quired by the Act itself. According to the Act's "Declaration of Policy," the warning was required so that "the public may be adequately informed that cigarette smoking may be hazardous to health." The day after the Act was signed into law, the FTC is- sued an order vacating its trade regulation rule (FTC 1965). The Federal Cigarette Labeling and Advertising Act also required that the FTC trans- mit annually to Congress a report on the effectiveness of cigarette labeling, current cigarette advertising and promotion practices, and recommendations for legislation. In its first report to Congress, submitted in June 1967, the FTC recommended that the health warning be extended to cigarette advertisements and be strengthened to read: "Warning: Cigarette Smoking Is Dangerous to Health and May Cause Death from Can- cer and Other Diseases" (FTC 1967). On May 20, 1969, just before expiration of the congressionally imposed moratorium on its action, the FTC announced a proposed rule that would have required all cigarette advertisements "to disclose, clearly and prominently, . . that cigarette smoking is dangerous to health and may cause death from cancer, coronary heart disease, chronic bronchitis, pulmonary emphysema, and other diseases" (FTC 1969a). During this time, hearings were being held in Congress on cigarette labeling and ad- vertising issues. On April 1, 1970, the Public Health Cigarette Smoking Act of 1969 (Public Law 9 I-222). which banned cigarette advertising on television and radio, was signed into law. The labeling provisions of this law, like its predecessor's, were less stringent than the FTC regulations they preempted. The Act (effective November 1, 1970) did strengthen the health warning on cigarette packages to read: "Warning: The Surgeon General Has Determined That Cigarette Smoking Is Dangerous to Your Health." However, it continued to prohibit any other health warning requirement for packages and to prohibit the FTC (through June 1971) from issuing regulations that would require a health warning in cigarette advertising. In late 197 1, after the second congressionally mandated moratorium on its actions had expired, the FTC announced its intention to file complaints against cigarette com- panies for failure to warn in their advertising that smoking is dangerous to health. Sub- sequent negotiations between the FTC and the cigarette industry resulted in consent or- ders on March 30, 1972, requiring that all cigarette advertising display "clearly and conspicuously" the same warning required by Congress on cigarette packages (FTC 1981b). The 1972 consent order specified the type size of the warning in newspaper, magazine, and other periodical advertisements of various dimensions. For billboard advertisements, the size of the warnings was specified in inches (PIG 1972). In 1975, the U.S. Government filed a complaint in the U.S. District Court for the District of Columbia for alleged violations of the consent order, including failure to display the health warning in some advertising, billboard warnings in letters smaller than required, and improper placement of the warning in some advertisements (FTC 1982). This ac- tion ultimately led to judgments in 198 1 by the U.S. District Court for the Southern Dis- trict of New York against the six major cigarette companies (U.S.A. v. Liggett et 01. 198 1; U.S.A. v. R J. Reynolds 1981). Among other things, these judgments required the cigarette companies to use larger lettering in billboard advertisements. Under this settlement, the format and size of the warning for advertisements of various dimensions 476 TABLE 2.-Major legislation related to information and education about tobacco and health in the United States Law Date Labeling requirements Major provisions and Federal agency affected Congressional Advertising reporting requirements Other Federal Cigarette Labeling and Advertising Act (PL 89-92) 1965 Health warning on cigarette packages Preempted other package warnings Temporarily preempted any health warning on cigarette advertisements F-w Annual report to Congress on health consequences of smoking (DHEW) Annual report to Congress on cigarette labeling and advertising (FTC) Public Health Cigarette Smoking Act (PL 9 I-222) 1969 Strengthened health warning on cigarette packages Preempted other warnings on packages Temporarily preempted FTC requirement of health warning on cigarette advertisements" F-m Prohibited cigarette Annual report to Congress on advertising on television and health consequences of radio (DOJ) smoking (DHEW) Preempted any State or local requirement or prohibition based on smoking and health with respect to cigarette advertising or promotion Annual report to Congress on cigarette labeling and advertismg (FTC) Little Cigar Act (PL 93-109) 1973 Extended broadcast ban on cigarette advertising to "little cigars" (DOJ) were specified in acetate exhibits that am maintained on file at the FTC. The Com- prehensive Smoking Education Act of 1984 (Public Law 98-474) again increased the size of the letters, but in the case of billboard ads, it did so only by requiring that all let- ters be uppercase. This Act was the first to codify into law the requirement for and the sizes of the warnings on ads. In 198 1, the FTC sent a staff report to Congress that concluded that the warning ap- pearing on cigarette packages and in advertisements was no longer effective. The report noted that the warning did not communicate information on the significant, specific risks of smoking and concluded that the warning had become overexposed and "worn out" (FTC 198 1 b). The report recommended changing the shape of the warning to a circle-and-arrow format (for example, see Figure 1), increasing the size of the warning, and replacing the existing warning with a system of short rotational warnings. FIGURE l.-Health warnings required for smokeless tobacco advertisements (except billboards) Some of these recommendations were enacted by Congress as part of the Comprehen- sive Smoking Education Act (Public Law 98-474), which was signed into law on Oc- tober 12, 1984. Effective October 12, 1985, it required cigarette companies to rotate four warnings on all cigarette packages and in advertisements (see Table 3). This was the first time that health warnings on cigarette advertisements were the result of legis- lative rather than regulatory action. The four warnings mandated for cigarette adver- tisements on outdoor billboards were slightly shorter versions of the messages required in other advertisements and on packages. The Act did not amend the existing prohibi- tion of any other health warnings on cigarette packages and the preemption of Stateac- tion, but it did not impose a similar preemption of other health warnings byFederal authorities in cigarette advertising. The Comprehensive Smoking Education Act of 1984 required each cigarette manufacturer to obtain FTC approval for its plans to implement the rotational warning 479 TABLE 3.-Health warnings required on tobacco packages and advertisements in the United States. CIGARETTES Warning(s) Effective dates Applicability Packages Advertise- ments CAUTION: Cigarette Smoking May Be Hazardous to Your Health. January 1, 1966- October 31.1970 X WARNING: November I, I970- X The Surgeon General Has Determined That October I 1.1985 Cigarette Smoking Is Dangerous to Your Health. 