Evaluation of Antismoking Advertising Campaigns

SINCE 1989, California,¹ Massachusetts,^{2 - 4} and Arizona⁵ have implemented large-scale, paid, tobacco-control campaigns to discourage people from starting to smoke and to encourage smokers to stop. (Michigan has a smaller campaign.) California uses a general market approach with strong, antitobacco industry and secondhand smoke components.⁶ Massachusetts uses a more youth-oriented approach and, although it has sought to discredit the industry, its spots are less confrontational with the industry than California's advertisements. The Arizona program is the most narrowly focused, limiting itself to youth and pregnant women,⁷ with no messages attacking the tobacco industry. This article reviews the qualitative marketing research used to develop antitobacco media campaigns and compares the overall cost-effectiveness of the California and Massachusetts campaigns. We found that more aggressive campaigns are more effective.

California,^{8 - 13} Massachusetts,^{14 - 16} and Michigan¹⁷ provided the focus group research that their advertising agencies used to develop their campaigns. Arizona's advertising agency, the Riester Corporation, refused to share its research with us. The Centers for Disease Control and Prevention provided a copy of its Media Campaign Resource Book,¹⁸ which contains information from the advertising agencies and state health departments on their advertisements, including target audience, optimal placement, key message, and focus group results.

We obtained 118 advertisements and the results from 186 focus groups involving more than 1500 children and adults that evaluated both produced advertisements and additional concept advertisements that were not produced. We categorized the advertising strategies as industry manipulation, secondhand smoke, addiction, cessation, youth access, short-term effects, long-term health effects, or romantic rejection.

Because focus groups are qualitative research and are subject to methodological limitations such as group dynamics and small sample size, broad quantitative conclusions should not be drawn from the findings. Nonetheless, focus groups are widely used by advertising agencies in designing new advertisements.¹⁹ Conclusions about effectiveness or ineffectiveness of various strategies are based on the responses of the participants in these focus groups, not testing in other ways.

California's advertisements have been on the air since 1990, and Massachusetts' since 1994, which allows a rough comparison of the cost-effectiveness of these campaigns. Using per capita cigarette consumption from the Tobacco Institute's The Tax Burden on Tobacco,²⁰ we examined the difference in per capita consumption of cigarettes between California and Massachusetts and the rest of the United States (excluding California and Massachusetts) as our outcome variable. The population estimates for computing the per capita numbers in The Tax Burden on Tobacco are based on the number of potential smokers, obtained by dividing reported total consumption by reported per capita consumption. We then computed a simple linear regression with time as the independent variable for the differences over time beginning the fiscal year before media program expenditures occurred (1989 for California and 1993 for Massachusetts). The slopes of the resulting regressions are a measure of the difference in the rate of decline in per capita consumption in California or Massachusetts compared with the rest of the United States. By taking this difference as the dependent variable, we partially account for changes in pricing and other trends in the national environment, without requiring the assumption that there is a linear decline in national (excluding California and Massachusetts) cigarette consumption. We then divided the slopes of these regressions by the total per capita expenditures (in 1996 dollars) on media over the lives of the programs^{4 ,21 - 22} to obtain an estimate of the rate of per capita decline in cigarette consumption per year and per per capita dollar spent.

Paid media is most effective when used as part of a multifaceted approach to reduce smoking, including community programs,²³ higher taxes, and school-based programs.^{24 - 28} Because the various program elements are designed to work together, it is difficult to separate the effects of paid media from other contemporaneous tobacco control interventions. Nonetheless, there is considerable evidence that paid antismoking advertisements are effective in reducing cigarette consumption.

Examination of tobacco consumption data shows a relationship between the presence of the media campaign and declines in consumption in California. The first wave of the media campaign began in April 1990 and lasted until the following year. It was gradually phased out between March and June 1991. Pierce et al^{29 - 30} found that tobacco consumption in California declined 13.7% from September 1988 to May 1989 (5 months after the imposition of the $0.25 per pack tax), then increased by 3% between October 1989 and March 1990. Between April 1990 and March 1991, roughly the same period as the first wave of the media campaign, consumption again decreased, this time by 12.2%. Because the effects of the tax seemed to dissipate after May 1989 and no other Proposition 99–related tobacco control interventions were in effect during the first wave of advertising, the second dramatic drop in consumption can be attributed to the media campaign. Cigarette consumption declined again by 12% between February 1992 and April 1993.³⁰ Because the second phase of the media campaign did not begin until October 1992 and the other Proposition 99 programs went into effect in 1992, this decline could reflect both the media campaign and the other interventions.

