Creating the socially marginalised youth smoker: the role of tobacco control


Address for correspondence: Katherine L. Frohlich, Department of Social and Preventive Medicine, IRSPUM, Université de Montréal, PO Box 6128, Centre Ville, Montreal, H3C 3J7, Canada


We discuss how the tobacco control discourse on youth smoking in Canada appears to be producing and constituting socially marginalised smokers. We analyse material from a study on social inequalities in Canadian youth smoking. Individual interviews were conducted in 2007 and 2008 with tobacco control practitioners specialising in youth smoking prevention in British Columbia and Quebec. We found that the discourse on youth smoking is creating a set of divisive practices, separating youths who have a capacity for self-control from those who do not, youths who are able to make responsible decisions from those who are not – with these distinctions often framed as a function of social class. Youths who smoke were not described simply as persons who smoke cigarettes but as individuals who, through their economic and social marginalisation, are biologically fated and behaviourally inclined to be smokers. This ‘smokers’ risk’ discourse obscures the social structural conditions under which people smoke and reproduces the biological and behavioural reductionism of biomedicine. The collision of risk and class in the discourse on poor youth who smoke may not only be doubly burdening but may intensify social inequalities in youth smoking by forming subcultures of resistance and risk-taking.


The war on smoking in Euro–American societies has yielded one of the greatest success stories in the history of public health. Canada, for example, has seen a 50% decline in smoking prevalence since the mid-1960s. Today approximately 4.9 million Canadians, representing 18% of the population aged 15 years and older, are current smokers (Health Canada 2010). While the population-level impact of tobacco control policies is undeniable, we are concerned that social inequalities are increasingly differentiating tobacco users from non-users, with smoking prevalence and incidence following a progressively steeper social class gradient (Harman et al. 2006, Reid et al. 2010, Smith et al. 2009). Furthermore, we are concerned that smoking prevalence among younger smokers is not reducing at the same rate as the rest of the population (Graham et al. 2006), with young adults in their early twenties having the highest smoking prevalence of all age groups (Health Canada 2010, Nichter et al. 2006).

Numerous reasons are posited for the increasing inequalities in smoking – amongst them, the socially differentiated effects of population-level interventions (Frohlich and Potvin 2008) and increasing levels of social inequality in society. With regard to youth smoking, it has been suggested that levels remain high due to the unstable nature of this period of life where dramatic changes in social networks, living arrangements and social and work settings create vulnerabilities to smoking in this group (Hammond 2005).

We add, furthermore, that tobacco control itself may be playing an unintended role of enhancing, or possibly creating, social inequalities in youth smoking. Based on our analysis of interview materials collected from Canadian tobacco control practitioners working in adolescent smoking prevention, we show that current tobacco control discourse on youth smoking and smokers describes them as socially marginalised and largely unable to govern themselves. While it has been shown that smoking is becoming a practice of the disadvantaged (Reid et al. 2010, Smith et al. 2009), our focus here is on the way in which young people who smoke are discussed and known by tobacco control. The aim of this article is not to criticise tobacco control practitioners or the work that they do. Rather, we aim to draw attention to how tobacco control may be helping to constitute the very problem and subjects it seeks to eliminate or reduce.

Youth smoking in Canada

In recent years it has become clear that Canadian tobacco control faces a particular challenge with regard to youth smoking, with the prevalence of smoking in Canada highest (21%) in people aged 20–24 (Health Canada 2010). Stopping smoking initiation among young people is one route to a reduction in its prevalence in the whole population and yet, in many countries, reductions in prevalence have been weakest in younger age groups. The popularity of smoking among young people challenges the progress of cessation and prevention campaigns by continually supplying a fresh cohort of smokers to the smoking pool (Twigg et al. 2009).

Not only are there age differentials with regard to smoking prevalence but interventionists are becoming increasingly concerned with the imbalanced effect of their programming across socially differentiated youth (Fagan 2008, Niederdeppe et al. 2008, Vallone et al. 2009). Many studies note an uneven effect with regard to receptivity to anti-smoking media messages, with socially disadvantaged youth being less aware of or receptive to successful media interventions, such as the ‘truth® Campaign’ in the USA (Vallone et al. 2009).

