Justice Everywhere

a blog about justice in public affairs

21st Century Smoking


At the British Medical Association’s (BMA) annual representatives meeting this week, doctors voted overwhelmingly to push for a permanent ban on the sale of cigarettes to those born after 2000.* What are the different reasons that might motivate, and potentially justify, the state intervening in citizens’ smoking behaviour? Broadly speaking, the main distinctions are those drawn between: (1) welfare- (both individual and collective) and autonomy-based reasons; (2) ‘harm to self’ and ‘harm to others’, that is, for the sake of smokers versus for the sake of non-smokers generally; and, relatedly, (3) an aim to increase tobacco use cessation (i.e., stop smokers smoking) versus an aim to reduce tobacco use initiation (stop people from starting to smoke in the first place). Accordingly, an initial taxonomy of reasons might have the following six cells:

Welfare-based reasons
Autonomy-based reasons
Welfare of smokers
Autonomy of smokers
Welfare of non-smokers
Autonomy of non-smokers
Potential smokers
Welfare of potential smokers
Autonomy of potential smokers

The main focus of the proposed ban is the ‘potential smokers’ category, those who are currently aged 14 and under, with the ultimate aim being to essentially phase out smoking over the course of several generations (e.g., by 2070 it would only be those born before the turn of the millennium that would be able to buy cigarettes, that is, those who are over 70 years-old). So what are the main arguments in favour of this?

Welfare: Smoking is the primary cause of a whole host of preventable illnesses and of premature death, with half of all life-long smokers dying prematurely (losing ten years of life on average). The welfare-based argument in favour of the ban is fairly strong, then: it seeks to protect the objective welfare of potential smokers (i.e., health, financial opportunity costs) without having to weigh this against the subjective welfare that smokers derive from the practice (though this type of welfare is on shaky ground too, since various studies show that the majority report getting relatively little subjective pleasure from smoking and would rather not smoke). [NB: The kind of potential welfare costs not often taken into consideration, though, are relational ones, such as those expressed in this article (para. 4).]

Autonomy: Liberal anti-paternalist orthodoxy states that, as long as she is not impaired by ignorance or involuntariness, an individual’s choices should not be interfered with in the name of her own good. But although Millian liberals query whether one’s own good is sufficient warrant for interference in her behaviour, they would be the first to concede that it might be, if the behaviour in question is not (fully) voluntary: “If it is autonomy that we are trying to protect in opposing paternalistic legislation in general, then the same values that lead us to oppose such legislation in general will lead us to welcome it in those particular cases where what we are being protected from is something that would deprive us of the capacity for autonomous choice” (Goodin, 1989). Given that a very significant proportion of current smokers – estimates range between 66-80% – began smoking in their teenage years (a time when the state does not recognize them as autonomous), two main factors challenge the voluntariness of potential smokers’ decisions: peer/ social pressure and the addictive nature of nicotine. The autonomy of potential smokers, therefore, might actually be protected and promoted by such a prohibition.

Nevertheless, there are a number of ethical and practical problems with the proposed ban on the sale of cigarettes to those born after 2000, some of which are highlighted below:

(i) Much of the force of the argument is based on the idea of protecting those who are 14 and younger from taking up smoking; however, in four years’ time, the children born in 2000 will be 18 and there will then be an asymmetry in the legal rights that adults have concerning the purchase of cigarettes: tobacco will still be a legal product, but it will be either legal or illegal for adults to purchase it depending on an arbitrarily-set age threshold. Can this be justified?

(ii) In practical terms, the proposal appears to have forgotten one of the key facts on which the moral argument is based: if the vast majority begin smoking in their teenage years now, when it is illegal for those under 18 to purchase cigarettes, then it seems that there will be nothing (extra) stopping this from happening in a world where cigarettes are banned for those born after 2000. The result, then, would be a group of adults who are addicted to smoking, but who are prohibited from buying legal cigarettes; and it seems fairly inevitable that, within these circumstances, the illicit trading of tobacco products will rise to meet the demand, which would bring negative consequences for the whole of society. This point also highlights a contradiction in the pro-ban argument: the reason for selected the 2000-threshold is that there are humanitarian-cum-utilitarian reasons for continuing those who are addicted to smoking to service their addiction legally; but, with many teenagers still likely to start smoking even if this prohibition is passed, potential problems regarding the fairness of the above asymmetry become even weightier.

What are your thoughts on this?

*Although this signifies only the beginning of the lobbying efforts and public debate concerning this proposed paternalistic intervention, it is worth noting that the BMA has been successful in its efforts to ban both smoking in public placesand smoking in cars carrying children, following similar votes in 2002 and 2011 respectively.



