II. The Prospective Argument

In this section, I present The Prospective Argument for age-differentiated paternalism in two steps. First, I illustrate that paternalism is considered easier to justify the more good it promotes (including the more harm it prevents) for the individual interfered with. Second, I defend the view that the benefits of paternalism to the individual generally diminish with the individual's age.

It is a widespread view that the justification of paternalism is sensitive to considerations pertaining to the significance of the interests at stake. For example, Godwin speaks of ‘the paternalistic calculus’, which says that

Using a similar calculus, other authors argue that paternalistic measures that involve limited infringements with people's autonomy seem to be warranted when these are essential to preventing irrevocable situations of insufficiency (Pedersen and Midtgaard Reference Pedersen and Midtgaard2018; see also Casal Reference Casal2007: 322).Footnote ¹⁰

Although paternalism is often considered to be expressive of disrespect, some theorists have argued that such expressive or respect-based considerations sometimes speak against anti-paternalist inaction and in favour of interfering with the voluntary choices of competent agents (Hojlund Reference Hojlund2021; Pedersen Reference Pedersen2021). The problem with remaining passive when other people jeopardize their own interests is that, at least under certain conditions, it manifests ‘neglect, indifference, and unwillingness to give appropriate weight to the strong interests of others’ (Hojlund Reference Hojlund2021: 519). In this context, one of the relevant conditions is the significance of the good at stake, or as Hojlund (Reference Hojlund2021: 528) puts it, ‘it is arguably more expressively objectionable if the state through inaction fails to secure a very important good than if it fails to secure a rather trivial good’.

It might be argued that the above arguments are only relevant to people who accept hard paternalism. According to hard paternalism, it is a good and relevant (not necessarily decisive) reason in favour of interfering with a competent person's voluntary choice, without the person's consent, that such interference promotes the person's own interests or well-being (see e.g. Feinberg Reference Feinberg1986: 4; Hanna Reference Hanna2018: 4). By contrast, soft paternalists reject that this reason is ‘good or relevant’. In their view, interference with a person's conduct with the motive of promoting the person's own interests or well-being can be justified ‘when but only when that conduct is substantially nonvoluntary, or when temporary intervention is necessary to establish whether it is voluntary or not’ (Feinberg Reference Feinberg1986: 12). Accordingly, irrespective of the severity of the interests or well-being at stake, soft paternalists reject that paternalistic interference with a competent person's voluntary choices can be justified.

However, even soft paternalists rely on a certain ‘paternalistic calculus’. Specifically, they understand voluntariness as a matter of degree. On their view, paternalistic interference is never warranted above the voluntariness threshold, whereas they do not reject such interferences below the relevant threshold. Where to set the threshold depends on considerations pertaining to the good at stake. As Shafer-Landau (Reference Shafer-Landau2005: 171) puts it, ‘[t]he greater the threatened self-harm, the higher the probability of such harm eventuating, the more irrevocable the results of a choice, the greater the voluntariness required for legal immunity’. Accordingly, interventions that prevent high levels of harm will be easier to justify on soft paternalistic grounds than interventions that prevent lower levels of harm. The upshot is that both hard and soft paternalists consider interventions easier to justify the more is at stake (see also Feinberg Reference Feinberg1986: 117–24; Grill Reference Grill2020: 16–17; Le Grand and New Reference Le Grand and New2015). It follows, for example, that seatbelt mandates in cars are easier to justify than seatbelt mandates in trains if the former prevent more harm to people using seatbelts than the latter. To be clear, this applies regardless of the age of those being treated paternalistically.

The next step in the argument for age-differentiated paternalism is to defend the view that the benefits to a person of preventing the person's imprudent activity generally decrease with the person's age.Footnote ¹¹ This view finds support if we consider certain arguments in the literature on priority setting according to which it is permissible to give higher priority to treatments that primarily benefit young people. There are different justifications for this view. Due to the above-described focus on the net benefits to the subject of a paternalistic act, to begin with, I will focus on justifications that cite age-related differences in the magnitude of expected benefits. In this context, Bognar (Reference Bognar2015: 253) asks us to consider the following case:

If we think that we should save A rather than B, one possible justification of this is that saving A promotes more good for A than saving B does for B because of the differences in additional life expectancies between A and B. In this context, it might be argued that life expectancy as such is not a good in itself. According to this view, ‘lifetime is like a “container,” whose value depends on what it will allow for—that is, on what lies “within the container” (life events, activities, projects, etc.)’ (Fleurbaey and Ponthiere Reference Fleurbaey and Ponthiere2022: 357). If this is true, additional life expectancy is only an indicator of the good involved in increasing the duration of the life of the individual (see also Bognar Reference Bognar, Nagel and Lauerer2016: 166). But the conclusion in favour of giving the drug to A still stands, because a longer lifespan generally leaves more room for valuable ‘life events, activities, projects, etc.’ than a shorter lifespan. In other words, the magnitude of the good in question depends on both the quality and quantity of the extra life-years that the person gains (in this case, if she gets the drug). I refer to this as the Individual Good involved in benefitting the person, for example, through increasing the duration of the person's life. The Individual Good is the good that an intervention (e.g. giving the medicine) does the person.

