Lay off the fatties: they're not hurting anybody--maybe not even themselves.

Fat Politics: The Real Story Behind America's Obesity Epidemic, by F. Eric Oliver, New York: Oxford University Press, 228 pages, $20/$15.95 paper

The Diet Myth: Why America's Obsession With Weight Is Hazardous to Your Health, by Paul Campos, New York: Gotham Books, 299 pages, $15 paper

THE GOVERNMENT seems to have made tremendous strides in its War on Fat. In 2004 researchers at the U.S. Centers for Disease Control and Prevention (CDC) said "poor diet and physical inactivity" were killing 400,000 Americans a year, a number that was widely presented as an estimate of "obesity-related deaths." Just one year later, the estimate had been reduced to about 100,000. To cut the death toll by 75 percent in the space of a year, the anti-fat crusaders must be doing something right.

Or something wrong. Ascribing deaths from chronic diseases to specific lifestyle variables is a tricky, highly uncertain business, and the 400,000 figure, which was announced in The Journal of the American Medical Association by a team that included the director of the CDC, was suspect from the start. For one thing, the association between fatness and mortality disappears among Americans 65 and older, the age group that accounts for most deaths. According to the CDC's own data for the years 2001 to 2003, excluding older Americans leaves just 585,000 or so deaths a year, of which more than 180,000 are caused by accidents, suicide, homicide, lung cancer, HIV, influenza, pneumonia, and chronic lower respiratory diseases--none of which the CDC blames on obesity. To believe the 400,000 death toll, you'd have to believe that virtually all the remaining deaths, from causes such as heart disease, stroke, hypertension, cancer, and diabetes, are due to "poor diet and physical inactivity," a phrase public health officials and the press have treated as synonymous with fatness. (More on that later.) That would leave no room for risk factors such as smoking, stress, and heredity.

It did not take long for another set of government-employed statisticians to issue a new, much lower estimate of mortality due to excessive weight, this one also published in FAMA. By their reckoning, obesity--meaning a body mass index (BMI) of 30 or more, corresponding to a weight of 203 pounds or more for a man of average height (five leer, nine inches)--accounts for 112,000 deaths a year in the U.S. But in this study, people who were merely "overweight," with BMIs between 25 and 30 (meaning a weight between 169 and 202 pounds for an average-height man), did not have higher mortality than people of "normal" or "ideal" weight, which is in fact neither normal (since most Americans exceed it) nor, to judge by this study, ideal in terms of health. In fact, the death rate among chubby (but not obese) people in this study was lower than the death rate among thin (but not underweight) people, to the tune of 86,000 fewer deaths a year. Which makes you wonder exactly what it means to be "overweight" and why we should be worried about it.

University of Chicago political scientist J. Eric Oliver, author of Fat Politics, and University of Colorado law professor Paul Campos, author of The Diet Myth (published in hard-cover as The Obesity Myth), both take up this question, and they reach similar conclusions. First and foremost, they argue that, except for certain conditions associated with very high BMIs (starting around 4% which corresponds to a weight of 271 pounds for an average-height man), there is little evidence that extra weight per se causes health problems. To the extent that fatness is correlated with illness, they maintain, it is primarily because fatness is associated with "poor diet and inactivity"--factors that independently raise the risk of diabetes, high blood pressure, cancer, and cardiovascular disease. Fat people are less likely than thin people to exercise regularly, and they are more likely to skimp on fruits, vegetables, and whole grains while eating diets high in fat and refined carbohydrates. Oliver and Campos say these habits, which are more common among fat people but shared by many thin people, are the main problem. Campos also emphasizes the health risks of repeatedly losing and regaining weight. Overall, Oliver and Campos, both of whom say they accepted the conventional wisdom about weight at the outset of their research, make a persuasive case for their contrarian stance.

Having shown that the medical case against fatness is much weaker than government officials and anti-obesity activists claim, Oliver and Campos ask why it is pushed so aggressively and accepted so widely. They see motivations ranging from the rational (the vested interests of obesity researchers, public health officials, and the diet and pharmaceutical industries) to the irrational (a deep-seated cultural revulsion at fat people, disproportionately poor symbols of sloth who serve as stand-ins for minority scapegoats). Here, too, Oliver and Campos are pretty persuasive, but from a policy perspective their analyses are ultimately unsatisfying. Assuming they are right that fatness per se is the wrong target, that the real threats to our health are poor nutrition and sedentary lifestyles, the question remains: What is the government's proper role in addressing these threats?

While neither Oliver nor Campos seems inclined to favor interventions aimed at getting us to eat better and exercise more, neither do they take a clear, principled stand against them. In particular, they do

not directly challenge the slippery "public health" reasoning that treats risky behavior like a contagious disease, providing an open-ended excuse for government meddling in formerly private decisions. The same rationale that makes smoking, drinking, drug use, driving without a seat belt, or biking without a helmet a "public health" issue--the government's purported duty to discourage actions that may lead to disease or injury--applies with equal force to diet, exercise, and every other lifestyle variable that affects morbidity and mortality.

