Advertising in the mass media allows public health campaigners to repeatedly expose the population to messages about the negative effects of tobacco use and the benefits of quitting. This exposure occurs incidentally during routine media use,1 so it provides a way of reaching not just individuals who are already thinking about quitting but also those who are not yet at this point. Public education campaigns can directly influence individual decision-making about quitting as smokers view or hear campaign messages, gain new insights, and reflect on the implications for their own lives. Campaigns can also operate indirectly by influencing interpersonal discussion about tobacco use within family and friendship networks or by influencing social norms. Changes in broader social norms can, in turn, create a more favourable climate for tobacco control policies (such as smokefree areas or tobacco tax increases) thereby increasing the likelihood of policy adoption, and these policies might also prompt further quit attempts or prevent uptake. 2
If they are to be effective, public education campaigns must be noticed (using appropriate media channels and placement to reach the target group), persuasive (experienced by the target group as engaging, relevant and/or emotionally affecting), and remembered (seen often enough for them to be recalled and acted upon). Campaigns generally have effects that are small in size, but, because they reach large numbers of individuals within populations, the degree of change can be of high practical significance. 3
Several notable reviews examining the effect of public education campaigns on smoking behaviour, 2, 4-6 youth smoking, 7, 8 and across health behaviours 9, 10 have helped to consolidate knowledge from research and practice. These comprehensive reviews reach a similar conclusion: mass media campaigns can positively change smoking behaviour in adults and youth. Where they differ is in the strength of their conclusions. The most comprehensive review on the effects of media campaigns on smoking behaviour was a major monograph published by the US National Cancer Institute in 2008. 5 This 650-page document systematically reviewed scientific studies conducted in numerous countries across several decades. 11 This included scientific literature relevant to the role of advertising in forming opinions about and attitudes to smoking, 12 sources of evidence about the effects of advertising by tobacco companies and by health campaigns, 13 and the portrayal of smoking in the media and popular culture. 14 On the basis of its broad-ranging expert analysis of direct and indirect evidence from within health, communication and broader psychological fields, it found that there was sufficient evidence to conclude that anti-smoking campaigns can reduce population smoking and that both mass media advertising of tobacco products and portrayal of smoking in the media also can exert causal effects on population smoking among both youth and adults. 15 A 2012 review by Durkin and colleagues 2 provides the strongest evidence of effectiveness of public education campaigns in reducing smoking; confidence grows as study design improves and data accumulate from multiple campaigns in different places over many years. The content of this chapter draws largely on the content and conclusions of this 2012 review, supplemented with further information from later reviews and major population-level studies.
14.4.1 Population-level effectiveness
Isolating the effects of campaigns from other policies within a comprehensive tobacco control program is extremely challenging. 4 Most population-level studies rely on pre- and post- campaign comparisons of outcomes such as smoking prevalence, tobacco consumption, quit attempts and intentions within the same region, or comparisons to similar regions that were not exposed to the campaign.
With its long history of running mass media-led campaigns to discourage smoking, Australia’s contribution to the evidence base about population-level effectiveness has been an important one. Beginning with the ‘Quit. For Life’ Sydney–Melbourne pre–post trial in the early 1980s (see Section 14.0), 1 6 , 17 then the National Tobacco Campaign in the late 1990s, 18-20 and with successful exports of campaigns such as ‘Bubblewrap’ and ‘Sponge’, 21 Australia has developed an international reputation for excellence in developing and evaluating anti-smoking television advertising. Since the 2000s, Australian researchers have focused on identifying campaign elements that optimise effectiveness, such as themes and content, 18, 22-24 media type, 25, 26 media placement, 27, 28 and the intensity and duration of campaigns. 29-33
A growing body of research confirms that, to have an effect at a population level, campaigns need to be broadcast at sufficient intensity and duration (see Section 14.4.3). Public education campaigns have relatively short-term effects on quitting behaviour; evidence shows that repeated cycles of broadcasting are required to achieve sustainable population changes in smoking behaviour. Jurisdictions should aim for high reach and consistent exposure over time and favour messages that contain effective messages about the negative health effects of smoking (see Section 14.4.2).
