Objective The tobacco industry often relies on the assertion that smokers make ‘informed adult choices’. We tested this argument by exploring how young adults initiate smoking.
Methods Fifteen in-depth interviews with young adults who had started smoking since turning 18, the legal age of adulthood and tobacco purchase in New Zealand. We undertook a thematic analysis of the interview transcripts.
Results Although participants had a general awareness that smoking is harmful and knew some specific risks, they rarely saw these as personally relevant when they started smoking, and few had made a deliberate decision to smoke. Participants’ poor understanding of addiction meant most regarded smoking as a short-term phase they could stop at will. Initiation contexts discouraged the exercise of informed choice, as smoking onset often occurred when participants were influenced by alcohol or located in socially-pressured situations that fostered spur of the moment decisions.
Conclusions Young adults’ ability to exercise ‘informed choice’ at the time of smoking uptake is constrained by cognitive and contextual factors. We propose an updated informed choice framework that recognises these factors; we outline environmental changes that could make default adoption of smoking less common while promoting more ‘informed choices’.
- Public policy
- Priority/special populations
- Advertising and Promotion
Statistics from Altmetric.com
Tobacco companies have opposed policy interventions by arguing that smokers exercise ‘informed adult choice’ and start smoking after appraising the risks and benefits they may incur.1 By transferring responsibility for harm onto smokers themselves, tobacco companies reduce their potential liability and promote beliefs that tobacco control measures undermine consumers’ right to smoke.2 ,3
Given most adult smokers began smoking as adolescents, when by definition they are not making ‘adult’ choices, the industry's argument has obvious flaws. However, increasing smoking uptake also occurs among young adults in many countries. For example, in New Zealand, smoking prevalence remains high among those aged 18–25, despite reductions in adolescent smoking rates.4 ,5 This evidence suggests a need to test the industry's argument by exploring whether young adults make active, informed decisions to start smoking.
The beguiling simplicity of ‘informed choice’ arguments may overlook important cognitive and life stage factors that influence young adults’ decision-making. For example, optimism bias may lead them to believe they are exempt from harm or addiction,6 or able to control any risk they might face.7 Affect heuristics reduce concern about risks that occur in situations about which young people feel positive,8 while temporality biases privilege short-term outcomes and limit consideration of long-term effects.9 ,10
Behavioural contexts may also militate against ‘informed choice’. Where smoking prevalence is high, young adults may regard it as normal and associate it with desirable social benefits.11 Concurrent use of alcohol, a pervasive factor in smoking initiation among young adults, may further undermine risk–benefit assessments.12
Chapman and Liberman13 have proposed four levels of understanding that smokers should pass through before they can make an informed choice.
Having heard that smoking increases health risks.
Being aware that smoking causes specific diseases.
Accurately appreciating the meaning, severity and probabilities of developing diseases caused by tobacco use.
Personally accepting the risks inherent in levels 1–3 as relevant to themselves.
Chapman and Liberman proposed that this framework could identify smokers who qualify for a ‘license’ to smoke. However, this could arguably reinforce the industry's strategy of holding consumers responsible for harms.14 Nevertheless, the framework represents a useful starting point for considering what a truly informed choice to smoke might entail. We developed a study drawing on Chapman and Liberman's framework, and the cognitive biases and contextual factors outlined above, to investigate whether young adults make active, informed decisions when they begin to smoke.
We conducted in-depth interviews with 15 18–25-year-olds who had started smoking since turning 18. Participants were recruited using posters, online advertisements, word of mouth and public notices. Purposive selection ensured diversity by age, gender and smoking behaviour. Recruitment emails collected information about participants’ current smoking behaviour. We distributed advertisements to shopping areas and health providers where young people of varying backgrounds would go, as well as college campuses; most participants were current or recent students. The sample was therefore relatively well-educated. Category A ethics approval was obtained from the University of Otago's Human Ethics Committee. Table 1 contains details of participants’ demographics.
We used a semistructured interview protocol to explore participants’ smoking initiation and each component of Chapman and Liberman's conceptual framework. Specifically, we explored participants’ knowledge and risk acceptance when they began smoking, and their reflections on how informed their actions were. To extend the framework, we also examined participants’ understanding of addiction, explored evidence of risk assessments, and analysed the social and environmental contexts in which their smoking began. A full copy of the interview materials is provided as a online supplementary file.
