You are here

Article Text

Article menu
Research paper
Social smokers' management of conflicted identities
  1. Janet Hoek1,
  2. Ninya Maubach2,
  3. Rachel Stevenson2,
  4. Philip Gendall3,
  5. Richard Edwards4
  1. 1University of Otago, Dunedin, New Zealand
  2. 2Department of Marketing, University of Otago, New Zealand
  3. 3Department of Marketing, Massey University, New Zealand
  4. 4Department of Public Health, University of Otago, New Zealand
  1. Correspondence to Professor Janet Hoek, Department of Marketing, University of Otago, PO Box 56, Dunedin 9054, New Zealand; janet.hoek{at}otago.ac.nz

Abstract

Background Although social smoking has increased among young adults, it remains a poorly understood behaviour. The authors explored how young adult social smokers viewed and defined smoking and the strategies they used to reconcile their conflicting smoker and non-smoker identities. The authors also examined alcohol's role in facilitating social smoking and investigated measures that would decouple drinking and smoking.

Methods The authors conducted 13 in-depth interviews with young adult social smokers aged between 19 and 25 years and used thematic analysis to interpret the transcripts.

Results The authors identified four key themes: the demarcation strategies social smokers used to avoid classifying themselves as smokers, social smoking as a tactic that ameliorates the risk of alienation, alcohol as a catalyst of social smoking and the difficulty participants experienced in reconciling their identity as non-smokers who smoke.

Conclusions Although social smokers regret smoking, their retrospective remorse was insufficient to promote behaviour change, and environmental modifications appear more likely to promote smoke-free behaviours among social smokers. Participants strongly supported extending the smoke-free areas outside bars, a measure that would help decouple their alcohol-fuelled behaviours from the identity to which they aspire.

https://doi.org/10.1136/tobaccocontrol-2011-050176

Statistics from Altmetric.com

    Request Permissions

    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.

    Introduction

    Tobacco smoking causes many cancers, respiratory diseases and cardiovascular diseases and is internationally recognised as the largest single cause of preventable mortality.1 Several governments have therefore intervened to reduce smoking prevalence and protect groups at risk of smoking experimentation and addiction. These measures, which include removing mass media tobacco advertising and implementing age-based supply restrictions, have seen smoking prevalence decline sharply among adolescents. Recent New Zealand research found that 5.5% of 14–15-year-olds described themselves as daily smokers in 2010 compared with 15.6% of those surveyed in 1999, and the proportion that had never tried tobacco was double that of 11 years previously (64.3% cf. 31.6%).2

    Despite this evidence, and findings from population surveys that daily smoking prevalence in general has decreased,3 ,4 international evidence shows many young people have commenced smoking ‘socially’ on a non-daily or intermittent basis.5–7 Because social smokers generally do not see themselves as ‘smokers’, they are often not captured in population statistics nor reached by current cessation interventions.8–11 Epidemiological evidence suggests that social smokers face health risks, exhibit nicotine dependence well before the onset of daily smoking and may progress rapidly to regular smoking.12 ,13 These findings highlight the need to gain greater insights into this behaviour and interventions that could reduce its prevalence.

    At present, researchers know little about social smokers, how and where they smoke or how they view smoking; even definitions of social smoking vary, which further complicates attempts to reach this group. In particular, our knowledge of how social smokers manage their identity as non-smokers who occasionally smoke, or how smoke-free interventions could bring their perceptions of themselves as non-smokers and their behaviour into stronger alignment, remains unclear.

    Evidence to date suggests that social smokers may define themselves according to whether they smoke with others or by linking their behaviour to a social context, such as smoking at parties, bars or nightclubs, after having consumed alcohol.9 ,14 Some define themselves according to the control they exert over where and how frequently they smoke and use these definitions to create a ‘reverse rationality’ where they have no ‘need’ to quit.15 Overall, social smokers see themselves as other than a ‘smoker’ and use definitions that corroborate their view of smoking as a passing phase that does not put them at risk of health problems.5 ,9 ,16–18

    Social smoking occurs predominantly among young adults, and longitudinal studies suggest that this age group marks an important transition point where experimental smoking may progress to an addiction.11 ,12 Industry documents reinforce the importance of young adults as future smokers, as do the promotions tobacco companies target at this demographic.19 ,20 Analyses of industry strategies, evidence of normative ambiguity surrounding social smoking and findings that non-daily smoking has increased raise important questions about how social smokers interpret and rationalise their behaviour. Furthermore, as knowledge of social smoking comes largely from US samples of college students, research that examines these questions in non-US jurisdictions would both extend existing findings and their generalisability to other settings and populations.