1972-October I I, 1985 Xa SURGEON GENERAL'S WARNING: Smoking Causes Lung Cancer. Heart Disease, October 12, 198%present Emphysema, and May Complicate Pregnancy. SURGEON GENERAL'S WARNING: Quitting Smoking Now Greatly Reduces Serious Risks to Your Health. SURGEON GENERAL'S WARNING: Smoking by Pregnant Women May Result in Fetal Injury, Premature Birth, and Low Birth Weight. SURGEON GENERAL'S WARNING: Cigarette Smoke Contains Carbon Monoxide X Xb SMOKELESS TOBACCO Warnings Effective dates Applicability Packages Advertise- ments WARNING: This product may cause mouth cancer. February 27,1987-present X x' WARNING: This product may cause gum disease and tooth loss. WARNING: This product is not a safe alternative to cigarettes. `Required by Federal Trade Commission consent order. All other warnings required by Federal legislation. h-rhe four wammga mandated for cigarette advenrsements on outdoor billboards are slightly shorter versions of the same messages `The wammgs on adveni\ement\ must appear m a circle-and-avow format (see Figure I ). No warnings are required on outdoor billboards 480 system. Legislation was subsequently enacted that permitted certain smaller manufac- turers and importers to display simultaneously all four warnings on packages instead of by quarterly rotation (Nurse Education Amendments of 1985, Section 11, amending section 4(c) of the Federal Cigarette Labeling and Advertising Act, I5 U.S.C. 1333(c)). This practice is now followed by 20 to 25 small manufacturers and importers. More recently, Congress has extended requirements for warning labels to smokeless tobacco products. In early 1986, two national review groups, a National Institutes of Health Consensus Development Conference (US DHHS 1986a) and the Surgeon General's Advisory Committee on the Health Consequences of Using Smokeless Tobacco (US DHHS 1986c), issued reports concluding that smokeless tobacco can cause oral cancer and a number of noncancerous oral conditions. Between 1985 and 1986, the State of Massachusetts adopted legislation requiring warning labels on pack- ages of snuff, and 25 other States considered similar legislation (Connolly et al. 1986). The Massachusetts law was preempted before it took effect by the Federal Com- prehensive Smokeless Tobacco Health Education Act of 1986 (Public Law 99-252). which was signed into law on February 27. 1986. The Act requires one of three wam- ings to be displayed on all smokeless tobacco packages and advertisements (except billboards) (Table 3). It requires that the three package warnings "be randomly dis- played . . . in each l2-month period in as equal a number of times as is possible on each brand of the product and be randomly distributed in all parts of the United States in which such product is marketed." On advertisements, the law requires rotation of each warning every 4 months for each brand. The warnings on advertisements are required to appear in the circle-and-arrow format recommended earlier by the FTC for cigarette warnings (FTC 1981b) (Figure 1). The Act prohibits Federal agencies or State or local jurisdictions from requiring any other health warnings on smokeless tobacco packages and advertisements (except billboards). No other Federal, State, or local actions were preempted by the Act. The FTC issued regulations implementing the law on Novem- ber 4,1986 (FTC 1986b). Package inserts provide the opportunity to present more detailed information to the consumer than is possible with a warning label. They are a standard way of providing consumers with information about pharmaceutical products, but they have not been proposed for tobacco products in the United States. When used for prescription phar- maceuticals, patient package inserts have been generally effective in providing patients with information (US DHHS 1987d; Morris, Mazis, Gordon 1977) but have not been demonstrated to be effective in altering behavior (Dwyer 1978; Morris and Kanouse 1982). Information about smoking risks is included in the package insert for one class of pharmaceutical agents marketed in the United States. After several studies published between 1975 and 1977 reported that smoking increases the cardiovascular disease risks associated with oral contraceptive use (US DHEW 1978), the Food and Drug Ad- ministration (FDA) issued a regulation on January 3 1, 1978 requiring that as of April 3, 1978, packages of oral contraceptives contain a printed leaflet with the following boxed warning: Cigarette smoking increases the risk of serious adverse effects on the heart and blood ves- sels from oral contraceptive use. This risk increases with age and with heavy smoking (15 481 or more cigarettes per day) and is quite marked in women over 35 years of age. Women who use oral contraceptives should not smoke (FDA 1978). The information provided to consumers of another nicotine-containing product con- trasts with the information provided to consumers of tobacco products. The patient package insert for nicotine polacrilex gum, a nicotine-containing product approved by the FDA as an adjunct to smoking cessation programs, informs users of the addictive- ness of nicotine and its potential effects on the fetus (US DHHS 1988). The product insert does not mention the risks of cigarette smoking, but it does state: "Warning to female patients: Nicorette contains nicotine which may cause fetal harm when ad- ministered to a pregnant woman. Do not take Nicorette if you are pregnant or nursing." The insert also warns that dependence on Nicorette "may occur when patients who are dependent on the nicotine in tobacco transfer that dependence to the nicotine in Nicorette gum." Effectiveness of Cigarette Warning Labels In May 1987, the Assistant Secretary for Health, Department of Health and Human Services, transmitted a report to Congress on the effects of health warning labels (US DHHS 1987d). Based on a review of the research literature, the report reached three major conclusions. First, health warning labels can have an impact on consumers if designed to take account of factors that influence consumer response to warning labels (e.g., a consumer's previous experience with the product, previous knowledge of the risks associated with product use, and education and reading levels). Second, health warning labels can have an impact upon the consumer if the labels are designed effec- tively (e.g., visible format and providing specific rather than general information). Third, studies that have examined the impact of health warning labels in "real world" situations have concluded that the labels did have an impact on consumer behavior. The report cautioned, however, that the results of these studies "cannot be regarded as conclusive evidence that health warning labels are necessarily effective in all situa- tions." This Section reviews evidence related to the effectiveness of cigarette warning labels in the United States. As noted above, the Federal Cigarette Labeling and Advertising Act of 1965 (Public Law 89-92), which required the first warning label on cigarette packages, stated that the health warning was required so that "the public may be adequately informed that cigarette smoking may be hazardous to health." More specific communications objec- tives were not set by legislation mandating warning labels. Generally, however, the goal of warning labels has been to increase public knowledge about the hazards of cigarette smoking. Such knowledge might deter individuals from starting or continu- ing to smoke. Despite the fact that cigarette warning labels have been required since 1966, there are few data about their effectiveness in meeting any objective. As described below, empirical evidence is available about the cigarette warnings' visibility to consumers, and it is consistent with analyses based on communications theory. However, there are no controlled studies to permit a definitive assessment of the independent impact of 482 cigarette warning labels on knowledge, beliefs, attitudes, or smoking behavior. In par- ticular, there has been little evaluation of the impact of the rotating warning labels re- quired since 1985. If warning labels are to have any effect, they must actually appear on packaging and in advertising as required by law. Available evidence indicates that the tobacco in- dustry has complied with disclosure obligations. For example, a study examining health warnings in magazine ads as an indicator of the industry's compliance with the 1984 labeling legislation found that the industry complied with the law (Davis, Lyman, Binkin 1988). The U.S. Department of Justice is empowered to enforce the disclosures required by the various labeling laws. According to the FTC (FTC 1967. 