The tobacco industry is important to the economy. In 1991, worldwide tobacco sales exceeded $59.8 billion and in 1992 the industry was rated as one of the top one hundred advertisers (Pechmann and Ratneshwar, 1994). However, there are high prices to pay - socially, economically, and personally - as a result of this industry. Annual mortality figures indicate that cigarette smoking is the number one cause of preventable death in the United States. An estimated 390,000 people die each year of smoke related illnesses, which is greater than the combined mortality for cocaine, crack, AIDS, homicide, suicide, and alcohol abuse (Botvin, G., Baker, Botvin, E., Dusenbury, Cardwell, and Diaz, 1993).
This paper will review research which suggests that social influences on smoking occur as a result of several external factors that adolescents face on a daily basis - parental factors, peer pressure, and advertising. Many current and past smokers began experimentation and use during their adolescent years. Social influence is a central component of all models of adolescent substance abuse. It is assumed to be a direct predictor of experimentation and initiation with cigarettes, alcohol, and marijuana. One of the major theories that is used to describe this phenomenon is the social control theory. This theory describes bonding factors, such as involvement, family attachment, job or education commitment, and belief in conventional social norms and values. Involvement refers to the idea that individuals who are deeply involved in non deviant activities lack the time to get involved in deviant acts. Attachment to others who value positive norms tends to decrease the likelihood of deviation from the norms as a direct result of the opinions expressed by others. Those with positive work and educational ethics are less likely to deviate as well as long as their commitment is related to conformity of social norms. Belief refers to the notion that persons who agree to social norms feel it appropriate to abide by them. Thus, they are less likely to engage in deviant behavior. They all have been found to have either a negative or positive influential effect (DeFronzo and Pawlak, 1993).
To illustrate, several studies have identified social controls whose absence has caused adolescents to experiment and initiate in tobacco use. Starting at home, the influence of parental attitude and behavior toward adolescent smoking has a major impact on adolescent smoking. Newman and Ward (1989) sampled 735 students from 12 schools in and around one moderately sized Midwestern city, 18.5% of the sample were smokers. In this study, Newman & Ward asked the students questions via a questionnaire in order to rate the parental attitudes. One question asked was, “With regards to my smoking cigarettes, my parents/guardian would: threaten to punish me if I smoked; haven’t told me how they feel if I smoke? ; have told me they don’t care if I smoke” (Newman and Ward, 1989, p. 150). Two-thirds of the students reported that both parents would be upset if they smoked. An interesting note was that about two-thirds of the nonsmoking adolescents reported parental disapproval versus one-half of the smoking adolescents. The analysis of the data revealed that when neither parent smoked and both disapproved of it, about 10 percent of their adolescents smoked. When the attitude toward smoking was held constant, but both parents were smokers, the percentage of adolescent smoking almost doubled. When neither parent smoked and had indifferent attitudes toward adolescent smoking, almost 20 percent of their adolescents smoked. Again, when the attitudes were held constant and both parents smoked, approximately 30 percent of their adolescents smoked.
These results suggest two important points according to Newman and Ward (1989). First, parental attitude, when expressed, appears to be important in moderating adolescent smoking behavior regardless of the parents’ smoking habits. However, since many parents are unable to tell their children not to smoke when they themselves smoke, adolescents will not feel any influential pressures against smoking and thus, the smoking rates might increase. Second, there is an apparent increase in permissiveness of parents. “Today, even more than in 1983, the dangers of cigarettes are documented by the scientific community, and accepted by parents and community leaders . . . [this leads] parents to feel the anti smoking message is clear and they no longer need to state the obvious to their youngsters” (Newman and Ward, 1989, p.153). As seen parents indeed play an integral role with regards to influencing their children, especially cigarette smoking.
Males (1995) reviewed the results of the Los Angeles Survey that was conducted in 1993 and found that parental smoking was a significant indicator for adolescent smoking. The study started by administering a baseline survey on smoking to students in health classes at four northern Los Angeles junior high schools. The survey asked questions regarding the students’ basic demographic information; smoking habits, if any; parental smoking habits. A total of 407 students was surveyed. Only 320 students then immediately heard an anti smoking presentation that included graphic details of health risks involved with smoking. Immediately following the presentations, the students were re-administered the survey. The other 90 students were not shown the presentation, but were retested.
