Article Text

Download PDFPDF

How low can you go? Justified hesitancy and the ethics of childhood vaccination against COVID-19
  1. Stephen David John
  1. History and Philosophy of Science, University of Cambridge, Cambridge, UK
  1. Correspondence to Stephen David John, History and Philosophy of Science, University of Cambridge, Cambridge CB2 3RH, UK; sdj22{at}


This paper explores some of the ethical issues around offering COVID-19 vaccines to children. My main conclusion is rather paradoxical: the younger we go, the stronger the grounds for justified parental hesitancy and, as such, the stronger the arguments for enforcing vaccination. I suggest that this is not the reductio ad absurdum it appears, but does point to difficult questions about the nature of parental authority in vaccination cases. The first section sketches the disagreement over vaccinating teenagers, arguing that the UK policy was permissible. The second section outlines a problem for this policy, that it faces justified vaccine hesitancy. The third section discusses three strategies for responding to this problem, arguing that there may be no simple way of overcoming parents’ reasons to resist vaccinations.

  • COVID-19
  • child
  • ethics- medical

Data availability statement

No data are available.

This article is made freely available for personal use in accordance with BMJ’s website terms and conditions for the duration of the covid-19 pandemic or until otherwise determined by BMJ. You may use, download and print the article for any lawful, non-commercial purpose (including text and data mining) provided that all copyright notices and trade marks are retained.

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Should we vaccinate 12–15 year-olds against COVID-19? The UK’s Joint Committee on Vaccination and Immunisation (JCVI) said not, but the UK government adopted this policy anyway. This gap between experts and policy-makers has created controversy.1 Such controversies will become more heated given proposals to lower the age range for vaccination further.2 This paper uses this case to explore the ethical issues around offering COVID-19 vaccines to children. My main conclusion is rather paradoxical: the younger we go, the stronger the grounds for justified vaccine hesitancy and, hence, the stronger the practical argument for enforcing vaccination. Some readers may think this counts as a reductio ad absurdum of the project of offering vaccines to children. I am uncertain this is true, but I do suggest that childhood vaccination raises unfamiliar questions about parental authority and ethical obligation.

Prudence and offering vaccines

From January 2021, the UK government pursued mass vaccination against COVID-19, initially targeting older, ‘more vulnerable’ groups.3 In September 2021, the JCVI—an arm’s length body which makes evidence-based recommendations the Secretary of State is statutorily obliged to consider4—responded to proposals for extending the vaccination programme to younger age groups

The health benefits from vaccination are marginally greater than the potential known harms. However, the margin of benefit is considered too small to support universal vaccination of healthy 12 to 15 year olds at this time… Given the very low risk of serious COVID-19 disease in otherwise healthy 12 to 15 year olds, considerations on the potential harms and benefits of vaccination are very finely balanced and a precautionary approach was agreed.

I interpret this statement in terms of whether vaccination is in the ‘prudential’ interests of younger teenagers. Vaccination is in an individual’s prudential interests when the expected value to the individual of being vaccinated is greater than the expected value of not. For example, the "triple" vaccine against measles, mumps and rubella (MMR) is in the prudential interests of nearly all infants because it reduces a risk of severe disease at the cost of a very small risk of minor side effects.5

Using this concept, we can interpret the JCVI’s slightly convoluted conclusion in terms of two claims. The first is an ethical claim: it is impermissible to offer vaccinations to individuals when they are not in those individuals’ prudential interests. This claim can be understood as stemming from a ‘non-maleficence’ principle, according to which it is wrong to impose risks of medical harm on individuals in the absence of countervailing greater chances of benefit.6 The second is more complex: we should act as if the vaccine is not in younger teenagers’ interests. We can understand this ‘prudential interests claim’ as also grounded in a concern about ‘non-maleficence’: given uncertainty around whether vaccination is in teenagers’ prudential interests and the importance of avoiding causing harm, policy-makers should ‘assume the worst’. (I return to this reasoning in the third section.)

Brusa and Barilan set out two reasons for vaccinating children against COVID-19: that this does clearly benefit children—for example, through reducing school absences (in effect, denying the prudential interests claim); that vaccination benefits other members of the community (in effect, denying the ethical claim).7 Both sorts of concerns were raised against the JCVI’s recommendations.8 Ultimately, the UKs' Chief Medical Officers recommended going ahead with childhood vaccination.

