I. Introduction
Voluntary euthanasia (VE) is the intentional shortening of a patient's life by a doctor at the patient's request in order to end the patient's suffering. Physician-assisted suicide (PAS) is the intentional assistance by a physician in a patient's suicide in order to confer the same benefit. Whether either or both should be legalised is one of the most consequential and controversial questions in contemporary medical law and ethics. Lord Judge has described it as “the great moral and legal problem of our times”.Footnote 1
Proposals to permit some form of “assisted dying” are a regular feature of the legislative programme at Westminster. The Assisted Dying Bill, introduced by Baroness Meacher, is the latest in a series of bills to permit PAS for capacitous and “terminally ill” patients. The Second Reading debate took place in the Lords on 22 October 2021.Footnote 2 Several attempts have also been made over the past twenty years to persuade the courts to uphold a right to PAS or VE,Footnote 3 attempts that, like the bills, have attracted enormous media attention.
One argument advanced against changing the law to permit VE is that there is a logical link between VE and non-voluntary euthanasia (NVE) – the intentional shortening of the life of a suffering patient who lacks the capacity to request euthanasia. This logical argument (LA), an argument that focuses attention on the underlying moral principles thought to justify VE, is obviously of importance for legislators or judges who are invited to liberalise the law. Judges in particular are well versed in ascertaining the underlying principles on which arguments are based and in determining where acceptance of those principles would logically lead.
Though the LA tends to be invoked by those opposed to VE, it takes no position on the merits of VE. It simply runs that acceptance of VE logically requires acceptance of NVE. The argument could therefore equally be used by a supporter of VE to show why the law should permit not only VE but also NVE. The LA should, therefore, be of importance to all those engaged or interested in the debate, whether or not they support VE in principle. Surprisingly, however, the argument has not received anything like the attention it deserves: it did not, for example, feature in the eight hour Second Reading debate on the Meacher bill, despite frequent references to euthanasia laws overseas. Proponents of legalisation generally ignore it, perhaps because they fear that drawing attention to it would play into the hands of opponents, and opponents have tended to focus on the practical risks of slippage involved in legalising VE. Several prominent scholars of medical law and ethics have, however, addressed the LA, either directly or indirectly, and have rejected it or at least not accepted it. They are: Gerald Dworkin and Roger Frey; Margaret Battin; John Griffiths; Hallvard Lillehammer (in this Journal); Stephen Smith; Robert Young; Emily Jackson; and Steven Daskal. Their responses to the LA merit careful consideration for if they are persuasive a significant buttress of the case against legalising VE falls away. This is the first paper to evaluate the responses of these scholars to the LA.
The paper proceeds as follows. Section II explains the essence of the LA, pointing out how it has been deployed by several leading experts in medical ethics and law on both sides of the Atlantic. It also observes how a related version of the argument, identifying a logical link between legalising PAS and legalising VE, has been accepted by a number of judges in England and the US.
Section III, the main body of the paper, sets out and evaluates seriatim the responses of the above scholars to the LA. This part of the paper should not be thought of merely as a descriptive literature review: fairness to those scholars (and to the reader), requires that their thinking about the relation between VE and NVE be set out at appropriate length as a necessary precondition to an analysis and evaluation of their reasons for rejecting, or at least not accepting, the existence of a logical link between VE and NVE.
II. The Nature of the Logical Argument
The historic debate about whether it is ever right in principle for a physician intentionally to end the life of a patient seems no more likely to attract consensus today than it has in the recent past. That is not, however, the central question in the contemporary legal and political debate. That question is the feasibility of legal control: could the law prevent a slide down the “slippery slope”, especially from VE to NVE?
There are generally thought to be two “slippery slope” arguments against legalising VE.Footnote 4 The first is the “empirical” argument (EA). It runs that drafting and enforcing safeguards against mistake and abuse is not feasible. Legal guidelines will inevitably suffer from imprecision, such as about the meaning of “terminal illness” or “unbearable suffering”, and will afford physicians extensive discretion. That discretion will resist effective enforcement, especially given the confidential nature of the doctor–patient relationship.
The second argument, the concern of this paper, is the LA. It runs that although the ethical case typically advanced for VE focuses on patient autonomy, that case is seriously incomplete without appeal to the principle of beneficence. Central to the ethical case for VE is the idea that death would benefit the patient, typically because it would put an end to serious suffering, and that physicians have a duty to confer that benefit (at least if they have no conscientious objection to so doing). Proposals for legalising “assisted dying” in England and the United States require not only that the patient make an autonomous request for a hastened death but also that the patient be suffering from a “terminal illness”. Baroness Meacher has written that her bill “would enable terminally ill, mentally competent adults whose suffering is beyond the reach of palliative care to die on their own terms”.Footnote 5 The LA holds that if a doctor can judge that hastening the death of a capacitous patient in such circumstances would benefit the patient then the doctor can, logically, make that same judgment in relation to a patient in the same circumstances who lacks capacity.