In Canada, youth tobacco control programmes have played an important role in public health through school programming and media interventions. Two examples of social media programming help to illustrate the framing of youth smoking in Canada. ‘Quit4Life’ is currently the largest national initiative for youth smoking in Canada. This intervention includes a handbook and website (Health Canada n.d.) designed to assist youths who wish to quit smoking. The approach of this initiative is illustrated in a passage from the website’s introductory page:

About Quit4Life:

Did you know that it takes more than will power to quit smoking? It takes skills and support, as well as your own motivation to quit. The new Quit4Life (Q4L) web site can help you to quit smoking … for life. Q4L is organised around 4 central steps: Get Psyched, Get Smart, Get Support, Get On With It … Each week, you will be asked to complete 3–5 activities on-line which will help you learn about why you smoke, how to quit and how to stay quit once you get there.

In Ontario, ‘’ is a well-known provincial programme targeting youth smoking. As with Q4L, instructions on the webpage ‘Project Stupid, Book 1’ clearly illustrate the particular framing of youth smoking issues:

Why should YOU take a stand against tobacco? Well, that’s up to you. The lists of reasons are long. And the more you learn about the harmful effects of tobacco, the more reasons you’ll have. By taking a stand you don’t have to start your own group, put up posters and try to change the world (although that would be great too). All you have to do is tell just one person about what you learn here. And that’s the first step towards making a difference – knowing the facts … Well, you’re in the right place. is filled with tons of information. But when it comes to tobacco, you can never have too much knowledge. (Ontario Ministry of Health and Long-Term Care n.d.)

Both the Q4L and campaigns frame smoking cessation as an individual responsibility. They suggest that with the right kind of knowledge, young people can be made responsible and thereby make the right choices: that is, with the help of these tools, self-controlled and smart youth will be able to stop smoking. Concerned by growing inequalities in smoking amongst youth and the framing of smoking in Canadian social media, we sought to examine how tobacco control practitioners themselves described youth smoking.

Governmentality, risk and social class

To support the analysis of our interview material, we draw on Foucauldian theory of governmentality and extend it to explore its intersection with the social relations of class. Foucault introduced the concept of governmentality in his writings on the emergence of a new mode of government in 18th century France, which was principally concerned with regulating the health and welfare of a newly constituted object of knowledge: the population (Foucault 1979). Foucault also used the term to refer to the art of government, pertaining to modes of thought around the governance of human conduct in and through multiple social sites.

Studies informed by a governmentality perspective typically make one or more central analytic moves. At the risk of oversimplification, these include unsettling state-centred conceptions of power, an emphasis on the exercise of power as productive rather than negative or limiting and an abiding interest in the central role played by calculative and other forms of expertise in encouraging practices through which people govern their own conduct. A concern with self-governance has been particularly prominent in work on the art of government relating to matters of health (Burchell et al. 1991).

A range of Foucauldian scholars has extended Foucault’s biopolitics of population – the use of normalisation to encourage people to engage in self-surveillance – to the contemporary emphasis on risk discourses, which encourage self-responsibility for health and wellbeing through various forms of risk-monitoring (Coveney 1998, Lupton 1995, McDermott 2007, Peterson and Bunton 1997). Our analysis is informed by these 20th century explorations of health promotion. In The Imperative of Health,Lupton (1995) argues that public health discourses both constitute and regulate such phenomena as normality, risk and health. She then notes (in her collaboration with Peterson) that, as a political practice, neoliberalism emphasises approaches to health that are increasingly individualised and focused on ‘the self who is expected to live life in a prudent, calculating way and to be ever-vigilant of risks, self-regulating and productive’ (Peterson and Lupton, 1996: xiii, 12). Lupton further emphasises how health promotion in neoliberal times operates as moral regulation, encouraging people to modify their bodily activities in the pursuit of good health. Of particular concern is how people are enjoined to identify and manage a host of risk factors as part of what Monica Greco (1993) has called the ‘duty to be well’. We orient to tobacco control as an expertise that promotes self-governing, ‘healthy’ subjects by exhorting them to conduct themselves in accordance with expert advice about the health risks of smoking.

Our interest here, however, is with the effects of such efforts on class relations, a concern typically removed from a Foucauldian analysis of power. By (social) class, we refer to economic or cultural arrangements of groups in society distinguished by levels of power. Traditionally, Foucauldians write counter to Marxist notions of the significance of class for social organisation. As such, research on public health that draws on Foucault has largely ignored how public health, as a form of governance, is class-stratifying in its effects. Governmentality-inspired critiques of public health tend to be staged at high levels of abstraction, with little empirical focus on actual social processes and the ways in which public health practices are conducted. The idea of the self-governing subject becomes all-encompassing, with scant concern for the often contradictory ways that health promotion messages are delivered and taken up by actual people, sometimes based on social class (Mykhalovskiy 2008, Poland and Holmes 2009). We will argue that the meaning and significance given to certain types of youth who smoke reflect and reinforce social differentiation through descriptions of the inevitability of smoking amongst socially and economically disadvantaged youth.

