The Economic Costs of Smoking-lnduced Illness


Although speculation regarding the effects of smoking dates back to the sixteenth century, when tobacco was introduced to the Old World, clinical and epidemiological investigation has positively linked smoking with specific illness and death only within this century. The economic effects of smoking have been relatively neglected, however. This report is intended to shed additional light on them.

The calculation of the costs of smoking is most important, not only because of the nation’s present preoccupation with health care costs in general but also because of political decisions that are being made daily in both public and private sectors concerning the allocation of resources, including specifically government regulation, health education, and the role of health prevention. Anti-Smoking advertising, for example, can be justified only by assuming that the benefits of the campaign outweigh the costs of the operation. But decisions can be made more rationally if we know (1) the actual costs of smoking, and (2) both the costs and the effectiveness of different methods of intervention.

Cost of disease considerations

Both direct and indirect health care costs are included in the economic impact of smoking. Direct costs are those associated with the prevention, detection, and treatment of illnesses attributable to smoking. Indirect costs are earnings lost through morbidity and mortality; consequently, they measure the value society places on an individual’s contribution to the economy. But the less obvious considerations, such as smoking-caused fires with their direct and indirect health care and property costs, operational expenditures of anti-smoking organizations, and cost of purchasing cigarettes, etc., must be taken into account as well. The sum of these disadvantages must be weighed in turn against the definite benefits, such as tax revenues, and the possible benefits, such as lowered anxiety levels in the smoking population and the health implications of associated weight loss.

Although estimating these costs is conceptually simple, most of the relationships have never been adequately defined. In fact, very little information is available concerning health care utilization that is directly attributable to smoking, since most of the research to date has merely linked smoking habits with particular diseases, correlated death with smoking, or isolated causative agents (e.g., carcinogens) within the smoke.

Review of the literature

The literature linking smoking to disease is well known.. It might be helpful, however, to acknowledge some of the more important works. The Surgeon General’s report of 1964 made the smoking-health controversy a public issue by correlating the higher death rates by various diseases with differing smoking habits, and by linking smoking with bronchopulmonary disease.

Since 1964, the literature periodically compiled by sources such as the U.S. Public Health Service in its The Health Consequences of Smoking (1965) and the Royal College of Physicians’ Smoking and Health Now (1971) has linked a rather significant roster of conditions to smoking (Med. J. Australia – Special Supplement 1975):

1. Cardiovascular disease

2. Chronic obstructive bronchopulmonary disease

3. Cancer

  • a. lung
  • b. larynx
  • c. oral cavity
  • d. esophagus
  • e. urinary bladder
  • f. pancreas

4. Pregnancy complications

  • a. decreased fertility
  • b. increased spontaneous abortions
  • c. increased still-births

5. Peptic ulcer

6. Infancy respiratory disease

7. Oral disease (noncancerous)

8. Accidents

  • a. fire
  • b. automobiles

Unfortunately, most of these studies do not lend themselves to generalized economic analysis. Instead of reporting that a specified portion of Disease X is believed to be caused by smoking, the results are usually couched in terms such as: “Males who are heavy smokers and are between the ages of 30 and 45 are Y times more likely to develop Disease X.”

Several studies have linked increased health care utilization to smoking status, but the results do not lend themselves to direct economic analysis (). Still other authors have attempted to estimate the economic health costs of smoking, but their figures are in need of updating and refining. Soper () estimates that the total economic health cost of smoking (medical care, lost income, and property loss due to fires) in 1966 was $5.3 billion, but his figures were based on an earlier Canadian study, and were determined by such crude measures as the CNT ratio of Canada to that of the United States,

Williams and Justus () estimate that the 1970 health costs attributable to smoking are $4.23 billion. However, they derive their figure from a 1958 source that states that the total cost of respiratory disease in the United States was $2 billion (), which they had to inflate to 1970. They then apply it to a “best estimate” from yet another Canadian study which states that 70 percent of chronic bronchitis and emphysema is due to smoking ().

