Tobacco Documents Online

approximately equal numbers of males and females at each of six age levels (6-7, 8-9, 10-11, 12-13, 14-15, 16-17). • Based on their findings, the researchers report that a majority of children between the ages of 6 to-17 years have some understanding of sponsorship by late primary/secondary years. A third of the 10-11 year-olds and over half of the secondary school children sampled • were able to name cigarette brands and sponsored sports. Of the primary school children, 47% pointed to sponsored brands as being liked by someone who likes excitement and fast racing cars. The authors argue that "associations between brand names or visual cues and exciting sports can be elicited [by simple advertisements] in 0 children who did not have a critical awareness of the- purpose of • • ! • 0 commercial sponsorship" (p. 209). No significant differences between smokers and non-smokers in awareness of brand sponsorship were found. It should be made clear that suggesting that advertising factors might be associated is not equivalent to demonstrating that they are causally linked. Approval of ads is not necessarily related to the initiation or continuation of smoking behavior. Moreover, the authors argue that linkages/associations between brand names (or their visual cues) and exciting sports are often unconscious. This, however, was merely observed at the level of recognition or recall, suggesting a low involvement model. The researchers report that sponsorship is a form of promotion, treating this obvious fact as if it were some sort of discovery. Of course sponsorship is perceived to have economic benefit; if it did not, companies presumably would not bother to engage in the practice. The fact that children also understand this concept is an indicator of their cognitive/perceptual capacity; it is not an indicator of the persuasive capacity of sponsorship messages per se, nor is it proof that such messages either target children specifically or purposely promote consumption amongst children. The researchers want to argue - inappropriately and without sufficient evidence - that recognition indicates a host of other impacts, including imitation, persuasion, consumption, and the like. The authors cite 20 references for their research. Not one of these is an acceptable publication, based on the criteria established earlier o ~ • .~ © ~ - o~ ~ ~ o

• in this statement. The sources cited include the following: Ten to Fourteen-Year-Olds and Alcohol, British Journal of Social and • Clinical Psycholo&, Social Science and Medicine, Health Education Journal (3 citations), British Journal of Addiction, Journal of Epidemiology, Appendix H of the British Code of Advertising Practice, The Young Smoker, Cigarette Advertising & Smoking: A Rev_iew of Evidence, American Journal of Public Health, Teenag_ers 0 and Alcohol, a Department of Health and Social Services document entitled Smoking Among Secondary School Children: A Survey of Prevalence, Consumption, Attitudes, and Knowled~e, Community Health Studies, Health Education and the Media, Football Targe, ting Research: Summary Report, Smoking Attitudes and Behavior, • International Journal of Epidemiology, and Health Education Research. 0 • • • • (3) In short, the authors found that children are aware of advertisements and brands. Recognition and recall were assessed; these have nothing to do with advertising impact or the decision to initiate smoking behaviors. Chapman, S. & Fitzgerald, B. (1982). Brand Preference and Advertising Recall in Adolescent Smokers: Some Implications for Health Promotion. American Journal of Public Health, 72(5), 491- 494. As part of an anti-smoking school education program, the goal of this research (published in a relatively brief form of less than four pages) was to study brand preference and ad recall in 1,195 seventh and eighth grade adolescents in Australia. There were three hypotheses: (1) outstanding patterns of brand preference would be found in the sample, (2) smokers would be more familiar with cigarette ads than non-smokers, and (3) the ads most familiar to smokers would be for the brands most smoked. The authors point out that of 30 secondary schools randomly selected from a list of all schools in the Sydney area, only 19 agreed to run the educational program in 1980. They write: "any conclusions arising from the study may not warrant generalization to school children at large, but to children in schools with teachers interested in running such a program" (p. 491). ~ 0 ~ © -12- ~ W ~ • ' N+