Doctors make judgments every day about which treatments would benefit patients, capacitous or incapacitous. They do not, for example, decline to perform a life-saving nephrectomy because the patient cannot consent. While patient consent is required in the case of a capacitous patient it is obviously not required, indeed cannot by definition be obtained, in the case of a patient who lacks capacity. If a hastened death can (like a nephrectomy, or palliative drugs, etc.) be a benefit, why deny that benefit to a patient merely because the patient is incapable of requesting it? If, then, the principles of autonomy and beneficence justify intentionally administering a lethal injection to a suffering patient who requests it, the principle of beneficence by itself justifies administering a lethal injection to an identically suffering patient who is incapable of requesting it. (We assume that the patient has not refused such an injection in advance of incapacity and that there is no good reason to believe the patient would refuse if they had capacity.) In short, the absence of autonomy does not cancel the duty of beneficence. It is uncontroversial that physicians owe their patients a duty of beneficence, not least in relation to the alleviation of pain. For example, in Cox, Ognall J. made it plain that it is a doctor's duty to do all that is medically possible to alleviate a patient's pain and suffering, even if so doing carries an obvious risk that, as a side-effect of palliative treatment, death is rendered likely or even certain.Footnote 6 According to the LA, then, anyone who supports euthanasia for capacitous, suffering patients is committed in principle, whether they realise it or not, to supporting euthanasia for incapacitous, suffering patients. Not surprisingly, many (and very possibly all) leading writers on medical law and ethics who advocate VE also endorse NVE.Footnote 7
One of the first scholars to identify a logical link between VE and NVE was Yale Kamisar, in his classic paper arguing, on utilitarian grounds, against legalisation.Footnote 8 Kamisar would go on to become as eminent a jurist as the scholar to whom his article responded: Glanville Williams. For it was the publication of Williams’ landmark book The Sanctity of Life and the Criminal Law in 1957Footnote 9 that prompted Kamisar to counter that book's argument for the legalisation of VE. Since that famous debate,Footnote 10 several scholars of bioethics and law have also endorsed the LA, including Daniel Callahan,Footnote 11 Leon Kass,Footnote 12 the author,Footnote 13 David JonesFootnote 14 and Neil Gorsuch.Footnote 15
Gorsuch is now, of course, a Justice of the United States Supreme Court. In 1997 the court declined to uphold a constitutional right to PAS.Footnote 16 Delivering the judgment of the court, Chief Justice Rehnquist invoked another logical slope, that from PAS to VE. He noted that, although the Court of Appeals had upheld the right of capacitous, terminally ill adults to obtain lethal drugs from their doctors, the state had argued on appeal to the Supreme Court that the impact of this holding could not be so limited, and that if there were a constitutional right to suicide, it must be enjoyed by everybody. The Court of Appeal's expansive reasoning, added the Chief Justice, supported the state's concerns. That reasoning noted that the decision of a surrogate was as valid as that of the patient; that some patients might be unable to self-administer lethal drugs and that administration by a physician was the only alternative, and that not only doctors but also family members would inevitably participate in the decision-making process. “Thus,” he concluded, “it turns out that what is couched as a limited right to ‘physician-assisted suicide’ is likely, in effect, a much broader license, which could prove extremely difficult to police and contain.”Footnote 17
Similarly, in Pretty,Footnote 18 in which the Law Lords rejected the argument that the right to life under Article 2 of the European Convention on Human Rights contains a right to self-determination in relation to issues of life and death, including a right to assisted suicide, Lord Bingham noted the same logical link. Counsel for Pretty had submitted that a right to assisted suicide could be distinguished from a right to VE, but Lord Bingham disagreed. If Article 2 contained a right to self-determination extending to the decision when to die, there was no logic in drawing a line between the two:
If article 2 does confer a right to self-determination in relation to life and death, and if a person were so gravely disabled as to be unable to perform any act whatever to cause his or her own death, it would necessarily follow in logic that such a person would have a right to be killed at the hands of a third party without giving any help to the third party and the state would be in breach of the convention if it were to interfere with the exercise of that right.Footnote 19
These two cases concerned a logical link between PAS and VE, and illustrated the importance of judges accurately identifying the principle underlying a proposed right to assisted suicide and of seeing where acceptance of that principle logically led. Whether judges will be as alert to see where a proposed right to VE logically leads remains to be seen. If the Canadian experience is anything to go by, not all judges will. In Carter, in which the Supreme Court of Canada controversially upheld a right to VE and PAS under the Canadian Charter of Rights and Freedoms, the trial judge dismissed the LA as “requiring speculation” and as “only tangentially” relevant because the patients challenging the law in that case were capacitous .Footnote 20 The argument was also sidestepped by the Supreme Court.Footnote 21 The LA has, moreover, been either rejected or sidelined by several health lawyers and bioethicists whose work is prominent in the contemporary euthanasia debate: Gerald Dworkin and Raymond Frey; Margaret Battin; Hallvard Lillehammer; Stephen Smith; Emily Jackson; and Steven Daskal. Let us now evaluate in turn their responses to the LA.
III. Responses to the Logical Argument
A. Gerald Dworkin and Raymond Frey
In an edited volume debating euthanasia and PAS, Professors Gerald Dworkin and Raymond Frey put their case in favour, each contributing two essays.Footnote 22 Dworkin, in his first essay, responded to the LA as deployed by Kass and by Callahan. Dworkin wrote that the arguments from autonomy and from beneficence “are most naturally and plausibly understood as providing necessary conditions that, in suitable circumstances, may be together sufficient” to justify doctors intentionally killing patients. He recognised that if either argument were considered by itself it would have “unwanted implications” but that the “most plausible view” was that it was both the patient's choice and the patient's condition that made a doctor's killing permissible. That is why, he observed, “the view is called voluntary euthanasia – voluntary to indicate choice of the patient, euthanasia to indicate that death is ‘good’”.Footnote 23 He added a footnote to Glanville Williams’ “classic treatment” of the subject in The Sanctity of Life, in which Williams wrote that it had long seemed to some people that euthanasia was both permissible and indeed mandatory when performed on a dying patient with his consent, and where it was the only way of relieving his suffering. Dworkin continued that Kass's position – that it was the patient's great pain or terminal condition that, choice or no choice, justified termination – was held by few “at least with respect to competent adults”. And, to Kass's point that it was not autonomy but the patient's miserable condition that justified killing the patient, Dworkin replied that while the patient's condition was a necessary condition “for justified physician-assisted suicide” this did not mean it was a sufficient condition.Footnote 24
Dworkin claimed that Callahan, by arguing that the two justifications for euthanasia need not be linked and revealed serious problems when considered independently, fell into the same error as Kass. Dworkin's response was that although they were not linked, both were necessary to make the moral case. Callahan was, he analogised, like someone arguing against the position that two conditions were required to justify destroying a painting Dworkin owned: that Dworkin gave permission and that it was not a great work of art. Although these grounds could be considered independently, both were needed, Dworkin maintained, to justify the painting's destruction.