1969b, 1974, 1982, 1986a, 1988a,b) no actions have been brought by the Department of Justice for violations of labeling regulations, and the Commission has brought no action for failure to include the warnings in advertising (with the exception of the billboard and transit advertising enforcement proceedings discussed above). As of October 1988, no action had been sought against a cigarette manufacturer for a violation of the Comprehensive Smoking Education Act of 1984. Despite the industry's compliance with the required warning labels, there is empiri- cal evidence that the public did not pay much attention to the pre- 1985 labels in adver- tisements; little information is available about the visibility of warning labels on pack- aging. In a Starch Message Report Service test of 24 different magazines in 1978, only 2.4 percent of the adults exposed to the cigarette ads read the pre-1985 Surgeon General's warning in those ads (FTC 198 1 b). Similarly, a study of seven Kool ads con- ducted in 1978 for the Brown and Williamson Tobacco Company found that only 2.4 percent of the respondents read the entire warning; the average time spent examining the warning was less than 0.3 seconds. In an advertising copy test conducted for the Liggett and Meyers Tobacco Company in 1976, no respondents read the entire wam- ing (FTC 198 1 b). More recent studies of later cigarette and smokeless tobacco adver- tisements suggest that little attention is paid to the post-1984 health warnings. An eye- movement study examined the rotational cigarette warnings in magazine ads in a sample of 61 adolescents. Over 40 percent of the subjects did not view the warning at all; another 20 percent looked at the warning but did not read it (Fischer et al, 1989). Similarly low levels of warning recall were found for the recently introduced smokeless tobacco warnings (Popper and Murray 1988). These findings are consistent with analyses of the visual imagery of tobacco adver- tising, which note that the structures of the ads draw consumers' attention away from the warnings contained in the ads (Richards and Zakia 198 1; Zemer 1986). It has also been argued that the sheer volume of cigarette advertising, all applying the basic themes of product satisfaction, positive image associations, and risk minimization (Popper 1986b), overwhelm the in-advertisement warnings (Schwartz 1986). In some advertising media, the cigarette warnings may not be readable. In a study of cigarette advertisements on 78 billboards and 100 taxicabs, Davis and Kendrick (1989) compared the readability of the Surgeon General's warning with recognition of the content of the cigarette advertisement. Under typical driving conditions, they found that a passing motorist could read the warning in about half of street billboard advertisements and in only 5 percent of highway billboard advertisements. The wam- 483 ing could not be read by a stationary observer in any of the taxicab advertisements. In contrast, the brand name could be read and notable imagery in the advertisements could be identified in almost all cases. Cullingford and coworkers (1988), using a model to assess the optical limits of the eye, showed that only about half of the health warnings on 37 billboard cigarette advertisements in Australia were legible to passing motorists; on the other hand, 98 percent of the brand names were legible. Despite these findings, a national survey conducted by Lieberman Research, Inc. (1986) showed moderate recall of the post- 1984 warnings 9 months after they began to appear on packages and advertisements. In this random survey of 1,025 Americans 18 years of age and older, 64 percent of all respondents and 77 percent of cigarette smokers said they recalled seeing one or more of the new warnings on cigarette packages. Lieberman concluded that this "represents a high level of penetration in a relatively short time period." Respondents were also asked whether they recalled seeing each of the four warnings as well as the pre- 1985 warning and a fictitious warning ("Smoking reduces life expec- tancy by an average of 6 years"). Recall of the true warnings ranged from 28 to 46 per- cent of a11 respondents (40 to 55 percent of smokers); recall of the carbon-monoxide warning was lowest among the four. Recall of the pre- 1985 warning was substantial- ly higher (85 percent of all respondents, 94 percent of smokers). Recall of the fictitious warning was 10 percent for the total sample as well as for smokers. Because the fic- titious warning differed in style from the true warnings by presenting quantitative in- formation, it is possible that stated recall of the fictitious warning was lower, at least in part, because of inferences made by respondents (as opposed to genuine differences in recall). The proportion who believed that a particular warning was "very" or "fairly" effective in convincing people that smoking is harmful ranged from 40 percent for the carbon-monoxide warning to 76 percent for the warning about lung cancer, heart dis- ease, emphysema. and complications of pregnancy (the corresponding proportion for the pre- 1985 warning was 56 percent). Analyses of the wording and format of mandated health warnings have identified reasons why their impact may be limited even if they are noticed and read. Use of con- ditional words such as "can" or "may" anywhere in the warning can dramatically reduce the effect of the entire warning (Linthwaite 1985). Two of the current rotational wam- ings include the word "may." The other two warnings ( "Quitting Smoking Now Great- ly Reduces Serious Risks to Your Health" and "Cigarette Smoke Contains Carbon Monoxide") are not warnings but statements of fact; linguistically. consumers might be expected to minimize their impact (Dumas, in press). Furthermore, information in the current warnings is presented technically and abstractly rather than in a concrete and personal manner. A reader is more likely to read and learn information that is made personally relevant as opposed to that which is abstract and technical (Fishbein 1977). Researchers who have addressed the format of warnings have found that consumers' attention will be most effectively caught by novel formats (Cohen and Srull 1980). This line of study has suggested that the communications effectiveness of the post-1984 warnings may have been diminished because the same rectangular shape of the pre- 1985 warnings was maintained (Bhalla and Lastovicka 1984). 484 The analysis of time trends in national survey data provides an opportunity to assess the effect of health warning labels on public knowledge of the health risks of smoking. As described in Chapter 4, public knowledge of these health effects has increased since 1966. when the first health warning label was required. Because warning labels were only one of a number of educational influences during this period, most JeseaJCheJS have concluded that it is impossible to isolate the effect of the warnings from other in- formation sources (US DHHS 1987d; FTC 1974; Murphy 1980). Similarly, it is im- possible to determine any independent effect of health warnings on aggregate cigarette sales (FTC 1967. 1969b). In sum, there are insufficient data to determine either the in- dependent contribution of cigarette warning labels to changes in knowledge or smok- ing behavior or the precise role played by warning labels as part of a comprehensive antismoking effort. Perhaps the most powerful indirect index of the effect of health warnings. along with other sources of information, is the number of smokers and consumers in general who remain unaware of the health risks of smoking. After a comprehensive review of studies on health risk awareness, including publicly generated studies and those con- ducted by the tobacco industry, the FTC concluded that significant numbers of con- sumers in general and even higher numbers of smokers were unaware of even the most rudimentary health risk information about smoking (FTC I981 b). It was this lack of consumer awareness that led the FTC to call for revised and expanded rotational warn- ings for cigarettes. More recent data reveal that a substantial minority of smokers still does not believe that smoking causes lung cancer, heart disease, emphysema, and other diseases, and the majority of smokers underestimate the degree of increased health risk posed by smoking. (See Chapter 4.) Summary As a result