be required to post a "danger to health" warning on packs of cigarette and in advertisements. Those hearings were an outgrowth of the Suregon General's report on the health hazards of cigarette smoking (Smoking and Health, 1964). A lot of political words have been spoken since then, and a law requiring a warning on the pack has gone into effect. But the situation with respect to cigarette advertising has not changed radically up until now (1969). Therefore I have not thought it necessary or sensible to update the material, but rather to leave it much as it appeared then except for some added newer data.

Many individuals and groups favor a warning requirement. Some want cigarette advertising prohibited completely. However, no responsible person has suggested outlawing the manufacture or sale of cigarettes themselves, mostly because of the sorts of dangers that arose during liquor prohibition in the 1920s. People who oppose the warning and the ban on advertising base their opposition on grounds of legality as well as of economics. This chapter considers only the economics of a warning requirement or a ban. It does not consider other economic alternatives such as an increase in cigarette taxes. (In the latter connection see Lyon and Simon, 1968.) The chapter discusses the possible effects on cigarette use, and the consequent economic impacts, of these two proposals on the groups within the economy and society that have a stake in what happens. Mostly it refers to the abolition of all cigarette advertising, rather than the affect of a warning, because the effect of advertising abolition is better understood. The effect of a warning requirement would probably be much less than that of an advertising ban, but of the same general nature.

Effect on cigarette-consumption rate

Opponents of a warning or ban say that forbidding cigarette advertising, or requiring a danger warning, would have "practically no effect" on consumption. Supporters argue, however, that advertising has a "substantial" effect in influencing people to start smoking, and in keeping them smoking. Where is the truth?

It is perfectly clear that advertising has the power to influence the purchase of particular brands of cigarettes; the $200 million spent annually for cigarette advertising as of 1963 (perhaps $260 million in 1967) is positive proof of that. But we are not interested in the power of advertising to shift smokers from one brand to another. We want to know how cigarette advertising as a whole starts people smoking or keeps them smoking.

Borden (1942) examined the role of cigarette advertising in the astounding growth of cigarette smoking starting about 1900, when the annual consumption of cigarettes per person over the age of 15 was 49. By 1962 the yearly rate had risen to 3,598 cigarettes per capita (U.S., F.T.C., Bureau of Economics, n.d., p. 3; 4,280 cigarettes per person over the age of 18 in 1967: Statistical Abstract, 1968, p. 737). Borden did not say that advertising caused the rise in cigarette consumption. He argued that if the public had not been ready to take up cigarette smoking, advertising could never have caused such a large increase in consumption. Nevertheless, Bordon concluded that advertising was an important factor in the size and speed of increase in cigarette smoking.

Basmann (1955) studied the rise and fall in cigarette advertising from year to year in the United States, and its apparent effect on cigarette consumption, using sophisticated regression techniques. He found that for each 1 percent change into total cigarette advertising, the number of cigarettes smoked changed 1/20 of 1 percent. In an earlier study Schoenberg (1933) arrived at a similar estimate, between 1/20 and 1/12 of 1 percent. In other words, the consumption of cigarettes is affected by the amount of advertising, but it takes a big change in the amount of advertising to make much of a difference in consumption. This is typical of an industry once it has become well established, but it may also result from the degree to which the smoking habit takes hold of people and the fact that nothing elso is a good substitute for smoking.

1 Almost all of the statistics mentioned in the chapter apply to the year 1963. Some of the estimates are extrapolations because only earlier data were available when the paper was written in 1964. Other estimates are my own compromise judgments resulting from several conflicting estimates. In many cases all the sources for the estimates of cigarette advertising will not be cited individually because there were so many of them and because the available data were extrapolated and adjusted by the author. Sources from which data were taken include various issues of Broadcasting, Advertising Age, Printer's Ink, and the Statistical Abstract. In many places I have cited only the source of the raw data which I then fiddled with. In the case of the footnoted estimates above, the most important sources are estimates in Advertising Age (Aug. 31, 1964, pp. 36-37) of $252 million and $207 million for the expenditures in unmeasured and measured media, respectively, of the six biggest cigarette manufacturers.

What would happen if all cigarette advertising were cut off? An extension of the Schoenberg-Basmann finding (using Basmann's estimate) suggests that if there had been no cigarette advertising in any past year, consumption would have been about 5 percent less than it was. If the ban on advertising continued, one might expect further decrease in the amount of consumption each year, but the absolute decrease would be less each year. The predictions are subject to many technical reservations, and they go far beyond the data. But they are the best that one can do at this time.

