Tobacco Documents Online

r • • diminishes those other risks.12 Ironically, the 1981 FTC Staff Report said that warnings should be personally relevant to the viewer, and not assert abstract propositions.13 "A is greater than B" is a paradigm of an abstract proposition. Warnings that cigarettes are akin to a drug may in fact make smoking more enticing and more adventuresome to the target teenage group. Cigarettes are much more readily accessible and less expensive than cocaine, marijuana, and other illegal drugs. Attaching the drug label to the smoking activity may give it an allure that smoking would not otherwise have had. Warnings that smoking is addictive also are potentially counterproductive. If youths have begun to experiment with cigarettes, it would be a mistake to give them a rationale for continuing by telling them they are addicted. Warnings regarding addiction may consequently deter people from quitting smoking once they have started, as the government has informed them that, once started, their smoking is beyond their control. These various difficulties may not all prove to be of great practical importance, though certainly some of them are likely to be pertinent. However, what they do 12 Indeed, one focus group participant believed that this warning fostered the impression that alcohol and drug use were a preferable alternative to tobacco use. See Transcript of Focus Group Studies, November 9, 1995, 6:00 p.m., San Francisco, CA, at 47-48 ("PARTICIPANT: These are really good except for the one about tobacco kills more each year then AIDS, alcohol, accidents, murders, suicides, illegal drugs, and fires, because then that's just endorsing safe sex, drinking and driving, killing people, jumping off a cliff, dealing drugs. MODERATOR DAVIS: Endorsing it in what way? PARTICIPANT: Well, it's saying, you know, well, hell, I'm going to go drink and, you know, snort some cocaine. MODERATOR DAVIS: Oh, I see what you're saying. So those are less risky than smoking. PARTICIPANT: Yeah. I mean, hell, my parents smoked cigarettes, but I'm going to be the one to say, you know, drink and drive or something. "). 13 See the Federal Trade Commission Staff Report on the Cigarette Advertising Investigation 4-15 to 4-16 (1981). -11-

0 • • • • • • highlight is the fact that designing and issuing hazard warnings is not a trivial and innocuous enterprise. Even well-intended hazard warnings can have adverse effects. The Hazard Warnings Vocabulary FDA's proposed warnings may also undermine the integrity of hazard warnings systems in general. The design of any warnings policy must ensure consistency with the generally accepted warnings vocabulary. FDA has, for example, been a leader in developing a highly structured hazard warnings vocabulary for patient package inserts. These warnings follow a standardized structure with respect to the provision of information regarding different kinds of properties of the drug, such as uses, contraindications, warnings, and other information. In addition, use of the human hazard signal words, such as "danger" and "warning, " as well as use of various mechanisms to alter the structure of the information presented, such as bold type and boxed warnings, also follow a rigid, standardized practice. The purpose of this standardization is to enable recipients of the information to make comparable judgments across products. If we use bold type and extremely prominent warnings for distant or contingent risks, then we have no available mechanism for communicating much more immediate perils. If we attempt to impress upon people the seriousness of distant or contingent risks by speaking about them in the terms ordinarily used to describe immediate perils, we necessarily undercut the • seriousness with which people will regard the immediate perils -- because people will come to think of them as being in the same class as the distant or contingent risks. The warnings policy favored by participants in the FDA-sponsored focus groups • fails to ensure consistency on a number of dimensions. First, the risk information . ~ provided (e.g., "Tobacco causes shortness of breath, coughing, yellow teeth, and p csc ~ ~ C..~ ~ -12- ~

• • • • • wrinkles. ") is without a source. Is this a message from the cigarette companies, the state legislature, Congress, the Surgeon General? From Mom and Dad? 14 Unattributed warnings are a departure from the current approach in which it is the Surgeon General providing the information, and the information provided has the weight of an official pronouncement. Is any change from that approach desirable? Does it devalue the Surgeon General's warnings? Is one to be credited but not the other? Does anyone seriously believe that adolescents will pay more attention to a statement that seems to come from a cigarette manufacturer than one that comes from the government, especially at a time when they are bombarded from all sides with the message that cigarette manufacturers are untrustworthy? FDA provides no scientifically rigorous evidence on how this aspect of the warning will affect its impact. Under FDA's proposal cigarette advertising would now have to include two warning messages. One warning message would be one of the standard rotating messages in which the Surgeon General is providing information. The second warning would presumably be a jagged-edged, boxed warning without any authorship. Given the juxtaposition of these two warnings, the potential for confusion is greater than in the normal warnings case in which no responsibility for the warning statement is assigned. Second, as was noted above, the contemplated hazard warnings may be in color and, based on the recommendations of the teenage respondents, possibly in neon. Color is sometimes used or required for warnings. However, use of color is restricted to very 14 This proposed warning -- an obvious effort to manipulate behavior by producing disgust or fear in the child -- is all too reminiscent of parental warnings regarding the perils of gumchewing and MTV. Such warnings are notoriously counterproductive. Indeed, some youths may regard the most dire warnings about the risks of smoking as a dare. During one of the focus group sessions, a participant raised this objection to the idea of using scare tactics along the lines of "you take a smoke, you'll die." See Transcript of Focus Group Studies, October 25, 1995, 6:30 p.m., Charlotte, NC, at 74 ("MODERATOR DAVIS: I mean, would that just -- maybe kids would read that and say, well, that's obviously not true so -- PARTICIPANT: Well, they might try just to see if it happens, you know. "). -13-