Although neither Oliver nor Campos launches a broad attack on this agenda, they do an effective job of questioning the reality of the "obesity epidemic," beginning with the very definition of overweight and obese. "Fat," says Campos, "is a cultural construct, not a medical fact." In 1985, Oliver notes, a consensus conference convened by the National Institutes of Health (NIH) recommended that men and women be considered "overweight" at BMIs of 27.8 and 27.3, respectively. In 1996 an NIH-sponsored review of the literature found that "increased mortality typically was not evident until well beyond a BMI level of 30" Yet two years later, the NIH yielded to a World Health Organization recommendation that "overweight" be defined downward to a BMI of 25, with 30 or more qualifying as "obese." Oliver says "the scientific 'evidence' to justify this change"--which made millions of Americans overweight overnight--"was nonexistent" since "there is no uniform point on the BMI scale where all these diseases [linked to weight] become more evident."

Moving beyond correlation to causation, things get even more complicated. Campos and Oliver complain that obesity researchers are so eager to demonstrate a link between fatness and sickness that they routinely make the sort of statistical adjustments that strengthen the association but rarely make the sort that would weaken it. For example, studies generally control for smoking, an independent cause of disease that is more common among thin people, and pre-existing illnesses, which might make thin people look unhealthy because people tend to lose weight when they're sick. But obesity researchers usually do not take into account fitness levels, nutrition, yo-yo dieting, and the side effects of weight loss drugs.

"Nearly all the warnings about obesity are based on little more than loose statistical conjecture," says Oliver, adding that there is no plausible biological explanation for most of the asserted causal links between fatness and disease. "The health risks associated with increasing weight are generally small," says Campos, and "these risks tend to disappear altogether when factors other than weight are taken into account" For example, "a moderately active larger person is likely to be far healthier than someone who is svelte but sedentary." Campos cites research finding that obese people "who engage in at least moderate levels of physical activity have around one half the mortality rate of sedentary people who maintain supposedly ideal weight levels." Lest you think these facts have been noticed only by political scientists and law professors, Campos and Oliver draw heavily on the work of biomedical researchers such as Case Western nutritionist Paul Ernsberger, University of Virginia physiologist Glenn Gaesser (author of the 1996 book Big Fat Lies: The Truth About Your Weight and Your Health), and Steven Blair, the physician/epidemiologist who heads the Dallas-based Cooper Institute.

That's not to say Oliver and Campos don't sometimes overstate their case. "In reality," says Oliver, "we have no clear idea whether any deaths at all can be attributed solely to a person's body weight. "Yet elsewhere he says "there are only two medical conditions that have been shown convincingly to be caused by excess body fat: osteoarthritis of weight-bearing joints and uterine cancer that comes from higher estrogen levels in heavier women." Women do sometimes die from uterine cancer, don't they? And Oliver hedges a bit by mentioning that "fat distribution may actually be a better predictor of mortality than body weight." While "fat on someone's hips and thighs seems to have little or no relationship to the risk of death," he writes, "fat in the belly ... seems to be problematic."

Oliver also says "about the worst thing that comes from being heavy is that it puts great pressure on people's joints and inhibits their ability to exercise." But if, as Oliver and Campos both suggest, an unreasonable fear of fatness should be blamed for the bad health effects of anorexia, dangerous diet drugs, fluctuating weight, and even smoking ("a common weight loss and weight maintenance strategy," Campos notes), surely obesity can be blamed for deterring the exercise necessary to keep fit. Beyond the fitness issue, at a certain point obesity seriously compromises a person's ability to get around and participate in everyday activities.

Yet none of this contradicts the main scientific point of these two books, which is that the public health establishment, abetted by a credulous and alarmist press, has greatly exaggerated both the strength of the evidence linking fatness to sickness and the level of risk involved. Oliver cites a 2004 New York Times story headlined "Death Rate From Obesity Gains Fast on Smoking," based on the highly implausible 400,000-death estimate that was later repudiated by the CDC. He also mentions a 2003 A.P. article that announced "Obesity at Age 20 Can Cut Life Span by 13 to 20 Years." He notes that "the obesity in question was at a BMI of 45 [305 pounds for an average-height man], which affects less than I percent of the population." In a passage that could have been lifted from a critique of U.S. drug policy, Campos says "the basic strategies employed by those who profit from this war are to treat the most extreme cases as typical, to ignore all contrary data, and to recommend 'solutions' that actually cause the problem they supposedly address."

Campos and Oliver emphasize that, while people lose weight all the time (over and over again, in fact), keeping it off over the long term is rare. "Despite a century-long search for a 'cure' for 'overweight,'" says Campos, "we still have no idea how to make fat people thin. "That's a bit of an overstatement, since even Campos concedes that a determined, persistent effort to reduce calorie intake and increase calorie expenditure (an approach he derides as "chronic restrained living") can make fat people thin. But it's true that our bodies resist weight loss, an evolutionary defense against famine that in circumstances of abundance tends to make us chubbier than we might like. Each person seems to be genetically predisposed to a certain weight range. Although it's possible to overcome that predisposition, it requires more effort than most Americans are willing to expend, judging from their flabby guts and quivering thighs. Campos and Oliver say, in essence: Don't bother. Not only is there little evidence that weight loss per se (as opposed to the lifestyle changes that accompany it) improves one's health, but it can be harmful, especially if it involves weight cycling, drugs with dangerous side effects, or radical surgery with high complication rates.

