Media headlines in the UK are widely reporting Rishi Sunak’s announcement of a proposal to ban smoking for younger generations. Under the proposal, the legal age of smoking would increase by one year every year so that, eventually, no-one would be able to buy tobacco.

The proposal has proved to be controversial, and it has prompted a number of different arguments. This is unsurprising; the proposal represents a classic conflict between individual well-being, liberty, and third-party interests. As the BBC reports, some commentators have also highlighted an apparent inconsistency in Sunak’s own position, since he recently pushed back part of the government’s anti-obesity strategy, because of “people’s right to choose”. Again, the BBC reports that Sunak’s own response to this consistency argument has been that there is an important difference between the two policy positions, because ‘there is no healthy level of smoking’, whilst one can enjoy unhealthy foods as part of a healthy diet.

However, the claim that there is ‘no healthy level of smoking’ can be used to respond to this consistency argument and support the proposed smoking ban in quite different ways. Whether we support or oppose the proposal, it is crucial to be clear about the precise moral arguments that both supporters and opponents are making.

One useful way to begin is by thinking about whether or not the proposed ban is paternalistic.

Paternalism and Mill’s Harm Principle

Paternalism can be understood as an interference with another person, against their will, in order to promote that individual’s own interests.

This last clause is important; not all restrictions of freedom are paternalistic on this definition, because we might sometimes restrict an individual’s freedom in order to prevent them from harming others, and not because doing so will promote their own interests. For instance, we might non-paternalistically restrain someone who is behaving violently towards others.

Indeed, a staunch anti-paternalist can support John Stuart Mill’s Harm Principle, outlined in his book On Liberty, which states:

The only purpose for which power can be rightfully exercised over any member of a civilized community, against his will, is to prevent harm to others.

So, one initial question we can usefully ask about the smoking proposal is whether it is being justified in paternalistic or non-paternalistic terms; is it intended to protect smoking individuals from their own choices, or is it intended to protect non-smokers from the choices of smokers? Or is the justification meant to be some mixture of the two? Let’s first look at what arguments are relevant to the non-paternalistic justification.

Non-Paternalistic Justifications

Mill’s harm principle is often invoked in discussions of public health interventions. However, considered in itself, it leaves a number of important questions unanswered about the sorts of interferences that might be justified on its basis.

One crucial question we should begin with is ‘What sort of harm to others might the proposed smoking ban prevent?’ One relevant harm is the direct health risks to others that arise as result of passive smoking, or ‘second-hand’ smoking. Notably, smoking is already banned in the UK in most enclosed public spaces, and this ban was largely justified by appeal to the harms of passive smoking.

Another relevant harm to consider is the pressure that smoking-related health needs put on the NHS – if we reduce the prevalence of smoking, then the thought might be that this will reduce the burden on an over-stretched healthcare system that non-smokers might also need to use. On this reading, the problem with there being ‘no healthy level of smoking’ is that anyone who smokes is thereby harming others by somewhat increasing the expected burden on a shared healthcare system.

However, establishing that individual choices might pose some risk of harm to others does not alone entail that it would be justifiable to interfere with those choices in order to prevent that harm. We also need to consider some other questions that the harm principle alone does not answer.

The first is whether the interference will be effective in preventing the harm in question. Notably, some commentators in the media have suggested that a complete ban on smoking would lead to a black market for cigarettes. The likely effectiveness of the prohibition in reducing the prevalence of smoking is ultimately an empirical question; however, it is a crucial issue for the non-paternalistic justification.

Let’s assume that a prohibition would be at least somewhat effective in reducing the number of people smoking. The second question we should then ask is whether an outright prohibition is necessary for achieving this goal. Are there other things we can do to reduce the harms of passive smoking and the pressures on the healthcare system that involve lesser interferences with liberty? Finally, we can ask is whether the prohibition would be proportionate; is the harm to others that we are seeking to prevent sufficiently grave to outweigh the harms involved in the interference of liberty that the prohibition involves?

In answering the last two questions, it is crucial to be clear about which harm to others the prohibition is intended to prevent. Consider first passive smoking. Notice that a complete prohibition of smoking would prevent at least some people from smoking in environments where they pose no risk to others of passive smoking; for instance, someone smoking alone in the privacy of their own home. It seems hardly necessary to prohibit such a person from smoking in order to prevent the harms of passive smoking; we might sensibly ask whether a less restrictive prohibition (such as banning smoking in outdoor and indoor public places) would be sufficient to achieve the aims of the proposed outright ban. Moreover, since the existing ban on smoking in enclosed public spaces has significantly reduced the harms of passive smoking in the most high-risk environments, it is far from clear that the additional benefit of lowering the risk even further with a complete ban would be proportionate, given the significant interference of liberty the prohibition involves.