Results of the baseline study concluded that adolescents whose parents smoked, where twice as likely to have tried cigarettes, three times more likely to have smoked within the past week, and two and one half time more likely to indicate future intent to smoke. Furthermore, they were four times more likely to be smoking weekly by the age of twelve than were children of nonsmoking parents. Males were significantly more likely to smoke than females if one or more parent smoked. DeFronzo and Pawlak (1993) likewise found that being female was positively associated with belief and attachment bonds that inhibited smoking and alcohol abuse. The after survey indicated that students who came from homes where both parents smoked, almost half of the students stated that they would or might smoke in the future - twice the proportion of students with one smoking parent and four times that of students with nonsmoking parents. Again, this research confirms previous research in that parents do play a key role in influence when in comes to smoking, even when other factors are figured in as in this study - the anti smoking presentation.
Stepping outside of parental attitudes and behavior, DeFronzo and Pawlak (1993) examined the effects of social bonds and childhood experiences on smoking. The sample of 595 subjects (34.8% were smokers) had responded to Ballots B and C on the 1988 General Social Survey conducted by the National Opinion Research Center. Overall, there were 17 questions ranging from smoking issues to family connections to commitment-related questions to religious beliefs that were analyzed.
Evaluating the results based on the social bond theory, the results of the analysis of the data revealed that religious belief, moral conformity, commitment, and involvement had significant negative effects on smoking; however, attachment was not significant. Adolescents who were beaten and had partial families were more likely to engage in smoking, but the relationship between the two variables was insignificant. An interesting note about the study involves adolescents who live within non-traditional families at the age of 16. DeFronzo and Pawlak found that there was a significant increase in smoking between these adolescents and attributed this influence to a different casual mechanism. “Rather than increasing one’s desire to escape, being a child in an incomplete family may increase one’s freedom, making it relatively easy to conceal deviant behavior, such as smoking” (DeFronzo and Pawlak, 1993, p. 634).
Also concerned with social bonds and childhood experience, Epstein, Botvin, and Diaz (1999) examined inner-city adolescents and the social influences they faced based on four domains: socioeconomic background, social influences to smoke, social and personal competence, and individual differences. The study was also based on the social learning theory and the problem behavior theory. These theories state that behaviors, like smoking, are socially learned, purposeful, and functional and result from the interactions of social and personal factors.
Over the two-year study, 2192 junior high students form the New York City schools participated. All of the students completed a questionnaire each year over a four-week period that measured self-reporting smoking and cognitive, attitudinal, and psychological characteristics hypothesized to be related to smoking initiation. To ensure accuracy on the self-reporting questionnaire, two measures were taken: 1) the students were guaranteed that the information was confidential, and 2) each student was administered a carbon monoxide breath test. Questions ranged from gender and age to smoking habits, if any to assertiveness and decision-making skills to self-esteem and self-efficacy. An eleven-point smoking index assessed smoking frequency. The question asked, “How often do you currently smoke?” was followed by eleven possible answers. Choices ranged from one (never smoked) to eleven (a pack or more a day). All of the remaining questions were based similarly to that of the cigarette question. At the end of year two, the data was analyzed using statistical means.
The significant findings in this study evolved while looking at the second year results. Two social influences, mothers smoking and friends smoking both predicted smoking one a year later. This study, contrasting with earlier studies, revealed that the father’s smoking and/or the sibling’s smoking did not influence adolescent smoking. Friends turned out to be the leading social influence in smoking initiation and experimentation. In contrast to the other studies mentioned, Botvin and Diaz concluded that adolescents spend much of their time with similar friends and that they are socially influenced by them; however, perceptions that “everyone is smoking” between peers and adults did not predict subsequent smoking.
Another study supporting the role of social learning was completed by Harton and Latané (1997) who examined social influence and adolescent lifestyle attitudes during a two-year study of nine to fifteen year olds. They hypothesized that over time, the attitudes of the adolescents would change as their social influences change. Students who completed the data over the two-year study totaled 193. Students were tested in groups of two to twenty-two in the computer testing room at their school in November and March and on days that did not follow a weekend or holiday, producing a time span of five months between tests. Once the students were randomly seated, they were told that they were being tested for attitudes and friendships. The questions, intermixed with others, centered around the sociometric measures and on thirteen lifestyle attitudes (ranging from healthy to deviant activities). The attitude questions all started off with, “Do you think it is O.K. for students in your grade to . . . ?” (Harton and Lantané, 1997, p. 204). Six sociometric questions were also asked regarding whom the students talk to in class, liked in class, and respected in class.
