“We all have these mental models in our heads. Often they are an inch deep and a mile wide. We construct them based on whatever information we have in order to have some coherent way of managing the world . . . whether that information comes from Gwyneth [Paltrow] or comes from a scientist or it comes from an advertisement.”
“Consumers are faced with tens of thousands of food-related communications every year. They are literally swimming in a sea of messages.”
Moderated by Fergus Clydesdale of the University of Massachusetts at Amherst, planning committee member, and Sylvia Rowe of SR Strategy, Session 2 had as its goal to explore how scientific information is communicated, including the credibility of the source and of the communicator, the clarity and usability of the information, misconceptions and misinformation, the impact of scientific communication on policy makers, and the role of policy as a macro-level channel of communication. This chapter summarizes the Session 2 presentations and discussion.
The session opened with a presentation by Timothy Caulfield, University of Alberta, on how popular culture, celebrities in particular, influence consumer decisions about food and nutrition. He suggested that what he described as “pop culture nutrition noise” has created a gap between sci-
ence and people’s food-related behaviors. For example, an estimated 4.3 million Canadians have gone gluten-free, which he opined was a “remarkable” number “given what the science says.” He observed that although celebrities may promote ideas that may have some emerging science behind them, they often cherry-pick the data and push an idea until it builds and is picked up by social media. He believes that many food-related decisions are as much (or more) about identity as about nutrition: people want to be identified in a certain way, and celebrities help create those identities. He expressed hope that knowledge of how celebrities influence decisions can help inform strategies to influence healthier choices.
William Hallman, Rutgers, the State University of New Jersey, described the current state of scientific education in the United States. “Many Americans lack the foundation in basic science to put new scientific information into any kind of context,” he suggested. Most nonscientists rely on words or pictures to tell stories, he observed, yet most scientists communicate using numbers, with scientists from different disciplines using different types of numbers. He echoed Grier’s and Baur’s calls in Session 1 (Chapter 1) to start with the end user, or consumer, and encouraged workshop participants to remember the mantra, “mental models matter.” That is, people construct mental models to have some coherent way of managing the world. Hallman explained that whether the information used to construct these models comes from celebrities, science, the Internet, or elsewhere, everyone has them, and communicators need to consider these mental models when thinking about how to translate scientific information into popular thought.
Sally Squires, Powell Tate, suggested that everything discussed thus far in the workshop related to the issue of trust. Twenty years ago, she noted, television and newspaper reporters earned a mix of trust and distrust with regard to communicating risk information related to food. Today, she said, they are among the least trusted authorities, with the most trusted professionals being nurses, followed by doctors. This declining trust, in her opinion, reflects a change in the media landscape. Media have changed more in the past 10 years than at any other time in history, she said. Both digital and social media are growing, and people are paying more attention than ever before to what their friends are saying. Squires’s take-home message was that having expertise does nothing for communication unless that expertise is accompanied by characteristics known to be associated with trust. Trusted sources are those that are concerned with public welfare, provide understandable and relatable information, and admit to uncertainties.
The growth of digital and social media has contributed to what Carol Byrd-Bredbenner, Rutgers, the State University of New Jersey, described as a “sea of messages” in which consumers are swimming. Yet even with all
these messages and even though consumers are actively seeking information, she said, people have reported finding it easier to do their taxes than to understand how to eat healthfully. She suggested that what she termed “communication friction” is interfering with effective communication about food and nutrition. Communication friction comes from what she described as “flabby” writing style, that is, writing that is difficult to read and uses unfriendly vocabulary. Inconsistency—whether in terminology, in formatting, or in information itself—also creates communication friction, she noted, while even simplicity can be problematic when it relies on definitive language, which creates skepticism. She listed several additional sources of communication friction and ways to eliminate it.
The sea of messages consumers receive are being delivered not just through digital and social media but also from food products themselves according to Craig Andrews, Marquette University. He described the very difficult communication environment consumers face while shopping, with the multitude of nutrition-related claims and symbols on food packages. Whether nutrition disclosures actually work depends on many factors, he explained, including whether the message being sent is the right one for the audience and what its goal is (e.g., exposure, comprehension, behavior). Disclosure may fail, he said, when a message is not personally relevant or noticed, when consumers are already familiar with the brand, when they lack the necessary nutrition knowledge, or when they become desensitized (e.g., after repeated false alarms or when messages are more extreme than necessary). He emphasized the importance of pretesting messages with the target audience.
Scot Burton, University of Arkansas, reiterated that consumers are exposed to a broad array of nutrition claims, icons, and information on a daily basis, sometimes with ambiguous and unintended effects. Although nutrition disclosures can have positive effects, he expounded on Andrews’s observation that a variety of individual and contextual factors impact their overall effectiveness. He described research on the effectiveness of front-of-package disclosures showing that effectiveness depends on the processing task for which consumers are using the information (e.g., evaluating a single brand versus comparing different brands). Additionally, he described a study on calorie labeling at restaurant chains showing that the effect of such labeling varies among different segments of the population and in different contexts.
Jeff Chester, Center for Digital Democracy, began by saying, “We are entering a new era of mass personalized communications. [Companies] are able to track you and target you anywhere, anytime.” He described how food and beverage companies are partnering with digital media companies, collecting and using big data, and redefining marketing to shoppers. While collecting data, he said, these companies are also actively engaged in shap-
ing the “shopper journey.” He observed that marketing now is not just the sale of a single product but a comprehensive and multichannel effort to promote brand loyalty and continuous consumption. Digital marketers have helped create what they call a “path to purchase” that entails constantly promoting products using an expanding array of data-driven social media, online video, and mobile phone apps. A new generation of YouTube celebrities known as “influencers,” for example, are helping marketers integrate their products with entertainment to help trigger brand loyalty among youth. Chester called for a greater understanding of how this new era of marketing is changing the relationship between consumers and scientific and accurate product information.
Vivica Kraak, Virginia Polytechnic Institute and State University, reiterated what several speakers had said previously about the crowded food messaging environment. Exacerbating this problem, she noted, as Chester had, is that many companies marketing to children under 12 years of age have yet to align their brand mascots or licensed media characters with uniform nutrition criteria. Likewise, no food, beverage, or restaurant company has yet pledged to align its celebrity endorsements targeting teens with those criteria. Today’s food messaging environment calls for comprehensive, consistent, and smart policies, Kraak said—not necessarily new policies, but revisions to existing policies. Products that are unhealthy need to be disincentivized, she asserted, and healthy products incentivized. She emphasized that both private and public policies play important roles; that policy change is an iterative, not linear, process; and that the decision-making processes of policy makers are very different from those of scientists, with policy makers valuing nonscientific as well as scientific information. She encouraged scientists to be more aware of the cultural differences between science and policy, to ask more policy-relevant questions in their research, and to communicate their findings more effectively to diverse audiences.
In his closing presentation, Joseph Levitt, Hogan Lovells U.S., LLP, described three case studies illustrating the gap between science and public perception. One was the use of Alar in apples in the late 1980s. At the time, the greatest food safety concern was carcinogenicity. Although the U.S. Environmental Protection Agency (EPA) had determined that the use of Alar in apples posed no significant cancer risk, 60 Minutes aired an exposé on the carcinogenicity of Alar in apples, and Meryl Streep made a series of public appearances offering the same information. The U.S. Food and Drug Administration (FDA) would later call on a consumer communications expert who explained that the real issue underlying the Alar scare was not the science, but the triggering of a series of “consumer outrage” factors: that Alar was intentionally added, that it was hidden, that it was being fed to vulnerable populations (i.e., children), and that the purported risk (i.e., cancer) was considered a serious one. The lesson learned, Levitt said, was
that “science by itself cannot control public perception. There are other factors that come into play.” Levitt encouraged a greater understanding of the anxieties underlying public concerns about food.
To close the first day of the workshop, David Freedman, contributing editor of The Atlantic, reflected on the information and opinions expressed thus far. He observed that the problem being addressed was how to support the public in receiving and embracing scientifically valid information about healthy eating in a way that will lead to healthier choices, healthier behaviors, and ultimately healthier living. But immediately, just in defining the problem, he said, another problem emerges—which information? Different scientists reach different conclusions about what is causing any given problem, he noted. Even if it is decided that the most immediate problem is obesity, scientists have yet to agree on what exactly needs to be done to combat obesity. And even assuming that they do agree on what needs to be done, yet another problem immediately emerges: What actually can be done? Freedman observed that people are getting caught up in “pop messaging” and locked into beliefs that are not scientifically valid and that end up short-circuiting any effort to deliver accurate information. If someone has decided that calories do not matter, for example, what good does it do to improve at sending messages about calories? And, Freedman continued, even if experts can cross that gap and reach consumers, which message should they send? He encouraged going beyond theory and getting out into the public to determine how to tackle the problem.
In his presentation on the role of popular culture in framing decisions people make about food, Caulfield focused on celebrity influence. He warned the workshop audience that he would be engaging in some speculation in his inferences about the true power and influence that pop culture and celebrity branding exert with respect to consumers’ food and nutrition decisions.
Pop Culture Nutrition Noise
Caulfield referred to Sonya Grier’s description of the growing discussion in popular culture about nutrition, food, and wellness (see Chapter 1 for a summary of Grier’s presentation). He showed a picture of Michael Douglas discussing his gluten-free diet with Jimmy Fallon, when Douglas told Fallon
1 This section summarizes information presented by Professor Caulfield.
that having gone gluten-free is why he (Douglas) “looks so great.” Upon watching the televised interview, Caulfield went on Twitter and tweeted, “Michael Douglas talking gluten-free pseudoscience on @jimmyfallon tonight. Frustrating this health bunk gets this kinda profile. Science!” “This prompted a strong tweeter reaction from gluten-free enthusiasts,” he said. Gluten-free diets are an extremely popular trend now, he observed, with an industry study reporting that 4.3 million Canadians have gone gluten-free or tried to reduce gluten in their diets. This is a “remarkable” trend, in his opinion, given what emerging science says about gluten-free diets and nonceliac gluten sensitivity.
As another example of celebrity influence, Caulfield mentioned Gwyneth Paltrow and the “incredible hesitancy” in the general public regarding genetically modified organisms (GMOs). According to a 2015 Pew Research Center study, 88 percent of scientists surveyed said genetically modified foods are safe to eat, but only 37 percent of the public agrees with that view (Pew Research Center, 2015). This represents “one of the biggest gaps between scientists and the general public,” according to Caulfield. In his opinion, celebrities have played a role in creating that gap.
Another example of celebrity influence on decisions about food is Katy Perry’s focus on cleansing and detoxing and what Caulfield described as the “incredible” rate at which this trend has grown in the past several years. If he had asked 8 years ago how many people in the audience had heard of cleansing and detoxing, he suspected only a handful of hands would have been raised. Today, numerous cleanse and detox books and products are on the market. Some estimates suggest it is a $5 billion industry, according to Caulfield. In one survey of naturopathic physicians, 92 percent of respondents reported using clinical detoxification (Allen et al., 2011). This trend is “scientifically absurd on numerous levels,” Caulfield stated. He cited a lack of evidence for the need to detox and suspects that celebrity culture has played a major role in popularizing the trend.
Organic foods are another example of celebrities pushing the nutritional value of certain foods despite what Caulfield said was little evidence indicating that, in this case, organic foods have a nutritional advantage over nonorganic foods (Dangour et al., 2009; Smith-Spangler et al., 2012). Juicing is yet another example. While juicing may not be harmful, there is no evidence suggesting it is “needed,” Caulfield remarked, yet again, it is a very large industry. “We could go on and on,” he said, with other examples.
Celebrities also are influencing food choices through their direct endorsements of food products. In a well-known study demonstrating the impact of celebrity endorsements, Boyland and colleagues (2013) found that such endorsements increase consumption even when the celebrities do not talk about but are merely associated with food. These endorsements do not have “a constructive impact in general,” Caulfield said. He also mentioned
Dr. Oz, who, Caulfield maintained, “says a lot of less than scientifically robust things,” but has a major impact on what people eat.
Celebrities are influencing not just food choices but other choices as well, Caulfield observed. He noted that studies consistently demonstrate that celebrities impact people’s decisions (Hoffman and Tan, 2015). He cited speculation that humans are evolutionarily predisposed to follow people with prestige or with specific skill sets (Tehrani, 2013). The notion that humans cannot help but follow and that they do so unconsciously builds on Kahneman’s work on cognitive biases, he explained. (Cynthia Baur had mentioned Kahneman’s Thinking, Fast and Slow during the Session 1 panel discussion; see Chapter 1.)
Examples of celebrity influence outside of the realm of food include cancer screening (e.g., Metcalfe et al., 2011) and, back in the 1920s, tanning. Before the 1920s, it was not considered fashionable to have a tan. But when Coco Chanel went on vacation, accidentally got a tan, and came home, then “boom,” Caulfield said, tanning was invented. “Despite the health implications of tanning, it is still with us,” he said. One of his favorite examples of celebrity influence and, in his opinion, one of the most powerful is cosmetic surgery (Swami et al., 2009). “People are altering their bodies in a semipermanent way based on norms that are almost entirely created by celebrities,” he said. The most rapidly growing form of plastic surgery today, augmentation of one’s buttocks, was influenced by one celebrity, Kim Kardashian.
The Acceleration of Celebrity Impact Through Social Media
Caulfield cited Katy Perry as an example of how celebrity impact is accelerated through social media. Perry tweets about what she eats, including her 26 supplements per day, Caulfield explained. With 75 million Twitter followers (more than follow President Obama), Caulfield asked, “How can CDC [the Centers for Disease Control and Prevention] compete?” Emerging evidence suggests that with Twitter and other similar platforms, such as Instagram, people feel as though the individuals posting messages are, as Caulfield put it, “just around the cyber corner” and that “they are speaking to us” (Stever and Lawson, 2013).
Additional emerging evidence suggests that social media also are having an influence on food choices through what Caulfield called the “Prius effect,” with people making food (or other product) choices as a way to express to the world who they are (in the case of the Prius, that they care about the environment) (e.g., Wansink et al., 2014). Nichter and Thompson (2006) report that people take dietary supplements for health reasons but also for identity reasons, that is, because they relate to what Caulfield described as a “new age kind of approach to life.” Von Essen and Englander
(2013) likewise suggest that people who eat organic food do so because they want to indicate, as he put it, “this is who I am.” Celebrities play a role in creating these identity packages, he suggested.