of policies described in this Section, a system of rotating health warning labels is currently required for all cigarette and smokeless tobacco packaging and ad- vertisements in the United States. This system, established by congressional legisla- tion in I984 (for cigarettes) and 1986 (for smokeless tobacco products), achieves a por- tion of one of the Health Objectives for the Nation for 1990: By 1985, the present cigarette warning should be strengthened to increase its visibility and impact, and to give the consumer additional needed information on the specific multiple health risks of smoking. Special consideration should be given to rotational warnings and to identification of special vuhrerabte groups. The 1984 Act provided the consumer with some of that "needed information," al- though the four mandated warnings provide less information than would have been provided by the 16 warnings described to the U.S. Congress in the 1981 FTC Report (FTC 198lb; Keenan and McLaughlin 1982). There is no legislated mechanism for monitoring the visibility or communications effectiveness of existing warning labels, and there are insufficient data to determine whether the visibility and impact Of the warnings have increased as a result of the 1984 Act. Furthermore, current legislation does not provide a mechanism for updating the content of labels to reflect advances in 485 knowledge about health effects and smoking behavior. One example of changing knowledge is the growing scientific awareness of the addictive nature of tobacco use, which was the subject of the 1988 Surgeon General's Report (US DHHS 1988). In that Report, the Secretary of Health and Human Services, the Assistant Secretary for Health, and the Surgeon General recommended that a new health warning label on the addic- tive nature of tobacco use be required on cigarette and smokeless tobacco packages and advertisements. On the day of the Report's release (May 16, 1988), legislation was in- troduced in the U.S. Senate that would require a warning to read: "Smoking is addic- tive. Once you start, you may not be able to stop" (S. 2402). Other bills that include provisions calling for a warning label on addiction have also been introduced in Con- gress. As of November 1988, this legislation was not enacted. Currently. labels are not required on cigarettes made for export or on cigarettes manufactured abroad by U.S. tobacco companies. Federal law does not require wam- ing labels on other tobacco products, such as cigars, pipe tobacco, and roll-your-own cigarette tobacco, despite the established health risks associated with cigar and pipe smoking (US DHEW 1979; US DHHS 1982a, 1984; Chapter 2). During the early 1970s there was particular concern about the health risks for individuals who smoke "little cigars" (US DHEW 1973). In its 1974 report to Congress (FTC 1974), the FTC recommended that the following warning be required on little-cigar packages: "Wam- ing: Smoking Little Cigars May be Dangerous to Your Health if Inhaled and Smoked in the Same Quantities as Cigarettes." The Little Cigar Act of 1973 (Public Law 93- 109) extended the broadcast advertising ban for cigarettes to little cigars, but neither this Act nor subsequent legislation extended requirements for health warnings to little cigars (Table 2). A warning label will appear on cigars and pipe tobacco sold in California, as a result of an agreement reached on October 18, 1988, between tobacco manufacturers and the State of California. Twenty-five tobacco manufacturers, along with eight retailers, had been sued by California's Attorney General for failing to comply with the State's Safe Drinking Water and Toxic Substances Enforcement Act, which requires warnings on all consumer products containing chemicals known to cause cancer or reproductive toxic effects (Wilson 1988a; Kizer et al. 1988). Because existing distribution systems for cigars do not easily permit the labeling of cigars destined only for California, the president of the Cigar Association of America indicated that most cigars sold in the United States would carry warning labels (Wilson 1988a). As of October 1988, the effect of the settlement on warning labels for pipe tobacco sold outside California was unknown. Tobacco labeling requirements in other countries (Roemer 1982,1986) provide com- parisons for current labeling practices in the United States. Outside the United States, six countries (Finland, Iceland, Ireland, Norway, Sweden, and the United Kingdom) have enacted a rotational warning requirement. A Swedish law, adopted in 1976, re- quires the rotation of 16 warning statements on cigarette packages. Ireland requires the rotation of three brief, direct statements on cigarette packages and advertise- ments: `SMOKING CAUSES CANCER," "SMOKERS DIE YOUNG," and "SMOK- ING KILLS!" In the United Kingdom. one of six rotated warnings indicates smoking- attributable mortality: "More than 30,000 People Die Each Year in the UK from Lung 486 Cancer." Since 1985, Iceland has required the rotation of pictorial warnings (Figure 2). Several countries also require health warnings on packages of cigars and pipe tobac- co. On packages of cigars, cigarillos, and pipe tobacco, for example, Ireland requires the warning: "SMOKING SERIOUSLY DAMAGES YOUR HEALTH." On June 29, 1988, Canada's House of Commons enacted a new labeling law as part of a comprehen- sive package of smoking restrictions, the Tobacco Products Control Act (House of Commons of Canada 1988). Canada's current cigarette warning labels will be replaced by a mandatory package insert that details all known health risks of smoking. 4. If you map rmokin "0" mlPmw YOU, heah ind minw~your lila exp.ctAncy. 5. Smokmg II I heallh problem you C." hap to s&o. rlimhlibw- FIGURE Z.-Health warnings on tobacco packages in Iceland according to regula- tion no. 49911984 SOURCE: Blondal and Magnusson (1985). 487 Disclosure of Tobacco Product Constituents History and Current Status The FTC has also been concerned with the disclosure, on packaging and in advertis- ing, of information about the constituents of tobacco smoke (e.g., tar, nicotine, and car- bon monoxide). More recently, there has also been growing interest in the identity and amounts of other ingredients added to tobacco products during the manufacturing process. The first industrywide regulation occurred even before the release of the first Sur- geon General's Report. In the mid- to late 1950s. many cigarette advertisements made conflicting claims for the tar and nicotine levels of various brands. This period became known as the "Tar Derby" (Wagner 197 1 a; Whiteside I97 1). On September 15, 1955, after a year of conferences with the cigarette industry, the FTC promulgated cigarette advertising guidelines "for the use of its staff in the evaluation of cigarette advertising" (FK 1964b). These guidelines, among other things, sought to prohibit cigarette ad- vertising that made unsubstantiated claims about the level of nicotine, tars, or other sub- stances in cigarette smoke. By 1960, the FTC obtained agreements from the leading cigarette manufacturers to eliminate from their advertising unsubstantiated claims of tar and nicotine content (FTC 1964b). As the previous section noted, the FTC proposed three rules addressing cigarette labeling and advertising shortly after the release of the 1964 Surgeon General's Report (FK 1964a). The third proposed rule provided that: No cigarette advertisement shall contain any statement as to the quantity of any cigarette- smoke ingredients (e.g., tars and nicotine) which has not been verified in accordance with a uniform and reliable testing procedure approved by the FTC. This recommendation was not among the final regulations promulgated by the FTC nor in subsequent congressional legislation. Shortly after passage of the Federal Cigarette Labeling and Advertising Act of 1965, the FTC identified a uniform testing system for measuring the tar and nicotine yield of cigarettes (Pillsbury et al. 1969; see Chapter 5). The FTC determined that meaningfui disclosure of tobacco product constituents required the availability of accurate information