Tobacco Documents Online

PROFESSOR MICHELLE A. WOLF • f PUBLIC POLICY DECISIONS SHOULD BE BASED ON SOUND SOCIAL SCIENCE RESEARCH, NOT SPECULATION OR POLITICAL MOTIVATION. The research regarding tobacco advertising and youth smoking initiation and behaviors relied on by the FDA is fraught with flaws in research design, clarity of results, and the conclusions that are reached. The points I am advancing here can be generalized to the vast majority of the research that is cited in the FDA Federal Register document as support for its argument against the tobacco industry's advertising practices and the conclusions regarding initiation and practice of smoking behaviors. 1. Many of the sources used as references in the Federal Register are not acceptable examples of social science research. Advertising and media effects research should be conducted by advertising and communication scholars and should be published iii respectable, peer-reviewed journals in those fields. a. One of the problems with the literature cited in the FDA document is that a substantial number of studies were not conducted and published by researchers with appropriate backgrounds and qualifications. While the absence of appropriate professional qualifications does not automatically render the studies invalid per se, it does serve to alert the reviewer to more critically evaluate the rigor of the methodology and the conclusions based thereon. William Krumske Jr. addresses the investigator biases of much of this research. He points out that these investigators begin their research with an"abhorrence of tobacco smoking and their collective desire to protect, at any cost and by all means, all people, especially the young, from tobacco smoking" (p. 2). Krumske states that another investigator bias in this research concerns specific tobacco industry marketing objectives. He contends that the mission of such research is "to demonstrate that unintended audiences are exposed to advertising which, to some degree, influences them" (p. 2). Examples of researchers without appropriate advertising/consumer behavior qualifications follow. (1) Fischer, Paul M., Schwartz, Meyer P., Richards, John W. Jr., Goldstein, Adam O. & Rojas, Tina H. (1991). Brand Logo Recognition by Children Aged 3 to 6 Years. Mickey Mouse and Old Joe the Camel. Journal of the American Medical Association, 266(22), 3145-3148. FISCHER, SCHWARTZ, RICHARDS, AND GOLDSTEIN ALL IDENTIFY THEMSELVES AS MEDICAL DOCTORS.

• (2) Pierce, J.P., Gilpin E., Burns, D.M., Whalen, E., Rosbrook, B. Shopland, D., & Johnson, M. (1991). Does Tobacco Advertising Target Young People to Start Smoking? Evidence From California. Journal of American Medical Association. 266(22), 3154-3158. JOHN P. PIERCE IS A PH.D. DAVID M. BURNS IS A MEDICAL DOCTOR. ELIZABETH WHALEN HAS AN MA AND BRADLEY ROSBROOK HAS AN MS. DONALD SHOPLAND WORKS AT THE NATIONAL CANCER INSTTTVI'E IN MARYLAND AND MICHAEL JOHNSON IS A MEDICAL DOCTOR. Goldstein, A.O., Fischer, P.M. Richards, J.W., Jr. & Creten, D. (1987). Relationship Between High School Student Smoking and Recognition of Cigarette Advertisements. The Journal of Pediatrics, 110(3), 488-491. ADAM O. GOLDSTEIN HAS A BA, PAUL M. FISCHER AND JOHN W. RICHARDS ARE MEDICAL DOCTORS, AND DEBORAH CRETEN HAS ONLY A BA. • • (4) 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), 491-494. SIMON CHAPMAN IS THE COORDINATOR OF THE ANTI- SMOKING PROJECT GROUP IN NEw SOUTH WALES AND A LECTiTRER IN THE DEPARTMENT OF PREVENTATIVE AND SOCIAL MEDICINE AT THE UNIVERSITY OF SYDNEY, AUSTRALIA AND BILL FITZGERALD IS A STATISTICIAN AT THE UNIVERSITY OF SYDNEY. Huang, P.P., Burton, D., Howe, H.L. & Sosin, D.M. (1992). Black-White Differences in Appeal of Cigarette Advertisements Among Adolescents. Tobacco Control, 1, 249- 255. • PHILIP P. HUANG AND DANIEL M. SOSIN WORK FOR THE DIVISION OF FIELD EPIDEMIOLOGY AT THE CENTER FOR DISEASE CONTROL IN ATLANTA GEORGIA. DEE BURTON ~ WORKS FOR THE PREVENTION RESEARCH CENTER AT THE ~ CTc O -2- W ~ O . ~

