CONTEXT: The 1998 Master Settlement Agreement (MSA) restricted tobacco industry advertising practices that targeted teens.
OBJECTIVE: To assess whether cigarette-advertising campaigns conducted after the MSA continue to influence smoking among adolescents.
DESIGN, SETTING, AND PARTICIPANTS: Participants were a national longitudinal cohort of 1036 adolescents (baseline age: 10–13 years) enrolled in a parenting study. Between 2003 and 2008, 5 sequential telephone interviews were conducted, including the participant's report of brand of “favorite” cigarette advertisement. The fifth interview was conducted after the start of RJ Reynolds' innovative “Camel No. 9” advertising campaign in 2007. Smoking outcome reported from the fifth survey.
RESULTS: The response rate through the fifth survey was 71.8%. Teenagers who reported any favorite cigarette ad at baseline (mean age: 11.7 years) were 50% more likely to have smoked by the fifth interview (adjusted odds ratio: 1.5 [95% confidence interval: 1.0–2.3]). For boys, the proportion with a favorite ad was stable across all 5 surveys, as it was for girls across the first 4 surveys. However, after the start of the Camel No. 9 advertising campaign, the proportion of girls who reported a favorite ad increased by 10 percentage points, to 44%. The Camel brand accounted almost entirely for this increase, and the proportion of each gender that nominated the Marlboro brand remained relatively stable.
CONCLUSIONS: After the MSA, adolescents continued to be responsive to cigarette advertising, and those who were responsive were more likely to start smoking. Recent RJ Reynolds advertising may be effectively targeting adolescent girls.
WHAT'S KNOWN ON THIS SUBJECT:
A body of research in the 1990s demonstrated that innovative marketing campaigns for cigarette brands had effectively encouraged adolescents to start smoking. The public health response included the 1998 MSA, which restricted tobacco marketing to underaged youth.
WHAT THIS STUDY ADDS:
This national cohort entered adolescence years after the provisions of the MSA went into effect. However, the 2007 Camel No. 9 campaign was effective in encouraging young girls to start smoking.
In 1998, the Master Settlement Agreement (MSA) was enacted between states' attorneys general and the tobacco industry.1 After convincing evidence that tobacco marketing encouraged adolescent smoking,2,3 the agreement restricted several tobacco industry advertising practices,4 and tobacco manufacturers agreed not to “take any action, directly or indirectly, to target youth.”5 The MSA also funded a national youth education campaign aimed at reducing adolescent tobacco use.4 The Monitoring the Future study6 has reported that 30-day smoking in US high school seniors declined from 35.1% in 1998 to 24.4% in 2003 (by 2.1% per year) and then declined further to 20.4% in 2008 (by 0.8% per year). The slowing rate of decline7 has raised concern about weakened tobacco industry compliance with the MSA marketing restrictions.8
An important early strategy in product marketing is to encourage nonusers to try a new product; regular use may require additional strategies.9 Mass media image-advertising has been used to effectively increase experimentation, as evidenced by both the Virginia Slims10 and Joe Camel11 cigarette campaigns. Product placement in movies has also been demonstrated to be an effective marketing strategy.12
Such tobacco industry marketing campaigns are persuasive communications.13 As such, well-established theory has noted a hierarchy of effects: target audiences need to be exposed to the message; however, not all those exposed will attend to the message. Those who do attend and develop an emotive response such as liking the message will be the most likely to experiment.14,15 In accordance with this theory, nomination of any favorite cigarette advertisement by never-smokers predicted future experimentation in a sample of young Californian never-smokers.16 In confirmatory follow-up studies of 2 separate California cohorts, never-smokers who were receptive to tobacco advertising at the ages of 12 to 15 years were more than 50% more likely to be adult smokers 6 years later.17 It should be noted that enactment of the MSA was associated with a considerable decline in the proportion of adolescents who nominated a favorite cigarette advertisement.18
In the United States, cigarette-smoking initiation occurs between the ages of 12 and 24 years19,20 and is preceded by a period when the never-smoker is cognitively susceptible to experiment.21,–,23 Smoking experimentation is known to vary with race/ethnicity, parental educational level, family composition, and state of residence. Most studies have shown that adolescents who are exposed to other smokers (friends, siblings, parents) are at higher risk of experimentation, and poor school performance is a marker for increased risk.24
In 2007, the market share for RJ Reynolds' Camel cigarettes increased sharply after years of slower growth.25 Industry analysts attributed the increase to the innovative fashion-themed Camel No. 9 marketing campaign aimed at young women that was launched in February 2007.26 Later in 2007, attorneys general from California and Ohio initiated investigations, and a formal complaint was filed with the Federal Trade Commission alleging that the Camel No. 9 campaign violated MSA provisions on marketing to children.27
We investigated whether nominating a favorite cigarette advertisement remains a significant predictor of future experimentation in the post-MSA era by using a longitudinal study of adolescents (from 2003 to 2008). We hypothesized that RJ Reynolds' Camel No. 9 marketing campaign in 2007 coincided with increased nomination of a favorite cigarette advertisement among adolescent girls but not boys. We also hypothesized that the proportion of girls, but not boys, who specifically nominated Camel as their favorite ad has increased and that proportions remained relatively stable for those nominating Marlboro, a brand with broad appeal and strong market share that did not launch controversial ad campaigns during the study period.