Caulfield noted that when something has been categorized a certain way—for example, when eating organic is perceived as a “good” thing to do—this categorization can actually alter the experience of taste. Bratanova and colleagues (2015) found that people who think they are eating organic food, which they perceive as “ethical” food, also think the food tastes better, with the improved taste reinforcing the idea that they should be buying this food. According to Caulfield, blinded studies have provided little evidence that organic foods actually do taste better; indeed, at least one study (Zhao et al., 2007) found that conventionally grown foods taste better. Moreover, he mentioned a study reporting that people who ate organic foods volunteered less time to help a needy stranger (Eskine, 2012). In a similar study, Karmarkar and Bollinger (2015) showed that people who bring their own shopping bags to stores are more likely to buy environmentally friendly items, but they also are more likely to buy junk food. In both studies, the authors speculate that people keep what Caulfield called a “moral balance sheet.” If they are doing some things they consider moral, they believe they do not have to worry about doing other moral things, or they can balance their sheet by buying junk food.
Again, in Caulfield’s opinion, celebrities have an impact on whether people perceive a food or behavior to be moral. He showed an image of Gwyneth Paltrow carrying a basket of green leafy vegetables and carrots. Even if people do not think of Paltrow as a credible source of scientific information, she is the source of an image, specifically an “old timey, natural” image, he explained. Her image helps people shape their identity packages and reinforces the perceived moral relevance of particular behaviors, he suggested.
In sum, Caulfield speculated that new ideas about nutrition, which may or may not have some basis in emerging science, become part of an identity package when they gain cultural currency and that celebrities play a major role in creating this identity package. A celebrity identity package is not only picked up by social media, he noted, but also reinforced by commercial marketing, becoming even more powerful. When this happens, he said, it is no longer possible to critique the idea the celebrity is spreading. Even if one critiques the science behind the idea, one is perceived as critiquing the person and the identity package associated with that person. Moreover, said Caulfield, as has been demonstrated with vaccine promotion (Nyhan
et al., 2014), when a concept becomes part of a person’s identity package, it becomes very difficult to change the person’s mind based on science alone.
Caulfield echoed the comments of other workshop speakers and participants on the complexity of decision making about food. “It’s not just about science or facts,” he said. “You could have a million studies on the impact or lack of impact of a diet, but when you have Beyoncé looking so fabulous and in shape, who are you going to believe?” Yet while celebrity influence complicates the information people receive about food, he expressed his hope that it can also inform strategies to influence healthier choices.
Scientific Education as a Starting Point
“Starting points really matter,” Hallman began, echoing Cynthia Baur’s earlier call to “meet people where they are” (see Chapter 1). A starting point with many Americans is that they lack the foundation in basic science to put new scientific information into any kind of a context, he explained. For most Americans, formal science education ends at high school. Even then, data collected by the National Math and Science Initiative (www.nms.org) suggest that only one-third of students who graduate from high school are ready for college-level science. Furthermore, most Americans do not have a college degree: fewer than 29 percent of Americans over the age of 24 have a bachelor’s degree or higher.
In Hallman’s opinion, moreover, colleges do a poor job of science education among the minority of students who do make it to college. According to U.S. census data, he noted, only about 10 percent of Americans graduating from college have a degree in a STEM (science, technology, engineering, and mathematics) discipline (U.S. Census Bureau, 2012). Although non-STEM students often are required to take one or two science courses, most introductory science classes are designed with the expectation that they are the first in a series of science courses the student will be taking and require students to memorize detailed scientific facts (e.g., segments of the Krebs cycle) for use in those later courses. This approach serves to frustrate, humiliate, and alienate many students whose primary interest in taking the required classes is to pass them, leaving the students to conclude, Hallman said, that “science is too hard to understand.” After they leave college, he continued, they are unprepared to engage with scientific topics or arguments and lack the skills needed to make decisions as informed citizens.
Therefore, Hallman asserted, most Americans depend on curators and
2 This section summarizes information presented by Dr. Hallman.
interpreters of scientific information. Celebrities are one such source, he said; others include health professionals, science communicators, authors and journalists, websites and blogs, social media, museums, and interactions with other people. He noted that although many people say they go to their health professionals for information about nutrition, most people in fact do not ask their physicians anything about the subject. And most physicians have little training in nutrition anyway, he added.
Literacy, Graphicacy, Numeracy, and Ecolacy
Hallman referred workshop participants to a book written by Garret Hardin in the mid-1980s, Filters Against Folly (Hardin, 1985). Hardin’s three “filters of reality” are literacy, numeracy, and ecolacy. Hallman added a fourth, graphicacy, and described each in turn.
Hallman described literacy as the ability to understand written and spoken words and stories and as the way normal people learn and communicate. Most people grow up learning culturally specific stories, anecdotes, examples, metaphors, and analogies.
Graphicacy, Hallman explained, is the ability to understand graphical information, or visual communication through sketches, photographs, diagrams, charts, maps, symbols, and other nontextual formats. As with literacy, he noted, the interpretation of nontextual information frequently is culturally constrained. People often think they can use pictures to communicate to others who cannot read their language or cannot read at all, but symbols do not necessarily have universal meanings, he observed. He showed an image of a skull and crossbones on a bottle. The bottle also had the word “poison” written on it. But the skull and crossbones symbol has no intuitive meaning and is a poor symbol for poison, he said; indeed, children need to be taught that it symbolizes poison. He told the workshop audience how he had showed an image of a skull and crossbones to his 6-year-old daughter and asked her what it was. She replied, “Daddy, that’s what pirates drink.”
Numeracy is essential to science, Hallman continued. Unfortunately, he said, much of the American public struggles with mathematical concepts, including very small and very large numbers, fractions, proportions (e.g., parts per billion), percentages, and probabilities. He noted that nonscientists and most journalists communicate using stories and often use pictures to illustrate those stories, while most scientists communicate using numbers. The result is a communication barrier, he explained. Additionally, different scientific disciplines use different kinds of numbers, creating another type of barrier.
Finally, Hallman explained, ecolacy is the ability to see “the big picture,” to envision both intended and unintended consequence. It is about
“getting it.” Figuratively, it is the ability to see both the forest and the trees. People can be extraordinarily literate, graphicate, and numerate, Hallman noted, but that does not automatically make them ecolate. Nor is seeing the big picture a matter of merely observing all the details and adding them together, he suggested. “Simply educating people about scientific details does not lead to a greater comprehension of the big picture or to their ability to necessarily make informed decisions,” he said. He also added that underlying all of these ideas is a fundamental issue of trust: as many of the risks and benefits associated with food are invisible, consumers have to trust the information presented without seeing physical evidence.
Mental Models Matter
“If you get nothing else out of this talk,” Hallman said, “I hope you remember this mantra: Mental models matter.” He reiterated the importance of starting points and knowing how people think about things. “We all have mental models in our heads,” he explained. “Often they are an inch deep and a mile wide. We construct them based on whatever information we have in order to have some coherent way of managing the world.” Whether that information comes from a celebrity such as Gwyneth Paltrow, a scientist, or an advertisement, he explained, we take it all in and use it to construct mental models.
Hallman used foodborne illness to illustrate the type of food-related mental models many Americans have. A series of studies conducted for the U.S. Department of Agriculture on how to get people to respond better to food recalls and to engage in better food safety practices revealed that most Americans know little about foodborne illness (Cuite et al., 2007, 2008; Hallman and Cuite, 2010; Hallman et al., 2009). People underestimate the incidence of foodborne illness, cannot identify groups of people who are particularly at risk for such illness, cannot identify its symptoms, and do not recognize it even when they personally experience it. The CDC estimates that one in six Americans gets sick every year with a foodborne illness, Hallman reported. Yet, in a 2008 study, he and his colleagues found that only 18 percent of respondents reported ever having been made sick as a result of eating contaminated food (Hallman et al., 2009).
The mental models people have around foodborne illness lack feedback loops, Hallman continued. People usually blame foodborne illness on someone else and do not believe they are the cause of their own illness. Even fewer believe they have made anyone else sick as a result of their poor food safety practices. Many people also believe that symptoms become evident shortly after one has eaten a tainted food, which Hallman said is typically wrong; many symptoms do not appear until 12, 24, or 48 hours or more after the food is ingested. As a result, he noted, people usually do not con-
nect their actions with the consequences. He explained that this disconnect creates a problem when communicating not only with home cooks but also with people who work in food service.
Many people have what Hallman described as “sympathetic magic” mental models concerning food. As an example, he noted that many people believe in “psychological contagion,” so that foods that come into contact with or are associated with things viewed as “dirty” become stigmatized even if they have been made clean after the contact. In a study conducted for the CDC about 10 years ago, he and his colleagues asked 1,100 American adults how often they engaged in particular behaviors (Hallman, 2008). They found that very few people who found a stranger’s hair in their food would be willing to take the hair out and keep eating. The same is true of finding an insect in one’s food; very few Americans would be willing to keep eating the food.
Hallman continued by noting that more than half of those surveyed said they would throw food out after its sell-by date “because of germs.” Many people believe that “germs come from other people,” he added. He explained that cellophane was first marketed for food products as a way to keep food separate from the “germs” of other people. He noted that people also tend to anthropomorphize germs, particularly when communicating to small children (see Figure 2-1). Based on the same CDC dataset described above, he and his colleagues found that nearly one-quarter of Americans agree that germs can sense when people are nearby, about one-third agree that germs can sense which people are most vulnerable, and more than half agree that germs move to places that make it easier for them to infect people.
Hallman reported that studies on intuitive toxicology have shown that when thinking about the risks of toxins in their food, many consumers fail to take dose into consideration and are unfamiliar with or do not believe in the existence of exposure thresholds. For most people, he said, what matters is whether they are exposed, not the amount of exposure. Additionally, he observed, most people have no idea how toxins cause harm. Many think that all chemicals are poisons; others think that particular ingredients are poisons; and some think everything is dangerous. Hallman showed several images of book covers with titles ranging from Milk, the Deadly Poison to Slow Death by Rubber Duck: The Secret Danger of Everyday Things. As far as who is doing all this poisoning, he said, “blame is easily placed.” He showed more book covers, one title being The Crazy Makers: How the Food Industry Is Destroying Our Brains and Harming Our Children.
Finally, Hallman noted that “how to” advice is plentiful, again being the subject of many books. He showed covers of some of the same detoxification books that Caulfield had cited.
FIGURE 2-1 Visual examples of how germs are anthropomorphized.
SOURCES: CDC, 2015a; Fight Bac! and Partnership for Food Safety Education, 2016.
Where Does the Misinformation Come From?
Hallman stated that much of the misinformation that people are using to construct their mental models comes from the Internet. As one example, he showed an excerpt from an article, “The Top Five Cancer-Causing Foods,” that was published on a Natural News website. The article states, “The truth is that most people give themselves cancer through the foods, drinks and products they choose to consume. In my opinion, over 90 percent of cancers are easily preventable.” The author of the article, “the Health Ranger,” goes on to say that cancer tumors develop, in part, by
feeding on sugar in the bloodstream. As another example of misinformation on the Internet, Hallman showed the image of a graph titled “Why do you need to detox and cleanse your body?” The graph illustrates a relationship between the consumption of “junk food” and illness. As yet another example, Hallman showed the cover of a book titled How to Cure Almost Any Cancer at Home for $5.15 a Day. The book tells readers how they can change their body chemistry from “cancer-friendly acidic” to “cancer-killing alkaline” for “pennies a day.” As a final example, Hallman showed an image of Oprah, the “ultimate expert.”
Hallman and colleagues conducted a study of FDA-approved qualified health claims versus “common knowledge.” They asked respondents how familiar they were with the relationship between certain dietary components (e.g., olive oil) and particular health claims (e.g., prevention of heart disease) (Hallman, 2015). They found that people were as familiar, or more so, with common folk wisdom claims as with FDA-approved qualified health claims. Moreover, people were applying that folk wisdom. Among 1,300 American adults aged 55 and older, for example, 35 percent reported consuming dark chocolate for heart health (Hallman, 2015). It is not just Dr. Oz who is popularizing the idea that dark chocolate can have health benefits, Hallman observed. He showed an image of a WebMD webpage with a list of top “superfoods” offering “super health protection.” The American Diabetes Association also lists on its website “diabetes superfoods.” “We are getting a confluence of this idea that there are in fact superfoods,” Hallman said.
People are open to simple heuristics, Hallman said. He mentioned Michael Pollan’s book Food Rules; examples of these rules include “eat organic” and “whole foods are best.” Pollan’s is one of many books on food rules, according to Hallman. These rules depend on what he called “intuitive plausibility.” For example, the rule “Don’t eat breakfast cereals that change the color of the milk” has absolutely no science behind it. “But,” Hallman said, “doesn’t it sound like it is something that is true?” He recommended that communicators think about the intuitive plausibility of their messages when trying to translate scientific information into popular thought. How does the message connect with where people actually are? How can the invisible be made visible for people?
In conclusion, Hallman stated, “We cannot simply continue to lament lack of knowledge and action. We have to be proactive. The public needs and deserves better science communication about food safety, nutrition, and health. The question is, ‘Whose responsibility is it to do that?’” He encouraged workshop participants to consider the role they can play.
“Trust has really been underlying everything that we have been talking about,” Squires asserted. She cited findings from what she thought was a very interesting study conducted 20 years ago in the United Kingdom (Frewer et al., 1996). The study, which was government-funded, was conducted by a team of psychologists who examined topics surprisingly similar, she observed, to the topics being addressed at this workshop. One of the central questions investigated was why some individuals and organizations are viewed as trusted sources of risk information related to food, while others are not. The researchers found not only that medical professionals were among the most trusted sources, but also that no one mistrusted them. Scientists also were trusted, although less than medical sources, nor were they mistrusted. Other sources of information were both trusted and mistrusted to various extents. Television and newspapers, while among the most trusted sources of risk information related to food, were also among the most mistrusted. Squires found the combination of great trust and great mistrust in television and newspapers particularly interesting, especially given that this survey was conducted before the rise of social media.
More recently, the 2014 Gallup Poll on Honesty and Ethics found that registered nurses (RNs) are the most trusted professionals (80 percent), followed by doctors (MDs) (65 percent), members of the clergy (46 percent), bankers (23 percent), lawyers (21 percent), advertising executives (20 percent), business executives (17 percent), and car salespeople (8 percent). Scientists were not included in this particular poll. Not only are nurses viewed as the most trustworthy professionals, but they have held that position for the past 11 years, according to Squires. Nurses are considered trustworthy because they are believed to have very high or high standards of honesty and ethics, she explained.
Other Gallup poll data from 2004, 2007, and 2010 showed military officers and grade school teachers to be highly trusted (65 to 74 percent) and day care providers and judges less so (both in the 44 to 53 percent range). Auto mechanics, nursing home operators, television reporters, newspaper reporters, local officeholders, and state officeholders all were cited as trustworthy by less than 30 percent of the surveyed populations. State officeholders held the lowest level of public trust (24 percent in 2004, but only 12 percent in 2007 and 2010). Television reporters were trusted by only 23
3 This section summarizes information presented by Ms. Squires.
percent in all three survey years, while newspaper reporters were trusted by only 21 to 22 percent.