Media campaigns can also be effective in influencing smokers' decisions to quit smoking. Popham et al³¹ surveyed adult Californians who had quit smoking during the first wave of the California media campaign in 1990 to 1991. In response to uncued questions, 6.7% of smokers cited an advertisement that they had seen or heard as a factor in their decision to quit smoking. When asked direct questions about the media campaign, an additional 34.3% of smokers in the survey replied that a tobacco-control advertisement had been influential in their decision to quit. This result translates to 33000 former California smokers for whom the 1990 to 1991 media campaign was a significant factor in their decision to quit and an additional 173000 former smokers for whom the advertisements contributed somewhat to their decision to quit smoking.

Popham et al³² also evaluated the effect of the first media campaign on California students before the campaign began and 3 times during the campaign. To determine whether the media campaign met the goal of reducing tobacco use, Popham et al³² evaluated the campaign's effects according to 5 indicators: campaign awareness, tobacco use, smokers' intention to quit, nonsmokers' intention to start smoking, and attitudes regarding smoking. Over the course of the campaign, awareness of the campaign increased, smoking prevalence decreased, nonsmokers thinking about starting decreased, and health-enhancing attitudes about smoking increased. When Popham et al³² compared campaign-exposed and campaign-unexposed students from the fourth wave of data gathering, they found mixed results: health-enhancing attitudes about smoking were stronger among the exposed students, but the percentage of nonsmokers thinking about starting was also higher among exposed students. Popham et al³² recognized, however, that the method they used to measure nonsmokers' intentions was experimental and may not have been accurate.

Glantz³³ estimated that prior to the passage of Proposition 99, total cigarette consumption was falling by 45.9 million packs per year in California. After enactment, the rate tripled to 164.3 million packs per year. In 1992, the rate of reduction slowed to 19 million packs per year at approximately the same time the media campaign was suspended³³ by Gov Pete Wilson (based on claims that it was not effective). The American Lung Association sued Wilson and the media campaign was restored in 1992³⁴ ; cigarette consumption began to decline again. From 1995 to 1997, when there had been little tobacco-control advertising in California³⁵ and no new advertising produced, tobacco consumption was essentially flat (Figure 1).

Hu et al³⁷ conducted an econometric analysis of cigarette consumption in California between 1980 and 1992 using quarterly data and controlling for time, price excluding taxes, state tax, federal tax, and the media campaign. They estimated that cigarette sales were reduced by 1.33 billion packs from the third quarter of 1990 through the fourth quarter of 1992. They attributed 232 million packs of this decline (7.7 packs per capita) to the media campaign. Hu et al^{37 - 38} probably underestimated the impact of the media campaign because it does not take into account the tobacco industry's increased use of promotional activities to counter the media campaign.³⁹

In November 1992, residents of Massachusetts approved a measure similar to California's Proposition 99 that increased the tobacco tax by $0.25 per pack with the funds devoted to antitobacco activities.^{2 - 3} In 1993, when the tax went into effect, the tobacco companies reduced their wholesale prices to the 1992 pretax level.⁴⁰ These changes in wholesale pricing essentially eliminated the price increase associated with the tax. Massachusetts' tobacco control media campaign began in October 1993, after the price cut. Even so, per capita consumption in Massachusetts continued to decline from 1992 to 1996.⁴⁰ This evidence suggests that the media campaign in Massachusetts played a role in reducing cigarette smoking.

Similarly, an evaluation of the media campaign compared Massachusetts youth to their counterparts in states without antismoking media campaigns and found encouraging results.⁴¹ Massachusetts youth had significantly more knowledge about tobacco use, were more likely to cite additional reasons (other than health) not to smoke, and held stronger antismoking attitudes than youth in other states.

The focus groups we reviewed provided information on what strategies were thought to be effective or ineffective by the participants.

Tobacco advertising portrays smoking as glamorous and smokers as attractive and appealing. The industry manipulation strategy seeks to delegitimize the tobacco industry⁴² and deglamorize smoking. Industry manipulation advertisements make the industry the problem by exposing its predatory business practices. The message is, "Tobacco industry executives use deceitful, manipulative, dishonest practices to hook new users, sell more cigarettes and make more money."¹⁶

Industry manipulation advertisements carry different messages for adults and youth. Most adult smokers recognize the negative social and physical consequences of their smoking and are frustrated by their addiction. Industry manipulation advertisements help them redirect their feelings of guilt over their own smoking toward anger at the tobacco industry and its desire to profit from a deadly product.^{11 ,43 - 44} The advertisement that opened the California campaign, "Industry Spokesman," depicts tobacco industry executives sitting around a conference table in a smoke-filled room discussing ways to entice new smokers. This advertisement made all targeted adult groups angry and resentful and focused those feelings on the tobacco industry rather than on individuals.¹¹ Another advertisement, "Nicotine Soundbites," presents real industry executives denying that nicotine is addictive before Congress (Figure 2).