With regards to attitudes toward smoking, the results revealed that fourth grade approval had a mean of 1.72 (1 = not at all approved and 5 = very much approved). By the eighth grade, the approval rating was 2.15. Furthermore, there was a linear relationship between the attitude and increasing grade of 0.58. Girls were less approving in fourth grade; however, they were quick to catch up to the boys by the eighth grade.
Other factors, such as popularity between same sex and opposite sex were used in calculating social and attitudes. Popularity over time did not increase among the same sexes for males; however, influence females were more popular among the same sexes. The major increase in popularity was seen in the opposite sexes - attributed to heterosexual dating in older students. Students who were more popular were more approving of smoking and thus, were found to be more influential on others as well.
Further interpretation of the results showed that during early adolescence, children may begin to respect and emulate their older peers and siblings more than they do their parents. This evidence was supported by that fact that popular children (opposite sex) were more approving of the sixteen attitudes that were questioned in the study. In the language of social impact theory, “other-sex children may be higher strength, influencing children more than do their same-sex peers on these developmental attitudes” (Harton and Lantané, 1997, p. 214).
This study did not explicitly state that cigarette smoking was socially influenced by same sex or other-sex relationships and popularity; however, an assumption can be made as to the relationship. More than likely, popularity among same sex peers as well as other-sex peers plays a role in the attitude of smoking. It is natural that if one wants to be popular with a given group, he or she will conform to that group for approval. Also, since approval ratings increased over the grades, one can infer that social influence by older peers does play a key role in smoking attitudes, according the social impact theory. “Parents and family may be more influential when children are younger and friendships are based on activities rather than on intimacy, but as children begin adolescence, older siblings and teenagers [peers] may become the primary objects of respect and, thus, more influential” (Harton and Lantané, 1997, p. 198).
To further examine the notion that attitude changes over the adolescent years, a study completed by Morgan and Grube (1989) was conducted. This study examined three thousand students ranging in age from 13-17 years old. This study examined three key social areas - normative factors, beliefs about consequences, and social bonding. First, normative factors are subdivided into perceived approval and behavioral norms. Perceived approval consists of beliefs about the approval or disapproval of others for a particular behavior, while behavioral norms consist of beliefs about the extent to which significant others engage in the behavior themselves.
This study utilized a survey by way of questionnaires on two separate occasions. The questionnaires asked about a range of variables relating to cigarette smoking, particularly frequency of smoking behavior during the previous month, normative beliefs, beliefs about consequences, social bonding, and background characteristics. The questions were set up similar to the previous studies in that the answers could range from very good to very bad. Once completed, the questionnaires were matched across phases and the data was analyzed.
The results showed a curvilinear relationship from age 13 to 17. Further data analysis found that peer smoking was far more significant in influence than peer approvals; 0.613 and 0.294 respectively (both were curvilinear over age groups). In contrast, parental approval was more influential than parental smoking; 0.349 and 0.083 respectively (both were stable over age groups).
Beliefs about consequences consist of two subdivisions as well, perceptions of the likelihood that a behavior will have specific personal effects and evaluations of these effects. Smokers are less likely than nonsmokers to believe that their own smoking will bring about negative consequences like lung cancer, offending others and bad breath (Morgan and Grube, 1989). Smokers often feel that smoking will make them more popular, feel relaxed, and will help them concentrate. Again, the results showed a curvilinear correlation with age. Furthermore, overall relationships between beliefs about consequences and smoking were much stronger in the case of perception of the likelihood of negative consequences and evaluation of them than for positive consequences and evaluations. That is, participants who perceived smoking as unhealthy were more likely to understand that concept and evaluate smoking as a bad habit as compared to those who felt smoking was not detrimental to their health. Interestingly noted, the attitudes toward negative consequences and evaluations increased with age and decreased with positive consequences and evaluations.
Social bonding was found to be related to the social control theory previously mentioned. Morgan and Grube’s (1989) data did not support a curvilinear relationship from age 13 to age 17, however. The relationship remained fairly stable, contrasting the previously mentioned studies. Three out of the four social bonds examined were found to have significant correlations with smoking - bonds to schools, religion, and family. In contrast to previous studies mentioned, social bonds to peers were unrelated to smoking behavior.
