Responsibility, Healthcare, and Harshness

Written by Gabriel De Marco

Suppose that two patients are in need of a complicated, and expensive, heart surgery. Further suppose that they are identical in various relevant respects: e.g., state of the heart, age, likelihood of success of surgery, etc. However, they differ on one feature: for one of these patients, call her Blair, the need for the heart surgery is due to her lifestyle (suppose she was a smoker), whereas the other, Ingrid, has not had this lifestyle, nor any other that would lead to the need for the surgery.

Some people think that:

1. We can be responsible and blameworthy for our actions and their consequences.

Some of those people also think that:

2. We can, or should, take this into account when making decisions about how to distribute healthcare resources.

For the purposes of this blog post, let’s assume 1 and 2 are true. Commonly, it is thought that, in order to be blameworthy for something, one must be responsible for it. Further, it is commonly thought that, whatever the appropriate response is to blameworthiness for something or other (assuming that there is an appropriate response), it will be negative in some sense or other. Now further suppose that Blair is blameworthy for her illness. Given 1 and 2, this fact about Blair, combined with the fact that Ingrid is innocent with regard to her illness, suggests that, at least in some contexts, we should treat them differently (or at least it would be permissible to do so). Call a healthcare policy that adopts, and reflects, 1 and 2 a Responsibility-Sensitive Policy, or RSP for short.

Proponents of an RSP might still disagree on what differences in treatment are required (or permissible). This is because they might disagree on what “responsibility” or “blameworthiness” mean, or at least, what they might imply (or, perhaps, people might agree that there are multiple senses of “responsibility” or “blameworthiness” yet disagree on which sense of it gets us true versions of 1 and/or 2). Further, even if we agree on what being responsible implies, we might still disagree on what sorts of RSPs are justified.

On some views of responsibility, when we say that someone is blameworthy for something, this implies that it would be non-instrumentally good – it would be good regardless of its effects – for the blameworthy agent to suffer or be harmed in some way as a response to what the agent is blameworthy for. This is a thought that we encounter often, particularly when we talk about retributive justifications for punishment (though whether this is essential to a retributive view will depend on how one defines “retributivism”). Call this the Strong View of responsibility.

Although many will adopt this view when it comes to punishment, I do not know of anyone who adopts it for the purposes of an RSP. Such a policy would be quite difficult to justify. When it comes to distribution of healthcare resources, an RSP that adopts an extreme version of the Strong View might imply that we are justified in expending resources for the purposes of causing harm or suffering to someone like Blameworthy Blair. Perhaps on a slightly weaker RSP, we would not be justified in expending resources to ensure harm and/or suffering, but the fact that someone is blameworthy for their illness might instead imply that we should perhaps ensure that the patient suffers or undergoes harm (even if we ultimately end up treating her). For example, we might need to ensure that Blameworthy Blair suffers a bit before we operate, or make sure that the operation would not be as good as the one we do on Innocent Ingrid (in the extreme case, we would not operate on her at all).

Another thing we might be thinking when we discuss responsibility is that, when we say that someone is blameworthy for something, this implies some sort of reduction thesis: that the agent’s claims to certain goods have been reduced, that we should not weigh their interests equally (or as heavily), or that the strength of their rights has been reduced to some extent (whether these are different will depend on one’s view of how these concepts and/or properties are related to each other. I’ll just be focusing on reduced claims to goods). Accepting a reduction thesis is consistent with holding the Strong View. For our purposes, we can say that something is a Moderate View of responsibility if it holds a reduction thesis along these lines and it rejects the Strong View.

The Moderate View will justify treating blameworthy agents differently than non-blameworthy ones. However, it will not justify doing some of the things that might be justified on the Strong View. The Strong View would provide us with reasons to go out of our way to cause harm and/or suffering to a blameworthy agent, or at least ensure that the blameworthy agent suffers and/or is harmed. The Moderate View, on the other hand, would never provide us with such reasons (though having such reasons is consistent with holding the Moderate View).

Different versions of the Moderate View can differ on how strong or weak they are, and one of the ways they might do this is by differing on the nature of the reduction thesis. On one version, blameworthiness of some health outcome can result in one’s completely losing one’s claim to some healthcare resources; call this an unlimited reduction thesis. On a different view, blameworthiness can reduce one’s claim to some healthcare resources, but one cannot lose the claim in virtue of blameworthiness for the outcome; there is some non-zero limit to how much one’s claim can be reduced. Call this a limited reduction thesis. A view on which Blameworthy Blair’s claim to surgery can be lost in virtue of her blameworthiness – one which accepts an unlimited reduction thesis – will be stronger than a view on which her blameworthiness might make it such that she has less of a claim than Ingrid, yet retains a claim nevertheless – one which adopts a limited reduction thesis.