    Contradictory identities

    Social identity theory (SIT) provides a helpful framework on which to build an understanding of social smoking; it suggests that individuals identify with others who share common characteristics, a process that reinforces positive self-concepts. SIT suggests that individuals' social identity arises from their group memberships; members comply with group norms until belonging to the group no longer positively contributes to their social identity.21 Many social smokers state that they smoke in response to their peers' normative behaviours; smoking thus provides a common connecting behaviour that enables them to share the group's identity and define themselves as different from other groups.22 However, as social smokers define themselves as non-smokers, they belong to groups whose behaviours are mutually exclusive and cannot be reconciled. SIT suggests that they manage this complexity by smoking only in situations where their peers will affirm and validate their behaviour and where the group identity outweighs any dissonance they might otherwise experience.23

    Social smoking achieves complex interactional goals by enabling young adults to structure “social time and space that would otherwise be ambiguously defined” (see Stromberg et al, page 124). As well as facilitating identification with their social smoking peers,25 smoking may also alleviate social awkwardness as it represents something individuals can do in contexts they feel ill equipped to manage more directly. For example, offering or accepting a cigarette or a light creates a connection that obviates the need for conversation, yet still facilitates social interaction.26 Smoking thus has a social and situational utility, both of which consolidate smokers' identity.

    Yet evidence that smokers quickly grow to regret smoking questions how easily social smokers manage their dual identities, and the longevity of the social benefits they initially enjoy.27 ,28 Increasing social disapproval of smoking led Schroeder29 to describe smokers as “stranded in the periphery” (p. 2284), again focusing attention on their conflicted identities, which prevent them from belonging fully to any group. To provide some insight into these contradictions, we used SIT to interpret social smokers' explanations of their behaviour, particularly their simultaneous adoption and rejection of smoker identities and the inherent conflict between these positions. Specifically, we addressed the following research questions:

    RQ1: How do young adult social smokers view and define their smoking, and reconcile conflicting smoker and non-smoker identities?

    RQ2: What factors facilitate or catalyze social smoking, and what interventions could ameliorate these?

    Methods

    We collected data from 13 in-depth interviews conducted with self-defined young adult social smokers aged between 19 and 25 years who were recruited via Facebook, poster advertisements placed in cafes, supermarkets and on community notice boards and snowball sampling. Our advertising materials sought volunteers who ‘smoke socially’, thus they focused on the behaviour of interest rather than advancing potentially artificial definitions. Participants responded to an email address and were then contacted by a researcher who provided information about the study, clarified any questions potential participants had and set up a time for an interview.

    Interviews took place in two New Zealand cities, Auckland and Dunedin, in January and February 2011. As table 1 shows, participants were predominantly men and most identified as New Zealand European. Participants' education background varied, and the age at which they first recalled smoking a cigarette ranged from 12 to 19.

    View this table:
    Table 1

    Participants' characteristics

    The interview protocol comprised open-ended questions that enabled detailed discussion of participants' perceptions and behaviours.30 We began by exploring participants' smoking identity, patterns in their smoking, smoking contexts and their perceptions of smoking. A supplementary file contains the protocol used. All interviews were conducted in person, in a location nominated by participants, and lasted from 40 to 70 min. Participants received a $25 gift voucher (a book token) to recognise their assistance in the research. An internal reviewer with delegated authority from the central university ethics committee reviewed and approved ethical issues in the research prior to any data collection. Consultation was also undertaken with an indigenous people's committee. Participants received a full information sheet and were also advised orally of their right to withdraw from the research at any time, ask a question at any time and decline to answer any question. Each participant gave written consent prior to participating in the research.

