The problem of unequal treatment

Imagine cigarettes were invented today. Should their sale be allowed? It is often said, with good reason, that regulatory agencies would never allow them onto the market.10 Put simply, the extreme health risks and highly addictive nature of smoking stand in an unfavourable relation to the potential benefits for consumers. This thought experiment suggests that a society without cigarettes would be preferable to our status quo. But even granting this claim, not every attempt to ban cigarettes can be morally justified. Withdrawing the right to purchase cigarettes from existing smokers, for example, might inflict so much suffering on them as to make this approach unjustifiable. The idea of a tobacco-free generation bypasses this worry by granting current consumers the right to continue smoking, and simultaneously withholding it from those who have not yet picked up the habit.

However, this proposal also faces an important ethical challenge, namely that of unequal treatment. This challenge has been pressed against various concrete proposals to introduce a generational ban. For instance, one observer of the (unsuccessful) Tasmanian plan to prohibit the sale of cigarettes to those born after 2000 argued as follows:

 There is no rational reason why, at January 1, 2030, a 30-year-old person can buy cigarettes, but a 29-year-old person (perhaps born only an hour earlier) cannot. Both are otherwise competent adults, but only one is legally allowed to buy tobacco. The legal distinction has no public health basis and is solely concerned with their age.11

Commenting on the more recent UK government’s plans, another found:

The ban infantilises one cohort of adults, discriminates on the basis of age and raises issues of intergenerational unfairness. Even people who have no intention of ever buying tobacco may feel discriminated against if someone a few years older than them has a right that has been withheld from them.12

In other words, the worry is that the proposal would involve the unequal treatment of people who are similarly situated in every respect except the morally arbitrary fact of their date of birth. This might distinguish it from other forms of unequal treatment that are commonly accepted. Some rights and liberties are age-differentiated, such as the right to drive. Some depend on choices and behaviours; for example, the right to vote may be withheld from some competent adults if they have been convicted of serious crimes. The generational ban, on the other hand, appears harder to justify. By withholding liberties from some people based merely on when they were born, the ban may discriminate, stigmatise, dominate or otherwise disadvantage or disrespect them.

Proponents of the ban have rejected this claim. Yvette Van der Eijk and Gerard Porter, for example, argue that the unequal treatment of birth cohorts is innocuous because it has ‘a sound and reasonable basis: to protect future generations from the health effects of active and passive smoking’.13 After all, many policies that treat people unequally—say, policies that offer some types of routine cancer screening only to those above a certain age—can be justified if there is a reasonable and objective justification. Similarly, the claim here would be that the generational ban does not involve objectionable unequal treatment because the goal it pursues is important enough to justify it.

The general point is true enough, but it is simply assuming what needs to be shown: namely that the policy is justified when the considerations for and against it are weighed up. We know that on one side of the ledger, the proposal promises to dramatically reduce deaths, morbidity and welfare losses. Given the unequal socio-economic distribution of smoking in society, it could also reduce social inequalities in health and longevity. These are significant distributive outcomes. But the open question is whether they outweigh the presumptive moral costs of unequal treatment. Clearly, even the best distributive outcomes must sometimes be passed up if attaining them requires violating key moral principles. Hence, we need a better sense of precisely what, if anything, is wrong with unequal treatment.

In recent decades, philosophical discussions of inequality have increasingly focused on the nature and quality of social relations. According to relational egalitarianism, the hallmark of a just or good society is the ability of its members to interact and relate to one another as equals. This view suggests that we should be concerned not just with distributive outcomes, important though they may be, but also with how people are treated by others and by social structures, which may be even more crucial.i A particular focus for egalitarians should thus be to avoid oppressive relationships: domination, marginalisation and stigmatising differences in status.14–17 Historically, it has often been these forms of inequality that have motivated struggles for social justice, including the fights against inherited class hierarchies, caste systems, sexism or racism. Relational egalitarians therefore contend that we should be wary of introducing policies that threaten to undermine respectful and egalitarian relations among individuals.

