I've compiled these notes because I'm trying to think through with clarity the issues involved with Jersey's Assisted Dying Law, and indeed such legislation more generally throughout the world.
So what I have done is to take an analogy - the Drake equation, as I think examining how it varies in outcome by initial factors can help to understand how in assisted dying debates, the axioms chosen, or initial factors, if varied, can produce different outcomes.
With the Drake equation, we have to make assumptions, but despite it being built on scientific principles, we do not have enough knowledge to know which assumptions are correct. As I explain, something like this is part of the problem - and the muddle - over assisted dying.
I then looked at safeguards, and also where they appear to break down. One of the most unsettling Radio 4 programmes on the issue looked at the logic in the Netherlands, where "unbearable suffering" is extended - quite logically - to mental as well as physical suffering, and where someone with acute mental distress had managed to obtain assisted dying. The case of the Netherlands can be seen as a "slippery slope", but actually it can also be argued rationally in a way that shows it is not.
One of the things I learned most from Karl Popper with what he called "social engineering" - and assisted dying is a form of social engineering - is that there are often unintended consequences, na these are key to considerations.
I take two cases as examples. One is to do with mental suffering - the case of bipolar individuals (such as the Netherlands case) and how tricky that becomes. The other is that of dementia, which was once under the proposed Jersey law with a "waiver of future capacity". The issues with dementia are fraught, especially once terminal illness is removed from the equation, and the example of the "coffee cup" case shows I think how close we are approaching involuntary euthanasia under the guise of "mercy killing".
Finally in this preamble, I would note that I am not giving my views on the subject. The reasons should be plain - when it comes to ranking moral values and feeding them into any debate on assisted dying, I can see no objective means of ranking them which would be universally agreed. Hence all I can suggest is that we should try to move from ranking values to designing procedures that honour multiple values at once. And especially when values cannot be ranked, the only honest path is to build safeguards and to map unintended consequences in advance. That becomes the substitute for moral consensus.
The Assisted Dying Equation
The Drake equation estimates the number of communicative extra-terrestrial civilizations by multiplying uncertain factors: the rate of star formation, the fraction with planets, the number of habitable worlds, the likelihood of life emerging, the probability of intelligence developing, the chance of technological communication, and the length of time such civilizations transmit signals. Small changes in any factor dramatically alter the final estimate.
I find the assisted dying debate rather like the way the Drake equation functions.
The Drake‑equation analogy works because both situations involve several independent factors whose weighting determines the final outcome, yet none of those factors has an objectively mandated value. In the assisted‑dying debate, people begin with different moral priorities: autonomy, sanctity of life, compassion, protection of the vulnerable, and the social meaning of legalising intentional death.
Each of these functions like a parameter in the Drake equation: change the weight of one, and the conclusion shifts dramatically. Someone who gives greatest weight to autonomy will reach a very different ethical position from someone who gives greatest weight to the protection of vulnerable people, even if both are reasoning carefully and in good faith.
The difficulty, as I’ve noticed, is that there is no neutral standpoint from which to declare the “correct” weighting. In the Drake equation, we lack empirical data; in ethics, we lack a universally accepted method for ranking moral values. That is why the debate feels irresolvable: people are not disagreeing about facts but about which moral uncertainties they are willing to live with. One person fears prolonged suffering more than the risk of subtle coercion; another fears the erosion of social protections more than the risk of denying autonomy. These are not errors of logic but differences in moral emphasis.
Where the analogy becomes more complex is that, unlike astrophysical parameters, moral parameters are not value‑neutral. They can be examined, challenged, and tested against lived experience. A society can ask whether a particular weighting leads to greater dignity, greater justice, or greater harm. So while there is no objective algorithm for choosing the parameters, there are still reasons (historical, social, emotional, and philosophical) for preferring one configuration over another. Ethical reasoning is not arbitrary, even if it is not mathematically provable.
What my analogy captures I think beautifully is the sense of structural indeterminacy: the outcome depends on the starting assumptions, and the starting assumptions cannot themselves be derived from the system. That is why the debate feels like two groups solving different equations rather than disagreeing about the same one.