Participants were 10 to 13 years old when enrolled onto a trial on parenting practices. In 2003, a random-digit-dial telephone screening survey identified a nationally representative sample of households, oversampled for black participants. Of household members who responded to the screening survey, 3079 met the eligibility requirements of a family with an oldest child aged 10 to 13 years. Over the next 14 months (distributed not to overload the study's intervention team), eligible families were invited to volunteer for a parenting study and a telephone contact was attempted to complete parent and teenager baseline surveys. We enrolled 1036 families with an estimated efficiency of 36% of families enrolled among eligible families. A detailed protocol for the parenting study has been published and included baseline characteristics of the sample.28 The intervention focused on helping parents achieve 3 parenting goals: build positive teen behaviors, set effective limits, and maintain a responsive relationship with their teenager.
We compared our sample to national population estimates for households with an oldest child aged 15 years from the 2001/2002 Current Population Survey.29 As reported,28 we slightly overenrolled black participants (18% of the sample; 16% of the US population) and 2-parent households (75% of the sample; 67% of the US population), the latter to account for teenagers living with a parent or guardian. Parental education was similar to national averages. Twenty-five percent of the sample and 23% of the US population resided in the Midwest, 40% of the sample (39% US population) in the South, 16% of the sample (18% US population) in the Northeast, and 19% (22% US population) in the West. To minimize selection bias, the baseline sample was weighted to be representative of the US population using these national population estimates, after initial base weights accounted for household selection probabilities. Weighting variables included gender, race/ethnicity (non-Hispanic white, non-Hispanic black, and Hispanic/other), household composition (2-parent versus single-parent), and parental education (some college versus no college attendance). Survey weights were constructed by using a standard raking procedure.30 Each adolescent follow-up survey was similarly weighted to these totals to adjust for loss to follow-up.
We interviewed the target adolescent 5 times at ∼8-month intervals after receiving parental permission on each occasion. Baseline interviews (n = 1036) were completed from September 2003 to October 2004 when the average respondent age was 12 years (median interview month: February 2004). Subsequent surveys were conducted as follows: survey 2, April 2004 to Oct 2005 (median interview month: January 2005); survey 3, May 2005 to July 2006 (median interview month: January 2006); survey 4, April 2006 to April 2007 (median interview month: July 2006); and survey 5, July 2007 to September 2008 (median interview month: February 2008). Seventy-two percent of the original sample (n = 744) was retained through survey 5. Adolescents born in 1990 (ie, 13 years old at baseline) were less likely to be retained for all surveys (64% retention rate), as were teenagers from nonwhite families (63%), from a single-parent home (59%), or with a parent who was a smoker (63%).
Adolescents who had reached 18 years of age by survey 5 (n = 27) were excluded from the analyses because they were no longer underage for purchasing cigarettes. Among 937 retained baseline never-smokers, 681 were followed across all 5 surveys (retention rate: 73%). The average age of respondents at survey 5 was 15.7 years.
Tobacco-use items included, “Have you ever smoked a cigarette?” and, if not, “Have you ever tried or experimented with cigarette smoking, even a few puffs?” A positive response to either question classified the person as “having smoked.”16,23 The remaining adolescents were classified as never-smokers. Susceptibility to smoking among never-smokers was assessed as in previous research17,21 with the questions, “Do you think that in the future you might experiment with cigarettes?”; “At any time during the next year do you think you will smoke a cigarette?”; and “If one of your best friends were to offer you a cigarette, would you smoke it?” Any response other than “Definitely not” to any question classified the adolescent as a susceptible never-smoker.