Dworkin's reasoning was questionable. He did little more than assert that autonomy and beneficence were necessary justificatory conditions. Why was this the most plausible view and why was the LA implausible? His position was scarcely strengthened by his observation that when the patient consented to be killed the appropriate definition was “voluntary euthanasia”. This observation invited the retort that when an incapacitous patient was killed the appropriate definition was “non-voluntary euthanasia”. He recognised that the argument that beneficence justified the killing of suffering, incompetent patients was held by more than a few people but he failed to offer any cogent reason why they were mistaken. Note also how he conflated capacitous and incapacitous patients when he wrote that the patient's consent and condition were both necessary “for justified physician-assisted suicide”. Of course they were: PAS implies a patient with capacity. Moreover, his citation of Glanville Williams did not assist. Williams was stating the obvious when he wrote that some thought it right to kill a suffering patient on request. But some, indeed many, who support VE also think that it is right to end the life of a suffering patient who is incapable of making a request. Indeed, Williams’ book made it clear that he himself had no objection in principle to euthanising infants with disabilities and old people with dementia.Footnote 25
Further, Dworkin's analogy with the destruction of a painting was unconvincing, for two reasons. First, if A destroys B's painting without consent, it is a trespass, a wrongful interference with B's right to enjoyment of B's property. A is not conferring a benefit on B; quite the contrary. That is why it is wrong for A to destroy B's property without B's consent (and perhaps even with B's consent if the property is of value to the state). But a key justification for euthanasia is precisely that death will benefit the patient. This important consideration of benefit also explains why A would be justified in destroying B's property even without B's consent, as where B is absent and A destroys B's garden shed to prevent a fire in the shed from spreading to B's adjacent mansion. Similarly, A could justifiably push B out of the path of an oncoming car, without B's consent, to save B's life. A second problem with the painting analogy is that, unlike strangers, doctors are under a duty to benefit their patients, including their incapacitous patients.
Finally, in his introduction to his and Frey's essays, Dworkin noted that the reasons for PAS were not difficult to determine. He wrote that they consisted mainly of the interests that dying patients had in the process of dying being as painless and dignified as possible, though they also relied on the interest of patients in determining the time and manner of their deaths.Footnote 26 So, of the two reasons he gave for PAS the main reason was dying in as painless and dignified a manner as possible. He also wrote that what counted was whether the patient was in great pain, advanced senility, or the like and that if the doctor acted on the basis of these conditions then of course the doctor would be acting benevolently, and it was the patient's condition that justified intervention not the doctor's motive.Footnote 27 But incapacitous patients, including those with “advanced senility”, may be dying in great pain. Why would beneficence alone be insufficient to justify their termination? Dworkin did not explain. Let us now turn to Professor Frey.
Frey's second essay in the book explicitly addressed “The Fear of a Slippery Slope”. Having asked why an incapacitous person should not be euthanised if he or she had left an advance directive requesting termination, Frey moved on to discuss the case of incapacitous patients who had not made such a directive. His response to the LA was, like that of Dworkin, unsatisfactory. He merely observed that the only possibility of a slide was if morally serious people found that the differences between cases were not “significant and morally noteworthy”.Footnote 28 He then asked whether there were such differences. He wrote that if PAS were allowed and that, “as much of the PAS literature supposes”, a surrogate should be allowed to choose death for an incapacitous patient, this would be regarded as a significant and morally noteworthy difference.Footnote 29 He asked whether this should be permitted but did not provide an answer, merely stating that it was a question about which we needed to think carefully, and that one way was to ask how far new cases were like more familiar and resolved cases. He then considered whether it would be right to euthanise a patient who had lost capacity before they could request PAS. Frey wrote that the case was exceedingly controversial but that it was arguable that there were cases where appeals to the best interests of the person sometimes carried great weight, such as infants with a poor prognosis who had undergone an extensive series of operations to eke out life. While, he added, there was no suggestion that we should pass from voluntary to non-voluntary killing in the absence of more discussion and thought, there was equally no suggestion that we should ignore arguments relating to the patient's best interests. He continued that perhaps it would turn out that NVE would come to be permitted but this could not be assumed simply by claiming that NVE was similar to VE. We needed to examine the reasons or factors that marked differences between cases and see to what extent the reasons or factors present in the new case led us morally to rethink the earlier case for erecting a barrier. To forego this examination out of concern for a slippery slope would, he continued, be to yield to a fear that the rethinking of cases and of what we take to be significant and morally noteworthy features could result in moving the barrier. He concluded that if the arguments advanced by himself and Dworkin in favour of PAS established a moral claim by individuals in certain conditions to “assistance in dying” then the burden was on the proponents of the slippery slope to provide evidence showing it was likely that a slide would occur.Footnote 30 Echoing Frey, Dworkin noted in his second essay that NVE was a “more morally dubious step” and one that they were not advocating, but he added that it was noteworthy that the courts had ruled that life support could be removed from various incapacitous patients in their best interests.Footnote 31 (It is surprising that neither Dworkin nor Frey mentioned the Dutch courts’ approval of NVE, two years before the publication of their essays, in the form of infanticide.Footnote 32)
Neither Dworkin nor Frey, then, raised a principled objection to NVE. Moreover, we will recall that Dworkin noted that the main reason for PAS/VE was beneficence, and that a doctor who killed someone with “advanced senility” would be acting beneficently; that Frey observed that appeals to the “best interests” of a disabled infant in being killed carried great weight, and that both noted without demur that courts had upheld the removal of life support from incapacitous patients in their best interests. The LA – one of the key arguments against VE, and an argument foregrounded by Kass and Callahan – challenged Dworkin and Frey to explain why beneficence did not justify NVE. They did not rise to the challenge.
Further, one of their main moral arguments for PAS and VE was the moral similarity they perceived between a physician withdrawing life-prolonging treatment at the patient's request and a physician supplying a prescription for lethal drugs at the patient's request.Footnote 33 A physician, they argued, was hastening the patient's death in both cases, so why should the former be permitted but the latter not? For example, in his discussion of a physician's withdrawal of feeding tubes at the patient's request, Frey asked: “To be prepared to see the patient dead; to take the step that will assuredly produce death; to know as a certainty that death will ensue or be hastened: is this not morally equivalent to intending the patient's death?”Footnote 34 Frey regarded such withdrawal as equivalent to VE.Footnote 35 However, if withdrawing tube-feeding at the patient's request was a case of VE, why was withdrawing tube-feeding from an incompetent patient not a case of NVE?
In sum, Dworkin and Frey not only omitted to offer a satisfactory answer to the LA, and voiced no objection to NVE in principle, but they apparently failed to recognise that one of their main arguments for “assisted dying” – that there was moral equivalence between withdrawing life-prolonging treatment at the patient's request and VE – generated an additional logical argument for NVE.
