Does MAID Create Perverse Incentives?
A look at Noah Smith's arguments
Perverse Incentives
Smith’s argument is that medical assistance in dying creates perverse incentives, the largest of which is the potential for MAID to save money. According to Smith, “The perverse financial incentives of euthanasia seem both insidious and pervasive.” Here’s the story: healthcare spending is concentrated in a very small percentage of the population. If a policymaker were looking to save some money, that population would be the place to start, and having people in that group die sooner would cut costs.
People have been making this point for a long time, but the possibility of perverse incentives doesn’t on its own mean that the incentives are a major factor. To show this further claim, we need evidence that the incentives are being acted on because of the incentives. Smith doesn’t offer this evidence and I’m not aware of any.
For example, people in the last year of life account for around ten percent of healthcare spending in most jurisdictions, and this is the population that makes up almost all MAID deaths in Canada (96.5 percent in 2022). But this isn’t strong evidence of the government trying to save money, since being close to natural death is strongly correlated with wanting an assisted death, and a terminal illness is a requirement in all U.S. states that allow MAID.
A potential piece of evidence Smith mentions is the large increase in assisted deaths in Canada, which could be the result of perverse incentives. I’ve previously argued that there’s no fact of the matter regarding what the correct number of assisted deaths should be. Absent other evidence, the best explanation is that people are accessing it because they want it.
There are other areas of healthcare where perverse incentives might apply, yet we correctly don’t worry about them. When France recently announced its proposed assisted dying legislation, it also announced a $1.09 billion increase in palliative care. Since palliative care is cheaper than aggressive treatment, isn’t this a perverse incentive? As Smith says, “there’s the possibility that people could simply be denied life-saving treatments if health providers consider MAID a cheaper, acceptable alternative.”
Shouldn’t we oppose palliative care for the same reasons Smith is worried about assisted dying? The reason no one argues this is that palliative care is a good option that people can choose if they want it. Smith’s point applies—it’s possible that the government could pressure people into choosing palliative care—but absent any evidence that this is happening, we shouldn’t jump from possibility to actuality.
If policymakers were trying to save money with MAID, there are various strategies they might use. One would be to incentivize physicians and nurse practitioners to offer it. In practice, this isn’t what we find. For one, there’s no evidence that the government is encouraging people to become MAID providers, and the actual incentives don’t match the predicted perverse incentives. James Downar, a Canadian physician, puts it as follows: “As someone who does ICU and Palliative Care, I’ll say that MAID is the most difficult thing I have experienced emotionally, and I can make more money doing literally anything else.”
MAID Isn’t the Only Choice
Smith mentions the story of Allison Ducluzeau, a woman in British Columbia who was denied cancer treatment because her prognosis was poor. She then paid for the treatment herself in the U.S. According to Ducluzeau, the surgeon told her “[Y]ou have a life span of what looks like to be two months to two years. And I suggest you talk to your family…talk to them about your wishes…you know, whether you want to have medically assisted dying or not.”
Ducluzeau had to wait months to see the surgeon, which is unacceptable and is too common in Canada. But, given the surgery’s high cost and its limited benefit, it’s unclear if she should have been able to have it. Smith thinks she should have. He imagines how it would have gone if she had been unable to pay for the treatment herself:
She would have had little choice but to accept euthanasia instead. It’s not 100% certain that financial incentives were involved in the Canadian health care system’s decision to deny Allison Ducluzeau the treatment that might have saved her life, and to instead urge her to kill herself.
This is a stretch. Smith is saying that MAID likely caused a perverse incentive, so banning MAID would prevent this problem, but this doesn’t follow. First, it’s false that “she would have had little choice but to accept euthanasia”, since she could have chosen palliative care or no treatment instead. Denying someone cancer treatment doesn’t force them to choose MAID. But the bigger problem is that MAID doesn’t create the perverse incentive here. If the government wanted to save money, they could have denied Ducluzeau treatment regardless of MAID being available. The perverse incentive is the much bigger one that healthcare costs money and governments have a limited amount of it so, insofar as there’s a problem, Smith is focusing on the wrong source.
It’s also a stretch to conclude from the quotation Ducluzeau provided that the surgeon “urge[d] her to kill herself”. Telling a patient to talk to her family about whether she wants an assisted death or not is hardly urging her to get MAID. And the surgeon has no incentive to urge this choice. Regardless of what she chooses, he won’t be paid more or benefit in any other way. Once again, there’s no clear link between Smith’s claim and the facts.
Depression
Smith says the following:
In other words, some people simply aren’t in their right mind, and thus aren’t equipped to make the choice of whether to die. Lots of people intuitively realize this, which is why Canada has temporarily delayed its plan to authorize MAID for the mentally ill.
First, this isn’t why Canada postponed this change. The given reason is that the provinces claimed they weren’t ready. Second, I’ve written about mental disorders and decision making at length before. Smith is correct that depression and other mental disorders can interfere with decision making, but he’s mistaken when he implies that people with depression can never make informed decisions about ending their own lives. The evidence on capacity doesn’t support this. There are also many mental disorders, so we shouldn’t reason from depression to all of them.