Receptivity to tobacco advertising was measured as in our previous research13,16,17 and focused on whether the teenager had a favorite cigarette advertisement by asking, “Think back to the cigarette advertisements you have recently seen. What is the name of the cigarette brand of your favorite cigarette advertisement?” Respondents who did not name a brand were also asked, “Of all the cigarette advertisements you have seen, which do you think attracts your attention the most?” We did not record when interviewers used this probe. Those who responded “Don't know” or said they had no favorite ad were grouped as “none,” and for all others, the name of the favorite brand was recorded.
Teenagers were asked to rank their performance in school as “much better than average,” “better than average,” “average,” or “below average.” This measure has consistently predicted later smoking behavior in our studies.13,16,17,21 Smoking in the respondent's social network was assessed by the following 2 questions: “Have any people that you live with now smoked cigarettes in the last year?” and “How many of your best friends smoke?” Single-parent family composition and parent educational status were assessed on the baseline parent survey.
Analyses were conducted by using SAS 9.2 (SAS Institute, Inc, Cary, NC); all estimates were weighted by the survey weights to be representative of the US population and to minimize nonresponse bias. We used logistic regression to assess whether presence of a favorite ad among never-smokers at baseline would predict experimentation by the fifth survey, and we adjusted for the randomized group assignment as well as the baseline variables of gender, age, race (white versus nonwhite), geographic region, self-assessed school performance, living in a single-parent home, parental education, presence of friends who smoke, presence of other smokers in the house, and susceptibility to smoking.
Characteristics of the Sample
At study enrollment, 30% of the participants in the sample (n = 1036) were aged 10 or 11 years and 70% were 12 to 13 years old (Table 1). No gender imbalance was observed among the 10- and 11-year-old smokers (3%–4%). However, among 12- to 13-year-olds, almost twice as many boys as girls had experimented (11.1% vs 6.5%). For both boys and girls, among non-Hispanic white adolescents (67% of the sample), a lower proportion had experimented than other races/ethnicities. Parental college education (any college attendance) was associated with lower initiation during this early adolescent period for both boys and girls, but particularly so for girls. Teenagers from single-parent families were more likely to have experimented than were other teenagers (boys, 12.0% vs 8.0%; girls, 8.3% vs 4.3%). Those who reported having below-average grades in school were more likely to have tried smoking than were others (boys 12.6% vs 6.5%; girls 10.3% vs 2.4%). The proportion who reported having smoked seemed to be higher for boys who lived in the southern or midwestern United States compared with those who lived in either the northeast or the western United States.
Predictors of Smoking Initiation
By survey 5 (completed in 2008), 27% of baseline never-smokers reported having smoked. A logistic regression assessed which baseline variables predicted later smoking (Table 2). Nonsignificant variables included age, gender, race/ethnicity, parental education, geographic region, and whether the parent was randomly assigned to receive the parenting intervention. Teenagers from single-parent families (29%) were ∼70% more likely to smoke than those from 2-parent families (adjusted odds ratio [aOR]: 1.7 [95% confidence interval (CI): 1.1–2.6]). Teenagers who indicated at baseline that they had below-average grades in school (40% of sample) were twice as likely to have smoked by survey 5 (aOR: 2.0 [95% CI: 1.4–2.9]). Only 8% reported having friends who smoked at baseline (when they were 10–13 years old); however, among those who did, 75% had smoked during the follow-up time (aOR: 5.6 [95% CI: 3.0–10.6]). Teenagers who lived with a smoker were also more likely to have smoked (aOR: 1.6 [95% CI: 1.1–2.4]). After accounting for the above-listed variables, teenagers who reported having a favorite cigarette ad at baseline (32% of sample) were more likely to have smoked by survey 5 than those who did not have a favorite cigarette ad (aOR: 1.5 [95% CI: 1.0–2.3]; P = .0393).
Changes in Reporting a Favorite Cigarette Ad
Among baseline never-smokers, the proportion of participants in each survey who reported that they had a favorite cigarette ad is presented in Fig 1. For girls (panel A), 31% to 34% indicated a favorite advertisement across the first 4 surveys. However, on survey 5, conducted from July 2007 to September 2008, the proportion of girls having a favorite cigarette ad increased to 44%. This increase reflected increased nominations for Camel as a favorite cigarette advertisement, and nominations for Marlboro were relatively stable over the 5 surveys. During the first 4 surveys, ∼10% to 13% of girls chose Camel as their favorite ad (year 4 estimate: 12.2% ± 3.3%). In survey 5, this estimate was 9.3 percentage points higher at 21.5%, with no overlap in confidence limits. Nearly half (45%) of the girls who nominated Camel as a favorite ad on survey 5 reported no favorite cigarette advertisement on survey 4, and only 3% had previously reported Marlboro.