    Interviews were audio-recorded and transcribed verbatim, then checked for accuracy by the interviewers and reviewed using thematic analysis. We followed Braun and Clarke's approach to thematic analysis, which is a flexible and iterative method for identifying, analysing and reporting patterns, as themes, within qualitative data.31 The analysis was driven by participants' narratives rather than a theoretical position. Two coders independently undertook several close readings of the transcripts, identified idea elements, plotted initial relationships and proposed preliminary themes. We then reviewed and evaluated these groupings, compared them with other coded and uncoded items and agreed on final themes. Throughout this analysis, we tested the data interpretation against validity criteria by Whittemore et al.32 On-going reviews throughout data collection confirmed saturation had been achieved, and we undertook iterative data reviews to identify, revise and test the themes outlined below. We further tested the themes against the research questions to ensure they addressed these, against the wider literature to test their logical congruence with current knowledge and, using Kuzel's expertise criterion, against our own knowledge and past experiences of young adults' smoking behaviours.33

    Results

    After analysing the interview transcripts, we identified four key themes related to the research questions: demarcation strategies that avoided smoker identities, the risk of alienation, alcohol as a smoking facilitator and rationale and internal identity conflicts.

    Demarcation strategies

    Most participants had never been daily smokers, strongly rejected the label of ‘smoker’ and used several strategies to differentiate themselves from addicted smokers, on whom they sometimes looked down. Points of differentiation included the claim that they never smoked alone: “they'd [social smokers] never smoke when there's no one around. It's only when they're with a group of friends. Like right now they'd never go outside and have a ciggie on their own.” For one participant, smoking with others defined an important boundary that he used to dissociate himself from smokers: “I've never actually had a cigarette when I'm just by myself sort of thing so I don't see myself as a smoker but I see myself as a social smoker. It's almost like they're almost mutually exclusive.”

    Others identified characteristics, such as perceived control, that they thought addicted smokers lacked and used these as differentiating attributes: “well an addiction is something that's seen to be needed all the time. And if you can choose when and where and how often you do something it can't really be an addiction, can it?” Participants who limited the contexts in which they smoked, for example, by refraining from smoking during the day time or at home or work, used this evidence of control to reinforce differences between themselves and daily smokers: “I see a social smoker as someone who only wants a cigarette when they are having a few drinks and don't smoke by themselves or have cravings during the day. Could go without. Someone who could go without.”

    Using similar logic, some participants defined smoking as a temporary phase, which they controlled and saw as a short-term indulgence; this distinguished them from smokers, who they saw as permanently addicted. For example, very few participants thought they would smoke in 5 years' time and some actively rejected smoking as part of the lifestyle to which they aspired: “Because when I see my future I don't see myself as a smoker, I don't see myself even as a social smoker. Like, … when I see successful people, none of them smoke and …I want to be successful.”

    A smaller number reasoned that if they did not perform associated behaviours, such as purchasing cigarettes, they could not be a smoker: “so yeah, that's part of my rationalizing, if I'm not buying them, I'm not a smoker. If I'm only getting them off people, then it's not an issue. Because I'm not wasting my money.”

    Only one participant reflected on how his definition had evolved to ensure he did not have to confront the possibility he could be a smoker: “At the start it's like oh I'm a social smoker and I only smoke when I drink. And then, it's like I'm a social smoker and like I only smoke when … I'm round 4 pints, and now it's like I'm a social smoker because I've never had one by myself (laughs)… for me it's changed to always include me.” Most participants showed little awareness of how they rationalised their behaviour to arrive at definitions they found acceptable.

    Avoiding the risk of alienation

    Although participants were careful to avoid defining themselves as similar to daily smokers, many nevertheless associated with daily smokers. This association created pressure to smoke, and some participants reported smoking to maintain their affinity with others in the group and avoid rejecting a ‘gift’ that had value within the group: P: I'd do that but it's more just kind of … I dunno … fit in and people offer it to you. So you feel bad if you say no. I: And why do you think you feel bad? P: I dunno. It's just a courtesy thing I suppose. I'm not really sure. There's usually a few of us that aren't really smokers so you usually half one with someone or something like that.