14.4.1.1 Population-level effectiveness: adults
One of the first large-scale examinations of the effectiveness of a national tobacco public education campaign, Australia’s 1997 National Tobacco Campaign, found that exposure to the campaign increased quitting intentions among smokers and prevented relapse among former smokers. 34 Evaluation studies were able to demonstrate that the first phase of the National Tobacco Campaign achieved a reduction of 190,000 smokers. 35, 36 Since then, a large body of international research has examined the effectiveness of anti-smoking public education campaigns. This research differs widely in terms of the measures of effectiveness used—smoking prevalence, tobacco consumption, quit attempts, quitting intentions, and quitting-related behaviours such as engaging with quitting services—and the methods used. The studies reviewed here examine the effects of adult-targeted anti-smoking campaigns on adult smoking behaviour.
A considerable body of research since the 2000s has related commercial TV ratings data (that is, potential exposure to the campaign among the population) with changes in individual-level smoking behaviour. This research has consistently demonstrated numerous desirable outcomes. In a prospective population study in Massachusetts, greater population exposure to televised media campaigns among adult smokers was associated with a higher likelihood of quitting at a two-year follow-up. 37 Higher levels of population exposure have also been associated with increased calls to quitlines, 23, 38-44 , greater campaign website visits, 45 increased quitting behaviours among adults 37, 46-50 and decreased adult smoking prevalence. 31 A 2012 study compared measures of self-reported recall of a campaign and potential exposure to that campaign, and found both exposure measures were significantly associated with the likelihood of making a quit attempt. 51
More recently, researchers have used time series analyses to examine the effects of public education campaign exposure on smoking prevalence and other quitting behaviours over long time periods. A time series analysis of monthly smoking prevalence in Australia over 11 years to 2006 found that greater population exposure to televised media campaign advertising was associated with a faster decline in adult smoking prevalence, after adjusting for variation in tobacco prices, smokefree restaurant laws, tobacco marketing restrictions and availability of smoking cessation products. 31 These findings were supported by a follow-up analysis from 2001–11. 52
Further evidence for the effectiveness of public education campaigns comes from economic evaluations. 53, 54 A 2015 systematic review identified ten cost-effectiveness studies of both adult and youth campaigns. 55 All ten studies noted favourable cost-effectiveness, and were of sufficient quality (though variable methodology) to conclude that anti-smoking public education campaigns offer good value for money at a population level. These studies were all from high-income countries, and the authors have cautioned that the findings are not necessarily generalisable to low and middle- income countries.
Indeed, most of the research on the effectiveness of public education to discourage smoking comes from high-income countries such as Australia, the US and UK. Using data from the Global Adult Tobacco Survey, a 2013 study examined the effects of noticing anti-smoking public education campaigns across seventeen countries, many of which were low- and middle- income. 56 Intentions to quit were significantly associated with campaign awareness on one type of media channel (for example, television, radio, or print) in nine countries, while campaign awareness across multiple channels was associated with greater quitting intentions in fourteen countries.
Several systematic reviews have also synthesised available international population-level evidence. These reviews are weighted heavily towards studies using conventional control-group study designs. In 2008, the National Cancer Institute 5 identified 11 eligible studies that examined government-funded public education campaigns. All eleven demonstrated positive effects on smoking prevalence and/or per capita tobacco consumption.
A later systematic review by the Cochrane Collaboration in 2017 examined 11 public education campaigns to assess associated adult smoking behaviour changes. 4 This review concluded that comprehensive tobacco-control programs that included mass media campaigns could positively change smoking behaviour in adults, but noted that studies were variable in their methodological design and quality. The authors further concluded that intensity and duration of mass media campaigns might influence effectiveness, however, most studies were not of sufficient duration following exposure to assess whether these changes are enduring. The review found no consistent differences in campaign effectiveness by demographic factors such as age, gender, or socio-economic status.
Summary
- Adult-focused mass media campaigns have consistently been shown to be effective in reducing smoking prevalence and prompting quitting behaviours.
