Up in Smoke


Smoking: Making the Risky Decision, by W. Kip Viscusi, New York: Oxford University Press, 170 pages, $24.95

Are smokers crazy? This is what public-health officials and anti-smoking activists seem to think, although they never quite say it directly.

They do come close. The Department of Health and Human Services says, "Tobacco use is a disorder which can be remedied through medical attention." Leaving aside the dubious proposition that behavior is a disease that doctors can cure, the government's message is that no person in his right mind would voluntarily choose to smoke. People start smoking because they are tricked; they continue smoking because they are hooked. The possibility that there might be other reasons for smoking—that people might actually like it—is rejected out of hand. "There is no positive aspect to it," says Scott Ballin, former director of the Coalition on Smoking OR Health. "The product has no potential benefits….It's addictive, so people don't have the choice to smoke or not to smoke."

In Smoking: Making the Risky Decision, W. Kip Viscusi demolishes this view of smokers, which has long been a basic tenet of U.S. health policy. Familiar from decades of anti-smoking propaganda, the stereotype of smokers as nicotine slaves often plays a role in calls for higher tobacco taxes and restrictions on cigarette advertising. Viscusi observes that the "stylized smoker" is passive, generally impervious to risk information, and in any case unable or unwilling to act on it. But do smokers actually behave this way?

Viscusi, an economics professor at Duke University and founding editor of the Journal of Risk and Uncertainty, looks at evidence of what people believe about smoking, the choices they make, and the relationship between the two. He finds that, far from underestimating the hazards of smoking, both smokers and non-smokers tend to exaggerate them, especially the risk of lung cancer. Furthermore, risk perceptions have the sort of effect on smoking decisions that you would expect if people were acting rationally: The higher the perceived risk, the less likely a person is to smoke.

But Viscusi notes that the decision whether to smoke also depends on tastes and preferences—on the value that people attach to the pleasures of smoking, the strength of their desire to avoid health risks, and the trade-offs they are willing to make. He bolsters this point by examining how smokers and non-smokers approach other risky decisions. The upshot is that smokers are crazy only in the sense that their tastes and preferences do not correspond to those of public-health officials and anti-smoking activists.

Viscusi's main source of data is a nationwide telephone survey of attitudes toward smoking conducted by the New York research firm Audits & Surveys in 1985. He supplements this information with several surveys that he conducted in 1990 and 1991. Among other things, the surveys asked people to predict how many smokers in a group of 100 would suffer a given kind of health damage because of their smoking. Viscusi finds that the vast majority of people think that smoking is considerably more dangerous than the scientific research indicates.

In the 1985 survey, the average respondent said a smoker runs a 43-percent risk of contracting lung cancer because of his habit. Smokers, on average, estimated a 37-percent risk. By contrast, the surgeon general at that time put the risk between 5 percent and 10 percent. Viscusi's own surveys found similar overestimation of the lung-cancer risk and less dramatic but still substantial overestimation of other health risks. For example, he asked: "Among 100 cigarette smokers, how many of them do you think will die from lung cancer, heart disease, throat cancer, and all other illnesses because they smoke?" The average response was 54; smokers said 47, while non-smokers said 56. Based on the evidence available in 1991, the actual number is between 18 and 36.

As Viscusi notes, it's not surprising that people overestimate these hazards. Individuals tend to exaggerate risks that are highly publicized, especially if the outcome is dramatic or dreaded, as with airplane crashes, hurricanes, and lung cancer. During the last few decades Americans have been bombarded by anti-smoking messages from the government, the news and entertainment media, schools, and health organizations. These efforts to teach people about the hazards of smoking, Viscusi notes, generally emphasize "outcomes and body counts," ignoring probabilities. One of the surgeon general's warnings, for example, simply states that "Smoking Causes Lung Cancer, Heart Disease, [and] Emphysema." This implies a 100-percent risk for these diseases: If you smoke, you will get them.

