Hostname: page-component-8448b6f56d-c4f8m Total loading time: 0 Render date: 2024-04-24T14:33:51.590Z Has data issue: false hasContentIssue false

Hard Choices: How Does Injustice Affect the Ethics of Medical Aid in Dying?

Published online by Cambridge University Press:  16 October 2023

Brent M. Kious*
Huntsman Mental Health Institute, Department of Psychiatry, University of Utah, Salt Lake City, UT, USA Center for Health Ethics, the Arts, and Humanities, Department of Internal Medicine, University of Utah, Salt Lake City, UT, USA Department of Philosophy, University of Utah, Salt Lake City, UT, USA
Rights & Permissions [Opens in a new window]


Critics of medical aid in dying (MAID) often argue that it is impermissible because background social conditions are insufficiently good for some persons who would utilize it. I provide a critical evaluation of this view. I suggest that receiving MAID is a sort of “hard choice,” in that death is prima facie bad for the individual and only promotes that person’s interests in special circumstances. Those raising this objection to MAID are, I argue, concerned primarily about the effects of injustice on hard choices. I show, however, that MAID and other hard choices are not always invalidated by injustice and that what matters is whether the injustice can be remediated given certain constraints. Injustice invalidates a hard choice when it can, reasonably, be remedied in a way that makes a person’s life go better. I consider the implications of this view for law and policy regarding MAID.

Research Article
Creative Commons
Creative Common License - CCCreative Common License - BY
This is an Open Access article, distributed under the terms of the Creative Commons Attribution licence (, which permits unrestricted re-use, distribution and reproduction, provided the original article is properly cited.
© The Author(s), 2023. Published by Cambridge University Press


Many bioethicists think that if the background circumstances informing a person’s decision to accept some significant personal risk or injury are sufficiently bad or inequitable, then others should not enable that decision. For example, when Kious and BattinFootnote 1 argued that there is at least some reason to permit psychiatric medical aid in dying (MAID), many of their commentators objected about the poor social conditions faced by persons with psychiatric conditions.Footnote 2, Footnote 3, Footnote 4 We should not allow psychiatric MAID, these commentators said, because access to mental healthcare is too limited, mental health conditions are worsened by stigma, and many mental illnesses result from trauma, discrimination, and other forms of mistreatment. To the extent that people want to receive MAID primarily (or even partly) because of difficulties resulting from deficient social conditions, these critics implied, they should not want this, and we should not help them do it. In its 2022 report, Canada’s Expert Panel on MAiD and Mental IllnessFootnote 5 addressed this question, considering whether the existence of what it called “structural vulnerabilities” affecting persons with psychiatric illness should preclude psychiatric MAID. It concluded that they should not.

Parallel criticisms are sometimes offered regarding proposals to allow MAID by advance directive for persons with dementia: Critics claim that we should not permit this if persons developing dementia request MAID because they fear being a burden, or not being able to access sufficient care; these fears are due to the fact that we fail, wrongly, to devote enough resources to dementia care.Footnote 6, Footnote 7, Footnote 8, Footnote 9, Footnote 10, Footnote 11 Indeed, criticisms of this form are applied to MAID, in general. It has been argued that if there were enough access to palliative care, good pain control, and family support, people would not want MAID, implying that, in the absence of those commodities, we should not permit it.Footnote 12 This is particularly true for disability rights criticisms of MAID. Longmore wrote, in his description of David Rivlin’s decision to withdraw ventilator support and end his life, of how the decision seemed to have been necessitated by inadequate social supports:

[U]ntil people with major disabilities are guaranteed their rights to self-determination, independent living, equal access to society, and appropriate psychological counseling, medical professionals must never support or assist the suicide of a disabled person. To do so in the present societal circumstances of devaluation, discrimination, and segregation is simply the ultimate act of oppression. Footnote 13

As against Longmore and others, I will defend the intuition that deficient social conditions do not necessarily make MAID or similar choices impermissible—though they can in some cases. In arguing this, I will be engaging in a sort of non-ideal theory—attempting to discern what morality requires when we can assume, at best, partial compliance with moral principles.Footnote 14, Footnote 15 I begin by trying to clarify the critics’ concern, which I take to be that injustice or inequality makes MAID impermissible. The idea that injustice makes MAID “impermissible” involves at least two claims: a claim that at an individual level, it is wrong for a person to receive MAID (or for others to provide it) if she has been significantly affected by injustice, and a claim that at a social level, injustice means that MAID should not be permitted in law or policy. I then problematize these concerns by considering an array of cases about which we have conflicting intuitions: sometimes injustice seems to make MAID wrong, but sometimes it does not.

Focusing first on the individual level, I consider and reject several ways of explaining the cases and conclude that what matters most is whether injustice is reasonably remediable by appropriately motivated parties to the decision. But if this is correct, it is not injustice that invalidates an agent’s request for MAID, but instead the fact that the person’s circumstances can reasonably be improved, so that MAID does not really promote self-interests; injustice is only an inconsistent marker of this. Thus, somewhat surprisingly, whether injustice invalidates a choice like MAID is a consequentialist matter. I then turn to consider when the presence of injustice in a society should preclude MAID and again argue for a consequentialist approach that considers the comparative effects of different policies.