The Changing Media Landscape
Squires said she could not think of a time when media have changed more than they have over the past decade. “For those of us who were trained to be journalists,” she said, “we viewed it as a calling . . . we went into it because we felt it was important in a free society to be that extra [set of] eyes and ears.” That situation has changed dramatically in recent years, in her opinion, although she expressed hope that the field is moving forward.
According to 2015 findings from the Pew Research Center, key audience trends include a 5 percent increase in network news viewership and a 3 percent increase in local news viewership. Cable news viewership, in contrast, declined by 8 percent, and newspaper readership by 3 percent. With respect to online traffic, 39 of the 50 news sites included in the survey receive more traffic from mobile devices than from desktop computers, a trend that Squires said was not surprising. However, mobile visitors spend more time than desktop visitors per visit for only 10 of the 50 news sites. With respect to social media, slightly older (2013) data collected by the Pew Research Center (N = 960) indicate that the leaders in use are millennials, with a median of 250 Facebook friends, followed by gen x’ers (200), younger boomers (98), and older boomers (50). The 2015 study found that while growth in Facebook and other social media sites is being driven by millennials, participation among adults aged 55 and older is increasing. “We are paying more attention to what our friends on social media say about things,” Squires said.
Harvard University’s 2015 Institute of Politics study, much of which was focused on voting, also examined participation in social media platforms (Harvard University IOP, 2015). The researchers found that Facebook continues to drive a large proportion of social media participation (used by 83 percent of respondents), followed by Instagram (44 percent), Twitter (39 percent), Pinterest (34 percent), and Snapchat (33 percent). Of these, Squires said she especially likes Twitter. She noted that it was through Twitter that NBC News learned of the plane landing on the Hudson River, after someone on the ferry picking up the plane’s passengers tweeted about the incident. Participation in all these social media platforms has become a significant means of conveying messages and obtaining information, she remarked.
Results of a Sullivan Higdon & Sink FoodThink survey on the use of digital and social media to make food choices indicated that 1 in 10 consumers was engaging with grocery-related brands (Sullivan Higdon & Sink, 2014). Squires was surprised that the figure was that high. Twenty-
two percent of fathers reported being engaged in grocery-related brands on social media, followed by 18 percent of mothers, 17 percent of millennials, 13 percent of “good cooks,” 7 percent of boomers, and 7 percent of “bad cooks.” In Squires’s opinion, the higher percentage of fathers following grocery-related brands reflects their growing role as nutritional gatekeepers.
With respect to what consumers are doing with this information, results of the same FoodThink survey indicated that 76 percent of consumers were engaging in some activity on the Internet related to grocery shopping. Squires told the audience that she herself uses her iPad to check recipes and see what ingredients she needs to purchase. Twenty-five percent of the surveyed consumers were checking product prices online, 14 percent were using mobile coupons while shopping, and 11 percent were scanning Quick Response (QR) codes at grocery stores.
Other data from the same FoodThink survey showed that people were using social media to make online restaurant ordering decisions as well, with 42 percent of boomers and 68 percent of millennials accessing nutritional information. Not only were they using the information to help with online ordering, but they also were sharing information with others. Overall, Squires reported, 14 percent of users surveyed were sharing a restaurant experience with others. Twenty-four percent of millennials were sharing restaurant information, and 23 percent were following a restaurant on social media, compared with 8 percent and 7 percent of boomers, respectively. Fathers and mothers were highly engaged as well, Squires explained, with 33 percent of fathers and 17 percent of mothers sharing information and 31 percent and 21 percent, respectively, following a restaurant.
What Fuels Trust and Mistrust?
Squires elaborated on some additional findings from the 20-year-old study that she had mentioned at the outset of her talk. Frewer and colleagues (1996) found that trust was fueled by being responsible, trustworthy, accountable, and accurate; having a good track record; being concerned with public welfare; and having knowledge and facts. Mistrust was fueled by a perceived vested interest, self-protection, exaggeration, and distortion. The most mistrusted information was on natural toxins (e.g., aflatoxin), genetic engineering (i.e., GMOs), and pesticides, while the most trusted information was on high-fat diets, microwave ovens, food poisoning, food irradiation, and alcohol. The conclusions reached by the authors, according to Squires, were, first, that improving scientific literacy is not just a matter of teaching better science. The bar is too high for many people to understand what they need to know in order to evaluate information; moreover, the public lacks the motivation to learn what needs to be learned. Second, trusted independent sources are those that provide information that is
understandable and relatable. Trust also was linked to admitting uncertainties and helping the public understand that science is a process and that scientific knowledge changes.
Squires’s take-home message was that trust is linked to perceptions of accuracy, knowledge, and concern with public welfare. Having expertise or the freedom to talk does not lead to trust, she said, unless it is accompanied by these other characteristics. Distrust is associated with perceptions of deliberate distortion of information by the source and a history of providing erroneous information, she concluded.
The Food Communication Environment
Consumers are faced with tens of thousands of food-related communications every year and are “literally swimming in a sea of messages,” Byrd-Bredbenner began. The messages are coming from many different places, including educational sources, professional organizations, and media of all sorts. Topping the list is the Internet, which is consulted by three of every four people for health information every year, according to Byrd-Bredbenner. In 2013, there were 40 million downloads of mobile apps related to diet, a figure she suspects is even higher now. According to a 2000 poll conducted by the Academy of Nutrition and Dietetics, in 2000, only 19 percent of those surveyed agreed with the statement, “I actively seek information about nutrition and healthy eating.” By 2011, that figure had more than doubled to 46 percent. “People are actively looking,” Byrd-Bredbenner said.
In a survey conducted by the International Food Information Council (IFIC), Byrd-Bredbenner reported, 94 percent of respondents reported that they thought about the healthfulness of foods and beverages they consumed, with 48 percent thinking about it “a lot” and 44 percent thinking about it “a little.” Eighty-four percent of respondents indicated that they thought about the safety of foods and beverages, with 39 percent thinking about it “a lot” and 45 percent thinking about it “a little.” Not only are people thinking about the healthfulness and safety of the foods and beverages they consume, Byrd-Bredbenner explained, but according to this survey, most also are trying to do something about these concerns. Fully 96 percent of respondents indicated that they were trying to control the
4 This section summarizes information presented by Dr. Byrd-Bredbenner.
healthfulness of their diet, and 94 percent said they were trying to control the safety of foods and beverages they consumed.
In sum, Byrd-Bredbenner said, there is a great deal of interest in the healthfulness and safety of food. But people also receive a multitude of messages about these topics, she observed. According to another IFIC survey, half of the U.S. population believes it is easier to do one’s own taxes than to figure out how to eat healthfully.
With the persistence and escalation of high body mass indices (BMIs), high blood pressure, high blood cholesterol levels, and other health-related problems, Byrd-Bredbenner suggested that perhaps all of the messaging about food nutrition and food safety is not having the desired effect. She asked, “What is getting in the way?” The answer, she said, is “communication friction.”
“Friction is anything that slows you down,” Byrd-Bredbenner said. Communication friction can cause consumers to “grind to a halt,” she asserted, and not pay attention to a message, veer off in a different direction, or use information in a way that does not benefit them. She listed several sources of communication friction, the first being what she described as “flabby and convoluted” writing style.
Flabby writing is difficult to read, Byrd-Bredbenner explained. She suggested that consumers are too busy to continue reading writing that is convoluted and unduly complex. This is especially true for the 4 of 10 Americans who have basic or less than basic reading skills, that is, at the third- to fifth-grade level and lower, she noted.
As an example of flabby writing, Byrd-Bredbenner read the first two sentences of a ChooseMyPlate.gov tip for increasing physical activity: “Children and adolescents should do 60 minutes or more of physical activity each day. Most of the 60 minutes should be either moderate- or vigorous-intensity aerobic physical activity.” After reading these sentences, she said, “I am tired already. I will bet the average mom that is reading this is going to flip to the next page.” In her opinion, the information could be written in a much more readable, user-friendly way, making it more accessible to consumers, and doing so would allow more people to act on the information. She observed that not every page on the ChooseMyPlate.gov website is as difficult to read. More broadly, however, studies have shown that the average reading level required for health and nutrition websites is eighth grade and for printed nutrition education communications is ninth
grade. Byrd-Bredbenner interprets this to mean that a significant number of consumers are being left out when it comes to written communications.
Qualified health claims are another example of flabby or convoluted writing, Byrd-Bredbenner stated. She mentioned that a forthcoming study (at the time of this workshop) on the readability of qualified health claims found that their average required reading level was 12th grade, with one health claim having what would be estimated as a grade 30 reading level (Berhaupt-Glickstein and Hallman, 2015). Fewer than 1 percent of qualified health claims were considered easy to read. Inaccessible qualified health claims cannot achieve the goal of helping consumers do a better job of choosing healthful foods, Byrd-Bredbenner remarked.
Unfriendly vocabulary is another type of communication friction, Byrd-Bredbenner said, usually taking the form of using terms that are more appropriate for health professionals than for consumers (see Table 2-1). For example, health professionals talk a great deal about the importance of fiber and eating more whole grains. Yet fewer than 10 percent of consumers are familiar with the terms “insoluble fiber,” “soluble fiber,” and “functional fiber,” according to Byrd-Bredbenner, while 65 percent are familiar with the term “fiber.” She suggested that perhaps health professionals should just talk about “fiber.”
TABLE 2-1 Health Professional “Talk” Versus Consumer Vocabulary
|Health Professional Talk||Consumer Talk|
|diet||foods you eat|
|females of child-bearing age||women who might get pregnant|
|dark, green, leafy vegetables||greens|
|complex carbohydrates||starchy foods|
|serum glucose||blood sugar|
SOURCE: Presented by Carol Byrd-Bredbenner on September 3, 2015.
Inconsistency is another type of communication friction, Byrd-Bredbenner said. When health professionals write, give presentations, and talk to people, she suggested, they like to use varied language to “keep it interesting.” But doing so can backfire, she noted. An example is use of the word “legume.” Dietary guidance documents use “legumes” to refer to dried beans as well as dried peas, while consumers know these foods simply as “beans.” “We need to be talking about beans if we want them to eat more of them,” Byrd-Bredbenner said. The same is true of whole grains, she suggested. Sometimes health professionals talk about “whole grains” and sometimes about “fiber-rich whole grains.” When consumers hear these two different phrases, they can become confused. Consistency in terminology is therefore important, Byrd-Bredbenner stated.
Byrd-Bredbenner explained that not only inconsistent terminology but also inconsistent formatting creates communication friction. Nutrition Facts panels have helped improve the consistency with which information on calories and nutrients is presented, in her opinion. But she suggested that a great deal of clutter on food packages gets in the way of people’s finding and using that information to make good dietary decisions. She noted that inconsistent formatting is also a problem with qualified health claims. There are 36 different variations in how the 53 currently enforced qualified health claims are presented. Some tell how many studies back the qualified health claims, while others do not; some provide evidence for a claim and then give the claim, while others give the claim and then the evidence. According to Byrd-Bredbenner, this lack of consistency makes it difficult for consumers to know where to look and how to parse the claims.
Changing Story Lines
Changing story lines are another source of communication friction, according to Byrd-Bredbenner. She noted that three of four consumers say they see a large amount of contradictory information. She suggested that consumers become frustrated when they think scientists are changing their minds and begin to believe that scientists do not really know what people need to be eating to stay healthy.
Byrd-Bredbenner asserted that keeping a message simple and streamlining a scientific story can also confuse consumers (Fiscella et al., 1999; Jensen et al., 2011; Nagler, 2014; Vardeman and Aldoory, 2008). When consumers see a headline such as “eat less fat,” she noted, they get the
message, but often they end up eating not only more low-fat and nonfat products but also more calories. When this happens, she suggested, either the scientists or health professionals did not tell the full story—that is, that all sources of calories count—or consumers did not stick with the story long enough to hear it in its entirety. Either way, she said, the “eat less fat” message is confusing.
Streamlining also often leads to inadequate context, Byrd-Bredbenner said. She reiterated what earlier speakers had said about how many people do not understand the evolving nature of science. Often, she noted, consumer communications about nutrition present current science but do not put it in the context of the science that preceded it. In these cases, scientists and health professionals miss an important opportunity to help consumers weigh the evidence and adjust their “mental models.” (See the summary of Hallman’s presentation earlier in this chapter for a discussion of mental models.)
Byrd-Bredbenner observed that streamlining also leads to the use of definitive language, such as “consumers need to do x” or “this study’s findings clearly demonstrate x,” which introduces skepticism. And skeptical consumers, she suggested, are less likely to make the changes being advocated. She opined that hedging, a linguistic tool, although not used often in scientific communications to consumers, could be helpful. Hedging is conveying caution or tentativeness with respect to one’s findings. Scientists use it when communicating to each other in peer-reviewed journals; entire sections of papers may be focused on the limitations of the study being discussed, with warnings to readers that results should be viewed with caution. Yet, Byrd-Bredbenner noted, despite research indicating that hedging improves trustworthiness, it is not used in scientific communications to consumers. She believes that expressing uncertainties would improve the trustworthiness of scientists, which is positively correlated with the likelihood that consumers will pay attention to their recommendations. Hedging reduces the likelihood of what she called “nutritional backlash,” that is, consumers backing away from information and saying, “I am just going to eat. You guys go figure it out.”
In sum, Byrd-Bredbenner said, “We have to use streamlining cautiously.” While tight communication is important, she suggested, it is necessary to tell the whole story and in a way that avoids confusion, skepticism, and backlash. Otherwise, she stated, people will not be able to perform the recommended behavior.
Another source of communication friction is lopsided coverage (Allen, 1991; Berger and Milkman, 2010; O’Keefe, 1999; Verbeke et al., 2008;
Winter et al., 2015), which Byrd-Bredbenner explained refers to the valence of the presentation, that is, whether it is positive, negative, or neutral. Positive presentations tend to lead to positive attitudes and negative presentations to negative attitudes, she noted, but neutral, or balanced, presentations tend to lead to less skepticism and greater likelihood that consumers will think about the message. Two-sided presentations, she observed, can be presented in two ways: (1) descriptive, with both pros and cons, or (2) refutational, where the positives are presented and then refuted with the negatives. Refutational messages are more powerful and persuasive than both one-sided messages and two-sided descriptive messages, she explained, yet such messages are hardly ever seen in nutrition communications.
Vague Call to Action
Another cause of friction in communication, according to Byrd-Bredbenner, is a vague call to action. Consumers often are unsure what they need to do when they read a nutrition message, she observed. Instead of saying, for example, “Get more calcium,” she suggested the more specific call to action, “Have yogurt for a snack.”