Because B will only live for a few more years if she gets the drug, whereas A will live for many years, the Individual Good of treating B is presumably less than the Individual Good of treating A. Accordingly, a focus on promoting Individual Good as much as possible supports giving the drug to the 20-year old person.Footnote ¹² Similarly, because of differences in additional life expectancies, a straightforward justification of age-differentiated paternalism is that preventing an imprudent activity, e.g. smoking or driving a motorcycle without a helmet, performed by a young person generally promotes more Individual Good than preventing the same imprudent activity by an older person.

It might be objected that, in contrast to the above cases, most paternalistic policies do not necessarily save lives; they may merely prevent diseases, financial losses or other sorts of non-fatal harms. While I grant that the argument seems strongest in situations where the possible consequences of an imprudent action are significant and either irreversible or long-lasting, it is not only applicable in ‘life or death’ cases. Imagine, for example, a person who hits their head and must live with a chronic or long-lasting headache. Say that this accident decreases the person's quality of life to a certain extent each life year for the rest of the person's life. The Individual Good involved in preventing this calamity will be greater the more years the person has left to live.

To summarize the argument of this section: it is a widespread view in the literature on paternalism that paternalism is easier to justify the more good it promotes for the individuals interfered with. Since paternalism targeting young people generally protects more life years than paternalism directed at older people (because the additional life expectancies of young people are generally longer than the additional life expectancies of older people), age seems to be a proxy for the Individual Good promoted by paternalism. It follows that paternalism towards young people is generally easier to justify than paternalism towards older people.Footnote ¹³

However, the above argument for prioritizing young people in scarcity decisions, which focuses on differences in how treatment benefits individuals, has implications that some are unwilling to accept. The argument implies that young people should be prioritized, but it also implies that other groups who might benefit more from treatment (e.g. healthy or privileged people) should be prioritized. This seems unfair. In the next section, I discuss whether The Prospective Argument for age-differentiated paternalism is vulnerable to a similar objection.

III. An objection to The Prospective Argument for age-differentiated paternalism

Now, I will present and discuss an objection to the argument for prioritizing healthcare to young people in scarcity decisions that I described in the previous section. This objection may also threaten The Prospective Argument for age-differentiated paternalism. Both arguments use age as a proxy for additional life expectancy (which is an indicator of the good promoted for the individual), but additional life expectancy not only varies with age – it also seems to follow other traits, such as socio-economic status. This challenges the argument for prioritizing healthcare to young people that focuses on where the greatest benefits can be achieved. For example, Nielsen (Reference Nielsen2021) asks us to consider a prioritization case like the one suggested by Bognar above, but now we must choose between:

In this case, saving C (henceforth referred to as ‘Privileged’) rather than D (henceforth referred to as ‘Disadvantaged’) involves the greatest Individual Good. According to Nielsen (Reference Nielsen2021: 232), the implication that treatment to Privileged should be prioritized because this would generate more benefits than prioritizing Disadvantaged ‘disturbs our intuition of fairness as it pro tanto implies discriminating based on class’ (Nielsen Reference Nielsen2021: 232). That is, prioritizing Privileged because Privileged benefits more from the treatment does not seem to be a fair response to the case.

Is The Prospective Argument for age-differentiated paternalism vulnerable to a similar objection? It follows from the argument that paternalizing Privileged is more justified than paternalizing Disadvantaged (all else being equal, e.g. assuming that the burdens associated with the intervention are the same for Privileged and Disadvantaged). Is this implication of The Prospective Argument problematic from a fairness perspective? One way to assess this is to consider whether there are any relevant differences between the two arguments (the health-care prioritization argument and The Prospective Argument) that render the differential treatment of Privileged and Disadvantaged unfair in the health-care context, but not in the paternalism context.