14.4.1.2 Population-level effectiveness: youth
Whether tobacco-control campaigns should primarily focus on youth (because most people start smoking before 18 years of age) or on adults (who make up the vast majority of current smokers) has previously been the focus of extensive debate. 57, 58 Comprehensive reviews of the evidence conclude that there is no need to choose between the two options: effective adult-targeted campaigns are effective in communicating with youth as with adults. 8, 15, 59 This may be because they change broader social norms about smoking, including the perception that smoking is common and accepted, while also educating young viewers about the harms of smoking. 10, 57, 59 For example, evaluations of the effects of the 2012 national adult-targeted ‘Tips from Former Smokers’ campaign in the US have shown that, among teens, self-reported exposure to and awareness of the campaign was high, with exposure highest among smokers or at-risk never smokers. 60 A subsequent study found exposure was associated with greater intentions to quit among adolescent smokers and lower smoking susceptibility among non-smokers. 61
Reviews of early field trials provided some support for the effectiveness of media interventions combined with school programs, 8 and cumulative population-based research on adult- and youth-focused mass media campaigns from the 2000s, has provided convincing evidence that these campaigns can reduce youth smoking. 15, 60-66 The context or circumstances in which youth are exposed to messages also seems to be important in that discussions about such advertising can enhance the impact of anti-smoking messages on quitting intentions and quit attempts in adolescents. 27, 67 Dunlop 68 found that for smokers, talking about the ‘truth’ campaign was associated with beliefs, attitudes, and intentions in greater agreement with campaign messages, but only if associated with positive campaign evaluation. For smokers who evaluate a campaign negatively, talking about that campaign was associated with beliefs and attitudes counter to the campaign messages. However, a 2016 study from the US suggested that adult or general-targeted public education campaigns were not associated with youth smoking prevalence between 1999–2005, while particular types youth-targeted campaigns were associated with reduced youth smoking (see also Section 14.4.2.2). 69
In 2012 the US Surgeon General released a report on reducing tobacco use among youth and young adults. This included a systemic review that examined 17 reviews of mass media campaigns and youth smoking published up to the mid-2008. 8 This review concluded that there was sufficient evidence ‘to infer a causal relationship between adequately funded anti-smoking media campaigns and a reduced prevalence of smoking among youth’. 8, p.691 Further, adult-targeted campaigns were linked to reduced smoking among youth. A subsequent review of research published between 2008 and 2013 identified relevant 21 studies of adult- and youth-targeted public education campaigns. 59 Fourteen of these studies reported positive campaign effects on youth smoking behaviours and intentions, and no negative effects were found in any studies.
The 2017 systematic review by the Cochrane Collaboration identified eight mass campaigns targeted at youth aged under 25 years that were evaluated using controlled trials or time-series analyses. 7 Methodological approaches and quality of these studies was variable. The public education campaigns in question also varied, including those delivered via traditional media channels such as television and radio but also school-based campaign interventions. Three of the eight studies showed significant reductions in youth smoking associated with campaign exposure. Characteristics of these effective campaigns (but also some of those campaigns that showed no effect) included extended duration and greater intensity of exposure, for example, at least four weeks of campaign activity across multiple channels. Combined approaches, such as school-based programs paired with media messages, were also associated with greater campaign success. 7
Youth-targeted campaigns can also be effective in maintaining reductions in youth smoking. 69-72 Further studies examining the exposure of youth to anti-smoking advertising have shown that higher levels of exposure are associated with improved smoking-related attitudes, beliefs and behaviours among youth. 62, 70, 73, 74 Exposure to the first youth-targeted national public education campaign funded by the US government, ‘The Real Cost’, was associated with significantly lower likelihood of initiating smoking than those with little or no exposure to the campaign. 75, 76 Economic evaluations of this and a previous US youth-targeted campaign have shown these campaigns to be highly cost-effective, with savings primarily achieved through preventing smoking initiation among teens. 77, 78 However, a US study among a somewhat older population of young adults aged 20–30 years between 2001–08 found that increased potential anti-smoking advertising exposure over a two-year period was associated with increased odds of quitting smoking, but not related to smoking uptake. 79
Summary
- Adult- and youth-targeted public education campaigns can positively influence youth smoking, in terms of preventing initiation and maintaining reductions in prevalence.