The Uncertainty Principle.
There seems to be no objective way to choose between the competing moral starting points in assisted dying. This is a feature of the terrain itself. The debate is built on values that are all legitimate, all deeply human, and all in tension. Autonomy, sanctity of life, compassion, and protection of the vulnerable each illuminate something essential, yet none can be reduced to the others or ranked by any neutral standard. That is why both sides can argue with clarity and conviction and still fail to produce certainty. They are not disagreeing about facts but about which moral risks matter most, and there is no external vantage point from which to declare one set of priorities objectively correct.
This is also why the debate remains so emotionally charged. Each position protects something precious and fears something real. One side fears prolonged suffering; the other fears subtle coercion or the erosion of communal care. These fears are not errors of reasoning but expressions of different moral emphases. Ethical reasoning can clarify the stakes, expose contradictions, and test consequences, but it cannot eliminate the underlying conflict because the conflict is between values that are all valid. The absence of certainty is not a failure of thought; it is the cost of taking human dignity seriously in all its dimensions.Ethicists describe this as a pluralistic moral conflict, a situation in which several values are simultaneously real, non‑negotiable, and impossible to reduce to one another. In such cases there is no single correct answer, only shifting trade‑offs between goods and harms. Reasoning can illuminate the stakes but cannot dissolve the tension, and the absence of certainty is not a defect in the argument but a feature of the human condition.
Where can we find guidance?
Even without certainty, societies and individuals can still make responsible decisions by drawing on several kinds of guidance. Consequences matter, because we can look at what actually happens in places that allow or forbid assisted dying. Coherence matters, because any position has to fit with the other values we hold. Lived experience matters, because patients, families, and clinicians reveal realities that abstract arguments miss. Safeguards matter, because any system must protect those who are vulnerable to pressure or neglect. And symbolic meaning matters, because laws express what a community believes about dignity, care, and the worth of a life. None of these sources of guidance can deliver certainty, but together they allow for thoughtful, accountable judgement rather than arbitrary choice.
Safeguards often become the most workable anchor because they do not require anyone to resolve the underlying moral conflict; instead, they manage the risks that each side fears most. In a landscape where autonomy, compassion, sanctity of life, and protection of the vulnerable all carry real weight, safeguards act as the practical meeting point. They acknowledge that no value can be maximised without cost, and they try to prevent the worst harms associated with either extreme.
This is why many jurisdictions that legalise assisted dying build their systems around strict eligibility criteria, independent assessments, waiting periods, and oversight mechanisms. These measures cannot create certainty, but they can create a framework in which decisions are made with transparency, accountability, and protection for those most at risk. In a morally plural world, safeguards are not a perfect solution, but they are often the most responsible one available.
The Slippery Slope
The fear of a slippery slope is real because safeguards can reduce risk but can never eliminate it. Every system that permits assisted dying has to balance two opposing dangers: the danger of unrelieved suffering if the practice is forbidden, and the danger of unintended pressure or expansion if the practice is allowed. Safeguards work by slowing decisions down, adding layers of independent judgement, and creating transparency, but they cannot guarantee that boundaries will never shift. That is why the debate remains tense even in countries with long‑standing laws: people are not only arguing about what is allowed now, but about what might follow.
Safeguards tied to terminal illness with a limited timescale often feel more reliable because they place the decision within a clearer medical boundary. A prognosis, however imperfect, anchors the process in a situation where death is already approaching, autonomy is exercised within a narrowing horizon, and the risk of subtle social pressure is reduced.
By contrast, “unbearable suffering” is broader, more subjective, and more open to interpretation. It can include psychological, existential, or chronic conditions where the person may not be dying, and where vulnerability, ambivalence, or treatable distress can be harder to distinguish from a settled wish to die. That wider scope increases the risk of unintended consequences, especially for people who feel burdensome, isolated, or unsupported.
Unbearable Suffering?