Some communications are “just out of tune,” Byrd-Bredbenner continued. About 8 of 10 consumers want to hear what they should eat, she suggested, not what they should not eat (IFIC, 2015; Martin-Biggers et al., 2015). According to data from the Academy of Nutrition and Dietetics, a declining percentage of people are reporting that it appears they are always hearing information about what not to eat rather than what to eat. Still, Byrd-Bredbenner said, there is much room for improvement in this regard.
According to Byrd-Bredbenner, poker-faced communications, that is, those that do not elicit emotion, also can cause friction. Emotion-laden communications get people interested and invested in a topic, she said. She mentioned a study of The New York Times articles demonstrating that emotion-laden articles were more likely to be read and shared than articles with neutral tones (Berger and Milkman, 2010). Moreover, articles that were more positive in their content were more likely to be read and shared than those that were negative. In their own research, Byrd-Bredbenner and colleagues have found that mothers of preschoolers are more motivated to read short communications if they elicit emotions related to happiness/fun,
unique/special, and quick/urgent themes (see Figure 2-2) (Martin-Biggers et al., 2015).
Byrd-Bredbenner observed that communication friction can result from impersonal messages—generic messages that are broadcast to a general audience and end up resonating with no one. Picture an audience of women, she suggested. Not only do women differ demographically and by age, she noted, but different women define quality of life in different ways. And it is quality of life, she suggested, that really motivates people to make decisions about their health. “If we want our messages to resonate, we need to do more audience segmentation,” she said. She explained that tailored and targeted nutrition messages are more likely to be read and remembered, rated as attention catching, saved and discussed with others, and perceived as personally relevant (Brinberg et al., 2000; Hingle et al., 2013; Kreuter and Wray, 2003; Petty and Cacioppo, 1986).
Inattention to Taste
Messaging needs to take into account that when it comes to food, consumers’ highest priority is taste, Byrd-Bredbenner suggested. According to data from the Academy of Nutrition and Dietetics, she noted, the number one reason for not eating healthier foods is that people do not want to give up foods they like.
Communications that do not define the personal benefit of a food also create friction, Byrd-Bredbenner observed. Consumers want to know, she said, “What’s in it for me?” Consumers consistently report not seeing a personal benefit in messaging. Yet ample research indicates that knowing the benefits of a food correlates with improved diet quality (Aldrich, 1999; Beydoun and Wang, 2009; Byrd-Bredbenner and Finckenor, 2001; Moon et
FIGURE 2-2 Words that motivate consumers to read short communications because of the emotions they elicit.
SOURCE: Presented by Carol Byrd-Bredbenner on September 3, 2015; modified from Martin-Biggers et al., 2015. Reprinted by permission of Taylor & Francis Ltd., http://www.informaworld.com.
al., 2005; Smallwood and Blaylock, 1994). People who know the benefits of a food are more likely to respond in a positive way and to incorporate that food into their diet, Byrd-Bredbenner explained.
Theory and Testing
Finally, Byrd-Bredbenner remarked on the thousands of studies that have shown how the constructs and processes of behavior change theory really do make a difference in whether people accept and make a behavior change. Yet many communications and programs still are not applying those methods, she said, nor are the communications being tested with pilot audiences. She asserted that cognitive testing is essential to determine whether a message will resonate, motivate, or be understood but, she suggested, is done much too infrequently.
Byrd-Bredbenner closed by suggesting that messaging about food safety and nutrition could be improved by eliminating the above sources of communication friction:
- Create tight, accessible, friendly communications.
- Use consistent terms, formats, and story lines.
- Tell complete stories (create context and use hedging).
- Provide balanced and refutational coverage.
- Include an explicit call to action.
- Be positive and emotive.
- Keep food taste in mind.
- Identify clear personal benefits.
- Make sure the communication is grounded in behavior change theory.
- Road test the message.
She concluded by challenging workshop participants to “try to put these all in your next communication related to food and nutrition.”
Summarizing 45 Years of Experimental Research on Consumers
Consumers often are confronted with a “dizzying array” of nutrition symbols, icons, and facts in stores, Andrews began. First is the wide variety of front-of-package nutrition symbols, including both reductive or nutrient-specific symbols (e.g., Facts Up Front in the United States, traffic lights in the United Kingdom) and evaluative or summary symbols (e.g., Model Front-of-Package Symbol System [IOM, 2012]) (Andrews et al., 2014; Newman et al., 2014). Added to these are three different types of nutrition claims: (1) nutrient content claims, which describe the actual nutrients (e.g., “low-fat,” “natural,” “gluten-free,” “organic,” “non-GMO”); (2) health claims, which link a nutrient to a particular health benefit (e.g., “low in saturated fat, reduces coronary heart disease”), 12 of which are backed by significant scientific agreement and have been approved by the FDA in 2015; and (3) structure function claims (e.g., “high in calcium, helps build strong bones”). Added to these front-of-package symbols and nutrition claims is the Nutrition Facts panel, which Andrews noted was undergoing changes (at the time of this workshop). “It can be chaotic for consumers,” he said. “It is a very difficult environment.”
Andrews then posed the question of whether, in this environment, nutrition information disclosures and claims work. Based on his review of 45 years of research (Andrews, 2011), he said, “It depends.” Many factors come into play, including whether the disclosure or claim matches the appropriate communication objectives for the target audience. Is the objective exposure, attention, comprehension of the nutrition information, or behavioral change? Second, what is the message content? Third, what is the message modality? And finally, what are the effects on the receiver (e.g., altering initial beliefs)?
The elaboration likelihood model of persuasion provides a helpful conceptual framework for thinking about the persuasiveness of nutrition information, Andrews suggested. This model accounts for variation in consumers’ demographics (e.g., medical condition, age, gender), as well as the initial opinions they can bring to the processing of communications (Andrews and Shimp, 1990; Batra and Ray, 1986; MacInnis and Jaworski, 1989; Petty and Cacioppo, 1986). The key component of the model, Andrews pointed out, is “the receiver’s motivation, ability, and opportunity to process the message.” For example, he said, motivation, ability, and opportunity to process a message can be limited by many distractions,
5 This section summarizes information presented jointly by Dr. Andrews and Dr. Burton.
but if the receiver’s motivation, ability, and opportunity are all high, there is a strong likelihood that the message will lead to longer-term attitude and behavior change. He noted that another helpful conceptual tool for examining the effectiveness of nutrition disclosures and claims is the communication–human information processing model, which identifies stages of information processing from attention to behavior (Wogalter, 2006).
Andrews reiterated that a key question to consider is whether the goal of communicating nutrition information is exposure, comprehension, and/or behavior change. Most experimental research conducted over the past 45 years has focused on message claims and receiver effects (Andrews, 2011). In Andrews’s opinion, much more work could be done on what he termed “destination issues”—for example, whether the type of behavior targeted by nutrition information is focused on prevention or cessation, total or situational use, immediate or long-term effects, and so on.
Information disclosures can fail for several reasons (e.g., Stewart and Martin, 1994), Andrews suggested. For example: (1) people are not paying attention to the disclosures; (2) the information is not personally relevant; (3) consumers are already familiar with the information, and it addresses a routinized purchase; (4) consumers are distracted; and (5) consumers are desensitized after repeated exposures (they say, “I know my brand. I am not going to even look and analyze things.”).
Based on his experience, Andrews noted the importance of the Federal Trade Commission’s clear and conspicuous standards for televised ad disclosures, established in 1970 (see also Hoy and Andrews, 2004). An important standard, in his opinion, is dual modality, that is, communicating messages via both audio and video. He also underscored the importance of visual communication, especially for children.
Andrews listed several biases that impact whether health and nutrition claims may work. The first is positivity bias, which occurs when consumers give a product a better rating merely because of the presence of a health claim. Second is the halo effect, in which the presence of a health claim induces consumers to rate a product higher on other attributes not mentioned in the claim. For example, when a claim indicates “zero or low cholesterol,” consumers may rate the product as having a low fat level as well. Third, Andrews explained, the magic bullet effect is the attribution of inappropriate health benefits to a product. For example, eating a “healthy” Subway sandwich or some other “healthy” food item may be perceived as a magic bullet against “less healthy” foods eaten later. Fourth, interactive effects are judgment biases due to ambiguous information, such as when prior knowledge of nutrition interacts with the information in a health claim. Fifth, truncation behavior is the tendency for consumers to cut short their search for nutrition information (i.e., on the Nutrition Facts panel) when a health claim is provided on the front of a package (Roe et al., 1999). Finally,
Andrews observed, portion size and branding effects occur when people eat more either because of the serving or container size (Wansink and Kim, 2005) or because of the brand’s “healthy” claim (the “Subway effect”) (Chandon and Wansink, 2007), even when the food does not taste good.
In summary, Andrews reiterated the importance of understanding the consumer’s motivation, ability, and opportunity to process nutrition information. In his opinion, opportunity may be where the value of future research lies. “Just look in the stores,” he said. Consumers are often stressed and distracted, which limits their opportunity to process nutrition facts and information.
In closing, Andrews shared some research on advertising that he and his colleagues conducted. First, they administered a pretest and rated products based on whether they were perceived by consumers as not nutritious (e.g., “margarine”) or nutritious (e.g., “soup”). Then they conducted separate studies on margarine (Andrews et al., 1998) and soup (Andrews et al., 2000). In both cases, they found that consumers overgeneralized nutrient content claims (e.g., “low cholesterol” was overgeneralized to a perception of healthfulness) and that there was a halo effect (e.g., when a “low cholesterol” claim was made, consumers perceived other nutrients, such as fat, also to be at low levels when they were actually high). Andrews reported that these misleading halos were reduced only when the claims were accompanied by an evaluative disclosure (e.g., characterizing a per-serving level of margarine as “high” as evaluated by the FDA). Interestingly, with margarine, the halo effect was reduced when accompanied by an evaluative disclosure regardless of the consumer’s level of nutrition knowledge. With soup, however, because people perceive it as being “good for you” even when it contains high levels of a negative nutrient (i.e., salt), Andrews and colleagues (2000) observed the opposite: the effect of evaluative disclosures depended on the consumer’s level of nutrition knowledge.
Front-of-Package Nutrition Disclosures
Burton reiterated Andrews’s key point that the provision of accurate nutrition information often does not have the unambiguous results desired. He noted that food choices are affected by many variables that differ among individuals (e.g., goals, health consciousness, health knowledge), contextual influences, inferences beyond objective information (e.g., health halos and “health horns”), environmental effects, and other factors. He examined some of these differences for their effects on, first, front-of-package disclosures and, second, calorie labeling in restaurant chains.
A basic research question, according to Burton, is the most effective way to communicate front-of-package nutrition information. As Andrews had mentioned, such information is of two different types: (1) reductive,
also considered objective, in which information from the Nutrition Facts panel is condensed, or reduced, and placed on the front of the package; and (2) interpretive or evaluative, in which information is qualified in some way to indicate the relative healthfulness of the product (see Figure 2-3).
The effectiveness of front-of-package icons depends on whether consumers are engaged in a comparative versus noncomparative task when they view that information, Burton argued. Are consumers comparing the healthfulness of products of different brands, or are they evaluating a single brand? While the goal of the Nutrition Labeling and Education Act (NLEA)—to “provide information that would assist consumers in maintaining healthy dietary practices”—is relevant to both types of tasks, Burton suggested, the nature of the processing involved is very different for these tasks.
FIGURE 2-3 Different types of nutrition information messaging on food packages. (a) The Nutrition Facts panel, which is usually located on the back of a package. (b) An example of reductive, or objective, front-of-package labeling, with information taken from the Nutrition Facts panel. (c) Examples of interpretive, or evaluative, front-of-package labeling.
SOURCE: Presented by Scot Burton on September 3, 2015.
For comparative tasks, in which a consumer is evaluating several products, Nutrition Facts panels are not very practical, Burton asserted. With respect to front-of-package information, he stated that, based on resource matching theory, an evaluative icon should be a better match for consumers in a comparative processing context, while a reductive icon should be a better match in a noncomparative processing context (Newman et al., 2016). In a pilot study on icon attributes, he and his colleagues asked consumers a series of questions about their perceptions of different types of front-of-package icons. Consumers responded using a 7-point scale. Objective icons, such as the Guideline Daily Amounts (GDAs), were perceived as more detailed, specific, and quantitative than evaluative icons and as more appropriate for single-product evaluations. Conversely, evaluative icons were perceived as more interpretive in nature and more appropriate for product comparisons.
Burton elaborated on the methods used in a nutrition labeling study designed to evaluate different consumer processing contexts (i.e., whether consumers are evaluating a single product or comparing multiple products). In a comparative processing context, he explained, consumers look at different products and evaluate them simultaneously in a comparative manner, and then base their evaluation of the healthfulness of the single objectively healthiest product relative to the set of other products. In a noncomparative processing context, he continued, consumers are shown only the single objectively healthiest product, and their evaluation of the healthfulness of the product is based on their impression of that one product. In both processing context conditions, consumers are shown either reductive or evaluative icons. When conducting this type of study, Burton explained, in addition to gathering information on consumers’ perceptions of the objectively healthy product (e.g., “not at all nutritious” versus “highly nutritious,” “very unhealthy” versus “very healthy”), researchers can measure processing fluency (e.g., “it is easy to determine how healthy this product is,” “information about this product is easy to process”) and purchase intentions for the healthy product (e.g., “very unlikely” versus “very likely,” “not probable” versus “very probable”).
Purchase intention for a more objectively healthy product has been shown to be greater for evaluative front-of-package cues (e.g., Model Front-of-Package Symbol System [IOM, 2012]) in comparative than in noncomparative processing tasks, Burton explained. The effects are reversed with reductive cues (e.g., the GDA icon), with purchase intention for the objectively healthy product being greater when the consumer is engaged in a noncomparative processing task than when engaged in an evaluative task. Again, Burton said, “The nature of the task makes a difference in the performance of the different [front-of-package] cues.”
In an expanded experimental design, Burton continued, investigators
can examine product choice more closely by exposing consumers to situations in which they see either no icon, just reductive icons, just evaluative icons, or both types of icons on a set of products. Crowding information on food packages is not desirable, but often is the case in the marketplace today, he noted. What he and his colleagues tend to see when they conduct this sort of study is that the interaction between the evaluative icon and objective product healthfulness has a significant effect on purchase intention. This means that when the objective nutrition level of a product is high, adding an evaluative icon to the front of the package, such as the Model Front-of-Package Symbol System (IOM, 2012), increases overall purchase intention, he explained. The opposite occurs when the objective nutrition level of a product is low: adding an evaluative icon on the front of the package reduces people’s purchase intentions.