Was this the right decision? If vaccination is clearly in children’s prudential interests, as the Chief Medical Offiers seemed to imply, presumably so. More interesting is whether it could be the right decision even if vaccination is not in children’s own prudential interests, but is in the interests of others in the community. Should we hold the ethical claim?

I suggest that, as well as prudential interests, children have ethical interests—interests in acting in ways which avoid harming, or which promote the well-being, of others. Given that vaccination affects one’s likelihood of transmitting a virus, then we each have an ethical interest in being vaccinated. Note that I assume that this claim is true regardless of whether we have an all-things-considered enforceable obligation to vaccinate, much as it is true we have an ethical interest in giving money to charity even if we lack an enforceable obligation to do so. Asking people to donate money to charity is permissible, even if this is not in their prudential interests, because donation is in their ethical interests. So, too, I suggest, when people have some ethical interest in being vaccinated, then it is permissible to offer them a vaccination.

Giubilini et al seem to dispute this claim, arguing that teenage vaccination programmes for COVID-19 are impermissible because they treat children as a ‘mere means’.9 I suggest that this objection is misplaced. While I agree that forcing children to get vaccinated for the sake of adults might treat them as a ‘mere means’, it does not follow that offering a chance to get vaccinated does the same. Compare two cases: in the first, I force you to donate to charity, in the second I ask you to donate. In both cases, there is a sense in which my actions—forcing and asking—treat you as a means to the end of charitable fund raising. However, in the first case I treat you as a mere means, because I bypass your will, whereas in the second by leaving the choice with you, I respect your ethical agency. Of course, as I discuss below, issues are more complicated when we are dealing with children, but we should not assume that all cases where we instrumentalise others are tantamount to treating them as a mere means.

The problem of justifiable hesitancy

When we offer vaccines, people are sometimes reluctant to accept. Unfortunately, vaccine hesitancy—that is, ‘delay in acceptance, or refusal, of vaccines despite availability of vaccination services’ (Eskola et al)—is particularly likely in our case.10

To explain, note a gap between why we (legitimately) want people to get vaccinated and how we encourage them to get vaccinated. We introduce both vaccination programmes and cancer screening programmes on the grounds that uptake will improve population health. However, we have reasons to care about vaccine hesitancy more than screening hesitancy. Public policy typically is, and ought to be, guided by a Millian principle according to which we should not interfere with individuals’ autonomous self-regarding medical choices unless and until those choices threaten harm to non-consenting third parties.11 Someone who does not get screened might be imprudent (arguably), but her choices threaten no harm to others. By contrast, someone who does not get vaccinated threatens others.12 Therefore, encouraging vaccination is both ethically more pressing and more easily justifiable than tackling screening hesitancy.

Nonetheless, even if our primary reason to care about vaccination is ethical, typically the reasons we offer are prudential: we encourage people to vaccinate because this is good for them. Consider responses to initial fears about the risks of myocarditis associated with the COVID-19 vaccine. These focused on explaining that, overall, the risks from contracting COVID-19 were higher than the risks of clotting associated with the vaccine, not that vaccination is the ethical choice.13

This gap is not necessarily problematic. Consider a hypothetical analogous case. Every day, Jan stands outside your office window for his cigarette break, annoying you. You think Jan has good ethical reasons to stop, but he is selfish. So, one day, you explain to him that smoking risks his health. In offering Jan a prudential reason to do something which, ethically speaking, he ought to do anyway, you may treat Jan as a means to an end, but this action seems permissible just insofar as what you say is true and your intention is to ensure he performs the ethically correct action. What would be problematic would be to deceive Jan—by telling him smoking is bound to kill him—hence denying his agency, treating him as a ‘mere’ means.14

Of course, even when vaccination is in individuals’ prudential interests, selling vaccination in these terms can be difficult.15 However, communicating about the COVID-19 vaccine for adolescents is even more complex, because we all have excellent reasons—contained in an official expert report, no less—to believe the vaccine is not in teenagers’ prudential interests. A General Practitioner complained that the JCVI’s decision ‘has led to confusion and mistrust and is feeding into vaccine hesitancy’.16 This is puzzling, insofar as the JCVI’s report implies that at least some members of the community have reasons to be vaccine hesitant!

I will say that vaccine hesitancy among some group is ‘justified’ when there are strong reasons to think that vaccination is not in the prudential interests of that group. Unfortunately, even when hesitancy is ‘justified’, it might still be problematic from the perspective of promoting population health, because vaccinations play a role in stopping transmission. So, I assume, as I argued in the first section, that it is permissible to offer a vaccine even if it is not in individuals’ prudential interests (at least, as assessed in ‘medical terms’), because we each have an ethical interest in being vaccinated. How, then, might the State legitimately encourage vaccination under these circumstances?