B. Margaret Battin
Professor Margaret Battin, one of the leading international advocates for “assisted dying”, outlined her moral case for PAS in an essay in 1998.Footnote 36 That case, she wrote, was grounded, first, in respect for autonomy, “the right to live one's life as one sees fit, subject only to the constraint that this not involve harm to others”. Because, she added, living one's life as one chose must also include living the very end of one's life as one chose, the matter of how to die was as fully protected by the principle of self-determination as another part of one's life: “Choosing how to die is part of choosing how to live.” The second component of the moral case for PAS was respect for beneficence and non-maleficence, or what in the end-of-life context she preferred to refer to as the principle of mercy: “the principle that one ought both to refrain from causing pain or suffering and act to relieve it.”Footnote 37 Self-determination and mercy did not function as independent principles: both must apply for a patient to have any substantial claim on the physician's assistance and for the physician to have a corresponding obligation to assist.Footnote 38
Battin's argument raised two obvious questions. First, if autonomy entitles one to end one's life as one chooses, provided doing so does not harm others, why does it not confer a right on any person autonomously to request assistance in suicide from their physician? Second, if the principle of mercy requires that one ought to refrain from causing pain or suffering and act to relieve it, why does it not require (or at least permit) physicians to end the lives of suffering, incapacitous patients? Surprisingly, Battin addressed neither question. She did, when outlining what she took to be the case against PAS, discuss what she termed the “slippery slope argument” but this turned out to be the EA, not the LA.Footnote 39 She wrote that physicians and patients alike recognised the obligations of mercy: “Physicians and patients alike also recognise obligations of mercy: relieving pain and suffering is a central part of the physician's task, as well as what the patient seeks from the physician.”Footnote 40 But, the reader might ask, was this not a central part of the physician's duty whether or not the patient had capacity? She also wrote that the stronger the patient's wish for death and the greater the patient's experience of unrelievable pain and suffering in the process of dying, the stronger the physician's duty to provide “assistance in dying”, but that where the patient did not want to commit suicide and was not suffering from a terminal illness, the physician had a duty not to provide such assistance. Intermediate points, she went on, yielded a range of weaker to stronger claims on the physician.Footnote 41 But what of cases where the patient was suffering gravely but lacked capacity? Why would it be wrong for the physician, applying the principle of mercy, to put the patient out of their misery? The reader was left to guess. She went on to discuss the situation of the patient who would perhaps have requested PAS but who had become incapacitous, delirious or comatose. She wrote that self-determination and mercy ceased to be relevant. The incapacitous patient was no longer autonomous “and, if unconscious, the patient is no longer capable of experiencing suffering or feeling pain”.Footnote 42 But what if the patient, although lacking capacity, were in great pain? By no means all incapacitous patients are unconscious of pain. Infants, for example, can suffer intense pain from certain conditions. Yet again, Battin omitted to explain why incapacitous patients should be denied a merciful release.
She also wrote that two widely-employed end-of-life means were the withholding or withdrawal of respiratory support or artificial nutrition and hydration, and the “over-ample” use of pain-relieving drugs. These two ubiquitous means, she commented, did not satisfy the doctor's duty to the dying patient who sought assistance in suicide, “since the physician's obligation is rooted not just in the principle of mercy, which these means of negotiating death might provide” but also in autonomy.Footnote 43 However, doctors regularly used these means in the case of incapacitous patients, relying solely on the principle of beneficence. Why, then, should they not, relying on the same principle, intentionally administer lethal injections to incapacitous patients?
Further, Battin, a leading defender of Dutch euthanasia, referred to the experience of the Netherlands, including the widely reported “1,000 cases” in 1990 in which, Dutch researchers found, patients (who were mainly incapacitous but a significant minority of whom had capacity, either wholly or partly) were intentionally given a lethal injection by their doctors without having made an explicit request, in breach of a key requirement for lawful euthanasia. She also mentioned the government-appointed commission (the Remmelink Commission) that reported on those and other statistics generated by the Dutch researchers.Footnote 44 Oddly, however, she made no mention of that commission's endorsement of NVE among the 1,000 cases, an endorsement grounded on the patients’ suffering or, in other words, on the principle of beneficence.Footnote 45 Nor did she mention the endorsement of NVE in 1993 by the Dutch researchers themselves, who invoked the principle of “universalizability”: if euthanasia was justified for the suffering with capacity then it was also justified for the suffering without capacity.Footnote 46 Nor the rulings of Dutch appellate courts in 1996 declaring lawful the NVE of disabled, suffering infants, again on the ground of beneficence.Footnote 47 Peculiar omissions indeed.
We may also mention a later essay of Battin's in which she discussed slippery slope arguments.Footnote 48 She repeated her assertion that the arguments from autonomy and beneficence did not operate independently and that it could not therefore be claimed that permitting PAS would require providing it for lovesick teenagers who wanted to die but who were not terminally ill. Likewise, she added: “it cannot be claimed that permitting physician-assisted dying on the basis of the principle of mercy would require involuntary euthanasia for someone in pain who nevertheless wants to stay alive.”Footnote 49 But this was a straw person. The claim made by many opponents of VE is that endorsement of VE logically involves endorsement of NVE, not that it logically involves endorsement of involuntary euthanasia (IVE: the intentional killing of a patient with capacity who does not want to be killed). She again failed to respond to this claim. The failure was again surprising, even more so because the claim featured in her own list of concerns voiced by opponents of legalisation. She quoted the concern of the Canadian Medical Association that if VE or PAS were permitted there may be “legal challenges … to extend these practices to others who are not competent”,Footnote 50 an extension that would be the slippery slope that many feared. She also quoted the concern of the American Medical Association that euthanasia could readily be extended “to incompetent patients and other vulnerable populations”.Footnote 51 She asked whether such fears were well founded but her answer concerned the EA not the LA. She mentioned, as in her previous essay, the notorious 1,000 cases in the Netherlands. She commented that these cases reflected the physicians’ perceptions of the moral urgency of ending someone's agony when he or she has become incompetent and is no longer able to express a request. It was “a mercy death for someone dying in severe suffering”.Footnote 52 She again omitted to note that the Remmelink Commission endorsed these cases of NVE on the ground of beneficence. In this article she did mention the Dutch medical profession's endorsement of infanticide by the “Groningen Protocol” but again, instead of recognising that it also illustrated the LA in action, merely commented that infanticide was not common.Footnote 53
Battin has, moreover, openly endorsed NVE. A collection of her essays published in 1994 contained an essay in which she defended NVE.Footnote 54 She wrote that if a person was healthy and without pain we would interfere to keep him or her alive, as by preventing suicide, but “if his or her life means only pain, we act for the person's sake by causing him or her to die (as we should for certain severely defective neonates who cannot survive, but are in continuous pain)”.Footnote 55 Turning from the principles of autonomy and mercy to the principle of justice she noted that we were coming to use the term euthanasia not just for “pain-sparing deaths” but also for “resource-conserving deaths” and to apply it to cases in which the person was neither suffering nor capable of choosing to die, such as someone in an irreversible coma.Footnote 56 It had been argued that justice permitted the termination of such lives to permit a fairer distribution of resources. She wrote: “The argument from justice demands that these patients, since their claims for care are so weak as to have virtually no force at all, be killed, not simply allowed to die.”Footnote 57 She argued that we should favour a distributive principle that would allocate medical resources to those who wanted treatment, where wanting was interpreted as a “realistic desire”: the patient must understand the treatment's purposes, side-effects, probability of success and the possible end condition to which it would lead.Footnote 58 Life-prolonging care given to the “permanently comatose, decerebrate, profoundly brain-damaged, and others who lack cognitive function” was not realistically desired, even if they were requested by the patient in an advance directive. Since such patients could not want such care, or even feeding or routine hygiene, they were not entitled to it. Such patients “should be allowed – or perhaps caused – to die”.Footnote 59 She concluded her essay advocating VE and PAS “supplemented by nonvoluntary euthanasia only when the patient is permanently comatose or otherwise irretrievably inaccessible”.Footnote 60 In short, she thought that NVE could be justified: by justice in the case of the permanently comatose and by mercy or beneficence in the case of suffering neonates.
