What's Wrong With Fat? (25 page)

News reports on the Eating-to-Death study overwhelmingly blame individual choices for obesity: “Americans eat the wrong things, don’t exercise enough and smoke too much—and it’s killing them.” 93 Bad food and activity choices are often evoked: “According to Dr. William Healy, of Cuts Fitness For Men(TM), ‘increasingly sedentary lifestyles set up a self-defeating sequence of behavior. Almost all men can get in a rut—it’s simply easier to be out of shape than it is to exercise. If they are not active, they are likely sitting around—and eating becomes a convenient “activity.”
Eating the wrong foods—often the case—makes the problem worse.’” 94 As is shown in figure 4.6, news reports on the Fat-OK study, however, are also less likely to discuss a range of different solutions to the obesity crisis, reflecting that this group of news reports is less likely to frame fatness as a problem in the first place. Only 16 percent of news reports on the Fat-OK study, compared to 83 percent of news reports on the Eating-to-Death study, discuss individual-level solutions of getting individuals more physically active. Three percent, compared to 60 percent of articles on the Eating-to-Death study, discuss individuals eating better, and 2 (compared to 34 percent) discuss individuals changing their “lifestyle.” Similarly, only 3 percent of articles on the Fat-OK study, compared to 66 percent of news articles on the Eating-to-Death study, discuss regulating the food industry, and 3 percent of articles on the Fat-OK study, compared to 54 percent of news articles on the Eating-to-Death study, discuss regulating school lunches. None of the news articles on the Fat-OK study discuss either regulating food advertisements or removing vending machines from schools, compared to 14 and 17 percent, respectively, of articles discussing the Eating-to-Death study.

Figure 4.5:
Percentage of News Samples Discussing Various Causes of Overweight/Obesity.

In this chapter, we have examined how scientific research and the news media jointly frame fatness as a public health crisis brought on by bad personal behavior. We have seen how news media routines amplify scientific claims about an obesity crisis brought on by individual failings.
We saw how the publication in 2005 of a study, showing that “obesity” is associated with fewer deaths than previously thought and that “overweight” people actually live longer than “normal weight” people, provided an opportunity for the news media to air the possibility that concern about body size is overblown. However, the news media subjected this study to greater scrutiny than studies affirming the commonsense notion of an obesity crisis.

Figure 4.6:
Percentage of News Samples Discussing Various Solutions for Overweight/Obesity.

But so what? What effect does this have on social processes and people’s lives? What difference would it make if fatness were instead understood as beautiful, healthy, and/or a basis for civil rights claims? What difference would it make if body size were understood to be largely beyond personal control?

In the middle of conducting research for this book, I enlisted the assistance of a talented undergraduate student, whom I will call Liz. One of the tasks entrusted to Liz was to read the news reports on the Eating-to-Death and Fat-OK studies, discussed in the previous chapter, and to copy and paste quotes that illustrated some of the quantitative results. Liz began with the 35 news reports on the Eating-to-Death study, in which authors estimated that overweight, obesity, and poor nutrition resulted in 400,000 excess deaths in the year 2000. 1 As we have seen, news reports on this study typically framed overweight and obesity as a major public health crisis, while suggesting that individual bad behavior was the root cause. After reading news report after news report of this study, Liz arrived in my office one day looking out of sorts. She told me that this reading had led her to look up her own BMI and that, even though it was within the “normal” range (at about 24), she worried that she was too fat.
Looking somewhat uncomfortable, she told me that she had started a weight-loss diet for the first time in her life.

I felt terrible. I told Liz about the Fat-OK study, which had found that overweight and obesity combined were associated with fewer than 30,000 excess deaths in the year 2000, in large part because being overweight but not obese was associated with almost 90,000 fewer deaths that year. 2 Liz spent five hours in the following week reading news reports of the Fat-OK study and pulling out illustrative quotations. Afterward, Liz returned to my office looking more relaxed. She told me that she had gone off her diet and was feeling better about her body.

We all notice patterns like this. You may remember having gone on a diet after reading a series of articles about the health risks of obesity or limiting your consumption of fish after reading a news story about rising mercury levels. Another university professor has written that her students reported increased fears of becoming fat after being exposed to a steady stream of popular books and news media reports on the alleged obesity epidemic, as part of her class. 3

For social scientists, such observations can provide the basis for hypotheses that can subsequently be systematically tested. After speaking with Liz, I wondered whether these sorts of news reports would have similar effects on other people. Based on her experience, as well as other things I knew through my research, I began to hypothesize that exposure to different kinds of news framing of fat would systematically affect not only people’s views about health risk but also their body image and expression of fat prejudice.