In summary, Burton said the FDA and other groups are interested in the potential effectiveness of different front-of-package formats (Andrews et al., 2014; Newman et al., 2016). He asserted that the recent research summarized above suggests that processing context should be considered when evaluating these formats: What situation is the consumer in? Is she or he evaluating a single product or deciding among a set of products? Relative to only a Nutrition Facts panel, Burton noted, there appears to be some value in both types of front-of-package icons. Generally, however, as the processing difficulty increases for the consumer, the importance of the evaluative, or interpretive, component becomes greater, he said.
Calorie Disclosures for Restaurant Chains
Burton went on to discuss some of the work he, Andrews, and other colleagues have conducted over the past 10 to 15 years on calorie disclosures for restaurant chains. Since the NLEA was passed, people have been eating away from home more often. According to Burton, American consumers now spend close to 50 percent of their total food budget on food prepared outside of the home, up from 25 percent in 1970.
Burton told the workshop audience how, a number of years ago when his children were young and before information about calories or nutrient levels for foods purchased in restaurant chains was publicly available, he would take his children to a certain restaurant once every 2 or 3 weeks. He would always order nachos from the appetizer section of the menu. He knew that the meal, with its cheese and sour cream, was not healthy. Still, when the nutrition information was made available online, he was surprised to learn that every time he ate his plate of nachos, he was eating 2 days’ worth of total fat, 3 days’ worth of saturated fat, and 1.5 days’ worth of sodium. These figures were far beyond his expectations about just how unhealthy the meal was.
By 2009, a growing number of states and localities were requiring disclosure of calories for restaurant chains. As Burton explained, however, different states and counties wanted different types of disclosures, leading to the 2010 passage of a national menu labeling law that overrode the state differences and established a single set of ground rules. The initiative was strongly supported by the National Restaurant Association, according to Burton, because of the importance of standardization for the industry. As of December 2016, all restaurant chains with more than 20 stores nationwide will be required to comply with this law.
Researchers who study calorie disclosures for restaurant chains consider a number of assumptions, Burton explained. The first is that there is a segment of consumers who want to make healthful food choices when eating outside the home, but there is another, generally larger segment who care little about the calorie and nutrient content of restaurant foods. Another assumption, Burton continued, is that many consumers will misestimate calories for foods they consume away from home, as he did with his nachos. Yet another assumption is that changes to the information environment could have the greatest impact on those consumers who not only are misestimating some items but also are motivated to make more healthful food choices. Finally, and key according to Burton, is that there are many possible interactions between these factors and contextual influences. While many laboratory studies have shown that providing calorie information affects what people say they will purchase, he noted, the actual effects vary among different segments of the population and in different contexts and conditions.
As an example of the type of study design Burton and his colleagues have used in their research on calorie disclosures for restaurant chains, he showed several items pulled from a dinner house restaurant menu. He explained how one group of consumers would see the items with no calorie information, while a second group would see the same menu but with objective calorie information added. A typical finding of this sort of study, he observed, is that people provided with calorie information choose meals that have, on average, about 250 fewer calories overall. However, when these kinds of data are examined using what is known as a health orientation value, the difference in calories in meals ordered does not become statistically significant until the health orientation reaches a value of about 5 on a 7-point scale.
Results from this type of study contrast with what has been observed in marketplaces in cities where menu labeling has existed for a few years (e.g., New York City, Philadelphia, Seattle), Burton explained (e.g., Elbel et al., 2009, 2011; Harnack and French, 2008; Long et al., 2015). This discrepancy is not too surprising, in his opinion, given that calorie labeling should have effects only for some items (i.e., those for which calories are
misestimated), for some people (i.e., those who are health literate, motivated, and knowledgeable), and for some occasions.
Burton used a funnel to illustrate the effect of the many factors that impact consumer food choices and consumption (Burton and Kees, 2012). The consumer segment that is actually impacted in a favorable way by calorie labeling is probably relatively small, he suggested.
In summary, Burton reiterated that many individual-difference variables and contextual factors impact, in this case, what a consumer orders for any particular dining occasion. Motivations for food consumed outside the home differ substantially across consumers, with different people valuing taste, amount or quantity, and convenience to varying degrees (Glanz et al., 1998; IFIC, 2011). According to Burton, low-calorie or health-related disclosures often lead to negative inferences about taste, satiation, and quantity, at least for some consumer segments. Moreover, he suggested, because taste, quantity, and convenience often trump nutrition and health information, calorie disclosures may actually lead to increased calorie consumption for some consumers. He referred workshop participants to Brian Wansink and colleagues’ research on sensory, emotional, and normative drivers of food consumption (Wansink, 2014; Wansink and Chandon, 2014). He suggested that, when considering effects of calorie disclosures for restaurant chains, policy makers and health researchers should consider chains’ reformulations and new product offerings as part of the aggregate effects of these market changes.
In conclusion, Burton emphasized, again, that consumers are exposed to a broad array of nutrition claims, icons, and information on a daily basis, resulting in many inferences and questionable conclusions. The effects can be large and not necessarily what is desired, he suggested. While nutrition disclosures and communications have many beneficial effects, he said, a variety of factors impact their overall effectiveness and, ultimately, consumption behavior. In his opinion, given the complicated and dynamic nature of today’s marketplace, there is much opportunity for future research.
Chester began by asserting, “We need to understand the narrative of digital media in our lives. If we are to understand and effectively respond to the dramatic changes that have transformed and will continue to [trans-
6 This section summarizes information presented by Mr. Chester.
form] how we live our lives in the digital era, including how we decide what products to consume, we need to pay attention to this transformation.”
Food and beverage companies are increasingly able to influence consumers throughout the day, both offline and online, according to Chester. Not just product producers but also retail and grocery stores are at the forefront of these changes, he noted. He stressed the global nature of the industry and what he described as the “very powerful” digital communication techniques being used to transform media, marketing, and sales and to trigger and influence brand loyalty and behavioral response. One of his takeaway messages was that this emerging system is growing more powerful and sophisticated every day, with ads targeting individuals based on what is known about their race, their income, where they live, and what they buy.
Chester elaborated on the many ways in which food and beverage companies have evolved to become more than product producers, marketers, and sellers, not the least of which is that they have become what he described as big data specialists. He cited @WalmartLabs, a Walmart spin-off in Silicon Valley, as an example. Its neuroscientists use the latest techniques and tools to better understand how the human mind works and how brands and product messages can be inserted into people’s unconscious minds. Food and beverage companies also are increasingly focusing on entertainment and information, Chester noted, specializing in creating experiences, activating individuals, and engaging in storytelling. Many have social media newsrooms with daily programs and, increasingly, online music channels. In Chester’s opinion, food and beverage companies are becoming “community organizers” as well, with highly developed social media marketing strategies that take advantage of consumers’ mobile and social relationships and locations. Finally, he observed, these companies have become venture capitalists in the new media, investing in startups to ensure that their brands and products are featured in the online apps that people, especially children, use. He also noted that, through their partnerships with the most powerful digital media companies, including Google, Facebook, and others, food and beverage companies have redefined shopper marketing in the 21st century. This reality must be addressed when one is thinking about the kinds of interventions that can help people live healthier lives, he emphasized.
According to Chester, one of the food and beverage companies leading the way is Mondelēz International. He described its 2014 Oreo campaign, which was a finalist for a 2015 Effie Award in advertising and marketing. The goal was to have each customer who bought one pack of Oreos buy a second pack. The ad targeted mothers with children younger than 12. The campaign addressed every touch point along the path to purchase, Chester observed. Mondelēz used social media (including Facebook), digital banner ads, and an online instant win game and promoted the campaign in a
weekly print circular and in digital coupons. It also used in-store sampling. Chester reported that sales grew 7 percent, with 105,000 people visiting the instant win game webpage and with data being collected on almost 67,000 of those visitors. Food and beverage companies are constantly creating new campaigns like this one, he said. They have responded to the growing availability of individualized data by investing significantly in highly sophisticated data management platforms and related services. Moreover, they are recognizing the opportunity to communicate with single individuals regardless of what device they are using. Chester cited Kellogg’s client relationship with Krux, a leading data marketing company, as another example of how food and beverage companies are using information about consumers in unique and powerful ways.
Advertising that uses super-fast computers to target individuals using their data profiles is known as programmatic advertising, Chester explained, and represents about 40 percent of the overall digital advertising market. He showed recent media headlines providing a sense of what some food and beverage companies are doing with such advertising: “How Kellogg’s Partners with Publishers on Programmatic” (May 2015); “Mondelēz Taps TubeMogul for Programmatic Video” (June 2014); “D3 Studios Is a New Digital Agency Serving Iconic Brands with the Frito-Lay Portfolio” (August 2015); “WFA Releases Programmatic Media Guidelines for Brands and Unveils Coca-Cola, Johnson & Johnson, Boehringer, Mastercard as Part of ‘Taskforce’ to Drive Take-Up” (September 2014); and “Here’s How Unilever Leverages Programmatic Buying for All-Inclusive Mobile Push” (February 2014).
Chester observed that programmatic advertising can be used to either target or reject a consumer in milliseconds based on collected data indicating whether the consumer has shown an interest in a certain product. Children aged 12 and younger are protected by 1998 federal legislation—the Children’s Online Privacy Protection Act (COPPA), but sadly, he said, there is no real privacy legislation for individuals aged 13 and over. These new forms of data targeting to sell food and beverage products also are being aimed at Hispanics, African Americans, Asian Americans, and low-income consumers, he noted.
In addition to programmatic marketing, food and beverage companies have invested heavily in what is called “neuromarketing,” Chester continued. They are using such techniques as functional magnetic resonance imaging (fMRI), electroencephalograms (EEGs), eye tracking, and galvanic skin responses to ensure that their messages reach the subconscious parts of consumers’ minds. The goal, Chester said, is to increase dopamine levels in individual consumers, and he described neuromarketing as a global phenomenon that has taken off in the past 5 to 6 years.
“Social media surveillance” is the term Chester used to describe the
way food and beverage companies are continuously tracking everything consumers say and do on social media sites such as Facebook, Twitter, and YouTube. They are analyzing the flow of information and immediately responding by creating countermessages and strategies to program the social media environment, he explained. Because consumers have become dependent on mobile phones, he said, companies are able to reach them at anytime, anywhere.
Additionally, Chester observed, companies are using their collected data to identify and map consumers’ locations and track their movements. They know when a mother goes to school, to the playground, and to the grocery store, he said, and they are using what is known as hyperlocal targeting, or advertising, to send coupons in real time. A woman may be driving down the street, for example, on the way to a competitor’s store, and because she has downloaded a particular app on her mobile phone, a company can send her a coupon in real time to encourage her to shop across the street instead. Even inside stores, Chester noted, mobile coupons can pop up and direct a consumer to a certain aisle. “All of this is happening,” he said. “None of this is science fiction.” Industry use of mobile coupons needs to be addressed, he suggested.
Increasingly, moreover, consumers will be making payments through their mobile devices. When Apple Pay was first introduced, McDonald’s was a partner. In the near future, Chester predicted, a consumer will be able to order food at the press of a button such that payment and waiting time will be seamless.
Young people in particular are being targeted by this “incredibly powerful digital marketing machine,” according to Chester. As an example, he showed a video of Walmart working with Coca-Cola to target teens through an award-winning campaign that involved the use of digital media. And it is not only teens who are being targeted, he said. When Google launched its YouTube app for children under 5 years of age in early 2015, it was filled with ads, including many for foods and beverages. In fact, Chester said, a whole new generation of YouTube celebrities is promoting fast food and beverages. As an example, he showed a YouTube video in which one of these new celebrities, EvanTube, promotes new flavors of Pringles potato chips (https://www.youtube.com/watch?v=KVDHl26K7yQ [accessed March 17, 2016]). The video is not a 30-second commercial, but a 19-minute segment. According to Chester, companies are not just selling products; they are creating environments that nurture deep brand loyalty. “This is what we are up against now,” he said.
In concluding, Chester suggested the need for new rules governing cross-platform marketing to children, policies to protect adolescents, fair marketing practices for the digital era, enforcement of COPPA, and effec-
tive self-regulation enforcement. He stressed, again, “This is something that we really need to address now.”
The Dietary Guidelines for Americans (DGA) is one of the major policy documents influencing much of what is done to promote healthy food environments and food literacy in the United States, Kraak began. The DGA in use at the time of this workshop was the 2010 version, with the 2015 version expected to be released by December 2015. Another major policy document influencing the promotion of healthy food environments and food literacy, Kraak continued, is the U.S. Department of Agriculture’s (USDA’s) MyPlate, which is based on the DGA. MyPlate not only sends about eight different messages, from “make half your plate fruits and vegetables” to “switch to skim or 1% milk,” but also is available in several versions, including a Spanish version, a children’s version, and a SuperTracker version for people who want to individualize the recommendations. Added to these, noted Kraak, is the Harvard School of Public Health’s Healthy Eating Plate version of MyPlate. So just with the DGA and MyPlate alone, she said, “We have a lot of diet-related messages.”
In addition to the DGA and MyPlate, observed Kraak, is the wide range of food and beverage product and restaurant menu labeling, plus the Nutrition Facts panel. Not only are there very large numbers of messages, she suggested, but the Nutrition Facts panel is highly numeracy based. The Facts up Front labeling system launched in 2011 by the Grocery Manufacturers Association and the Food Marketing Institute also is a numbers-based system. Yet, Kraak remarked, as Cynthia Baur had discussed (see the summary of her presentation in Chapter 1), a high percentage of U.S. adult Americans struggle with numeracy.
Kraak asserted that the crowded messaging environment is made even more so by the growing number of sustainability, eco-friendly, and ethical food labels as more consumers seek to know who produced their food and how. And added to these numerous different messages and labels in the U.S. food information environment, she continued, are the many brand mascots and media characters being used to market to children. She noted that of the 14 companies examined that participated in the industry self-regulatory program called the Children’s Food and Beverage Advertising Initiative (CFBAI), not one has yet pledged to align its brand mascot with uniform nutrition criteria. This is an area in which companies could be
7 This section summarizes information presented by Dr. Kraak.
pressed to perform better as a matter of corporate responsibility, she suggested. In addition to brand mascots, entertainment companies license their cartoon media characters to food, beverage, and restaurant companies for marketing products to children and teens. While some companies have made voluntary pledges under the CFBAI, the use of these characters on packages and toy premiums in food retail settings is not covered by these pledges, Kraak observed. Again, industry can do important work to improve the healthfulness of the food marketing landscape for young people, she suggested. She described celebrity endorsements of foods and beverages, including restaurant meals, as “the wild, wild west.” Again, she said, most companies have made no voluntary pledges to align their celebrity endorsements with healthy criteria that target adolescents in particular.