A required danger warning in the advertisements would be a type of negative advertising. One cannot estimate how much the warning would cut smoking, but certainly the effect would not be as drastic as a ban on advertising, or else no firm would continue to advertise. Our inability to come up with any better prediction is testimony to how little scientific knowledge we have about the effect of different forms of advertising copy. But it should certainly be possible to protest advertisements that contain warnings, just as other advertisements are protested, in order to obtain an estimate of the effect of a warning.

Effect on life expectancy

Now let us estimate the health effect of an advertising ban and its resulting reduction in cigarette consumption.

1. People who are kept from starting to smoke will live, on the average, 4 to 5 years longer than if they had started smoking. For each cigarette smoked, someone's life is shortened by five to nine minutes. We shall figure seven minutes per cigarette (Simon, 1968a; 1968b). This assumes that all of the relationship between smoking and life expectancy is causal.

2. About 500 billion cigarettes were smoked in 1963 (Bureau of Economics, n.d., p. 3; 527 billion in 1967: Business Week, Dec. 21, 1968, U.S., F.T.C., p. 68). A decrease of 5 percent in consumption for just one year would therefore mean an increase of human life in the United States of about 180 billion minutes, or 340,000 years of life. Remember, this is the amount of lifetime increased by a decrease of 5 percent in smoking for just one year.

Effect on employment and local economies

An estimated 225,000 people make a substantial part of their living in tobacco agriculture (Statistical Abstract, 1963, p. 617). Cash receipts were $1.3 billion (Bureau of Economics, n.d., p. 13), providing earnings of approximately $600 million last year (ibid, p. 634), of which about $450 million came from cigarettes (USDA, 1963, p. 112). Some 31,000 factory workers earned $150 million from cigarette manufacture in 1963 (Statistical Abstract, 1963; estimates jibe closely with those of Hill and Knowlton, 1964, and are much the same for 1967). In total, then, cigarette purchases put about $600 million into the pockets of workers and farmers. How would a ban or a warning affect them?

A drop in consumption would have no immediate effect on farm earnings because of the government subsidy program. Unless the government removed the subsidy, the taxpayers at large rather than the farm population would take the loss. But assume that the subsidy were cut. If so, the effect of a loss in earnings would probably be worse than the figures show, because the effect would be concentrated in a few states that are already economically backward. Many tobacco farmers are already poor and would find it hard to find new jobs. For example, North Carolina is an agricultural state and almost half of its farm income comes from tobacco.

Using the estimates above, employment and earnings would be cut by 5 percent at most during the first year of an advertising ban. In subsequent years, the further cut in jobs and/or dollars would be less. I say "5 percent at most" because there is good reason to believe that an important proportion of smokers who quit smoking cigarettes, or young people who never start, would use other forms of tobacco instead. Table 10-1 shows that cigarettes largely replaced other forms of tobacco and did not create much new demand for tobacco. To the extent that smokers switch to pipe tobacco, cigars, chewing tobacco, and snuff, the damage to tobacco farming would be reduced, even though cigarette tobacco is a more expensive product than other types of tobacco.

2 By now we seem to have solid evidence that negative advertising can reduce smoking: a mounting barrage of negative publicity. has been taking its toll. This year (1968), for the first time since 1964, producers sold fewer smokes than they did the year before... a decline in the total number of smokers and a decrease in consumption per capita, plus signs that fewer young people are taking up the habit.. One of the strongest influences in the antismoking campaign is the profusion of new TV spots by health groups" (Business Week, Dec. 21, 1968, pp. 68-70).

35-881-69

TABLE 10-1.-CONSUMPTION OF TOBACCO PRODUCTS, SELECTED YEARS, 1900 TO 1960 (PER PERSON AGED 15 YEARS AND OVER IN THE UNITED STATES)

Source: U.S. Department of Agriculture, Economic Research Service, as shown in "Smoking and Health," 1964, p. 45

Furthermore, some or many tobacco workers who are thrown out of work would get other jobs, so one overestimates greatly when one assumes that the equivalent of lost cigarette-industry wages would be lost to the economy as a whole. But the worst (or close to it) is assumed for the sake of argument. Later we shall look at the potential effects on employment again, when the overall picture is considered.

Effect on cigarette companies

To understand the effect of a ban or a warning requirement on the cigarette companies, one must first understand the nature of advertising as a business investment. When a firm spends a dollar in advertising a brand of cigarettes this year, the advertising bought with that dollar increases cigarette sales this year. But it also increases cigarette sales next year, and the year after, and in subsequent years. Customers get into the habit of buying a given brand, a habit that may continue for many years. To say it another way, a dollar of advertising may create some good will or brand loyalty that persists long into the future, though each year the effect of that single dollar of advertising is less than the year before. Cigarette advertising is really an investment, just like an investment in a new machine that will produce for many years after it is bought. (Fuller statements of the "distributed-lag" property of advertising may be found in Jastram, 1955; Telser, 1962b; and Palda, 1964.)

