Legal Strategies in Childhood Obesity Prevention: Workshop Summary (2011)

5

Using Regulations and Taxes to Prevent Obesity

Regulations on the sale and advertising of foods and beverages and taxes on particular foods and beverages have the potential to promote healthy eating. Two speakers at the workshop examined these broad approaches to childhood obesity prevention. Although regulations and taxes are inevitably subject to limitations, these policy options can be applied in a surprisingly

large number of ways to change the foods and beverages people purchase and consume.

THE USE OF REGULATIONS TO CHANGE FOOD AND BEVERAGE CONSUMPTION

In the United States today, the easy choice is not the healthy choice, said Jennifer Pomeranz, director of legal initiatives, Yale University’s Rudd Center for Food Policy and Obesity. In supermarkets, for example, shoppers generally must pass through the store to reach the necessities located toward the back. Food marketers refer to end-of-aisle bulk displays as ground zero for in-store decision making because shoppers must pass by them to enter an aisle. Moreover, checkout aisles are lined with unhealthy food products that shoppers are encouraged to purchase on impulse while waiting to check out. According to a report from the Grocery Manufacturers Association (GMA), 73 percent of shoppers make impulse purchases in grocery stores, most often of candy, cookies, and chips (GMA Sales Committee, 2009). A retail report suggests that shoppers can avoid consuming thousands of calories a year by going through self-checkout aisles that lack displays of candy and other items (Schuman, 2006).

Shelf position is also big business in the retail environment. Companies pay large sums of money to have their products in certain locations. A study from the University of Chicago Graduate School of Business showed that shelves in the vertical center, which is the natural resting eye position for adult shoppers, are best for default purchases (Dreze et al., 1994). When products—whether canned tuna, cereal, or toilet paper—were moved to different shelf locations, the items on the vertically central shelves were purchased the most. One former retailer described product location as “eye level equals buy level.” High shelves, in contrast, are reserved for slow-moving products and products sought by destination shoppers (Cardello, 2009).

Retail outlets also rely heavily on point-of-purchase and package promotions using spokespeople, licensed characters, celebrities, games, graphics, premiums, and movie tie-ins. According to a Federal Trade Commission (FTC) study, expenditures on these promotions are second only to those on television advertising (FTC, 2006). Furthermore, Pomeranz noted that a Yale study conducted from 2006 to 2008 found that 79 percent of all such food promotions are targeted to children under the age of 12 (Harris et al., 2009). More than 80 percent of the food promoted was found to be unhealthy, and over the 3 years of the study, the nutritional quality of the promoted foods declined, despite pledges made by the food and beverage industries as part of the Children’s Food and Beverage Advertising Initiative (described in Chapter 4).

Another study of “fun food” packaging, which excluded snack items,

found that 89 percent of food in supermarkets packaged in such a way was unhealthy (Elliot, 2008). In its study of food marketing, the Institute of Medicine (IOM) (2006) found that children as young as 2 to 3 years of age recognize food packaging. Children prefer the taste of food that has recognizable characters on the packaging, even when the food is identical to samples without that packaging.

Pomeranz spent much of her talk analyzing what is legally permitted and not permitted in food and beverage sales and marketing. The First Amendment to the U.S. Constitution protects speech. The U.S. Supreme Court has applied the First Amendment to commercial speech, defined as speech, such as advertising, that promotes a commercial transaction. The First Amendment also protects expressive conduct, which consists of action that has a communicative component; an example is flag burning to protest war. Pure conduct (lacking a communicative component) is not protected by the First Amendment.

The 2002 case Lorillard v. Reilly involved a Massachusetts regulation that banned advertisements for tobacco within 1,000 feet of schools and playgrounds on billboards and less than 5 feet from the floor inside stores. The state also banned self-service displays of tobacco, requiring that tobacco products be held behind the counter so that only salespersons could access them. The Supreme Court struck down the first two advertising restrictions, ruling that they were unconstitutional under the legal standard applied to commercial speech restriction, called the Central Hudson test.¹

However, the Supreme Court upheld the ban on self-service displays of tobacco. This regulation was analyzed under a different standard, known as the O’Brien test, which is the test for expressive conduct.² The lesson learned from this case, said Pomeranz, “is that regulating conduct—and, in this context, sales practices—is achievable, and [often] is preferable.”

The O’Brien test asks whether a restriction is within the constitutional

____________

¹ Originating in the 1980 U.S. Supreme Court case Central Hudson Gas & Elec. Corp. v. Public Serv. Comm’n of N.Y., the Central Hudson test is used by courts to determine whether government restrictions on commercial speech are constitutional. The court first must determine whether the expression is protected by the First Amendment, meaning it must concern lawful activity and not be misleading or deceptive. Second, the court must ask whether the asserted government interest is substantial. Third, the court must determine whether the restriction directly advances the government interest asserted, and fourth, it must determine whether the restriction is more extensive than is necessary to serve that interest.

