There is a mini-scandal in the nutrition research community as Disney pressed a scientific journal to pull a study it funded:
The Walt Disney Company urged an academic journal to withdraw a nutritional study of children’s meals at Disney World last fall — a study it had funded — amid a public backlash over corporate involvement in scientific research, according to newly obtained emails.
Disney wasn’t concerned about the study’s findings, but feared being publicly associated with one of its main authors, James Hill of the University of Colorado School of Medicine.
Hill’s work last summer drew an outcry among scientists who felt his project, funded by Coca-Cola, played down the impact of sugary drinks in obesity.
Normally it is considered important to publish all research to make sure the funders — or the researchers — are not cherry-picking those studies that prove their point. But that is not the case here. The study spoke well of Disney and its efforts to improve nutrition in kid’s meals, but Disney did not want to be associated with this particular researcher who became controversial on another project.
We have written about and engaged with Brian Wansink, Cornell University’s Professor of Marketing at the Dyson School, many times, including these:
Fortunately, he was one of the editors (along with Koert Van Ittersum, Professor of Marketing and Consumer Well Being at the University of Groningen in the Netherlands) of the first edition of a new journal, the Journal of the Association for Consumer Research. The issue was focused on The Behavioral Science of Eating, and Professor Wansink refused Disney’s request to withdraw the piece, pointing both to logistical and ethical difficulties.
The study was basically designed around these parameters: Disney changed its default options on kids’ meals so that, in theory at least, consumers would have to specifically request French Fries rather than carrots, apple slices, unsweetened apple sauce or grapes; the consumers would also have to specifically request soda rather than water, 100% juice or low-fat milk. Would offering this default option change what children eat and, if so, how? The abstract explains the study this way:
A retrospective study of kids’ meals purchased at Walt Disney World was conducted to determine acceptance rates for healthy sides and beverages. Purchase data from all 145 Walt Disney World restaurants were analyzed using a log-linear model and a Poisson regression.
Across all restaurants, 47.9% and 66.3% of guests accepted healthy default sides and beverages, respectively. Acceptance rates of sides and beverages were higher at quick-service restaurants (49.4% and 67.8%, respectively) compared to table-service restaurants (40.3% and 45.6%, respectively).
The healthy defaults reduced calories (21.4%), fat (43.9%), and sodium (43.4%) for kids’ meal sides and beverages. This study contributes by examining the use of kids’ meal healthy defaults in quick-service and table-service restaurant formats at the world’s largest theme park, a previously unstudied setting, and by providing the largest ever healthy default data set.
The results suggest that healthy defaults can shift food and beverage selection patterns toward healthier options.
Marion Nestle, Professor of Nutrition, Food Studies, and Public Health at New York University, has also often been featured on the Pundit, including in pieces such as these:
Special Edition IX — Salmonella Saintpaul Outbreak…Concerns About FDA/CDC:
Pundit’s Mailbag — Plenty of FDA Incompetence to Go Around But TheProblem Is A Matter Of Politics, Says Marion Nestle, Esteemed Food Policy Professor
And Dr. Nestle had quite an exchange with the authors of the study:
In 2006, the Walt Disney Company announced a new initiative to improve the nutritional quality of meals served to children at its theme parks. The company would be changing the default kids’ meals — the components that come without having to be ordered separately — to include low-fat milk, juice, or water rather than soft drinks, and sides such as apple sauce or carrots rather than French fries. Parents who wanted sodas or fries for their children would have to ask for them, something many might not bother to do. Health groups had long advocated for this policy change (Wootan 2012).
As I commented to a reporter at the time, ‘going to Disney World is an excuse for eating junk food… Disney or its advisers must be feeling they have some responsibility” (Horovitz and Petrecca 2006). Indeed, the healthier defaults were part of a larger effort by Disney to deal with its contribution to obesity in America. After ticket prices, food is the second greatest source of revenue at Disney World.
Although reducing the amount of food consumed at the parks might help create a less “obesogenic” food environment, revenues might fall. But the default change might be revenue-neutral. By 2008, Disney could report that two-thirds of U.S. customers ordering kids’ meals had accepted the default, with no loss in sales. In Hong Kong Disney parks, nearly all customers accepted the default. The report, however, did not include data on the numbers or proportions of customers ordering kids’ meals (Walt Disney Company 2008).
Disney’s more recent summary of its child health initiatives states that it is funding investigators at the University of Colorado to conduct a more formal evaluation of use of the default options (Walt Disney Company 2015). The paper by Peters et al. (2016) in this issue of the Journal presents the results of that research. Their work confirms the ongoing effectiveness of the strategy. Nearly half the customers ordering kids’ meals accepted the healthy default side dishes and two-thirds accepted the healthier beverages. These choices resulted in significant reductions in the calories, fat, and sodium in purchased kids’ meals, but not sugar (Peters et al 2016).
The authors argue that gentle nudges like these are preferable to more coercive policies that smack of nanny-statism. Such reductions help, but are they enough to make a real difference? To answer this question, it would help to know what else the children were eating along with the drink and side dishes. Although the authors were given raw sales data, Disney did not permit them to use this information as part of the overall analysis. The company also refused to provide information about the number of children who visited the park or the number of kids’ meals sold.