Interviews were conducted in late 2012 and early 2013 and the average interview length was 50 min; each interview was audio recorded and then transcribed verbatim. We prepared vignettes of each participant to summarise their attributes, demeanour and key themes.
All interviewers made independent notes following the interviews. The first author undertook an intensive review of the interview transcripts and developed an initial descriptive classification, based on the research protocol.15 ,16 Subsequent iterative analyses identified patterns and themes within the transcripts beyond those addressed directly in the research questions. At a workshop convened to develop and cross-validate overall themes, all research team members discussed and agreed on the themes reported below.
We identified themes that corresponded generally to Chapman and Liberman's theoretical framework. We also identified themes that extended the original framework; these reflected understanding of addiction, the degree of active decision-making and cognitive and situational factors that influence risk acceptance and the likelihood of making informed choices.
Levels 1 and 2: awareness of general and specific health risks
All participants had received information about smoking and felt they had known it posed health risks:
I mean everyone sees those ads on TV… Everyone knows how bad it is for you… I couldn't see a person going what, smoking's bad for you? What is this?. (Harriet, 19, social smoker)
Most could identify some specific risks caused by smoking, particularly those currently promoted by on-pack pictorial warnings. Yet while they now knew of more risks, participants recalled they had less understanding of these when they started smoking:
Before I started smoking I wouldn't have a clue what gangrene toes look like… I probably wouldn't say I knew a lot of the consequences. (‘Bethany’, 20, daily smoker)
While their current perceptions and experiences of smoking varied, all participants had started smoking even though they understood at the time that smoking was generally risky.
Levels 3 and 4: personal acceptance and understanding the meaning of risk
Despite the ubiquity of information about health risks and participants’ general awareness that smoking is harmful, participants reported that messages about smoking often lacked impact and salience. Participants recounted their limited understanding of this information, and their indifference to it:
I had no clue when I started smoking. Other than the general ‘it's bad’ kind of message that I'd been given, I hadn't really listened and I didn't really care. (‘Bella’, 19, daily smoker)
The heuristics participants drew on affected whether and how they used risk information. Some expressed scepticism and attributed illnesses to other factors:
They go on about how it causes blindness and it causes cataracts but to me old age and cell degeneration can be just as big as a cause of, you know, any degenerative diseases. (Natasha, 24, recently quit daily smoker)
Participants often relied on ‘obscure-cause’ attributions and used evidence that non-smokers had developed smoking-related diseases, to question the validity of health risks:
They say it increases your chances of like cancer and things like that, but… people that don't smoke and don't drink at all can still get lung cancer. And I think if you, you know, add an extra four cigarettes or five cigarettes a week to that, I don't think it's going to… increase your chances of getting anything. (Harriet, 19, social smoker)
While many had heard the statistic that ‘50% of all smokers are killed by smoking’, few understood the severity of illnesses smokers suffered and most believed these only affected ‘hard-core’ smokers, who they saw as very different to themselves. By establishing self-exempting ‘risk thresholds’, participants, particularly ‘social’ smokers, diminished the dissonance their risk awareness created:
It's only mainly the pack-a-day smokers who kind of turn out like that, and not (people who have) just a couple while drinking. (‘Henry’, 20, social smoker)
However, those who had smoked for longer acknowledged they could no longer deny that risks applied to them:
I used to associate it really positively when I was younger, but the older I get, the more aware I become of, actually, this is going to have quite a severe impact on my health. (Kat, 23, daily smoker)
Most participants had also experienced or observed some shorter term risks, such as struggling to keep up when playing sport, noticing cosmetic effects on their skin, teeth or hands and waking up coughing. Risks experienced directly or vicariously had greater relevance, and were more likely to elicit concern and avoidance actions:
That's the one thing I've always said, that as soon as my fingers start changing colour, then I'll stop. (‘Cameron’, 22, daily smoker)
Despite not considering the risk of serious harm when they started smoking; subsequent experiences had led some to feel more vulnerable:
You don't really think about (the risk) until something bad happens. So it's not really an informed adult choice until you've actually had to feel something like that yourself. (Luke, 20, social smoker)
Luke summarises a conundrum central to this study: can informed choices about smoking be made by people who have no experience of the future harms they may face?