In Kraak’s opinion, the very crowded food and beverage messaging environment calls for comprehensive, consistent, and smart policies—not necessarily new policies, but revisions to existing ones. She stressed the need to disincentivize the marketing of products that are unhealthy and to incentivize and increase the production of healthy food and beverage products.
Public Policy and Food Literacy
What Harold Lasswell said about the difference between politics and policy some 80 years ago still holds true today, in Kraak’s opinion. That is, “politics is a process of who gets what, when, and how,” while policy is a law, procedure, or standard that dictates and guides how government, businesses, and organizations operate and how citizens live (Lasswell, 1936). Public policy, Kraak continued, is what public officials in government, and the citizens they represent, choose to do or not do about public problems. Importantly, public policy is not a linear process, she emphasized. Rather, it is iterative, with policies sometimes taking one step forward, then two steps back, she explained.
Another important point to keep in mind, in Kraak’s opinion, is that most researchers in the field of food and nutrition policy spend their time thinking about the details of the issues (UK Food Ethics Council, 2010): Exactly what happens? Is the problem trans fat? Is it sugar? Is it salt? Is it food insecurity or obesity? Kraak urged more focus on the rules of engagement and the terms of the debate, that is, why and how things happen (UK Food Ethics Council, 2010). She emphasized that there also are many stakeholders at different levels, with different interests in food and nutrition issues and with varying levels of power and influence regarding how funding is used (Bryson et al., 2011).
Finally, Kraak emphasized that both public- and private-sector policies impact food, nutrition, health, and media literacy (see Figure 2-4). Public-sector policies are important, she said, “but we also need to be reaching
FIGURE 2-4 The food, nutrition, health, and media nexus.
SOURCE: Presented by Vivica Kraak on September 3, 2015.
out to the private sector to say, ‘You could be doing a lot more with your voluntary efforts in order to be marketing food and beverage products that support a healthy diet much better than you currently are doing.’”
When Science Clashes with Public Opinion
Kraak posed the question, “What do we do when science clashes with public opinion?” For example, many authoritative bodies—including the FDA; the National Academies of Sciences, Engineering, and Medicine; the European Commission; the World Health Organization (WHO); the American Medical Association (AMA); and the American Association for the Advancement of Science (AAAS)—say that GMO foods are safe to consume. Yet according to a 2014 study conducted by Health Focus International (Watson, 2015), 87 percent of consumers think non-GMO foods are healthier. Although there is some variation in this regard among Americans, Europeans, and Asians, Kraak noted, GMOs still are ranked among the top five issues of concern among global shoppers, and a whole industry has developed around non-GMO-verified labeling. She reported that while 2013 annual sales of foods labeled as non-GMO amounted to just over $3
billion, the sale of third-party-verified products represented about $8.5 billion in annual sales (Mayer, 2015). Now there are even third-party auditors, such as FoodChain ID, that advise on and certify non-GMO food products for companies. Additionally, some food companies use non-GMO food marketing as part of their product marketing profiles, according to Kraak.
The gap between what eight different institutions have said about the safety of GMO foods and the growing industry around non-GMO food products demands attention, in Kraak’s opinion. For her, the question is how this disconnect can be addressed with policy. The state of Vermont passed a GMO food labeling law in 2014 that would have required such labeling beginning in 2016. In July 2015, however, the U.S. House of Representatives passed H.R. 1599, the Safe and Accurate Food Labeling Act of 2015, which, if it becomes law, will supersede the Vermont law and give the FDA the authority to mandate labeling for foods with altered nutrition profiles, such as the presence of allergens, and give it the authority not to mandate GMO food labeling. Kraak noted that supporters of the “Just Label It!” campaign call H.R. 1599 the “Denying Americans the Right to Know” (DARK) Act. In her opinion, what is “really” going on, which she said was well expressed in a recent National Public Radio (NPR) editorial (Lowe, 2015), is that GMOs are a proxy for other uncertainties that people have about the food environment. She mentioned local and regional food systems as examples to counter consumer concerns, and urged greater consideration of these concerns.
Cultural Differences Between Researchers and Policy Makers
Political decision making is very different from scientific decision making (Brownson et al., 2006), Kraak explained. She listed the most important things scientists need to know about policy making if they want to be effective in transmitting their messages (Tyler, 2013):
- Formulating policy is difficult, and no policy will ever be perfect.
- Policy makers are not a homogeneous group and can be experts, too. Many have Ph.D.s and other professional credentials.
- Policy decisions are subject to extensive scrutiny, which is why they are sometimes watered down in their final form.
- Developing policies from scratch is rarely an option for policy makers. Usually, they must build on considerable work done by others.
- Economics and law, not health and nutrition, are priorities in policy advice.
- Public opinion matters to policy makers, particularly if they want to get reelected.
- Again, policy and politics are not the same thing.
- Policy and science operate on different time scales (Brownson et al., 2006). Political decision making has a much shorter timeline, during which policy makers must make political judgments and build support so they can get reelected. Scientific researchers work over a much longer timeframe and can look at issues in a very different way.
- As Kraak had mentioned previously, policy making is an iterative, not linear, process.
- Policy making is an evolving art and science.
- Policy makers are not interested in science unless it helps them make better decisions.
- “We need more research” is the wrong answer. That is not what policy makers want to hear, according to Kraak. She said, “You have to give them the best available evidence, not the best possible evidence.”
With respect to how policy makers use scientific evidence, Kraak emphasized that they have a different perception of the value of research for informing policy. “We need to be aware of that and appreciate and work with that consideration,” she said. Policy makers sometimes may use scientific research, such as systematic evidence reviews, cost–benefit analyses, and modeling, she acknowledged, but they also use a great deal of nonresearch evidence. And while they rely on expert opinions, she said, they also rely on the opinions of the public and their constituents, as well as local knowledge, information about political feasibility, personal stories and anecdotes, and political principles.
Kraak encouraged researchers to understand the diverse roles they can play in the policy process and to take advantage of political windows of opportunity to become involved. She cited qualitative research with policy elites in Australia, in which Haynes and colleagues (2011) found that policy makers use expert advice from researchers to galvanize ideas, clarify and advise on issues, persuade others, and defend positions. Policy makers seek robust dialogue and creative thinking from experts and value expert opinion even when research is limited, the researchers concluded.
Additionally, Kraak encouraged scientists to learn how to communicate policy-relevant information effectively and to cultivate relationships with legislative staffers. She urged that they share their personal experiences and anecdotes; provide brief, 1-page fact sheets, not 20-page peer-reviewed articles; and cultivate political champions. She noted that the retirement of Senator Tom Harkin of Iowa, who served from 1985 to 2015, represented the loss of a champion for promoting nutrition and health. Finally, she urged researchers to ask policy-relevant questions and conduct policy-relevant research.
Kraak listed several ways to track policy-relevant research outcomes and measure their translation into practice (Wilsdon et al., 2015). In terms of dissemination, one can measure the number of times such information is shared through social media or mentioned in the popular press; the number of times it is viewed online, heard through podcasts, or downloaded; and the number of times audience members at events or exhibition viewers engage with the information. Additionally, one can measure how often the research is discussed in public debates; referenced by journalists; cited in reports from government, industry, foundations, and nongovernmental organizations (NGOs); and mentioned in legal arguments or used as evidence in case studies. Finally, one can measure how often such research is used by academics serving on corporate or NGO boards, government advisory committees, or professional organizations; used by researchers engaged in paid or contracted research; and used in teaching materials, taken up by professional organizations, or built on to improve performance.
To conclude her presentation, Kraak highlighted five key points:
- The current U.S. food environment is highly saturated with diet-related messages. It not only fosters “information overload” but also sends inconsistent messages about healthy, available, and affordable dietary choices.
- Smart, consistent, comprehensive policies are needed to transform unhealthy U.S. food environments and to support food, nutrition, health, and media literacy.
- These policies need to be based on coordinated input from many stakeholders. While it may be the government’s role to develop policy, all individuals need to voice their views for policies to be implemented and evaluated.
- Scientists need to be aware that policy makers value both scientific and nonscientific evidence.
- Scientific research can support policy development, implementation, and evaluation.
Levitt presented three case studies of gaps between public policy and public perception. These case studies were based mainly on his experience
8 This section summarizes information presented by Mr. Levitt.
working with the FDA, which began in 1978. Through the 1980s, he spent most of his time in the FDA commissioner’s office. He finished his 25-year tenure as director of the Center for Food Safety and Applied Nutrition (CFSAN).
Alar and Apples
Levitt’s favorite case study of the gap between public policy and public perception, he said, is a food safety example involving Alar and apples in the late 1980s. At the time, the greatest safety concern was cancer in the food supply, and the FDA was trying to determine how to implement the Delaney clause.9 Understandably, Levitt remarked, public anxiety on the subject was high. Alar was a chemical used as a ripening agent to help apples maintain their color and keep them on the tree longer. But it was regulated by the EPA as a pesticide. A 60 Minutes exposé, followed by a series of public appearances by Meryl Streep telling people to “throw away your apples,” fueled growing public concern that Alar was potentially carcinogenic, Levitt noted. In response to this growing public concern, the EPA, which had conducted the usual risk assessments and found that Alar posed no risk, contacted the FDA, which declared, “Apples are safe.” But the public did not listen. Instead, there was what Levitt described as “public hysteria” around Alar and apples.
Reacting to this response, the FDA invited Ned Groth of Consumers Union to the commissioner’s office. Levitt told the workshop audience how he still remembers Groth’s presentation more than 20 years later. Groth talked about Sandman’s (1987) “consumer outrage factors.” The number one factor contributing to consumer outrage, according to Sandman and as conveyed by Groth, is whether something is intentionally added or naturally occurring. Alar was intentionally added. Groth recalled speaking with an NBC News reporter, and being quoted on the Evening News by Tom Brokaw as saying that acrylamide occurred naturally as a result of the cooking process. Although Groth spoke with the reporter for about 20 minutes, the show capitalized on that one statement, Levitt said. The public sentiment was, “It occurs naturally . . . we cannot blame anyone . . . there is no victim . . . there is no villain.” Acrylamide never triggered massive public outrage.
Levitt explained that the number two factor contributing to consumer outrage, according to Sandman (1987) and as conveyed by Groth, is whether something is transparent or hidden. Because pesticides are used on a farm and not seen by consumers, they are, Levitt said, “on the wrong
9 The Delany clause, part of the Food Additives Amendment of 1958, bans the use of chemicals in food that are known to be carcinogenic.
side of the ledger.” Other factors contributing to the outrage over Alar, he suggested, were the fact that apples are fed to a vulnerable population—children—and that cancer is considered a very serious risk.
The number one lesson for the FDA from the Alar experience, according to Levitt, was that science by itself cannot control public perception; other factors come into play as well.
Shortly after the Alar case, another similar situation arose. Levitt recalled, “We all sat around and said, ‘Here we go again. What can we do differently?’” First, the FDA distributed as much information as it could, as quickly as it could. Because of that transparency and because, as Levitt said, “there was nothing diabolical seemingly going on behind the scenes,” the public did not display the outrage that characterized the Alar case. Based on this experience, he concluded that consumer outrage factors can be managed with some thought, although sometimes, he said, they “creep up on you,” and policy makers need to take them into account.
Food Nutrition Labels in the Early 1990s
Levitt was involved with development of the current Nutrition Facts panel, which is now more than two decades old. Today, the label is viewed as an icon, but it also produced some unintended consequences, he said. He noted that it was based largely on reports by the surgeon general and other experts declaring, in essence, “fat is bad, eat less.” The FDA heard that message and tried to implement it through policy by highlighting fat on the Nutrition Facts panel and specifying total fat, saturated fat, and calories from fat. “Miraculously,” Levitt said, the food industry responded and developed a large number of low-fat, reduced-fat, and fat-free products. However, he observed, no one had really thought about the consequences of removing fat from foods—that to market those foods, the fat would need to be replaced with something else that would provide full texture and taste. That replacement was carbohydrates, which have a great deal of sugar, and hence calories, so many of the new low-fat products on the market ended up being higher in calories than their high-fat counterparts. As a result, people who were buying low-fat products because they were healthier ended up gaining weight instead.
Based on this history, Levitt reported, the FDA was proposing to change the Nutrition Facts label (at the time of this workshop), placing more emphasis on total calories and providing a more nuanced fat message. So the landscape has shifted, he observed, but the shift has been slow. It has been almost 25 years since the current Nutrition Facts label, with its emphasis on fat, was developed. The lesson learned, Levitt said, is that it is important to think not only about intended consequences but also about potential unintended consequences.
Foods Derived Through Biotechnology
The difference between Alar and food biotechnology, in Levitt’s opinion, is that at least a serious potential risk was associated with Alar. But that is not the case with food biotechnology, he said. He stated that virtually every major reputable U.S. scientific organization is convinced that biotechnology-derived foods are safe. But when Europe, with its very different perspective, banned importation of foods derived through biotechnology in around 2000, the issue “mushroomed,” he recalled. At the time, he was CFSAN director. Several of Sandman’s (1987) consumer outrage factors were at play, he observed. The biotechnology was hidden. It was intentional, as opposed to naturally occurring. And while the associated risk was not considered serious, Levitt noted, an unknown risk can be as bad as or even worse than a known serious risk in people’s minds. In addition, he said, foods derived from biotechnology had a dedicated counter-force—environmentalists, who effectively branded them in a negative way, initially as “frankenfoods” and later as “genetically modified organisms.” He explained that the FDA conducted some focus group studies and found that people viewed the term “genetically modified organism” as pejorative. Today, a decade later, the term is still used, and in fact has become part of the lexicon.
Levitt explained that the FDA’s response to the increased concern about food biotechnology was to begin holding public meetings around the country. “You would be surprised,” he said. “You get a lot of public good will just by listening.” Many times Sandman’s public outrage factors arise because consumers feel that nobody is listening to them, he observed. In addition to the public meetings, the FDA proposed stronger regulation to make the review of any new plant varieties mandatory. A voluntary review system had been in place, but making it mandatory made it perceived as a stronger program, Levitt suggested. The agency also issued draft guidance on voluntary labeling for products that do not contain ingredients derived through biotechnology. Together, these steps “calmed things down,” Levitt said, as least for a period of time.
A couple of years ago, however (more than a decade later), public concern again emerged, Levitt continued. Although the scientific community remains confident that food biotechnology is safe, he suggested, consumers want transparency. There has been some state legislation on labeling, he noted, and federal legislation is pending.