0 • • • unusual circumstances in which the risk is not only substantial but imminent. Bright orange is, for example, used in the standard industrial warning for TNT. However, this is a severe explosive risk that poses immediate catastrophic implications possibly to large numbers of individuals. Extremely potent carcinogens and other industrial chemicals do not receive the kind of neon treatment that is potentially envisioned for cigarettes. This aspect of the policy is clearly inconsistent with accepted warning practice across products. Such a reckless misuse of warning symbols and content could undermine the efficacy of warnings in other dimensions and for other activities for which we would like to promote sound risk-averting decisions.15 The jagged border around the warning is an unprecedented innovation and is subject to similar difficulties. All we know at this juncture is that the teenagers in the focus groups seem to like the jagged borders. Perhaps teenagers would be pleased to wear the design on tee-shirts.16 We do not really know, however, whether the presence of these jagged borders would have a statistically significant effect on either smoking risk perceptions or behavior. Moreover, using a jagged border here could dilute the impact of 15 The focus group studies' stated goal of identifying the "most compelling" warning message created an incentive for participants to favor extreme warning symbols and conventions. See Transcript of Focus Group Studies, November 9, 1995, 6:00 p.m., San Francisco, CA, at 74 ("MODERATOR DAVIS :... Would it matter if you had a different symbol? PARTICIPANT: Biohazard thing. MODERATOR DAVIS: Biohazard thing? PARTICIPANT: Yeah, like -- PARTICIPANT: You know, the triangle. PARTICIPANT: Yeah. PARTICIPANT: That would catch any kid's attention, really. "). 16 At least one participant in the focus group studies liked the jagged-edged border because it apparently reminded him of the artistic convention for depicting sound effects in comic books. See Transcript of Focus Group Studies, November 1, 1995, 6:30 p.m., Houston, TX, at 68 ("PARTICIPANT: Yeah, that's good. MODERATOR BRYANT: The jaggedy edges there? PARTICIPANT: Yeah. PARTICIPANT: Yeah, like Batman, you know. MODERATOR BRYANT: Like Batman? PARTICIPANT: Like, bang, you know. "). -14-

! ! • • ! • • merely boxing warnings for over-the-counter drugs, pesticides, and other hazardous products. The icon with an exclamation point surrounded by a triangle is subject to similar criticisms. There are two dimensions to the concerns with respect to being consistent with the hazard warnings vocabulary. First, at any point in time it is essential to maintain a consistent vocabulary so that individuals can distinguish relative riskiness across products or activities. Second, the teenagers of today who would be reading cigarette warnings in neon will eventually grow into the adults of the future, many of whom will work with hazardous chemicals on their jobs and almost all of whom will utilize potentially dangerous prescription drugs or household chemicals. If they were to become accustomed to extremely bold neon warnings of cigarette risks, then they would be likely to become passive with respect to many other classes of risk for which we would subsequently like to attract their attention. The prospective new cigarette warnings will devalue other important warnings both now and in the future. FDA should exercise restraint and forethought in developing new elements of a warnings vocabulary. In my past studies for the EPA, I have found that once warnings have reached a sufficient degree of readability, additional nuances such as large type and bolder boxing of the warning do not have a significant effect on the receipt and processing of the risk information. Although it may be that there is no constructive role that may be • played by the jagged edges, this format would disturb the accepted practices for presenting • warnings. Should, for example, all manufacturers of potentially hazardous products also add jagged edges to their warnings instead of simply boxing them in order to avert product liability suits? Will jagged edged boxes be used for the Physicians' Desk Reference listings? Once jagged edges become the norm, jagged edges will no longer be a ~ O distinctive feature so that the initial rationale for its distinctive character will be gone. Cit O ~ -15- ~ 00