A different approach was used by Sussman, Hahn, Dent, Stacy, Burton, and Flay (1993) that revealed similar results to the Morgan & Grube (1989) study. This study, different from the others, was conducted utilizing naturalistic observation of adolescent tobacco use. The main purpose behind this study was to observe behavioral events as they occur to explore the naturally occurring social context of adolescent tobacco use including whether or not smoking occurred in a group context. Furthermore, Sussman et al, observed whether or not offers were made of tobacco products to others in the group. Forty-one students were observed using tobacco in fifteen different grade levels and school settings (rural, urban, hang out spots, etc...). The observers used a single-sheet checklist to minimize visibility and to maximize speed of data collecting. Each student was assigned a letter and then the group was diagramed on the sheet. Various information was also recorded about the persons: approximate age, gender, sex, ethnic group, offers to smoke, refusals of offers, and clothing style, to name a few. Each observation lasted for 30 minutes or until the group dispersed.
The results were analyzed and some were found to coincide with other studies. Many smoking groups were smaller than nonsmoking groups. Cigarette sharing and offers to smoke within the groups were made. Sussman, Hahn, Dent, Stacy, Burton, and Flay (1993) identified that the offers to smoke were minimal and that the groups did contain both smokers and nonsmokers. Of six group observations made, only two groups exhibited offers to smoke to fellow smokers. No refusals were observed. One of these two groups where cigarettes were being offered included five students who were not smoking and who were not offered cigarettes. The remaining groups in which no offers were made contained nonsmokers. Sussman et al, found that the predominant effect of these groups was positive. Thus, Sussman et al, concluded, “If future research obtains the same results, perhaps less emphasis should be placed on assertion refusal training in prevention programming and more emphasis should be placed on informational sources of social influence to use tobacco” (p. 807). Furthermore, 22% of the students were observed smoking alone as compared to previous studies (through interviews) that showed 12% smoked alone. Sussman et al, also speculated that the oldest among the youth who were smoking alone may have had cravings even when they were not in a group setting. This illustrates that social influence to smoke in a crowd among fellow, respected peers may not be the only reason for smoking. Also, solitary smokers may be smoking for personal reasons, such as stress relief or boredom.
To this point, social influences based on family and friends have been discussed. However, social influence in one last area needs to be investigated - the media. Cigarette ads are seen by most on a daily basis in magazines, on billboards, sports sponsorships, and on television (indirectly), to name a few. Researchers have found that ads that depict attractive, sexy models who are engaged in exciting and cool activities may suggest to impressionable youths an association between the desirable traits of the cigarette models and people who smoke. This concept can also be used with icons, such as Old Joe Camel.
Pechmann and Ratneshwar (1994) examined the effects of three different advertisements - control ads (unrelated ads to smoking), anti smoking ads, and cigarette ads, in that order. The unrelated, control ads were used to examine the subjects’ prior judgements of a smoker. All of the ads were rated based on the traits in the following order: smart, intelligent, healthy, grown-up, mature, good-looking, attractive, exciting, adventurous, popular, has friends, and cool. They were then asked whether they might personally like the person depicted in the ad. The responses were rated on a scale from one (definitely not) to seven (definitely yes). The subjects rated smokers and nonsmokers the same, except when it came to common sense and like-ability, which were rated as relatively lower in smokers. The anti smoking ads had only reiterated what the subjects had already known about the health effects of smoking. However, the subjects judged the smoker even lower on both common sense and personal appeal. Also, the smoker was judged lower on maturity, glamour, and traits that even subject had no negative preconceptions about for smokers. The cigarette ads depicted sexy, cool, popular, and exciting layouts. The results showed that even after viewing anti smoking ads, subjects interpreted the smoker’s traits to be more positive and had more favorable thoughts about the smoker. However, the subjects did not rank the smoker higher on common sense, personal appeal, glamour, or maturity. Pechmann and Ratneshwar concluded that ads did indeed have an effect on the subjects. Anti smoking ads worked well if they were used more often and in conjunction with school and home education. The cigarette ads slightly influenced the subjects’ judgements of the smoker, but their effect was not as profound as were the effects of the anti smoking ads, thus, more anti smoking ads and lessons need to be implemented.