    Some recognised the price they paid to retain group membership, but felt the risks of saying ‘no’ exceeded the health risks they knew existed: “like I know it's bad for me but I can just do it and fit in or I can say no and run the risk of being out-casted or something like that.”

    For these participants, smoking represented a behaviour that defined their group membership to the extent that some described it as a norm: “Um, I dunno … I suppose you feel sort of left out if you're like standing there and everyone's drinking… It's just something else, so I suppose it's like, like, being included. You know the social norm.” They used smoking to maintain their group membership while out drinking and avoid the uncertainty that would come if they were left on their own: “well everyone else will be out smoking. You don't want to sit inside on your own and just drink so you're like ok, I'll jump up and have a ciggie with you.”

    Smoking also relieved social pressures they might otherwise face and enabled them to rely on shared behaviour rather than dialogue to foster social connections: “it's not that I can't talk without a cigarette, it's just that something else to do all the time. It kind of makes the conversation flow a little bit more easily.” As well as meeting their own needs to ‘belong’ participants recognised their friends' desire for group membership: “Well, no one wants to smoke on their own, do they, so give me a ciggie and they'll be like, ‘come have one with me’ […] I'll take one just to keep them company.”

    However, while smoking provided an access point to the group that enabled participants to avoid social alienation, many recognised the wider disapproval they may incur: “Dad's always been real against them [cigarettes and smoking] … … [saying] you're better than that.” Perhaps because of the tension between immediate social acceptance and societal disapproval, these participants did not enjoy smoking or anticipate smoking occasions with pleasure; instead, it was the price they paid to avoid a less desirable outcome. They resolved the reluctance implicit in their comments by drinking, which both facilitated smoking and social interactions, and provided them with a means of rationalising behaviour they regretted in retrospect.

    Alcohol as facilitator and excuse

    Nearly, all participants saw smoking and drinking as highly paired behaviours that went ‘hand in hand’. Consuming alcohol liberated social smokers from the constraints of their non-smoker identity: “Um, you're more relaxed so you don't really mind smoking … I dunno, somehow it makes it … how do you say that, makes it easier to smoke somehow, I'm not sure how, yeah.”

    Some participants recognised the paradox implicit in their drinking environment and used this to support their behaviour. Just as inside areas have become defined as smoke free, areas outside bars have become smoking zones where smoke-free behaviours are incongruous: “At the same time, it's anti-social, but it's very sociable to people that are smoking. I guess you're drawn to it. So you're out with people that are smoking there.” The ambiguity of bar settings sanctioned and reinforced social smokers' dual identities and helped them maintain the difficult balance between these. The quotation highlights the connection between the alcohol and alienation themes by illustrating how alcohol reduces rational barriers to smoking (participants are ‘drawn’ rather than deciding) at the same time as it promotes social interactions.

    Several participants reported that alcohol was more than a mere social facilitator; drinking induced physiological reactions: “I just don't have any cravings unless I've had like, if I'm out having a few drinks and then I do feel like one, but otherwise I just never feel like one.” These cravings threatened social smokers' belief in their control, and they quickly attributed them to drinking: “Must be something that the drink triggers. Something that makes the cravings stronger, yeah.”

    Participants used alcohol so they could not be held rationally responsible for their actions: “… some nights I can smoke fourteen/fifteen ciggies or a pack while I'm drinking, but I can never do that without alcohol”. Alcohol thus provided participants with an explanation of behaviour they could not explain logically and enabled them to excuse their actions: “When I'm drunk, I guess because the care factor goes down, that's what … yeah… the care factor goes to zero. Not zero, but goes down, like who cares about smoking.”

    Internal conflicts

    Nevertheless, while alcohol may have reduced participants' ‘care factor’, it did not prevent either immediate conflicts or retrospective remorse. By actively selecting the public situations in which they smoked, participants attempted to manage the internal conflicts they experienced: “Um, depending on the person and their place in society. Like if they were someone who I wouldn't want to look bad in front of, then I'd probably feel bad [smoking]… Employer, future employer, past employer, parents, some family.” Trade-offs between their desire to belong to an immediate social group and how their wider social network might perceive their actions created considerable tension, as did subsequent self-reflections.