Additional themes: factors that complicate the exercise of informed choices
Degree of active decision-making
Although ‘informed choice’ implies an active and conscious decision, participants reported acting impulsively when they started smoking:
When I first started having smokes like every now and then I didn't really… well I still don't really think it through or anything. Just… yeah, just do it I guess. (Ethan, 20, social smoker)
Few had thought seriously about how smoking could affect their health when they started smoking:
Maybe it's just because I was young and I didn't… care about it, maybe I didn't recognise the risks of it. (Cameron, 22, daily smoker)
The realisation they had become smokers occurred insidiously and few reported having made an active decision to become a smoker. The transition between identifying as someone who occasionally smoked and as a ‘smoker’ rarely involved conscious consideration:
Nah, I don't think it was really a tipping point for me… it was more so that, it just like, slowly happened. Like you'd have one, and then you'd have two or three… then so far down the road I began thinking oh yeah, well I may as well just think of myself as a smoker. (Ben, 22, daily smoker)
A minority reported having made a deliberate choice to smoke. However, these choices were strongly influenced by pre-existing beliefs that smoking would foster social integration, enhance personal image or promote stress control.
We were going to a gig and I just decided I would buy a pack… it's the social thing cause they all… go outside and otherwise you're left there by yourself. (Kimberley, 20, social smoker)
Influence of social and environmental factors
Smoking initiation while drinking was common. Alcohol undermined participants’ capacity for rational assessment:
I can't even remember the transition, I guess because when you start drinking and smoking … you know how when you get drunk you start smoking… (Bethany, 20, daily smoker)
Social pressures had a more complex influence. Most participants did not report being actively pressured by their peers to smoke and saw ‘peer pressure’ as something that applied to younger teenagers, and not to them. However, many reported that they started smoking to participate socially.
I'm here for like four years or something, studying my degree, and just trying to be as social as possible, and (smoking) kinda seems to be one of the factors in being … social. (Henry, 20, social smoker)
This comment indicates how internalised social pressure influenced decisions to smoke.
Understanding of addiction and future smoking
Nearly all participants saw smoking as something they were unlikely to continue in the long term, a view that obviated any need to review risks. Smoking was part of a lifestyle phase where their priority was maximising novelty and social opportunities. Participants acknowledged their decisions paid little heed to longer term considerations:
You just don't think about the future like that. You know it's like a decision then and there: do you buy it or do you not? It's not “if I buy it will this happen?”, it's just “sale or no sale”, I think. (Luke, 20, social smoker)
Although participants did not regard smoking as part of a future involving settled jobs, relationships and responsibilities, they saw no immediate reason to quit:
Because (after university) you start your whole life, you've got your career, you've got kids, you've got everything … and it's expensive… I reckon I'll get sick of it, because when I finish uni I don't want to be a smoker. (Bethany, 20, daily smoker)
Many participants reported they had not understood the addictiveness of smoking when starting to smoke, although some still believed they would quit easily. These participants had a strong need to feel in control:
I like to have control over my actions, I think it needs to be an informed, like, conscious decision, not something that I have to do. (Kimberley, 20, social smoker)
However, older, more frequent smokers who now had some personal insight into addiction often looked back at their original beliefs with a sense of irony:
I can quit any time like you know everyone thinks. It's ok, I'm not a real smoker, I'm just a social smoker (laughs). I underestimated the addiction side of it, cause I always thought I could stop any time…I'll be different, you know, and everybody says that as well. I really did underestimate it. (Sara, 21, daily smoker)
Participants who had already tried and failed to quit were more cautious about their chances of quitting, although even they saw cessation as a mental rather than physiological challenge:
I think I can quickly get it under control. And by under control I mean eliminating it completely… it's all a mental thing. I'm not that weak mentally and I can get over it. (Ben, 22, daily smoker)
Viewing addiction as a psychological construct or character flaw, and not a matter of physiological dependency, allowed participants to rely on self-assessments that they were ‘strong-willed’ enough to resist addiction. This rationalisation disassociated them from those who they saw as ‘weak’ enough to have become addicted, although some reflected ruefully on the irony of their reasoning:
I think it was more... I'm not gonna get addicted, like, I'm a strong person, I'm not gonna get into this thing. But honestly, it's not hard to get addicted. (Kaine, 21, daily smoker)
