientist (like me) is that most people have an opinion about the effectiveness of public health strategies, such as food labeling and pricing. This makes sense; most people buy food regularly and therefore have ideas about how people (themselves!) can be encouraged to choose healthier food. However, as we know, there is quite a difference between what people think will work to make them choose healthier and what actually makes them choose healthier. Luckily there is science to help us out.
The gold-standard method to measure effectiveness of interventions is randomized controlled trials (RCTs). In this design you give one group the intervention (say a new food label) and the other group a placebo (say the current nutrient information panel) and measure food purchases in both groups. Systematic reviews and meta-analysis combine the results from multiple RCTs to give an overview of all existing evidence. So, what does this evidence look like for food pricing and food labeling?
First, food labeling: the two most common types of labeling are calorie labeling in fast food restaurants and front of pack labeling (FOP) of supermarket products. Since 2010, all U.S. restaurants and food outlets with 20 or more locations are required to list calorie content for standard menu items. Does this work? Evidence to date suggests it doesn't. A systematic review including outcomes from studies comparing calorie labeling versus no labeling concluded that its impact was modest at best. Likewise, a recent meta-analysis concluded that labeling menus with calorie information alone had no effects on consumed calories. One potential explanation for this lack of effect is that misinterpretation persists, despite the calorie label. Many menu items appear with energy ranges instead of specific energy values because items are customizable. For example, total energy will differ for sandwiches depending on the selected fillings. These energy ranges might confuse consumers, leading them to misestimate the total energy value of their meal. Also, consumers might confuse the energy range with healthiness, believing that the item with the lowest energy content is the healthiest instead of the item with the least ingredients. Finally, consumers who are not health literate might still have problems interpreting the calorie information when it comes with no further explanation. The meta-analysis by Sinclair et al suggests that this can be improved if additional interpretive information is added, such as the daily recommended calorie intake.
So what about FOP labeling on supermarket foods? There is now considerable evidence showing that consumers understand interpretative labels (using graphics, symbols, or colors) much better than quantitative (text) nutrition labels. However, evidence of the effects on actual purchases, sales, or consumption is still inconclusive. There are only a handful of studies that have looked into these effects, and none of them found strong evidence for the effectiveness of FOP labeling.
Conclusively, we know that food labels alone cannot solve the obesity epidemic. The food industry spends billions of dollars to promote their products—it would be surprising if one simple label could overturn these effects. However, this doesn't mean we shouldn't introduce food labeling. Quite the contrary, clear food labeling is arguably the very minimum we should do to support healthier food choices. Another good reason to support food labeling is that it might incentivize producers to develop healthier products. Nevertheless, it is important to realize that food labeling is not the egg of Columbus, and additional strategies are needed to curb the obesity epidemic.
One strategy with much potential to enhance the effects of labeling is food pricing. The reasoning behind this is that unhealthy (calorie-rich) foods are relatively cheaper than fruits and vegetables. Also, there is evidence showing that price (alongside taste) is the main determinant in food choice and that high prices form a barrier for lower income consumers to buy healthy food.
There are different types of food pricing strategies to consider. Most importantly, we differentiate subsidies on healthier foods (for example a fruit and vegetable subsidy) or taxes on unhealthier foods (for example a fat tax). It is important to distinguish health-related food taxes from other food or beverage taxes. For example, many American states impose sales taxes on soft drinks. However, these taxes are generally too low to create an effect on consumption and are implemented for revenue purposes, not health.
The basis of food pricing strategies is the first economic law of demand: if you increase the price of a product, purchases will decrease and vice versa. This effect is known as own-price elasticity. There is now strong evidence showing that food taxes and subsidies are indeed effective in changing purchases of the taxed product. For example, a recent randomized controlled trial showed that a 13 percent tax on soft drinks was effective in reducing soft drink purchases. Also, two RCTs have shown that subsidies on fruits and vegetables encourage fruit and vegetable purchases significantly. The most recent systematic review on food taxes and subsidies concludes that the majority of studies show pricing strategies to be effective in increasing purchases of healthier foods and reducing purchases of foods high in sugar, fat, sodium, and calories. The authors state that the tax or subsidy should at least be 10-15 percent to effectively change diets and ultimately decrease the burden of non-communicable disease. Also, the highest effects are observed when taxes and subsidies are used in tandem.
One of the biggest remaining concerns about taxes and subsidies, outside being potentially regressive, are cross-price elasticity effects. Cross-price elasticity refers to the effect from price changes on purchases of products that are not directly impacted by the tax. For example, the consumption of sugary foods might change following the introduction of a fat tax. Or the consumption of diet drinks might change following the introduction of a soft drink tax. These effects are often much harder to measure, because you need to keep track of everything people buy and be able to separate people affected and not affected by the tax. However, high-quality evidence on these effects is crucial because substitution effects can change the overall effectiveness of the tax. For example, experiments on healthy food subsidies have shown that people tend to buy the subsidized products on top of what they normally buy (as opposed to instead of) thereby increasing their total calorie purchases. While extra fruit and vegetable consumption is a favorable outcome, additional calorie consumption is unfavorable in light of the obesity epidemic.
The Danish Example
Another complicating factor when considering food pricing strategies is the implementation process. Take the Danish example. In 2011, Denmark was the first country to introduce a food tax on saturated fat. The tax was abolished within one year of its introduction. Opponents were very quick to say that this was proof of the tax not working. However, more recent evidence shows that the Danish fat tax was in fact effective in improving health outcomes. An analysis of parliamentary debates, expert reports, and media coverage; key informant interviews; and a review of studies about the effects of the tax on consumer behavior revealed that the reasons for abolishment didn't have anything to do with the tax's effects on health. The evidence indicated that the tax wasn't well designed for health outcomes in the first place, and was introduced mainly to increase public revenue. The tax also suffered from having no strong proponents and many influential adversaries.
Finally, it is important to realize that the price changes imposed by the tax or subsidy might not be fully passed on to the consumer. Companies can make decisions to absorb parts of the tax and thereby limit the impacts of the price change. For example, soft drink companies might choose to increase the price of both regular and diet soft drinks following a sugar sweetened beverage drink tax to balance the price increase throughout their product range. Companies can also find ways to circumvent the tax altogether. A retrospective example of this comes from the 1980s, when U.S. farm policies made sugar more expensive, expecting that this would result in higher soft drink prices and subsequent decreased consumption. However, the Coca-Cola Company started using high-fructose corn syrup as a substitute for sugar, saving them around $25 million per year and thus making Coca Cola both cheaper and more profitable.
If we really want to address obesity and non-communicable disease, the sales of some key food groups (fast food, soft drinks, sugar, etc.) have to drop. This does not sit well with our commercialized food system, a multi-billion dollar industry which many national economies rely on to grow their food exports. When so much effort is put into selling food, it can hardly be a surprise that people become obese. My colleague Boyd Swinburn once claimed in a landmark Lancet paper that obesity is in fact a robust sign of commercial success.
Food labeling and food pricing do not do much to change this food system; they only address the symptoms. To really make a difference to health (and environmental sustainability), we have to change the system itself. If we do it well, industry will have the incentives to change the business model from selling many cheap calories to selling real food sustainably. This way, we no longer have to educate consumers about choosing healthier foods; healthier food will be the most attractive alternative to begin with.