Transcript: Assisted Dying 

Tim Phillips talks to Al Roth 

This episode discusses assisted dying, end-of-life choices, and suicide. Some listeners may find the content distressing. 

We can’t avoid thinking about trade-offs, and we require evidence to think about them clearly, and we should not hesitate to look at evidence and gather it where we can. 

Death comes for us all, but few of us choose the time or the manner in which we die. In many countries, the law permitting medical assistance in dying is changing, or is about to change, but there is strong disagreement about how, or whether, assisted dying should be legalised. My guest today is the leading scholar on how we can use economics to resolve hard controversies like this one. So today on VoxTalks Economics, What economics can tell us about how we choose to die? When two sides of a debate can’t reach an agreement on moral or religious grounds, can we design policies and markets that mean we don’t allow everything or ban everything? Al Roth of Stanford has written a book called ‘Moral Economics’ that deals with these issues. Not just assisted dying, but also access to contraception, IVF and abortion. Should we be able to purchase drugs, and which ones? Do we pay people to donate blood plasma, or even a kidney? Ahead of the launch of the book, I spoke to Al about how the debate about how we die is an example of how we try to balance the right to pursue our own interests, while protecting the most vulnerable from harms. Al, welcome to VoxTalks Economics. 

I would like to start with a quote from your book. ‘I am still active, enjoying many things in life, except the daily news, and will die a happy man, but my kidneys are on their last legs; the frequency of mental lapses is increasing; and I am 90 years old. It is time to go.’ Who wrote that? Danny Kahneman wrote that in an email to his close friends before going to Switzerland to end his life in a medical aid-in-dying clinic there. Why did he have to fly to Switzerland to do that? 

Well, there aren’t so many places that allow medical aid-in-dying. And he lived in New York, which didn’t yet allow it. They may allow it this year. But New York, like most American places that do allow medical aid-in-dying, only allows it for people who have terminal diagnoses. And Danny’s unusual decision was he wanted to end his life while he was still firmly on his feet. 

Now, you call medical aid in dying a repugnant transaction. It’s a very specific term with a very specific meaning, isn’t it? What does it mean? 

Okay, so you’re right. I use it as a technical term. And one reason it’s not in the title of my book is it was in the working title, but people would say, “Oh, you think medical aid in dying is repugnant?” And I’d say, “No, no, I’m not forming a judgment. I’m saying it’s a controversial market which some people object to.” So that’s what I mean by repugnant. A repugnant transaction is a transaction that some people would like to engage in, and other people think they shouldn’t be allowed to for moral or religious reasons. And without obvious negative externalities to the people objecting. So the way I put it in the book is, it’s a repugnant transaction if some people want to do it, other people don’t think they should be allowed, even though those other people can’t tell if it has happened unless someone tells them. 

Medical aid-in-dying is an example of this. It’s an example that’s on a lot of people’s minds at the moment. And I guess if your moral position is that ending your life voluntarily is wrong, then there’s no room to make policy about this. Do we know how many people think in this way and why they think in this way? 

Well, there are longstanding religious objections, and there are objections in the medical community: in the 5th century BCE, Hippocrates formulated an oath in which his students promised not to participate in medical aid-in-dying, promised not to help people end their lives. So it must have been controversial even in Hippocrates’ time. He wouldn’t have had to include that in the oath if doctors weren’t being asked to move the death process along in its final stages. But among doctors, sometimes the objection is, you know, our goal, we doctors, our goal is to save lives, not to end them. And it will complicate our relationship with our patients. And in the religious community, sometimes the feeling is that this is a godlike decision, you know, to give life and to end life, and people shouldn’t intervene. There’s also some sense sometimes that suffering has a purpose. But the reason more and more jurisdictions are allowing medical aid in dying is the feeling that medicine is supposed to help alleviate suffering and that long, painful, extended, difficult deaths are something to be avoided. 

One of the reasons to discuss end-of-life care is that, when we expect to live longer, it affects the decisions we make. In February, we spoke to Martin Ellison and Julian Ashwin about what decisions seniors will take about their later years, and whether policy can accommodate both their abilities and their needs. Follow VoxTalks Economics to listen to our episode, ‘The Economic Consequences of Living Longer’. 