obtained by standardized testing methods. In 1966, the Commission sent a letter to U.S. cigarette manufacturers approving their factual statements of tar and nicotine content in advertising, if based on tests conducted using the approved method. In 1967, the FTC activated its own laboratory to analyze the tar and nicotine content of cigarette smoke. At the request of the Chairman of the Senate Commerce Committee. the FTC began to test and report periodically to Congress the tar and nicotine content of various cigarette brands (FTC 1981a). In 1981, the FTC first published carbon monoxide yields, based on its own laboratory tests, along with data on tar and nicotine yields (FTC 1981a). In 1983, the FTC determined that its testing procedures may have "significantly un- derestimated the level of tar, nicotine and carbon monoxide that smokers received from smoking" certain low-tar cigarettes and sought comments pursuant to modifying its testing procedures (FTC 1988a). One cigarette brand, Barclay, manufactured by the 488 Brown and Williamson Tobacco Company, was permanently enjoined from including in its advertising, packaging, or promotion the tar rating the brand received using the FTC test methods because of problems with the testing methodology and consumers' possible reliance on that information (FTC v. BIXMYI and Williamson 1983). On April 15, 1987, the FfC announced the closing of its in-house laboratory that tested cigarettes for tar. nicotine, and carbon monoxide levels. The FTC attributed its decision to the cost of running the laboratory and the fact that the information was avail- able from the cigarette industry's laboratories, whose methodology was identical to that used by the FTC. The FTC stated that it would collect tar, nicotine, and carbon monoxide ratings from the industry for inclusion in its annual report to Congress pur- suant to the Federal Cigarette Labeling and Advertising Act (FTC 1987; MacLeod 1987). As a result of these actions, a mechanism has been in place whereby information about tar, nicotine, and carbon monoxide yields of cigarettes becomes part of the public record. However, this information is not as readily accessible to consumers as it would be if it were disclosed on all packages of tobacco products or in advertising. Recom- mendations for uniform disclosure of cigarette constituents have been made previously by the FTC and the Department of Health and Human Services, and a specific goal was set by the Public Health Service's 1990 Health Objectives for the Nation (US DHHS 1986d): By 1985, tar, nicotine, and carbon monoxide yields should be prominently displayed on each cigarette package and promotional material. In 198 1, the Department of Health and Human Services (DHHS) recommended that "manufacturers should list yields of `tar', nicotine and other hazardous components on their packages and in their advertising with appropriate explanatory information on the health significance of these measurements" (US DHHS 1981a (transmittal letter)). As early as 1969, the PTC (FTC 1969b) recommended that disclosure of tar and nicotine yields be required on cigarette packages as well as in advertisements. The next year, the FTC proposed a regulation requiring cigarette companies to disclose the tar and nicotine content of cigarette brands in their advertisements, based on the most recent FTC test results (FTC 1970). The FTC suspended this proceeding to allow the major manufacturers to implement a voluntary plan for such disclosure. Since 1971, all manufacturers have complied with this plan and voluntarily disclose the tar and nicotine content of cigarette brands in advertisements (FTC 198 I b). There is no industrywide disclosure of tar and nicotine content on cigarette packages; such disclosure is often made voluntarily for cigarettes yielding 8 mg or less of tar but rarely for higher tar brands. (unpublished data, Office on Smoking and Health 1988). Carbon monoxide yields are neither required nor voluntarily disclosed on packages or in advertising, despite a 1982 FTC recommendation that they be required on cigarette packages (Muris 1982). Currently, there are no government requirements for the dis- closure of tobacco smoke constituents to consumers, although, as noted above. levels of some constituents are disclosed voluntarily in advertisements and on some packages by cigarette manufacturers. 489 In addition to tobacco, tobacco products contain other ingredients added in the process of manufacture. The identity of these additives is regarded as confidential in- formation by manufacturers. The Comprehensive Smoking Education Act of 1984 and the Comprehensive Smokeless Tobacco Health Education Act of 1986 required, for the first time, that the manufacturers, packagers, and importers of cigarettes and smokeless tobacco products provide annually to the Secretary of Health and Human Services a list of additives used in the manufacture of these products. The Secretary is required to treat the lists as "trade secret or confidential information," but may report to Congress on research activities about the health risks of these additives and may call attention to "any ingredient which in the judgment of the Secretary poses a health risk to cigarette smokers" (Public Law 98-474, Public Law 99-252). However, the Secretary is granted no specific authority to regulate any such hazardous products. Regulations describing the procedures for protecting the confidentiality of this information have been published (US DHHS 1985a). Analysis of the information on cigarette additives is in progress. Federal legislation on smokeless tobacco (Public Law 99-252) now requires that manufacturers provide to the Secretary of Health and Human Services a specification of the nicotine content of smokeless tobacco products, but it does not require that nicotine content be listed on packages or in advertisements. Currently, one brand of smokeless tobacco is marketed as "light" snuff, and the nicotine content is disclosed on its packaging and advertising. Effects of Disclosure of Tobacco Product Constituents Current Federal law neither requires the disclosure of tobacco product or tobacco smoke constituents on packages and advertising, nor provides for the monitoring of communications effects of voluntary disclosures. The principal public health rationale for requiring disclosure is to inform consumers about the amount of hazardous substan- ces to which they are exposed, so that consumers will be better informed and so that those who do not abstain completely may be able to reduce their health risks by select- ing a brand with a lower concentration of hazardous substances. There is some information that this has occurred. As noted in Chapter 5, the rapid growth in the market share of cigarettes with reduced tar and nicotine yields during the 1970s indicates that consumers can and will make choices based on information about tobacco constituents (US DHHS 198 la). However, there is no clear evidence of sub- stantial health benefits to consumers who switch to lower tar and nicotine cigarettes. The potential health benefit to smokers of making such discriminations is at best limited, because there is no known safe level of tobacco product consumption (US DHHS 198 1 a). As mentioned in Chapter 5, concerns about low-yield cigarettes center around: (I) compensatory smoking behavior among smokers who switch to low-nicotine brands, which might even increase total tobacco smoke intake in some smokers; (2) the increased use of additives with possible adverse health effects in low-yield cigarettes; and (3) the possibility that some smokers who believe these cigarettes to be safe or less hazardous will be less inclined to quit. It is also possible that if smokers saw a more complete listing of the harmful con- stituents of tobacco on packages or in ads, some would stop smoking rather than mere- 490 ly choosing a different brand. Evidence to test this hypothesis has not been collected. The impact of informing smokers about the identity of tobacco product additives, about which consumers know little, is unknown. It is possible that this information might en- courage