For youth, the industry manipulation strategy succeeds for a different reason. Young people begin smoking to express independence by rebelling against their parents and others who admonish them not to smoke.⁴⁷ They believe that they can make their own decisions, including the decision to smoke.⁴⁸ By making youth aware of the industry's calculated attempts to manipulate them, these advertisements tell young people that they are not acting independently. They also transform a low-interest topic, smoking, into an attention-getting, emotional issue,^{8 ,49} and reconfigure the parent–rebellious child dynamic by giving both youth and adults a common enemy—the tobacco industry.^{8 ,50}

Massachusetts found that the most effective positioning statement for young people is one that shows the industry as "money-hungry companies that intentionally and willfully target very young and vulnerable kids with manipulative and deceptive tactics in order to get them addicted to cigarettes at an early age so they become customers for life (or until tobacco kills them)."¹⁴ Focus groups showed youth disliked being manipulated by the tobacco industry.¹⁸

One California advertisement that tested well with youth, "Hooked," portrays a man fishing and tossing caught fish onto the dock. With the image of fish fighting to get away in the background, the narrator reveals that tobacco companies are using nicotine to hook more smokers since so many quit or die each day. The advertisement ends with the line, "The Tobacco Industry. They profit. You lose." This spot was effective at communicating to youth that the industry is relentless in its pursuit of profits.¹⁰

The most successful industry manipulation advertisements specifically attack the tobacco industry by name, rather than using a vague "they" or "them." The Centers for Disease Control and Prevention⁴⁸ conducted interviews with teenagers to test its antismoking advertisements and found that they misunderstood the reference to "they" in the advertising campaign, even though the cognitive content of the advertisements was clearly antitobacco. Only 10% of participants understood that "they" referred to the industry while a majority thought "they" referred to their friends or peers. Thirty-eight percent of the participants even thought the advertisements were promoting smoking.

Secondhand smoke advertising seeks to convince smokers that their smoking endangers others; to help denormalize smoking by portraying the dangerous effects of secondhand smoke on nonsmokers; and to motivate smokers to quit. To counter the industry's use of patriotic concepts like liberty and freedom to choose whether to smoke, this strategy shows that many people involuntarily breathe secondhand smoke at work and in public places and that children breathe their parents' smoke. Among youth, secondhand smoke messages can awaken a "sense of injustice for the little guy."¹¹ "Living Room," a California advertisement, portrays a brother and his much younger sister watching television. As the brother smokes, the sister begins coughing and smoke comes out of her mouth. This advertisement was effective among both adults and youth because it showed the child as a helpless victim and made people aware of the effects of their smoking on others.¹⁸

Addiction advertisements make both smokers and potential smokers aware that nicotine is an addictive drug and that the tobacco industry is using nicotine to hook smokers. This message angers current smokers and deters nonsmokers from starting. Adolescents believe cigarettes are addictive.⁴⁸ Coupled with the finding that youth do not want to feel they are being manipulated, the Centers for Disease Control and Prevention determined that an advertisement that emphasized the industry's deliberate use of nicotine to keep people smoking could be effective.

The California campaign uses the issue of nicotine addiction in combination with its industry manipulation strategy. "Industry Spokesman," "Nicotine Soundbites," and "Hooked" all conveyed the idea that nicotine is addictive and that the tobacco industry understands and uses this information to recruit new smokers and to keep old ones. These advertisements were some of the most effective of the California campaign (B. Silverman, interview, May 5, 1997).^{10 - 11} Youth remembered "Nicotine Soundbites" (Figure 2) and when they were asked to visualize the tobacco industry, they described a negative image of the chief executive officers in the tobacco industry (B. Silverman, interview, May 5, 1997).