Telser studied the pre World War II cigarette market in considerable detail (1962). He found that only 15 to 20 percent of the advertising investment is used up in the year in which the advertisements appear. This means that for each $1 of sales created in the advertising year, much more than $3 of sales will be created in subsequent years. (However, because of the chaos in the postwar cigarette market, investment is probably used up faster now than Telser's estimate.)

Therefore, even if all cigarette advertising were stopped tomorrow, the established cigarette brands would continue to sell well for many years, though at continually diminishing rates. During that time the cigarette companies would be recouping the investments they had already made. Furthermore, since all the firms would have to stop advertising, the investments already made would not be used up as fast, a situation which would give the cigarette companies a better return on their invested dollars than they expected to earn when they made the investments.

The totoal effect, then, wuld be that in future years the sales of any brand would gradually decrease. But the gross profits from a brand would be at a very high rate for a while, because the firm would not be making any further investment in advertising. The cigarette companies would have a fine opportunity to "milk" their brands for profit.

The cigarette companies already know how to milk a brand after they cease advertising it. For example, substantial quantities of nonfilter Old Golds have been sold in the last couple of years despite the fact that Lorillard has practically stopped advertising them.

If advertising were stopped, the cigarette companies would generate large amounts of cash each year, which they could either liquidate to stockholders or use to diversify. The former is not likely because of our tax structure and because no executive likes to liquidate himself out of a job. In the latter case, (a process which had already begun in earnest by 1969) much of the capital would go to create new jobs in other industries.

Either way, I would guess that a cigarette stock would have a very solid value if advertising were banned. The same type of predictions would apply to a warning, but the effects would not be as sweeping.

Effect on advertising media 3

As of 1964, the advertising media had already been hit by the Surgeon General's report. Some radio and television stations voluntarily restricted cigarette advertising to certain hours of the day, while others cut it off completely. Some magazines and papers have always refused to accept tobacco advertising, notably the Reader's Digest. And the cigarette advertisers set up an authority to regulate copy and media.

A warning requirement would not hit the media as hard as a ban, of course. But a warning that really affected consumption would make advertising less profitable for the cigarette firms, and the firms would therefore advertise less. Television would lose more than $120 million in advertising revenue ($220 million in 1967), about 7 percent of its total revenue in 1963. But that would not represent a dead loss to television stations and networks. Television time is limited, especially on networks, and the available time is therefore rationed among potential advertisers. If cigarette advertising were banned, the television time could be sold to other advertisers, though at a somewhat lower price.

Television stations are charged with the public interest to a greater extent than are other communications media, because they are given a free franchise for a channel. This franchise gives them some monopoly power. Therefore, the television people should be particularly slow to complain about the loss of cigarette advertising revenue if it is in the public interest.

Radio would lose an estimated $20 million in cigarette-advertising revenue as of 1963. Other advertisers would not replace this revenue, because unlike television there are never more advertisers who wish to buy time than there is time available at current rates. But radio stations also have a franchise granted free by the public.

The $34 million loss (very slightly more in 1967) to general and farm magazines would be a complete loss, about 7 percent of their total revenue, and costs remain almost the same no matter how much advertising they carry, which makes the loss even more serious. The magazines would not find other advertisers to replace cigarettes, and some magazines would feel a considerable strain. But since it would hit them all, they could all be expected to reduce their editorial cost somewhat, without fear of losing advertisers or circulation to competition. This might cushion the impact somewhat.

The $18 million lost to newspapers (less in 1967) would be only 1⁄2 of 1 percent of their advertising revenue.

Effect on advertising agencies

The advertising-agency business would take a beating if cigarette advertising were banned. Agencies would also be hurt if a warning in the advertisements were required, because in that case total cigarette advertising would decrease. Madison Avenue type agencies would lose approximately $200 million billing ($266 million in 1967; Advertising Age, June 17, 1968, p. 8) of their total of perhaps $4 billion (more in 1967), about 5 percent of their total.* (Actually, only 15 percent of the $200 million-$30 million-stays with the agencies. The rest goes to the media.) Perhaps a thousand copywriters, account executives, and other agency people would be scurrying about looking for jobs, and the job market for advertising men would be glutted for a while.

It is interesting to note that after the Surgeon General's report came out, some major advertising agencies said that they would refuse to handle cigarette advertising because they considered it immoral. None of these agencies had a cigarette account at the time. Their statements do mean something, nevertheless. Effect on the economy as a whole

The total cigarette market was about $6.8 billion in 1963 (Advertising Age, August 30, 1965, p. 259), and $8.9 billion in 1967 (Business Week, April 28, 1969, p. 82). Excluding taxes, the industry accounts for $3.6 billion ($5.2 billion in 1967), much less than 1 percent of the Gross National Product.