Your book takes many of these issues from the point of view of how an economist might think about it. If we’re thinking about it this way, we might not have a moral objection to medical aid-in-dying, but we might still have concerns about it. And I guess we should then call on evidence, what we know about how this would work in practice so far, from where can we find that evidence at the moment? 

So there are a number of countries and American states that allow medical aid in dying under different rules and regulations. So I think we’re starting to get a good deal of incremental evidence about how it works. In the U.S., there are 12 or 13 jurisdictions now, you know, states plus the District of Columbia, that allow medical aid-in-dying. And almost all of them, maybe all of them, require an imminent diagnosis of death, a terminal diagnosis. So I think typically they say you have to be reasonably sure that death will come within six months, something like that. So that, of course, wouldn’t have suited Danny Kahneman. There’s Canada, which has a medical aid-in-dying that’s quite a bit more liberal. They don’t say you have to have a terminal diagnosis. They say you have to have an irremediable grave condition that can’t be fixed, so that you’re suffering, but might not die in six months, but maybe want to die now. So the Canadian law is a little different than the American laws. The American laws are really about aid-in-dying, about altering the death process a little bit, whereas the Canadian law is about ending life sometimes. Both in Canada and the U.S., most people who take advantage of these laws are elderly, late-stage cancer patients. 

Sure, sure. And if we flip this on its head… and think about the places in the world in which medical aid-in-dying is not permitted, do we have evidence that because it is not permitted, it does not occur? 

On the contrary. In other words, one of the big themes of my book is that when you put legal bans in place for a repugnant transaction –  and remember, part of the definition of a repugnant transaction is some people want to do it – when you put legal bans in place, you often get black markets, and depending on how the legal ban works, these may be black markets run by criminals, or they may just be covert markets of various sorts, and I think there’s a good deal of evidence that there is medical aid-in-dying, even where it is not legal,  because the same drugs that relieve pain can end life. So when people are dying in pain, treating their increasing pain is something that also can end their life. And when I talk to American physicians, they seem very familiar with the idea that sometimes painkillers are given to permanently end the pain in cases of terminal illness. Incidentally, you began by saying if you have a moral objection to something, then there’s no further discussion. And I’m not sure that’s true for exactly this reason. In other words, sometimes we have a moral objection to something – the best example from my book is like heroin addiction. Nobody likes heroin. So we have, at least in the United States, we have very strict laws against it. When we find dealers or even addicts, we imprison them for a long time. Some years we have 100,000 overdose deaths a year. That is, we don’t like heroin. It’s against the law, but we have it. I think selling heroin is morally wrong, but I can’t help but notice that making a law against it doesn’t make it go away. So I think one of the things that economists bring to the discussion is sometimes you have to think about trade-offs when you have no alternative. And putting some morally contested market in the hands of criminals to run is not always the best outcome. 

I might worry, even if I have no moral objection to medical aid-in-dying, that if we have a policy where it is permitted, People might be in a state of mind, for example, or a temporary state of pain and seek death. And then they would regret it later if they had the chance, which of course they don’t. Is there evidence on that? 

There’s certainly concern about that. And so many of the states and places that allow medical aid in dying require waiting periods and multiple medical committees, things like that. So it’s not meant to be a treatment for acute depression where you wake up feeling terrible and decide to end your life. And indeed, suicidal ideation is sort of a psychiatric diagnosis. You know, if you’re a psychiatrist, if you’re a doctor and someone comes to you and says, I feel terrible, I’m thinking of killing myself. One of the things you as a medical person are obligated to do and can do as you say, it might be a chemical imbalance in your brain. You know, it might be something we can treat, you know, give us a chance. And often that works. And there are plenty of people who’ve contemplated suicide and then live full, happy lives. So, in the places where medical aid in dying is restricted to people with imminent terminal diagnoses, that’s less of an issue. In other words, if you’re already near death, and now we’re talking about managing the process. But in Canada, in Belgium, in Switzerland… They certainly have to also think about these other things. 

Mental health for a long time was perhaps not considered on a path, certainly for treatment, as physical health problems were. How does medical aid in dying as a policy cope with people who have… mental health problems, and wish to die. 