This perspective compels us to scrutinise the proposal for a generational ban, and in particular the way it treats members of different birth cohorts unequally. To be sure, the ban’s broader goals of improving population health and reducing health inequalities are valued by relational egalitarians.18 But prohibiting an activity for some individuals and not others, based on seemingly arbitrary factors, might create power imbalances, signal a form of disrespect or otherwise undermine egalitarian relations. For this worry to be warranted, however, we need a better understanding of what it means to treat individuals unequally—an idea that can carry many different meanings. In what follows, I examine various dimensions in which the generational ban may run afoul of egalitarian commitments. Once we get a better understanding of these various concerns, however, it becomes clear that the ban need not threaten the ideal of a society of equals.

The denial of moral agency

Egalitarians insist that people have equal status: nobody has higher or lower moral worth than anyone else. One way to deny this status is to question the moral agency of a person. ‘All competent adults are equally moral agents’, Elizabeth Anderson writes, ‘everyone equally has the power to develop and exercise moral responsibility, to cooperate with others according to principles of justice, to shape and fulfil a conception of their good’.14 Applied to the generational ban, the objection might thus be that depriving some adults of liberties that others enjoy—being able to buy cigarettes—misrecognises their moral agency. The ban would be disrespectful and demeaning because it paternalistically assumes that some people are not able to form a valid conception of the good, so that we must intervene on their behalf.19

Now, a possible reply to this worry is to deny that paternalistic interventions are always disrespectful.ii Few people believe that seatbelt laws or the prohibition of duelling insult their moral agency. In some circumstances, intervening for a person’s good may even be required by a commitment to equal respect, namely when not doing so would express insufficient concern for the intended beneficiary’s inherent moral worth.20 For example, not attempting to stop a suicidal person from drinking bleach seems callous and uncaring. A society that normalises such indifference would hardly reflect fraternal relations.

But even if we thought that paternalism is always disrespectful, this worry does not specifically single out a problem of unequal treatment. It expresses a general rejection of paternalism that would presumably hold even if the ban were applied universally, that is, if it did not discriminate between those born before and after the generational cut-off date. So the objection cannot be simply be a restatement of the familiar concern that paternalism is disrespectful. Instead, it must be that the policy in question singles out a specific social group and expresses disrespect by associating it with negative attributes, namely a supposed lack of competency in self-regarding matters. The ban infantilises the birth cohorts who are not trusted to look after their own health, or so the objection.

However, this objection wrongly assumes that the justification of the ban requires seeing members of certain birth cohorts as less competent than others. The proposal is not based on the idea that their judgement is worse, their rational capacities inferior or that they are any less able to independently form a life plan than those who came before them. In fact, it is precisely the feature that is sometimes held up to demonstrate the discriminatory nature of the proposal—the arbitrariness of the generational cut-off date—that shows that the generational ban is not premised on the assumed inferiority of one group to another. Either groups are arbitrary, or they track some independent notion of worth, but both things cannot be true at the same time. Rather than a particular view of certain birth cohorts, then, what is driving the proposal are the benefits of banning cigarettes, combined with the fact that a universal ban is hard to justify to existing smokers.

An analogy: many countries introduced seatbelt laws in the 1960s, mandating seatbelts in newly registered cars, but exempting (‘grandfathering’) older vehicles. This meant that throughout the 1960s and 1970s, some drivers had to buckle up and others did not. Did this express a disrespectful attitude specifically towards new drivers? No—a reasonable observer could hardly attribute offensive meaning to this form of unequal treatment.iii Note that my argument here is not that seatbelt laws or smoking bans are justified, all things considered. I am simply making the more modest claim that if they can be justified, then the differential treatment involved in phasing them in need not express an anti-egalitarian attitude, or a disrespectful view of any group’s moral agency.

Discrimination

Discrimination is a paradigmatic violation of the ideal of a society of equals. Even if the generational ban does not deny the moral agency of those whose freedom to smoke it restricts, it may still discriminate against them by denying them freedoms that others enjoy. And this may be the case even when the affected individuals do not intend to make use of these freedoms. At its worst, the generational ban may create a discriminatory, two-tiered society—or as the former British Prime Minister Boris Johnson put it, a ‘smoking apartheid’.21 However, discrimination is a complex concept: not all forms of unequal treatment are discriminatory, or at least not wrongfully so.