² Originating in the 1968 U.S. Supreme Court case United States v. O’Brien, the O’Brien test is used to determine whether a government restriction on expressive conduct is constitutional. This test asks whether the government restriction is within the constitutional power of the government, whether it furthers an important or substantial governmental interest, whether that interest is unrelated to the suppression of free expression, and whether the incidental restriction of alleged First Amendment freedoms is no greater than is essential to further that interest.

powers of the government, and states and localities have great leeway to impose regulations for public health and safety purposes, Pomeranz observed. The O’Brien test also asks whether a regulation furthers a substantial interest, and the Supreme Court has ruled that governments must have leeway to experiment with solutions that can further such interests. Regulations cannot target expression, but in the case of foods, they can target the nutritional profile of a product. “We know that the foods marketed to children are so unhealthy that all we have to do is target [their] very poor nutritional profiles … and we will be addressing a big problem within our retail environment,” Pomeranz suggested. Also, any incidental restriction on speech caused by a regulation cannot be beyond what is essential to further the substantial government interest, but this will not be a problem if calorie-dense and nutrient-poor food products are targeted, Pomeranz said.

In the past, the Supreme Court has struck down restrictions on speech aimed at protecting the public from harmful products, such as tobacco and alcohol. But the Supreme Court also has offered many ideas regarding what the government can do to regulate conduct. These ideas include increasing the price of a product, limiting per capita purchases of a product, banning or limiting harmful products or ingredients, and imposing an age limit on the sale of a product. In practice, all of these actions have yielded important benefits, Pomeranz observed. Taxes on cigarettes have made people less likely to purchase tobacco. Limiting per capita purchases of ephedrine-containing products has reduced illicit uses of these products. Placing tobacco and ephedrine-containing products behind the counter, which many states have done, has helped control consumption. And minors are not allowed to buy alcohol or tobacco or enter gaming institutions.

Pomeranz described how these approaches might be taken with sugar-sweetened beverages. Tax increases (discussed in the next section of this chapter) could be used to limit purchases and consumption. Access could be restricted, as has occurred in schools with the removal of products from vending machines. Access also could be restricted near schools or in other locations. Age limits could be imposed on the purchase, possession, or use of sugar-sweetened beverages.

Such approaches could work in the retail environment as well. Junk-food-free checkout aisles could reduce impulse purchases. Specially marked aisles could contain all of the unhealthy foods in the market, so that parents could avoid those aisles. End-of-aisle displays and vertically central shelves could contain healthy rather than unhealthy foods to make the default choice the healthy choice. Location criteria could be based on nutritional requirements rather than industry payments, with unhealthy items being located at the tops of shelves. Food samples to minors could be banned. Pomeranz observed that practices that make unhealthy foods attractive to children not only affect public health and nutrition, but also raise safety

and allergy concerns. Finally, inclusion of foods in vending machines could be based on nutritional criteria.

The pledges made by the Children’s Food and Beverage Advertising Initiative do not cover in-store promotions, and almost no pledges cover front-of-package labeling. The failure of self-regulation in these areas argues for government involvement, Pomeranz argued. “I have analyzed all these [potential regulations for retail environments] under the O’Brien test, and they passed,” she said. “I would encourage governments around the country to institute them.”

THE USE OF TAXES TO INFLUENCE FOOD AND BEVERAGE PURCHASES

Individuals respond, to varying degrees, to economic incentives. Therefore, the use of taxes to increase prices can affect the consumption of foods and beverages, which may translate into changes in weight.

Lisa Powell, senior research scientist, Institute for Health Research and Policy, and research professor, Department of Economics, University of Illinois at Chicago, described her research on the effects of taxes on food consumption and some of the policy implications of that research. Using a variety of national data sets, Powell and her colleagues have been examining the associations among state-level soda taxes, consumption, and weight outcomes. These taxes vary by state and change over time. The group has looked at adolescents in particular, since their consumption of soda and sugar-sweetened beverages has greatly increased. “These calories account for up to 10 percent of their total calories per day, so that is a lot of extra discretionary calories,” Powell said.

The group has used both cross-sectional and longitudinal models to analyze the effects of tax rates, and it has analyzed both the absolute rate of soda taxes and the soda tax rate compared with the general food tax rate. Soda consumption and weight outcomes depend on many different factors, and analysts must control as many of these factors as possible to measure the effects of taxes. Powell and colleagues’ analysis showed that a 1 percentage point increase in the soda tax rate would reduce household consumption over 3 months by only about 9 ounces. That amount sounds small, said Powell, but it is also in response to a small change in taxes. Today taxes on soda average about 4.36 percent. If the tax rate were increased in all states to 7 percent, household soda purchases would be an estimated 3.6 percent lower.

