Skip to main content

Verified by Psychology Today


A Serious Questioning of the Science of Smoking

The effects of smoking are less settled than what we are told to think.

Key points

  • Smoking is certainly personally risky, but does not seem to cost society.
  • Secondhand tobacco smoke has not been shown to be dangerous.
  • Every scientific conclusion is more or less uncertain.
Abscent Vector/Shutterstock
Source: Abscent Vector/Shutterstock

As the new millennium dawned, I noticed that smoking, very much on the wane since its 1960 peak, was suddenly getting a lot of alarming publicity. “Smoking causes lung cancer, bronchitis, emphysema, heart disease and cancers in other organs including the mouth, lip, throat, bladder, kidney, stomach, liver, and cervix…” said a prestigious British report,1 “Half of all smokers will die prematurely” it added. Environmental tobacco smoke (ETS) is almost as dangerous as mainstream smoke: “ETS has been shown to cause lung cancer and ischaemic heart disease, and probably to cause COPD, asthma, and stroke in adults. ETS is harmful to children, causing sudden infant death, pneumonia and bronchitis, asthma, respiratory symptoms, and middle ear disease” said the Tobacco Advisory Group of the Royal College of Physicians in July 2005. Even third-hand smoke is hazardous, apparently. And what’s more, smokers’ healthcare costs everyone else.

This alarmism made me curious. I grew up in a family where all three adults smoked. I never saw any ill effects and all eventually passed away from other causes at advanced ages. Is my family a miraculous exception or is there something wrong with ‘the science’?

Smoking science is necessarily weak. It would be very difficult to prove smoking causes many of the problems blamed on it simply because they are both uncertain—not everybody will suffer from these problems—and invariably long delayed. Research on phenomena like this is obviously dodgy: you can't do real experiments because you can't give people things that may make them sick, and very few researchers can wait thirty years for their experiments to show results. All you have are correlations that might or might not be causal.

My own research into the science behind smoking yielded conclusions that were quite different from those which are broadcast through all levels of society, even down to cigarette packages themselves. There was, for example, little or no evidence that secondhand (passive, environmental) smoke posed any kind of health risk.2 Smokers actually cost less in lifetime health care than nonsmokers, not more, as both politicians and health authorities proclaimed.3 The life-years lost by heavy cigarette smokers are fewer than generally believed (in 2002 scientific estimates of that loss were in the range of six to eight years. The public’s risk perception was greater, as men believed the life-expectancy loss is 10.1 years and women believed that it is 14.8 years).4

There is a discrepancy between what is actually known about smoking, even after all these years of research and advocacy, and what is publicly affirmed. Annual ‘deaths caused by smoking’ have risen from 314,000 in 1985 to 480,000 now, even as many fewer people smoke. The smoking issue resembles the way the science around COVID has been treated. In these cases, necessarily weak and uncertain science is presented as ‘settled,’ not to inform, but to sway the public’s decision-making. Our adult populace is intelligent enough to understand that science is an incremental discipline, and one rooted in fundamental uncertainty. When the citizenry is treated like children who must be told ‘scientific’ fairy tales to get them to behave properly, it can have unintended consequences, including a building distrust of the institution of academia and science itself.

A confirmation

In 2013 I published a short book, Unlucky Strike: Private Health and the Science, Law and Politics of Smoking, summarizing the research that led me to these conclusions. In 2020 I published an article on science-based regulation that contained some discussion of a Harvard science historian’s biased-but-conventional account of the passive smoking issue. This caught the attention of an experimental psychologist colleague, Alan Silberberg, at American University. I had not seen Alan for decades and had no idea he was interested in the smoking issue. It turned out that not only was he interested, but in 1999, more than a decade before my book, he had written several chapters of a carefully researched book on smoking. The book was not finished and it was never published. He sent me the manuscript, which was based on earlier data than my own book. Nevertheless, it came to essentially the same conclusions. The second edition of Unlucky Strike has an updated version of several chapters of Alan’s book as an appendix.

The received wisdom, both now and earlier, is that: (a) cigarettes kill almost 500,000 Americans per year; (b) second-hand smoke kills nonsmokers; (c) cigarettes are addictive; (d) cigars, pipes, and chewing tobacco are also substantial health threats; and (e) smoking imposes uncompensated costs on the American health system. In all these cases, weak science has supported a conclusion that remains questionable. In all these cases, for various reasons, the science has turned out to be shockingly simplistic, and in every case but one (the shortened life expectancy of heavy cigarette smokers) inconclusive or even wrong.

Silberberg and I decided to publish an updated second edition of my 2013 book with Alan’s chapters as an appendix: Unlucky Strike, Second Edition. Will it have any effect on the public perception of smoking? Perhaps the book will accelerate healthy skepticism toward the ‘believe science’ mantra and help restore public understanding of science as a mode of inquiry rather than dogma to be revealed to the ignorant masses. Every scientific conclusion is more or less uncertain. To promote a conclusion without some idea of its uncertainty is certainly irresponsible.


Public health: ethical issues. Nuffield Council on Bioethics, 2007. A report prepared by a Working Party chaired by Lord Krebs, p. 105.

Environmental tobacco smoke and tobacco-related mortality in a prospective study of Californians, 1960-98. James E Enstrom, Geoffrey C Kabat, BMJ VOLUME 326 17 MAY 2003. See also Environmental Tobacco Smoke and Coronary Heart Disease Mortality in the United States—A Meta-Analysis and Critique. Inhalation Toxicology, 18:199–210, 2006. The BMJ paper produced a storm of criticism because it questioned the link between secondhand smoking and illness. Anyone who thinks that smoking is treated dispassionately by the medical community should take a look at these comments: BMJ, 15 May 2003. See also Geoffrey C Kabat Hyping Health Risks: Environmental Hazards in Daily Life and the Science of Epidemiology. Columbia University Press, 2008, for an extended discussion of ETS. The study was first cited but not discussed and then omitted completely from two Surgeon-General reports.

Lifetime medical costs of obesity: prevention no cure for increasing health expenditure. Pieter H. M. van Baal, Johan J. Polder, G. Ardine de Wit1, Rudolf T. Hoogenveen, Talitha L. Feenstra, Hendriek C. Boshuizen, Peter M. Engelfriet, Werner B. F. Brouwer PLoS Medicine |, February 2008 | Volume 5 | Issue 2 |. This Dutch study compared smokers and the obese to a “healthy-living” group. Smokers had the lowest lifetime health-care costs, healthy non-obese the highest. See also Miller, L. S., Zhang, X., Rice, D. P., & Max, W. (1998). State estimates of total medical expenditures attributable to cigarette smoking, 1993. Public Health Reports, 113, 56-98.: Average per capita health costs are unrelated to the population percentage of smokers—see Figure A7 in Staddon, J Unlucky Strike, Second Edition (PsyCrit Press, 2022)

The New Cigarette Paternalism, By W. Kip Viscusi, REGULATION, Winter 2002 – 2003. A comprehensive survey, The Price of Smoking, by Frank A. Sloan, Jan Ostermann, Christopher Conover, Donald H. Taylor, Jr. and Gabriel Picone (MIT Press, 2004) summarizes the data up to 2002.

More from John Staddon, Ph.D.
More from Psychology Today