Who can be a victim of discrimination? On one influential view, discrimination is unequal treatment based on (perceived) membership in a socially salient group, that is, a group that affects social interactions in many social contexts.22 Race, class or gender are socially salient, and people are often at risk of being oppressed, marginalised or otherwise disadvantaged on account of these characteristics. Eye colour, in contrast, is not socially salient in most societies. Is being part of a birth cohort more like race, or more like eye colour? It is not clear that belonging to a particular birth cohort as such structures our social interactions. What is true, of course, is that age is socially salient: being young, middle-aged or elderly shapes the way we interact with others. Age is therefore a protected characteristic in many legal systems. Following from this, one could perhaps argue that the generational ban is a form of ageism against the young, who will be immediately affected by it. But this is not quite accurate: because birth cohorts eventually progress through all age groups, from the young to the old, the two categories do not overlap except contingently. At some future point, the birth cohorts affected by the ban will be middle-aged and elderly. Hence, the ban cannot simply be a case of ageism against the young.iv

On another possible view, prominent in American discrimination law, unequal treatment warrants special scrutiny when it is based on immutable traits such as race or ethnicity23 (pp. 22–23). Being born into a birth cohort is an arbitrary fact that cannot be changed, and thus, perhaps, a candidate for discrimination. However, not all forms of discrimination are based on immutable characteristics (eg, discrimination against people with certain political convictions), and not all forms of differential treatment based on immutable characteristics are discriminatory (eg, denying driver licenses to blind people). Hence, the burden of proof seems to lie with those who argue that differential treatment based on this specific immutable characteristic is discriminatory.

In any case, claims of discrimination are typically made only when differential treatment involves some form of moral wrong. Theories about the wrong-making feature of discrimination tend to fall into two main camps: disrespect-based and harm-based views. Among the former, some locate wrongness in the intention to deny a person’s equal moral status,24 or demeaning attitudes expressed in the discriminatory act.25 Because I argue throughout the paper that the generational ban need not involve these disrespectful features, let me focus on accounts that centre on harm or disadvantage.22 Critics of the ban claim that the affected birth cohorts are disadvantaged by losing the freedom to legally purchase cigarettes. However, even if this is true, it does not necessarily mean that they are disadvantaged, all things considered. Proponents of the ban argue that the loss of freedom is greatly outweighed by the health, material and possible autonomy gains that future birth cohorts will enjoy by never being exposed to the risk of becoming smokers.8 Given the high proportion of smokers who wish they had never started smoking, these benefits cannot simply be set aside when determining whether the ban is harmful. As such, at least one influential version of the discrimination objection hinges on claims that are difficult to establish.

Stigmatisation

Suppose that birth cohorts are not socially salient groups. Do we not make them so by treating them unequally? Compare: even if people with green eyes do not form part of a socially salient group that structures their social interactions, they would arguably become one if we denied them the right to vote. Plausibly, this would happen because disenfranchisement would stigmatise them. Hence, we need to consider whether denying the right to smoke would stigmatise birth cohorts affected by the generational ban, and thus create inegalitarian differences in status. By marking them as inherently different and subject to special restrictions, we might risk perpetuating a sense of exclusion and inferiority among the affected cohorts.

But while unequal rights can be stigmatising, much hinges on the social meaning of the rights in question. Stigmatising someone means, roughly, to treat them in negative ways by associating them with a characteristic that is seen as undesirable according to community norms.26 Voting is socially prized and losing this right would be socially undesirable. But the same is arguably not true of smoking: it does not seem plausible to say that being unable to buy cigarettes would be stigmatising. Indeed, if anyone has reason to fear community norms, it is those in older birth cohorts who continue to buy cigarettes—not those who are legally unable to do so. As the policy takes effect, smoking would gradually disappear from society as a ‘norm cascade’ is set in motion.27 It is therefore far-fetched to suggest that the inability to smoke expresses norm deviance, while simultaneously denormalising the act of smoking.