Powell and her colleagues also estimated the effects of a new 20 percent tax on soda, assuming that the effects of tax increases can be extrapolated from small to large increases. In that case, household regular soda purchases would be an estimated 33 percent lower. The extent to which this reduction would apply to all soda purchases depends on whether soda consumed out-

side the home is similarly responsive to tax increases. For example, people would perhaps be willing to spend more in restaurants than in supermarkets.

Translating changes in consumption to changes in weight outcomes is a “large leap,” Powell acknowledged, and depends on many factors. For example, if all people substituted water for soda in response to tax increases, weight reductions would be likely, but this is not a realistic scenario. Rather, people would probably substitute other drinks for soda. For example, they might drink milk, although they would be unlikely to drink 20 or more ounces of milk to quench their thirst. More likely, they would substitute drinks such as sugared iced tea or high-sugar fruit drinks, which is why it is important for taxes to cover all sugar-sweetened beverages, not just soda.

In data comparing soda consumption, school soda purchases, and weight change from third to fifth grade, Powell and colleagues found associations that varied by group. While a higher soda tax did not affect total consumption for the entire sample, it did have an effect on total consumption for low-income students. Higher taxes also reduced school consumption for low-income students, African American students, and heavy television watchers. With respect to weight, higher soda taxes had a small positive effect on the change in body mass index from third to fifth grade, but taxes had a stronger positive effect on children who were at risk of becoming overweight. Powell recounted how, in a simulation of the effects of an 18 percent sales tax using these data, she and her colleagues found that such an increase would correspond to about a 20 percent reduction in excess weight gain from third to fifth grade.

In a study of the link between soda taxes and weight among adolescents, the associations found were for the most part very small. Higher tax rates had the greatest effect on youth who were already at risk of becoming overweight. The same effect was found in a study of fast food pricing. This makes sense, said Powell, because youth at the upper end of the weight distribution may be consuming more calorie-dense foods. This is an important finding, she suggested, because it means that a tax would have more of an effect on people at the upper end of the weight distribution. “From a public health point of view, that is exactly the group whose behavior and weight outcomes you are trying to change,” she said.

Food taxes are regressive in that they disproportionately affect lower-income individuals. But lower-income individuals also are more responsive to price. Those who respond to higher prices by changing their behavior therefore will achieve a greater health benefit than those who do not. Also, one way to offset the regressive effect of such tax increases would be to use the revenue to subsidize the purchase of fruits and vegetables for low-income individuals. These findings already have implications for tax policy, said Powell. For example, Supplemental Nutrition Assistance Program (SNAP) recipients do not pay taxes on their food items, so taxes on sugar-sweetened

beverages would exclude this group. From a strictly public health perspective, it may therefore be necessary to think about applying these taxes to SNAP purchases so that there is no group exempted from this policy.

Food taxes generally have not been introduced with the aim of modifying consumption behavior as has been the case in other public health areas such as tobacco use, Powell noted. Where taxes have been imposed on selected categories of foods, such as soft drinks, candy, and snacks in grocery stores and vending machines, the tax rates have been quite low. In those states that do tax soda, rates are generally between 5 and 7 percent. As some jurisdictions adopt higher taxes, the greater variation among states will enable further work on the relationships among taxes, consumption, and weight.

Powell also commented on the importance of tax policy design, noting that excise taxes included in the shelf price of a product may be more effective than sales taxes. A sales tax is applied at the checkout counter, so consumers may be unaware of what taxes are being applied to which products. Also, taxes applied on the basis of liquid volume would be more effective than those applied as a percentage of the sales price since supersizing a drink usually costs relatively little, but the per unit tax would persist on the larger quantity.

DISCUSSION

In response to a question about the possibility of subsidizing healthy foods, Powell observed that such subsidies are less likely than taxes to raise political opposition. The problem is that most governments do not have the money to implement subsidies. A politically palatable option, noted above, may be to use tax revenues on unhealthy foods to subsidize healthy foods. Powell’s research indicates that such subsidies would increase the consumption of fruits and vegetables among low-income families. However, the subsidies probably would have to be limited to low-income families to be affordable.

Russell Pate, a member of the IOM’s Standing Committee on Childhood Obesity Prevention, asked Powell about building the political will to increase taxes on certain foods. Powell responded that, according to the polls she has seen, people are more likely to accept a tax if the money is dedicated to a specific purpose, particularly to reducing childhood obesity, rather than going into the general fund.

In response to a question about how the issues surrounding expressive conduct apply in schools, Pomeranz observed that schools have the ability to regulate speech and are also a nonpublic forum, which means they could be more heavily regulated than traditional public venues. That they are not

more heavily regulated probably has to do more with political than legal considerations, she said.

Pomeranz also addressed a question about the incentives for restaurants to offer smaller portion sizes, not just different types or categories of food. She replied that such incentives would constitute regulation of conduct similar to the other examples she had mentioned.