In a 2018 paper to which she contributed she reaffirmed her support for infanticide.Footnote 61 Replying to a hypothetical question from the Dutch Minister of Health whether the Dutch infanticide protocol should be extended to allow the killing of children from one to 12 years old, she wrote: “I generally support your proposed change in Dutch law governing eligibility for euthanasia.” She noted that in much of the world the word euthanasia had overtones of Nazism and connoted political killing “that had nothing to do with the interests of the person killed”, whereas in the Netherlands the word connoted mercy and compassion. She urged the Minister to make it clear that only euthanasia in the latter sense was being legalised: “That is, you want to permit the ending of life in a way that, given the unbearably sad circumstances of a child's dying, can make it gentler, easier, and more humane for both the child and for the parents in whose arms you can help that death to occur.”Footnote 62 She urged the Minister to extend the law: there were plenty of good reasons to do so, she concluded, and none against.
In sum, like Dworkin and Frey, Battin did little more than assert that both autonomy and beneficence are necessary to justify euthanasia, and her assertion was difficult to square with her open endorsement of NVE.
C. Hallvard Lillehammer
In 2002 Dr. Hallvard Lillehammer directly confronted the LA in a paper in this Journal, and rejected it.Footnote 63 He began by suggesting that those who employ the EA think that VE may be morally permissible in certain circumstances but that it should not be legalised because of the bad effects, such as the practice of NVE, that would ensue. This is misleading: one may employ the EA or the LA even if one thinks that VE is always immoral.
Responding to the LA, Lillehammer claimed that it rested either on a logical confusion or a misunderstanding of the value of autonomy, or both.Footnote 64 He wrote: “The defender of voluntary euthanasia should claim that moral weight attaches both to the patient's autonomous request…and to the doctor's competent judgment that would be a benefit to the patient.”Footnote 65 However, it is one thing to assert that both conditions are necessary, quite another to advance an argument that they are. Moreover, Lillehammer's reasoning was not always easy to follow. He seemed to think that the presence of an autonomous request was a good reason to kill that person, or at least may be: he seemed to accept that granting some requests would be “patently immoral” or “inconsistent with human well-being”.Footnote 66 He also seemed to think that the absence of an autonomous request was therefore a reason against killing an incompetent person, but this is a non sequitur.
Lillehammer addressed a hypothetical of mine concerning twin brothers, X and Y, suffering identically unbearable pain from the same terminal illness. X is competent and requests a lethal injection for himself and for Y, who is incompetent. The question raised by the hypothetical was: if the doctor agrees to give X the benefit of a merciful release, why should the doctor deny that same benefit to Y? Lillehammer reasoned that the absence of an autonomous request for euthanasia by Y entailed that the doctor's judgment that death would benefit Y was insufficient to justify euthanasia. He omitted to explain, however, why the absence of an autonomous request justified the physician in condemning Y to the prolongation of unbearable pain. Lillehammer continued: “if non-voluntary euthanasia is permissible in Y's case, this will not be for precisely the same reasons for which voluntary euthanasia is permissible in X's case” and it was the latter claim on which the LA depended.Footnote 67 However, the LA does not depend on showing that the reasons justifying VE and NVE are “precisely the same”, simply that if killing X confers a benefit because it ends unbearable pain, then so does killing Y. In short, Lillehammer's response to the LA was inadequate.
D. Stephen Smith
Three years later Dr. Stephen Smith also sought to rebut the LA.Footnote 68 Echoing Lillehammer, he misleadingly asserted that the “slippery slope” argument pointed to the risk of a slide from a morally acceptable practice to a morally unacceptable practice.Footnote 69 He also misrepresented the LA as running that by accepting VE one is logically committed to accepting both NVE and IVE.Footnote 70 The LA runs that killing an incapacitous patient is justified by beneficence alone but it does not run further, for it does not logically follow, that beneficence also justifies killing a capacitous patient when the patient has not autonomously requested death.
What, then, was Smith's response to the LA? He largely echoed Lillehammer who, he claimed, had produced a convincing refutation of the LA.Footnote 71 As we have seen, this was not so. Smith added that although it might be that a doctor granted a patient's request for VE because the doctor judged that patient's life lacked value, the doctor might approve the request for other reasons. The doctor might believe that his or her only duty was to give effect to the autonomous wishes of the patient because the patient was best placed to determine the value of his or her life. The doctor's only role was to determine whether the patient's judgment was autonomous. This was why, he continued (somewhat confusingly) that the doctor's involvement “is not often characterised as the doctor's judgment, but is instead considered to be the ‘mercy’ interest or beneficence”.Footnote 72 This interest was the ethic of doctors to prevent suffering and it was the prevention of suffering that was of importance in the case of VE, not that the doctor has decided that the patient's life had no value. But how, one may ask, could the doctor be acting beneficently unless the doctor judged that the patient was indeed suffering and was suffering sufficiently that death would benefit the patient? Beneficence is not simply doing what the patient wants, whether the request is for a hastened death or the removal of a kidney. That is precisely why proposals for VE standardly require that physicians must be satisfied that certain conditions in addition to an autonomous request must be satisfied. Smith failed to appreciate that no responsible doctor would kill a patient merely because the patient autonomously asked to be killed.