smokers to stop smoking, or at least to reduce their daily cigarette consump- tion . Mandated Education About Health Risks Government activities to educate the public on smoking and health are not limited to product-oriented warnings to the tobacco consumer. Government policy has required schools to educate students and teachers about the health hazards of tobacco use. Educational messages in the broadcast media were also mandated by Federal policy from 1967 through 1970. School Education Current Status Both public and private efforts to reduce the initiation of smoking by children have targeted schools. Education on tobacco and health may be provided voluntarily in school curricula or may be required by legislation or regulation. For the purposes of this review, such education is considered voluntary if it is based on a decision of the individual teacher or on an action taken by an individual school or school district. A "policy" refers to Federal or State legislation or regulation mandating instruction on tobacco and health. Voluntary initiatives on school education on smoking and health are considered in Chapter 6. Policies restricting smoking in schools by students and teachers are reviewed in Part III of this Chapter. The Federal Government has taken no action to mandate education on tobacco in the Nation's schools. Federal legislation was introduced in the 100th Congress (Adoles- cent Tobacco Education and Prevention Act, H.R. 3658; Atkins 1987) that would re- quire tobacco to be included in drug abuse and education programs established under Sections 4124-4125 of the Drug-Free Schools and Communities Act of 1986 (Public Law 99-750), but this legislation was not enacted. The Surgeon General, the Secretary of Health and Human Services, and the Assistant Secretary for Health have recom- mended that prevention of tobacco use be included, along with instruction on illicit drug use, in school health education curricula (US DHHS 1988). A number of States have enacted laws mandating education about smoking and health in schools. The usual content of mandated instruction is the health effects of tobacco use, often included as a component of general health education or a drugs-and-alcohol curriculum. Few school-based educational programs provide education on cessation methods for students who have already started to smoke (Chapter 6). Policies may re- quire the education of either students or teachers, the latter sometimes as a prerequisite to receiving a teaching certificate. 491 TABLE 4.Etate requirements for school health education on drugs/alcohol/tobacco (1974-81) and on tobacco use prevention (1987) State State requirement for instruction in State requirement for instruction in tobacco drugs/alcohol/tobacco' prevention 1974 1977-1978 1981 I987 Alabama Alaska Arizona Arkansas California Colorado Connecticut Delaware District of Columbia Florida Georgia Hawaii Idaho Illinois Indiana Iowa Kansas Kentucky Louisiana Maine Maryland Massachusetts Michigan Minnesota Mississippi Missouri Montana Nebraska Nevada New Hampshire New Jersey New Mexico New York 0 M M M M M M M M M M M M M M-S M M M M M-S M M-S M M M 0 M M M M M M M M M M M M M M M-S M M M 0 M M M-S M-S 0 M M M M M M M 0 M M M M M M M M M M M M M M M M M M M M M M M M 0 M M M M M 492 TABLE 4.Xontinued State requirement for State requirement for instruction in mstruction in tobacco drugs/alcohol/tobacco" preventlon 1973 1977-197X 19x1 1987 North Carolina Nonh Dakota Ohio Oklahoma Oregon Penn\yIvanra Rhode I\land South Carolina South Dakota Tennessee Texa\ Utah Vermont Virgmia Washington West Virginia Wisconsin Wyoming M M M M M M M M M M M II M 0 M M M M M M M M M M M M M M M TOTAL (mandatory) 35 35 39 20 NOTE: Thirty-four State\ required mstnlctton in drugs/alcohol/tobacco m 1985. The indwldual States were not identified in the repon (ASHA 1987) "M. mandated: 0. optional/permissive: S. secondary school level. Unless otherws noted, pohcie\ refer to both elementary and scondary levelb. SOURCE: ASHA (1976, 1979. 1981): Lovato. Allenwonh.Chan. in press. Surveys of State requirements for school health education for the years 1974, 1977, 1978,1981,1985, and 1987 have been conducted by the American School Health As- sociation (ASHA 1976, 1979, 1981, 1987; Lovato, Allensworth, Chan, in press). Ques- tionnaires were sent to State school health consultants, when identifiable. or to State commissioners of education or health. Between 1974 and 1985, the number of States (including the District of Columbia) mandating school education in the category labeled "drugs/alcohol/tobacco" varied from 34 to 39. with no clear trend over time (Table 4; data not shown for 1985, for which only the total number of States-34--was provided). In fact. several States apparently weakened or repealed preexisting require- ments. In most jurisdictions, the requirement pertained to both elementary and secon- dary school levels. The extent to which education in this broad category specifically 493 required tobacco education is unknown. The results do not suggest that the number of States requiring instruction on the health effects of tobacco use is increasing. In the 1987 survey, mandated curriculum on tobacco use was reported separately from cur- ricula on drug and alcohol use. The prevention of tobacco use is mandated curriculum in 20 States (Lovato, Allensworth, Chan, in press). A separate survey of State legislation enacted as of December 1985 reported similar findings. It found that 18 of 2 1 States providing data required elementary and secon- dary schools to include instruction on the dangers of using tobacco as part of their health education programs (Table 5) (US DHHS 1986e). Several States also require teacher training. Three States (Alabama, Connecticut, and Oklahoma) have directed their departments of education to establish and implement in- service training programs to educate teachers, school administrators, and other school personnel about the effects of nicotine or tobacco use. All educational institutions in Minnesota that provide teacher training must offer programs on the use of and depend- ence on tobacco. Connecticut law requires universities that train teachers to provide instruction on the effects of nicotine and tobacco use and on the best methods for in- structing students on these topics. To receive a certificate to teach or supervise in any public school in Connecticut, a person must pass an examination on the effects of nicotine and tobacco use (US DHHS l986e). Compliance and Effects Little is known about the level of compliance with these State regulations. A 1986 survey of a random sample of 2,000 school districts conducted by the National School Boards Association found that 61, 64, and 62 percent of school districts provide anti- smoking education in elementary school. middle orjunior high school, and high school, respectively (NSBA 1987). The generalizability of the survey is limited by a low response rate (36 percent). It is unclear to what degree this instruction is voluntary or the result of a State requirement. Even less is known about the content or quality of curricula developed to comply with government mandates. Evaluations of voluntary school-based smoking preven- tion programs (Chapter 6) suggest that they can be effective if done well. The extent to which government-mandated school education programs match these results is un- known. Consequently, it is impossible to determine the extent to which govemment- mandated school education has contributed to greater awareness by children of the health consequences of smoking or to reductions in the initiation of smoking. 494 TABLE S.