• • • • • • • • • UNIVERSITY OF ILLINOIS, SCHOOL OF PUBLIC HEALTH, CHICAGO, ILLINOIS. H.L, HowE IS WITH THE ILLINOIS DEPARTMENT OF PUBLIC HEALTH IN SPRINGFIELD, ILLINOIS. Even a casual glance raises the question of why medical doctors and public health officials are publishing studies about advertising. It is especially curious that almost all of these authors onIV publish articles about a single kind of advertising -- that for cigarettes. b. Another problem with the literature cited in the FDA proposal is that a substantial number of the studies are published in journals that focus on medical -- not youth and/or tobacco advertising -- research. Moreover, a great number of these studies have nothing to do with the relationship between tobacco advertising and the initiation of youth smoking behaviors. (1) Tye, J.B., Warner, K.E. & Glantz, S.A. (1987). Tobacco Advertising and Consumption: Evidence of a Causal Relationship. Journal of Public Health Policy, 8(4), 492-508. (2) 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. (3) 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. (4) Charlton, A. (1986). Children's Advertisement-Awareness Related to Their Views On Smoking. Health Education Journal, 45(2), 75-78. (5) DiFranza, J.R., Richards, J.W., Paulman, P.M., Wolf= Gillespie, N., Fletcher, C., Jaffe, R.D., and Murrary, D. (1991). RJR Nabisco's Cartoon Camel Promotes Camel Cigarettes to Children, Journal of the American Medical Association, 266(22), 3168-3173. 0 • C0 ~ 0 -3- ~ C.v o • ~

0 • • • • • • • • (6) 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] (7) 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), 8489. (8) 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. (9) 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 Pediatrics, 110(3), 488-491. (10) Huang, P.P. Burton, D., Howe, H.L. & Sosin, D.M. (1992). Black-White Differences in Appeal of Cigarette Advertisements Among Adolescents. Tobacco Control, 1, 249- 255. (11) Klitzner, M., Gruenewald, P.J. & Bamberger, E. (1991). Cigarette Advertising and Adolescent Experimentation With Smoking. British Journal of Addiction, 86, 287-298. (12) Pierce, J.P., Gilpin, E., Burns, D.M., Whalen, E., Rosbrook, B., Shopland, D., Johnson M. (1991). Does Tobacco Advertising Target Young People to Start Smoking? Evidence From California, Journal of the American Medical Association, 266(22), 3154-3158. (13) Tye, J.B., Warner, K.E. & Glantz, S.A. (1987). Tobacco Advertising and Consumption: Evidence of a Causal Relationship. Journal of Public Health Policy, 8(4), 492-508. ~ 0 • ~ ~ 0 -4- ~ W 0 • G~

• • • • • • • (14) Sarason, I.G., Mankowski, E.S., Peterson, Jr., A.V. & Dinh, K. T. (1992). Adolescents' Reasons for Smoldng. Journal of School Health, 62(5), 185-190. (15) DiFranza, J.R. & Tye, J.B. (1990). Who Profits From Tobacco Sales to Children? Journal of the American Medical Association, 263(20), 2784-2787. (16) Taioli, E. & Wynder, E.L. (1991). Effect of the Age at Which Smoking Begins on Frequency of Smoking in Adulthood. The New England Journal of Medicine, 325(13), 968-969. 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. (18) Feigherty, E., Altman, D.G. & Shaffer, G. (1991). The Effects of Combining Education and Enforcement to Reduce Tobacco Sales to Minors. Journal of the American Medical Association, 266(22), 3168-3171. (19) Klonoff, E.A., et al. (1994). The Problem and Sociocultural Context of Single-Cigarette Sales. Journal of the American Medical Association, 271(8), 618-620. (20) Popham, W.J., et al. (1993). Do Anti-Smoking Media Campaigns Help Smokers to Quit? Public Health Reports, 108, 510-513. (21) Popham, W.J., et al. (1994). Effectiveness of the California 1990-1991 Tobacco Education Media Campaign. American ~ Journal of Preventative Medicine. [in press] (22) Flay, B.R. (1987). Mass Media and Smoking Cessation: A Critical Review. American Journal of Public Health, 77, 153- 160. (23) 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 1148-1150. 0 0 -5- ~ o 1 +~

