Johns Hopkins University Press

Several ethicists have defended the use of responsibility-based criteria in healthcare rationing. Yet in this article we outline two challenges to the implementation of responsibility-based healthcare rationing policies. These two challenges are, namely, that responsibility for past behavior can diminish as an agent changes, and that blame can come apart from responsibility. These challenges suggest that it is more difficult to hold someone responsible for health related actions than proponents of responsibility-sensitive healthcare policies suggest. We close by discussing public health policies that could function as an alternative to contentious, responsibility-sensitive rationing policies.


Many serious health conditions are caused or exacerbated by lifestyle factors. This fact, combined with widespread limitations in healthcare resources, has given rise to public policy debates over whether (and if so, how) the State ought to take active measures to ameliorate the burden on the healthcare system caused by a growing number of people responsible (in part or in whole) for their own ill-health (Pillutla, Maslen, and Savulescu 2018; Resnik 2013; Campbell 2017). Call the problem of how best to answer this question the lifestyle illnesses problem.

Ethicists and public policy researchers have offered two kinds of positive proposals for public policy responding to the lifestyle illnesses problem, which we will call respectively responsibility-sensitive and responsibility-insensitive proposals. According to responsibility-sensitive proposals, existing criteria for the allocation of resources ought to be reformed across [End Page 53] a wide range of healthcare services, ranging from the allocation of vital organs to the allocation of vaccines in an influenza pandemic (Albertsen 2016; Littmann 2014). The proposed reforms involve making the criteria for allocation responsibility-sensitive: according to such criteria, people who have fallen ill through no fault of their own should receive priority over people who are responsible for their ill-health.

Responsibility-insensitive proposals, like responsibility-sensitive proposals, involve the adoption of active measures for reducing the amount of resources allocated to people with lifestyle-related illnesses. However, unlike responsibility-sensitive proposals, the measures proposed do not involve the passing of any judgment on the extent to which a given individual is responsible (in part or in whole) for their ill-health. They are not (or, at least, not necessarily) based on assertions about the blameworthiness of people for their patterns of consumption. Rather, these proposals seek to prevent individuals from engaging in risky behaviors that may endanger their health. For instance, one kind of responsibility-insensitive proposal involves an increase in State-sponsored preventative measures discouraging people from engaging in unhealthy lifestyles by heavily taxing cigarettes, alcohol, fast food, sugary foods, high-sugar soft-drinks, and other products likely to cause harm to a person’s health (Briggs 2019). These products are already taxed in many countries; what is distinctive about responsibility-insensitive proposals is that they use these and other measures as a means for reducing the overall allocation of healthcare resources to those with lifestyle illnesses. In this spirit, some have suggested an increase in taxation proportionate to the scarcity of healthcare resources available for treating the health problems that these products cause, and a redirection of the revenue from such taxation to the improvement of public healthcare infrastructure (Joyner and Warner 2013; Anonymous 2013).

It is our position that responsibility-insensitive proposals are superior to responsibility-sensitive proposals in a number of ways. Firstly, from an ethical perspective, we hold that responsibility-insensitive proposals provide a more rational approach to decision-making under uncertainty, for, by pre-empting rather than implicitly punishing irresponsible behavior, they do not run the risk of wrongly blaming someone for behaviors that they in fact ought not to be blamed for. Furthermore, from a metaphysical and metaethical perspective, we hold that responsibility-sensitive proposals are implicitly committed to false assumptions about the nature of responsibility, and that it is only responsibility-insensitive proposals [End Page 54] which succeed in avoiding any such implicit commitments, in virtue of their eschewal of measures which implicitly pass judgment on individuals’ responsibility for past actions. In short, we hold that responsibility-insensitive proposals provide means for addressing the problem of lifestyle illnesses which are ethically, metaethically and metaphysically less controversial.

While it is thus our position that responsibility-insensitive proposals are superior to responsibility-sensitive proposals in addressing the problem of lifestyle illnesses, the aim of this paper is not to establish this conclusion. Rather, our aim is merely to take a step towards it, by providing needed clarification of the metaphysical and meta-ethical assumptions underlying responsibility-sensitive proposals. It is perhaps tempting to think that both responsibility-sensitive and responsibility-insensitive policy proposals are metaphysically and meta-ethically on par (Brown and Savulescu 2019). However, in this paper we argue that responsibility-sensitive proposals alone are committed to the possibility of attributing responsibility to an individual for actions they have committed in the past. This commitment is far from being metaphysically innocent. As Brown and Savulescu note:

“attending directly to the diachronic nature of responsibility for habitual, health-related behaviour will require a commitment to some view of personal identity: the extent to which it is continuous over long periods of time, in what situations it becomes discontinuous and the ways in which identity interacts with responsibility”