One potentially relevant difference is scarcity. The argument for prioritizing healthcare to young people suggests how we should distribute healthcare under scarce resources. In contrast, The Prospective Argument in favour of age-differentiated paternalism does not seem to focus on how scarce resources should be distributed between individuals. At least in standard situations, the paternalistic agent does not have to choose between interfering paternalistically with, e.g. the young and the elderly, for scarcity reasons. In fact, it is often argued that policies that prevent people from making imprudent decisions will save others in society from the costs associated with treatment, emergency aid, etc. (see e.g. Bou-Habib Reference Bou-Habib2006; Le Grand and New Reference Le Grand and New2015: 60; Stemplowska Reference Stemplowska2009: 252–53). However, sometimes, there may be a limit to how many people a paternalist can interfere with. We can imagine a situation where a paternalist must choose between protecting Privileged or Disadvantaged (e.g. for practical reasons). As far as scarcity is concerned, this case is relevantly similar to the case where we can only provide treatment for one of the two patients. And in such a situation, The Prospective Argument tells the paternalist to protect Privileged.

A seemingly more relevant difference between the two arguments is that The Prospective Argument for age-differentiated paternalism is about an intrapersonal value conflict that is not pertinent in the health-care prioritization argument. Specifically, The Prospective Argument focuses on whether paternalism's beneficial effects for the interests or well-being of the individual justify interfering with the individual's autonomy. According to the argument, a paternalistic intervention towards a young person is likely to benefit this young person comparatively more than a similar intervention towards an older person would benefit this older person. For this reason, the aim of promoting the person's own interests or well-being is, all else being equal, more likely to justify unwelcome interference with the person's autonomy the younger the person is (although some paternalistic interventions towards the older person may also be justified – and some interventions for the young would be unjustified).

Because of the intrapersonal value conflict involved in paternalism, one might challenge the view that paternalizing Privileged, but not Disadvantaged, is problematic from a fairness perspective. There is a potentially relevant difference between, on the one hand, prioritizing treatment to Privileged and, on the other hand, paternalizing Privileged but not Disadvantaged. In the former case, both Privileged and Disadvantaged want the treatment, and they both seem to have a claim to receiving the treatment. In the latter case, both Privileged and Disadvantaged prefer their situation without the benevolent interference (this follows from the definition of paternalism). Accordingly, it is not clear that they have a claim to being treated paternalistically. And whether Privileged and Disadvantaged have a claim to being paternalized seems relevant to the question of whether it would be unfair to paternalize Privileged but not Disadvantaged. Consider, in this context, Broome's (Reference Broome1988: 61) description of fairness:

On this view, given that both Privileged and Disadvantaged have an equal claim to the medical treatment (or given that Privileged has a weaker claim because Privileged already has more than their fair share), it is unfair that only Privileged receives the treatment, even if this is what generates most benefits. Yet, unless Disadvantaged has a claim to paternalistic protection, it is not unfair to Disadvantaged if only Privileged is paternalized.

Since hard paternalism involves acts or policies that interfere with competent people's voluntary choices, without their consent, it may seem strange to say that people have such a claim to paternalistic protection. At least, it is often invoked as an objection to hard paternalism that those who introduce such paternalistic policies, or in other ways act paternalistically, fail to discharge their duty to respect the autonomy of the people interfered with (see e.g. Enoch Reference Enoch2016; Feinberg Reference Feinberg1986; Groll Reference Groll2012).Footnote ¹⁴ Based on the corresponding claim to have their autonomy respected, it is rather those who are subjected to paternalism who can object that they have been treated unfairly. It follows that, if someone is treated unfairly in the scenario where Privileged but not Disadvantaged is treated paternalistically, it seems to be Privileged.

But this is too quick. As Broome puts it, ‘[s]ometimes there can be a conflict of duties owed to a single person: the duty to be truthful may conflict with the duty to spare her distress, for instance. These duties we weigh against each other in deciding what to do’ (Broome Reference Broome1988: 62). Similarly, the duty to respect a person's autonomy sometimes conflicts with a duty (that at least some authors think we have) not to stand idly by when another person is about to act in a way that could lead to serious harm to the person concerned (Hojlund Reference Hojlund2021; Pedersen Reference Pedersen2021; Pedersen and Midtgaard Reference Pedersen and Midtgaard2018). Given that such a duty (and a corresponding claim) exists, this duty must be weighed against the duty to respect the autonomy of the person interfered with (cf. Broome Reference Broome1988: 62).