(24) Flynn, B.S., Worden, J.K., Seeker-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. Although the mere act of publishing advertising research in medical/health related journals does not, in and of itself, completely • discredit the research, it is a signal to the reader to be alert to evaluate the methodology, conclusions and, perhaps, the motivation for the research. c. A third problem with the literature cited in the FDA proposal is that • some of this literature merely consists of critical essays and reviews -- not reports of social science research. 0 • • • • (1) Washin on Post (1995, January 9). p. A5, col. 3. (2) Freedman, A. (1994, October 26). How a Tobacco Giant Doctors Snuff Brands to Boost Their 'Kick.' [sic] Wall Street Journal, p. A14, col. 5. (3) Wall Street Journal (1990, May 3). p. Bl, col. 3. .(4) Wall Street Journal (1989, October 19). p. B i. (5) Wall Street Journal (1994, December 27). p. B5. (6) The New York Times (1984, January 13). p. D4, col. 5. (7) Ecenbarger, W. (1993). America's New Merchants of Death. Reader's Digest,142(852), 50-57. ~ (8) Mintz, M. (1991). Marketing Tobacco to Children, The Nation, 252 (17), 577. (9) Archibald, C. (1993). Sale of Individual Cigarettes: A New Development. Letters to the Editor, Pediatrics, 91(4), 851. (10) Doxiadis, S.A., Trihopoulos, D.V. & Phylactou, H.D. (1985). Impact of a Nationwide Anti-Smoking Campaign. The Lancet, pp. 712-713. vc ~ 0 -6- v~ w v~

d. A fourth problem with the literature cited in the document concerns the apparent reliance on a number of government documents. These documents do not constitute primary social science research. Moreover, many of these documents do not even focus on areas of consideration and analysis that are relevant to issues surrounding relationships between tobacco advertising and youth consumption of tobacco products. In general, government documents are not regarded as original contributions to research. Rather than examine the original research scientifically, these reports often make mere conclusory references, many of which are based on third-party accounts of the original studies. Often, the findings of original research are camouflaged. For example, frequent research references are made to the 1994 Surgeon General's Report. In this document, research findings are mischaracterized and important points are omitted from research results and conclusions drawn because they do not support the points that are being advanced. In fact, of the 108 studies cited in Subpart D of the Federal Register document, only a single study -- the research of Beltramini (1988) -- meets the necessary criteria established above in that it: (1) addresses the appropriate subject matter (specific cigarette advertising related issues and youth smoking behaviors), (2) is published by an advertising or communication scholar in (3) a respectable, peer-reviewed journal in the field of advertising or communication. 2. Several additional problems emerge from a careful study of the research relied on by the FDA in the Federal Register. a. Problem statements are biased. b. Unwarranted generalizations are advanced. c. Conclusions drawn are not justified by results or findings. d. Researchers do not - and cannot - answer the questions that they ask. The weaknesses noted above can be found throughout the research cited in the FDA Register in support of the proposed regulations regarding tobacco advertising. To illustrate these problems, I have selected four examples of cited studies which illustrate these problems. In the following pages, the 7