It is important to delimit my argument’s goals and foundations. Although I happen to believe that MAID can sometimes be morally permissible, the view I will describe does not assume this, and I am not attempting to provide a comprehensive argument for the permissibility of MAID. My arguments should be understood as conditioned on MAID being permissible in other respects. I should also emphasize the subtle but important differences between the position I am critiquing and some other adjacent arguments. I will not consider the claim that we should not permit MAID because doing so would reduce or eliminate pressure to change the unjust status quo, nor the claim that MAID should not be permitted because doing so would worsen injustice by allowing us to siphon away even more resources from MAID recipients.Footnote 16, Footnote 17 Nor will I consider claims that permitting MAID would be wrong because it would increase pressure on socially marginalized persons to request MAID,Footnote 18 nor that it would increase the stigma they experience, nor that it would make them feel bad about themselves—for example, believe that they are burdens.Footnote 19 These are all important objections, but they differ from the idea that injustice itself makes MAID impermissible. I set them aside largely because I take them to be empirical matters.

The Critics’ Concerns

I assume that those who believe that background conditions or structural vulnerabilities can make MAID impermissible worry primarily about injustice rather than lack of resources simpliciter. Imagine a society that is extremely equitable and just, so that everyone has the same social goods and everyone’s rights are respected, but also imagine that it is, because of facts beyond anyone’s control, quite impoverished, so that no one has access to effective medical treatment, palliative care, adaptive aids for disability, pain control, or even sufficient food. These dire but equitable conditions do not themselves make MAID wrong. It was not, for Longmore, primarily that Rivlin lacked access to certain goods that made his decision to withdraw the ventilator invalid, but the fact that this lack of access was unequal and due to injustice that did so.

Critics of MAID like Longmore also presumably think that to invalidate a decision, injustice needs to have touched the decision maker and to have played something like a causal role in self-choice. If I request MAID but have never been the victim of injustice, the fact that injustice affects others in my society does not necessarily threaten the validity of my request. Likewise, if I have been a victim of injustice but the injustice is irrelevant to my request for MAID, such that I would have requested MAID even without the injustice, the injustice does not appear to threaten the validity of my choice.

Those who believe that injustice makes MAID wrong also seem to assume that MAID is prima facie harmful: that dying is typically bad for people, even though it may not be bad on balance for a specific person in certain circumstances (such as when it is the only way to alleviate severe suffering). This badness is evinced by an evaluation of counterfactuals: If it were possible for an individual who received MAID to have alleviated personal suffering without dying, that would usually be better. I will call actions like MAID that involve some significant risk of harm, injury, or loss, such that most of us are disinclined to engage in them unless there is some very significant compensating benefit, “hard choices.” Hard choices, when made for oneself, are prima facie self-injurious—they involve doing something that would typically harm one’s interests, were it not for some special circumstances. MAID is prima facie harmful since dying is typically bad for people, but (presumably) not in the special circumstances where they are suffering intolerably from a severe and terminal illness. That MAID is prima facie self-injurious is clearly relevant to the purported effects of injustice on its moral status. No one supposes that a person who has developed pneumonia because of injustice should not accept curative antibiotics since antibiotics are not prima facie harmful. Whether to get antibiotics to treat one’s infection is not a hard choice.

There are other biomedical examples of hard choices that seem to be invalidated by injustice. Consider paid organ donation: If a father chooses to sell his kidney because it was the only way he can feed his children, many will see this as morally problematic and think the transplant surgeon ought not participate.Footnote 20, Footnote 21 Likewise, paid maternal surrogacy contracts are not enforceable in the U.K., presumably because of concerns about exploitationFootnote 22, Footnote 23, Footnote 24: If a woman altruistically chooses to be a surrogate, it is unlikely that she is making this decision because of injustice; but if she does it for pay, it implies that injustice has made her desperate. Although I focus on MAID, I suspect that the conclusions reached here generalize other hard choices in biomedicine, too.

Some Cases

In some cases, injustice does seem to make MAID impermissible. Consider some examples:

Ivan has metastatic gastric cancer and is dying in great pain. He cannot achieve adequate pain control because, even though his doctor has prescribed plenty of pain medicines, his wife, who is addicted to opioids, keeps stealing them. Ivan knows this but does not want to disclose it to his doctor. He finds that he certainly cannot tolerate the pain much longer and sees few other reasons to go on living, so he places a request for euthanasia with his physician.

Sally has multiple sclerosis and is experiencing severe neurological deficits that make it hard to carry out tasks like bathing herself. She is also in pain because of problems with positioning her body and a worsening sensory neuropathy. A friend suggests that since she is suffering so much, she should consider MAID. Unfortunately, Sally’s MS has never actually been adequately treated: Although a variety of interventions that are likely to be effective and cause at least temporary remission of her symptoms exist, she has no insurance, is severely impoverished, and cannot afford to pay for them out of pocket.

Robert has developed ALS and can no longer work. He has great difficulty caring for himself. He has obtained state disability benefits, but they are insufficient to cover his reasonable expenses. He does not have enough income or savings to continue to pay his mortgage and cannot afford a nursing facility. In desperation, Robert considers MAID.

Eric has quadriplegia after a car accident. He needs constant care. His husband, who had always been abusive but who has begrudgingly taken responsibility for Eric’s care, tells him that if Eric does not request and receive MAID, then he will kill Eric and Eric’s mother.

