The "Social Cost" of Smoking

Numerous "studies" purport to show that smokers are costing "society" amounts which vary from 22 cents to $4.80 for each pack they smoke. That the estimates vary so widely is already a clue that the authors select data which pleases them and ignore the rest.

Apart from accidents, smokers and nonsmokers alike die mostly from heart disease, cancer and strokes. Smokers just get them sooner, on average. By living longer, non-smokers incur more in medical and old age expenses, not less. In addition, smokers' shorter lifespans mean they cost less in Social Security and pensions. Finally, they pay more taxes during their lives. Thus it can be shown that smokers, rather than costing money, actually pay into the system more than they take out.

Stanford economist Timothy Taylor, an anti-smoker, made these points in the San Jose Mercury News, March 7, 1994, and agrees that tobacco taxation cannot be justified by the social cost theory.

In a recent paper for the National Bureau of Economic Research, Duke University economist W. Kip Viscusi calculated a net saving to society of 83 cents per pack.

In 1995, the non-partisan Congressional Research Service updated a 1986 Manning study on smoking costs originally published in the Journal of the American Medical Association. The CRS recalculated Manning's 1986 dollars into 1995 dollars by using the GNP deflator for some costs and a medical services index for others, recognized increased cigarette taxes and added ETS effects. The result: a net saving to society of 15 cents per pack in 1986 became 33 cents per pack in 1995.

But what of the Centers for Disease Control's much-publicised figure of $50 billion? A proper economic analysis must include ALL external costs: health care, sick leave, life insurance, nursing care, pension, fires and effect (if any) of ETS. The CDC arrived at their $50 billion figure as follows:

    $26.9 billion, hospitals
    $15.5 billion, doctors
    $4.9 billion, nursing homes
    $1.8 billion, prescription drugs
    $900 million, home health care

    $50 billion


    $21.6 billion, or 89 cents per pack

What is important in this list is what is not there. No mention is made of offsetting taxes, nor of social security savings, nor of the extra cost of old age.

 Also, they produced the numbers above not by adding up what was actually paid, but by adding up the the total list price of procedures recommended by the book for such illnesses as they declared were caused by smoking. In real life many of the procedures are skipped, what is actually paid by insurance companies is a discounted amount, and some illnesses go largely or completely untreated. As a result, the "costs" they used bear little resemblance to those of Taylor, Viscusi, Manning or the CRS.

 A recent issue of the New Yorker magazine had a short article on smoking and health costs. Its take-off point was the recent Florida law which authorizes the State to sue tobacco companies for its cost of providing health care to smokers. Some of the arguments seem worth quoting:

 "...there is something slightly off about Florida's arithmetic -- and, more broadly, about the whole notion of health care as an exercise in bargain-hunting. Florida says that it has spend a billion two hundred million dollars over the past five years in Medicaid payments for smoking-related illnesses. But that figure is misleading. While smokers use a lot of state-sponsored health care, and about three and a half billion dollars a year of federal Medicare money, they also tend to die around five years earlier than nonsmokers. That means five fewer years of the heavy health-care burdens of old age, five fewer years of nursing-home care, and five fewer years of drawing a federal pension.

 "In fact, for those inclined to such ghoulish calculations, the easier case to make is that smokers save society money. Around a third of the Americans who die from smoking-related causes die before the age of sixty-five, and that means that after paying into Social Security throughout their working lives many of them will collect little or nothing from it. In effect, they subsidize the pensions of nonsmokers. How great is the subsidy? The shortened life span of just those smokers born in 1920, according to a recent study by three Stanford University researchers, will end up saving the Social Security system fourteen and a half billion dollars. All told, they conclude, smoking will reduce the obligations of the system by 'hundreds of billions of dollars.' And, while it's true that smokers are out sick more frequently than their more virtuous fellow-citizens, it's also true, alas, that illness and death, with their attendant costs to the health industry, come to us all."

 And the French have discovered the same thing:

 PARIS (AP) -- Smoking, long considered a costly health care burden, actually reaps the government an annual profit, according to a study released Friday by Paris' Institute of Political Science. High tobacco taxes and social security savings gained by premature deaths made the French government a net profit of 18.5 billion francs (dlrs 3.5 billion) in 1990, the study revealed.