14.4.2 Campaign content and theme
Anti-smoking public education campaigns vary greatly in their message content and style. 5 Experimental research on information processing supports the hypothesis that advertisements that evoke high arousal will receive greater viewer attention and will be remembered more readily than those that do not. 80 Further, negative content tends to produce higher levels of arousal than does positive content.
Recent work has tended to first categorise anti-smoking messages by purpose in order to identify the most effective message strategy. 81, 82 Several categories provide useful differentiation: ‘why to quit’ messages encourage cessation by advising smokers why they should quit smoking, usually through depicting the negative health effects of smoking; ‘how to quit’ message promote cessation by educating smokers how to quit or depicting those who have quit successfully; ‘keep trying to quit’ messages focus specifically on encouraging smokers to persist in their efforts to quit smoking; and ‘secondhand smoke health effects’ messages can include some elements of ‘why to quit’ and often depict the effects of secondhand smoke on children. Some messages reveal ‘tobacco industry manipulation’ or may be used to promote ‘public awareness' of new tobacco control policies. Advertisements may also be differentiated by the level and type of emotion elicited, and recent research has also examined message execution features, such as graphic imagery, simulated health effects, or personal testimonials.
14.4.2.1 Campaign content and theme: adults
The 2008 National Cancer Institute review demonstrated that messages that elicited negative emotions by describing the serious health consequences of smoking scored higher on ratings on perceived effectiveness 83 and memorability 84 and were more likely to be recalled by recent quitters who believed that anti-smoking advertisements had contributed to their quit attempt. 85 Exposure to messages with high levels of emotion and/or a personal testimonial (typically negative health effects advertisements) significantly increased the likelihood of being quit at two-year follow-up, whereas messages that were low in emotion and did not feature a personal testimonial were not related to the likelihood of quitting. 37 Robust evidence continues to find better performance for advertisements that arouse strong negative emotions than for those that do not. 5, 24, 37, 83, 86-89
Several large US studies have found that exposure to emotional or graphic ‘why to quit’ advertisements was associated with an increase in quit attempts at follow-up, while no effects were found for exposure to ‘how to quit’ advertisements, or those lacking strong emotions or graphic content. 51, 88 Similar Australian research found greater exposure to fear-evoking messages was associated with higher levels of quit attempts among all smokers, while the effects of ads that evoked multiple negative emotions (e.g. fear and sadness) were more variable. The effect of positive emotion ads, when aired in combination with high-fear ads, varied by smoker socio-economic status. 24
The effect on quitline call volumes generated by different message types also provides useful insights into the potential effectiveness of different styles of advertising. Population-level studies from Australia and the US have found ‘why to quit’ testimonial advertisements to be the strongest driver of quitline call volumes, 23 and most cost-effective at increasing quitline calls. 43 Other studies have shown ‘why to quit’ ads with graphic imagery prompted a higher proportion of smokers who were not ready to quit and had lower confidence in their ability to quit to call a quitline than other types of ads. 90 The authors concluded that these ads were prompting smokers who would otherwise not have called the quitline to do so.