Most of us will find the prospect that the persons described in these cases could pursue and receive MAID disconcerting. The problem seems to be that the suffering person’s hand is being forced by circumstances that are unjust. On the other hand, we sometimes have the intuition that even if circumstances are unjust, MAID (or some other way of hastening death) is permissible. Consider these cases, too:

Near the end of the 1992 film adaptation of James Fenimore Cooper’s book The Last of the Mohicans,Footnote 25 Cora, Hawkeye, and Major Heyward are captured by their enemies, but Heyward arranges to be burned alive to spare Cora’s life. As Heyward is tortured, he sees Hawkeye, who is perched on a nearby hill, from afar. He knows Hawkeye, who is an excellent marksman, can shoot him and end his agony. They share a glance, conveying mutual understanding. Hawkeye shoots Heyward, killing him.

Tenzin is suffering from cancer in rural Tibet. He is in great pain. He could obtain significant relief if pain medications were available, but because of unequal distribution of medical supplies from China, they are not. His physician, though deeply concerned, is unable to secure any pain medications. He does, however, have access to a decoction of hemlock that could kill Tenzin quickly and painlessly. Tenzin accepts the hemlock.

In cases like these, aid in dying is, as Battin has said, a way to achieve the least worst death. Footnote 26 Thus, it may appear that injustice sometimes invalidates hard choices and sometimes does not.

When Injustice Invalidates a Choice

An initial proposal: Coercion, oppression, and authenticity

A theory of how injustice invalidates hard choices needs to account for cases like those mentioned earlier. One simple proposal is that it does so because persons should not be required to bear personal costs to correct the harms injustice causes. We often think that social costs should be distributed as evenly as possible and that it would be unjust for us to expect or permit a person to take on additional costs when that person is already unfairly burdened. But this proposal is unhelpful in the second set of cases: Heyward’s and Tenzin’s requests for aid in dying involve them taking on significant prima facie costs, but their choices are still valid. Nor does this rule seem to hold in other contexts. Suppose, for example, that Mark has been arrested on fabricated drug charges while vacationing in another country and has been told by the police that he will be released from jail only if he pays $10,000. Mark can permissibly pay the bribe, even though he has already unjustly borne other costs. Likewise, we can permissibly help him pay the bribe, such as by arranging a transfer of funds.

A more plausible proposal is that injustice invalidates a hard choice because it undermines autonomy in some respect.Footnote 27 It is widely agreed that people should not receive MAID if the choice is clearly coerced; it must appear free of external pressure. Indeed, this is written in most guidelines and statutes governing the practice.Footnote 28 This is because of the doctrine volenti non fit injuria. Although there is a presumption against harming others, this presumption can be overturned with their autonomous consent; but coercion undermines autonomy, so consent does not overturn the presumption against harm. The idea that injustice invalidates a hard choice because it means that the decision maker is coerced seems to account for cases like Eric’s: Eric is coerced, the coercion undermines his autonomy, and this invalidates his decision.

Appeals to the autonomy-vitiating effects of injustice are insufficient, however. Coercion also seems to be involved in some cases where injustice does not invalidate a hard choice: Heyward is clearly coerced, and it is not hard to imagine that Tenzin is, too. In both cases, however, coercion does not invalidate their choices. Conversely, although Robert, Sally, and Ivan are victims of injustice, they are not obviously coerced in the sense that some threat is made in order to force their choices.

One could suggest that injustice invalidates a hard choice when it involves coercion in a looser sense that does not require a threat—or perhaps when it involves a similar sort of wrong, like “oppression.” The extensive literature on adaptive preferences and relational autonomy, for instance, suggests that oppressive social circumstances sometimes undermine autonomy by producing adaptive preferences—where the values pursued by oppressed persons reflect a normalization or internalization of the oppression.Footnote 29, Footnote 30, Footnote 31, Footnote 32 To borrow an example from David Enoch,Footnote 33 if a woman who has grown up in a rigidly patriarchal society chooses to make great personal sacrifices, such as starving herself, to slightly improve life for her husband, and she does this because she has internalized misogynistic social norms, we may think that her sacrifices are not fully autonomous. If she were later to request MAID or make another hard choice, like kidney donation, in order to promote her husband’s seemingly less significant interests, we might think her choice is invalid—she is only choosing this because she has been oppressed.

As Wiebe and Mullin have also recently argued,Footnote 34 however, this approach is also unhelpful. In the first set of cases, where providing MAID seems problematic, the person requesting MAID need not have internalized any oppressive norms; they may actually think quite clearly that they have been subjected to injustice and see it as wrong. Conversely, in the second set of cases, where aid in dying is permissible, it might still be permissible even if we assume that each of the requestors has internalized the injustice (suppose Heyward thinks he deserves to be punished, or Tenzin thinks that native Tibetans do not deserve access to good medical care).

A more general worry is that talk of autonomy, coercion, authenticity, and similar evaluations often serves primarily to rationalize our views about the respect-worthiness of a decision. Our concepts of autonomy, coercion, and the like are bidirectionally linked to our other moral intuitions, so that we cannot expect to fully clarify our moral intuitions by first determining whether these concepts apply (Kious, Brent M 2015).Footnote 35 To say that a choice is autonomous is sometimes to say little more than it is deserving of respect; but then to say that a choice is not worthy of respect because something has made it non-autonomous flirts with tautology. This is not, of course, to say that autonomy, coercion, and similar properties do not matter: If a hard choice is clearly coerced or is otherwise rendered non-autonomous by injustice, it is less likely to be valid. My point is only that we are unlikely to be able to delineate when injustice invalidates a hard choice, in all of the unclear cases, by appeal to such notions alone.