As participants grew to depend on nicotine, some had become concerned about addiction, though they continued to see themselves as having chosen to smoke initially:
It's definitely a choice the first time. Um, but then after that it's no longer a choice… people are addicted… it then becomes a choice for them to quit. And that's a very difficult choice to make. (Nathan, 18, non-daily smoker)
Participants were aware that smoking is risky and some could recall specific risks; however, their uncertainty about (and denial of) risks, limited personal experience, short-term focus, impulsivity and underestimation of addiction compromised their ability to exercise an informed choice. Most struggled to assess the probability of developing health problems caused by smoking, rarely saw risks as personally relevant, and often repeated industry arguments that disease complexity meant smoking's role could not be easily delineated.17
The perceived normality of smoking among their peers and disinhibiting effects of alcohol also undermined active risk evaluation and facilitated smoking uptake. As most participants greatly underestimated smoking's addictiveness and did not expect to continue smoking, they saw no need to consider future health risks. Few young adults appreciated what addiction might entail, even though understanding this concept was pivotal to making an informed choice.
Our findings support earlier research that found young people's behaviour reflects a desire for novelty, rather than consideration of future risks.18 ,19 Participants’ accounts revealed many cognitive biases; in particular, optimism bias6 ,20 ,21 enabled participants to diminish their own risk, maintain confidence in their ability to quit,22 and gain comfort from the rationalisation that only people who smoked more than they did faced any real risk.23
Smoking both marked and threatened the transition to a new life phase in which participants could assert their independence. Many repeatedly declared their ability to stop smoking when they wished and constructed identities where they were in control, aware of risks, but not yet affected by them.12 ,24 Addiction undermined this confidence and self-image by marking the failure of psychological willpower to outweigh physiological addiction.
Some participants believed they had made a rational choice to smoke in the short term, but those who had been smoking for longer tended to consider that their earlier choices about smoking had been irrational. Those who fulfil their expectation of smoking for only a few years, quit while still young, and avoid addiction, could arguably maximise the utility they believe smoking offers while minimising its potential risks.25 However, evidence that smokers vastly overestimate their likelihood of quitting successfully,26 suggest younger participants’ confidence they will avoid addiction is misplaced. Even those who reported having chosen to smoke were arguably not making a fully informed choice because they lacked understanding of smoking's addictiveness.
Young adulthood is a period of change during which people make many significant decisions, such as whether to study, marry or relocate, that have profound but uncertain long-term consequences. While fully informed decision-making may be difficult, we argue that smoking initiation without an informed choice is highly problematic. Unlike other life decisions, smoking initiation is a not a necessary part of adult life, and its future consequences are predictable and overwhelmingly negative. Yet our findings suggest that starting to smoke is often not treated as a significant decision. Few young adults make an active and considered decision to begin smoking, and many do so in social and environmental contexts that undermine decision-making.
Our findings suggest extensions to Chapman and Liberman's framework are required. Table 2 outlines an expanded framework that incorporates these factors.
The revised framework has implications for tobacco control policy and practice.
Although the framework outlines mainly individually-focused decision-making criteria, our results illustrate how social and environmental factors influence young adults’ capacity to exercise informed choice when they start to smoke. Individuals’ peers, social and community environments, and societal norms influence and facilitate their unthinking adoption of smoking. As a result, prevention strategies should include interventions that ameliorate these influences.27
If policymakers are to enable ‘informed choice’, they must address young adults’ initiation environments as these do not promote considered decision-making. The strong links between smoking and alcohol, and the ease with which tobacco may be purchased and consumed, highlight potential measures. Although many youth may first try smoking in private settings, young adults who establish a smoking pattern often do so in social drinking contexts.28 Creating smoke-free outdoor bar areas and restricting sales of tobacco where alcohol is sold could disassociate smoking from alcohol, reduce the risk of impulsive purchase and disrupt social supply to people who are not yet established smokers. Given older participants’ rueful reflections on their own naivety when experimenting with smoking, future work could also explore raising the age of tobacco purchase beyond 18. These measures could help young adults to avoid becoming smokers unthinkingly.