Foucauldian analyses of smoking and tobacco control are not new, but they are rare, despite the fertility of the analytic insights they offer (Poland 2000). We focus on tobacco control discourses of youth smoking as they draw attention to the ways in which knowledge and authority are produced to determine what counts as truth (Bacchi 2009). What is of greatest interest for us in this article is not the objective merits of truth claims embedded in these discourses, but rather their probable effects. We will argue that tobacco control discourses may not only be creating notions of what a youth who smokes is, but may also be creating a class-based notion of smokers.

Study description

The empirical work that informs this article was undertaken in two Canadian cities, Vancouver and Montreal, from 2007 to 2009. These cities were chosen for their differences both in their demographic smoking profiles and their tobacco policies. In demographic terms, in 2007, 17% of people aged 15–19 in the province of Quebec were daily smokers – one of the highest rates in the country (Health Canada 2007) – while British Columbia had one of the lowest levels of youth smoking in Canada with 9% of youth aged 15–19 reporting daily smoking (Health Canada 2007). With regard to tobacco policies, all indoor public places have been 100% smoke free in Vancouver since 1996. In contrast, Quebec was slower to enforce its tobacco policies, with initial policies to reduce public area smoking introduced in June 1998 and universal public banning introduced only in May 2006.

As part of our multi-component study, individual interviews were conducted with tobacco control practitioners specialising in youth smoking prevention in Vancouver (13) and Montreal (12) in 2007 and 2008. We defined a youth tobacco control practitioner as any health professional or programme developer who had the prevention or cessation of youth smoking as a major component of their job mandate. We stratified our sample by whether the practitioner worked in programme development or delivery (see Table 1).

Table 1. Practitioner sample characteristics
Practitioner Position title Organisation Type of programme Gender (M/F)
 BC 01Programme coordinatorNon-profit health sectorDevelopment and deliveryF
 BC 02Programme coordinatorEducationDevelopmentM
 BC 03DirectorEducationDevelopmentF
 BC 04DirectorHealth ministryDevelopmentM
 BC 05Prevention coordinatorNon-profit health sectorDeliveryF
 BC 06Programme coordinatorHealth authorityDevelopment and deliveryF
 BC 07ConsultantEducationDevelopmentF
 BC 08CounsellorEducationDeliveryF
 BC 09CounsellorEducation/health authorityDeliveryF
 BC 10ManagerHealth ministryDevelopmentM
 BC 11ManagerHealth ministryDevelopmentF
 BC 12CounsellorEducation/health authorityDeliveryF
 BC 13Programme coordinatorHealth authorityDevelopmentF
 QC 01Programme coordinatorNon-profit health sectorDevelopmentF
 QC 02DirectorNon-profit health sectorDevelopmentF
 QC 03Programme coordinatorNon-profit health sectorDevelopmentM
 QC 04DirectorHealth ministryDevelopmentF
 QC 05Academic and ConsultantPublic universityDevelopment and deliveryM
 QC 06Programme coordinatorSocial servicesDevelopmentF
 QC 07Programme managerAthletic associationDevelopmentF
 QC 08DirectorNon-profit social servicesDeliveryF
 QC 09Clinician/counsellor (school nurse)EducationDeliveryF
 QC 10Clinician/counsellor (school nurse)Health authorityDeliveryF
 QC 11DirectorVoluntary sectorDeliveryF
 QC 12CounsellorEducationDeliveryF

Once permission was granted, we engaged in one-on-one semi-structured interviews with each of the practitioner participants. They were asked to speak about the general context of their work in tobacco control and youth smoking prevention, the social context of smoking and youth who smoke, and their general work environment. The interviews lasted between 37 and 89 minutes with the average interview lasting from 45 to 60 minutes. All 25 interviews were digitally audio-recorded. The interviews were transcribed verbatim from the audio-recordings using well-developed transcription guidelines (Poland 1995).