Massachusetts found that youth do not believe industry claims that nicotine is not addictive and people smoke for pleasure. They responded well to a positioning statement that both refuted the industry's arguments and cited the statistic that "74% of all smokers aged 12 to 18 say they wish they could quit but can't because they are addicted to the nicotine."¹⁴ However, when youth were asked to rank the effectiveness of various positioning statements, this one was not rated as highly as others, such as, "The tobacco industry trades human lives for profit."¹⁴

The cessation strategy tries to convince current smokers to quit smoking, either on their own or with help from a local cessation program, by providing smokers with a rationale for quitting, such as health, money, and family.¹⁵ The ads may also emphasize that quitting is not simple and that many people are only successful after several attempts to quit. Michigan has used the slogan "Don't Quit Quitting" in several of its advertisements, presenting smokers with both reasons to quit and information about different ways to quit.⁵¹ California's advertisement, "Quitting Takes Practice," depicts a frustrated smoker who has tried to quit smoking but failed. He is standing in front of a steep incline, which represents his inability to quit, and when the announcer tells the smoker to take his time and try to quit slowly, for good, the ramp gradually flattens. This advertisement became part of local quit smoking smoking programs throughout California.⁵⁰ In early 1991, when California's advertising focused heavily on cessation, calls to local health departments and toll-free quit lines increased dramatically.⁴³

Youth access advertisements depict how easily youth can obtain cigarettes from vending machines, stores, parents, or siblings. These advertisements try to counter tobacco industry recruitment of underage smokers by convincing adults to reduce youth access to tobacco products. A California advertisement, "Vending Machine," shows children asking for various types of cigarettes. As the last child buys his cigarettes, the camera shows him walking away from a vending machine. This advertisement was tested with teenagers only, and many were concerned about the ease with which children can obtain cigarettes. After the advertisement aired, adults reacted angrily to how easily children can get cigarettes. However, some of the respondents saw the issue as a vending machine problem, not an underage smoking problem.¹⁸ In addition, adults in 2 other studies voiced concern that there was no way to keep youth from obtaining cigarettes if they really wanted them.^{9 - 10}

The purpose of this strategy is to counter the industry's portrayal of smoking as glamorous, attractive, and healthy by showing the immediate health and cosmetic effects of smoking and how uncool smoking actually is. These advertisements often use gross humor to focus on the physical consequences of smoking (such as yellow teeth and fingers, headaches, and unpleasant smelling clothes and hair). California's "Clifford" advertisements humorously portray a high school student in a smoke-filled school bathroom discussing the consequences of smoking, such as premature aging and bad breath. Adolescents understood the messages of these advertisements and liked the humor. In an unintended consequence, however, they found some of the advertisements unbelievable because of their use of humor and exaggeration. Many had friends who smoked, and these friends didn't look any different than nonsmokers.¹³

The evidence on the effectiveness of these advertisements is mixed. In their controlled study of smoking in movies, Pechmann and Shih⁵² found that students shown 1 of the Clifford advertisements before watching a movie with attractive lead actors who smoked did not find the smoking appealing. In another controlled study, Pechmann and Ratneshwar⁵³ showed short-term effects of magazine advertisements to nonsmoking seventh graders and evaluated their perceptions of peers who smoked, finding that students who saw these advertisements rated smokers lower on such traits as common sense, personal appeal, glamour, and maturity.

One California focus group, however, found that many youth reject the emphasis on short-term effects because it "trivializes the seriousness of smoking."⁸ Others believe that such problems would only occur among heavy smokers, many years in the future. These youth think that they do not smoke enough for the negative effects to occur, and that they will quit before they are harmed. The short-term effects advertisements, therefore, are not conveying the message they are supposed to—smoking has immediate, negative consequences.

This strategy involves detailing the potential long-term health consequences of smoking, such as lung cancer and emphysema. This strategy is one of the least effective, especially with youth, for 2 reasons: most already know the potential health hazards associated with smoking, which are printed on the cigarette packs themselves, and young people live in the present and believe they are invulnerable. Unless adolescents have personal experience with these smoking-related diseases (for instance, they know someone who has such an illness), they do not feel threatened.¹³ Reminding youth of the long-term health consequences of smoking and telling them that they will die prematurely in 20 years do not have much impact.^{48 ,54 - 56}

This strategy tries to convince smokers and those contemplating smoking that they will be undesirable if they smoke. The advertisements point out that the majority of people don't smoke and find smoking socially unacceptable. Like the short-term effects strategy, romantic rejection tries to counter industry advertising portraying smokers as sexy and alluring. Adult smokers found this message offensive and noted that their own personal experiences differed from the situation portrayed in the advertising. Others saw the message as a sign of the advertisers' superficiality.^{11 - 12} Young smokers responded similarly.^{8 ,11} For young nonsmokers, however, a person's smoking status was only relevant if that person was unappealing. The respondents were willing to overlook the smoking if the person was otherwise desirable.⁸

From 1989 through 1996, California per capita consumption of cigarettes fell 1.93 (±0.21) (SE) packs per year faster than the rest of the United States, excluding Massachusetts. California spent an average of $0.50 (in 1996 dollars) per capita per year on the media program during this time. Dividing the rate of decline in consumption by the average annual per capita media expenditure of $0.50 yields an estimate of a fall of 3.9 packs per capita per year for each per capita dollar spent on the media campaign.