What about the harms that widespread smoking poses to the healthcare system in general? There are again important empirical questions to address in establishing the extent of this harm, and whether a prohibition would be an effective and proportionate measure to prevent it. But we should also attend to the question of necessity. In addition to considering other less intrusive ways we might use to reduce the prevalence of smoking itself, we should also consider whether there are other ways in which we might seek to reduce stress on the health care system. Naturally, this leads to much broader policy questions about how a public healthcare system should be financed, and how such policy should balance considerations of liberty, equality and responsibility for healthcare need.

Paternalistic Justifications

We have seen that the non-paternalistic justification of the prohibition is far from straightforward. But Sunak’s comments can be read more paternalistically – perhaps the main problem with there being ‘no healthy level of smoking’, is that anyone who smokes is making a choice that is against their own interests, and we should interfere with such choices.

Such an interference would be paternalistic according to the broad definition I outlined above, and staunch anti-paternalists may simply claim that no such form of paternalism can be justified. If that is so, then questions about effectiveness are somewhat moot – no matter how effectively a prohibition might promote an individual’s interests, it may not be justifiable to interfere with liberty just to achieve that goal.

But there are different forms of paternalism, and some might be more palatable than others to those who are inclined towards weaker anti-paternalist views. Again, it is therefore important for both supporters and opponents of the proposal to be clear about the nature of the paternalistic justification at stake.

One distinction we can draw is between soft paternalism and hard paternalism. The former only permits interference with an individual’s involuntary choices, whilst the latter permits interference with an individual’s voluntary choices. Soft paternalism is sometimes claimed to be less problematic because it does not amount to an interference with real liberty, or the kind of liberty that we really value. Indeed, Mill himself in On Liberty suggests that if we saw another person attempting to cross a bridge which had been ascertained to be unsafe, we might seize him and turn him back, without any real infringement of his liberty, if we believe the person to be ignorant of the danger they are in. Ignorance, as Aristotle observed in Book 3 of the Nicomachean Ethics, can undermine the voluntariness of our actions.

Similarly, one question we might ask about the smoking prohibition is whether it might admit of a soft paternalist justification. In order for such a justification to succeed, one would have to show that individuals who choose to smoke are not making sufficiently voluntary decisions. Notice that in order to avoid the consistency argument, one would also have to establish that smoking behaviour is less voluntary than unhealthy eating behaviour. This is, I believe, a tall order – however, it is here where debates about the reasons why and when different people start smoking are perhaps of most moral relevance.

A second distinction we can draw is between strong and weak paternalism. Strong paternalism claims that we may interfere with people’s choices in order to enable people to achieve better goals than the ones they are choosing for themselves. In contrast, weak paternalists claim that it is only legitimate to interfere with the means that agents choose to achieve the goals they themselves want to prioritise in their own lives, if and when those means are likely to be unsuccessful.

So, if a person has the main goal of living a long and healthy life, then smoking is plausibly a poor means for achieving that end – and a weak paternalist will claim that it is permissible to interfere. Conversely, suppose someone wants to live a life that focuses on the short-term pleasure of nicotine consumption over other components of the good life; here, the strong paternalist may claim that we should intervene, whilst the weak paternalist will not.

Weak paternalism is often understood to be a more palatable form of paternalism because it at least acknowledges the moral significance of allowing individuals to set their own goals; it thus might plausibly honour the value of global autonomy, even if it involves overriding local autonomy. Notably, in her book Against Autonomy, Sarah Conly mounts a weak paternalist argument in favour of what she calls coercive paternalist public policy measures, including smoking prohibitions. One question for the weak paternalist is whether we can accurately judge which goals other people want to prioritise; if there are good reasons to think that some people will not prioritize longevity over the pleasures of smoking, then the weak paternalist justification will be more difficult to mount.

Concluding Remarks

I have intentionally not set out my own views on the proposed smoking prohibition in this blog. I have also not attempted to be exhaustive about all of the morally relevant factors in this debate.

Instead, my main concern has been with the clarity of some the arguments that can be raised in this context. If we are not clear about dialectic role that different considerations might have on this debate, the most likely outcome is that opponents and critics simply will end up talking past each other. My own view is that in order to develop coherent policy here, opposing sides should at least agree about precisely where they are disagreeing.