All members of our research team were requested to independently read a selection of the interviews. We then met to identify common themes and codes emerging from our reading of the interviews. Based on our reading and in-depth discussion of the material, we decided to analyse the interviews using the following six analytical questions: (i) How are youth who smoke positioned in tobacco control work? (ii) How do tobacco control practitioners know the youth? (iii) How does tobacco control work account for local conditions or subgroups? (iv) How does tobacco control explain and intervene on the (uneven) social distribution of smoking? (v) How do tobacco control practitioners understand the social context of smoking? and (vi) What knowledge base and discourses do practitioners use in their work?

The material was analysed using N-Vivo software. Several members of our research team were responsible for coding the material. To ensure consistency in applying the codes, two members of the team each independently coded the first two interviews and then met to discuss their results and approach. Through this process, codes were clarified and coding approaches standardised. Subsequent reviews and check-ins were scheduled between coders in the Vancouver and Montreal teams as well as across these groups in order to maintain inter-coder agreement at the team level.


A surprising key finding, which was not an explicit aspect of our initial analysis, was that the discourses distinguished, in various ways, youth who engaged in self-governance from those who did not. This separation, crucially, was drawn largely along social class lines. To substantiate and further explore this finding we organised our results into three categories: explanations of youth smoking; ways of intervening to reduce youth smoking and reflexive thoughts from practitioners. Because we were interested in dominant discourses we took particular interest in the framings of youth who smoke that were consistent across our two locations, suggesting this was strongly grounded in professional discourses.

Explaining youth smoking

The intersection of risk and marginalisation discourses:  Across the interviews it was common for practitioner accounts to slip quickly from causes of smoking to what causes social problems, as though smoking were part of a larger risk package. The relationship between smoking and other ‘social ills’ such as drug addiction, alcohol consumption and general delinquency was in this way voiced as a given, something to be expected. These issues are brought out cogently in the following segments from two Vancouver-based practitioners:1

BC09 [I]t’s a higher risk population that takes the chance, yeah. I have this one sheet that shows that smoking can be correlated to skipping, lateness, all kinds of high-risk behaviours, a whole page of them, early sexual activity, so all of those factors, although tobacco is also correlated to a lot of other things, like family issues.

BC03 I think there’s a lot of other issues of why kids take up smoking, you know, there’s a lot of risk-takers, there’s kids who get into high-risk … and there’s high-risk behaviour that goes hand-in-hand, of course, with smoking … they’re drinking on a regular basis. So it’s probably easier to pick up a cigarette or two, you know, start that other behaviour. I think that’s probably an area that we may find or you already have found that the high-risk youth are … you’re going to have youth risk behaviours … not just tobacco but others associated with marijuana.

The framing of youth smoking as part of a risk package suggests not only that youth who smoke are at risk for problems associated with smoking (such as cancer and heart disease) but that they are also at risk of being and becoming more deviant through other risky practices. Youth smokers are thus not only framed as smokers: smoking appears to engender other deviant social and behavioural tendencies. Ostensibly, youth who smoke are constructed as susceptible to a whole host of deviant practices, unable to self-manage in all sectors of their lives and set apart from what is considered appropriate and responsible behaviour. Not only were youth who smoke represented as socially and behaviourally deviant but more importantly, their behaviour was framed as a result of poor, misguided and morally questionable individual choices:

QC09 Like using drugs, being more likely to be involved in other risk behaviours in their sex lives, experimenting with alcohol, drugs, bad stuff. Once one is addicted to one important drug … we have to intervene to help them stop smoking. There are so many studies also that tell us that youth who smoke are also more likely to make other poor choices in their life as well.

QC03 I mean when she [the smoker] arrives home, there is smoke everywhere, where there are … questionable or just plain bad interpersonal relationships, let’s say it like that, if we think of our socioeconomically poor milieus, I am thinking of a young girl in Gaspésie last year … what she described, it is as though I was seeing a bad film in front of me.

Overall, the practitioner discourses begin to suggest that actual risk is constituted as much through behaviour (smoking) as it is through group membership (being part of the lower social classes) (Shoveller and Johnson 2006). Here the double meaning of ‘poor’ (as bad choices and as economic and social poverty) is telling. There is an implicit suggestion that smoking, along with the other ‘poor’ behavioural choices, is part of a risk package specifically accompanying membership of a low social class. Furthermore, in line with neoliberal notions of normalcy commonly framed as low risk, these youth who smoke are clearly positioned outside the neoliberal definition of responsible citizenry.