Now, perhaps the worry is that the generational ban will stigmatise existing smokers, who (unlike entire future birth cohorts) often belong disproportionately to socially disadvantaged groups in terms of class, race or ethnicity. By denormalising smoking, we would further marginalise these groups. This is a different type of unequal treatment than the one we have been considering so far—that is, not between birth cohorts, but between socioeconomic or racial groups—though of course no less important. But it is worth noting here that there is a strong egalitarian rationale for the generational ban, given that other, less restrictive tobacco control measures (such as higher taxation or workplace smoking bans) harbour the same risks of stigmatisation while potentially adding additional burdens on socially disadvantaged groups.28 Hence, a generational ban promises greater health equity among future generations, and reduced stigma among disadvantaged groups. In New Zealand, for instance, Maori leaders have been outspoken proponents of the tobacco-free generation proposal, suggesting it is seen at least by some socially disadvantaged groups as the most promising means to reduce racial and socioeconomic health disparities.29,v

But there is an important caveat. Things might look different if the generational ban is not fully effective, and some people born after the cut-off date end up smoking by buying cigarettes on the black market. Trade in illicit substances is dominated by organised crime, which threatens to associate consumers with socially condemned behaviour. If particular social groups, such as ethnic minorities or people of lower socioeconomic status predominate among these ‘illicit’ smokers, there is genuine risk of stigmatisation and marginalisation. At worst, it can also generate risks by exposing disadvantaged communities to racist law enforcement practices.30 However, this risk depends on how precisely the generational ban is implemented and enforced. It is significantly lower if criminal penalties apply only to sellers (as opposed to consumers) of illicit cigarettes.

Therefore, while the generational smoking ban harbours some potential for stigmatisation, especially among smokers from disadvantaged backgrounds, these worries hinge primarily on the practicalities of implementation and enforcement. But there is no principled reason why a tobacco-free generation should give rise to stigmatising inequalities.

Unequal recognition of interests

Let us consider another interpretation of what it means to treat people as equals. This view focuses on treating relevant interests equally. Clearly, we would fail to recognise the equal moral standing of citizens if we held that some people’s interests count for more when making collective decisions. As Samuel Scheffler puts the point:

In a relationship that is conducted on a footing of equality, each person accepts that the other person’s equally important interests—understood broadly to include the person’s needs, values, and preferences—should play an equally significant role in influencing decisions made within the context of the relationship.16

By taking away the option to smoke from some birth cohorts but not others, we may be impacting on their interests in unequal ways. Imagine, for example, that we decided to deny the option to buy cigarettes to half of all current smokers, based on some arbitrary characteristic (say, whether their surname begins with a letter in the first half of the alphabet). We would be frustrating the important interests of this group of smokers, assigning them less weight than those of the unaffected smokers, and thereby expressing unequal status. The question, then, is whether the differential treatment of birth cohorts raises the same worries.vi

However, what matters here is the content and strength of people’s interests. Taking away a highly prized option affects a person’s interests more than taking away an option they value little. Hence, it is possible that a policy treats two groups’ interests equally, while having differential effects on them. It would indeed be a problematic policy that banned cigarettes for half of all existing smokers, because many of them are likely to have a strong preference for being able to continue to smoke. But in the tobacco-free generation proposal, we assume that people in the affected birth cohorts will not have developed preferences for smoking in the first place, because they will never have had the opportunity to do so (this, again, will depend on the effectiveness of the proposal). In other words, we affect interests much more strongly by withdrawing an option from someone who has always enjoyed it and built their life plan, identity or community around it, than by withholding the same option from someone who has never had it.10

Now, of course, one might have an interest in something one has never experienced: an abstract interest. For example, someone who has never been able to go on holiday may still have a strong interest in being able to do so. We need not deny that there can also be an abstract interest in being able to smoke cigarettes. For example, some might derive pleasure from the soothing effect of nicotine. Especially among socially disadvantaged groups, nicotine may also provide an accessible way to cope with stress and adverse life circumstances.31 However, the question is whether for future birth cohorts, these abstract interests can outweigh the interests that would be advanced by a ban, including health, material and autonomy benefits. The thought experiment of cigarettes being invented today and failing to gain regulatory approval strongly suggests that the latter interests are stronger. This seems even more plausible under versions of the proposal in which other ways to consume nicotine (such as e-cigarettes) remain available.32 To claim that the ban disregards the interests of future generations assumes, controversially and implausibly, that the interest in smoking would always carry the day.