In this paper, we seek to clarify the implicit commitments underlying responsibility-sensitive proposals, by making explicit two sets of assumptions regarding the nature of personal identity and responsibility. In the course of this clarification, we show that at least some well-known accounts of personal identity are incompatible with the assumptions underlying responsibility-sensitive proposals; furthermore, we show that some prominent accounts of responsibility are incompatible with the conception of responsibility presupposed by responsibility-sensitive proposals. Our goal in doing so is clarificatory: while we regard the conceptions of identity and responsibility underlying responsibility-sensitive proposals to be themselves problematic, we seek to identify and clarify (rather than to argue for or against) these conceptions by way of contrast and comparison with rival accounts of identity and responsibility. Nevertheless, we hope this clarification will provide needed groundwork for further critical discussion of responsibility-sensitive proposals. [End Page 55]

In section one of this article, we provide a more detailed account of recent responsibility-sensitive proposals with regard to the problem of lifestyle illnesses, with an eye towards highlighting assumptions regarding responsibility and personal identity implicit in these proposals. In section 2.1, we discuss in more detail the assumptions regarding personal identity by contrasting them with a well-known alternative account of personal identity due to Derek Parfit. In section 2.2 and 2.3, we discuss in more detail the assumptions regarding responsibility by contrasting them with recent alternative conceptions of responsibility, namely, Fischer and Ravizza’s reflection-based conception of responsibility and Tannenbaum’s account of responsibility as separable from blame. We close by considering discussing health policies that could function as an alternative to contentious, responsibility-sensitive rationing policies.

To be clear, our aim is not to defend an alternative account of identity or responsibility. We do not, for example, endorse Parfit’s account of personal identity. Rather, the aim of our paper is to show that responsibility-sensitive proposals are more controversial than one may think. Without committing to some substantive alternative account of identity or responsibility, we wish to flag a series of potential concerns about responsibility sensitive criteria that are additional to the concerns about responsibility-sensitive criteria that have already been raised in the literature (Denier 2005; Gheaus 2017).

Before we begin, it is worth clarifying that the primary notion of responsibility this paper is concerned with is that of moral responsibility. By moral responsibility, we have in mind a person’s susceptibility to praise or blame for actions that they have committed. We will explore the meaning of moral responsibility in greater detail in section three of this paper. Suffice to say here that moral responsibility should not be confused with causal responsibility. Causal responsibility refers to a causal relationship that exists between an agent and an event. For example, someone may injure themselves by stubbing their toe against a rock. Importantly, causal responsibility is a necessary, but not sufficient, criterion for moral responsibility, since an agent may reasonably be unaware of the effects of her choices (for example, when one flicks a light switch that has been secretly rigged to set off a bomb).


Responsibility and rationing has been a central topic of discussion in healthcare ethics discourse in recent years. As medical technology improves [End Page 56] and people begin to live longer, we need to consider when and under what circumstances treatment should be provided to patients. When allocating interventions such as organ transplants or vaccines, for example, we may have to choose between one patient who has a lifestyle related illness and another who has an illness that is unrelated to lifestyle.1 One must decide whether personal responsibility for illness provides sufficient grounds for giving one patient less priority than others.

One intuition motivating responsibility-sensitive criteria for rationing is the idea that freedom comes with not only rights but also responsibilities— principal among which is the responsibility to accept the consequences of one’s own actions. Plausibly, we should provide healthcare—even expensive and resource-intensive care—for persons who have fallen ill or sustained an injury through no fault of their own. It is less clear that we have an obligation to provide care for persons who experience illness or injury as a result of their reckless decisions.2 One relevant difference is that persons who are responsible for their illness cannot appeal to misfortune in the same way that others might. Someone who is experiencing liver failure as a result of a genetic disorder, for example, is the victim of a genetic lottery, whereas someone with Alcohol Related End Stage Liver Disease (ARESLD) is not.3

Some egalitarian political theorists argue that equality requires that we provide healthcare to everyone who needs it (Nielsen 2013). Yet other ethicists take personal responsibility into account when determining how we should distribution resources such as healthcare. Luck egalitarian political theorists seek to neutralize the effects of luck (or, at least, certain kinds of luck) on distribution (Segall 2016). They differ from other egalitarians in that they give particular importance to the agency and responsibility of the agents involved. According to luck egalitarians, the responsibility that agents bear for their actions can be a reason to give greater or lesser importance to improving their current condition (Arneson 2004). In the case of healthcare, luck egalitarians state that people can weaken or entirely forfeit their own claim on medical care when they freely and knowingly engage in behaviors that are inimical to their health. Albertsen, for example, has argued that we may be justified in deducting points from the MELD scores of persons with ARESLD who are in need of liver transplant (Albertsen 2016).