Like Lillehammer, Smith responded to the hypothetical concerning patients X and Y. Smith's response was that we do not know what Y wants and “[s]ince we do not know what Y wants, he is not, nor can he be, subject to exactly the same treatment”.Footnote 73 But why not? The doctor's primary duty is to benefit the patient, whether or not the patient is capable of requesting that benefit and, unless the doctor has any reason to believe the patient would object to a benefit, the doctor is surely duty-bound to confer it. If the hypothetical involved palliative treatment rather than a lethal injection, would Smith object to Y being palliated on the ground that we do not know whether Y would want palliation? Smith speculated that Y might differ from X in various ways, such as being religiously opposed to VE. Again, were the hypothetical to concern palliative treatment, would Smith speculate – without any basis – that Y might have religious objections to palliation and would prefer to suffer? Speculation that twin Y might have objections to a hastened death could be met with the response that there is no basis for any such speculation or by amending the hypothetical to spell out that Y had no such objection.Footnote 74
E. Robert Young
In 2007 Robert Young joined the debate in his book Medically Assisted Death. Footnote 75 He wrote that although the LA might seem convincing at first sight, it exemplified a common error. That error lay in assuming that a patient's best interests could be judged entirely by reference to the patient's best medical interests, since it was only in relation to these that the physician could possibly lay claim to having a better vantage point than the patient from which to determine a patient's interests. As patients had other interests, including interests in shaping and directing their own lives, that is, in acting autonomously, it was erroneous to conclude that the physician should treat capacitous and incapacitous patients alike when assessing their best interests.Footnote 76
Young's argument was fragile. The LA does not assume that a patient's best interests are comprised entirely by the patient's best medical interests. A physician's judgment that death would benefit a patient will typically be informed by medical factors, such as the nature and extent of the patient's physical and mental condition, including the degree of the patient's suffering and alternative possibilities for its alleviation. But medical science will not tell the physician whether or not he or she ought to kill the patient, which is a moral judgment. As his book recognised, when a doctor performed euthanasia she did so because she had good reason to believe either that the effects of illness or disability had made the patient's existence “so bad that he would be better off dead” or that, unless she intervened, the patient would soon reach that point. The doctor's belief that euthanasia would benefit the patient “has to constitute a primary element in her motivation” because euthanasia was for the sake of the one who was to die.Footnote 77 Why, then, could a physician not act on the basis of the same belief and motivation in relation to a patient lacking capacity? This does not mean that doctors treat capacitous and incapacitous patients alike, for the former are able to express their views, including the view that they do not want a particular (or any) treatment. Moreover, while they have an interest in shaping and directing their lives and acting autonomously, incapacitous patients do not, leaving the physician to decide what is in their best interests.
Like several other critics of the LA, Young's position was odd because of his evident moral approval of NVE. He wrote that the legalisation of NVE was not necessary, even if NVE would on occasion be morally justifiable.Footnote 78 He claimed that only “qualitatively valuable human life” merited protection.Footnote 79 He wrote that the withdrawal of futile treatment from patients without capacity was, “properly understood”, a “form of non-voluntary euthanasia”.Footnote 80 There was, he added, good reason to believe it was a frequent occurrence that took place whenever those without capacity to ask for life-sustaining treatment to be withheld or withdrawn nevertheless had it withheld or withdrawn by physicians.Footnote 81 He also endorsed the intentional ending of the lives of infants who would be “better off dead”, in the interests of the infant and others.Footnote 82 It was sometimes in the best interests of infants and young children with conditions that offered no hope of recovery to be allowed to die, provided this could be managed without needlessly prolonging their suffering. He wrote that “killing, in such circumstances, would be morally no worse than letting die”.Footnote 83 However, he continued, there was “no need to give ammunition to those opposed to the legalisation of voluntary medically assisted death by insisting that the means for administering non-voluntary euthanasia to incompetent patients be no different than for administering voluntary medically assisted death to competent patients”.Footnote 84 So, he seemed to be saying: NVE by injection was morally equivalent to NVE by withdrawing treatment but the former should not be advocated lest people might realise they were morally equivalent, a realisation that would assist those who argued there was indeed a slippery slope from VE to NVE.
F. John Griffiths
The following year, Professor John Griffiths, a leading authority on and defender of euthanasia in the Netherlands, criticised the LA in his valuable book on euthanasia in Europe.Footnote 85 He doubly mischaracterised the slippery slope as involving a slide from something that was ethical, and as a factual inevitability.Footnote 86 More importantly, however, he proceeded to criticise the LA for reflecting “the common if ill-considered notion” that for any law there must be only one justifying principle, whereas the “assisted dying” laws in the Netherlands and Oregon were based on two principles: autonomy and beneficence.Footnote 87 Beneficence was, he added, necessary because, unlike cases of suicide, a second person, the doctor, was involved, and the behaviour of the doctor required a justification other than respecting the autonomy of the person who sought death. In the case of the doctor, this additional justification lay in the duty to relieve suffering. All this was, clearly, consistent with the LA. However, he also wrote:
Similarly, the other way around, the fact that euthanasia law is based (in part) on the principle of beneficence – needed to justify the involvement of a doctor – does not entail that the requirement of a voluntary request of the patient will inevitably be swept away by the logic of beneficence (since there are suffering people who cannot or have not asked to die). The requirement of a request will not be swept away precisely because it is based on an entirely different principle.Footnote 88
This was muddled. It does not follow from the fact that a doctor needs the consent of an autonomous patient in order to benefit the patient that a doctor may not benefit a patient who lacks autonomy. Autonomy is not in play in the case of an incompetent patient, but beneficence is. To put it another way, the requirement of a request will not be “swept away” by the principle of beneficence but by the simple fact that the patient is not autonomous.
Griffiths added that the LA was “intrinsically incoherent”: “If A is indistinguishable from B, then the one cannot be benign and the other abhorrent. If relevant distinctions can be made, they can be (and in Dutch and Belgian euthanasia law are being) maintained.”Footnote 89 Griffiths here further confused the issue. Of course if A (VE) is indistinguishable from B (NVE) the one cannot be benign and the other abhorrent. But the question is not about the moral judgment whether VE is benign and NVE is abhorrent; it is whether by endorsing VE one is logically committed to endorsing NVE, irrespective of whether one regards them both as benign or as abhorrent.