This line of questioning speaks to what cultural sociologists refer to as
cultural reception
, or the way in which people make sense of specific accounts.
Cultural sociologists have demonstrated that the way in which different people interpret cultural texts, such as newspaper stories or romantic novels, varies. 4 For this reason, it is not enough for a researcher to examine the
production
of these accounts by newspapers, novelists, or others, and to provide his or her interpretation of
content
; it is also important to see how people themselves interpret or
receive
specific texts.

My specific questions, however, were not just about how people interpret different news media frames but, more specifically, about how exposure to different news media accounts shapes specific attitudes. This, I realized, would require an experimental design—in which I would vary the news media accounts that people read and then measure their attitudes about health, body image, anti-fat prejudice, and other factors. 5 With psychologist David Frederick, I designed seven separate experiments, each of which enrolled between 99 and 779 participants, coming to a total of 2,379 participants. 6

With the exception of Experiment 3, all of the experiments relied on a sample of students at a large west coast state university, consisting of approximately 70 percent women versus 30 percent men, and 35 percent Asian, 30 percent white, 20 percent Latino/a, and 15 percent identifying as biracial, multiracial, or “other” ethnic group. The average BMI of these samples was lower than the national average, at 22 to 23 for women and 23 to 24 for men, depending on the experiment. This lower-average BMI echoes the fact that fat students are underrepresented in higher education, in large part because of discrimination on the part of teachers and parents. This means that our findings might not generalize to heavier participants. 7 For Experiment 3, we recruited adult participants by posting advertisements on Internet websites, generating a sample consisting of 78 percent women and 22 percent men, and an ethnic composition of 75 percent white, 8 percent Hispanic, 5 percent black, 4 percent Asian, and 6 percent biracial, multiracial, or another ethnicity. In Experiment 3, the average BMI was 26 for women and 27 for men. A table presenting the samples sizes and conditions for each of experiment is provided in the appendix.

In all of the experiments, we presented respondents with actual newspaper articles. Most of these had been published in the
New York Times
or other leading newspapers and were either formatted to appear as online version of a
New York Times
article or, in the case of the online survey, were accompanied with a mock weblink to give the appearance of authenticity.
We edited them so that they had the same title and author, to control for how these factors might affect perceptions. Other than those changes, however, we reproduced the articles faithfully, with a few minor exceptions. For instance, we deleted and substituted a fake name in an article that quoted me, as we thought this might bias the student participants who knew or had heard of me. Having participants read real news reports means not only that our experiments are arguably very generalizable to “real life” but also that we had less control over the experimental stimuli themselves.
After reading one or more news articles, participants were asked to answer a series of attitudinal questions. After highlighting some of the most striking findings, this chapter shifts to a broader discussion of what fat frames do.

EXPERIMENTAL FINDINGS

Contrary to our initial expectations, respondents did not answer differently questions about body image or propensity to diet based on which article(s) they read. It could be that Liz was especially susceptible, that our survey questions were not sufficiently sensitive, or that our respondents were not exposed to enough news articles for this to have an impact. We did find, however, that reading different news reports significantly shaped attitudes in other ways. Here, I want to focus on two main sets of findings: perception of health risk and expression of anti-fat bias. 8

HEALTH RISK

Experiments 1 and 2 had a series of questions designed to measure whether reading different news media framings of fat shaped views about the health risks associated with obesity. We measured this by asking respondents whether they agreed—on a 9-point scale, in which 1 was strongly disagree and 9 was strongly agree—with statements that higher BMI is associated with health risks and that rising population weights represent a public health crisis. 9

In Experiment 1, participants read either: (1) three articles developing a public health crisis frame
and
three articles developing a personal responsibility frame or (2) two long articles developing a health at every size frame
and
two long articles presenting a fat rights position. Compared to those who read the news articles framing fatness as potentially healthy and as a civil rights issue, those who read the news articles framing fatness as a public health crisis or as about personal responsibility were significantly more likely to strongly agree that being fat was unhealthy (7.9 versus 5.9, on a 9-point scale). To get a sense of how big an effect this is, we calculated an
effect size
. As a general rule of thumb, effect sizes of 0.2, 0.5, and 0.8 correspond with small, moderate, and large effects, respectively. 10 The effect size in this case was 1.83, or very large.

Experiment 1 showed large differences between the two test groups, but, because it did not include a control case, it could not tell us whether only one or both of the test cases were shifting attitudes. Moreover, because it paired the public health crisis and personal responsibility frame, on one hand, and the health at every size and fat rights frame, on the other, it could not tell us which of the two frames were having the greatest effect.

To get at this, Experiment 2 separated out each of the frames and included a control case, that is, a group of participants that read an article unrelated to body size. The two control articles discussed the benefits and risks of cancer screening and risk of delaying cancer treatment, respectively, and made no mention of body weight or lifestyle. The other participants read two or three long articles developing one of the following four frames: (1) public health crisis, (2) personal responsibility, (3) a health at every size, or (4) a fat rights. These articles were identical to those used in Experiment 1 (see Appendix).