In contrast, a UK study comparing exposure to positive versus negative emotion-eliciting campaigns found that quitline calls steadily increased as exposure to positive campaigns increased from 0 to 400 gross ratings points (GRPs). Call volumes increased in response to negative campaigns only after a threshold of 400 GRPs per month was achieved. 91 Notably, both types of campaigns positively influenced quitline calls, and both types of campaigns had strongest effects once the 400 GRPs per month threshold was reached. Positive-emotion advertising was also shown to have positive effects on tobacco use in a large UK study of 60,000 English participants over seven years. A 400 GRP increase in positive emotion advertising was associated with 7% lower likelihood of smoking one month later, while a 400 GRP increase in negative emotion advertising was linked to a 4% decline two months later. Negative, but not positive, emotion advertising was also associated with declines in consumption. 89
Advertisements evoking moderate or high levels of emotion may also be more likely to stimulate interpersonal pressure from family and friends on smokers to quit, and this pressure can increase the likelihood of quit attempts. Interpersonal pressure generated by highly and moderately emotional advertisements has also been positively associated with salient quitting thoughts. 2, 92
Message characteristics can also affect the adaptability of campaigns for audiences beyond those originally targeted when the ad was made. This was examined in a pre-testing study of anti-smoking television advertisement across ten low- and middle- income countries. 93 Ads that used graphic imagery were rated by smokers as most acceptable and had higher perceived effectiveness, followed by ads that simulated health effects, while personal testimonials and non-health effects ads performed weakest. Smokers’ ratings of graphic and simulated health effects ads were also less likely to vary by demographic factors such as age, gender, SES, and country. 94 The authors suggested that advertisements that do not feature identifiable cultural differences and focus on health effects are most suitable for adaptation and use in other countries.
Summary
- Negative health effects advertisements with high levels of emotion, including those with graphic imagery and/or personal testimonials, have the greatest potential impact on adult smokers.
14.4.2.2 Campaign content and theme: youth
Research and reviews of youth campaigns have begun to focus on which message characteristics work best, and the ideal level of exposure and which types of youth are most and least affected by mass media anti-smoking campaigns—see also Section 14.4.3.2.
Biener and colleagues 86 noted the types of advertisements that young people perceived as most effective were those that evoked a strong negative emotion such as fear or sadness and conveyed a thought-provoking and believable message about the serious long-term consequences of smoking. These were perceived as more effective than advertisements that were designed as humorous or entertaining, or normative advertisements that had low emotional content or generated low cognitive engagement. These perceptions were consistent between teenage boys and girls.
As noted in Section 14.41.2 population-based research also indicates that recall of campaign messages has been associated with reduced smoking behaviour in youth 60, 61, 63, 64 and a study by Biener and colleagues 87 of adolescents aged 12–17 years found that the level of emotional intensity of the advertisement was a significant predictor of recall.
While the ‘tobacco industry manipulation’ approach has not been extensively employed in Australia there is evidence of its effective use in the US. Evaluation of the American Legacy Foundation’s national ‘truth’ campaign, a strong ‘tobacco industry manipulation’ campaign, found significantly changed attitudes towards smoking among young people in a relatively short period. 95 The national campaign delivered stark facts about the tobacco industry and its marketing practices, and built on positive results achieved by the campaign in Florida. 96 Positive associations with the ‘truth’ campaign brand were associated with significantly lower past 30-day smoking and higher intentions to quit among youth aged 15–21 years. 97
The review of mass media campaigns in the 2012 US Surgeon General report on youth smoking noted that messages that emphasise the negative health effects of smoking and arouse negative emotions, and messages that focus on the tobacco industry’s deceptive practices and tactics, were particularly effective in reducing youth smoking. 8 Similarly, six out of ten studies in the 2008–2013 review that directly compared the effects of message themes among youth found that negative health effects campaigns were more effective, and that messages that elicited negative emotional responses were also generally more effective. 59
Many studies that have examined the effects of message characteristics on youth smoking have relied on smokers’ perceptions and intentions as outcome measures, not smoking behaviours. 98 However, a US study linking advertising volume with youth smoking prevalence found an overall positive effect for youth-targeted advertising, which was driven by campaigns depicting the health effects of smoking and anti-industry messages. The executional style of the ads (graphic imagery or personal testimonials) was not found to be related to smoking prevalence, nor were adult-targeted campaigns. 69
Summary
- Advertisements that evoke high levels of emotion by depicting the serious long-term consequences of smoking or by highlighting tobacco industry deception have been found to be most effective for youth.