-States requiring school health education on tobacco use effects State School health education In-service Instruction Instructional teacher required for material must training teacher certification be accurate Other Alabama X X Alaska X Arizona California Connecticut Florlda Georgia Idaho Illinois Indiana Iowa Massachusetts Michigan Minnesota Nebraska Ohio Oklahoma Oregon Utah Vermont X X X X X X X X X X X X X X X X X X X Xa Xb Wisconsin X dConnecticut law provides that no cemficate to teach or supervise shall be granted to any person who has not passed a satisfactory examination on the effects of nicotme and tobacco. Corm. Gen. Stat. Ann.. Secnon IO-145a (West Supp. 1964). bFlorida's Cancer Control and Research Act provides that proven causes of cancer, including smoking, should be publicized and should be the subject of educational programs for the prevention of cancer. Fla. Stat. Ann.. Section 381.2712(2)(c~ (West Supp. 196.5). SOURCE: US DHHS (1986el. 495 Broadcast Media History In 1949, the U.S. Federal Communications Commission (FCC) promulgated its Fair- ness Doctrine (FCC 1949). Under this doctrine, which the FCC repealed in August 1988, licensed broadcasters were obligated to encourage and implement the broadcast of all sides of controversial public issues over their facilities, over and beyond their obligation to make available on demand opportunities for the expression of opposing views (FCC 1987). This meant that, as a condition of retaining the required license, broadcasters were required to air both sides of a controversial issue if one side was presented. Subsequent decisions by the FCC indicated that the Fairness Doctrine could require a station to grant free time, even when one viewpoint was presented under paid sponsorship. The FCC did not, however, require that a broadcaster provide equal time for opposing views; only a "reasonable opportunity" for the presentation of opposing views was required (Columbia Law Review 1967). In January 1967, John Banzhaf, an attorney acting as a private citizen, petitioned the FCC to apply the Fairness Doctrine to cigarette advertising. On June 2, 1967, the Com- mission ruled that the doctrine applied to cigarette advertising on television and radio and required broadcasters who aired cigarette commercials to provide "a significant amount of time" to citizens who wished to point out that smoking "may be hazardous to the smoker's health" (FCC 1967). In a subsequent press interview, the FCC's chief counsel gave his informal opinion that a ratio of one antismoking message to three cigarette commercials seemed to him to constitute "a significant amount of time" (Whiteside 1971). The ruling applying the Fairness Doctrine to cigarette advertising went into effect on July 1, 1967. Thereafter, broadcasters began to air an array of antismoking public ser- vice announcements (PSAs). developed primarily by voluntary health organizations and government health agencies (Whiteside 1971). The time "donated" for the anti- smoking spots amounted to approximately 75 million dollars (in 1970 dollars) per year from 1968 through 1970 (Lydon 1970). As discussed in the next section, subsequent Federal legislation, the Public Health Cigarette Smoking Act of 1969, banned cigarette advertising on television and radio, effective January 2, 197 I. Once this occurred and cigarette ads were removed from radio and television, the Fairness Doctrine basis for requiring broadcasters to carry antismoking PSAs was eliminated. Antismoking mes- sages then had to compete for public service advertising time donated by broadcasters. As a result. the frequency of the antismoking spots declined dramatically. According to Lewit, Coate, and Grossman (1981). the number of antismoking PSAs declined by almost 80 percent after 1970, relative to the number aired in 1969, and they were shown at times when youths in particular were not likely to be watching television. 496 Effectiveness The antismoking messages mandated by the Fairness Doctrine might have been ex- pected to increase public knowledge and change public attitudes about smoking. In- directly, they might reduce smoking prevalence and tobacco consumption by stimulat- ing cessation and retarding initiation. The degree to which the messages achieved these goals has been assessed by measuring trends in public beliefs concerning the health hazards of smoking, in smoking prevalence, and in cigarette hales before. during. and after the 1968-70 period. PSAs were only one of a number of societal influences on smoking during that period. Because of the broad reach of the mass media. it is impos- sible to control for these concurrent influences by examining a group that was not ex- posed to PSAs. Consequently, changes in these indices cannot be unequivocally at- tributed to the presence of PSAs. Nonetheless, they offer strong circumstantial evidence for an effect of the PSA campaign. Survey data indicate that PSAs were in fact seen and recalled by large numbers of Americans. O'Keefe (1971) surveyed 621 students below 2 1 years of age and 300 adults in Central Florida. Ninety percent of the sample recalled seeing at least one an- tismoking PSA, and about half of them were able to recall a specific commercial. When asked about the effect of PSAs on their own smoking behavior. 32 percent of smokers reported that they had cut down, 37 percent said they thought more about the effects of smoking than before, and 1 I percent said they stopped smoking temporarily as a result of the commercials. This study, based on the self-reported smoking behavior of a small sample, does not provide definitive evidence for an effect of PSAs on knowledge or cigarette consumption. Analysis of trends in national survey data provides a stronger quality of evidence for the effects of PSAs on knowledge or behavior. National survey data collected before, during. and after the 1968-70 period show consistent but small increases in public knowledge of the health hazards of smoking (see Chapter 4). According to the Adult Use of Tobacco Surveys (AUTSs), the proportion of adults who believed that smoking is hazardous to health was already high before the airing of PSAs. It increased slight- ly during and after the period when PSAs were shown, from 85 to 87 to 90 percent in 1966, 1970, and 197.5, respectively. Similar trends were seen for public beliefs con- cerning the causal relationship between smoking and specific diseases, including lung cancer, heart disease, and chronic obstructive lung disease (Chapter 4). One might ex- pect that the personal and emotional messages in many of the PSAs (Whiteside 197 1) would have a particularly salient effect on personalized acceptance of health risks from smoking (Chapter 4). AUTS data show a larger increase in this factor. coincident with the PSAs. The percentage of smokers who were concerned about the effects of smok- ing on their own health increased from 47 percent in 1966, before the Fairness Doctrine, to 69 and 68 percent in 1970 and 1975, respectively. One must be cautious in attribut- ing these changes solely to the PSA campaign, because increases in public knowledge sometimes continued after the campaign ended and because other informational ac- tivities, such as cigarette warning labels, occurred concurrently in both the public and private sectors. 497 The effect of PSAs on smoking behavior has been assessed by analyzing trends in cigarette sales and smoking prevalence. Analyses of temporal trends in tobacco con- sumption, as measured by cigarette sales, provide evidence for an effect of PSAs in restraining smoking, at least temporarily. For the 3-year periods before (1965-67), during ( 1968-70), and after ( 197 I-73) the Fairness Doctrine PSAs, per capita cigarette sales increased by 2.0 percent, decreased by 6.9 percent, and increased by 4.1 percent, respectively (Chapter 5). Warner ( 1977) compared actual sales figures for the Fairness Doctrine period to projected sales figures (for the same years) based on the trend in sales during the period 1947-67. He predicted that in the absence of PSAs and sub- sequent publicity, consumption would have been 19.5 percent higher than it actually was by 1975. In a regression analysis of the effects of both cigarette ads and the Fair- ness Doctrine PSAs. Hamilton (1972) found that the antismoking messages retarded per capita cigarette consumption by 530.7 cigarettes per year, while the cigarette ads boosted it by 95.0 per year. Schneider, Klein, and Murphy (1981) concluded that the PSAs reduced per capita tobacco consumption by 5 percent. Findings from these and related studies are reviewed in Chapter 8. If PSAs had motivated