    Those who identified strongly as non-smokers regretted the consequences of smoking and experienced severe dissonance but felt resigned to the circumstances they believed induced them to smoke: “Well I hate it. Like if someone lit one up right now I'd probably vomit. I just makes me feel so sick. It's weird I just …yeah, well, after I've had a drink I just don't care.” Alcohol both facilitated binge smoking yet allowed participants to avoid responsibility for their actions: “It's a bit contradictory really because I know that it is disgusting but I guess my ability to say no sometimes is a little bit impaired.” Claiming that their behaviour was impaired allowed participants to express strong disapproval (in line with their non-smoker identity) but justify continued smoking since they did not make this decision rationally and so could not be held to account for it.

    Alcohol may have facilitated smoking, but smoking intensified hangovers, creating a powerful negative physiological consequence: “Um, physically the next day I feel like crap. Um, for the next couple of days.” Despite their knowledge and dislike of the physical after-effects they experienced, participants showed little willingness to modify their behaviour since doing so would remove them from their social group: “Um, it makes it all, you think why did I do this, you know, I don't need to do this, this is ridiculous. But then I just do it again anyway. I dunno. It's ridiculous, it's absolutely ridiculous.” Participants managed these reactions in the same way as they coped with the challenges to their dual identity, by reducing the salience of these, typically by: “not smoking until I get drunk.”

    Because alcohol played such a powerful role in facilitating, supporting and maintaining social smoking, we explored measures that would separate drinking and smoking, such as a requirement for completely smoke-free areas outside bars. All but one participant strongly supported this proposal as it would remove the social pressure they felt: “then I wouldn't have to smoke. Like, well not have to, but none of my mates would offer me them”. Creating a physical impediment to smoking would enable them to reassert their smokefree identity, since the need to choose between identities would diminish: “Oh, I probably wouldn't do it if there was nobody to get cigarettes off… I'm willing to go outside but I'm not willing to completely leave a club to go and find someone standing somewhere on the street down the road.”

    Discussion and conclusions

    Like all small-scale qualitative research, our study has some limitations. The sample contains only four women as female social smokers proved more difficult to recruit; this factor has limited our ability to explore potential gender differences in social smoking identified in earlier studies.7 Similarly, we could not explore differences in ethnicity or social class, although we note that social class is not usually a major determinant of New Zealanders' attitudes or behaviour. Further research, specifically a larger scale quantitative study is required to test our findings and explore whether and how gender or ethnicity affect social smoking behaviours and social smokers' self-perceptions.

    Nevertheless, despite these limitations, the findings highlight important rationalising strategies social smokers used to maintain the dual identities they desire. While they perceived themselves as non-smokers, they also complied with normative behaviours practised by their peers and sought the benefits of group identity, even though smoking threatened the personal façade they wished to preserve for other contexts. Social smokers thus simultaneously disparaged smoking, saw themselves as ‘better’ than smokers and yet smoked to retain membership of a social network.

    The tension between participants' immediate social context and wider social persona created a challenging paradox. While they viewed themselves as non-smokers and enjoyed the cachet and superior social status this position brings, they engaged in behaviour that attracts stigma and social disapprobation and overtly undermines the desirable position they wish to occupy. Because social smokers rejected ‘smoker’ labels and the negative connotations these have, they struggled to maintain mutually exclusive identities and could not ever fully adopt either since doing so would disbar them from the other. This ambiguous identity requires social smokers to sustain an inherently contradictory position; they cannot consider themselves as either addicted smokers or smoke free.34

    Participants coped with this conflict by smoking when drinking with friends since alcohol eroded their ability to reason, thus enabling them to share in a common behaviour and maintain their identity within a group. Drinking enabled social smokers to reconcile the dissonance they felt since drinking went ‘hand in hand’ with smoking and removed the rational constraints that normally maintained their smoke-free behaviour. For many, the fact they smoked when inebriated absolved them from responsibility for their behaviour and enabled them to maintain their ‘non-smoker’ persona and the benefits this brought.