14.4.3 Campaign exposure
Advertising research reveals that the effects of advertising linger over the days and weeks after broadcast ends. 99, 100 However, although people may recall anti-tobacco advertisements long after they are discontinued, behavioural effects decay relatively quickly. 29, 32, 62, 99, 101-103 This means that campaign advertising only exerts effects on smoking behaviour during periods of advertising activity, and effects rapidly diminish once advertising is withdrawn. This is why advertising is sometimes referred to as acting ‘like a spring, rather than a screw ’ 29, 104 in pushing down smoking prevalence. Identifying the level and duration of media placement required to achieve optimum population response is critical for campaign planners and funding agencies, informing decisions about the campaign placement strategy and the investment that will most efficiently result in behavioural changes.
14.4.3.1 Campaign exposure: adults
Studies that relate exposure to commercial television with data on smoking prevalence 31, 52 or other smoking-related outcomes 37, 47, 51, 62, 73, 74 have consistently demonstrated that an individual’s potential exposure to televised anti-smoking campaigns is related to a range of positive outcomes. In recent studies researchers have attempted to identify the optimum level and duration of campaign exposure, most commonly expressed as gross ratings points (GRPs) or target audience ratings points (TARPs). [1]
A time-series analysis of Australian smoking prevalence and monthly tobacco control GRPs provided some early guidance. 31 This study demonstrated that an increase in tobacco control GRPs was associated with an accelerated rate of decline in smoking prevalence two months later. It was estimated that to achieve a temporary 0.3 percentage-point decline in smoking prevalence, a 390-monthly GRP increase two months earlier would be required (equating to approximately one ad exposure per person in the population per week). Further, a cohort study 32 and a serial cross-sectional analysis of the aggregate effects of tobacco-control campaign exposure 30 concluded that advertising was associated with short-term increases in the likelihood of smokers making a quit attempt. The cross-sectional study found the association between increased advertising exposure and quit attempts was detectable only after an average of nine weeks of cumulative exposure. Findings suggested that exposures of more than 100 TARPs per week (equating to approximately one or two exposures per week) produce the greatest effects on quitting thoughts and quit attempts. 30
Similar levels of media exposure have been shown to be important for sustaining quit attempts. 103 A dose of approximately 1000 TARPs per quarter has been associated with a 50% higher likelihood of staying quit at 3-month follow-up. Each additional 100 TARPs per quarter increased odds of staying quit by 5%. An evaluation of the US ‘Tips from Former Smokers’ campaign concluded that exposure of 800–1000 TARPs per quarter was sufficient to increase quit attempt rates in the general population. 46 Also, as described in earlier sections, several studies have found 400 GRPs/TARPs per month to be a threshold for detecting population-level effects in smoking behaviours. 89, 91 Evidence of a necessary minimum threshold of GRPs for eliciting quitting-related behaviour was found in a US study, where weekly low-dose GRPs were ineffective at stimulating quitline calls, while 250 GRPs per week was associated with significant increases in call volumes. 105
Several studies from the UK have used variations in public education campaign expenditure as a measure of potential exposure. A time series analysis observed a marked decline in use of smoking cessation support services following a freeze on public education campaign spending. 106 Another UK time series analysis of expenditure on public education campaigns found mixed effects: a 10% increase in monthly expenditure was linked to greater quitting success rates, but no clear associations were found for smoking prevalence or quit rates. 107
McAfee and colleagues 108 examined the effects of boosting the dose of an anti-smoking campaign. During a 3-month period of the national ‘Tips from Former Smokers’ campaign, the media buy in 67 media markets was tripled on local media channels. Smoking behaviours in these high-dose markets were then compared to 123 standard-dose markets which received only the national buy of 761 GRPs (on average) for the three-month campaign. Compared to the standard-dose markets, campaign exposure was 16% higher and quit attempt rates were 11% higher in the high-exposure markets. However, another study comparing differing levels of campaign exposure (measured through GRPs) found that higher doses of the campaign were positively associated with quit attempts only in states with already high levels of tobacco control funding. 50
Emerging evidence from Australia has examined levels of potential campaign exposure and quit attempts across socio-economic group. This research suggests that in order to reduce inequity in smoking prevalence between lower and higher socioeconomic groups, exposure to at least four public education television campaigns is needed. At these levels, rates of quit attempts are greater among lower SES groups than higher SES groups. Fewer than two exposures per month can increase inequities (where higher SES groups make more quit attempts than lower SES groups), while 2–3 campaign exposures acts to maintain disparities between SES groups. 109
Summary
- Recognising that smokers are cycling through stages of readiness to quit at different times throughout the year, it is important that sufficient funding be provided to enable repeated cycles of advertising throughout the year to sustain high levels of quit attempts. The consensus of expert evaluators of anti-smoking media campaigns is that a dose 800–1000 target audience ratings points per quarter is necessary to achieve this. Emerging evidence suggests that at least four campaign exposures per month is needed to reduce disparities in smoking rates between lower and higher socioeconomic groups.