To further illustrate how the tobacco industry has impacted youth, a study by DiFranza, Richards, Paulman, Wolf-Gillespie, Fletcher, Jaffe, and Murray (1991) investigated the extent to which advertising influences children. This study, included 1,060 high school students from five states. In addition, 345 adults from one state were included. The researchers exposed subjects to various ads and famous icons and to determine if the subjects could identify them. The results were astounding. When the students were shown a picture of Old Joe Camel without making any reference to cigarettes, 97.7% recognized him as compared to 72.2% recognition by the adults. The students were also better at recognizing what Old Joe Camel sold - 97.5% versus 67% by the adults. The students also found the ad more appealing than did the adults. Studies conducted prior to the Old Joe Camel campaign indicated that only 2.7% of smokers aged 17 to 24 smoked camels. In the current study, 32.8% of the smokers under 18 years old smoked camels. Camel’s share in the market over its three-year campaign rose from 0.5% to 32.8%, which demonstrates that Old Joe Camel has been rather influential on the student population. Whether examining the experimentation and/or initiation of smoking through the social control theory, social bond theory, social impact theory, social learning theory, or social proof theory, social influence is a central component of all models of adolescent substance abuse. It is assumed to be a direct predictor of experimentation and initiation with cigarettes, alcohol, and marijuana. A reoccurring theme in all of the mentioned studies centers on normative influences. Most adolescents experimenting with cigarettes appeared to correlate with their social contacts in everyday life. If many of the adolescents’ contacts smoked, a positive message was “flashed” that smoking was not deviant and, thus it was acceptable. On the other hand, if the adolescents’ contacts were mostly nonsmokers, then they were unlikely to engage in smoking since smoking appeared to be non normative.
From parents to peers to advertisers, adolescents are exposed to cigarettes every day of their lives. It appears that smoking has become a social norm that only time and effort can change. It will take a lot more that telling adolescents that smoking is not healthy. It will require massive efforts from all areas to demonstrate to the adolescent population the negative aspects of smoking. Education inside and outside of the classroom coupled with anti smoking education is required.
Botvin, G. J., Baker, E., Botvin, E., Dusenbury, L., Cardwell, J., & Diaz, T. (1993). Factors promoting cigarette smoking among Black youth: A casual modeling approach. Addictive Behaviors, 18, 397-405.
DeFronzo, J., & Pawlak, R. (1993). Effects of social bonds and childhood experiences on alcohol abuse and smoking. Journal of Social Psychology, 133, 635-642.
DeFronzo, J., & Pawlak, R. (1993). Being female and less deviant: The direct and indirect effects of gender on alcohol abuse and tobacco smoking. Journal of Psychology, 127, 639-648.
DiFranza, J. R., Richards, J. W., Paulman, P. M., Wolf-Gillespie, N., Fletcher, C., Jaffe, R. D., & Murray, D. (1991). RJR Nabisco’s cartoon camel promotes Camel cigarettes to children. Journal of the American Medical Association, 266, 3149-3154.
Epstein, J. A., Botvin, G., & Diaz, T. (1999). Social influence and psychological determinants of smoking among inner-city adolescents. Journal of Child & Adolescent Substance Abuse, 8, 1-15.
Harton, H., & Latané, B. (1997). Social influence and adolescent lifestyle attitudes. Journal of Research on Adolescence, 7, 197-220.
Males, M. (1995). The influence of parental smoking on youth smoking: Is the recent downplaying justified? Journal of School Health, 65, 228-232.
McBride, C. M., Curry, S. J., Cheadle, A., Anderman, C., Wagner, E. H., Diehr, W., & Psaty, B. (1995). School-level application of a social bonding model to adolescent risk-taking behavior. Journal of School Health, 65, 63-69.
Morgan, M. & Grube, J. (1989). Adolescent cigarette smoking: A developmental analysis of influences. British Journal of Developmental Psychology, 7, 179-189.
Newman, I., & Ward. J. (1989). The influence of parental attitude and behavior on early adolescent cigarette smoking. Journal of School Health, 59, 150-153.
Pechmann, C. & Ratneshwar, S. (1994). The effects of anti smoking and cigarette advertising on young adolescents’ perceptions of peers who smoke. Journal of Consumer Research, 21, 236-251.
Sussman, S., Hahn, G., Dent, C. W., Stacy, A. W., Burton, D. & Flay, B. R. (1993). Naturalistic observation of adolescent tobacco use. International Journal of Addictions, 28, 803-811.