Most philosophers acknowledge that addiction greatly diminishes personal responsibility. Addiction suggests that someone is not acting in a rational way, but is rather acting based on compulsion. It is difficult to say that a heroin addict is responsible for their behavior in the same way [End Page 57] that a fully rational adult is responsible for their everyday considered decisions. Even still, there are theorists who argue that people in the throes of addiction can still be said to retain responsibility in certain ways. Addiction often begins with a conscious decision to engage in some kind of reckless activity, and often the agent is cognizant of the risks. Thus, Albertsen writes, “addiction does not rule out that one could be considered responsible for initiating the abuse leading to said addiction or responsible for whether or not one seeks counselling” (Albertsen 2016). Even if someone loses control after the psychological impacts of addiction set in, they can still be said to have been autonomous at the time at which they first engaged in the addiction-conducive activity.

In addition, Savulescu argues that “golden opportunities” may arise in the course of an addiction whereby an agent has the opportunity to adopt an objectively more valuable and less risky lifestyle (Savulescu 2018). Golden opportunities are moments where an agent has sufficient control and understanding of their actions so as to be held accountable for their choices. A drug addict, for example, may have the opportunity to voluntarily admit themselves to a rehabilitation clinic at a moment in which their cravings are less intense. This might be said to be a golden opportunity, as it constitutes a brief moment in which the addict has the autonomy to lessen the chance that they will engage in further substance abuse. According to Savulescu, agents who previously did not meet the conditions of responsibility but subsequently fail to capitalize on golden opportunities may well be considered responsible, even if up until that point they did not meet the conditions of responsibility. Responsibility, in this respect, “is dependent on whether [an agent has] a golden opportunity to change” (Brown and Savulescu 2019).

In a more recent paper, Brown and Savulescu have taken a closer look at the conditions under which we might hold someone responsible for their actions (Brown and Savulescu 2019. Specifically, the authors focus on whether an agent has met the epistemic and control conditions required for responsibility. The epistemic condition for responsibility focuses on whether an agent has sufficient awareness of the consequences of their actions, whereas the control condition focuses on whether an agent has an adequate degree of control over their actions (i.e., the actions of an agent are free and not constrained or determined by an outside force). Someone can be held responsible for a pattern of behavior, Brown and Savulescu argue, to the extent that they met the epistemic and control conditions for responsibility in all instances of that behavior. Brown and Savulescu also [End Page 58] consider whether we might hold other agents responsible for a person’s poor state of health. That is, they consider whether a significant other in someone’s life (a partner, parent, child or close friend) might be held at least partly responsible for the health of their loved one. The claim is that our significant others have some responsibility for our health, in addition to the responsibility that we have for our own health.

This should suffice for an overview of proposals endorsing the responsibilisation of healthcare. While we have not discussed the full spectrum of literature on this topic, we believe that the representative sample just presented gives adequate insight into the main arguments used to justify responsibility-sensitive criteria. We now turn to a consideration of some of the ethical issues raised by responsibility-sensitive proposals for healthcare rationing.


There are several challenges that have been identified in the literature to the implementation of responsibility-based policies.4 In the remainder of this paper, we focus on two challenges in particular, namely, that responsibility can diminish as an agent changes, and that blame can come apart from responsibility. Importantly, attributions of responsibility are not enough to justify limiting someone’s access to healthcare. One could argue that what really matters for accountability in healthcare is not past responsibility (whether an agent was responsible at the time of committing an act) but present responsibility (whether an agent can be considered responsible now). That is, we should only hold someone accountable to the extent that they remain culpable for their past actions. This is important, because agents can change in a variety of ways that affect responsibility. In addition, one must think about whether an agent can be blamed for their behavior, independent of whether they are responsible for their actions. It is often assumed that blame and responsibility are two dimensions of the same moral phenomenon; yet a closer consideration reveals that the two moral concepts may come apart. In section 2.1, we consider how agential change impacts upon responsibility; in section 2.2., we consider the implications that a conception of responsibility qua reflective endorsement has for an agent’s accountability for their past behavior; in section 2.3, we discuss the conditions that must be met to hold an agent not just responsible but blameworthy for their actions.

To reiterate, we do not necessarily endorse the views discussed in this section and the next. Rather, our aim is to show that responsibility-based [End Page 59] proposals are more controversial than one may think. We leave to one side the question of whether the views to be presented are in the last analysis correct. We do, nevertheless, believe that these views are sufficiently plausible so as to be taken seriously by proponents of responsibility-sensitive criteria for rationing.

