In the early 1900s, the American breakfast looked very different from what we think of as traditional today. Instead of artery-clogging platters of meat, eggs, and potatoes, most breakfasts consisted of little more than a roll, fruit, and coffee to wash it all down. As people left rural areas for the cities and for new opportunities brought about by the Industrial Revolution, they left heavy farm breakfasts behind them, too.
Those in the bacon business weren’t pleased. In the 1920s, Beech-Nut Packing Company contacted a man named Edward Bernays and asked him to find a way to get people to eat more of its product. Bernays went to a “famous New York doctor” and asked him, with a bit of a wink and a nudge, whether a hearty breakfast might be better for Americans than the paltry smattering of calories they were getting from coffee and cereal. The doctor said sure, why not? The body loses a lot of calories during sleep, and jump-starting the morning with a bigger meal would help get people going. Bernays had this doctor write to 5,000 physicians telling them, essentially, that the body lost energy during the night. Did they agree that eating a hearty breakfast would be a better way to recover?
“Obviously all of them — we got about 4,500 answers — all of them concurred,” Bernays said in a video for the Museum of Public Relations. He publicized the results of this survey to newspapers throughout the country, each one adopting a variation of the same headline: “4,500 physicians urge Americans to eat heavy breakfasts to improve their health.” The stories included the helpful note that there was no heartier meal than bacon and eggs. Rather than selling to the American public, Bernays let the public sell to itself. The idea that bacon and eggs is somehow good to eat first thing in the morning has been passed down through generations ever since. It’s so ingrained in American culture that while I can happily eat unusual lunches or dinners while visiting foreign countries, something feels off during breakfast when I’m served little more than a croissant in France or savory fish soup in Thailand.
You are not what you eat. Though you may have brought home bags of groceries from the store and put them together in the form of a meal (or accidentally left them to rot in the crisper like every other bag of salad I’ve ever purchased), your choices aren’t strictly your own. Bernays was only one of the first to realize how easily people can be swayed by the right kind of advertising — a kind that speaks to our unconscious minds and pulls at our emotions. It’s no coincidence that Bernays — these days known as the “father of public relations” to communications students across the country — was the nephew of Sigmund Freud.
In 2014, there was a huge outcry when it was revealed that Facebook had secretly tinkered with over half a million users’ News Feeds to make some of them see more negative posts and others more positive. The emotions were contagious — those shown happy posts got happier and vice versa. The reaction to Facebook’s psychological manipulations was uniformly swift and angry. Yet when a group of psychologists remodels a cafeteria or grocery store to see if it changes people’s eating habits, we hardly think twice about it.
Those types of changes have occurred for decades. Advertisers, economists, and even the government have been stepping in to shift what we put in our mouths. Sometimes these interventions mean well and try to make our eating more healthful. But not always. And as these “nudges,” as they’re known today, often appear in the form of advertising, there’s little incentive for any oversight on these forms of manipulation involving our meals and our minds. Especially when we’re the ones doing the manipulating.
There was something different about this Hardee’s in Champaign, Illinois. Notably, it had been split in two. One looked like every other Hardee’s in America, with utilitarian, spill-concealing floors and comfortable-but-not-too-comfortable metal chairs. Rectangles of fluorescent light flickered amid the ceiling panels. But on the other side, it was like a bizzaro fast food restaurant — a bistro. There were plants filtering indoor air through the restaurant and aesthetically pleasing art on the walls. Windows were tastefully covered with treatments while the light inside the restaurant was low and indirect. Romantic, even. Lunchtime diners were seated at tables covered in white tablecloth. There was no Top 40 filling the room — “jazz ballad instrumentals” played over the sound system instead.
As lunch goers were escorted into the restaurant, they were randomly seated in one of the two sides. Those in the classy jazz half had their food delivered to them at their table, and a waiter visited them to see if they’d like refills or to perhaps order another bite to eat. As the real diners ate, people surreptitiously placed around the room used stopwatches to record how long they stayed. Their orders were cataloged, calories were counted, and leftover food was weighed after the diners left.