These missing pieces raise red flags because this is a Disney-funded study that produced results that Disney can use to advertise itself as a company that cares about kids’ health, and to deflect attention from Disney World’s reputation as a junk-food paradise. Corporate funding of research introduces conflicts of interest and reduces the credibility of the results, not least because the biases inherent in such research are largely unconscious, unintentional, and unrecognized (Moore et al 2005)
The results of this study merit especially careful scrutiny. Taking them at face value, the default strategy worked well for the drink, but the sides are still a problem, and so are the sugars. They do not reveal much about what kids eat in a day at Walt Disney World.
Nudges like this default are an important part of strategies to counter childhood obesity. But are they enough to deal with the public health problem? To make a real difference, they need to be accompanied and supported by a range of policy approaches. Current thinking about such approaches recommends combining insights from behavioral research, economics, and public health to establish a food environment far more conducive to making the healthy choice not only an easy choice, but also the preferred choice.
Doing so is likely to require multiple actions — for example, regulation of nutrient content and marketing; incentives such as subsidies of healthier foods; disincentives such as taxes, warning labels, and nutritional rating systems for unhealthier foods; and education of adults and children (Hawkes et al 2015). Disney’s voluntary default is a small step in the direction of such policies, but many more are needed if we are to make real progress in reducing the prevalence of childhood obesity.
We found the study itself, moderate in its claims, and the researchers recognized many caveats in extrapolating the study. For example, Walt Disney World is a world unto itself. So whether consumers who accept something at Walt Disney World would accept it if there was active competition from other restaurants is unclear.
The basic gist of the study is to support arguments made in the book, Nudge: Improving Decisions About Health, Wealth, and Happiness by Richard A. Thaler, Professor of Behavioral Science and Economics at the University of Chicago Booth School of Business, and Cass R. Sunstein, the Robert Walmsley University Professor at Harvard and Administrator of the White House Office of Information and Regulatory Affairs From 2009 to 2012.
The basic idea is that we can encourage improvements in people’s lives by structuring choices in certain ways. For example, if you give a memo to everyone in the company when they are hired saying ”let us know if you want to start contributing to a retirement plan,” you get very different participation rates then if every employee is automatically enrolled and the employees have to actively opt out. Equally in a cafeteria, things such as what a consumer sees first alters what is selected.
Dr. Nestle’s comments focus on two things: First is whether the nudge analyzed in this study is sufficient to deal with public health problems related to childhood obesity. The answer to that it is clearly not, and in fairness the researchers never said or implied that. They just said this default would move things in the right direction.
Second, Dr. Nestle is very concerned over the fact that Disney helped fund the study. Certainly we agree that the funding of research can create conscious or unconscious bias. But we are uncertain why Dr. Nestle has concluded that “corporate” sponsorship is uniquely malicious in this regard. After all, governments, foundations and union interests all have their own agendas as well.
We have written about research sponsored by many organizations. We wrote about the Robert Wood Johnson Foundation:
And the Produce Safety Project — “an initiative of The Pew Charitable Trust at Georgetown University”
Just to name two, we could name many more.
We see definite patterns in what kind of research results come out of research funded by different foundations and organizations. So it is not sufficient to point out that corporate sponsorship can bias researchers; there is always the issue of biting the hand that feeds you whether that hand is a corporation, a foundation, a union or an individual.
To us, this is a fact of life. Researchers need support and so there is always the prospect of bias, but as analysts and reviewers, this just makes reviewing the study design imperative.
The real problem with this study is that it falls into the trap of many studies, in which the researchers study what is easy to gather data on, even if the end result is meaningless.
In this case, the data tells us virtually nothing of value because it does not get to the issue being studied – how do changes in the default option impact what children actually consume?
Children at Walt Disney World do not generally order their own food. Nor are they restricted to eating the particular things included in the kids’ meals.
To give an anecdote: The Jr. Pundits are quite the fans of Disney, and they were the target age for Disney kids’ meals when this default option was established.
Now Jr. Pundit Primo, aka William, and Jr. Pundit Segundo, aka Matthew, happen to not drink soda, so even before this change, they went for water. So switching the default option had no impact on their drinking habits. But they were both quite the fans of French Fries. Fortunately, they love grapes, carrots and apple slices, so they gladly accepted the new default option.
So on this study, it would show that the default option reduced their calorie intake, but this would not be true. Although the Jr. Pundit’s wanted their carrots, grapes or apple slices, they also wanted French Fries, and we bought them separate side orders of Fries. This study, however, ONLY looks at the actual Kid’s meal, parents could buy their kids a side cheeseburger and it wouldn’t show up because it is not part of the kid’s meal.
So the switch to the healthy default was a success if the criterion is “More Matters!” They did eat more fruits and vegetables than before. However, if the criterion is calorie reduction, the plan had the opposite effect. They still ate their French Fries, indeed because a full side order has more fries than was put on a kids’ meal, they ate more Fries than ever plus the calories from the apple slices, carrots and grapes!
In other words, the researchers assumed for the sake of the study that the kids ate the kids’ meal and nothing else. Since that is not necessarily true, we can draw no useful information from this study.
That would be true no matter who paid for the study.