Walker () editorialized: “It has been estimated that $11.5 billion is spent annually in the United States for health care costs resulting from cigarette smoking.” His source was a one-paragraph item in American Medical News (1974) attributing the statement to a physician addressing the American Lung Association.

Clearly the range of estimates and the imprecise analytical basis justify a need for better health cost estimates associated with smoking.

Methodology and results

In an authoritative article, Cooper and Rice () have published economic cost of disease data disaggregated into 16 diagnostic categories. Their analysis includes both direct and indirect costs, together with an in-depth explanation as to how these costs were derived. Table “Total economic costs of selected diseases: estimated direct costs indirect costs of morbidity and mortality, with present value of lifetime earnings discounted at 4 percent, by diagnosis, 1972 (in millions)” is extracted from this study to present the costs of the major disease categories associated with smoking.

Boden () reports figures from the working papers of the NIH Task Force on Prevention in Environmenta1 Health (1976) that estimate the percentage of major’ disease. categories’ due to environmental problems, including smoking. Fortunately, his disease categories parallel those of Cooper and Rice.

To determine the economic health costs attributable to smoking, the total economic costs () are multiplied by the corresponding estimated smoking factor (). Table “Total economic costs of smoking: estimated direct costs, indirect costs of morbidity and mortality, with present value of lifetime earnings discounted at 4 percent, by diagnosis, 1975 (in millions)” presents the results, inflated to the 1975 prices, and includes the cost of property fires caused by smoking (Med. J. Australia – Special Supplement 1975). The total direct cost of smoking is thereby estimated at $7.5 billion, which is a approximately 7.9 percent of all direct health care costs in the nation. The total (direct and indirect) smoking related economic cost of these diseases is $25.9 billion. This is an even larger proportion (11.3%) of the total cost of all diseases, probably due to extended morbidity and high mortality of the particular diseases considered (cancer, cardiovascular, respiratory).

A. Other health related costs

As mentioned earlier, there are other costs of smoking that should be considered, such as those associated with diseases of lesser economic significance (e.g., noncancerous oral diseases) or diseases in which cigarette smoking has a relatively small impact (e.g., peptic ulcer). These estimates are not yet available. Nevertheless their aggregate would no doubt be significant, and our estimates are correspondingly understated.

B. Tobacco and GNP costs

Not to be neglected in the economic analysis of smoking is the cost of tobacco and its associated taxes. In 1975, tobacco accounted for 4.6 percent of the value of all crops sold in the United States (U.S. Dept. of Agriculture 1976). Cigarette smokers smoked an incredible total of 607 billion cigarettes annually, which is an average of 4,121 (206 packs) per adult (18 years and older). American smokers paid $15.7 billion for all tobacco products ($14.4 billion for cigarettes) and of this, $5.8 billion was collected as taxes by all levels of government. If the estimated number of smokers were 60 million, as is reported by the National Cancer Institute (1976), then each smoker spent, on the average, $240 a year on 506 packs of cigarettes.

The net GNP effect is the total of (a) all direct and indirect health costs, (b) fire damage, and (c) tobacco sales. Table Total Costs of Smoking and Tobacco, 1975 (in millions) shows this total to be $41.5 billion, which is approximately 2.5 percent of the GNP, an average of $692 per smoker per year.

Table Total Costs of Smoking and Tobacco, 1975 (in millions)

Direct health care costs $ 7,507.l
Fire property damage 166.8
Lost earnings 18,209.6
Cost of tobacco (retail) 15,660.0
Total $41,543.5

Smoking results in a major drain of the nation’s economic resources regarding both direct health care costs and those costs associated with lost earnings due to sickness and death. Other costs that are considered are those associated with smoking-caused building fires and the purchase of tobacco products. In the absence of smoking, these resources would be reallocated to other sectors of the economy, and there would be more healthy individuals to share the respective benefits.