 The study, by Professor Jean-Jacques Rosa, attacks the idea that smoking creates "a considerable social cost." "Premature deaths of retired smokers benefit society because the government no longer has to pay their social security benefits or their health-related expenses," the professor said.

 According to the study, French smokers have a life expectancy of about 67 years, which is six years less than the average French citizen. The study estimated smoking-related medical expenses to cost the government about 26.5 billion francs ($5 billion).

 However, the combined profits from tobacco taxes and the annual savings brought about by premature deaths total about 44.9 billion francs ($8.5 billion). "Consuming alcohol and tobacco, just like reading or going to the movies, must be left to individual choice, and not be considered a social cost," Rosa said.

Finally, an article on the Viscusi study from the San Jose Mercury News (3/27/95):

 Study Looks at Who Pays for Costs Cigarette Smokers Impose on Society By David Ress, Richmond Times-Dispatch, Va. Knight-Ridder/Tribune Business News Mar. 27--If there's no such thing as a free lunch, is it possible to at least get a free smoke?

 Nope, says Duke University economist W. Kip Viscusi. In a recent paper for the National Bureau of Economic Research, Viscusi took a hard look at the question of who pays for the costs smokers impose on society.

 These include additional health-care costs, passed on to everyone in the form of higher health insurance premiums and taxes for Medicare. They include additional sick days at work; more fire risks and higher group life insurance rates.

 Viscusi calculated the additional health-care costs for smokers at the equivalent of about 55 cents for every pack of cigarettes bought in the United States.

 The added sick days, he figures, aren't that big a deal: he estimates they cost society the equivalent of a penny a pack.

 Viscusi pegs the extra fire risk at about two cents a pack. This risk comes from those who smoke in bed or never listened to what Smokey Bear said about pitching lighted cigarettes in the woods.

 The economist puts the added cost to group life insurance policies at 14 cents a pack.

 That adds up to social costs of 72 cents a pack. The federal government collects 24 cents a pack in excise taxes, and the states average another 29 cents. Taken together, that suggests 19 cents a pack of costs to society that cigarette taxes aren't paying.

 For Virginia, with its 2.5 cent-a-pack cigarette tax, the shortfall would be more like 45.5 cents a pack.

 But wait. Smokers don't live as long as nonsmokers. That, says Viscusi, means they spend less time than nonsmokers do in nursing homes. That's a saving to society equal to 23 cents a pack.

 Smokers don't collect pensions and Social Security for as long as nonsmokers do, Viscusi adds. There's a saving to society equivalent to $1.19 a pack.

 Society does lose out, though, in tax collections. If smokers lived longer, the additional taxes they'd pay on their income would translate to the equivalent of about 40 cents a pack.

 Overall, though, as Viscusi counts it, the cigarette taxes smokers pay more than compensate the rest of us for the additional costs they impose on society.

 (end quoted article)
Viscusi's bottom line: a net saving to society of 83 cents per pack.

And absenteeism? This is largely a result of excessive drinking and ensuing hangovers. Based largely on socio-economic differences, a higher proportion of smokers drink than non smokers. Accordingly, if you survey absent workers and ask only the question "Do you smoke?", smokers will appear to be absent more often. But it is alcohol which is primarily responsible. If only non drinking absentees were surveyed, there would be no correlation between smoking and absenteeism.

Apart from being wrong as to fact, however, the social cost arguments depend on a dangerous proposition: that when society, through taxes or insurance, shares costs and spreads risks it thereby becomes entitled to regulate, control and even prohibit behaviors deemed "costly". This is the perfect principle with which to transform a free society into a nation of brothers' keepers, since there is virtually no human activity to which it cannot be extended.


The climax of the government and media tobacco inquisition is now being realized in the addiction issue. This is ironic because the anti-smoking movement began as a drive to improve smokers' health - to save them from themselves. But since this kind of paternalism has never been very successful in this country, the focus soon switched to the effect of smoke on others. Now that doubts have been raised about the integrity of the science used to make that argument, the cause has again come full circle to saving the smokers. But this time, with the drug war fresh in everyone's mind, the antis are armed with a vicious new hate-word - ADDICT- to seize powers over individuals and corporations that they never had before.