Biotechnology in pharmaceuticals, in contrast to that in foods, was popular with the public, Levitt said, as was the case with medical diagnostics. The only difference, in his opinion, is the consumer benefit provided by those products. Pharmaceuticals and diagnostics give consumers something better than what they had before, he said. When offered a life-saving drug,
no one asks, “Was it genetically engineered?” By contrast, Levitt argued, biotechnology in foods provides no consumer benefit (it is viewed as a farmer’s tool, not something that improves consumer consumption), so the unknown risk weighs heavily in consumers’ minds.
To conclude his remarks, Levitt emphasized several points:
- Be simple. He recalled that when the FDA was developing the first food nutrition label, there was a great deal of discussion about how the label should be structured and how much information it should offer. Most consumers want something very simple, clear, and direct, Levitt suggested, and while the label that was developed had other issues, it definitely accomplished simplicity.
- It is difficult to explain complex issues. Levitt recalled someone at one of the food biotechnology meetings standing up and asking, “How can you tell me that adding a gene from a bug to my food is good?” There is no 30-second answer to that question, he said, except to say that scientists agree it is safe.
- Try to learn what is really causing the gap between public policy and public perception. Levitt agreed with Kraak that with food biotechnology, other concerns enter the discussion—for example, issues around sustainable agriculture and locally grown foods. He suggested that these other issues make it difficult to identify the real underlying motives driving the anxiety.
Session 2 ended with a panel discussion among all the speakers, with members of the audience invited to ask questions. This section summarizes the discussion that took place. Also included here, in the first section (“The Influence of Celebrity Culture”), is a summary of the brief discussion that took place immediately following Timothy Caulfield’s presentation.
The Influence of Celebrity Culture
Following his presentation, Timothy Caulfield fielded many questions from the workshop audience. First, Sarah Roller, a session moderator, asked him about Angelina Jolie’s endorsement of breast cancer screening, which Roller considered a “fairly science-based choice” compared with the other examples cited by Caulfield. She asked where the “germ” of an idea begins
when a celebrity decides that he or she is going to take on a particular case. How do celebrities decide that “this” is going to be “my cause”?
Caulfield replied that Angelina Jolie is often perceived as an example of the positive, or constructive, impact that a celebrity can have on decision making. But the data are mixed regarding whether the so-called Jolie effect has been beneficial for women’s health. Regarding celebrity endorsement decisions in general, he said, “They are desperate . . . to lose weight, to stay looking young . . . they are under tremendous pressure.” In his opinion, that is where it starts—they are willing to try anything. Moreover, he noted, most celebrities have a team around them that reinforces their behavior. Then, he said, “it takes off.” Often there may be some data behind whatever it is the celebrities are endorsing. If the science is contested, Caulfield noted, as it has been for nonceliac gluten sensitivity, they “cherry-pick” and “latch on” to that little bit of supportive data.
There is a deliberate infrastructure in the food marketing realm that is spending millions and millions of dollars to influence celebrities to endorse, Chester pointed out. He asked Caulfield whether any researchers have been studying that infrastructure, or network, of specialist companies that are working across platforms to influence celebrities in a deliberate way. Caulfield replied that there is continuum of celebrity endorsements, with some celebrities, such as the actress Shailene Woodley, being genuinely interested in what they are communicating, while others, such as Beyoncé with Pepsi or Kim Kardashian, who was recently promoting a drug on Instagram, are clear endorsers as part of orchestrated industry moves. He agreed with Chester that the power of social media was being increasingly leveraged. A question now, he suggested, is how it can be stopped. Regulation will be difficult, in his opinion.
An audience member remarked on the growing number of pro-science celebrities on social media, such as Bill Nye the Science Guy, and opined that, while a “great start,” these personalities are “preaching to the choir” and do not have the same reach that someone like Gwyneth Paltrow has. She asked whether Caulfield agreed with that assessment and what kind of messaging channels and personalities are needed to get pro-science and pro-evidence-based nutrition messages to people who are not already receptive to these messages. Caulfield agreed, noting that he has observed the same trend in Canada. “I wish I had an answer,” he said, “but I don’t.”
Fergus Clydesdale, co-moderator of the session, added that, first, scientists need to agree on which messages to convey. Caulfield concurred and encouraged the scientific community to become involved in social media and to become part of the discussion. If they do not become involved, then the messaging is being left to Dr. Oz and other celebrities, he asserted.
An audience member expressed interest in Caulfield’s focus on organic food, which the audience member pointed out does have some environmen-
tal benefits. In his opinion, however, most people would agree that today’s biggest public health crisis is obesity. He asked about the celebrity impact on obesity and expressed concern that celebrity endorsements are damaging in the way they encourage widespread consumption of very unhealthful products. He asked how that damaging effect can be counterbalanced and why there is not more focus on what has been driving the obesity crisis, including the celebrity role. Caulfield noted that the examples he had used were simply good examples of the role of celebrity culture. He agreed that celebrity culture has played a role in the obesity crisis through the marketing of sports drinks, pop, and similar products. More important, in his opinion, celebrity culture has had a more subtle impact via its emphasis on short-term solutions, extreme approaches, and aesthetic goals. Preliminary data suggest, he noted, that an emphasis on those kinds of goals, as opposed to working toward wellness or enjoyment or health, is associated with a lower likelihood of success.
Linda Neuhauser commented on the many good experiences she has had working in Hollywood over the past 12 years. She described how she and her colleagues have used celebrities alongside experts and parents in a “seamless” way to present parenting messages that are realistic, interesting to people, and science-based. She suggested studying that kind of successful leveraging of pop culture. It can be done, she said, but “it has to be done in an artful way.” Caulfield agreed and noted some interesting research that has shown how a good narrative, for example in a documentary, can have a sustained impact on public perceptions.
More on Celebrity Culture and Obesity
During the panel discussion, an audience member reiterated that the greatest food-related challenge for the Food and Nutrition Board is obesity. Noting that dozens of celebrities speak on behalf of Coke, Pepsi, McDonald’s, and other companies, he asked, “Where is the outrage about the fueling of the obesity crisis by celebrities that are pushing those [products]?” In his opinion, some of what has been discussed, particularly organic foods and GMOs, are “straw men.” He noted with respect to organic foods that most reputable organizations are not calling for people to eat such foods because they have greater nutritional value; rather, the concern is pesticide residues in conventionally produced foods and a belief that organic food production is more environmentally sound. He said the issue is similar for GMOs: it is not that they are unhealthy to eat, but that because 90 percent of corn and soybeans being grown are GMOs, the use of glyphosate has skyrocketed, and glyphosate was recently listed as a carcinogen. Again, he said, “There is an environmental issue there that does not seem to be recognized.” Finally, with respect to Alar, he observed that
the EPA terminated its use in 1989 because its use was found to be associated with an unreasonable risk. A special EPA committee revisited the issue in 1992 and again found that it posed an unreasonable risk. Subsequently, a National Research Council committee issued a report, Pesticides in the Diets of Infants and Children (NRC, 1993), concluding that the EPA had not sufficiently protected children from pesticides. “We have spent a lot of time on issues other than the most immediate risks to health, including the obesity issue,” he asserted. He asked the panelists how celebrities could be used to address the obesity crisis, rather than setting up strawmen.
Kraak replied that the first step would be to identify which celebrities have endorsement deals with which companies. She observed that few studies of celebrity food and beverage product endorsements have been conducted in the United States. One, from the Rudd Center for Food Policy & Obesity (Connecticut), found that 80 to 90 percent of such endorsements are for sugar-sweetened beverages and fast foods. “We need to do something about that,” Kraak said. Additionally, she suggested strengthening the CFBAI. “They need to do much better than they currently are to protect children from the marketing of energy-dense and nutrient-poor food and beverage products,” she said. However, she predicted that new legislation or regulation is unlikely in the near future. She suggested pointing out the “good players” and encouraging those companies to make pledges and then their peer companies to make comparable pledges.
Kraak’s mention of the CFBAI prompted Chester to point out that getting it to do anything will require political pressure. Instead, he suggested pushing for legislation and conducting studies to expose the “invisible network of influencers” that is working online to promote food and beverage products. As he had elaborated during his presentation, he reiterated that food and beverage companies have engaged the services of specialist companies and have partnered with Google, YouTube, and others to make it appear as though friends and other people are promoting their products when in fact an orchestrated promotion is being conducted. He noted that on the day before this workshop, the Federal Trade Commission had sanctioned a company for engaging in this kind of practice for Xbox on YouTube. “Unfortunately,” Chester said, “the answer was ‘more disclosure.’” Burton remarked that lessons can be learned from what some federal agencies, such as the CDC and the FDA, have done with celebrities and antismoking efforts.
Kraak added that some celebrity promotions of healthy foods have recently been seen. She mentioned the Fruit and Vegetable Promotion (FNV) campaign, launched in February 2015 by the Partnership for a Healthier America in conjunction with Michelle Obama’s Let’s Move! campaign. The FNV campaign is using up to 20 celebrities to promote fruits and vegetables to teens and millennial mothers, according to Kraak.
However, that number is small compared with the 90 percent of celebrity endorsers who are promoting sugar-sweetened beverages and fast foods, she acknowledged. She reiterated the need to strengthen industry’s self-regulatory efforts and hold it accountable. Additionally, she suggested doing whatever is necessary in the media strategically to change corporate behavior. She also commented on the Got Milk? campaign, which ran for 20 years and featured many iconic celebrities. Yet milk consumption and milk sales went down over that 20-year period—an outcome she found “interesting.” She cautioned, “If we are going to be using celebrities for a good cause, we need to make sure we are not sending mixed messages to our target audience.”
For Cynthia Baur, the question around obesity is whether it is a food literacy or health literacy issue. She suggested that while knowledge and skills may play a part, other factors are likely at play, including an emotional component, that are not typically topics of research but are clearly important to people’s experiences. She encouraged examining these other factors.
The discussion of obesity, and Baur’s comment in particular, led Sonya Grier to add that one factor often missing from discussions of food literacy is critical analysis. There is a great deal of discussion of knowledge, but less about that knowledge can be used in a critical way, she observed. When celebrities endorse a product, consumers need to know how to interpret that endorsement. For example, does that celebrity really drink five sodas a day while staying so thin? Part of being food literate, Grier suggested, is being able to analyze things that do not seem reasonable and truthful and to make more informed decisions.
Celebrity Culture and the Consumer
Craig Lefebvre responded to the celebrity issue by saying, “Throw the celebrities out the window.” He compared focusing on celebrities to chasing butterflies. In his opinion, the research question is, “Who are the people who listen to celebrities and respond to celebrities?” Marketers know the answer to this question, he said, and that is why they use those celebrities. The challenge is not to identify what cues celebrities are using or how many Twitter followers they have, but which people are following the celebrities and then themselves advocating for a ban on GMOs, low-fat diets, or whatever the trend may be.
Funding for Food Literacy Interventions
An audience member observed that the hyperlocal targeting described by Chester during his presentation looked “incredibly labor- and cost-intensive.” He asked Chester whether policy makers or anyone in the
National Institutes of Health (NIH) or any other government agency is working on developing counterstrategies and if so, whether they will need to partner with corporations to acquire the funding necessary to implement such strategies.
Chester reiterated that none of what he had described regarding the use of data and social media marketing was new. Food and beverage companies have been using these techniques for 4 or 5 years, he noted. In his opinion, the field could be doing a better job of tracking such developments and should have intervened earlier. He warned that the capacity to influence consumers to the extent that Coca-Cola and Mondelēz do requires a considerable amount of money. That said, the technologies these companies are using are “off the shelf” and relatively inexpensive. Chester suggested that counterstrategies could be developed, and that more research into the use of social media by these companies is needed to better understand how they are employing the technologies and determine the best set of counterstrategies. He would like to see the Federal Trade Commission subpoena food and beverage marketers to provide their data so researchers can study them and understand the marketers’ networks of influence, including who is being targeted and why. He reiterated that the “YouTube celebrity” is a “whole new version of celebrity,” some having millions of followers.
Additionally, Chester urged thinking about more holistic counterstrategies, that is, strategies that would counter not just what food and beverage companies are doing, but also what credit card, pharmaceutical, and other companies are doing. “It is the same system, same set of forces, at work,” he said.
An audience member questioned whether the digital marketing taking place is really new. “I think back to the old West,” she said, “and the guy that is out there selling the snake oil from the back of his carriage.” He knew where he could make a sale and was targeting communities in the same way. In response, Chester replied that although there is a continuum, today’s marketing is “a different kind of marketing.” He noted that many of the technologies being used by marketers are the same as those being used by the National Security Agency. The system, he said, “follows you wherever you go. It observes what you do with your friends and what they do with you. It analyzes you. It targets you everywhere, and it makes decisions about your future in milliseconds without allowing you to participate. It immerses and embeds you in a marketing commercial system that we have never seen before.” He referred workshop participants to a website, www.digitalads.org, that is based at the Berkeley Media Studies Group in Berkeley, California. This website catalogues what food and beverage companies have been doing with digital marketing over the past 10 years.
Thinking More Holistically
Hallman observed that one of the impressive things about the workshop thus far was that the various speakers had addressed the food literacy issue from slightly different perspectives, but that no one had contradicted anyone else. He asked, “Why don’t we have a more holistic perspective?” The answer, he said, is “because [the issue] is really complicated.” The easiest thing for a researcher to do is experiment with individual decisions, he noted, not environmental changes. Not only is it more difficult to change and evaluate environments, he suggested, but it also is more expensive. Indeed, sometimes evaluating whether an environmental change intervention has worked is more expensive than the actual intervention. To illustrate, Hallman mentioned a farmers’ market at Rutgers University, which he helped start and which is funded by Johnson & Johnson, that adds value to Supplemental Nutrition Assistance Program (SNAP) and Special Supplemental Nutrition Program for Women, Infants, and Children (WIC) vouchers. Whether the added value is an effective incentive in terms of changing behavior is unclear. To answer that question, he said, would cost more than the actual intervention. He and his collaborators would rather spend that money on purchasing more fresh fruits and vegetables for their clients.
Research on Food Well-Being
Clydesdale commented on the outstanding studies on the effects, or lack thereof, of front-of-package labeling, described by Andrews and Burton. However, he wondered why researchers are not spending more time examining food well-being. Grier replied that consumers respond to the notion of food well-being and that it should be promoted. She noted that, as Burton had stated during his presentation, different consumers have different goals, and a food well-being framework accommodates that variation. Andrews agreed that food well-being is a good model to keep in mind, but he cautioned that operationalizing it and understanding how it can be most effective will be challenging. He suggested that the number of factors involved makes the concept highly complex. Burton added that developers of the Nutrition Facts panel faced the reality that cues, formats, and the panel itself have different effects on different populations—the panel helps some people, but not everyone, he said.