Another possibility is that injustice invalidates a hard choice when it makes the agent “desperate”. I used the term desperation to describe Robert’s choice earlier—his financial dire straits and his inability to provide for his own basic needs because of illness make him desperate, which causes him to choose MAID. We might think that no one should have to make a hard choice when desperate, as desperation undermines the voluntariness or autonomy of a choice. While desperation is, perhaps, a condition that we should ameliorate when possible, it cannot itself preclude hard choices—since, seemingly, the persons who permissibly receive aid in dying in our cases (Tenzin, Heyward) are also desperate. Indeed, although desperation might come in degrees, I suspect that we could accurately describe almost all persons who choose MAID, even without the influence of injustice, as desperate in some sense—most commonly, perhaps, as desperate to avoid or relieve their own suffering.

An alternative account

Another feature that could distinguish the first set of cases, where injustice invalidates a choice, from the second group, where it does not, is that in the first set of cases, the injustice forcing the hard choice is subject to change in some way, but not in the second set. Thus, for instance, Ivan’s request for MAID is invalidated, but the injustice he experiences is, from one standpoint, easily fixed: If his wife stops stealing his pain medicine, he will not need to hasten his own death. In contrast, Tenzin’s situation is unjust but not easily fixed—at least from his standpoint or the standpoint of his physician since neither is likely to persuade the Chinese government to change its policies. Changing the narrative changes our intuitions, however. Suppose that Tenzin’s doctor happens to know a regional official who, for a small bribe, could easily get Tenzin enough morphine to control his pain. In that case, providing MAID to Tenzin rather than offering to get him the morphine starts to look problematic.

These observations suggest the following account: Injustice invalidates a hard choice, such as whether to receive MAID, if it is reasonably remediable by morally sensitive parties to the decision.

This needs explanation. First, when a hard choice is made under injustice, the injustice is reasonably remediable only if it is within the power of the agent or others involved in the decision to improve it, where improving the injustice would not itself involve more severe wrong-doing or imperil other equally important interests. In another revision of Tenzin’s case, we might imagine that his doctor could secure sufficient pain medication via a convoluted and dangerous process: He could obtain fake travel credentials for himself, cross the border illegally into India, buy morphine from dealers there, and smuggle it back into Tibet. This would involve terrible risks, significant costs, and a great deal of effort. It would, arguably, be unreasonable to expect the physician to do this, so Tenzin’s plight is not, in this instance, reasonably remediable.

For a different illustration, consider Eric again (recall: his abusive husband does not want to be responsible for his care, and so has threatened murder). Suppose Eric goes to his palliative care physician, who learns that Eric’s husband has coerced him. In most cases, the appropriate response from the physician would not be to offer MAID; it would be to call the police. Typically, however, calling the police would be reasonably expected to remedy the injustice. But imagine that calling the police would not be helpful: Eric’s husband has arranged things so that if any attempt is made to contact the authorities, he will immediately carry out his awful plan. In this instance, the pressure brought to bear by Eric’s husband might seem fixed and non-remediable, so that Eric and his physician must decide what to do in light of it. Indeed, if it were really true that Eric’s husband was very likely to kill him and his mother if Eric did not get MAID, and also really true that there was nothing anyone (other than Eric’s husband) could feasibly and permissibly do about it, then it would be permissible for Eric’s doctor to provide MAID.

My account also invokes the notion of the parties to a decision. The parties to a decision are, roughly, persons who already stand in a position to affect what happens to the agent, or who could be brought to such a position through permissible actions by others. Although we could imagine many ways things could go differently for Heyward—a wealthy merchant could decide to pay his torturers to release him, the British army could invade and rescue him, and so on—our evaluation of Hawkeye’s decision to kill him need only consider these possibilities if they are so likely that Hawkeye and Heyward themselves ought to have considered them in deciding what to do; as things stand, those other persons are not parties to the decision. This concept is admittedly vague—there are undoubtedly edge cases where it is unclear whether or not a person is a party to a decision. But this does not preclude us from having clear ideas about who counts as a party to a decision in many cases. Consider Tenzin again. As initially described, the parties to the decision whether he should drink the hemlock are Tenzin, his physician, perhaps some of Tenzin’s family members, and perhaps others in the community. We do not suppose that President Biden is a party to this decision, or that you and I are.

The parties to a decision can change. Suppose Tenzin’s doctor happens to be related to Xi Jinping (the current President of the People’s Republic of China) and could call him up and ask him to help; then, President Xi would become a party to the decision. Sometimes the parties can change by accident. Imagine that an American expat sympathetic to the plight of Tibetans happens to learn of Tenzin’s case at a party in Lhasa, when he overhears one of Tenzin’s cousins talking about it. The expat can easily get morphine for Tenzin. He contacts the physician. At this point, because he knows what is happening and is motivated to do something, he can be construed as a party to the decision; the fact that he is actually able to make a difference in what happens suggests it would be premature for Tenzin to seek MAID.

I have introduced the idea that the parties to a decision should be morally sensitive because we should not imagine that an injustice can be remedied by persons who, though contributing to the injustice, do not care to change it. As Heyward’s case is originally described, we naturally imagine that his torturers are not open to negotiation—and if there were reasons to think that they would have been sensitive, here, to moral considerations, this would change our evaluation of Hawkeye’s act. Likewise, we imagine that Eric’s husband (since he is clearly a sociopath) is not open to persuasion. What makes this case one where the injustice invalidates Eric’s choice, on my view, is that we are usually inclined to suppose that some other agent could make a difference.