Brand preference was cross tabulated with data from three questions that discriminated between light, medium and heavy smoking. • Subjects were presented with photographs of eight print media cigarette advertisements that had been edited to remove all identifying writing. Advertisements for five heavily advertised brands and three less frequently advertised brands were used to test identification. Slogans were edited and presented in different orders • to the edited photos to form slogan completion exercises. Four of more than 130 brands then retailed in Australia accounted for cigarettes smoked by 78.9% of the smokers. Brand preferences reflected those of adult smokers in Sydney found in another survey 0 taken three months later, but were different in proportion (chiefly due to a much greater proportion of children who smoked Winfield). Winfield smokers had a significantly greater proportion of heavy smokers. Smokers claiming no brand preference were more likely to be light, infrequent smokers. Smokers were nearly twice as likely to • correctly identify edited cigarette ads and slogans. Brand preference was considered to be an important descriptor of smoking profiles. • • • While the researchers concluded that tobacco ads are instructively seen as cues to smoking behavior, in fact what they tested was the ability of their subjects to perform recall tasks. They concluded that, since young smokers who smoke "any" brand tend more toward the light-medium range of smoking rates than do smokers who are brand loyal, part of the "career" of becoming a smoker is to adopt a preferred brand. The conclusion/belief that ads are cues to behavior is not supported by this research. Only five sources were cited by the researchers. Again, none of the references came from an acceptable publication, based on the criteria established earlier in this statement. The sources cited are: Media Council of Australia: Voluntary Advertising Code For Cigarettes in Australia, British Medical Journal, Consumer Beliefs and Behavior With Respect to Smoking: A Critical Analysis of the Public Literature (a report prepared for the FTC), Journal of Personality and Social Psychology, and International Journal of Health Education. • This study suggests a rather obvious point: Those who consume a product are more likely to be aware of advertising for that product. School children appear no different from adults in this regard. i~ 0 ~ 0 -13- ~ Ca . v~ ~ 0 i.~

However, it should be emphasized that while children who already smoke are attentive to cigarette advertising, it cannot be argued that • cigarette advertising is a causal factor in the initiation of smoking. (4) Klitzner, M., Gruenewald, P.J. & Bamberger, E. (1991). Cigarette Advertising and Adolescent Experimentation With Smoking. British Journal of Addiction, 86, 287-298. • • • • • • The authors state at the onset that their goal is to clarify the relationships between measures of the effects of advertising exposure derived from a cognitive advertising exposure model and youthful experimentation with cigarettes. The authors state two assumptions: (1) Experimentation with cigarettes would prompt subjects to attend to and retain information from cigarette ads (a selective exposure effect), and (2) Exposure to the ads themselves, mediated by the depth to which information from ads was processed, would prompt subjects to experiment with cigarettes (an effect of ad exposure). The sample included 295 adolescent subjects from three separate school districts in an east coast urban/suburban area. The grade strata included 5th and 6th, 7th-9th, and 10th-12th grades. Equal numbers males and females were sampled. The sample ethnic distribution included 25% black, 66% white, and 9% other (Hispanic/Asian). The subjects were from highly-educated families; 79% had a father and 60% had a mother with a college education. A total of 41% of the subjects had smoked at some point; 47% considered themselves to be smokers, and 27% smoked every day. Most began smoking in the 7th grade (29%). The 4% non-smokers believed they would begin smoking in 3-5 years. Advertisements from 37 general U.S. magazines, 2 major newspapers, and 5 youth readership magazines, covering a six-month time period from March to August 1987, were used. One issue per month of each magazine, and one issue per month of both newspapers, were used. The sampling universe consisted of 650 unduplicated ads (217 cigarette, 230 alcohol, 203 perfume). • Individual interviews in November-December 1987 were conducted by trained field staff. Four prompt cards listing possible choices of product and brand were used during the presentation of ad stimuli ~ O • . ~ ~ O -14- • ~ c,~

• • • • • • in each interview. The first listed the three generic products for a product recognition task. The remaining three prompt cards, one for each product, listed a choice of brands. Each card listed 20 possible choices. The 60 ads comprising each of the stimulus sets were presented sequentially in a manner that was counterbalanced, for order and stimulus set effects across subjects. Due to their considerable inattention to ads presented later in the task, 5th and 6th graders were presented with half of the 60 stimuli. The subjects were asked to identify magazine and newspaper ads for cigarettes, alcohol and perfumes from which product brand identification information was removed. Alcohol and perfume ads served as distractors. For each ad presented, subjects were asked: (1) whether they had seen the ad before, (2) which of the three generic products (cigarettes, alcoholic beverages, perfumes or colognes) they thought the ad was for, and (3) which brand of the identified product they thought the ad was for. Brand choice was requested whether or not product choice was correct. Demographic and smoking behavior data were also collected. Data were also collected on total exposure to magazines and newspapers (Exposure Opportunity measure). Total Magazine Exposure was assessed by subjects' responses to a checklist on which they were to indicate whether they had ever seen, looked at occasionally, or read regularly each magazine listed. Subjects were also asked how many other unlisted magazines they used. In general the research as designed was thoroughly unlike the actual experience of reading magazines and newspapers. The use of alcohol and perfume ads to serve as distractors, in particular, is also very different from the actual real experience of advertisement exposure in the practice of reading magazines and newspapers. ~ The researchers found that their subjects had substantial previous exposure to the ad stimuli. Using environmental and psychological measures of advertising exposure, they contend that they demonstrated that adolescents who experimented with cigarettes • were better able to recognize advertised products. Conversely, they say, those who could better recognize advertised brands were more likely to have experimented with cigarettes, an effect which they concluded was due to their exposure to cigarette advertising [emphasis is mine]. The authors contend that the size of the effect of ~ C ~ . ~ © -15- ~ . o~ ~ • ~