What about the interests of existing smokers? If the ban really has such great benefits, it may seem unfair not to extend them to those born before the generational cut-off date. There is some truth in this objection: it is important, from the point of view of equal treatment, that existing smokers in older birth cohorts are not (perceived to be) abandoned by public health efforts. This requires tobacco control efforts beyond the generational ban, such as cessation support or incentives. But given that some existing smokers strongly wish to retain the option to smoke, for reasons outlined above, a ban across the board cannot be said to advance their interests, as conceived by these smokers themselves. In short, we can maintain a commitment to treating people as equals even when we treat them differently based on the content and strength of their respective interests.

Political equality

Consider now a final way in which the generational ban may express an objectionable form of unequal treatment, one which is reflected in the process through which it comes about. On the face of it, the ban is an imposition of one set of birth cohorts on another, without the latter being able to have a say in the matter. This may be thought to conflict with a key requirement of a society of equals, namely democratic decision-making on equal terms. There are different ways to flesh out this worry. Some relational egalitarians emphasise an ideal of democratic citizenship in which people participate as equals in the political and civic sphere.14 15 Neo-republican theorists emphasise that free citizens should not be in relations of domination to other citizens or the government. This requires laws and arrangements subject to equally shared, popular control.33 34 The worry, then, is that the imposition of a cigarette ban on younger and future birth cohorts violates these commitments to political equality.

But once again, it is not clear that the generational ban falls foul of these commitments. Consider first equal participation rights. It is worth noting at the outset that the ban does not threaten political participation rights as such, for instance, the ability to protest, publish one’s opinions or stand for election. In other words, the freedom to smoke is not itself politically salient. More plausibly, the worry is that the ban comes about in a way that reflects unequal participation: the birth cohorts affected have had no direct electoral influence on the decision to implement the generational ban. However, there is nothing unusual about this form of unequal participation. It occurs whenever the effects of policies and laws outlast present generations. Indeed, the birth cohorts that decide to introduce the ban will, at some point in the past, have been subject to decisions that they themselves had no say in. Strictly speaking, what egalitarianism requires is that all birth cohorts enjoy equal democratic participation rights over their entire lifetimes, and not that all age groups exert equal influence at any given time.35 36

Of course, not everything goes. There are genuine cases of intergenerational domination where one set of birth cohorts exercises arbitrary power over another. For example, imagine a scenario where one birth cohort could unilaterally decide to enhance their own generous state pensions by reducing the pensions of those who come after them. Surely, the possibility of such decisions, which exploit the vulnerability of future generations, undermines the ideal of political equality among birth cohorts. What would make the exercise of power over future generations non-dominating? For neo-republicans, it is crucial that democratic institutions embody values such as inclusive deliberation, checks and balances, and shared control. Philip Pettit, in particular, emphasises the importance of ‘contestability’, meaning that decisions should be subject to challenge and revision by those affected.33 The goal is thus to design institutions that include younger birth cohorts and future generations in these processes. For example, even disenfranchised children should enjoy real opportunities for contestation (consider the recent School Strike for Climate movement). There should also be modes of representation that protect the interests and rights of future generations. Although it goes beyond the scope of this article to flesh out these ideas, there is no principled reason why legislation for a generational ban could not emerge from such non-dominating institutional arrangements. Note, in particular, that problems of intergenerational domination loom largest when the present generation can impose irreversible decisions on those in the future (think of burying nuclear waste, cutting down primary forests or, perhaps, gutting pension systems). Legislation for a tobacco-free generation is not like this. If it proves unpopular among future birth cohorts, the sale of cigarettes can be re-introduced. Future generations remain in charge of their political destinies.