On the other hand, the validity of a hard choice should reflect not only how all of the parties to the decision, including those whose actions constrain or force the choice, are in fact acting or proposing to act, but also how they could be brought to act through negotiation or permissible pressure. This is because, if others could be made to act better through easily executed actions by the agent or others who wish to promote his best interests, then better outcomes are readily accessible. Imagine again that Robert has ALS and has requested MAID in part because he has no family members or friends who are currently willing to provide care. If his physician knows that Robert has a son from whom he is somewhat estranged and suspects that the son could be persuaded to assist in Robert’s care, it would seem wrong for the physician to proceed with MAID, at least until the son’s availability can be ascertained.

Note that I am not suggesting that if one morally insensitive agent—Eric’s husband, for instance—can constrain someone’s options in a way that others cannot remedy, then his actions become justified; he does not get off the hook, morally. My point is only that it is permissible for morally sensitive persons who are appropriately concerned for the agent’s interests to help him make the best of a bad situation, despite injustice, if that injustice cannot reasonably be remedied. But injustice does not become justice simply by being beyond repair.

The view I am describing is fundamentally consequentialist. If injustice pushes someone to a hard choice, it makes him do something that, absent the injustice, would not be in his interests, and his interests would be better served by changing it. If the injustice is reasonably remediable, then we ought to remedy it because doing so promotes his best interests. On this view, however, it is not injustice itself that invalidates the hard choice it occasions; injustice is simply one reason things can go less well for someone. The intent of MAID is often to alleviate a patient’s suffering. If there are accessible ways to reduce the patient’s suffering that do not require that person’s death, then we should not provide MAID and should pursue alternatives. Likewise, if the injustice leading to a person’s suffering is remediable, then that person’s suffering is remediable, and the person should not receive MAID in the setting of that injustice. But if the injustice is not remediable, and if MAID is in the patient’s best interest given the injustice, MAID should be offered if the patient wants it.

Structural injustice

My view has implications for hard choices made in the face of structural injustices, including race- and gender-based injustices in healthcare. Although we have already considered one case of this sort—Tenzin is subject to a structural injustice—it may be worth examining this question more closely. Even here, I will claim that structural injustices make MAID impermissible only if they are reasonably remediable by morally sensitive parties.

Consider a case. Imagine that Rebecca has developed terminal, inoperable cancer. Rebecca is black, and her cancer is largely attributable to systemic, race-based injustices: She lived for many years in a racially segregated neighborhood that was contaminated by industrial waste, which caused the cancer; she was unable to get recommended screenings because she could not take time off of work, so the cancer was not caught early; she could not find an oncologist before the cancer spread because no oncologist worked near her home; and she did not get cutting-edge, effective chemotherapy because she had very limited insurance with a high deductible.

Now imagine that although she now has access to good care and lives in a more equitable place, Rebecca requests MAID because she is suffering severely from the effects of her cancer. Despite the fact that her request results from injustice, the account I have provided entails that the injustice does not invalidate it. Contributions of past injustices to her cancer have already been made, so there is now, definitively, no way to remedy them that makes a difference for her.

There are, of course, many cases where race-based or gender-based injustice does seem to invalidate a hard choice. Imagine that Sylvia is also suffering from cancer and contemplating MAID, but she is suffering mainly because implicit racial bias has led her physician to undertreat her pain.Footnote 36 Here, her request for MAID may be invalidated. But we are also apt to think that the injustice leading to it is remediable—whether her physician (who may well be morally sensitive despite his bias) starts to treat her pain more effectively, or she finds another, less biased, doctor.

It could initially seem that in Sylvia’s case, the injustice itself, and not its effects, is doing the moral work. But this is incorrect. What matters is that the injustice can be remedied in a way that makes a difference. To see this, compare a slightly different case: Jeanette has cancer, she is also not getting enough pain medication, and her pain has led her to request MAID. But she is not getting enough pain medication just because her doctor, who is of the same race, made an innocent mistake and has been giving her a dose that is too low. Here, there is no injustice, but Jeanette’s circumstances are also remediable, so supporting her decision to receive MAID would be wrong.

Note, too, that if an injustice has not adversely affected an agent’s interests, then it makes no difference to the respect-worthiness of her hard choice. Imagine that Melissa is suffering from severe pain and that her doctor is an inveterate racist and has tried to make her suffer. He has, however, also made a mistake and has accidentally given her the maximally effective, though still insufficiently effective, dose. Here, there is an injustice, but remedying that injustice will not improve Melissa’s situation. While morally condemnable, the injustice her physician has committed should not preclude Melissa from having access to MAID if she wants it.

Social Choices, Law, and Policy

In considering Rebecca’s case mentioned earlier, I claimed that although her cancer and associated pain and suffering are the products of multiple layers of systemic injustice, this should not disqualify her from receiving MAID if she wants it. It is possible, however, that while injustice should not preclude MAID at a moral, individual level, the kinds of injustice Rebecca and others have endured mean that it would be wrong to permit MAID in law or policy. This possibility is clearly relevant to the effects of systemic injustices on the health status of racial, ethnic, and gender minorities, and it is especially germane to psychiatric MAID since many persons with severe mental illness have been affected adversely by injustice. Stigma, discrimination, chronic and acute trauma, poverty, and inadequate housing all contribute to the development and maintenance of mental illness.Footnote 37 If these social ills—many of which reflect injustices—were corrected, then many persons who would otherwise request psychiatric MAID would not need to do so.