So there are different regulations in different jurisdictions. In the United States, that doesn’t count as a terminal diagnosis. You’re doomed to treatment or to taking matters into your own hands. You know, one of the interesting things to read in the Canadian Supreme Court decision that legalized medical aid in dying in Canada is they said, part of the reason for this is because Canadians have a right to life, to security in life, and if we take away the option of medical aid-in-dying, then there may be people who are still well enough to take their own lives who will feel compelled to do so because later in their disease, they may no longer be able to. By allowing medical aid-in-dying, said the Canadian Supreme Court, we’re trying to prevent people from feeling the need to take their own lives. 

Gosh, that is a remarkable way to think about it. What about the problem which is often raised that vulnerable populations might feel pressurized to die? 

Yep, that’s a real concern. I don’t know too much about that, but there are some papers on it. And there’s a medical ethicist named Ezekiel Emanuel at the University of Pennsylvania who’s written a paper with co-authors. And he says there’s no evidence on a population level that patients from vulnerable parts of the population are accessing medical aid-in-dying more than the rest of the population. So it doesn’t seem to be that medical aid in dying creates a focus on the vulnerable. That, of course, doesn’t mean that there aren’t individual cases where someone who’s dying might be pressed to end it earlier than they want to because of concerns of other people, of caregivers or family. Incidentally, one of the arguments that’s made against medical aid in dying is that sometimes people will choose to die or to die more quickly, when they’re dying, because of concerns about their family and caregivers. And it’s not obvious to me that that’s out of bounds. In other words, we all think about our family and their welfare, and that’s one of our concerns. So, you know, if you’re facing a long, difficult, painful death, that’s part of the pain, is that you’re causing pain to your family. 

I’m in the UK, and last week, a bill to permit medical aid-in-dying, has failed because our politicians could not resolve their differences. 

Soundbites from assorted UK MPs 

That’s how I’ve tried to approach this issue, as a human being, not as an MP, and a mixture of really sad, really upset, really disppointed, but, also a little bit angry as well….   

What’s going on here is a minority are engaged in a philibustering …  

Personally I think the law is fine where it is. We should debate it, we should keep debating it, but people really need to also understand the consequences of changes …. 

We’ve been doing seven days of committee so far. That’s about 5 hours a day – over 35 hours – and we haven’t got through section one yet.  

It is very, very difficult to define any kind of policy by debate. I know that these debates are also happening in many legislatures around the world at the moment. If we think about this as economists, could we consider the costs of a ban on MAID to an individual or the family? How might we think about those costs and how might we quantify them? 

Well, being prevented from doing something you want to do is a cost that economists recognize, right? I mean, if when you’re optimizing your decisions, you decide to do something, but it turns out to be illegal, that… That adds cost. Economics is about making trade-offs. People have particular concerns. Some of them, as you say, are inalterably opposed to medical aid-in-dying or to treating drug addicts as patients rather than criminals. But some are willing to consider the consequences. And I think as economists, it’s natural for us to consider consequences. That is, we would like there to be no drug addiction. We would like no one to die agonizingly in pain, but we aren’t able to manage those things. And so the question is, what are the alternatives? And some of the alternatives might be less desirable than a regulated legal market in these repugnant transactions. 

So is it legitimate to think about the costs to society of a ban on medical aid-in-dying? 

It’s expensive to keep people alive when they’re near death and trying to die. And if we’re talking about very near death, then we’re talking about no prospect of bringing them back to health. So there’s a lot of discussion in the United States, at least, about the end-of-life medical costs. And that’s a confusing discussion because some end-of-life medical costs are thrown down the drain. We do lots of complicated medical procedures to someone who seems to need them and they die anyway. But of course, sometimes they don’t die. Sometimes we do complicated things like heart transplant, say, and the patient is restored to life. And 20 years later, they’re at the weddings of their children and all of that. And then the costs that were paid are well worth it in a wealthy society like ours. So talking about the end of life medical costs is a complicated issue. But talking about the end of life medical costs when someone has a terminal disease is clearer. And if in addition, they’re in agony, then I think not only does society save medical costs in allowing them to exit more quickly and with more dignity than they might otherwise have, but but it might also be good for them. We’re talking about something that is a patient choice. In particular, right now, among the regulations is medical aid in dying is not available to people with dementia. Yeah. You have to be able to give clear consent, which is something debatable. That is, you could imagine writing your medical power of attorney to say, I don’t wish to continue to live when the people who will take over my medical care when I can no longer supervise it believe that I’m in great pain. You know, so even though I’ll no longer [be] able to consent, I direct them to go ahead with a medical aid-in-dying. But right now that can’t be done. It has to be first person clear consent. 