Anti-fat activists such as Yale psychologist Kelly Brownell agree with Campos and Oliver that substantial long-term weight loss is nearly impossible, which is why they emphasize social engineering to change the "food environment" (and the exercise environment), thereby preventing people from getting fat to begin with. Someone who believes fatness itself is not much of a health problem might nevertheless support such policies, most of which are aimed at getting people to eat better (as well as less) and exercise more, goals Campos and Oliver consider worthwhile. Yet Oliver, who agrees with Brownell that the ready availability of cheap, tasty food is the main reason for rising BMIs in the U.S. (because it led to an increase in snacking), is refreshingly skeptical about Brownell's proposal for "junk food" taxes, which he correctly sees as fundamentally unworkable. He likewise dismisses other anti-fat nostrums, including advertising restrictions, bans on soda in schools, and beefed-up physical education, saying none is likely to work. "Getting Americans to really change their eating and exercise patterns would require a level of totalitarianism that would make even Kim Jong Il blush," he writes. "The very rationale of a liberal system such as ours is that individuals are best left to decide for themselves which choices to limit, particularly as long as such decisions do not infringe on the safety or well-being of others."

Unfortunately, that "well-being of others" exception is elastic enough to justify "a level of totalitarianism that would make even Kim Jong Il blush." Every would-be regulator of every heretofore private matter argues that it affects the well-being of others. For example, anything that compromises people's health, including poor diet and lack of exercise, has the potential to raise the cost of taxpayer-funded medical care--an argument that is frequently heard in the obesity debate but that Oliver and Campos leave curiously unaddressed. Similarly, Oliver notes that public health specialists in the U.S. "needed new problems to tackle in order to justify their existence" after their triumphs over communicable diseases in the late 19th and early 20th centuries. He also criticizes an influential CDC PowerPoint presentation that made obesity look like a plague sweeping the nation. But he does not nail down the crucial distinction between true public health problems like tuberculosis and air pollution, which involve risks imposed on people against their will, and "public health" problems like smoking and overeating, which involve risks people voluntarily assume.

This omission may be due to Oliver's discomfort with the language of choice. Both he and Campos blame the unjustified obsession with weight and the cruel vilification of fat people on capitalism, which, they say, prizes self-discipline and stigmatizes those seen as lacking it. To be fair, Campos more specifically blames a procapitalist Protestant asceticism that encourages the pursuit of wealth but frowns on those who enjoy it too much. There's an element of truth to this analysis; a similar ambivalence regarding pleasure helps explain American attitudes toward sex, drugs, and gambling. But it does give you pause when you consider that the obesity obsessives also blame capitalism, for precipitating the current crisis by making food plentiful, inexpensive, appealing, and convenient. New York University nutritionist Marion Nestle, for example, blames America's adiposity on "an overly abundant food supply," "low food prices," "a highly competitive market," and "abundant food choices," while Kelly Brownell claims restaurants exploit consumers when they give them more for less, since "people have biological vulnerabilities that promote overeating when large portions are available, a strong desire for value, and the capacity to be persuaded by advertising."

Although they talk a lot about giving people more "options" (such as the option to eat a salad rather than a cheeseburger, of to walk rather than drive to the grocery store), what the anti-fat crusaders really want to do is limit people's options (by taxing the cheeseburger or redesigning cities to discourage driving).While he rejects their prescriptions as impractical, Oliver seems comfortable with the idea of enhancing freedom by restricting it. "Our increasing affluence and consumerism seem to have trapped us," he writes. "As the obesity epidemic shows, maximizing our choices does not necessarily maximize our freedom or power.... The expansion of choices is no longer making our lives any easier; in fact it may be making them harder."

Despite their anti-market instincts, Oliver and Campos do an important public service by dissecting what Campos correctly identifies as another in a long line of "moral panics," revealing the value judgments, aesthetic reactions, prejudices, and emotions beneath the veneer of objective science. But in case you have any illusions that putting the health risks of extra pounds in perspective is all it will take to call off the War on Fat, you might want to have a look at the CDC's list of "Frequently-Asked Questions About Calculating Obesity-Related Risk." The CDC explains that "many chronic diseases are affected by obesity," and mortality "is an important indicator of the severity of a public health problem." So does the recent dramatic reduction in the estimate of deaths caused by excessive weight "mean that obesity is less important than CDC once thought?" the FAQ sheet asks. "Not at all," says the CDC.

In short, obesity-related deaths are an important measure of how serious the problem is, but reducing the number by three-quarters does not make the problem any less serious. That's because the purpose of government-generated "public health" statistics is to alarm the public and expand the government. On no account should the numbers be considered reassuring. To the guardians of our health, nothing could be more alarming.

Senior Editor Jacob Sullum (jsullum@reason. com) is the author of For Your Own Good : The Anti-Smoking Crusade and the Tyranny of Public Health (Free Press).