The proportion of boys who nominated a favorite cigarette ad was 37.8% on survey 1 and essentially unchanged at 39.2% on survey 5 (panel B). The nomination of Marlboro was unchanged across surveys (12.6% [survey 1] vs 13.6% [survey 5]). The nomination of Camel was unchanged across the first 4 surveys (13.0% [survey 1] vs 14.1% [survey 4]) but increased to 18.2% on survey 5, an increase that was less than half the increase observed among girls and was not statistically significant, as evidenced by overlapping CIs.
More than one-third of this national youth cohort nominated a favorite cigarette advertisement when they were aged 10 to 13 years in 2003–2004. This was similar to the proportion of same-age Californians who nominated a favorite advertisement in 2002 and was ∼40% lower than same-age Californians in the mid-1990s,31 evidence of the continuing effect of enforcement of the MSA on reducing, but far from eliminating, this measure of tobacco industry influence. As in our earlier studies conducted before the MSA,16,17 adolescents who had never smoked but who reported a favorite cigarette advertisement had a 50% increase in probability of future experimentation with smoking.
The proportion of adolescents nominating a favorite cigarette advertisement was fairly constant (33% for girls, 38% for boys) across the 4 surveys administered between September 2003 and June 2007. However, survey 5 recorded a large jump in the proportion of adolescent girls who nominated a favorite advertisement, with almost all the increase attributed to Camel nominations (a 43% relative increase). This was more than double the percentage-point increase observed among boys of the same age, whereas nominations for Marlboro remained stable for both boys and girls. The theory of persuasive communication15 predicts that this increase in receptivity should translate into increased smoking initiation in the remaining 5 years of the smoking-initiation age window.
Our measure of favorite ad has always included a probe for adolescents who hesitate to nominate a favorite ad, asking them to nominate the ad that most attracts them, as indicated by theory.15 Without this probe, much lower rates of receptivity have been reported, and the association of favorite ad with later smoking has been questioned.32,33
Considerable lay and business press accompanied the 2007 launch of the Camel No. 9 marketing campaign, which noted that this sub-brand was particularly attractive to young women, including underaged girls.34,35 Looking much like a fashion spread, the Camel No. 9 campaign (Fig 2) ran in 5 of the top 10 US magazines as ranked by teen readership.36,37 The first advertisement in Vogue magazine led to a series of letters to the editor objecting to the ads as targeting teenagers.34,35 In June 2007, 41 members of Congress wrote to the publishers of 11 leading women's magazines and to the relevant federal authorities asking them to stop accepting tobacco advertisements targeted at teenagers, particularly those for Camel No. 9, because they seemed to violate the MSA. In 2008, RJ Reynolds announced that it would discontinue print advertising of Camel No. 9 but not related sponsored events or point-of-purchase or Internet marketing.38
According to the RJ Reynolds Web site, Camel is the company's largest and fastest-growing full-price brand and has a track record of consistent share growth and marketplace momentum. It is one of the few full-price brands in the industry that has shown sustained growth.25,26 This study suggests that Camel's increased market share may reflect the company's reverting to strategies39,40 that were effective before the MSA, when RJ Reynolds' Joe Camel advertising campaign was shown to target preadolescents.41,42
A strength of this study is its longitudinal design and its national sampling frame. Unlike previous studies that investigated the role of tobacco advertising, this study represents every region of the country. Five separate surveys collected information both on receptivity to cigarette advertising and smoking behavior, starting at the youngest adolescent ages. However, respondents to the national random-digit-dial survey were invited to participate in a study of parenting to prevent teen problem behaviors. Those who volunteered might be different from those who did not, although the demographics of the sample reflected those of the general population.