Although the cost of illness computations of Cooper and Rice are generally considered authoritative, the reader is cautioned that direct application of these estimates to calculate the costs of smoking is difficult .because the proportions of each illness which are attributed to smoking () have only been roughly estimated by the NIH Task Force on Prevention in Environmental Health, and have not been produced by rigorous analysis of empirical data. Consequently, empirical research is needed to improve these estimates and other research is needed to determine the costs and effectiveness of differing anti-smoking techniques. Once this information is available, it can be related to similar data pertaining to other health problems. Only then can rational decisions be made concerning the distribution of health resources within a program for the prevention of disease. The fact remains, however, that on the basis of methods delineated here, smoking related diseases cost the nation $25.9 billion annually and account for 11.3% of the total economic cost of all diseases.

The Economic Costs of Smoking-lnduced Illness: Discussion

These figures are all derived from secondary sources. Their accuracy is not certain since they are an update of a previous study done in the Sixties. The cost data were developed for the Commission on Heart Disease, Cancer and Stroke, although they are more refined here, and they were collected by Rice and Cooper.

It would be easy to calculate attributable risk for those cancers that are related to smoking using data from the third national cancer survey. The cost data were based on a rather large sample of the twenty million people surveyed. Cancer incidence was &rived from a 10% direct interview sample, examination of hospital records and from patients’ individual receipts for costs. Those tapes are available from N.C.I., and they cover all costs, not just those of hospitalization.

We are spending $140 billion per year on health care now, and by 1980 this figure will top $200 billion. Success in preventive medicine depends on the economics of health care. Tobacco use produces what might be termed negative benefits – immediate benefit to the individual, but long-term negative effects on health and the economy. Phillip Abramson says that when someone else pays for something the costs are infinite. We might reflect on this fact in planning approaches to these problems. It would be useful to do a hospital census and estimate the number of in-patients and out-patients who would not require care if they had never smoked. Possibly such a study could be done, however, priorities don’t seem to be in this direction.

Attributable risk estimates are available currently which show substantial risks associated with smoking for a wide variety of disease categories. It is, perhaps, unwise to assume that the entire excess risk of disease in smokers is a result of their smoking. Enidemiologically some of the diseases associated with smoking-have-not been shown to be etiologically related. Others, such as lung cancer, clearly are causally related to smoking. Blanket assumptions that every disease associated with smoking is caused by smoking create a credibility gap because they are subject to successful refutation by industry sources. For this reason, it would seem important in estimating the dollar costs of tobacco use to separate costs which are clear and definite results of smoking from those which assume causality where it has not been proven to exist.

Another issue worth considering is what would happen to health costs if, say, cancer were eliminated. Probably the effects would not be great. It has been estimated, for example, that mean life span at age 60 or 65 would be increased by about 1.4 years, and that two-thirds of deaths would then be cardiovascular, a condition which is more expensive to care for than cancer. This viewpoint was very controversial, and several discussants rejected it on the grounds that treatment of cardiovascular disease wasn’t more expensive than cancer and that this selective statistic focused only on people in their sixties, ignoring the impact of younger persons.

A number of suggestions were made concerning the fact that smokers do not pay the actual costs of tobacco use. Some felt that if health resources were to be reallocated that they should focus not on hospitals or medical care, but on the health insurance system. Persons who smoke should pay for the health costs of smoking.

The aim of a health care system is often forgotten. Many persons die in their sleep in their eighties, never having cost society or themselves much in terms of health care or social support. This is the idea that must be the aim of all health plans.


Selections from the book: “Research on Smoking Behavior”. Murray E. Jarvik, M.D., Ph.D., et al., editors. State-of-the-art of research on smoking behavior, including epidemiology, etiology, socioeconomic and physical consequences of use, and approaches to behavioral change. From a NIDA-supported UCLA conference. National Institute on Drug Abuse Research Monograph 17, December 1977.