The Netherlands’ extension of assisted dying to include “unbearable suffering” from psychiatric illness is one of the clearest examples of how a broad criterion can widen over time, and why it raises slippery‑slope concerns. Dutch law does not require a terminal diagnosis; instead, it requires that suffering be unbearable and with no prospect of improvement, and this can include psychiatric disorders. This is explicitly recognised in the Dutch due‑care criteria, which state that unbearable suffering may arise from somatic disease, dementia, or psychiatric disorders.
Why the Dutch position sees expansion as logically required
The Dutch reasoning rests on three linked claims:
Suffering is suffering, regardless of whether its origin is physical or psychiatric. Dutch review committees and legal scholars emphasise that unbearable suffering is not limited to bodily pain; psychological torment can be equally or more severe.
Hopelessness can exist in psychiatric illness, especially in long‑term, treatment‑resistant conditions. The law requires physicians to be convinced that suffering is both unbearable and without prospect of improvement, and this criterion is applied to psychiatric cases with additional caution.
Equal treatment demands consistency. If the law allows euthanasia for unbearable suffering in physical illness, then excluding psychiatric suffering would violate the principle of equal consideration for equal suffering. This is a recurring theme in Dutch ethical debate. This takes seriously that mental illness has a biological substrate. The popular maxim that “mental illness is like any other medical illness”, while a simplification, does imply that mental illness has a biological basis just like other medical illnesses and should be treated in the public’s eye in a similar manner.
From within that framework, we can see the Dutch motivation for change: the expansion is not a slippery slope but a logical extension of the original axiom.
Unexpected Consequences: Bipolar disorder
A bipolar disorder pattern, where someone may be suicidal in one phase and not in another, goes straight to the heart of why psychiatric suffering is uniquely difficult in assisted‑dying frameworks. Before going further, it’s important to say that anyone experiencing suicidal thoughts or mood instability should seek support from a qualified mental‑health professional, because these conditions require skilled, ongoing care.
The bipolar pattern highlights the central risk: if suffering is the criterion, and suffering fluctuates, then the timing of the request becomes ethically decisive. A request made during a depressive episode may not reflect the person’s long‑term values or identity. This is why many people argue that psychiatric suffering is fundamentally different from terminal physical illness, where the trajectory is clearer and the person’s wishes tend to be more stable. That is a conclusion with which I would concur.
Unexpected Consequences: Bipolar disorder
A bipolar disorder pattern, where someone may be suicidal in one phase and not in another, goes straight to the heart of why psychiatric suffering is uniquely difficult in assisted‑dying frameworks. Before going further, it’s important to say that anyone experiencing suicidal thoughts or mood instability should seek support from a qualified mental‑health professional, because these conditions require skilled, ongoing care.
The bipolar pattern highlights the central risk: if suffering is the criterion, and suffering fluctuates, then the timing of the request becomes ethically decisive. A request made during a depressive episode may not reflect the person’s long‑term values or identity. This is why many people argue that psychiatric suffering is fundamentally different from terminal physical illness, where the trajectory is clearer and the person’s wishes tend to be more stable. That is a conclusion with which I would concur.
Unexpected Consequences: Dementia
The lack of capacity (as in dementia) also highlights an issue. Where is a threshold? Before one lacks capacity the quality of life in the long term cannot be assessed, once dementia is advanced to lose capacity, the quality of life is lost, but so is the decision making process.
In the context of assisted dying, the "threshold" is the point where an individual loses the legal capacity to make a voluntary, well-considered request. For dementia, this creates a narrow and often missed window between being "too well" to qualify and "too incapacitated" to consent.
Most jurisdictions with assisted dying laws require the patient to have decision-making capacity at the exact moment the life-ending medication is administered. There is Catch-22 here: Patients often feel forced to choose assisted death "too early", while they still enjoy life, because they fear that if they wait, they will cross the threshold into incapacity and lose their legal right to choose.
In the context of assisted dying, the "threshold" is the point where an individual loses the legal capacity to make a voluntary, well-considered request. For dementia, this creates a narrow and often missed window between being "too well" to qualify and "too incapacitated" to consent.