    Yet sobriety brought an uncomfortable reality; participants frequently expressed disgust and remorse at how they felt after binge smoking and drinking. These feelings of revulsion temporarily fractured their smoke-free identity and caused participants considerable discomfort. However, many had developed strategies to rationalise the wider questions these feelings raised. Furthermore, they quickly set aside this unpleasant reality once the group recommenced its cycle of alcohol consumption, and familiar environmental and social cues overwhelmed any uneasiness they felt.

    Because alcohol plays such a pivotal role in facilitating social smoking, extending smoke-free areas to the outside of bars would decouple drinking from smoking in this environment. This finding is consistent with other studies documenting support from smokers, and local and national policy makers for smoke-free outdoor areas35–38 and with practice in an increasing number of jurisdictions.39 Future work estimating smokers' and non-smokers' support for smoke-free outdoor bar areas would strengthen the existing evidence base. Introducing smoke-free outdoor bars could reduce social smoking by removing cues that stimulate this behaviour and changing the environment that facilitates it.17 ,26 Such a policy would eliminate the current intersection between smoke-free and smoking spaces and create a physical barrier that, for some, would make accessing the smoking zone too difficult. Nearly, all participants supported this proposal, agreed it would lead them to reduce or cease smoking and seemed relieved the environment that supported an identity they rejected might instead enable one whose benefits they sought.

    What this paper adds

    • Social smoking has increased among young adults, most of whom do not regard themselves as smokers or see themselves at risk of harm. Social smoking thus remains ambiguous, and little is known about measures that might reduce the growing prevalence of this risky behaviour.

    • Our qualitative study found social smokers differentiated strongly between themselves and addicted smokers, yet often smoked in response to group norms set by more frequent smokers. Excessive alcohol consumption enabled participants to absolve themselves from responsibility for smoking, maintain their identity within a group and rationalise the regret they subsequently experienced.

    • Nearly, all participants strongly supported a proposal to extend smoke-free areas to include the external grounds of bars, a measure that would decouple smoking and drinking and create a barrier to social smoking. Although this finding requires wider testing, it represents a straightforward policy response to increased social smoking and would help resolve the conflicted identities social smokers currently experience.

    Acknowledgments

    We are grateful to the participants who agreed to be interviewed and whose detailed responses provided the basis of this article. RS worked on the project as a summer scholarship supported by funding from the Health Research Council of New Zealand.