large numbers of smokers to quit smoking, one would expect to have observed a decline in the prevalence of cigarette smoking, as well as in tobac- co consumption, during the period when they were shown. Prevalence data have some limits compared with cigarette consumption data. Estimates of smoking prevalence are based on individuals' self-reported behavior in national surveys, which is a less objec- tive measure than consumption estimates based on sales data. Furthermore, data on prevalence are collected less frequently than are sales data, making prevalence a less sensitive index of short-term effects. Data on the self-reported prevalence of cigarette smoking from 1965-U show a highly consistent linear trend downward during the en- tire period (Chapter 5). These data do not provide evidence for an independent effect of the PSA campaign on overall smoking prevalence and contrast with the cigarette consumption data cited above. However, Lewit. Coate, and Grossman (1981), who analyzed the effect of PSAs on the smoking prevalence of teenagers, reported an effect in that age group. They found that the teenage smoking rate was 3.0 percentage points lower during the Fairness Doctrine period than during the 16-month period prior to the Doctrine: most of this effect occurred during the time when PSAs were shown. Warner ( 1978) compared cigarette sales data to self-reported cigarette consumption for the years 1964-75. He found that the ratio of self-reported cigarette consumption to cigarette sales ("consumption ratio") decreased from a level of 72 and 73 percent in 1964 and 1966, to 66 percent in 1970, and to 64 percent in 1975. The decrease between 1966 and 1970. years spanning the Fairness Doctrine period, was statistically sig- nificant. Between 1966 and 1970, actual aggregate sales dropped 1 percent, while reported consumption dropped 9.5 percent. One explanation for this decline is a greater underreporting of current smoking because of growing awareness of the health hazards of smoking and the declining social acceptability of smoking (Chapter 5). Warner sug- gested that the Fairness Doctrine PSAs, by causing changes in knowledge and at- titudes, may have been responsible for increased underreporting. More recent data from 1974-85 show that the consumption ratio has remained stable at approximately 72 percent, despite further reductions in the social acceptability of smoking (Chapter 498 5). As mentioned in Chapter 5, the decrease in the consumption ratio reported by Warner may be related to the fact that the self-reported data for 1970 and 1975 were collected by telephone surveys, while the 1964 and 1966 data were collected by in-per- son interviews; the latter technique generally provides slightly higher smoking prevalence estimates than do telephone surveys. In summary, both per capita cigarette consumption changes and regression studies comparing actual cigarette sales to projected sales based on prior trends are consistent with the conclusion that the Fairness Doctrine PSAs affected smoking behavior, at least in the short term. Changes in public knowledge about the health effects of smoking as assessed in national surveys also occurred during the period PSAs were aired. Because of other social influences on smoking during this period, it is impossible to attribute changes in cigarette consumption or public knowledge solely to the airing of PSAs. However, as described further in Chapter 8, they were a prominent component of an- tismoking activities, which in the aggregate had marked effects on smoking prevalence and tobacco consumption in the 25 years since the release of the 1964 Surgeon General's Report. It is unclear whether and to what degree any short-term effects could have been sustained with an ongoing campaign. If PSAs had continued, it is possible that their short-term effects could have been sustained only with the types of message variation, pulsed media placement patterns, and ongoing communications measurement TABLE 6.-Cigarette advertising and promotional expenditures, 1970436 ($ millions) Year Advertising Promotional Total Total in Advertising constant as percentage ( 1986) dollars of total 1970 314.7 46.3 1971 25 1.6 NA 1972 251.6 NA 1973 247.5 NA 1974 306.8 NA 1975 366.2 125.1 1976 430.0 209. I 1977 552.0 247.5 1978 600.5 214.5 I979 749.0 334.4 1980 829.9 412.4 1981 998.3 549.4 I982 1040.1 753.7 1983 1081.0 819.8 I984 1097.5 997.1 1985 1075.0 1401.4 1986 931.8 1450.6 361.0 1019.4 87.2 NA NA NA NA NA NA NA NA NA NA NA NA 491.3 1000.9 74.5 639.1 1231.0 67.3 799.5 1446.6 69.0 875.0 1470.6 68.6 1083.4 1636.6 69.1 1242.3 1653.0 66.8 1547.7 1865.9 64.5 1793.8 2037.6 5X.0 1900.8 2091.9 56.9 2095.2 2211.7 52.2 2476.4 324.1 43.4 2382.4 2382.4 39.1 NOTE: NA, not available. SOURCE. Warner (19Xhb): Federal Trade Commission (I9XXb) 499 and tracking characteristics of ongoing national advertising campaigns (Aaker and Meyers 1987). including those of the cigarette companies themselves. Restrictions on Tobacco Advertising and Promotion Cigarettes are one of the most heavily marketed consumer products in the United States (FTC 198 I b; Davis 1987). Cigarette advertising and promotional expenditures totaled 2.4 billion dollars in 1986 (FTC 1988b). In both actual and constant dollars, these expenditures increased consistently between 1975 and 1985 but fell slightly in 1986, the last year for which data are available (Table 6). A study reviewing 1985 data found that cigarettes were the most heavily advertised category of products in the out- door media (e.g.. billboards), the second most heavily advertised category in magazines (after passenger cars), and the third most heavily advertised subcategory in newspapers (after passengercars and airlines) (Davis 1987). All six of the major cigarette manufac- turers were included among the 100 companies with the highest advertising expendi- tures in 1985 (Davis 1987). According to FTC reports to Congress for the years 1982 and 1983, the major advertising themes associated cigarette smoking with high-style living, healthy activities. and economic, social, and professional success (FTC 1985). Tobacco advertising includes both traditional advertising (in newspapers and magazines, on billboards, and in transit facilities) and promotional activities. Promo- tional activities are diverse and include the distribution of free product samples, coupons for price reductions, and offers for discounted products (often bearing the name of the cigarette brand). Promotional activities also encompass industry sponsorship of cul- tural, sporting, and entertainment events, and sponsorship of community or political or- ganizations. Incentives paid to distributors or retailers are another form of tobacco promotion. Over the past decade, the balance of expenditures has shifted from tradi- tional advertising to promotional activities (Davis 1987). so that by 1986, promotion- al expenditures accounted for 60 percent of the tobacco marketing dollar. compared with only 25 percent of the total in 1975 (FTC 1988b) (Table 6). This Section reviews previous. current, and proposed government policies to regu- late tobacco advertising and promotion. It considers the central public health issue- whether advertising and promotion increase tobacco consumption-and reviews avail- able evidence on this question. The focus of this review is on cigarette advertising and promotion because cigarettes account for the vast majority of both tobacco use and ad- vertising/promotional expenditures. The effects of advertising for other tobacco products have not often been studied. The discussion includes coverage of the smaller body of information about promotional activities beyond traditional advertising be- cause of their growing importance in tobacco marketing. Effects of Tobacco Advertising and Promotion Public health concern about tobacco advertising and promotion is based on the premise that these activities encourage the initiation of smoking and stimulate tobacco consumption, especially by children, while retarding cessation efforts, particularly by adults. It has been suggested that ads promoting low-tar and -nicotine cigarettes may