3 Estimates in this section are from data contained in the Statistical Abstract 1963, p. 847: 1968, pp. 783-785. and The World of Advertising (Chicago: Advertising Age, Jan. 15. 1963, p. 29). One may make rough estimates for 1967 by multiplying each 1963 figure by 1.5.

Estimates by the author for 1963 based on yearly Advertising Age statistics on total big-agency billing, and earlier-cited estimates of cigarette advertising.

In the long run Americans tend to spend a fairly constant percentage of their total yearly income, year after year. This suggests that a decrease in cigarette sales would lead to a compensating increase in other spending. If so, the effect on the economy as a whole would be lessened. Exactly how much the first impact would be, we cannot say. It would be somewhere between no effect and $180 million (i.e., 5 percent of $3.6 billion).

On the other side of the ledger, the "multiplier effect" would magnify the ill effects of whatever decrease in spending does take place, by a factor of two or three. This effect is due to the spending of money again and again by people in the business chain. In other words, if people saved half of the $180 million drop in cigarette sales, the drop in national income would then be between $180 million and $270 million.

1. The effect on the avertising media will not be profound. There should be no effect on television programming, and slight effect on the station stockholders. Newspapers will not be affected much.' Some magazines and radio stations might be hurt painfully, but they are not likely ot be forced out of business.

2. The economy's progress and overall vitality would hardly be affected. There would be a disturbance in tax revenues at federal, state, and municipal levels, but it would cost no more in the long run to levy other taxes.

3. A clutch of advertising agencies would be hit hard. But the overall magnitude of the effect would be infinitesimal-maybe a thousand jobs and about $33 million net revenue lost.

In sum, then, we must balance the expected effect on health against the expected effect in employment and earnings.

4. The above estimate of the decrease in consumption was in terms of number of cigarettes, and not in number of smokers. We have no idea whether a 5 percent decrease in cigarette consumption would mean 5 percent fewer smokers or 10 percent fewer smokers or zero percent fewer smokers. We predict only that the number of cigarettes smoked will decrease by 5 percent.

5. Using the data from the Surgeon General's report, we estimated that for each cigarette smoked, someone's life expectancy is decreased by seven minutes. Putting together the previous estimates, one can say that it takes a reduction of 880 cigarettes to produce a drop of $1 in a tobacco-worker's earnings. And a drop of that many cigarettes means that someone's life expectancy goes up by 880 x 7 minutes 104 hours. The drop in both consumption and earnings would be less in subsequent years. But they would stay in step with each other, so the same type of dollars-for-hours-of-life relationship would hold.

=

In any case, a small yearly decrease in cigarette sale and cigarette advertising, made even smaller by a shift to other forms of tobacco, would not be even a drop in the bucket for the economy as a whole. (The aforegoing discussion is best read in the light of Chapter 8.)

Cigarette smoking affects the federal economy and the economies of the states and some cities, too, by way of taxes paid on cigarettes. Federal excise taxes amount to $2 billion, state taxes are above $1 billion, and municipal taxes are $40 million (Bureau of Economics, n.d., p. 12; Statistical Abstract, 1963; Statistical Abstract, 1968, p. 409). These taxes are important to the tax-collecting bodies. But at first the loss would only be 5 percent of taxes that represent 2 percent of total government revenues. Furthermore, if taxes are not collected one way they can be collected another way, at the same total cost to the public.

On the other hand, cigarettes may cost the economy far more than they contribute. Louis Dublin estimated that cigarettes cost the nation $10 billion annually in the lost services and earnings of men killed prematurely by cigarettes (personal correspondence). My own estimate is a loss of more than $3.6 billion, based on a year of life lost by the average smoker before the age of 65 (Simon, 1963, p. 117), half of the men in the United States being smokers, and an annual payroll of $322 million. Furthermore, deaths caused by smoking decrease consumption spending. In the hundred or so hours lost by each dollar cigaretteindustry earnings, a live person would spend more than $20. This consumption spending may be important to the economy." (Harry Skornia pointed out to me

5 A. James Heins chided me for mentioning only the benefits of people living longer and not mentioning that if people live longer beyond the time they stop working they consume what otherwise would be consumed by others. To get a rounded answer about the net effect on the economy of reduced cigarette consumption and increased life expectancy one would need to construct a complete model of the economy and run it with two sets of parameters, one set appropriate for continued cigarette consumption and the other for reduced consumption. Crucial inputs would include the extent to which men (who are more important in production than women) die relatively earlier than women on account of smoking; the amount of work people do after the nominal retirement age; the amount that