Most jurisdictions with assisted dying laws require the patient to have decision-making capacity at the exact moment the life-ending medication is administered. There is Catch-22 here: Patients often feel forced to choose assisted death "too early", while they still enjoy life, because they fear that if they wait, they will cross the threshold into incapacity and lose their legal right to choose.
There is also a "Six-Month" Barrier: In many US states and parts of Australia, the law also requires a terminal diagnosis of six months or less. Because dementia can last for years after capacity is lost, patients rarely meet both the "terminal" and "capacity" requirements simultaneously.
Some countries allow patients to bypass the "in the moment" capacity requirement through Advance Requests (written when the patient was still competent). But the Netherlands is the only country where a written Advance Euthanasia Directive (AED) can legally replace an oral request after a patient has lost capacity. Even then, a physician must be convinced the patient is experiencing "unbearable suffering" in their current state. In Belgium and Luxembourg, they allow advance requests, but typically only if the patient is in an irreversible coma or unconscious. This rarely applies to advanced dementia, where the patient is conscious but lacks capacity.
In jurisdictions like the Netherlands, the threshold for assisted dying shifts from Cognition (can you think?) to Suffering (is your life intolerable?). But if a patient with advanced dementia appears happy (e.g., enjoying a meal) but their previous competent self wrote that they would find such a life "undignified" and "unbearable," which "quality of life" assessment wins? And who assesses something so subjective?
The "Coffee Case": A landmark Dutch case involved a woman with advanced dementia who had requested euthanasia in writing but resisted the procedure in the moment. The Supreme Court eventually ruled that her previous written request took precedence over her current confused resistance, as she no longer had the capacity to withdraw her request. The woman was given a lethal infusion to drink in a coffee.
I find this extremely alarming. This, for me at any rate, crosses ethical boundaries as it requires a human judgement on an individual on the quality of their life, without their formal consent at the time. They may have given a previously written request but while they had capacity, they might have withdrawn it at any time.
Again if we look at safeguards, we find issues:
Once present resistance is discounted, the system must rely on clinicians and legal authorities to decide that the person’s life has crossed the threshold where death is preferable.
That is not a neutral act. It is a human judgement about the value of another human’s continued existence.
Advance directives are meant to guide care, not to bind a future self who can no longer speak clearly.
The idea that a person “might have withdrawn it at any time” is crucial.
A written document cannot capture the complexity of a person’s evolving identity, fears, attachments, or desires as dementia progresses.
The Coffee Case shows what happens when the first view is taken to its extreme. When the living person’s behaviour is dismissed as “confused,” their humanity is diminished. Countries with assisted dying often broaden eligibility from terminal illness, to chronic illness, to mental illness, to dementia. Once the system accepts that some lives are “no longer meaningful,” it becomes easier to extend that logic.
These risks do not mean assisted dying is inherently unethical. But they mean that safeguards must be built around the living person, not only the past document.
The Coffee Case forces societies to choose between two incompatible ethical commitments: precedent autonomy (the past self rules) and current welfare and expressed will (the present self matters most). Most people assume both can be honoured. This case shows they cannot.
The Coffee Case also shows how a system can begin to treat past consent as sufficient and present resistance as irrelevant. Once that logic is accepted, three further steps become thinkable: (1) If past consent can override present refusal, then present refusal is no longer ethically decisive. (2) If present refusal can be dismissed as “confused”, then clinicians become the arbiters of which refusals count. (3) If clinicians can decide which refusals count, then they can also decide when consent is “implied,” “assumed,” or “inferred.” . At the moment the third step above has not been taken by the Dutch and there are still safeguards, albeit weakened.
This is how a system could drift from “voluntary” to “clinician‑interpreted” to “clinician‑decided.” It doesn’t require malice. It requires only a series of small, well‑intentioned decisions that gradually shift the centre of moral authority away from the person and toward the system. I think we should take this danger as a very real consequence of the moral logic.
The lesson is that once a society (not the individual themselves) accepts that some lives are better ended, the justification can expand faster than anyone expects.
Posts