    References

    1. ↵
      World Health Organization. Global Health Risks: Mortality and Burden of Disease Attributable to Selected Major Risks. Geneva: World Health Organization, 2009.
    2. ↵
      ASH New Zealand. National year 10 ASH snapshot survey, 1999-2010: trends in tobacco use by students aged 14-15 years. In: Zealand AN , ed. Report for Ministry of Health, Health Sponsorship Council and Action on Smoking and Health. Auckland: ASH, 2011.
    3. ↵
      Ministry of Health. A Portrait of Health. Key Results of the 2006/07 New Zealand Health Survey. Wellington: Ministry of Health, 2008.
    4. ↵
      Ministry of Health. Tobacco use in New Zealand: key findings from the 2009 New Zealand tobacco use survey. In: Ministry of Health, ed. Wellington: Ministry of Health, 2010.
    5. ↵
      1. Schane RE,
      2. Glantz SA,
      3. Ling PM
      . Nondaily and social smoking: an increasingly prevalent pattern. Arch Intern Med 2009;169:1742–4.
      OpenUrlCrossRefPubMedWeb of Science
    6. ↵
      1. Pierce JP,
      2. White MM,
      3. Messer K
      . Changing age-specific patterns of cigarette consumption in the United States, 1992–2002: association with smoke-free homes and state-level tobacco control activity. Nicotine Tob Res 2009;11:171–7.
      OpenUrlAbstract/FREE Full Text
    7. ↵
      1. Nichter M,
      2. Nichter M,
      3. Lloyd-Richardson EE,
      4. et al
      . Gendered dimensions of smoking among college students. J Res Adolesc 2006;21:215–43.
      OpenUrlCrossRef
    8. ↵
      1. Schane R,
      2. Ling P,
      3. Glantz S
      . Health effects of light and intermittent smoking: a review. Circulation 2010;121:1518–22.
      OpenUrlFREE Full Text
    9. ↵
      1. Schane RE,
      2. Glantz SA,
      3. Ling PM
      . Social smoking: implications for public health, Clinical practice, and intervention research. Am J Prev Med 2009;37:124–31.
      OpenUrlCrossRefPubMedWeb of Science
    10. ↵
      1. Ridner SL,
      2. Walker KL,
      3. Hart JL,
      4. et al
      . Of Discrepancies, issues for measurement and intervention. West J Nurs Res 2010;32:434–46.
      OpenUrlAbstract/FREE Full Text
    11. ↵
      1. Choi Y,
      2. Choi SM,
      3. Rifon N
      . “I smoke but I am not a smoker”: phantom smokers and the discrepancy between self-identity and behavior. J Am Coll Health 2010;59:117–25.
      OpenUrlCrossRefPubMedWeb of Science
    12. ↵
      1. DiFranza JR,
      2. Rigotti NA,
      3. McNeill AD,
      4. et al
      . Initial symptoms of nicotine dependence in adolescents. Tob Control 2000;9:313–19.
      OpenUrlAbstract/FREE Full Text
    13. ↵
      1. DiFranza JR,
      2. Savageau JA,
      3. Rigotti NA,
      4. et al
      . Development of symptoms of tobacco dependence in youths: 30 month follow up data from the DANDY study. Tob Control 2002;11:228–35.
      OpenUrlAbstract/FREE Full Text
    14. ↵
      1. Waters K,
      2. Harris K,
      3. Hall S,
      4. et al
      . Characteristics of social smoking among college students. J Am Coll Health 2006;55:133–9.
      OpenUrlCrossRefPubMedWeb of Science
    15. ↵
      1. Scheffels J,
      2. Lund KE
      . Occasional smoking in adolescence: constructing an identity of control. J Youth Stud 2005;8:445–60.
      OpenUrlCrossRef
    16. ↵
      1. Seigers DK,
      2. Terry CP
      . Perceptions of risk among college smokers: relationships to smoking status. Addiction Res Theor 2011;19:504–9. doi:10.3109/16066359.2010.545155
      OpenUrlCrossRef
    17. ↵
      1. Amos A,
      2. Wiltshire S,
      3. Haw S,
      4. et al
      . Ambivalence and uncertainty: experiences of and attitudes towards addiction and smoking cessation in the mid-to-late teens. Health Educ Res 2006;21:181–91.
      OpenUrlAbstract/FREE Full Text
    18. ↵
      1. Levinson AH,
      2. Campo S,
      3. Gascoigne J,
      4. et al
      . Smoking, but not smokers: identity among college students who smoke cigarettes. Nicotine Tob Res 2007;9:845–52.
      OpenUrlAbstract/FREE Full Text
    19. ↵
      1. Ling PM,
      2. Glantz SA
      . Why and how the tobacco industry sells cigarettes to young adults: evidence from industry documents. Am J Public Health 2002;92:908–16.
      OpenUrlCrossRefPubMedWeb of Science
    20. ↵
      1. Sepe E,
      2. Glantz S
      . Bar and club tobacco promotions in the alternative press: targeting young adults. Am J Public Health 2002;92:75–8.
      OpenUrlPubMedWeb of Science
    21. ↵
      1. Tajfel H
      . Social identity and intergroup behaviour. Soc Sci Inform 1974;13:65–93.