Now, consider a situation where Disadvantaged and Privileged both act imprudently and it applies to both that their claim to paternalistic protection outweighs their autonomy claim against being paternalized. Suppose also that a paternalist must choose between interfering with Privileged or Disadvantaged. The paternalist decides to paternalize Privileged because this will benefit Privileged more than paternalism towards Disadvantaged would benefit Disadvantaged. Since the paternalist focuses on protecting as much Individual Good as possible, Disadvantaged has no chance of being selected. Accordingly, the paternalist fails to discharge the duty not to stand idly by owed to Disadvantaged. Because of this, the disadvantaged person, who also has a claim to paternalistic protection, is treated unfairly in this case.

To summarize and conclude the discussion in this section: The Prospective Argument implies that paternalism towards Privileged is more justified than paternalism towards Disadvantaged. That is, the argument suggests differential paternalistic treatment of Privileged and Disadvantaged. The question is whether this implication is unfair towards Disadvantaged. The conflict between respecting individual autonomy and catering to the individual's well-being or interests is pertinent in the paternalism context (and less so in the health-care prioritization context). This complicates the fairness considerations pertaining to The Prospective Argument for age-differentiated paternalism. Whether it is unfair to paternalize Privileged, but not Disadvantaged, plausibly depends on the relative strength of the two conflicting claims that each of them has. In situations where both Disadvantaged and Privileged have a claim to paternalistic protection, which outweighs their autonomy claim against paternalistic interference, it is unfair towards Disadvantaged if only Privileged is treated paternalistically because this is what generates most benefits.

This shows that fairness and The Prospective Argument sometimes pull in opposite directions. It also shows that paternalism is not only about balancing conflicting intrapersonal values; it is also about mediating the claims of different people (cf. Broome Reference Broome1988: 61).Footnote ¹⁵ In the next section, I propose an additional argument for age-differentiated paternalism that accommodates the latter interpersonal considerations. This argument does not appeal to the amount of Individual Good to the person interfered with. Instead, it focuses on the moral value associated with promoting this person's interests or well-being from a fairness perspective.

IV. The Retrospective Argument for age-differentiated paternalism

Consider a variant of the priority case where two patients need a life-saving drug, but you only have one dose of the drug to give. Now, you must choose between:

In this case, there are no differences in additional life expectancies, which means that a consideration to promote as much Individual Good as possible will not give us reason to prioritize A over B. This seems wrong. At least in the literature on priority setting and age, many will consider it unfair if the 70-year-old patient is prioritized (see e.g. Bognar Reference Bognar, Nagel and Lauerer2016: 166; Kappel and Sandøe Reference Kappel and Sandøe1992: 313–14).

There are different possible explanations of this view.Footnote ¹⁷ For example, treating the 70-year-old in this case might be considered unfair because of the inequality between the two patients. The young patient has had less quantity of life and thus less time to pursue and realize their life goals and plans. For example (depending on their specific age, of course), young people have had less time to get an education and establish themselves in the labour market, and they have had less time to start a family, to have children, to have grandchildren, etc. For such reasons, the young patient is disadvantaged compared to the older patient who has lived for more years. Other things being equal, it does not seem implausible to suspect that most of us would prefer (i) living until age 70 and then dying due to lack of treatment over (ii) getting the drug at age 20 and then living until age 30 (Kappel and Sandøe Reference Kappel and Sandøe1992: 314). Unless the quality of the older patient's 70 years of life has been very poor compared to the young patient's 20 years of life, increasing the young patient's life with 10 years involves decreasing the inequality between the two patients.

To address such inequalities, it is sometimes suggested in the literature on health-care priority setting that treatments are allocated in a way that give some priority to the worst-off individuals (see e.g. Bognar Reference Bognar2015: 259–60; Bognar Reference Bognar, Nagel and Lauerer2016: 167). For example, it is possible to assign weights to benefits. The weight that is assigned to benefitting a person depends on how much worse off the person is and the degree of inequality aversion that is built into the chosen model (cf. Fleurbaey and Ponthiere Reference Fleurbaey and Ponthiere2022: 375).Footnote ¹⁸ We can describe this as the Social Good promoted by benefitting the person, for example, through increasing the duration of the person's life. In the above case, since the younger patient is worse off, increasing the younger patient's life with 10 years involves a greater Social Good than increasing the older patient's life with 10 years.Footnote ¹⁹

Again, something similar can be said about paternalistic cases of harm prevention. A paternalistic policy that prevents a person with a life expectancy of 30 years from dying at age 20 seems to involve a greater Social Good than an equivalent policy that prevents a person with a life expectancy of 80 from dying at age 70, because the 20-year-old has had less quantity of life (and therefore, other things being equal, are worse off in terms of lifetime well-being than the 70-year-old). Given that the benefits of interfering with the young person are similar to the benefits associated with interfering with the older person, there are equally strong Individual Good promoting reasons for interfering with the two persons. However, if equalizing lifetime well-being via paternalism is a meaningful goal, this is an alternative Social Good promoting reason for age differentiation.