Brian Wansink of Cornell’s Food and Brand Lab and Koert van Ittersum of the Netherlands’ University of Groningen, the coauthors of the experiment and resulting study, were surprised by what they found in the data from this restaurant. While both “sides” of the restaurant ordered similar amounts of food, the people with the upscale service ate less and lingered longer. Later analysis showed that people lost interest in their food when they took longer to eat it — perhaps because it takes roughly 20 minutes for your body to tell your brain that you feel full — but those in the fine-dining side also said the food tasted better despite it being, of course, from the exact same kitchen. A Hardee’s kitchen, no less. There were two important sides to these findings: The first is that restaurants that want better Yelp reviews should pay attention to their atmosphere; the second is that people who want to eat less might be able to do so if they just slow down or, as the study said, “The way to ‘have your cake and eat it too’ may be to enjoy the atmosphere instead of the cake.”
Even if you aren’t aware of the revolutionary effects of a field now known as behavioral economics, you may have been unwittingly nudged by it. The field started out as a way to use psychology to explain people’s economic decision-making. Over time, researchers realized that that same psychology could be used to prompt people to make one decision over another. Sometimes it’s the government trying to tweak what we eat or companies designing forms meant to assist us in purchasing a certain health insurance or retirement plan. In the world of food, behavioral economics is used to convince us to buy more, eat less, and even enjoy a fast food burger as much as a fine-dining experience.
Until about 50 years ago, economists treated the human players in their models as largely infallible, rational creatures. This so-called Economic Man was a careful decision-maker who was capable of weighing costs and benefits without pesky emotion getting in the way. The Economic Man would stop desiring a flashy car as the price rose (because, as anyone who has taken Economics 101 knows, demand falls as cost rises). The Economic Man would never have spent $100,000 on Beanie Babies and later gone bankrupt thanks to the plush toys — but real people did. There are probably still people somewhere with a closet full of Beanie Babies, waiting for the day the prices skyrocket again. These people are definitely not the Economic Man. But neither are any of us.
Today the most successful person in food psychology isn’t a shadowy advertising executive working behind the scenes; it’s Wansink, a quirky academic whose press photos often involve him pointing at zucchinis or throwing handfuls of popcorn gleefully into the air. He may not be a household name, but he is the puppet master behind many of the least painful weight-loss tips and tricks of the past two decades. It was the beginning of the Super Size Me era when Wansink decided to look at portion sizes. Namely, he wanted to know whether a person with a fixed amount of hunger would eat more food from a small package than they would from a big one. The answer was yes, and Wansink realized he could apply these findings to help people who wanted to go on a diet. In 1996, he published the results of the study and approached food companies with the simple idea that people might be willing to pay more for a diet-friendly food, even if it contained less actual food inside. Food companies were cautiously optimistic. Just one year later, Wansink founded the Food and Brand Lab at the University of Illinois (in 2005 it relocated to Cornell University). During its history it has become easily the most widely cited producer of insights into consumer behavior around food. After Wansink suggested that Kraft consider making foods with smaller serving sizes, the company introduced fun, 100-calorie packs in 2004. They were an immediate success, with over $75 million in sales in their first year.
A few months ago, Wansink made one of his many media appearances in a segment on Rachael Ray. He was armed with plates, Ziploc bags, and a color wheel ready to show audiences how changing their environment can help them lose weight without counting a single calorie. “The color of your kitchen can make you fat,” Wansink tells the audience in an exuberant voice, one that stammers when it can’t keep pace with his own excitement. “If it’s too bright you eat too fast, if it’s too dim what happens is you linger longer and you end up eating more.” (He recommends repainting the dining area with a neutral earth tone.) Wansink is a likable presence on television shows — he uses years of the Brand Lab’s studies to back up pithy weight-loss how-to guides, yet seems like someone you might meet at a barbeque. His bio lists the fact that his findings have been featured in articles in The New York Times and The Wall Street Journal as well as segments on shows like 20/20 and BBC News. It also describes him as a “former bad open-mic comic and rock sax player.” Even briefly watching him on television makes it clear that he is exactly the kind of man who would put those details into his biography.