Moreover, lest his reference to Dutch law maintaining “relevant distinctions” be understood (as it would naturally be understood) as an assertion that Dutch law has held the line between VE and NVE, the reader will recall that the Dutch courts have, on the ground of beneficence, permitted NVE in the case of disabled infants,Footnote 90 a fact which, far from undermining the LA, illustrates its force. As Griffiths himself candidly recognised: “The applicable norms in the Netherlands have assuredly changed in the direction of open acceptance of the legitimacy of termination of life of severely defective newborn babies.” He added that the influence on these changes of the way euthanasia had earlier been legalised was obvious and that in this sense “one might speak of a normative slippery slope”.Footnote 91
Further, Griffiths voiced no moral objection to this slide, whether the one that had already taken place the case of newborns or the one that also seemed also to be taking place in the case of incompetent adults, again as a result of Dutch courts, entirely logically, extending the application of the principle of beneficence.
G. Emily Jackson
In 2012 Professor Emily Jackson offered the following argument against the LA.Footnote 92 She wrote that, given that “assisted dying” involved another person helping to bring about a patient's death, it was not unreasonable to suggest that both unendurable suffering and an autonomous request should be necessary before a doctor was entitled to conclude that assisting a patient's death would be compatible with the doctor's legal duty of care towards the patient and the doctor's ethical responsibility to do no harm. She continued that although it was true that doctors were sometimes entitled to treat a patient without consent, as where the patient lacked capacity, it would be possible for “assisted dying” to be a special case, which could be carried out only following a patient's explicit request.Footnote 93
This requires only a brief response. It would indeed be possible for the law to make “assisted dying” a special case, but Professor Jackson offered no sound reason why the law should do so. The reason she suggested – that it involves another person, a doctor – was unsatisfactory. Doctors are involved in benefiting patients who lack capacity in many other situations. Her argument also sat uneasily with her condonation of intentionally ending the lives of the incapacitous.Footnote 94
H. Steven Daskal
In 2018 an extended response to the LA was offered by Steven Daskal.Footnote 95 Daskal wrote that supporters of the LA and defenders of VE had both grounded their analysis in a mistaken understanding of the doctor–patient interaction in cases of VE. He rejected the argument that to approve VE a doctor needed to judge that a patient would be better off dead; VE “merely requires physicians to judge patients as within boundaries of appropriate deference”.Footnote 96 He identified two ways – the “autonomy-oriented approach” and the “evidential approach” – of defending cases of VE that fitted his model. The first focused on the idea that there were good reasons to respect the patient's autonomy that are independent of considerations of well-being; the second was focused on the idea that a patient's request for euthanasia constituted significant evidence regarding what is best for the patient.
On the “autonomy-oriented approach”, when a patient requested euthanasia, that expression of autonomous will provided a physician with a good reason to comply. If the physician thought that the patient's well-being would be promoted by continued life, the physician would have to weigh that consideration against the importance of respecting the autonomous request. This, he argued, did not mean the physician ought to grant euthanasia on request, such as by jilted lovers. The physician “must judge that the patient is within the boundaries of appropriate deference”. One way to establish the boundary was in terms of “the amount of well-being, or the length and quality of life, that the patient is sacrificing”.Footnote 97 It was reasonable, he continued, for physicians to comply with requests from patients with a low quality of life and short life expectancy and yet deny requests from patients who appeared to have “relatively long and valuable lives ahead” on the ground that promoting well-being outweighed autonomy. The difference was “the amount of well-being at stake”.Footnote 98
On the “evidential approach” the patient's request for euthanasia operated to provide evidence as to what was best for the patient. The patient might not always be correct about what was best for him or her, but was generally more likely to be so than anyone else. Although there was no independent normative significance in respecting the request, the request was “a significant source of evidence that the physician must take into account” in determining whether to perform VE. As on the previous approach, the physician would need to exercise his or her own judgment before complying and would need to assign weight to the evidence provided by the patient's request and to balance it against competing evidence to determine if the patient was within the boundaries of appropriate deference. Daskal admitted that this approach would lead a physician to perform VE only if the physician believed there were some circumstances in which a patient was better off dead, but Daskal did not think this would result in an “inevitable slide from this way of vindicating voluntary euthanasia to an endorsement of non-voluntary euthanasia”.Footnote 99 It made sense to require “exceptionally weighty or compelling evidence” before concluding that death was a benefit and one might think that in the absence of a request from the patient a physician could never point to sufficient evidence.Footnote 100
Daskal added that his argument garnered independent support from uncontroversial cases involving possible courses of treatment. It was standard medical practice for a physician to describe various courses of treatment and to let the patient choose between them. A physician could provide a requested treatment without endorsing the patient's choice as correct: the physician merely needed to think that the patient's choice was within bounds of appropriate deference.Footnote 101
Daskal's argument, though interesting, proved unsuccessful. As a preliminary, we may note that he incorrectly described the LA as an argument that a slide from VE to NVE will “inevitably” occur.Footnote 102 The argument is not about inevitability: one may in practice resist the logical implications of one's position. The argument is that what is crucial to justifying VE – the judgment that the patient would be better off dead – also justifies NVE. More seriously, the core of his argument, that it is the patient's request that is decisive in justifying VE, either because of the weight due to it as an autonomous request or as evidence that euthanasia would be consistent with the patient's well-being, was flawed.
It is not the supporters of the LA who misconceive the doctor–patient interaction in cases of VE, but Daskal. His “autonomy-oriented” approach is wayward, for VE is not an exercise in self-determination but an act performed by another person: the physician. The physician has a duty to act for the good of his or her patient. The physician is not subject to the authority of the competent patient except to the extent that the physician may not treat the patient without consent. There is nothing in the status of a patient that entitles a patient to require that a physician act in a way incompatible with the physician's duty to act only for the good of the patient. There is no way a physician, conscious of his or her duty to the patient, can avoid making an independent judgment about a patient's request to be killed; about whether the patient would indeed be better off dead. Indeed, Daskal envisages the doctor taking into account “well-being” considerations, but this must mean that, despite the request, the physician may judge that the patient would not be better off dead. So, at the end of the day, the decision still turns on the physician's judgment about whether or not the patient's life is worth living.