a • r • • • • Conclusion FDA's proposal to require additional warnings on cigarette advertising is methodologically unsound, unlikely to prove effective and may produce unintended adverse consequences. The Agency developed its warnings policy without any regard for the usual scientific standards that are pertinent for work in the field and without considering the insights provided by the hazards warnings literature. Relying on experts and scientific evidence rather than conjectures by children to design warnings does not imply that the prospective recipients of the information should be excluded from the warnings design process. Indeed, my studies for EPA have drawn on hundreds of adults to assess the efficacy of different warnings. However, reputable academic journals simply would not treat as being scientifically valid a conclusion based solely on the results of focus group discussions. The Agency has not established the need for any additional warnings about the risks of smoking. There is simply no lack of information about the risks of smoking that an additional warning could remedy and therefore no reason to believe that the warning would reduce underage smoking. In light of this fact, it is all the more irresponsible of FDA to propose the warnings that it has. These warnings may have unintended adverse effects on behavior and may impair the integrity of the hazards warnings systems more 11 • • generally, all without generating any offsetting benefits in the form of reduced underage smoking. ~ O ~ O ~ -16- ~ ~ - ~ 0 ~

• • BACKGROUND APPENDIX Since the 1970s Professor W. Kip Viscusi has published academic studies on how people learn about risk and make subsequent choices under uncertainty. In part because of this research, he was asked in the early 1980s to settle the dispute between OSHA and OMB over the proposed OSHA hazard communication standard, which had been appealed to then Vice-President Bush. Both the New York Times and the Washington Post credited his analysis, which supported the regulation, with being the pivotal factor that led to the issuance of the regulation. l As a result of his involvement, he undertook a study of the effect of hazard warnings for workplace chemicals in four chemical plants, which was subsequently published in the 1984 American Economic Review.2 He has also been engaged in a series of studies for the U.S. Environmental Protection Agency to assist them in the design of hazard warnings for household chemicals and pesticides. These studies have been original scientific investigations involving experiments to assess the relative efficacy of different warning labels. The two studies based on this work, which he coauthored with Wesley A. Magat, were subsequently published as books by Harvard University Press and MIT Press.3 In addition to these two labeling studies, Professor Viscusi has also undertaken three controlled experimental field studies for the U.S. EPA on the effect of 1 For a detailed discussion of Professor Viscusi's involvement in the development of this regulation, see in particular the Washington Post Magazine June 9, 1985, pp. 11-13 and 36-41. 2 See W. Kip Viscusi and Charles O'Connor. "Adaptive Responses to Chemical Labeling: Are Workers Bayesian Decision Makers?" American Economic Review Vol. 74, No. 5 (Dec. 1984), pp. 942-956. 3 W. Kip Viscusi and Wesley Magat. Learning About Risk: Consumer and Worker Responses to Hazard Information Cambridge: Harvard University Press, 1987 and Wesley Magat and W. Kip Viscusi, Informational Approaches to Regulation, Regulation of Economic Activity Series No. 19, Cambridge: MIT Press, 1992.

• • communicating different kinds of risk information, particularly information dealing with ambiguous risks. He also served for EPA as a consultant/reviewer of their study of environmental tobacco smoke and continues to serve on two U.S. EPA Science Advisory Boards. Professor Viscusi has also published studies relating to cancer warnings on food and other aspects of hazard warning policy, served as the Associate Reporter for the American Law Institute Report on tort liability reform, for which he was a principal author of the hazard warning sections,4 and taught several continuing education courses to U.S. federal judges regarding the principles for judging the adequacy of effective warnings. His warnings work has also involved the design of hazard warnings for several nationally marketed consumer products, including the design of the current Dexatrim label. In connection with his work on alcoholic beverage warnings, Professor Viscusi was retained as one of the two principal consultants to Macro International -- the same consulting firm hired by FDA to prepare the study that served as the basis for the December 1, 1995 FDA notice. The Macro International report was subsequently issued by the U.S. Department of Health & Human Services as a critical overview of the warnings literature, where his work is discussed within the context of experimental approaches to warnings assessment.5 He has also testified before the U.S. Senate on alcoholic beverage warnings. 4 See Enterprise Responsibility for Personal Injury -- Reporters' Study, Vol. I: The Institutional Framework and Vol. II: Approaches to Legal And Institutional Change, Philadelphia: American Law Institute, 1991. 5 U.S. Department of Health and Human Services, Review of the Research Literature on the Effects of Health Warning Labels, Washington, D.C., June 1987. 2