One could think, with Longmore, that so long as injustices ever affect persons from marginalized groups, MAID should never be open to them—or perhaps even to anyone else since allowing MAID for some groups but not others would likely be wrongfully discriminatory.Footnote 38 But this absolutist standard is too stringent. If our society were far more equal than it is, so that it almost always respected the rights and interests of persons from currently marginalized groups, but was still afflicted by some particular residual injustice—perhaps, for instance, black men were still more likely to receive harsher penalties for drug-related crimes than white men—this would not be a compelling reason to prohibit MAID, as the injustice is only remotely related to MAID decisions, if at all related. Or for a different example, imagine that we have expanded access to healthcare and other social goods so that virtually all persons with serious disabilities have their needs met in a highly equitable fashion. Still, roughly one physician out of a thousand succumbs to significant implicit bias against disabilities, leading them to provide suboptimal treatment when persons with disabilities enter their care. Although in any individual case, if a person with a serious disability requested MAID because his physician’s implicit bias led to suboptimal treatment, this might mean that his choice was invalid (depending on the criterion described earlier), the overall social setting does not clearly justify prohibiting MAID in policy. Consider, for comparison, that we do not think now that the fact that injustice sometimes affects members of currently privileged groups means that we should never, as a rule, permit hard choices like MAID.

A more moderate position, then, is that for MAID to be permitted, the treatment of persons in currently disadvantaged groups must cross some minimum threshold of equality—so that even if we have not achieved perfect equality, we must do much better than we are presently doing. But this moderate approach still faces problems. First, it needs to specify how much closer to equality we need to be, if perfect equality is unnecessary; and it is deeply unclear how this standard can be articulated. Second, the moderate account (like the “never” account) seems to unduly discount the ability of individuals from disadvantaged groups to make decisions for themselves. Even when they have been directly affected by injustice, it is not obvious that persons in disadvantaged groups need what is effectively paternalistic protection from hard choices—especially when the protection itself does not promise to make their lives better (since prohibiting MAID does not necessarily entail ameliorating structural injustices). There is no clear justification for limiting the options of persons who have otherwise been oppressed if the limits imposed will not resolve the oppression.

We might, in light of these considerations, conclude that the correct legislative approach to MAID and similar hard choices is to permit them irrespective of the existence of background injustice, relying on only the individual-level standard. But this is too quick. If a significant fraction of persons who request psychiatric MAID did so because they were homeless, for example, it might also be true for most of them that their homelessness is not reasonably remediable by morally sensitive parties to their individual decisions and that it is only remediable en masse through policy change. Then MAID would be individually permissible for many of them, despite their homelessness. Critics of MAID might sensibly object and insist that we should not permit MAID before adopting policies that would significantly reduce homelessness. They might ask do we really want to live in a society that helps people with psychiatric illness end their lives rather than helping them find a place to live?

While this question implies a false dichotomy—there is no a priori reason to think that the choice must be between permitting MAID or ameliorating the injustices that sometimes occasion it—the critic has still hit on something important. At the individual level, injustice invalidates a hard choice if it is reasonably remediable. Something similar is true at the level of law and policy: whether MAID should be permitted in law, given that some persons will request it because of injustice, should depend on whether the injustice in question is reasonably remediable.

On this view, whether MAID should be permitted depends in turn on a number of facts, such as how often hard choices are made in the setting of injustice and how much the injustices in question contribute to those choices. It is conceivable, after all, that many persons who would request psychiatric MAID are victims of discrimination in employment, but that even if employment discrimination were remedied, most of them would still request MAID because of other considerations, like the severity of their symptoms. Whether MAID should be permitted depends, too, on how easily law and policy can be changed in order to ameliorate the injustices in question. Changes in law and policy are rarely quick and almost never costless—they take time, political will, resources, and infrastructure. Imagine that many persons in Canada with psychiatric illness who are now suffering a great deal would not be suffering if, for the past several decades, Canada’s government had invested more in mental health research, training programs for mental health providers, community mental health resources, and housing. Suppose that this long track record of underinvestment is unjust. Suppose, too, that if the allocation of resources had been just, many persons with psychiatric illness who now wish to request MAID would not do so. Law and policy could change now, making circumstances better. But the effects of these changes may not be realized for years, even decades, and so are not likely to reduce the suffering of persons who have been harmed by these injustices so far. The fact that this injustice cannot be remedied in a way that quickly improves their circumstances suggests that they should still be permitted to access MAID.

Another important factor is the expected effect of alternative policies. Given the existence of injustice, would it harm more people, to a greater degree, to permit a certain hard choice than to prohibit it? If MAID is prohibited, will this tend to reduce the overall amount of harm to which individuals are subjected, given that some of them have been affected by injustice, or increase it? If our alternatives are, on the one hand, permitting MAID while also striving to rectify many of the injustices that push people to request it, and on the other hand, prohibiting MAID while simultaneously striving to remedy those injustices, which is likely to make more persons’ lives go better?

These considerations are fundamentally consequentialist, and one might be tempted to raise a standard objection to consequentialism here that it is wrong to violate the rights of even one person and wrong to sacrifice that person’s interests, in order to benefit others. If we permit MAID on consequentialist grounds, are we not sacrificing some to benefit others? Are we not proposing to violate their rights? But this objection does not really pertain to the question we are trying to answer. If we permit hard choices like MAID in the setting of injustice, this will not itself violate individuals’ rights—it will only mean that their choices are exercised and influenced by injustice, which is not the same. Nor do we sacrifice their interests to benefit others, unless we have strong reason to think that our options really are limited to either permitting MAID or rectifying the injustices that make it appealing.