Starting Where the Audience Is
Referring to the emphasis of several speakers on the importance of “starting where your audience is,” Lefebvre commented that many people want to eat healthy foods so as to be energetic. “I don’t hear us talking
about energy,” he said. When he raised this issue in the past with federal government agencies, the immediate response was that there is no science behind the claim that people are more energetic when they eat healthier foods. In his opinion, it is an “interesting conundrum” that consumers want to eat healthier foods to be more energetic, yet “the system” does not allow for any discussion of the issue because the relevant research has not been funded.
Lefebvre referred to Andrews’s observation during his presentation that people sometimes misinterpret messages about “healthier” foods, and recalled reading recently in the newspaper that Campbell’s was making healthier soups by reducing sodium, but was unable to gain any traction in sales from this effort. He compared the situation to what happened with the Got Milk? campaign. He asked the panelists how they would advise Campbell’s on ways to increase sales of their healthier soups.
Kraak observed that Campbell’s is the parent company for Bolthouse Farms, which has been highly innovative in promoting fruits and vegetables and fruit and vegetable drinks. While Campbell’s appears to be reorganizing and rebranding itself as a producer of healthier foods, it also has been subject to pressure from its board of directors to generate more revenue. That its low-sodium soups are not great sellers creates an argument, in Kraak’s opinion, for a more aggressive national salt reduction initiative in the United States. She noted that England and Australia have made great progress in getting their industry players to meet specific salt-reduction guidelines voluntarily. If Campbell’s is the only company reducing sodium levels in soups, people will not want to buy that brand. “They are going to stick with the brands that taste good,” she said. Such change needs to be made across the board, in her opinion, with an accountability mechanism in place to ensure that companies that fail to participate pay some sort of penalty.
Burton acknowledged that, not just with sodium but with other nutrients as well, taste has been a challenge for many years for companies trying to sell healthier products. Based on a previous study (Burton et al., 2014), he added that in general, people significantly underestimate sodium levels in restaurant meals. While some special populations are aware of and concerned about the issue, it does not resonate with most consumer segments. Burton suggested that when something like a low-sodium product is being marketed and there are no other interventions to encourage people to be aware of the issue addressed by the product and its consequences, the company doing the marketing needs to think about how to combine that message with others that do resonate with consumers.
Family, Community, and Organizational Influences on Food Literacy
Neuhauser observed that most of the workshop discussion thus far had focused on either individual- or macro-level (e.g., public policy) responses and barriers to food literacy. She asked the panelists about their thoughts on the influences of family, community, and trusted organizations (e.g., schools, health care organizations, faith-based organizations). She was particularly curious about how those influences might be effective with respect to interventions.
Grier stressed the importance of understanding the self-interest of these organizations and using that understanding to engage them in promoting messages or participating in interventions. Faith-based organizations have very different self-interests from, for example, those of mothers’ blogger groups. A mistake people often make when trying to get organizations to collaborate or partner, she suggested, is not understanding their needs and how an intervention can help them, not just how their collaboration or partnership can help the intervention. She mentioned having recently heard about a mothers’ blogger group with more than 12,000 members exerting a great deal of power in Washington, DC, in terms of promoting or shaming products. If an intervention beneficial to children also satisfies the goals of an organization, as it would such a blogger group, collaboration will be more viable and sustainable for both partners, she suggested. The same is true, in her opinion, in working with companies.
The blogger group Grier mentioned is called MomsRising, Squires noted. She suggested that a lesson to be learned from political campaigns is that seniors, people with type 2 diabetes, or any other group can do what MomsRising is doing. “That’s the beauty of the Internet,” she said. One can share information through social media and “galvanize around anything.” Additionally, Squires mentioned SparkPeople.com and its 14 million unique users who are focused on losing or keeping weight off. Joining is free, its users are having online conversations, and they are really helping each other, she explained. She mentioned that she herself uses a Fitbit, and while the company that makes Fitbit probably has a great deal of information about her that she wished it did not have, the way its product is quantifying what users are doing for physical activity provides reinforcement or helps them change their behavior in a positive way.
Byrd-Bredbenner agreed that social networks are important and observed that they can be large, like MomsRising, but they also can be very small—for example, within a family unit. According to Byrd-Bredbenner, good data indicate that such networks do influence behaviors. Thus they represent, in her opinion, an opportunity for health communicators to provide accurate information to be shared within the networks. She mentioned
some recent work with school wellness policies and how that information has helped advance nutrition.
More generally, Byrd-Bredbenner agreed with Neuhauser that health communicators need to think more about the environments in which people live. “For too long, we have put the onus on the individual to make the changes,” she said. In her opinion, although health communicators are beginning to think about environmental factors, they have many opportunities to pay even more attention to such factors and to provide the information individuals need to make more informed decisions.
Chester remarked that stores nationwide are installing beacons that can track individuals’ movements—walking across the street to a McDonald’s, for instance, or inside a Walmart. He encouraged seizing the opportunity to go to churches, stores, and other neighborhood gathering places and explain to young people and others that they are going to be targeted in this way and can have a voice to “rewire this system in a different direction.”
An audience member questioned why more is not being done in schools to encourage health literacy and strengthen what she called “STEM acumen.” Baur responded that, although national health education standards exist, many health professionals do not know about or use them. Those standards, she noted, were designed to help develop skills, as opposed to specific topical knowledge, recognizing that teaching children “a bunch of facts” would not be helpful in the long run. Another part of the problem, she said, is that health education has been crowded out of the K–12 curriculum. Education is a local issue, she observed, and communities need to place a high value on health education and decide how their K–12 curricula can be used to deliver health education. People need to express a demand for their children to have a certain skill set with regard to their health and their bodies, she argued. “That has not happened,” she said. Baur suggested that this might be a good topic for a future National Academies of Sciences, Engineering, and Medicine workshop.
Hallman remarked that he frequently is asked to present talks on many different subjects at scientific conferences and that regardless of the subject, someone in the audience always asks why schoolchildren do not know more about it. The answer, he said, is also always the same: “We need to get this in the curriculum.” But that is a difficult thing to do, he suggested, requiring the will to give up something else in the curriculum. He agreed that health education is being crowded out of the K–12 curriculum, and even when it is included, food is not necessarily a priority.
“Is there a way we can dovetail what we want to teach?” Byrd-Bredbenner asked. Given that health education often is crowded out but science is still taught, she suggested using food and nutrition as examples when communicating about scientific concepts. She mentioned working on
a project in which she has been talking to editors of science textbooks to encourage them to do that.
Reflecting on what Squires had discussed about trust during her presentation, an audience member asked how nonprofit organizations can get their audiences to trust the information they are communicating. She asked whether it was better to get that information into the hands of doctors or to use social media as a platform for connecting directly with one’s audience. Squires replied that, sadly, nutrition education in medical schools has been lacking for several decades. However, she has worked with a number of different medical groups, including doctors, nurses, and pharmacists who are seeking continuing medical education (CME) credits, and has found that working with those groups is an effective way to inform health professionals. She suggested that educating these “point people,” especially nurses, may be a way to get this information out. Additionally, social media could be used to amplify messages through different communities.
To close the first day of the workshop, Freedman was invited to share his reflections on the presentations and discussions that had taken place thus far. He began by suggesting that, if asked to define the problem workshop participants were trying to solve, he would define it as follows: “Figure out ways to support the public in receiving and embracing scientifically valid information about healthy eating in such a way it would lead to them making healthier choices and having healthier behaviors and, of course, ultimately being healthier.” Yet, he observed, immediately upon defining the problem, another problem emerges—what information? While most people attending this workshop would probably agree that the goal is to shift the public toward scientifically valid thinking, there would probably be much disagreement about just what science has actually “proven.” As reflected in scientific journals, scientists have come to very different conclusions, or conclusions that at least appear to be different, regarding what is causing the problem.
This is the case just with regard to obesity, Freedman argued. He agreed with other speakers that obesity is such a “killer” that it “swamps everything else.” Yet, he said, “we cannot come to any sort of agreement on exactly what it is that science says we should be doing about it.”
Freedman suggested pretending for a moment that experts can agree on a message. But then another problem arises: What can be done about it? Many people refer to the case of tobacco use and argue that something can be done about obesity because something was done about that problem.
10 This section summarizes information presented by Mr. Freedman.
Freedman pointed out, however, that in 1950, 40 percent of the American public smoked. Today, even after 60 years of the U.S. government, scientists, activists, and others sending the single, strong message that “smoking will kill you,” 20 percent of Americans still smoke. Smoking remains the single greatest behavioral killer, according to Freedman, although obesity has just about caught up and will probably surpass it. So even if experts can reach agreement on what message to send about obesity, he predicted it will take another 60 years of sending that message before the proportion of the American population that is overweight or obese can be reduced from two-thirds to one-third. “I think we really ought to plan on doing much better than that,” he said. In his opinion, what was done with tobacco is not a role model for what should be done with healthy eating.
The problem is highly complex, Freedman continued. He said he liked Grier’s discussion of the embeddedness of food in people’s lives and the many factors that cause them to make poor eating choices. The socialization element is particularly challenging, he said. As several speakers had discussed, decisions about food are not based solely on information, logic, cognition, or evidence; what he called the “lizard brain aspect” plays a role as well. If the problem is going to be solved, he suggested, it will need to be attacked from many different angles, including at the societal level.
Nor is the problem simply that people are making bad decisions that lead to unfortunate beliefs and choices, Freedman continued. As Baur had pointed out, the battle is against not just bad decisions but also no decisions. “A lot of people out there,” Freedman said, “could not care less about changing their food habits.” Those are the people who need the help, he suggested. He told the workshop audience that people who approach him for advice about eating are almost always people who are already healthy.
If what Byrd-Bredbenner had said about people actively seeking information is true, Freedman continued, “that is a good start.” Unfortunately, as several speakers had discussed, when people do seek information, they are more likely than not to encounter “some silly belief” based on pop messaging, marketing, or bad journalism. As Hallman had pointed out, the American population is poorly educated in general. The vast majority of Americans do not have a college degree, let alone a STEM education. Freedman referred to Kraak’s discussion of GMOs and how people get lost in the crowded messaging environment and end up with beliefs that directly conflict with science: while most scientists say that GMOs are safe, most people think they are deadly. “You could not have a more stark example,” Freedman stated. Moreover, he asserted, people become “locked into” these silly and scientifically invalid beliefs. He described these beliefs as “incredibly tenacious,” even “religious.” Based on his experience, this is especially
true with beliefs about food and health. He has found it easier to question people about their religious beliefs than about their food beliefs.
These beliefs and their tenacious nature “short circuit” efforts to provide good, useful information, Freedman explained. He noted that Baur had discussed how most people could not make sense of nutrition labels, and that Burton had described challenges associated with communicating calorie information on nutrition labels. But if someone has decided that calories do not matter, he asked, “What good does it do for us to get really good at giving out messages about what has more calories than something else?” The same is true with exercise, he suggested. If someone does not believe exercise matters, what good does it do to communicate solid information about ways to start and maintain an exercise program? The problem is not just that experts cannot help consumers win the race toward making healthy decisions, Freedman observed. It is much worse than that, in his opinion. It is not even clear yet how to get consumers into the stadium where the race is taking place. “That is how far behind we are on this right now,” Freedman said.
The question for Freedman is how credentialed experts—not just scientists, but others as well—can attack this problem. As Caulfield had elaborated, pop culture is winning the war right now. Celebrities such as Katy Perry and Michael Douglas are more influential than any of the credentialed experts who attended this workshop. This means it may not really matter what credentialed experts say, Freedman argued. As Baur had asked during her presentation, how can the gap between what experts say and what the public believes be bridged? Freedman suggested pretending for another moment that experts knew how to bridge that gap. But again, which message, and which experts?
Yet another problem, Freedman noted, is that consumers often treat journalists as experts. Many journalists are good at sounding what he described as “very sciency.” But, he suggested, if one looks at The New York Times, for example, and what some of its leading writers have written about science topics for The New York Times Magazine in particular, much of it is “delusional.” He mentioned Gary Taubes, Michael Pollan, Michael Moss, Tara Parker-Pope, and Mark Bittman. According to Freedman, Taubes has made such claims as “calories don’t matter.” Pollan has written such statements as “all processed food is obesogenic and toxic.” Moss has written about food companies plotting to addict the public chemically to food. Parker-Pope has written about the biological impossibility of sustained weight loss. And Bittman has claimed that a McDonald’s salad will kill a person, but home-cooked bacon is healthy. These messages are “absolutely swallowed up” by the Times’ readers, Freedman said. They sound rational and hit people emotionally. They resonate, and people love them. They have the intuitive plausibility Hallman had discussed, and they have none
of what Byrd-Bredbenner had called communication friction. These writers are “masters at lubricating stories,” Freedman said.
Freedman reiterated that the problem is big, important, and widespread, and it will be very difficult to fix. Even if one breaks off a small piece of the problem—for example, by examining nutrition claims on packages, as Andrews and Burton have done—a “minefield” of effects take place. Kraak had described the multiple versions of MyPlate, and even if the MyPlate message were simple, clear, and loved by everyone, it would still get lost in the food information environment. If one tries to break off a bigger piece of the problem—for example, food company marketing—the challenge is even greater, Freedman observed. “Regulation is so against most of what America believes,” he said, “it is almost silly to talk about that as a solution.” He suggested competing directly with industry messaging. He warned, however, as Chester had shown, that the food and beverage industry is extremely good at what it does, and competing with those companies will be difficult.
Freedman suggested further that the widespread notion that intense food marketing is “bad” is a flawed assumption. He mentioned having heard someone at this workshop recommend, as an alternative strategy that he views as brilliant, doing what is necessary to get food companies on the “right” side of the problem so that they are producing healthier options and then “turn[ing] them loose with all their marketing magic.”
In conclusion, Freedman emphasized that whatever message experts decide to send must be clear. It was sobering for him to hear Byrd-Bredbenner say that people think it is easier to do their taxes than to make healthy eating decisions. But again, even with a clear message, how far can one get with people who are not open to receiving that message? People are being bombarded at all times and from all directions with “silly” but understandable, intuitively plausible, and frictionless information, Freedman argued. While sending a single, clear, strong, understandable message may get one into the competition, he said, “it does not help you win it.” He also wondered whether sending such a message might corrupt science in the process. Levitt had touched on this issue in his discussion of the low-fat message, which was simple and clear and worked, but was wrong.
Despite the complexity of the food literacy problem and its challenges, Freedman believes it will be possible to agree on a message and propagate it in such a way that it actually makes a difference. He suggested that the workshop presentations and discussions thus far had done a good job of characterizing the nature of the problem. He expressed his hope that the second day of the workshop would focus on tools that can be used to attack the problem, and he encouraged moving beyond theorizing and beginning to develop hands-on approaches.