Broadly, there are three strategies in response to this problem:

  1. To argue that, despite the JCVI’s claims, the vaccine is, in fact, in 12–15 year-olds’ interests (ie, to argue that, although vaccine hesitancy is justified given the JCVI’s statement, it is not justifiable all things considered).

  2. To explicitly appeal to ethical reasons to vaccinate.

  3. To change the ‘costs’ and ‘benefits’ associated with vaccination such as to make the previously justified hesitancy no longer justifiable.

As a matter of fact, policy has tended to muddle along, relying on public fears, demonisation of ‘anti-vaxxers’ and vague appeals to civic duty. As ethicists, however, we should ask how we should respond, both to assess policy and as a way of grappling with larger ethical issues. Specifically, I will now argue that thinking through these options reveals a deep problem with adolescent vaccination.

Responding to hesitancy

Before going on, note a complexity: typically, consent to childhood vaccination is not given by the child, but a surrogate (typically, a parent). Therefore, in asking about hesitancy, we need to ask about both the child’s prudential and ethical interests and about parental obligations when choosing for children. This can get complex. Therefore, in discussing each option, I will first ask how we might convince a competent child to consent to vaccination, before relating these justifications to the adult’s choice.

Ambiguous interests

The first possible response to justifiable vaccine hesitancy is to tell younger teenagers that, despite the JCVI’s apparent claims otherwise, vaccination is, in fact, in their prudential interests. There are two options here: one is to say, as Brusa and Barilan do, that if we look beyond medical interests to include, for example, mental health effects of missing school, vaccination is in children’s all-things-considered prudential interests; the second is to dispute the JCVI’s assumption that we should reason as if vaccination is not in children’s interests. The first response is clearly sensible, but simply raises the question of whether keeping children away from school is justified.17 Here, I focus on the second option.

The JCVI’s statement both acknowledges that benefits are ‘marginally greater than the potential known harms’, but then claims that we should assume the opposite. We can spell out the brief account of this reasoning given in the first section more fully as follows: the evidence about both the potential costs and benefits of vaccination is highly uncertain, ambiguous and provisional. As such, we are in a situation of ‘ambiguity’: where our evidence suffices to assess that the true probability of some future event lies within some range, but not to choose any particular estimate.18 So, for example, while we might be relatively certain that the probability of suffering a side effect is in some range, we lack sufficient epistemic grounds to choose any estimate in that range as ‘the’ correct decision-relevant probability. I interpreted the JCVI’s statement as claiming that, grounded in its distinctive concerns to avoid doing harm, it adopted ‘worst case’ estimates, implying vaccination is not in teenagers’ interests. (In turn, this ‘do no harm’ concern might be grounded on a more fundamental ethical principle or on practical grounds, such as maintaining confidence.)

However, even if the JCVI can be justified in adopting the ‘worst case’ probability, it can be equally reasonable, given the same evidence, to assume that vaccination is in children’s interests.19 Indeed, we might even argue that, when making choices about future health, it is rational to use less conservative estimates. So, we may legitimately be able to tell younger teenagers that the situation is not as dire as it seemed: given the JCVI’s data, and a less extreme degree of risk aversion, getting vaccinated is probably in their interests.

This is an elegant solution to the problem of what we can say to competent decision-makers. Unfortunately, it faces a serious problem when we think about the case of parents choosing for children. Plausibly, in virtue of their duties to protect their children, parents have an obligation to be risk averse in considering evidence about their child’s best interests.20 If so, even if it is true that, were she choosing for herself, an individual can (maybe even should) assume that vaccine is in her interests, it does not follow a parent, using the same evidence, must assume the same when choosing for her child. Even if I would be justified in taking some gamble with my health, it does not follow I should be willing to take the same gamble with my child’s health.