Fay Niker

Fay is a PhD student in political theory at the University of Warwick, working on the ethics of behaviour change policy. In general, her work focuses at the intersection of psychology, public policy, and political philosophy.



Does systemic injustice justify Robin Hood Strategies?


Taxing FIFA: Asides from having their board room modelled on the war room from doctor strangelove, are there any other reasons as to why FIFA should be taxed?


  1. Hi Fay, thanks for this post! Let me raise two questions. The first concerns the notions of autonomy and paternalism used here. Can individuals make a conscious and well-informed decision to opt for smoking, and to forego additional years of life? Would they make a mistake that the state should correct (that would mean that the argument is based on some objective, rather than subjective, notion of what the right choice would be)? I don't know if there is data available on how smokers think about their addiction, but would that make a difference to the argument at all? I assume that one of the problems here is adaptive preferences, which makes it harder to judge the problem, but prima facie I would assume that there can be at least *some* adults who make a fully-informed, autonomous choice to smoke, so there would definitely be a reduction of freedom for them (which may or may not be outweighed by other arguments).
    My second question concerns your last paragraph. Surely there are areas where the fact that there will be some violations of a law does not imply that the law should not be passed. For example, it will never be possible (or only at an unrealistically high price) to completely ban murder. But that does not mean that we should not have a law against murdering innocents. So this alone cannot be a reason against a having a law against smoking. Nonetheless, one often hears this demand-creates-its-own-supply-whether-legal-or-not-argument, also in debates about prostitution or about having caps on interest rates (and it played some role in the discussion we had about maximum rents for flats). Could you maybe clarify how exactly we should understand this argument in the context of anti-smoking-policies?

  2. Thanks very much for the post, Fay. I have a question that has some similarities to Lisa’s first query above. Essentially, I was hoping you could elaborate how you see your arguments applying to individuals at different ages across their lives. One thing that seems striking about the BMA proposal is that it targets, in the long-term, people across all ages. That is, although the group who will be affected immediately are presently under 18, they will, as you note, be over 18 at some juncture, but still banned from purchasing cigarettes. I wondered how, if at all, that difference in age affects your autonomy case. Some of the points you make concern protecting individuals pre-18 (or some relevant age of autonomy) from factors that influence the voluntariness of their choices (e.g., peer pressure). But I guess that those who advocate autonomy-based arguments will find such worries less compelling as a case for restricting the actions of those who are post-18 and hold the relevant information and competency. In this case, might they be more comfortable with tighter age restrictions, say along the lines of the challenge-25 model for purchasing alcohol, than the prohibition proposal? You mention that smoking also has the problem of addiction undermining autonomy. Such a concern looks relevant to protecting autonomy of those post-18. But many decisions we make as adults have effects (albeit, perhaps, of different kinds) on our long-term autonomy, ranging from diet to career choice. So, if we were to harness that line, would we extend it to some of these other spheres? I mean these thoughts only as possible examples. The broad question to which they are intended to speak is: could you say a little more about the specific application range of the different autonomy cases and whether they will support the BMA proposal or, rather, something closer to other mechanisms already used in relation to smoking, alcohol, and so forth?

  3. I really liked this post Fay, thank you. I don't know enough about the psychology of smoking, but anecdotal evidence tells me that often smoking is a way of coping with stress. I don't claim that it is a good, i.e. mid- or long-term effective, way, or that its benefits compensate its costs – yet it is one readily available to smokers. If so, it seems to me that both the principled justification of a ban and its practical one will depend, to some extent, on giving alternative coping mechanisms to those who are subject to this ban. On the principled side, I assume one can appeal both to welfare and autonomy-enhancing lines (since being at a loss in coping with stress will undermine agency.) On the practical side: if it is true that (some/many) people smoke to cope with stress, and if this coping strategy is taken away from them with no replacement: (a) the law will be likely to be broken and result in unintended criminal activity as you point out and (b) more importantly, other, even more destructive, coping mechanisms may evolve.

  4. Thanks for this Fay. One thing that makes me uneasy with the BMA's proposal is that it creates a division between citizens born before and after 2000. So in terms of the law you would have (a) pre-2000 citizens and (b) post-2000 citizens. Post-2000 citizens would I think have a legitimate complaint against not having the same rights as pre-2000 citizens, since its basically an arbitrary distinction.

    Additionally I worry how the division plays out socially, since you could have groups of friends of very similar ages where some of them are allowed to smoke (<1999) and some aren't (>2000). And that kind of division wouldn't be temporary (like between the 17 year olds who can't buy alcohol and the 18 year olds that can), it would be permanent so that 63 and 64 year old friends in 2063 would still be unable to legally smoke together.