14.4.3.2 Campaign exposure: youth
Several studies of youth campaigns in the US provide more insight into the effectiveness of different levels and duration of media exposure on youth smoking. The first national study of the effects of state-sponsored campaigns across the US, found that it was important that youth were exposed to an anti-tobacco advertisement at a minimal mean exposure rate of at least one advertisement every four months to maintain anti-smoking attitudes and beliefs, and to reduce smoking behaviour. 62
A dose-response relationship between exposure to the ‘truth’ campaign and youth smoking in the last 30 days has also been observed. 73 However, after reaching very high levels of exposure (over 10,000 GRPs over a two-year period; an average of four exposures per month per individual) no additional gain in reducing youth smoking were found. A longitudinal study of young adults aged 20–30 years found that exposure to the equivalent of about 100–150 ads over a two-year period had the strongest association with quitting among all smokers compared to lower and higher levels of exposure. However, among daily smokers, gains in reduced consumption and quitting behaviour continued as advertising exposure increased. 79
Ceasing anti-smoking advertising can cause smoking behaviour to resume. Several studies have demonstrated a rapid decay of campaign effects when anti-smoking advertising is suspended. When funding was reduced for the youth-focused Florida ‘truth’ campaign in 1999, immediate reductions in anti-smoking beliefs were observed and smoking intentions began to increase. 99, 110 A study examining the impact of de-funding the Minnesota youth tobacco-use prevention program 101 found that a range of measures of susceptibility to smoking among youth—including openness to smoking, and beliefs, attitudes and intentions to smoke—consistently increased following the de-funding of the campaign.
A study of Australian adolescent smoking prevalence examined the effects of past three- and 12-month media exposure triennial national survey data from 1993 to 2008. 33 Exposure to at least 400 TARPs per month was associated with decreased smoking prevalence. However, at this level, a constant dose of monthly TARPs was needed to reduce smoking, while higher doses (average exposure of at least 800 TARPs per month) successfully reduced smoking prevalence with intermittent exposure. The findings also indicated that low levels of cumulative 12-month exposure (100–399 TARPs per month) were associated with increases in adolescent smoking. The authors recommended that cumulative exposure to at least 5800 TARPs over a 12-month period was necessary to reduce adolescent smoking prevalence. 33
The 2012 US Surgeon General report on reducing youth smoking 8 noted a dose-response relationship between campaign exposure and youth smoking, and observed that reductions in funding for public education campaigns were associated with diminished declines in youth smoking prevalence. The authors concluded that the effects of public education campaigns decrease rapidly when the campaign is off air. 8 Later evidence from Australian 33 and US 79 studies led the authors of the 2016 review to recommend that television public education campaigns should be aired at an aggregate intensity of 1200–1400 TARPs per quarter to achieve population-level reductions in youth smoking. This can be achieved through regular low-intensity activity, or shorter bursts of higher-intensity activity. 59
Summary
- Similar to adults, the impact of media on smoking behaviour among youth appears to be dose-related, requiring at least 400 TARPs per month to show population-level effects. Ceasing advertising can result in regression in youth smoking behaviour.
* Thank you to Dr Tom Carroll who wrote the first version of this document. Thank you also to Dr Sally Dunlop and Professor Melanie Wakefield for advice and comments on an earlier version this section, and to Dr Emily Brennan for assistance in identifying relevant research, and advice and feedback on the most recent update.