Wansink’s interest in changing Americans’ eating habits was sparked during a boyhood trip to his aunt and uncle’s farm in Iowa. Usually the summer ended with the stuff idyllic childhood memories are made of: a trip to a movie and an ice cream at the local dairy bar. “But in 1968, grain prices were low,” Wansink wrote in his book Mindless Eating. “When I innocently asked Uncle Lester why we weren’t seeing a movie that year he summarized the state of agricultural economics in seven words, ‘we would if people ate more corn.’” This moment forever connected the joy of seeing movies and enjoying a cold ice cream cone with getting Americans to eat more farm-grown produce. He realized that hidden influences — like the price of corn — had a strong effect on his life.
As he describes in Mindless Eating, he went from a master’s degree in communication research to a Ph.D. program in consumer behavior at Stanford. Wansink told the Ph.D. program in his application that he wanted to research how to “get people to eat more vegetables” — corn or otherwise. His childhood observation had become a lifelong obsession.
Wansink’s growing influence was notable, but he wasn’t alone in studying what we eat.
There were food labs popping up throughout the United States, starting in the 1990s and picking up steam in the 2000s. Some of them, like the Drexel Food Lab, focus on helping the food industry by logging trends and creating new products. Others, like Penn State’s Laboratory for the Study of Human Ingestive Behavior, have been studying the causes of obesity. Then there’s the Mann Lab in Minnesota, where psychologist Traci Mann studies self-control and dieting.
Mann says it isn’t surprising that there are so many people studying food. “It’s surprising that there were so few before.” As a psychologist, she says, she and her colleagues study human behavior, and “eating is one of the most important behaviors that people do. You do it every day. You can’t not do it.” There are ties to every branch of psychology — studying eating disorders in clinical psych, studying kids’ eating habits in developmental psychology, and so on. Considering the government spends more than $70 billion a year on the Supplemental Nutrition Assistance Program, more commonly known as food stamps, policy makers and economists should also be invested in how people eat. “[Food] fits within every area and has been pretty ignored until 20 years ago,” Mann said. “It’s a very important topic that’s only starting to get the attention it deserves.”
Contrary to their titles, most of these “labs” aren’t quite what they sound like. There is no rush of white-coated scientists scurrying from one study to another; there are no people eating sandwiches while covered in electrode pads measuring their brainwaves. Wansink’s food lab is little more than a room set up to look like a household kitchen, normal in every way except for a two-way mirror and hidden cameras.
These labs publish most of their work in academic journals, and they often rely on the press to bring their findings to a consumer audience. But both Mann and Wansink have written popular books — Mann’s Secrets From the Eating Lab and Wansink’s Mindless Eating and Slim by Design — to help bring weight loss or healthy eating tips directly to the people who want them most.
They couldn’t have come at a better time. People were starting to get an inkling that most weight-loss diets were miserable, expensive, and unsustainable. As these studies were published and shared through the media, the general public cheered. Researchers had dropped in out of nowhere to tell everyone exactly what they wanted to hear: It takes 3,500 calories of food to equal a pound of weight, so why not skip the diet and painlessly cut out 500 calories a day, lose 52 pounds, and actually keep it off? The research was media friendly because the studies offered a rare bit of good news in the field of dieting — “get that summer beach bod just by rearranging your kitchen” reads all too familiarly like a headline in a women’s magazine. Whether you wanted to classify it as behavioral economics, nudging, or consumer psychology, groups throughout America saw this new way of looking at our diets as the lord and savior of battling the bulge. What few people seemed to be asking was whether it was really the best solution.
Over the years, it’s been relatively easy for the media to get in touch with Wansink. But recently, with the exception of that Rachael Ray segment (which was filmed in January), he’s been impossible to find. Requests to interview him get denied or handed off to other members of the staff at the Food and Brand Lab. His very active Twitter feed has collected dust since he tweeted out a blog post on February 13. The post clarified a seemingly innocuous tale of a Turkish Ph.D. student — a story he first told last November — who was given a data set to work on when she arrived at the Food and Brand Lab. The data was from “a self-funded, failed study which had null results,” Wansink wrote. In other words, the result didn’t support the lab’s hypothesis. But he didn’t want to give up on it. “This cost us a lot of time and our own money to collect,” Wansink told the student. “There’s got to be something here we can salvage because it’s a cool (rich & unique) data set.”