Where biology meets economics:  When youth smoking is not framed in terms of individual lifestyle or social deviance, many practitioners sought explanations for it in terms of individual biology. These discussions ranged from individual genetic susceptibility to addiction to physiological changes believed to take place when foetuses are exposed to smoking:

BC09 Also I was just reading The Tipping Point– it has been out a long time but in, in the back of that talks about things like addiction and sort of certain personalities or certain biologies seem to have a very instant response to certain chemicals in the brain … some of the chemical things they come up with to help people quit smoking are actually probably the most effective for that particular section of people.

QC04 [T]here are starting to be certain studies that tend to show that well, a baby in his mother’s womb suffers from the influence of his mother’s cigarette smoking, that his smoke gets through the placenta and that there may be an early development of a dependence.

In both these quotes, practitioners offer the hypothesis that certain youths smoke due to a biological propensity, whether it is caused by early foetal exposure or due to personality types or other biological predispositions. These quotes do not in themselves make an implicit connection between biology and genetic predisposition and social class. They are, however, part of a constellation of perspectives that clearly draws links between flawed biology, social class and the likelihood of smoking:

BC06 I think it’s certainly the kids that are marginalised for whatever reason that they’ve … they’re likely coming from homes where there’s open smoking in the home, over their lifetime, so essentially they’ve been exposed to secondhand smoke and probably somewhat conditioned, you know, almost tolerant of it. And maybe even craving … if they have a bit of an addictive … propensity I guess. Then I think some of it’s culture and some of it is genetic … the jury’s out on all that stuff.

BC13 I think that it’s, I think that it’s adults in their environment who smoke so they’ve seen this from an early age, I think that they have much greater access to cigarettes, I think that they have fewer opportunities to engage in more positive activities, I think that, that we know that there is a link between mental health and tobacco use and I think that that may start very young and we’re seeing evidence too that relates, that use of tobacco not just as an environmental issue but use of tobacco in pregnancy impacting tobacco use … and also the genetic information we’re getting now about the genetic susceptibility to become addicted to nicotine, I think that’s fascinating that we haven’t really looked at that yet either.

What is most critical is the linking of a flawed biology to social class, and in turn, the propensity to smoke. Earlier we learned that smoking was viewed as part of a risk package involving multiple social and behavioural risks. Here an additional risk is added: biology. This additional qualifier of smoker is significant in that it creates yet another reason to believe that socially and economically marginalised youth are unable to govern themselves: flawed biology precipitates smoking due to insufficient self-control (the propensity to become addicted to smoking).

Preventing youth smoking

Despite a level of consciousness with regard to the social class backgrounds of many youth who smoke, tobacco control practitioners tended to discuss their prevention practices through the well-established, individual-level recipe of interventions most commonly used in public health. These interventions largely focus on knowledge of the dangers of smoking and lack of self-control in relation to poor lifestyle choices, juxtaposed with the disciplined, self-governed non-smoker’s body.

Intervening on knowledge and addiction:  Consistent with their view of smoking as part of a risk package involving other social ‘vices’, it was not uncommon for practitioners to frame smoking interventions in the same way they framed other addictive behaviour: as part of a lifestyle choice. Suggested prevention methods invoked the language of addiction to encourage youth to stop or not take up smoking:

BC09 And usually they get a little core group of kids that come to that and then [an organisation] has developed, they’re using engagement process in another meeting where they invite kids who have been into drugs and alcohol to come and discuss whatever they would like to discuss and they, they may choose to go down to, I think of the field trips, they’re going to do it down … to a treatment centre in the downtown eastside. So they help prepare lunch and then they see these people and a lot of the people that are there who are in recovery will still be smoking cigarettes, it’s one of the last things a person in recovery can give up as well. So it’s kind of interesting for kids to realise wow, we might have started with cigarettes and then the very last thing that if we had drug dependency that we would be able to give up. So it’s always kind of subtly there, tobacco.

BC01 So, I don’t know, and increasingly what I’ve learned from workshops on cessation is that if you’re going to treat tobacco as an addiction then you have to talk addiction language, so, we talk about the nicotine delivery device, the cigarette. We talk about free-basing nicotine when you’re smoking. And that just puts it in a slightly different light to them. And … brings to it a degree of reality that they may not have appreciated.