Likewise, for Massachusetts, between 1993 and 1996 per capita consumption of cigarettes fell 1.28 (±0.90) (SE) packs per year faster than the rest of the United States, excluding California. Dividing this decline in consumption by the average annual per capita media expenditure⁴ of $2.42 yields an estimate of a fall of 0.5 pack per capita per year for each per capita dollar spent on the media.

Thus, the California media campaign appears to be about a factor of 7 more cost-effective than the Massachusetts campaign. These results are consistent whether one uses media campaign or total program spending. Similar results were obtained using several other analytical models (available from the authors).

The type and target of antitobacco advertising messages matter (Table 1). Industry manipulation and secondhand smoke are the most effective strategies for reaching all audiences. The industry manipulation strategy denormalizes smoking and delegitimizes the tobacco industry. By showing to what lengths the industry will go to recruit and keep new smokers, these advertisements have sparked interest in smoking and opened people's minds to other antitobacco messages. Secondhand smoke advertisements also denormalize smoking and heighten interest about smoking among both smokers and nonsmokers. The addiction and cessation messages can be effective, but work best when used in combination with, or rotated with, the more powerful industry manipulation and secondhand smoke strategies.

Youth access, short-term effects, long-term health effects, and romantic rejection all have limited effectiveness. Youth access advertisements can send a mixed message to youth by showing them how to obtain cigarettes and can reinforce the tobacco industry's advertising by portraying smoking as an adult activity.⁵⁷ Both the short-term effects and long-term health effects strategies can lose their impact among youth who believe that they do not smoke enough to suffer any negative consequences or that they will quit smoking before the cigarettes harm them. The romantic rejection strategy fails to have a significant impact because both adult and youth smokers find the message offensive, while youth nonsmokers see a person's smoking status as irrelevant if the person is attractive.

The tobacco industry clearly understands the power of antitobacco advertising and works to limit its effectiveness. An RJ Reynolds document,⁵⁸ which discusses the industry's reaction to California's aggressive campaign, states that "the California campaign, and those like it, represents a very real threat to the intermediate term. . . . Impact on self-esteem, social acceptance and smoking utility will ultimately influence business" [emphasis in original]. This same document summarized research on various California advertising strategies. Industry manipulation advertisements were generally seen as "believable, even among many smokers" and such an advertisement "presents risk of demotivating smokers." The industry also developed a sophisticated strategy, including working through other organizations, in an effort to eliminate funding for the media campaign or reduce its aggressive tone.^{59 - 60} Continuing industry awareness of the effectiveness of the power of antiindustry messages is reflected in the settlement of Florida's Medicaid suit against the tobacco industry, which provided $200 million for an antitobacco campaign, but explicitly prohibited attacking specific tobacco companies or advertisements.⁶¹

The comparison of the cost-effectiveness of the California and Massachusetts campaigns suggests that California's more aggressive approach is about an order of magnitude more cost-effective than Massachusetts' approach, using as an outcome measure the rate of decline in per capita consumption. In drawing this conclusion, however, it is important to remember that the statistical model we used is relatively simple. It does not model price changes, income distributions, age distributions, underlying economic conditions, or the precise patterns of air time purchased in the 2 states. This simplicity is dictated by the relatively limited data set that are available (annual data from The Tax Burden on Tobacco²⁰ ) and annual appropriations for the media campaigns.^{4 ,21 - 22} The fact that the differences hold up under a variety of different modeling approaches, combined with the large differences we find, suggests that there is a real difference in the cost-effectiveness of the 2 campaigns.

Two strategies are effective in reaching all audiences: industry manipulation and secondhand smoke. The addiction and cessation strategies can also be effective. Advertisers should refrain from spots that focus on youth access, short-term effects, long-term health effects of smoking (for youth), and romantic rejection. To compete with tobacco industry advertising, antitobacco advertisements need to be ambitious, hard-hitting, explicit, and in-your-face. Unless the advertisements grab and hold people's attention, their messages will be lost amid other advertising. Advertisements must clearly refer to the tobacco industry, rather than to "they" or "them."

A strong media campaign is a key element of any tobacco control effort. The tobacco industry has consistently tried to limit the size and scope of these campaigns by focusing them on children.⁶² Perhaps because of the power of paid media to shape public attitudes toward tobacco and the tobacco industry, public health advocates need to be prepared for a continuous battle to defend the existence and quality of the media.^{34 ,63 - 65}