Going ethical

The second possible solution to justifiable vaccine hesitancy is to concede that vaccination is not in younger teenagers’ interests, but to encourage vaccination by appealing to ethical considerations (much as we might encourage you to donate to charity out of concern for others). Indeed, vaccination is often presented as a civic duty. It is, of course, controversial whether we have an obligation to vaccinate or not. Note, however, that all I assume is that, even if we do not have an ethical obligation to vaccinate, we have ethical reasons to do so.21 As such, offers of vaccination simply enable children to act as ethical agents.22

Even before considering the parental perspective, there are two worries about direct appeals to ethical principles in response to justifiable vaccine hesitancy. First, this strategy carries ethical risks: presenting vaccination as an ethical duty can hide the tensions between the interests of the young and of the old in pandemic responses.23 Second, it simply displaces problems of hesitancy: just as there can be disagreement over the prudential benefits of vaccination, so, too, there can be reasonable disagreement over the grounds, nature and extent of ethical reasons to vaccinate. Therefore, there might be as much ‘ethical hesitancy’—hesitancy based on a disagreement about ethical interests—as hesitancy based on disagreement about prudential interests.

Furthermore, even if children have ethical reasons to get vaccinated, it does not necessarily follow parents have an obligation to vaccinate them. Consider an analogy: maybe it is true that a child should not eat stolen food. In general, a parent should enable the child to meet this ethical interest. However, a parent also has an obligation to do what is best for her child. Plausibly, this might justify the parent in stealing food to give to her child. Even if children have ethical interests—or even strict obligations—in getting vaccinated, their parents may have other duties, such that they should not help the child fulfil her ethical interests.

Changing the cost/benefit calculus

The third option to respond to justified vaccine hesitancy is to shift the prudential costs and benefits of vaccination, making it the case that previously justified hesitancy is no longer justified. For example, systems of ‘vaccine passports’—or even more restrictive attempts to encourage or even mandate vaccination—are often proposed as tools to encourage vaccination.24

However, any such schemes can only be justified if they do not violate ethical constraints. For example, even if the function of a vaccine passport is to encourage vaccination, such a restriction on freedom of movement can be considered only because it seems a fitting or proportionate response to the risks of harm the unvaccinated impose on others. Even if ‘screening hesitancy’ depresses population health, it would be inappropriate to respond to such hesitancy by denying the unscreened access to cinemas.

If I am right that children have ethical reasons to vaccinate, then there is some justification for responding to hesitancy with measures which drag prudence and ethics back together. These responses solve the problem discussed above of parents’ obligations to their children, because they ensure that parents have no choice, based on their child’s interests, but to vaccinate. Still, they seem to work only by creating unpalatable trade-offs.

How low can you go?

In short, vaccinating 12–15 year-olds is puzzling. Regardless of their prudential interests, there are good ethical reasons why members of this age group ought to vaccinate, grounded in the risks they otherwise pose to others. So, there are good reasons to offer them vaccinations. This is a familiar situation. Giving to charity might be against my prudential interests, but in my ethical interests; it might be wrong to force me to donate, but fine to encourage me to donate. What is hard about our case is that the people we must encourage are not necessarily children themselves, but their parents. In turn, parents might have excellent prudential and ethical reasons to resist these entreaties. The only way in which to ensure childhood vaccination campaigns succeed in their aims might, then, be to make them (quasi)mandatory. This seems strange. Such policies do look like they might treat children as a mere means; furthermore, as Williams has argued, they seem to generate an inequitable distribution of the risks of vaccination.25 Therefore, some readers might choose to read this conclusion as a reductio ad absurdum of the entire project of childhood vaccination against COVID-19.

My conclusion is, however, more cautious. Rejecting childhood vaccination out of hand is to ignore the possibility that, in not vaccinating, children (non-culpably) threaten others. What we face, however, is a question about parental authority. Consider the contrast with parental hesitancy towards the MMR vaccine.26 In that case, we face an ethical problem because we believe that parents who are reluctant to vaccinate their children act against the child’s prudential interests; we have a question about whether the harm they risk causing their children justifies over-riding parental authority.27 In the COVID-19 case, by contrast, parents who believe that the vaccine is not in their child’s best prudential interests are very likely justified. Rather than the familiar problem of whether we can over-rule parental authority on grounds of the child’s best interests, we have the problem of whether we can over-rule parental authority on grounds of the child’s obligations. Unfortunately, we lack any theory of how to resolve this dilemma and, hence, any account of how low we can go.

Data availability statement

No data are available.

Ethics statements

Patient consent for publication

Ethics approval

This study does not involve human participants.



  • Contributors SDJ wrote the paper entirely himself, and acts as guarantor.

  • Funding The authors have not declared a specific grant for this research from any funding agency in the public, commercial or not-for-profit sectors.

  • Competing interests None declared.

  • Provenance and peer review Not commissioned; externally peer reviewed.