Lest you think that this is an entreaty to persuade you to smoke, let me begin by stipulating that smoking is bad for you. It may kill you. Smoking is really stupid. Don't do it.
That out of the way, we can turn to our subject: the conflicting goals of public health and science.
The question of the danger that a practice entails is an empirical question, one amenable to a scientific answer. For example, the probability that taking a two week hike this summer will get you killed by lightning could be established empirically-let's say it's one in three million, vs. one in 60 million odds of being hit by lightning during the same period under normal circumstances.
If science had established these numbers, going on the hike would increase 20-fold your risk of being hit by lightning. What should you do? That's not a scientific question. It is a question of the strength of your particular impulses, desires, fears and other subjective factors.
The same is true of medicine: Your doctor can tell you that doing X will increase chances of dying young. He may be able to tell you how much likelier you are to die prematurely if you do X (e.g., five times as likely if you don't do X) or the probability that doing X will kill you prematurely (one in 20). But the doctor can't tell you if doing X is the right or wrong choice. Only you armed with the empirical data, can make the subjective decision of whether doing X is worth the risk.
Which brings us to smoking. Over the years, I've noticed that you can't get through the day without being warned that you smoke, you're Y times as likely to get lung cancer as if you don't smoke- (According to the American Cancer Society, the figure is 20 times.) But I've never seen anyone cite the absolute likelihood (i.e., if .you smoke, you run a one-in-Z chance of getting lung cancer).
For a long time I assumed, without thinking, that this meant "they" didn't know the absolute likelihood. Then, thinking back to my graduate statistics courses, I realized that you can't ascertain the relative risk unless you know the absolute risk.
So I wrote to several public-health organizations that were citing relative-risk figures, asking for the absolute figures. Without exception, they wrote back that they didn't know the answer, or that the answer was too difficult to calculate. I knew that this wasn't true, and wondered why they were telling me this. In due course I figured it out.
Scientists who work for these organizations are devoted to saving lives, which means they try to provide smokers with the evidence that will persuade them to kick the habit. But the goal of persuasion is not always congruent with the demands of scientific accuracy.
To understand why, suppose that someone who smokes a pack and a half of cigarettes a day has a one-in-eight chance of contracting lung cancer. To scientists and most nonsmokers, anyone who would take such a risk just to smoke must be one sandwich short of a picnic. However, the heavy smoker has been so traumatized by the incessant message that smoking is bad—and has had to stand in the rain outside the office so often in order to get his fix—that he'd consider himself lucky if he had a 50-50 chance of avoiding death by tar and nicotine.
Tell a smoker the truth, and he'll be thrilled to hear he has a seven-in-eight (or whatever) chance of avoiding lung cancer. Where the scientist sees death, the smoker sees life. And so the scientist, understanding this, tells you that he can't answer the simple question: What is the probability that a smoker will get lung cancer?
It is a commandment of pure science that an empirical finding should be presented without concern for the human consequences. Usually our ability to predict those consequences is so pathetically limited (take my word for it—I'm a sociologist) that this is only a theoretical problem.
But we can now predict with virtual certainty that the more people smoke, the more people will die prematurely. Scientists working in this area would have to be inhuman to put aside their desire to save lives in the name of pure science. Thus, I am not suggesting that they stop suppressing the figure that would reassure the smoker. But I am suggesting that it be made clear this is what the scientist is doing.
If the issue were just smoking, none of this would be worth mentioning. But there is an increasing tendency to let considerations of consequence determine the presentation of findings—a tendency that has become the standard in much of social science. In these cases the skewing is based not on an understanding of an unarguable causal relationship like that between smoking and cancer, but on an ideological view of the world that may or may not reflect reality. This, in the long run, is a threat to society that deserves to be taken as seriously as smoking.
Mr. Goldberg is chairman of the sociology department at New York's City College.
Material presented on this home page constitutes opinion of the author.
Copyright © 1997 Steven J. Milloy. All rights reserved. Site developed and hosted by WestLake Solutions, Inc.