Cool is a word Wansink likes to use, and he uses it a lot. The number of times he repeated it when I interviewed him about one of his studies in 2016 was high enough to be memorable even a year later. In his book Slim by Design, he uses “cool” as a descriptor on at least 19 separate occasions. What Wansink probably didn’t find cool was the response to his post about the Ph.D. student, which was intended to be a life lesson in productivity for researchers. He wrote about how she submitted five papers only six months after beginning her work at his lab — wrangling four papers from just one data set. (The post, which has been archived, was quickly deleted from Wansink’s website.) To nonresearchers, it may have seemed like little more than advice on par with “work smarter not harder,” but to those in the academic community it sounded a lot like p-hacking, once fairly common in the social sciences and now a deeply frowned-upon practice.
In short, p-hacking is a practice in which a researcher takes a data set and “mines” it for a significant outcome. Say a researcher wants to prove that people who drink milk are healthier, and the initial results of their study are inconclusive (and would never get published in a prestigious journal). So the researcher changes the definition of health from a specific BMI range to a specific heart rate. That still doesn’t work, so then they try to measure health by how often the subjects exercise or how happy they are in their relationships instead. When the tweaking is finally “successful,” researchers are left with a “statistically significant” study that isn’t true. With enough p-hacking, data sets can be used to prove a lot of different points (as FiveThirtyEight proved with an interactive data set that allowed users to prove that having either Democrats or Republicans in office could be construed as “best for the economy”). Once it came out that researchers were commonly p-hacking, there was an outcry in the scientific community and over the past decade the practice has died down, though clearly it didn’t die out entirely.
After Wansink’s post, researchers went digging through his papers, finding 150 errors in four of them alone. Then they kept digging. As of today, people have found 45 publications that contain “minor to very serious issues” ranging from self-plagiarizing to data duplication, or outright data or statistical issues within the studies.
In April, Wansink posted a statement about research saying that “like all researchers, I welcome the careful review of my colleagues as they attempt to replicate and expand on my work. This is how science advances.” In cases where he found errors in past work, he wrote that he reached out to six journals “to alert editors to the situation” and developed a new set of standard operating procedures for his lab to ensure that these “oversights and errors” wouldn’t happen again. One of those papers was retracted. Cornell University also released a statement saying that it had conducted an internal review and that while it found “numerous instances of inappropriate data handling and statistical analyses … such errors did not constitute scientific misconduct.” When asked about the fallout from the retraction and data errors, Adam Brumberg, deputy director of the Food and Brand Lab, declined to answer any questions, referring me instead to the University and lab’s statements.
Last month, another round of major corrections was made to a 2005 study. Among the corrections are changes to tables, a change in the methodology description, and a note saying that because original data was no longer available in some cases, “the true values are unknown.” This article had been cited 81 times before being corrected. Despite the numerous data errors, the lab has maintained that the final conclusions still hold, which has left some researchers outraged. Paul Brookes, professor of anesthesiology at the University of Rochester, commented on the website Retraction Watch, “While science is indeed a search for the truth, the path to arrive at the truth matters.”
It’s one thing for a bacon company to convince the population of the United States that eggs and bacon are the perfect way to start the morning in the early 1900s. But shouldn’t that kind of thing be harder to do in 2017?
Andrew Gelman, a professor of statistics and political science at Columbia University, picked up on the news of the Wansink scandal quickly and has been following it since the first discrepancies appeared in December. “Some of it looks like actual fraud and other parts look like sloppiness and not caring,” he says. “But the idea that it’s been happening in plain sight all these years — it’s quite a story.”
If Wansink had been studying, say, climate change, there’s no way his work would have gone unchecked for so long, Gelman says. “A political science study claiming UN peacekeeping is effective is going to get critiqued,” but that’s not the case with a lot of food or psychology articles (and Wansink’s are, in many ways, both). Sites that write articles about his research see it as a fun tidbit of information — so what if it’s not totally correct? Readers will click nonetheless.