It's possible, of course, that smokers exaggerate the risks in theory but underestimate them in practice, thinking "it can't happen to me." Viscusi rejects this hypothesis. His statistical analysis indicates that higher risk perceptions make a person less likely to smoke (it's not merely that people lower their risk estimates after they begin smoking). The effect is substantial, and it holds across all age groups. "Indeed," Viscusi writes, "if smokers perceived the lung cancer risks from smoking as being comparable to what scientists estimate those risks to be rather than their current risk perceptions, societal smoking rates would rise by 6.5 to 7.5 percentage points."

Notwithstanding the claims of the anti-tobacco lobby, teenagers and young adults do not ignore smoking-related health risks. Indeed, Viscusi finds that respondents between the ages of 16 and 21 tend to overestimate the risk of lung cancer even more than older respondents do, probably because they have been raised in an environment where anti-smoking messages are ubiquitous. "These findings strongly contradict the models of individuals being lured into smoking at an early age without any cognizance of the risks," Viscusi writes. (Since anti-smoking activists emphasize the supposed vulnerability of female teenagers especially, it's interesting to note that girls tend to have higher risk perceptions than boys.)

It may nevertheless be true that young people are likely to give less weight to health considerations than older individuals do. Even if a young person is acting rationally given his or her current tastes and preferences, he or she may not foresee how those tastes and preferences will change over time. As Viscusi puts it: "Does the 20-year-old smoker fully recognize how his or her future self will value health as compared with smoking?"

The harder it is to stop smoking, the more important this question is. Anti-smoking activists argue that most smokers start at a young age and are unable to stop when they decide they should. Viscusi demystifies the issue of addiction, treating the unpleasant consequences of quitting simply as "costs of change." He writes: "We make choices throughout our lives that are costly to reverse—getting married, choosing a profession, selecting a place to live, and purchasing a car. The fact that reversing such decisions is costly does not imply that such choices are incorrect." He notes that the problems involved in quitting—weight gain, unsatiated cravings, irritability—will be decisive only when they outweigh the perceived benefits of giving up the habit. Many people obviously consider the benefits to be worth the cost: About as many Americans have quit smoking as currently smoke, and more than 90 percent did so without formal treatment.

To illustrate the strength of nicotine addiction, anti-tobacco groups cite polls finding that most smokers say they want to quit. It's likely that many smokers are simply giving pollsters the socially approved answer. Viscusi offers an additional explanation, noting that "almost half the residents of Los Angeles indicate a desire to move out of the city, but do not do so. Almost one-third of blue-collar workers would like to leave their jobs but do not. Statements such as these may indicate that there is some dissatisfaction with one or more attributes of a particular product or activity. Survey statements in which individuals indicate that they would like to quit smoking, for example, might mean that they would like to smoke without risk. However, the fact that they have continued to smoke even with the availability of chewing gum with nicotine suggests that these statements…should not always be taken at face value."

Anti-smoking activists seek to explain smoking as a chemical compulsion mainly because they do not recognize that their values and priorities are not universal. Viscusi's research suggests that smokers simply have different attitudes toward health risks than non-smokers do. In a study of Oregon workers, he found that smokers demand a much smaller wage premium for risky jobs. (Smokers who don't wear seat belts are even more willing to accept hazardous work.) These findings indicate that smokers are acting in accordance with tastes and preferences that are consistent across situations.

At a fundamental level, the public-health establishment has difficulty accepting these tastes and preferences. Consequently, Viscusi notes, government policies aim to deter smoking—indeed, to achieve "a smoke-free society"—rather than to encourage rational decisions. Policy makers seem to assume that risk perceptions and cigarette taxes can never be too high. And government officials have actively discouraged innovations that make cigarettes safer, such as reduced tar and nicotine and smokeless brands like the now-defunct Premier, which HHS derided as a "nicotine delivery device."

Viscusi argues that the government should encourage competition for safer cigarettes and disseminate more information about the relative hazards of different brands. In general, "our purpose should not be to deter smoking but to provide information concerning the variety of smoking hazards, their magnitude, and their consequences for individual welfare. Thus the policy objective should not be a smoke-free society but rather a society with informed risk-taking decisions."

The moderate approach that Viscusi suggests means allowing for the possibility that some perverse individuals will continue to flout social norms by engaging in a habit that others consider foolish, disgusting, and unhealthy. This will require more than accurate information. It will require tolerance.

Jacob Sullum is associate editor of REASON.