Given the implicit suggestion that smoking is only one of the problematic forms of behaviour that marginalised youth pursue, it is not surprising that interventions developed for smoking prevention and cessation draw heavily on other addictions. Interestingly, the links drawn by tobacco control between smoking and other addictions are being cited by people who smoke themselves, as shown by Bell et al. (2010a: 919):

A number of interviewees made comparisons between tobacco and illicit drugs, particularly crack cocaine, expressing the view that in Vancouver tobacco use was seen to be on par with smoking crack.

Conflating the links between flawed biology and social class, some practitioners believe that poor youth who smoke require different interventions, specifically because they deem them unable to self-manage their addictive behaviour. The following response demonstrates how some practitioners believe that behavioural and educational interventions are insufficient with these youth; they suggest pharmacological interventions as a key solution:

BC10 The one thing that we most desperately need in this province is funding for NRTs, nicotine replacement therapy or pharma, pharmacotherapy – [Zyban, Champex] we need that because it was … it was a lot easier helping people quit that we would sort of call the low-hanging fruit. We could easily help that population, but now we’re getting into the really hard to reach populations, people that have got a lot more problems and issues, mental health and addictions, lower socioeconomic status and so these people really need something more than just a counselling session or a brochure. They need medication.

The implications of such a discourse are clear. Not only are low social class youth who smoke constituted as being at greater risk of other social and behavioural problems but these problems can only be resolved medically. The faith in the critical role of medicine and health care in the resolution of lifestyle-related problems is typical of neoliberalism (Wheatley 2005). While the so-called low-hanging fruit were responsive to tobacco control in a rational, self-regulatory and responsible fashion, low class youth, believed to have a biological propensity for smoking linked to poverty, are non-compliant and therefore require specific forms of pharmacological intervention to stop them from smoking.

Learning self-regulation:  A second frequently discussed approach to prevention focused on skills development and decision-making. Many smoking prevention programmes emphasise self-esteem and social skills building, based largely on the belief that youth start smoking because they lack self-confidence (Haines et al. 2009). This discourse implicitly frames individuals as rational actors, with smoking being a decision, a choice, a volitional, individual act. Those youths who choose to smoke are seen as lacking the right set of social, emotional, psychological and biological tools to resist pressures to smoke. It is assumed that once youth have the proper skills and knowledge they will be able to make the right choice: not to smoke:

BC03 [W]e talk about beer ads or toy advertisements or whatever, but just to start that critical thinking around how media affects decision-making. But the programme is around skill development, all the important skills, 'cos kids need … the focus is more on how do you solve problems, how you make decisions, what are some coping techniques you can use in certain situations, refusal skills are a big, that’s a big problem. Yeah so the focus is not the drug, specifically it’s more, skill development.

BC07 [W] e need to help children understand that there are dangers in smoking and help them figure out how to make decisions. So we really look at it as a decision-making model more than anything … and how do you say no. You know if you get into situations, so it’s really based on critical thinking and decision making … Well, I think what works well is really letting kids get the information and having … giving them some decision-making techniques, and helping them understand that they’re the ones that make the choices.

What is otherwise a well-intentioned focus on self-esteem by tobacco control becomes a discourse that ultimately implies that some types of youth who engage in smoking (read: low social class youth who smoke) must possess character or biological flaws or personality deficiencies resulting from low self-esteem (Shoveller and Johnson 2006). Practitioners see these shortcomings as disabling low social class youth from self-government, like their more socially fortunate peers. Fundamentally, this discourse obscures and disregards the larger social and structural conditions that constrain choices amongst socially marginalised youth (Hansen and Easthope 2007, Wheatley 2005).

Thoughts from reflexive practitioners

A number of interviewees were clearly aware of some of the (unintended) effects of their discourses and practices, particularly with regard to socially marginalised youth. Here, practitioners reflected on ways in which the framing of smoking and smokers in current popular and medical discourses demonised people who smoke, with some serious consequences:

QC11 I think that it is more the … not lobbying, I am using the wrong term, but this kind of demonisation that we have done of smokers that has made people stop smoking. The smoker is the ‘big bad guy’, after the paedophile comes the smoker practically, these days, in our society, the bad guys. You see a smoker outside smoking a cigarette, children, ‘Oh’ and they look at the smoker with big wide eyes as though he was going to kill a baby seal in Alaska. It’s the same thing for them, it is really the demon. They are really viewed, we marginalise, we really, really do marginalise smokers, the more we do, the less place smokers have.