If food pushed to the back of the refrigerator often gets forgotten about until it goes bad, it stands to reason that putting fruit out on the countertop would make you more likely to eat it. That’s one of Wansink’s classic examples — and it reads like common sense. Most of Wansink’s studies have the advantage of feeling true to most people.
But the paternalism that we often let guide our decision-making extends beyond Wansink’s findings. Maybe you hate carrots because your parents hated them or because ads on television made fun of carrots or because your parents loved them and you had an opposite reaction. Each of these explanations are completely different and, if an authority figure tells you one is true, it would be reasonable enough to accept as fact. As Gelman says, when you’re in a field where the message is “you are manipulable,” that manipulation can run in many ways.
Traci Mann, who says she’s “reserving judgment and hoping for the best” when it comes to the Wansink scandal, points out that it’s not his “big and important papers” that are under investigation. Yet it is worrying when a major figure in a field is accused of misconduct. “If there’s suspicion it reflects on all of us in the field,” she says. “But [Wansink] really popularized this kind of work, and that won’t go away no matter what.”
Wansink’s studies have been replicated by other labs, meaning that even if some of the data is a bit fuzzy, many of his findings are still correct. Wansink’s Smarter Lunchrooms program reorganizes lunchrooms (and doesn’t just remove cookies and chocolate milk altogether) in an effort to nudge kids into making better food choices. Suggestions include things like varying what fruits and vegetables are offered to make it more likely that kids can find something they want to eat, prominently featuring milk on easy-to-reach shelves, and using creative naming to turn “rice and carrots” into the more enticing “jasmine rice with roasted local carrots.” By all accounts, it’s been a success; the federal government was excited enough by the program that the Department of Agriculture gave it a $5.5 million grant in 2014 in addition to funding some of his other work. It’s tempting to say the Smarter Lunchrooms program wouldn’t have made it into more than 29,000 schools if some part of it weren’t working, but that’s exactly the line of thinking that allowed D.A.R.E. to fail at keeping kids off drugs. Even if Smarter Lunchrooms is succeeding, it doesn’t mean it’s the be-all and end-all to getting kids to eat more healthfully in school. We won’t know unless we keep criticizing, testing, and studying the program.
In other words, the Wansink scandal may not be a case of “the emperor has no clothes” as much as “the emperor went out in shorts and a tank top on a frigid day.” No one forced him to dress appropriately, so he didn’t. His current case of hypothermia is just fallout from two decades when behavioral economics was too popular, too easy, too marketable to be criticized.
Jayson Lusk, distinguished professor and head of the Department of Agricultural Economics at Purdue University, says that any fallout from the scrutiny of Wansink’s work will only make the field better. “What I mean by that is that reviewers will look at this work with a little more skepticism.” And consumers should be more skeptical as well. “There’s this message that we’re subject to unseen forces and there’s obviously some truth to that,” Gelman says. “But you have to be careful because you can’t have this many unseen forces all having large and predictable and consistent effects.”
Much like how allergists can’t tell what food is making someone sick without testing them one at a time, we can’t know whether we’re being nudged in one consistent direction or springing through our daily lives like we’re trapped in a nudge-themed pinball machine. There are hundreds and thousands of unseen forces and subliminal nudges acting on us every day. Real life, in other words, is not a controlled food lab.
Nutrition is such a complex field, with so many variables affecting it — gender, weight, health, metabolism — that figuring out what food is “good for you,” even with a scientific study and millions of dollars in funding, is next to impossible. Are paleo or gluten-free diets better? Should we all be eating fewer carbs? Or maybe it’s more carbs? No one has answers, and so junk science that gives us even a bit of advice into how to eat spreads like a virus through the internet.
Two years ago science journalist John Bohannon was responsible for a study that made headlines around the world. The study was titled “Chocolate with high cocoa content as a weight-loss accelerator” and was published in the International Archives of Medicine. Bohannon recruited real subjects and used control groups and data that achieved statistically significant results. As Bohannon wrote in a post for io9 about how the study came about, “It was, in fact, a fairly typical study for the field of diet research. Which is to say: It was terrible science. The results are meaningless, and the health claims that the media blasted out to millions of people around the world are utterly unfounded.” He published it under the name Johannes Bohannon, purportedly a researcher at the fictitious “Institute of Diet and Health” in Germany. For a payment of 600 euros, he got the study published in a mere two weeks.