• • • • • • • • • research objectives, methods, results, and conclusions (which are often, quite different from the results) are summarized and evaluated. (1) 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 the Camel. Journal of the American Medical Association, 266(22), 3145-3148. The objective here was to measure product logo recognition by subjects aged 3 to 6 years. It is important to recognize at the onset that logo recognition has nothing to do with intent to smoke or smoking behaviors. The authors used a convenience sample of 229 children attending 10 preschools in Augusta and Atlanta, Georgia; it was balanced for race and Socio Economic Status (SES), and included ages 3 to 6 years. Notably, this sample is not generalizable. Subjects matched 22 logo cards to one of 12 products pictured on a game board. Each of 12 products was named, and there was a demonstration of matching with a sample logo card. Subject responses were graded as correct or incorrect. The authors found a high recognition rate ranging from 91.7% for The Disney Channel to 25.3% for Cheerios. Joe Camel had the highest (over 50%) recognition rate among cigarette logos. Other cigarette logos were correctly recognized at a rate of 18.0% to 32.8%. Logos for adult products were recognized by 16.2% to 54.1% of the sample, with automobile brand logos having the highest rate. Three-year-olds recognized Camel at a rate of 11.1% and Joe Camel at 30.4%. Six- year-olds recognized Camel at a rate of 43.5% and Joe Camel at 91.3%. Logo recognition was found to be highly associated with age. Race and gender were not associated with the scores. It is not at all surprising that logo recognition is highly associated with age. Children recognize brand logos for many different products. However, recognition and recall are not the causes of consumer behavior; consumer decision factors are. The influence of peers and family, along with such factors as past experience, lifestyle, social class, and culture, have been shown to be very important influences. The study has several limitations. First, the authors used a convenience sample. It is unknown how the sampling method might ~ o ~ O - 8 - • ~

• • • have influenced results. Other problems include the reliance on the ability of children to match the logo cards and the arbitrariness of the selection of specific logos. In fact, the authors themselves conclude: "It is obviously impossible to predict how the exposure of children to environmental tobacco advertising might influence their later smoking behavior. .." Nevertheless, the authors go on to conclude "cigarette advertising may be an important health risk for children" (p. 3148). Despite acknowledging that this research does not establish any connection between logo recognition and smoking, the authors wrongly conclude cigarette advertising "may" be a risk. In fact, my own very recent research provides evidence that Joe Camel, while well recognized, symbolizes a product that children say is for adults and is bad for you (see Henke & Wolf, 1995). Beyond this, the researchers 'fail to consider other factors that affect "recognition." The Disney logo looks like a mouse, and matching it to a picture of Mickey Mouse might not demonstrate product recognition any more than matching any similar shapes, colors or objects would. The recognizability of Joe Camel might well be a routine outcome of any high visibility campaign. Most important is the fact that brand (character) recognizability has not been shown to determine smoking behavior. The researchers apparently assume, without benefit of supporting data, that brand awareness in childhood not only influences adult brand preferences but, indeed, causes consumption. That proposition is hardly tested here. The conclusion on p. 3148 that "RJ Reynolds ... is as effective as The Disney Channel in reaching 6-year-old children" is misleading. The conclusion is severely limited by, among other things, the researcher's narrow operationalization of "effectiveness," which is measured as the ability to match product logos with generic graphics (i.e., "effective" simply equals "recognition"). Measuring that ability is a far cry from measuring more complex cognitive and behavioral impacts. Finally, a total of 22 references are cited to support the foundation for this research. T'ne publications listed (in the exact citation form used by the authors) include: J Law Econ, Rev Econ Stat, J Peditar Am J Public Health, Br J Addict (2 citations), Health Promotion, Med J, N Eng J Med (2 citations), MMWR, Surgeon General, 9