As an example of your expertise, which is mechanism design, when we do design a mechanism for this process, are there consistent ways in which it is designed: particular steps and guardrails that are put in place? 

Well, when you look at the legislation that allows medical aid in dying, it tends to require more than one doctor giving the terminal diagnosis. It often requires that the patient take the medicine him or herself, right? Which sometimes precludes people who become too sick to swallow pills, for example. But the idea is to separate medically assisted dying from homicide. And one way some of the laws, particularly in the United States, tend to do that is they say the patient must self-administer the drugs. There are other jurisdictions where they allow the doctor to administer the drugs. And is it also legitimate 

to think about the choice architecture here that we often think about in other contexts? For example, when is this offered? How is it offered? 

Yeah. So there’s been a recent New York Times story about a patient in Canada who had intense back pains and went to the emergency room and was asked, would you like to consider medical aid-in-dying? 

Yeah, yeah. And this is just inappropriate. I mean, this is not part of ordinary medical treatment where the idea is to find out why there was a fractured bone in her, somewhere in her back. And she was cured and she traveled and danced and did other things. But that’s a mistake. Some junior doctor presumably was jumping to a conclusion about this somewhat elderly woman who came into the emergency room and he may have thought that she was living in intense pain all the time, as opposed to that she just suffered a bone break. So that’s a mistake. But I don’t think you want to ban medical assistance for dying people because sometimes it’s offered to people who aren’t dying by mistake by an inexperienced physician. You should help train the physicians better so that they don’t do that. I mean, that clearly would be worrisome if you came to the emergency room and they said, are you coming because you want ordinary medical care or because you want to die? 

Yeah. So as you say, if there are particular rules, contexts in which it can be mentioned, then I guess this is a policy that many people would defend. 

I think that’s right. New Zealand, I remember looking at one of their government web pages and it had a cheerful message that said something like, medical aid-in-dying is a new service offered in New Zealand. 

And you know, you go, oh, I mean, that’s good to know. You might not want to know it if you’ve just been carried into the emergency room because something bad just happened to you. 

This is a topic that really does make you think twice about many aspects of medical care. Is this a topic that you think should be debated by experts or should everybody be informed, have a point of view, join in this debate? 

I think everyone should join in the debate, but it should be informed, at least in part, by expert discussion of what it means to die, what it means to have… cancer in your bones that can’t be treated and that is painful. Similarly, my book is full of medical kinds of questions about donating a kidney, for example, or donating blood plasma. I mean, blood plasma is an interesting example. In many places, including Britain, it’s thought to be somewhat immoral to pay people to donate blood plasma. And besides, you don’t have to ask people to do that because you can buy as much as you need from the United States. And that’s where Britain gets most of its blood plasma. And that’s because we pay plasma donors. So there’s an element sometimes of ‘out of sight, out of mind’ to some of these moral contests. 

Al, it’s fascinating talking to you about this. I’m sure this has caused many people to think again about this. The best thing I can advise them to do is to read your book, which is on the one hand, very easy to read, but on the other hand, not an easy read in many parts, just because of the topics you take on. Thank you for taking them on. Thank you for talking to us about it today. 

I hope that people will read the book and think about it. And one of the things I say is that we can’t avoid thinking about trade-offs, and we require evidence to think about them clearly, and we should not hesitate to look at evidence and gather it where we can. 

The book is called ‘Moral Economics, What Controversial Transactions Reveal About How Markets Work.’ The author, of course, Alvin Roth, published on the 21st of May, 2026, by Basic Books. 

VoxTalks Economics is a Talk Normal production. The assistant producer is Megan Bieber. And our editor is Andrei Zagaryan. 

Next week on VoxTalks Economics, the economic and social implications of a decade of migration.