The total sample size of 16 people included 18 measurements of variables like weight, cholesterol, and sleep quality. Through some p-hacking, Bohannon was able to prove that chocolate could help you lose weight. Let the chocoholics of America rejoice! News about the study was published in multiple newspapers and featured on news and talk shows in Texas and Australia.
A quick Google search should have told anyone reporting on the study that neither the person nor the institute existed. Journalists shouldn’t be expected to replicate entire studies before reporting on them, but taking simple measures — like looking at the sample size, actually reading the study and not just the abstract, or double-checking the integrity of the publication to make sure it’s not pay-for-play — would do a lot to weed out the worst misrepresentations. It’s worth mentioning, however, that none of these actions would have caught the data issues with the Wansink studies. That’s why academics have to become more critical, too — even of the biggest names in their field.
“I think of there being a continuum between actual fraud and ignorance and willful ignorance,” Gelman said. “I’m not sure where [Wansink] falls.”
Even when behavioral economics is right — because either the study was done properly or the researcher haphazardly stumbled onto the correct answer — its shortcuts may not always present the best solutions. Some of the biggest names in behavioral economics are starting to question the way their findings have been used by governments. In 2010, George Loewenstein and Peter Ubel — both well-respected behavioral scientists — wrote an opinion column for The New York Times acknowledging the limits to the field of study. “As policymakers use it to devise programs, it’s becoming clear that behavioral economics is being asked to solve problems it wasn’t meant to address,” they wrote. “Indeed, it seems in some cases that behavioral economics is being used as a political expedient, allowing policymakers to avoid painful but more effective solutions rooted in traditional economics.”
Loewenstein and Ubel specifically mention the way behavioral economics was being used to curb obesity — a problem that has steadily increased in magnitude, from a 13.4 percent obesity rate in American adults in the 1960s to a rate of 38 percent as of 2014. As a result, weight loss in the U.S. is an enormous industry ($64 billion in 2014); in 2012, there were more than 100 million dieters in the country at any given time. Healthy eating has become a government imperative: Every five years, the United States puts out updated dietary guidelines. But even when government officials get the best nutritional minds together to come up with what they think people should do to eat healthfully, politics gets in the way.
One would expect that the federal government’s dietary guidelines consist purely of the most up-to-date nutrition recommendations, but there are frequently huge behind-the-scenes fights with industry groups that, for example, may not like the government telling people to eat less beef. Just to put a cherry on top, the members of the departments of Agriculture and Health and Human Services who finalize the guidelines aren’t disqualified over conflicts of interest. Members have received money in the form of grant funding or other payments from everyone from the nut industry to Pepsi subsidiaries.
Things like calorie labeling on restaurant menus have become popular ways to “inform” consumers about what they’re eating. But obesity, Loewenstein and Ubel write, isn’t caused by a lack of information; one of the major causes of obesity is that unhealthful, processed foods are cheaper than the good stuff. So those pesky, politically unpopular solutions like soda taxes or subsidizing fruits and vegetables really are a better solution than calorie labeling. “But because we lack the political will to change the price of junk food, we focus on consumer behavior,” they write.
Wansink built his career on teaching people kitchen cheats based on research, but reached a dead end when the culture of shortcuts he fostered began informing the way he conducted his research. If scientific research were about getting the right answer, methods be damned, math teachers wouldn’t have to make students show each step of long division to make sure they came to “35 divided by 7 equals 5” through skill and not luck.
Scientific advice for how to eat better may not to be as controversial as a study about climate change or gun control, but that doesn’t mean the results don’t matter to millions of Americans. These studies spread because Americans are worried about their weight and health and want someone to tell them how to eat better. When no one reputable is able to do that, junk science stands up to take its place. If one man could nudge generations of people into wholeheartedly believing in the American dream of bacon and eggs for breakfast, what could be possible with the resources of entire industries or governments behind it?
“Yes, behavioral economics has been oversold,” says Lusk, “but that doesn’t mean it doesn’t have value. The challenge is finding where to use it.”