The change that Canada postponed would have allowed people who have a grievous and irremediable medical condition to access MAID, regardless of whether their medical condition is physical or mental. To qualify, a person with depression would need to show that his condition is irremediable, meaning that treatment attempts hadn’t succeeded. Thankfully, many people with depression do respond to treatment, but some don’t. Treatment-resistant depression exists, and psychiatry already harms people by pushing treatments that don’t work. And as I’ve argued, the evidence shows that treatments for many mental disorders just aren’t that effective.
The delayed proposal would have allowed a very small percentage of people with a mental disorder to qualify for MAID. Having depression on its own wouldn’t do it.
Smith concludes by offering some proposals, most of which are sensible. It’s worth mentioning that, to be consistent, he should advocate for a more restrictive view than the one he does. One of his proposals is “Keeping the rule that mental illness alone is not grounds for euthanasia.” Right now, having a mental illness doesn’t disqualify someone from MAID if they meet all the criteria.
However, if Smith thinks that depression makes informed choice impossible, he should go further by advocating that having a mental disorder should automatically disqualify someone. Right now, the law recognizes that people with mental disorders can still have the capacity to make informed decisions, which the evidence supports, but is inconsistent with his claims. My view aligns with current law: people should have the right to demonstrate their ability to make decisions. Some people won’t meet the criteria, but they should be permitted to be assessed.
Ultimately, perverse incentives aren't something we can dismiss out of hand. Like many things in healthcare, abuse is possible. But Smith fails to show that anyone, from governments to providers, is acting because of perverse incentives. Therefore, there isn’t evidence to support his claims.
Quick Hits
A Timely Reminder that the Courts Need to Keep their Noses Out of Individuals’ Medical Decisions
The Pope once again proving he’s a moral leader by opposing surrogacy and trans surgery: Gender-affirming surgery threatens ‘unique dignity’ of a person, Vatican says
What are the key findings of the NHS gender identity review?
I too found his discussion of incentives in health care somewhat lacking. The neglected part is that the "cost" of the health care is actually a source of revenue for the people and organizations that are mostly delivering health care. The incentive for the actors in the system (doctors, hospitals, drug companies), where there is one, is actually to deliver as much health care as possible, long past the point where the patient wants it or stands to benefit.
The groups who stand to benefit from reducing costs (e.g. government/taxpayers in Canada, insurance companies in the US) are often actually those with the least ability to alter individual patient behaviour. It is weird to me that so many discussions of MAiD amount to "yes I agree it should be available, but it should be really bureaucratic and complicated so we can make sure that fewer people get it".
There's quite a bit of evidence to support the argument that clinicians are influenced by advertising, gifts, visits from pharmaceutical representatives, etc. Clinicians will choose the more expensive, less good thing for their patients under these influences. However, I'm not sure there's any evidence to suggest that clinicians, on a patient-by-patient basis, make decisions to steward healthcare resources, e.g., "Wow. Healthcare costs are out of control. I'm going to order the cheaper medicine for this patient today." This is why, I think, there's so much discussion around system-level controls to mitigate healthcare costs (from policies and formularies to nudges within electronic medical records).
I doubt it would play well to see an organization, whether private or governmental, to make a policy obviously favoring assisted suicide/euthanasia (ASE) over any medical intervention.
However, I imagine two avenues of perverse *disincentive*:
1. If there is a population of people that present a bedeviling problem to the healthcare system - maybe their care requires a lot of time, money, or resources - and many or most of them avail themselves of ASE. Insofar as there's an impetus to provide these resources, that impetus becomes less forceful or urgent when more people die before pressing their needs upon the system. How could you prove that this is a perverse disincentive? It would manifest as a negative; a lack of societal, political will to improve conditions. Now, healthcare the world over appears like a wicked problem even before the introduction of ASE, so I don't think you could establish to what degree it degrades the political will to improve the system, but I doubt it can make it better.
2. Most people live by default. They're not waking up every day choosing not to kill themselves. When they're offered death, they can choose to live or choose to die, but one thing they can no longer do is live by default. The mere offer of the choice imposes upon them the task of weighing living vs. dying. They'll need to justify their choice to live (if only to themselves), something they may not have done before. This is quite a burden to foist on someone who is already vulnerable. I call this a perverse disincentive because it may subtly dissuade a patient from longevity-focused healthcare when they fail to justify their continued to living to themselves (e.g., Hardwig's "duty to die"). I'll say, in my neck of the woods, this is particularly potent as many people report they "do not want to be a burden to their loved ones."
The challenge of a starkly preference-based view ("It's what they want") is that fails to acknowledge that it's our culture that taught people the content of these beliefs (e.g., about perceived burdensomeness), and their choices will go in to make the culture in which we live. It's hard to establish how any given choice or policy will change a culture at some point in the future. But given what we know about the influence of perceived burdensomeness on "classic" suicidal ideation (e.g., in depression), these considerations should give us pause when considering perverse disincentives.