This type of marginalisation has been described by some as a spoiled identity. The act of smoking is undesirable and smokers are vilified. As a result, being a smoker constitutes an undesirable social identity in contemporary western societies (Bell et al. 2010b). Another practitioner from Vancouver went so far as to suggest that tobacco control discourse and policy may even be creating a bond among those leftover people who smoke. Interestingly, the shaming, ostracising and marginalising effects of the current discourse were viewed to be creating a kind of a shared (shameful) identity between people who smoke:

BC10 I think that there’s always, with policy, an unintended consequence, and I think we see it with youth as a whole range of smokers is how we galvanise them into a particular subculture by ostracising, because we have, whether you want to call it social shaming or whatever … you know, whatever that kind of language it is, that’s something that they can bond over.

This shared identity has been described as the internalisation of a sense of outsiderness, with people who smoke describing themselves as weak-willed, stupid, gross or dirty (Thompson et al. 2007). This identity clashes with the preferred neoliberal individual who makes good health choices, is responsible and is concerned about health risks. It is entirely possible, through tobacco control discourse ascribing this spoiled identity to youth who smoke, that this identity becomes embodied, encouraging smoking amongst lower class groups.


Tobacco control is a powerful social institution. Its members act within and in response to a discourse, to which they contribute, that normatively constitutes youth smokers and smoking. By exploring tobacco control professional discourse at this time we gain insight into how tobacco control practitioners understand the particular ‘problem’ of youth smoking and how they constitute the youth smoker as an object of discourse, intervention and public health practice. We found, as did Shoveller and Johnson (2006), that discourses on youth smoking are creating a set of dividing practices that separate youth into those who are able to control themselves and those who are not; youth who are able to make responsible decisions and those who are not – often basing these distinctions on social class. As Skeggs (2004) notes, the social meaning given to a practice (such as smoking), and its ability to signify social (and class) distance does not derive from some intrinsic property of the practice but from the position it has in the system of objects and practices. The meaning ascribed to smoking by our practitioners reflects and reinforces the attribution of youth smoking as inherently classed.

The discourse of governmentality figured prominently in practitioners’ interviews. Practitioners founded their explanations of youth smoking in their perceived lack of self-management, whether this was due to social, behavioural or biological failings. Socially and behaviourally, youth smokers were seen to be making the wrong choices. They were also frequently viewed as being unable to control themselves from smoking due to deep biological dispositions. Youth smokers were therefore exhorted to engage in body maintenance activities (Featherstone 1991), including the improvement of their self-esteem, increasing their knowledge of smoking, and, as a last resort, using pharmaceutical options.

The moral implications of these discourses are of considerable importance. To engage in smoking was considered by these practitioners as a ‘failure of the self to take care of itself – a form of irrationality, or simply a lack of skilfulness’ (Greco 1993: 361). Indeed, in contrast to the civilised body (that which is self-controlled, autonomous and self-regulated), youth smokers were viewed as uncivilised, lacking self-control and self-discipline. In contrast with the modern subject associated with neoliberalism, (that is, the rational, autonomous and disciplined individual) (McDermott 2007), youth who smoke were seen to require particular forms of lifestyle and medical intervention; they lacked self-discipline and control, which they needed to learn through proper interventions. Strikingly, across the interviews the distinction between types of youth who were better able to govern themselves than others was drawn between youth belonging to a low social class and the rest of society.

Furthermore, youth who smoke were not simply described as individuals who smoke cigarettes, as in the past, but as individuals who are biologically fated and behaviourally inclined, through their social class position, to be smokers. This collision of risk and class in the discourse on poor youth who smoke creates a dual stigmatisation (Thompson et al. 2007) with potentially damaging consequences. Only certain social groups are signalled out as being ‘at risk’, a label that tends either to position members of these social groups as particularly vulnerable, passive, powerless or weak, or as particularly dangerous to themselves or others. In both cases, social attention is directed at these social groups, positioning them in a network of surveillance, monitoring and intervention (Lupton 1999, Poland and Holmes 2009).