The surveys were not designed to specifically investigate tobacco advertising and did not include multiple items to explore this question in depth, which are limitations to consider when interpreting our results. Also, although the timing of the surveys enabled exploration of issues related to the Camel No. 9 campaign, the average age of teenagers at survey 5 was 15.7 years. Experimentation with smoking continues at least through the age of 20.20 Hence, this study may underestimate the campaign's influence on smoking uptake, and additional follow-up is needed. Another limitation was that the study did not collect detailed information on exposure to smoking in movies, and such exposure seems to be a stronger association with outcome than having a favorite advertisement.33 An additional limitation is that the outcome variable is any smoking in adolescence rather than dependent smoking. However, in previous research, receptivity to tobacco advertising and promotions has predicted both any smoking behavior at 3 years16 and current established adult smoking at 6 years.17
This article presents credible evidence that the Camel No. 9 cigarette advertising campaign has targeted underaged girls. These findings suggest that the Camel No. 9 campaign may have been similar to the Joe Camel campaign in its targeting of underaged youth. Some 44% of adolescent girls nominated a favorite cigarette ad in 2007, and the evidence in this national sample is consistent with earlier studies showing that these girls are at a 50% increased risk of starting to smoke. Numerous public complaints have alleged that this Camel No. 9 campaign and other recent RJ Reynolds campaigns have violated the MSA. This study provides evidence to support these complaints.
This study was funded by National Cancer Institute grant CA093982, an American Legacy Foundation grant, and Tobacco Related Disease Research Program grants 17RT-0088 and 15RT-0238 from the University of California.
- Accepted November 10, 2009.
- Address correspondence to John P. Pierce, PhD, Cancer Prevention and Control Program, Moores UCSD Cancer Center, University of California at San Diego, La Jolla, CA 92093-0901. E-mail:
FINANCIAL DISCLOSURE: The authors have indicated they have no financial relationships relevant to this article to disclose.
Funded by the National Institutes of Health (NIH).
- MSA =
- Master Settlement Agreement •
- aOR =
- adjusted odds ratio •
- CI =
- confidence interval
- ↵National Association of Attorneys General. Master Settlement Agreement. Available at: www.ag.ca.gov/tobacco/pdf/1msa.pdf#xml=http://search.doj.ca.gov:8004/AGSearch/isysquery/467cabd0-2e70-4754-a167-033e20216b93/8/hilite. Accessed February 11, 2009
- DiFranza JR,
- Wellman RJ,
- Sargent JD,
- Weitzman M,
- Hipple BJ,
- Winickoff JP
- Lovato C,
- Linn G,
- Stead LF,
- Best A
- ↵American Legacy Foundation. Who we are. Available at: www.americanlegacy.org/whoweare.aspx. Accessed September 14, 2009
- ↵National Association of Attorneys General. Master Settlement Agreement. 1998. Available at: www.naag.org/backpages/naag/tobacco/msa/msa-pdf. Accessed February 10, 2009
- ↵Monitoring the Future. Available at: www.monitoringthefuture.org. Accessed February 12, 2009
- Neighmond P
- Alpert HR,
- Koh HK,
- Connolly GN
- Kotler P,
- Armstrong G
- ↵National Cancer Institute. The Role of Mass Media in Promoting and Reducing Tobacco Use. Tobacco Control Monograph No. 19. Bethesda, MD: US Department of Health and Human Services, National Institutes of Health; 2008. NIH publication 07–6242
- Evans N,
- Farkas A,
- Gilpin E,
- Berry C,
- Pierce JP
- Lindzey G,
- Aronson E
- McGuire WJ
- Lindzey G,
- Aronson E
- McGuire WJ
- Gilpin EA,
- Distefan JM,
- Pierce JP
- Messer K,
- Pierce JP
- ↵US Department of Health and Human Services. Preventing Tobacco Use Among Young People: A Report of the Surgeon General. Washington, DC: US General Accounting Office; 1994
- Beirne M
- Beirne M
- Collins M
- ↵US Census Bureau. Tobacco use supplement to the current population survey (2001–2002). Available at: http://aspe.hhs.gov/hsp/06/Catalog-AI-AN-NA/TUS-CPS.htm. Accessed January 15, 2010
- Izrael D,
- Hoaglin DC,
- Battaglia MP
- Gilpin EA,
- White MM,
- White VM,
- et al
- Sargent JD,
- Gibson J,
- Heatherton TF
- Noveck J
- ↵Mediamark Research Inc. Magazine Publishers of America teen market profile. Available at: www.magazine.org/ASSETS/905DE1D985A04046ABAC7754572F1163/teenprofile04.pdf. Accessed September 10, 2009
- ↵Campaign for Tobacco-Free Kids. Deadly in pink: big tobacco steps up its targeting of women and girls. Available at: www.tobaccofreekids.org/reports/women_new. Accessed September 10, 2009
- Elliott S
- Mizerski R
- Copyright © 2010 by the American Academy of Pediatrics