Reducin • • • • • • • the Health Consequences of Smokina: 25 Years of 0 Progress: A Report of The Surgeon General, J Public Health Policy, J Advertising Res, Curr Issues Res Advertisin~, Nonparametric Statistics for the Behavioral Sciences, Children as Consumers J Consumer Res (2 citations), Health Educ J, and World Smoking Health. Given the fact that the research concerns brand logo recognition and advertising, the overwhelming citation of research from medical sources is irresponsible. In short, the researchers' conclusion that cigarette advertising is a pediatric health risk is not substantively or rigorously supported by their data. (2) Aitkin, P.P., Leathar, D.S. & Squair, S.I. (1986). Children's Awareness of Cigarette Brand Sponsorship of Sports and Games in the UK. Health Education Research, 1(3), 203-211. The stated objective here was to explore children's awareness of cigarette brand sponsorship of sporting events in the UK. The authors begin by assuming, as stated in the introduction, that "under- age smokers are deriving some benefit from cigarette advertisements" (p. 203). This is based on findings (here the authors refer to research that is also problematic) that "under-age smokers tend to have more positive attitudes towards cigarette advertising than do non-smokers of the same age" (p. 203) and that "under-age smokers tend to rate cigarette advertisements more highly on rating scales describing attractive qualities (e.g., exciting eye-catching) than do non-smokers of the same age"(p. 203). It is important to point out that neither the notion that youth can identify brand imagery nor the assumption that they (presumably) have more positive attitudes towards cigarette ads is indicative of any "benefit" in the sense of showing that ads contribute to decisions to initiate smoking behaviors. In fact, on page 204 the authors themselves write: 'This of course does not necessarily mean that cigarette advertising plays a part in inducing children to start smoking." Nevertheless, they go on to contend that it "suggests that attitudes towards cigarette advertising . may play a part in the transition from non-smoker to smoker." This contention is not supported by the research. The sample included 726 primary and secondary school children between the ages of 6-17 years in Glasgow, Scotland. There were ~ O ~ 0 ~ O -10- W 0 • =~

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

0 • • 0 0 Henke, L.L, & Wolf, M.A. (1995). Parent Intervention to Prevent Smoking: Using Joe Camel to Teach Children About Smoking. Paper to be presented at the 1996 American Academy of Advertising Conference, Vancouver, B.C., March/April. Huang, P.P., Burton, D., Howe, H.L. & Sosin, D.M. (1992). Black-White Differences in Appeal of Cigarette Advertisements Among Adolescents. Tobacco Control, 1, 249-255. Klitzner, M, Gruenewald, P.J. & Bamberger, E. (1991). Cigarette Advertising And Adolescent Experimentation With Smoking. British Journal of Addiction. 86 287-298. Klonoff, E.A. et al. (1994). The Problem and Sociocultural Context of Single- Cigarette Sales. Journal of the American Medical Association. 271(8), 618-620. Krumske, Jr., W. When Medical Doctors Conduct Marketing, Research: JAMA and Old Joe, the Camel. (Unpublished paper). Mintz, M. (1991). Marketing Tobacco to Children, The Nation, 252(17), 577. Pierce, J.P. (PhD), Gilpen D., Burns, D.M. (MD), Whalen, E., Rosebrook, B. (MS), Shopland, D. & Johnson, M. (1991). Does Tobacco Advertising Target Young People to Start Smoking? Evidence From California. Journal of the American Medical Association, 266(22), 3154-3158. Popham, W.J., et al. (1993). Do Anti-Smoking Media Campaigns Help Smokers to Quit? Public Health Reports, 108, 510-513. Popham, W.J., et al. (1994). Effectiveness of the California 1990-1991 Tobacco Education Media Campaign. American Journal of Preventative Medicine. [in press] Sarason, I.G., Mankowski, E.S., Peterson, Jr., A.V. & Dinh K.T. (1992). Adolescents' Reasons for Smoking. Journal of School_ Health, 62(5), 185-190. Taioli, E. & Wynder, E.L. (1991). Effect of the Age at Which Smoking Begins on Frequency of Smoking in Adulthood. The New England Journal of Medicine. 325(13), 968-969. The New York Times (1984, January 13). p.D4, col.5. -21 -

• • • • 0 • • • Tobacco Advertising and Promotional Activities. (1994). Preventing Use Among Youna People: A Report of the SurQeon General. U.S. Department of Health and Human Services, Centers for Disease Control and Prevention, Office on Smoking and Health (Chapter 5). Tye, J.B., Warner, K.E. & Glantz, S.A. (1987). Tobacco Advertising and Consumption: Evidence of a Causal Relationship. Journal of Public Health Policy, 8(4), 492-508. Wall Street Journal (1989, October 19). p. Bl. Wall Street Journal (1990, May 3). p. Bl, col. 3. Wall Street Journal (1994, December 27). p. B5. Washington Post (1995, January 9). p. A5, col. 3. Washington Post (1995, January 9). p. A5, col. 3. i\Z - _ _- - - - -O ~ c.rt -22- ~ i~