advertisements on smoking is mediated by depth of cognitive processing. For the ads that did affect brand recognition rates at the mean level of observed recognition in their study, they argue that a substantial effect may be observed. The authors, themselves, point out: "In any cross-sectional study it is impossible to directly test the causal relationships between any set of variables. The temporal information necessary to establish antecedent causes is not available." They add that they can, however, determine the consistency of variables within their causal model, and that in their study such consistency was achieved. The authors conclude that their results supported "both of the arguments put forth in the Introduction" (p. 296) While they use the word "arguments" here, they are actually talking about the unsupported assumptions advanced in their introduction. They contend that experimentation does affect the extent to which individuals recognize information presented in cigarette ads (selective exposure effect). With increased depths of processing, they say, it appears that exposure to smoking ads does affect the likelihood that young people will smoke. 'That is, young people who retain more information from cigarette ads than others, for whatever reason, appear more likely to express an interest in smoking through their experimentation with cigarettes. In particular, young people who retain more brand-related information from cigarette advertisements are most at risk for smoking experimentation" (p. 296). As the authors themselves point out, the results of this study must be considered within the context of several limitations: (1) Inferring causality from correlational data is risky; if there is causality it may run in the opposite direction. (2) The statistical instrument used to represent smoking in the 2SLS analyses was not optimal. (3) The measure of ad exposure was not contemporaneous with incidents of subjects' experimentation with smoking. Experimentation might have occurred years before participation in the study. Any ads used would not be those to which these subjects were exposed at the time of the onset of any smoking behaviors. The authors conclude that: 'The results show that the association between exposure to advertising and consumption is complex, at least for cigarette smoking among adolescents. Rather than reflecting either an effect of selective exposure of smokers to advertising, or an effect of advertising on smoking, the results show -16-

that both processes operate simultaneously, though at different depths of processing" (p. 296). What the authors are claiming here is reciprocal causality, when in fact this is merely yet another correlational study. Smoking leads smokers to pay attention to advertisements; smokers pay attention to advertisements because they smoke. The subjects who smoked could correctly identify tobacco advertisements. No arguments for causality are supported by this research. The researchers cite the following to support their study: British Journal of Addiction (2 citations), Journal of Advertising Research, Journal of Communication, an unpublished doctoral dissertation, The Lung_Goodbve: A Manual of Tactics for Counteractingthe Tobacco Industry in the 1980's, Journal of Verbal Learning and Behavior, New York State Journal of Medicine, Introduction to Structural Equation Models, 'an FTC Report to Congress, a chapter in The Behavioral Aspects of Smoking (a NIDA Research Monograph), Journal of Pediatrics, Women and Health, Consumer Demand in the United States, The Theory and Practice of Econometrics, Cognitive Psychology and Information Processin~, Journal of Marketing, Journal of Farm Economics, Cognition and Reality Principles and Implications of Cognitive Psychology, Marketing Models and Economic Research, Chronometric Explorations of Mind, On the Determination of Advertising Effectiveness: An Empirical Study of the German Cigarette Market, New York State Journal of Medicine, Journal of Business, Agricultural Economic Research,, The Economics of Advertisin~, Economics and Alcohol: Consumption and Controls, Tabak Journal International, New En~land Journal of Medicine, A Smoking Gun: How the Tobacco IndustrX Gets Away With Murder, Journal of Public Health Policy, Pediatric Annals, A Study of Ci agL rette Advertising,Among Teenage Girls and Young_,Women: A Summarv of _Findings. This is a substandard list of publications for this type of research in advertising. In general, the propositions advanced by the authors have not been convincingly argued or supported. The limitations conceded by the authors are veU significant. It should once again be emphasized that while children who already smoke may be attentive to cigarette advertising, it cannot be argued that cigarette advertising is a causal factor in the initiation of smoking. Moreover, these correlational results cannot be used to support causal inferences. Recognition of ads is due to exposure to ads; smoking initiation is not.