A second strategy would pay greater attention to the temporal framing of messages, explore how to make risks appear more immediate, and focus on affective rather than cognitive dissonance.10 ,29 For example, messages that reframe smoking as undermining young adults’ desire for control and self-assertion could challenge self-exemptions and undermine the perceived utility of smoking.
However, communicating the risk and impact of addiction remains a significant challenge, not least because there was little evidence existing health promotion messages had influenced participants’ behaviour. Fully informing people so they both see and accept the risks smoking presents may require not just information, but fundamental changes to the way that information is processed.23 This change is likely to be difficult to achieve; for this reason, we suggest policy measures that change initiation contexts may have stronger and more immediate effects than educational strategies.
Our study had some limitations. First, our relatively well educated sample raises the question of whether smokers with less education might have different experiences, though it seems unlikely less educated groups would make more informed choices. Further, the major themes we identified align with those reported in other studies where participants’ education levels were more varied. As with all studies involving participant recruitment, the sample may have comprised people unusually motivated to talk about smoking. However, not all participants appeared to have thought deeply about smoking before the interview, and those who had strong opinions expressed diverse perspectives. If a self-selection bias was present, it appears to have been randomly distributed. Participants may also have conflated the knowledge they now have with the knowledge they recalled having when first smoking. While many recognised their knowledge had changed, the passing of time meant they could not recreate the knowledge they formerly possessed.
To define informed choice in relation to smoking, we suggest augmenting Chapman and Liberman's framework. Having accurate and comprehensive knowledge, and understanding and applying it personally, is important but not sufficient. Informed choice requires young adults to understand that, from the first time they smoke, their actions may lead to addiction and they may become one of the long-term smokers who will die prematurely of a disease caused by smoking. Informed choice also requires conscious, sober consideration, ideally free from contextual and social influences that facilitate smoking. If young adults passively discover themselves to be smokers, rather than actively evaluating and accepting the risks smoking entails, they have not exercised an informed choice. Our findings suggest the informed choice argument, like much of the rhetoric the tobacco industry uses to oppose policy interventions, lacks validity. Although the intuitive logic of assuming adult smokers make ‘informed choices’ is seductive, it is ultimately misleading, and holds smokers responsible for outcomes they cannot anticipate.
What this paper adds
This study is the first qualitative analysis to examine the validity of ‘informed choice’ arguments. It identifies contextual and cognitive factors that impede informed choices by young adult smokers, and produces an extended definition of the pre-requisites for an ‘informed choice’ about smoking.
The study identifies implications for environmental and health promotion interventions. Decoupling of smoking from drinking in young adults’ environments may reduce the ease with which people can take up smoking without thinking. Risk messages aimed at young adults need to recognise their short-term focus and desire for control over their lives.
The authors thank the study participants for their time and thoughts. A wider research group including Dr Heather Gifford, Stephanie Erick, Associate Professor George Thomson and Dr Hera Cook contributed to early thematic development. The authors also thank Professor Philip Gendall for his helpful comments on earlier drafts of the revised framework in this paper.
This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.
Files in this Data Supplement:
- Data supplement 1 - Online supplement
Correction notice This article has been corrected since it was published Online First. The author name Rebecca Grey has been amended to read Rebecca Gray.
Contributors RJG was responsible for drafting the manuscript and led the questionnaire design, data collection and data analysis. JH and RE conceived the study, supported the methodological development, analysis and interpretation and provided detailed feedback on all drafts.
Funding This study was supported by the Marsden Fund Council from Government funding, administered by the Royal Society of New Zealand, grant number UOO1126.
Competing interests None.
Ethics approval This study has been reviewed and approved by the University of Otago Human Ethics Committee (Project number: 11/297).
Provenance and peer review Not commissioned; externally peer reviewed.
Data sharing statement Only members of the wider research team working on this project have access to the raw data (transcripts) quoted in this study. Any requests for access to other materials relating to the study can be made directly to the authors, who will negotiate information sharing on a case by case basis.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.