This same positioning appears as the outcome of the ‘at risk’ labelling of Canadian youth who smoke. Essentially the risk discourses we heard with regard to Canadian youth who smoke create moral boundaries between good and bad (Wheatley 2005). More seriously, these discourses disregard the constraints that some people face in their efforts to not start smoking (or to stop doing it), and end up (unintentionally or not) blaming those who fail to live up to the standard of the non-smoker: the privileged subject who is self-regulating, rational and disciplined. The enhancing of skill sets, competencies and knowledge encouraged by several of the practitioners through their interventions aligns with neoliberalism in that responsibility was placed on individuals to improve their knowledge, self-esteem and decision-making abilities rather than on the structural factors affecting smoking, such as social class. These types of interventions obscure several facts. Lower social class is related to a reduced sense of personal agency and control over all aspects of life, not just smoking (Bosma et al. 1999, Lachman and Weaver 1998). Furthermore there are social structural conditions under which people start to smoke beyond the biological and behavioural reductionism of biomedicine (Poland et al. 2006, Wheatley 2005). While skills development and pharmaceutical treatment may be successful in solving some issues in certain individuals, they will largely fail to improve the health of youth whose smoking is produced, and reproduced, by forces and relations beyond their individual control.


We sought in this article to identify discourses mobilised by Canadian tobacco control and show how they are contributing to the construction of the identity of socially marginalised youth who smoke. Our goal is to destabilise assumptions about youth and their smoking that are part of the public health establishment by identifying the dominant discourse on youth smoking and showing how it contributes to assumptions about socially marginalised youth, which, we argue, may serve to maintain, or even exacerbate, social inequalities in smoking (Lupton 2000).

These discourses are creating a particular identity for youth who smoke based on social class – an identity that is problematic in part because it deviates from the dominant discourse on smoking based on the self-controlled, rational, health-seeking individual. The ‘truth’ about marginalised youth who smoke is being created through the descriptions and knowledge attributed to them. Further, when youth who smoke are framed as biologically, behaviourally and socially flawed, they may begin to believe it. Indeed, in focus group material gathered with low social class youth who smoke in a companion piece to this project, we often heard this to be the case. These youth often use terms such as loser, slacker and bad in relation to their ‘smoker’ status in ways that privileged youth who smoked did not. By adopting such discourses tobacco control practitioners may, inadvertently, be reinforcing and creating the very phenomena they wish to remedy.

Several practitioners were aware of the tensions in the current discourse and practice of tobacco control in relation to low social class youth who smoke. They pointed out, as do we, the importance of being conscious and critical of the dominant discourses. These practitioners understood and recommended that they be aware of their position as producers and reproducers of certain neoliberal discourses and practices as well as the values inherent in them (Lupton 1995).

Finally, our analysis demonstrates that tobacco control continues to focus on individually based behavioural approaches to smoking. Because smoking amongst youth is stratified by social class, individually based approaches will inevitably preclude treatment of the more complex social relations that link smoking to class. Specific alternative approaches for Canadian tobacco control would include the integration of life choices and chances and addressing their interplay (Cockerham 2005, Frohlich et al. 2001) across a range of diversely situated youth. Canadian tobacco control would do well to consider smoking in a broader structural context of unequal life chances, rather than as an atomistic, individualised behavioural trait (Poland et al. 2006).

In the case of Canada, rather than focusing on tobacco control interventions prescribing better knowledge, self-esteem and decision-making, focus should be turned to reducing differences in life chances and inequities across disadvantaged areas. Examples include the implementation of extracurricular activities in schools and communities with high rates of smoking to encourage adolescent participation in identity-affirming activities other than smoking. Structural differences in area based on social class composition can be further mitigated by community-based efforts to reduce violent, dangerous and often non-salutogenic living conditions in disadvantaged neighbourhoods and create safe and attractive places to play and mingle. Another strategy entails integrating tobacco control (prevention and cessation) into multiple services provided by governmental and non-profit organisations in disadvantaged areas. The health start tobacco cessation initiative, developed jointly by the Mailman School of Public Health at Columbia University and Legacy (an American anti-smoking foundation) – is an example of an approach that integrates cessation into social and human services to bring support specifically to socially marginalised communities (American Legacy Foundation 2010).


  • 1

     All interview material prefixed BC was collected in British Columbia; that prefixed QC was collected from the province of Quebec. The French language material collected in Quebec has been translated for the purposes of this paper


This work was supported by a Canadian Institutes of Health Research Operating Grant as well as a CIHR New Investigator and Humboldt Senior Scholar award to K.L. Frohlich. Thanks are due to Caroline Fusco and Thomas Schlich for careful critical readings of earlier drafts of this manuscript. K.L. Frohlich would also like to extend thanks to the members of the Public Health group at the Wissenschaftszentrum Berlin für Sozialforschung (WZB), for stimulating some of these ideas while she was on sabbatical leave. Finally, special thanks to the practitioners interviewed for their continued generosity and collaboration.