This brings us to Daskal's “evidential approach”, which recognises that VE does involve a judgment by the physician that the patient would be better off dead but which holds that the judgment rests on the evidence provided by the patient's request to be killed. But how can a bare request make death a good treatment option such as to justify the doctor killing the patient? In deciding whether or not to grant the request the doctor will necessarily be passing judgment on the reasonableness of the request, not taking the mere fact of the request as carrying a weight independent of a range of factors, and those factors are the same as those that are invoked to justify NVE. Moreover, killing a patient lies outside the goals of medicine, so is not comparable to medical treatment options that fall within the goals of medicine. It may be perfectly reasonable for a doctor to allow a patient to decide as between a range of medical treatments that the physician has judged beneficial, but the physician must equally judge that death would benefit the patient before offering that option to the patient.
There seems to be little practical difference between Daskal's two approaches for they both require the physician to arrive at an independent decision based on factors concerning the patient's “well-being”. The request by itself does not establish that it would be beneficial to grant it; that turns on these other, independent, factors. It is difficult to see how the physician's judgment can be other than a judgment about whether the patient would indeed be better off dead.
This is not to say that a doctor who carries out VE would necessarily do so in the case of a similarly-situated patient lacking capacity. It may be that a suffering patient who is capacitous refuses palliative treatment, (perhaps because of a dislike of unpleasant side-effects) and that the doctor therefore grants the patient's request for a lethal injection, whereas in the case of an incapacitous patient, where there is no such refusal, the doctor resorts to palliative treatment. But, all things being equal, as where palliative treatment will prove futile, and the doctor judges that a lethal injection will benefit the former by ending suffering, it will also benefit the latter by ending suffering.
I. Some Additional Arguments Against the LA
Could it not be argued that there are at least some requested procedures, such as cosmetic surgery, which doctors perform solely on the basis of a patient's request without making any independent judgment about whether it would benefit the patient? It does not follow from the fact that a surgeon reduces the size of a patient's bulbous nose at the patient's request that the surgeon is logically committed to performing similar surgery on patients lacking capacity. Could this not justify drawing a similar distinction between VE and NVE?
Not really. First, the alleviation of suffering is clearly a core purpose of medicine and extends both to patients who and request it and those who cannot. This is not the case with purely aesthetic and subjective “improvements” to one's face or body (which might even be regarded as completely outwith the purposes of medicine, properly understood). Second, even with mere beautification there are limits to the autonomous requests that doctors ought to respect: no responsible doctor would carry out a harmful procedure such as the removal of an eye because the person wanted to look like Cyclops.Footnote 103 Third, when cosmetic surgery concerns the correction of serious disfigurements, perhaps resulting from congenital abnormality or burns, doctors surely are duty-bound to provide those benefits to patients, capacitous or incapacitous.
What about “sex-change” operations and live organ donation? As to the former, not only do they raise questions about whether they are beneficial and whether they fall within the recognised purposes of medicine, but it is difficult to see how a diagnosis of gender dysphoria could properly be arrived at in relation to an incapacitous adult. (Such procedures in relation to minors are of course particularly controversial, not only because of their irreversible nature but also because the minor will be able to make their own decision in due course.) As for live organ donation, it is the autonomous request that makes the organ's removal a true donation or gift. To remove an organ from an incapacitous patient would be a case of taking rather than giving. In short, the fact that physicians would not perform some procedures without an autonomous request does not undermine the LA: it is undisputed that physicians are duty-bound to benefit their patients by alleviating their suffering, whether their patients are able to request that benefit or not.
There is another possible argument against the LA that merits consideration. It could be argued that just as the law draws a distinction between sexual intercourse with a consenting, capacitous person, which is lawful, and sexual intercourse with an incapacitous person, which is unlawful, it could sensibly draw a similar distinction between VE and NVE. However, although the law could indeed stipulate that VE but not NVE was lawful, following the typical proposals for legalisation, it would be difficult to offer a sound moral justification for the distinction. The rapist can hardly claim that his violation of the victim confers a benefit, but the physician who administers a lethal injection to a suffering patient lacking capacity can indeed claim that it benefits the patient by ending the patient's suffering. In short, if euthanasia is a benefit because it puts an end to suffering, it remains a benefit even if the patient is unable to request it.
IV. Conclusion
It is not surprising that most, and very possibly all, leading academic writers on medical law and ethics who advocate VE also support NVE. They appreciate that the argument from beneficence applies even if the patient lacks capacity. Incapacitous patients may suffer just as gravely as capacitous patients (and even more gravely), and physicians are duty-bound to alleviate the suffering of all their patients.
The attempts of those advocates of VE and PAS considered in this paper to rebut the LA prove unsuccessful. As we have seen, they either largely assert rather than argue that both an autonomous request and suffering are required to justify euthanasia, or they offer arguments that are lacking, such as the argument that VE could be regarded as an exceptional case. Where those advocates also conflate – as many of them do – intended and foreseen consequences and/or endorse NVE on the ground of beneficence, their position is not only weak but inconsistent.
The ongoing debate about whether VE should be legalised is too important for it to be clouded by confusion about or inattention to the underlying principles that are thought to merit legalisation, and of the implications of legalisation. This paper has suggested that, unfortunately, such confusion is not uncommon, even among prominent academic advocates of legalisation, as well as among judges who have upheld a right to VE. The LA merits much greater attention from legislators, judges and scholars, not to mention the media and the general public, than it has hitherto received.
Nor should the LA be dismissed as irrelevant to jurisdictions like England and the US where proposals for legalisation are currently limited (surely for tactical reasons) to PAS. For, as we noted in Section II, there is also a logical link between PAS and VE. If respect for autonomy and beneficence justify a lethal prescription, they equally justify a lethal injection, especially for those unable to kill themselves even with assistance. In short, the arguments for PAS logically justify VE, and the arguments for VE logically justify NVE. It is surprising that opponents of the PAS bills that have been introduced in Westminster over the past quarter of a century have tended to deploy the EA rather than the LA. If they were to demonstrate, by joining the logical dots, that a vote for PAS for the “terminally ill” is in principle a vote for PAS for the chronically ill (who have even longer to suffer); for VE, and also for NVE, they might well cause many supporters of “assisted dying” to think again.
Open access
