U.S. Surgeon General Richard Carmona AP

The dangers posed by secondhand smoke are debatable and likely to remain so given the limitations of epidemiology. It's well established that tobacco smoke can raise the risk of diseases such as lung cancer and heart disease. The question is how much it takes.

OUR VIEW: 'The debate is over'

Because the doses absorbed by non-smokers exposed to secondhand smoke are much smaller than those absorbed by smokers, any health risks would be so small that it is difficult to confirm them in studies comparing, say, the spouses of smokers with the spouses of non-smokers. The weak, statistically insignificant associations typically found in such studies are consistent with a low-level risk.

They are also consistent with no risk at all if, for example, factors associated with marriage to a smoker (such as poor diet, lack of exercise, or unreported smoking by subjects believed to be lifetime non-smokers) independently raise disease risks. The surgeon general's report mentions such problems but concludes they probably cannot fully account for the observed associations. The truth is we don't know for sure and probably never will.

Nonetheless, I doubt that the average person encountering tobacco smoke in a bar or restaurant objects to it because he thinks his tiny risk of lung cancer might go up slightly if he stays there for several decades. The main complaint, as always, is the immediate smell and discomfort. Even if there is an added element of anxiety about long-term health consequences, that does not justify imposing a one-size-fits-all solution on every business in the country. Whether secondhand smoke is a health hazard or merely a nuisance, people who want to avoid it can do so by avoiding businesses that allow smoking. A free society that respects diversity should make room for people with different preferences.

Jacob Sullum, a senior editor at Reason magazine, is the author of For Your Own Good: The Anti-Smoking Crusade and the Tyranny of Public Health.