Conclusion • • • • • • • • The weaknesses regarding bias in problem statements, unwarranted generalizations, conclusions that are not justified by research results, and research designs that cannot answer questions that are asked are all evident in the research examples analyzed above. As noted, these problems permeate the research cited in the FDA proposal in support of the proposed regulations regarding tobacco advertising. In addition, the vast majority of this research is tainted by the value judgments of the researchers, who will apparently stop at nothing to advance their ideological assertions and insert them in the reports of their research results in ways that are completely unwarranted and without justification. Michelle A. Wolf % Date ~ 0 0 ~ -18- W

S References • • • • • • • • Aitkin, P.P., Leathar, D.S., O'Hagan, F.J. & Squair, S.I. (1987). Children's Awareness of Cigarette Advertisements and Brand Imagery. British Journal of Addiction. 82, 615-622. Archibald, C. (1993). Sale of Individual Cigarettes: A New Development. Letters to the Editor, Pediatrics, 91(4), 851. Beltramini, R. (1988). Perceived Believability of Warning Label Information Presented in Cigarette Advertising. Journal of Advertising, 32, 26-32 Chapman, S. & Fitzgerald, U. (1982). Brand Preference and Advertising Recall In Adolescent Smokers: Some implications for Health Promotion. American Journal of Public Health, 72(5), 491-493. Charlton, A. (1986). Children's Advertisement-Awareness Related to Their Views On Smoking. Health Education Journal, 45(2),75-78. Crawford, C.T., Gramm, W.L., Weininger, B., Allen, R., Snyder, W.S., Rubin, P., Higgins, R S., Levine, B. C., Damato, D. G., Keenan, D. & Popper, E. T, (1985). Omnibus Petition for Regulation of Unfair and Deceptive Alcoholic Bevera e Advertising and Marketing Practices: Docket No. 209-46 (Memorandum). Recommendation of the Staff of the Federal Trade Commission. Washington, DC: U.S. Federal Trade Commission. Unpublished Paper. , Cummings, K.M., Pechacek, T. & Shopland, D. (1994). The Illegal Sale of Cigarettes to U.S. Minors: Estimates by State. American Journal of Public Health. 84(2), 300-302. DiFranza, J.R., & Tye, J.B. (1990). Who Profits From Tobacco Sales to Children? Journal of the American Medical Association. 263(20), 2784-2787. DiFranza, J.R., Richards, J.W., Paulman, P.M., Wolf-Gillespie, N., Fletcher, C., Jaffe, RD., and Murray, D. (1991). RJR Nabisco's Cartoon Camel Promotes Camel Cigarettes to Children, Journal of the American Medical Association. 266(22), 3149-3153. ~ O ~ O -19- ~ W ~ • po

0 • • • • • • • • Doxiadis, S.A., Trihopoulos, D.V. & Phylactou, H.D. (1985). Impact of a Nationwide Anti-Smoking Campaign. The Lancet pp. 712-713. Ecenbarger, W. (1993). America's New Merchants of Death. Reader's Digest, 142(852), 50-57. Feigherty, E., Altman, D.G. & Shaffer, G. (1991). The Effects of Combing Education and Enforcement to Reduce Tobacco Sales to Minors. Journal of the American Medical Association, 266(22), 3168-3171. Fischer, D.A. & Magnus, P. (1981). "Out of the Mouths of Babes ...": The Opinions of 10 and 11 Year Old Children Regarding the Advertising of Cigarettes.- Community Health Studies, 5(1), 22-26. [Australian study] Fischer, P.M., Richards Jr., J.W., Berman, E.J. & Krugman, D.M. (1989). Recall and Eye Tracking Study of Adolescents Viewing of Tobacco Advertisements. Journal of the American Medical Association, 261(1), 84-89. Fischer, P.M., Schwartz, M.P., Richards, J.W., Goldstein, A.O. & Rojas, T.H. (1991). Brand Logo Recognition by Children Aged 3 to 6 Years. Mickey Mouse and Old Joe Camel. Journal of the American Medical Association, 266(22), 3145-3148. Flay, B.R. (1987). Mass Media and Smoking Cessation: A Critical Review. American Journal of Public Health, 77, 153-160. Flynn, B.S., Worden, J.K., Secker-Walker, R.H., Badger, G.J., Geller, B.M., Costanza, M.C. (1992). Prevention of Cigarette Smoking Through Mass Media Intervention and School Programs. American Journal of Public Health 82, 827- 834. Flynn, B.S., et al. (1994). Mass Media and School Interventions for Cigarette Smoking Prevention: Effects Two Years after Completion. American Journal of Public Health. 84, 1148-115 0. Freedman, A. (1994, October 26). How A Tobacco Giant Doctors Snuff Brands to Boost Their 'Kick.' [sic] Wall Street Journal, p. A14, col. 5. Goldstein, A.O., Fischer, P.M., Richards, J.W., Jr. & Creten, D. (1986). Relationship Between High School Student Smoking and Recognition of Cigarette Advertisements. The Journal of Pediatric Medicine, 110(3), 488-491. ~ O ~ O -20- ~ ~ • ~,p