Moderate Views can also differ on other features of the reduction thesis. As stated, the reduction thesis is general; the agent’s claim to a relevant good will be reduced, and this can justify different treatment in a variety of contexts. An RSP that adopts this general reduction thesis will not need a tie-breaker scenario in order to take responsibility into account. Consider, for example, a variation of the case above. First, suppose that the situation is such that we can only operate on either Blair or Ingrid, but not both. Further suppose that, since she is blameworthy, Blair’s claim to some healthcare resources – perhaps those involved in treating the illness she is blameworthy for – is weaker than Ingrid’s; yet, Ingrid’s chances of a successful surgery are slightly lower than Blair’s (but we are still keeping everything else equal). Depending on how we should weigh strength of claims against chances of success, an RSP adhering to a general reduction thesis might suggest that we operate on Ingrid rather than Blair in this modified case.

On a different version of the reduction thesis it can, in effect, be conditional. Suppose we are considering an RSP on which a patient’s reduced claim to some healthcare resources would only be used as a tie-breaker – in cases where all other relevant features are equal, and we cannot give treat both patients equally. In the modified version of Blair and Ingrid, we would operate on Blameworthy Blair, given that the non-responsibility features are not equal (so we do not have a tie). In effect, this would implement what we can call a conditional reduction thesis: a blameworthy patient’s claim to some resources is reduced on the condition that she is in a tie-breaker scenario.

There are, of course, many other ways that RSPs can differ; but this discussion ought to be enough to make a brief point. Sometimes, people object to the adoption of an RSP on the basis that it would justify responding to blameworthiness in ways that are too harsh. This objection can take different forms, but the point is fairly intuitive. One might think, for example, that refusing treatment to Blameworthy Blair on the basis of her responsibility for her lifestyle and its negative health outcomes is too harsh of a response to her behavior. This would be a problem for the proponent of an RSP on which we would refuse treatment to Blair. And the harsher the response, in proportion to the item one is blameworthy for (and perhaps the degree of blameworthiness), the bigger the problem for the RSP. Further, the harshness of a policy can vary depending not just on the harshness of an individual case, but also how often it justifies a response that is too harsh (or perhaps the rate of responses that would be too harsh). If an RSP allowed overly harsh treatment very rarely, the harshness objection may not be as powerful as it would be for an RSP that often resulted in overly harsh responses. How powerful this objection is to a particular RSP, I suggest, will partly depend on how strong or weak the RSP is, as I have been using these terms.

Other things being equal (a qualification one can assume for the comparative claims I make below), the harshness objection will be more powerful against an RSP that adheres to the Strong View of responsibility than it will be against an RSP that merely adheres to the Moderate View. An RSP of the latter sort would not allow us to ensure that a blameworthy patient receive worse, or no, treatment, whereas an RSP adhering to the Strong View would. This difference, at the very least, would seem to result in a difference in how common overly harsh responses are.

For RSPs that adhere to the Moderate View, the harshness objection will be more or less powerful depending on the nature of the reduction thesis. An RSP adhering to a limited reduction thesis will not be as open to this objection as one that accepts an unlimited reduction thesis. This is because an RSP with a limited reduction thesis – on which blameworthiness can never eliminate one’s claim to healthcare resources – will not lead to as many harsh responses as a view on which blameworthiness can fully eliminate one’s claims to resources. Further, the harshness of the responses may vary as well, depending on where one sets the limit for the reduction thesis; that is, depending on how weak one’s claim can get on the basis of blameworthiness.

Finally, an RSP adhering to a general reduction thesis may lead to harsh responses more often than one that only accepts a conditional reduction thesis. Recall the modified case of Blair and Ingrid above: we can only operate on one, and outside of blameworthiness for the illness, the only difference between Blair and Ingrid is that Ingrid has lower chances of a successful surgery. An RSP with a general reduction thesis may allow us to operate on Ingrid rather than Blair on the basis of blameworthiness, whereas a reduction thesis that is conditional on tie-breakers would not. If not operating on Blair, in virtue of her blameworthiness, constitutes an overly harsh response in this case, then an RSP with a general reduction thesis would seem to get us more cases of harsh responses than one that accepts a conditional reduction thesis.