Our question is analogous to whether we should permit contracts between consenting adults. If we allow contracts, it will inevitably mean that some of us enter them under and because of injustice. But it would be misleading to say that by permitting contracts, we violate the rights of individuals who enter them because they have been influenced by background injustices. We do not suppose that the very possibility that individual choices about contracts can be influenced by injustice means contracts should not be permitted.

Consideration of contracts is illuminating in other respects. While the net benefits of a system of contracts are considerable, and while the very existence of such a system does not wrong persons who, under a pall of injustice, participate in it, it is still essential that our system of contracts contains safeguards to minimize its negative effects—including safeguards against the effects of injustice. Examples include prohibitions on excess interest rates for loans, lemon laws for automobile purchases, and outright prohibitions on certain kinds of contracts that we think are never good enough for their participants, such as voluntary slavery.Footnote 39, Footnote 40, Footnote 41 Thus, whether law and policy should permit MAID and similar hard choices should also depend on whether safeguards can be implemented that mitigate their negative effects. Criteria for MAID, though they vary by jurisdiction, already accomplish this to some degree. If a person has a severe and truly treatment-refractory condition, and if that person’s decision to receive MAID is fully capacitated, then it is unlikely that the choice is forced by a reasonably remediable injustice.


Injustice, whether structural or perpetrated intentionally by the wrongful acts of individuals, can imperil our ability to make hard choices—choices to do things that, in most circumstances, would be worse for us, and which are only reasonable because our other options are so bad. It initially appears that if our options are bad because of injustice, we should not be obligated, or even allowed, to bear the costs of a hard choice. This impression has led many to argue that certain hard choices in medicine, such as MAID, are invalidated by unjust social conditions, so that it is wrong for physicians and others to abet an individual’s decision to obtain MAID, while MAID also should not be permitted in law or policy.

I have argued, however, that injustice does not always invalidate hard choices. It only does so if it is readily remediable by persons who are appropriately responsive to moral considerations and able to affect what happens; in such cases, the hard choice is not in the best interests of the agents since their circumstances can be improved. When injustice is insurmountable, though, it should not constrain persons from attempting to promote their own best interests as they understand them; they should be permitted to make the best of a bad situation, and others can permissibly help them do so. This is true whether the injustice is imposed by wrongful acts or is a feature of the social structure more broadly.

Whether hard choices like MAID should be permitted in law and policy, given the existence of injustice, is a more complex question; the answer depends on how rampant injustice is, how feasible it is to address that injustice, and how much it worsens the lives of persons affected by it, among other factors. I have contended, however, that the mere existence of injustice should not prevent us from permitting some hard choices. And I have emphasized that permitting hard choices does not discharge our obligation to address the injustices that occasion them.


This work was supported by a Faculty Scholars Fellowship from the Greenwall Foundation. Thanks to Andrew Peterson for helpful feedback on the manuscript. I would also like to acknowledge helpful input from Leslie Francis and Peggy Battin, who commented on early versions of the project.

Competing interest

The author declares none.



1. Kious, BM, Battin, MP. Physician aid-in-dying and suicide prevention in psychiatry: A moral crisis? American Journal of Bioethics 2019;19(10):2939 CrossRefGoogle ScholarPubMed.

2. Campbell CS. The unbearable burden of suffering: Moral crisis or structural failure? American Journal of Bioethics 2019;19(10):46–7CrossRefGoogle Scholar.

3. De Vries, R. Moralities of method: Putting normative arguments in their (social and cultural) place. American Journal of Bioethics 2019;19(10):40–2CrossRefGoogle ScholarPubMed.

4. Ho, A, Norman, JS. Social determinants of mental health and physician aid-in-dying: The real moral crisis. American Journal of Bioethics 2019;19(10):52–4CrossRefGoogle ScholarPubMed.

5. Health Canada. Final Report of the Expert Panel on MAiD and Mental Illness. Ottawa, ON; 2022.

6. Bausewein, C, Calanzani, N, Daveson, BA, Simon, ST, Ferreira, PL, Higginson, IJ, et al. ‘Burden to others’ as a public concern in advanced cancer: A comparative survey in seven European countries. BMC Cancer 2013;13(1):111 CrossRefGoogle Scholar.

7. Beckford M. Baroness Warnock: Dementia sufferers may have a “duty to die.” Telegraph 2008 Sept 18.

8. Cassilly R. Medically assisted suicide sends message to elderly that they are a burden. Baltimore Sun 2019 Feb 26.

9. Foster, C. ‘Being a burden’: An illegitimate ground for assisted dying. In: Practical Ethics Blog. Oxford: University of Oxford; 2017 Google Scholar.

10. Hale J. We are told we are a burden. Legalising assisted suicide would further devalue our lives. The Guardian 2017 July 17.

11. Hardwig, J. Is there a duty to die? Hastings Center Report 1997;27(2):3442 CrossRefGoogle Scholar.

12. Gandsman, A.A recipe for elder abuse:” From sin to risk in anti-euthanasia activism. Death Studies 2016;40(9):578–88CrossRefGoogle ScholarPubMed.