Webster's defines both the verb addict and the nouns addict and addiction consistently in two ways: one denotes a psychological dependence or devotion, the other physiological. We can distinguish between the two by using habit (psychological) and addiction (physiological). Habits don't require medical intervention to overcome. Addictions do.

So which applies to smokers? Heroin addicts become extremely ill when they withdraw. They normally require medical supervision, as do cocaine users. So do many alcoholics, who suffer DTs. Coffee drinkers get headaches. Smokers who quit merely become irritable for a few days. While some use nicotine patches, with mixed results, most smokers who quit successfully simply quit.

Dr. Jack Hennigfield of the National Institute of Drug Abuse (NIDA) and Dr. Neal Benowitz of the University of California at San Francisco (UCSF) ranked several drugs on a scale of 1 to 6 (six being the least serious, 1 being the most) based on the following criteria: Withdrawl, Reinforcement, Tolerance, Dependence, and Intoxication.

Substance .... Wdl ... Rif ... Tol ... Dep .. Intoxication
Nicotine ..... 3,3 ... 4,4 ... 2,4 ... 1,1 ...... 5,6
Heroin ....... 2,2 ... 2,2 ... 1,2 ... 2,2 ...... 2,2
Cocaine ...... 4,3 ... 1,1 ... 4,1 ... 3,3 ...... 3,3
Alcohol ...... 1,1 ... 3,3 ... 3,4 ... 4,4 ...... 1,1
Caffeine ..... 5,4 ... 6,5 ... 5,3 ... 5,5 ...... 6,5
Marijuana .... 6,5 ... 5,6 ... 6,5 ... 6,6 ...... 4,4

  • Withdrawal refers to the severity of symptoms experienced upon cessation.
  • Reinforcement refers to the tendency to self-administer, and has been observed in laboratory rats.
  • Tolerance refers to an increase in the amount of a drug necessary to experience the same effect.
  • Intoxication refers to the degree to which functionality is impaired.
  • Dependence is not so rigorously defined. It appears to refer to the determination of the subject to continue using the substance in question, or perhaps simply to the pleasure experienced.



    It is important to note that the "dependence" criterion is new. Prior to their report, addictiveness had been assessed based solely upon the other criteria. While the other four can be quantified to some extent, the dependency rankings were subjective, based upon the doctors' personal opinions and experience. The list they produced has nicotine at the top because it was sorted on the dependency column. Using any of the other criteria, nicotine ranks much lower or at the bottom.

    Thus we can see that the very definition of addiction has been changed, not only in order to apply the label to nicotine but to characterize it as one of the most addictive drugs of all. But even the chart above suggests that any craving is not physiological, but psychological.

    The pleasure of tobacco goes far beyond what nicotine does to the system. It's a way of life. What the smoker enjoys is the whole experience: the routine of handling the pack and the cigarette, lighting up, gazing into the flame, the oral satisfaction of drawing, the taste and the smell. Eating and drinking are synergistic with smoking: they each enhance the taste of the smoke, and smoking enhances the contemplation of the food and drink.

    Nicotine plays a part, but a small one. It can no more substitute for a smoke than No-Doz tablets could replace a good cup of coffee. That's why nicotine patches and chewing gum aren't very effective when it comes to quitting. Of course it's hard to give up. So are many other things which are not physiologically addicting. Your right arm, for example. Or your spouse. If either is taken away you will experience a severe psychological withdrawal. Using "dependence" as a criterion, millions of people are addicted to Monday Night Football.

    In the not-too-distant past, smokers would freely admit that they were addicted and even joke about needing a "fix". Now, however, the most many of them will admit to is a habit. Whether it's a habit or an addiction would be merely a semantic argument, except that most anti-smokers seem to think that addiction gives them the moral right to step in and pass laws or otherwise control the "addict's" behavior without his consent. It doesn't.

    In the same way that the phrase "I could care less" came to mean, through common usage, "I couldn't care less," the meaning of the words "addict" and "addiction" have been fundamentally altered with recent usage. This is what they mean now:

    ad-'dic-tion (n) a condition entitling those not affected by it to control those who are.

    'ad-dict (n) one expected to surrender to the ministrations of those who despise him.

    And that is what is being denied.

    [Previous Page] [Top] [Next Page]