  5. This is a problem you have for many political decisions; for example, retirement rules change by cohort, and so a very small temporal difference can decide about what kind of benefits you get. I see that the social circumstances of smoking might make a difference here, but then isn't there a danger that if you make this argument too strong, you can *never* introduce a change into these legal areas.

  6. You're right Lisa, I didn't mean that's its a decisive argument against all policies that affect cohorts differently. But I think smoking is a particularly visible social activity that sets up a visible divide between people. Having different pension schemes (within certain limits) doesn't divide a group as obviously as smoking (though perhaps that reflects my age preferences!).

  7. I should add that I don't think that pension differences are unimportant (or even justifiable). Only that smoking is a particularly public division between people, in a way that pensions are not. Then again pension differences (like wage differences) create so many other public and more serious social divisions (where you live, eat, and hang out) that perhaps there isn't as much of a difference.

  8. Firstly, I must apologise for my slow response to your comment. Thanks for the helpful questions raised. On the first point, my sense is that people *can* make a conscious and well-informed decision to smoke and that, therefore, there are *some* fully-consenting adults for whom such an intervention would count as a reduction of freedom (though this does depend somewhat on how we are defining these terms due to issues such as hyperbolic discounting). There is some data stating that about 90% of smokers have tried to give up smoking at some point, which offers some more 'subjective' grounds for intervention; the interesting question, as you point out, is figuring out how these two points weigh in against other arguments in play. On the second point, my point was simply a practical one, namely, that this policy would be unlikely to stop what already goes on (minors starting to smoke even though they cannot legally purchase cigarettes) and that, the fact that they would not be able to then service their addiction legally would in fact worsen the personal and societal issues associated with smoking. It was meant more as a feasibility constraint than a stand-alone reason against the proposed ban.

  9. Thanks Andrew. Again, please accept my apologies for my late response. Firstly, I would just point out that the paragraph on the autonomy-based argument in favour of the ban is simply my attempt to present what I think is the thrust of BMA's position, rather than my own views. My own sense is that something more akin to the challenge-25 model would, as you suggest, be a better way to go. I'm not completely convinced by your point that those who advocate autonomy-based arguments would necessarily be less worried about voluntariness in smokers over 18; I think the key issue is at what age (and under what conditions) they first *started to smoke*, because those who are autonomy-minded would still be concerned by the smoking practices of informed, competent adults who took up smoking pre-18 precisely because of the issue of addiction – these adults may (now) know many of the (theoretical long-term) risks of smoking and may even want to give up because of this knowledge, because may not have sufficient motivational resources to overcome their addiction. This is likely to be different for those who started smoking post-18. I like your (slippery slope) point about other things that affect our long-term autonomy: is it that the addiction aspect of smoking is relevant here too? The addiction is what makes you smoke, and it also has longer-term autonomy-undermining effects (so it seems to undercut autonomy in two distinct ways); this doesn't seem to be the case for something like career choice, where only the longer-term effects would be felt. Perhaps, also, the distinction between cost and difficulty (Gerry Cohen) is relevant here?

  10. Thanks, Anca. This is a very helpful line of thought. So, unfortunately I also don't know enough about the psychology of smoking; one thought, though, might be that a relatively small proportion of smokers actually *start* smoking for stress-related reasons, but that – over time – this regular physical and often social activity offers a time to leave behind what one is doing and to 'relax' for a few minutes. Could it be, though, that a lot of what was stressful beforehand was the physical need to smoke (for those with a nicotine addiction), which is then relieved by smoking a cigarette? If this is the case, then smoking would be to cope with the stress that is (partially) caused by smoking itself. Whether this is the case or not, though, the justification of the ban would be partially dependent upon the alternatives that would be available and easily accessible. (The NHS offers a number of different options through its smoking cessation clinics, etc.). What type of alternative coping mechanisms do you have in mind?

  11. Thanks, Bruno. Yes, I agree that this arbitrarily-set asymmetry is probably the most disturbing aspects of the proposal. I'm interested in how you think this weighs against the other reasons for and against the proposal – would the reduced rights of post-2000 citizens (about a quarter of whom would want to smoke, from current averages), along with the reduced freedom for that section of smokers for whom smoking is a fully-informed and autonomous decision, be able to overweigh the overall (subjective and objective) welfare benefits, as well as the proposed autonomy-promoting effects of anti-smoking-policies of this kind?

  12. Also, how does the argument relating to the provision of alternative coping mechanisms relate to those in the two separate groups, i.e., pre-2000 and post-2000? Do we owe more to one group than the other with regards to this?

  13. I think the welfare benefits could outweigh the reduced freedom of some of the post-2000 citizens. But I think the effect of the social divide is more serious and outweighs the welfare benefits.

Leave a Reply

Powered by WordPress & Theme by Anders Norén