13. Longmore, PK. The strange death of David Rivlin. Western Journal of Medicine 1991;154(5):615–6Google ScholarPubMed.

14. Valentini, L. Ideal versus non‐ideal theory: A conceptual map. Philosophy Compass 2012;7(9):654–64CrossRefGoogle Scholar.

15. Francis LP. Feminist philosophy of law, legal positivism, and non-ideal theory. In: Garry A, Khader SJ, Stone A, eds. Routledge Companion to Feminist Philosophy. New York: Routledge; 2017.

16. Le Glaz, A, Berrouiguet, S, Kim-Dufor, D-H, Walter, M, Lemey, C. Euthanasia for mental suffering reduces stigmatization but may lead to an extension of this practice without safeguards. American Journal of Bioethics 2019;19(10):57–9CrossRefGoogle Scholar.

17. Lemmens, T. When a theoretical commitment to broad physician aid-in-dying faces the reality of its implementation. American Journal of Bioethics 2019;19(10):65–8CrossRefGoogle ScholarPubMed.

18. Potter, J. The psychological slippery slope from physician-assisted death to active euthanasia: A paragon of fallacious reasoning. Medicine, Health Care, and Philosophy 2019;22(2):239–44CrossRefGoogle ScholarPubMed.

19. Kious, BM. Burdening others. Hastings Center Report 2022;52(5):1523 CrossRefGoogle ScholarPubMed.

20. Delmonico, FL, Arnold, R, Scheper-Hughes, N, Siminoff, LA, Kahn, J, Youngner, SJ. Ethical incentives--not payment--for organ donation. New England Journal of Medicine 2002;346(25):2002–5CrossRefGoogle ScholarPubMed.

21. Siminoff, LA, Leonard, MD. Financial incentives: Alternatives to the altruistic model of organ donation. Journal of Transplant Coordination 1999;9(4):250–6CrossRefGoogle Scholar.

22. Wilkinson, S. The exploitation argument against commercial surrogacy. Bioethics 2003;17(2):169–87CrossRefGoogle ScholarPubMed.

23. Gov.UK. Surrogacy: Legal rights of parents and surrogates; 2023 June 3; available at,except%20for%20their%20reasonable%20expenses (last accessed 3 June 2023).

24. Osberg, B. For your first born child: An ethical defense of the exploitation argument against commercial surrogacy. Penn Bioethics Journal 2006;2(2):42–5Google ScholarPubMed.

25. Mann M. The Last of the Mohicans. In: 20th Century Fox; 1992.

26. Battin, MP. The least worst death. Hastings Center Report 1983;13(2):13–6CrossRefGoogle ScholarPubMed.

27. Mayo, DJ, Gunderson, M. Vitalism revitalized…. Vulnerable populations, prejudice, and physician-assisted death. Hastings Center Report 2002;32(4):1421 CrossRefGoogle ScholarPubMed.

28. Marina, S, Wainwright, T, Pereira, HP, Ricou, M. Trends in hastened death decision criteria: A review of official reports. Health Policy 2022;126(7):643–51CrossRefGoogle ScholarPubMed.

29. Westlund, AC. Rethinking relational autonomy. Hypatia 2009;24(4):2649 CrossRefGoogle Scholar.

30. Oshana M. Personal Autonomy in Society. New York, NY: Taylor & Francis; 2016.

31. Stoljar, N. Informed consent and relational conceptions of autonomy. Journal of Medicine and Philosophy 2011;36(4):375–84CrossRefGoogle ScholarPubMed.

32. Mackenzie C, Stoljar N. Relational autonomy: Feminist perspectives on autonomy, agency, and the social self. Oxford: Oxford University Press; 2000.

33. Enoch, D. False consciousness for liberals, part I: Consent, autonomy, and adaptive preferences. Philosophical Review 2020;129(2):159210.CrossRefGoogle Scholar

34. Wiebe K, Mullin A. Choosing death in unjust conditions: Hope, autonomy and harm reduction. Journal of Medical Ethics 2023:jme-2022-108871.

35. Kious, BM Autonomy and values: Why the conventional theory of autonomy is not value-neutral. Philosophy, Psychiatry, & Psychology 2015;22(1):112 CrossRefGoogle Scholar.

36. Hoffman, KM, Trawalter, S, Axt, JR, Oliver, MN. Racial bias in pain assessment and treatment recommendations, and false beliefs about biological differences between blacks and whites. Proceedings of the National Academy of Sciences 2016;113(16):4296–301CrossRefGoogle ScholarPubMed.

37. Allen, J, Balfour, R, Bell, R, Marmot, M. Social determinants of mental health. International Review of Psychiatry 2014;26(4):392407 CrossRefGoogle ScholarPubMed.

38. Rooney, W, Schuklenk, U, van de Vathorst, S. Are concerns about irremediableness, vulnerability, or competence sufficient to justify excluding all psychiatric patients from medical aid in dying? Health Care Analysis 2018;26(4):326–43CrossRefGoogle ScholarPubMed.

39. Lewison, M. Conflicts of interest? The ethics of usury. Journal of Business Ethics 1999;22(4):327–39CrossRefGoogle Scholar.

40. Archard, D. Freedom not to be free: The case of the slavery contract in J. S. Mill’s on liberty. Philosophical Quarterly (1950-) 1990;40(161):453–65Google Scholar.

41. Vogel, J. Squeezing consumers: Lemon laws, consumer warranties, and proposal for reform. Arizona State Law Journal 1985;1985(3):589676 Google Scholar.