      OpenUrlCrossRefWeb of Science
    22. ↵
      1. Fry G,
      2. Grogan S,
      3. Gough B,
      4. et al
      . Smoking in the lived world: how young people make sense of the social role cigarettes play in their lives. Br J Soc Psychol 2008;47:763–80.
      OpenUrlCrossRefPubMed
    23. ↵
      1. Scheffels J,
      2. Schou KC
      . To be one who continues to smoke: construction of legitimacy and meaning in young adults' accounts of smoking. Addiction Res Theor 2007;15:161–76.
      OpenUrlCrossRef
    24. ↵
      1. Stromberg P,
      2. Nichter M,
      3. Nichter M
      . Taking play seriously: low-level smoking among college students. Cult Med Psychiatry 2007;31:1–24.
      OpenUrlCrossRefPubMedWeb of Science
    25. ↵
      1. Kobus K
      . Peers and adolescent smoking. Addiction 2003;98(Suppl 1):37–55.
      OpenUrlCrossRefPubMedWeb of Science
    26. ↵
      1. Nichter M,
      2. Nichter M,
      3. Carkoglu A,
      4. et al
      . Smoking and drinking among college students: “It's a package deal”. Drug Alcohol Depend 2010;106:16–20.
      OpenUrlCrossRefPubMedWeb of Science
    27. ↵
      1. Fong GT,
      2. Hammond D,
      3. Laux F,
      4. et al
      . The near-universal experience of regret amoung smokers in four countries: findings from the international Tobacco Control Policy Evaluation Survey. Nicotine Tob Res 2004;6(Suppl 3):S341–51.
      OpenUrlAbstract
    28. ↵
      1. Wilson N,
      2. Edwards R,
      3. Weerasekera D
      . High levels of smoker regret by ethnicity and socioeconomic status: national survey data. N Z Med J 2009;122:99–100.
      OpenUrlPubMed
    29. ↵
      1. Schroeder S
      . Stranded in the periphery—the increasing marginalization of smokers. N Engl J Med 2008;358:2284–86.
      OpenUrlCrossRefPubMedWeb of Science
    30. ↵
      1. McCracken G
      . The Long Interview. Newbury Park, CA: Sage, 1988.
    31. ↵
      1. Braun V,
      2. Clarke V
      . Using thematic analysis in psychology. Qual Res Psychol 2006;3:77–101.
      OpenUrlCrossRef
    32. ↵
      1. Whittemore R,
      2. Chase SK,
      3. Mandle CL
      . Validity in qualitative research. Qual Health Res 2001;11:522–37.
      OpenUrlAbstract/FREE Full Text
    33. ↵
      1. Kuzel AJ
      . Commentary 1: the importance of expertise. Qual Health Res 2010;20:1464–65.
      OpenUrlFREE Full Text
    34. ↵
      1. Whitesel J,
      2. Shuman A
      . Social smoking: an untenable position. Socio Focus 2009;42:330–49.
      OpenUrlCrossRef
    35. ↵
      1. Thomson G,
      2. Wilson N,
      3. Edwards R
      . At the frontier of tobacco control: a brief review of public attitudes to smokefree outdoor places. Nicotine Tob Res 2009;11:584–90.
      OpenUrlAbstract/FREE Full Text
    36. ↵
      1. Thomson G,
      2. Wilson N,
      3. Weerasekera D,
      4. et al
      . Most smokers support smokefree council-owned playgrounds: national survey data. N Z Med J 2009;122:122–3.
      OpenUrlPubMed
    37. ↵
      1. Wilson H,
      2. Thomson G
      . ‘Balancing acts’: the politics and processes of smokefree area policymaking in a small state. Health Policy 2011;101:79–86.
      OpenUrlCrossRefPubMedWeb of Science
    38. ↵
      1. Wilson N,
      2. Blakely T,
      3. Edwards R,
      4. et al
      . Support by New Zealand smokers for new types of smokefree areas: national survey data. N Z Med J 2009;122:80–9.
      OpenUrlPubMed
    39. ↵
      American Non-Smokers' Rights Foundation. Municipalities with Smokefree Outdoor Dining Laws. Berkeley, CA: American Nonsmokers' Rights Foundation, 2011. http://www.no-smoke.org/pdf/SmokefreeOutdoorDining.pdf

    Supplementary materials

    Footnotes

    • Funding RS was supported by funding from the Health Research Council of New Zealand grant number 09/195R.

    • Competing interests Although we do not consider it a competing interest, for the sake of full disclosure, we note that JH, NM, PG and RE have undertaken work for the New Zealand Ministry of Health; JH, PG and RE have also undertaken work for tobacco control NGOs and have received funding for tobacco control research from the Health Research Council of New Zealand.

    • Ethics approval Ethics approval was provided by Department of Marketing Ethics Administrator using delegated authority from University of Otago Ethics Committee.

    • Provenance and peer review Not commissioned; externally peer reviewed.

    • Data sharing statement Our ethics approval states that the data will only be available to members of the immediate research team.