Moreover, the inclusion of Social Good considerations equips us to address the challenge to The Prospective Argument raised in section III: it does not follow from The Retrospective Argument for age-differentiated paternalism that paternalizing Privileged is easier to justify than paternalizing Disadvantaged. For example, The Retrospective Argument seems to support the well-known policy of raising the prices of cigarettes, since this policy has a stronger effect among members of low-income groups who are more responsive to prices compared to members of high-income groups (Sassi et al. Reference Sassi, Belloni, Mirelman, Suhrcke, Salti, Vellakkal, Visaruthvong, Popkin and Nugent2018). If raising the prices of cigarettes primarily benefits people with low incomes, this is to be considered an advantage of the policy.Footnote ²⁰ For example, as Chapman (Reference Chapman2012: 3) puts it, ‘[t]hose groups advocating keeping tobacco tax low perversely seek to “help” poor smokers by keeping tobacco affordable, which encourages use’.

Since the Retrospective Argument focuses on promoting Social Good rather than Individual Good, one might argue that The Retrospective Argument provides a defence of age-differentiated paternalism that eliminates the paternalism involved in the policies and actions under consideration.Footnote ²¹ While paternalism aims to promote the interests or well-being of the people interfered with, The Retrospective Argument is concerned with priority or inequality aversion. Specifically, The Retrospective Argument supports policies that protect the interests or well-being of those who have less because this would produce a greater Social Good. One may suggest that policies introduced with this (e.g. equality-enhancing) motive are not paternalistic.

But, as Bengtson and I have recently argued, whether this is true depends on why we must promote equality to begin with. One view is that inequality matters because it involves that people's interests are not equally protected. If one then introduces a policy to ensure a better protection of the interests of those who are worse off relative to others, this rationale seems to be of a benevolent kind (see Bengtson and Pedersen Reference Bengtson and Pedersen2023: 5–6).Footnote ²² Given that the policy also involves unwelcome interference with the intended beneficiaries, the policy seems to be paternalistic towards these people. Imagine, for example, a state that raises the prices on cigarettes because the state knows that this policy has particularly beneficial effects for ‘poor’ smokers and thus is likely to promote equality. Imagine also that the intended beneficiaries (smokers from low-income groups) prefer their situation without the policy in question. This equality-enhancing policy is a case of unwelcome benevolent interference with smokers from low-income groups, i.e. the policy is plausibly described as paternalistic towards these smokers (Bengtson and Pedersen Reference Bengtson and Pedersen2023: 5–6).Footnote ²³

Now, given that it is both true (i) that paternalism is easier to justify the greater good it promotes for the individuals interfered with (cf. The Prospective Argument), and (ii) that paternalism is easier to justify the worse off the intended beneficiary is, in absolute terms or relative to others (cf. The Retrospective Argument), it is not clear that a combination of the two arguments would deem paternalizing socially privileged people easier to justify than paternalizing socially disadvantaged people. In the case from section III where we must choose between Privileged and Disadvantaged, the two arguments point in different directions. Paternalizing Privileged involves more Individual Good, but less Social Good. The opposite is the case if Disadvantaged is paternalized. These values for Privileged and Disadvantaged may exactly offset each other in an overall evaluation, or they can tip to one side depending on the specific sizes.

In the case of age-differentiated paternalism, the two arguments point in the same direction: it is generally the case that young people benefit more from paternalism than older people, because young people generally risk losing more years of life or must live with the consequences of their imprudent decision for a longer time. In other words, the Individual Good promoted by paternalizing young people is generally greater than the Individual Good promoted by paternalizing older people. Moreover, compared to older people, young people are worse off in the sense that they have had less time to live than older people, which means that the Social Good promoted by paternalizing young people is also generally greater than the Social Good promoted by paternalizing older people.

However, this does not preclude the existence of other relevant factors. For example, suppose that some activities involve greater risks for older people because older people have a harder time recovering from, say, a crash with a motorcycle.Footnote ²⁴ In such cases, it is not unlikely that paternalizing the older person promotes more Individual Good than paternalizing the younger person.Footnote ²⁵ However, paternalizing the young person may still involve more Social Good. In such cases, whether or not paternalism towards the young person is still more justified than paternalism towards the older person for non-competence-based reasons depends on the differences in the Individual Good and the Social Good promoted by paternalizing each of them (which in turn depends on the degree of priority or inequality aversion built into the theory).