2.1 Responsibility, identity, and agential change

The basic claim at the center of this challenge is that agents change in ways that diminish if not extinguish responsibility for past, imprudent behaviors. An agent may be a very different person now than when they engaged in irresponsible, health-affecting behaviors. As such, the agent may not be accountable for their past actions in the same way that they were at the time when they committed those actions. We will here consider how changes in personal identity impact on ascriptions of responsibility.

Some theorists argue that it is often unclear whether someone is sufficiently psychologically connected to their past self to ground an attribution of responsibility. Rather, an agent in the present may only have a weak psychological connection with their past self, and, as such, may not be the appropriate object of praise or blame. Certain theorists would go so far as to deny that we are in fact dealing with the same person now than the person who engaged in irresponsible behavior in the past (Tomlin 2013).5 To be clear, our aim here is not to endorse a particular conception of the metaphysics of personal identity, in part because there is so much disagreement on this topic. Our aim is, rather, to draw attention to the fact that there are views on personal identity that conflict with the assumptions underpinning responsibility-sensitive criteria for rationing. To illustrate our point, it is instructive to briefly discuss Derek Parfit’s view of personal identity, and the implications that this view has for attributions of responsibility.

Derek Parfit famously argued that personal identity is not a deep fact but rather consists solely in the unique holding of a relation of psychological connectedness and/or psychological continuity. He called this Relation R. According to Parfit, we are just our brains and bodies, and personal identity is based on the fact that someone is psychologically connected and/or continuous with me, and uniquely so (i.e., there is no other who is psychologically connected and/or continuous with me). Relation R just is the psychological connection that exists between me and my past self. Thus, Parfit writes, [End Page 60]

“X today is one and the same person as Y at some past time if and only if X is psychologically continuous with Y [and] this continuity has the right kind of cause”

By the right cause, Parfit was referring to the manner in which mental phenomena such as memories are generated. He noted that memories can be generated in a normal way—such as by having an experience which leaves a memory in the brain—or an atypical way—such as by virtue of the testimony of others which may lead me to think that I have a certain recollection of a past event.

Importantly, Parfit rejected the view that there is one single self that persists across the course of a person’s lifetime. Rather, he conceived of persons as a series of overlapping “selves” that are joined to each other by psychological connections of differing strength. I may have very strong psychological connections with myself yesterday, while only having very weak psychological connections with myself 20 years ago. As Parfit observed,

“Between me now and myself twenty years ago there are many fewer than the number of direct psychological connections that hold over any day in the lives of nearly all adults. For example, while these adults have many memories of experiences that they had in the previous day, I have few memories of experiences that I had twenty years ago”.6

The upshot of all of this is that, while I am the same person as myself twenty years ago, I have only a weak psychological connection with my past self. This is to the extent that my relationship with my past self starts to look more like a relationship with the self of another person.

A Parfitian conception of self can be used to ground the idea that we are only very weakly connected to our past selves. If we have only very weak psychological connections with our past selves, then it is difficult to argue that I am—at present—just as susceptible to praise or blame for my past actions as I was the very day that I committed those actions. Rather, my responsibility for those actions is much weaker now, if it exists at all.7 The discontinuity of the self, in this sense, can be used as an argument in favor of rethinking our attitudes of blame for actions that someone has carried out in the distant past.

An interlocutor may respond by flatly rejecting the assumptions underpinning a Parfitian conception of the self. Yet one need not adopt a Parfitian conception of self to understand the basic point at stake. We need only consider the intuition of some philosophers that the self changes, and [End Page 61] eventually this process of change reaches a point where we are dealing with a very different individual (Tomlin 2013). Our memories are different now to what they were in the past. Our way of seeing the world may be fundamentally different, too. To this extent, we are a different self to the self we were before. Yet this problematizes attributions of responsibility, at least insofar as we take responsibility to be based on a relationship of identity between the person in the past and a person in the present. If psychological connectedness is what counts, then a person in the present may be for all intents and purposes an entirely different self to who they were in the past. One might argue that this clashes with the strong intuition that people are fundamentally the same over time. Indeed, one could argue that this is a more intuitive account of personal identity than the idea that the self changes radically over time. Our aim here, however, is not to say one way or the other whether the self actually changes over time. Rather, we are merely articulating an alternative intuition that motivates the diminished responsibility view.

Indeed, we, like many others, are not sympathetic to the Parfitian view. We do, however, believe that attributions of responsibility based on identity are not as simple as one might think. It is useful here to revisit Brown and Savulescu’s observation:

“attending directly to the diachronic nature of responsibility for habitual, health-related behaviour will require a commitment to some view of personal identity: the extent to which it is continuous over long periods of time, in what situations it becomes discontinuous and the ways in which identity interacts with responsibility”.8

We agree, and would note that more radical theories of personal identity may lead one to completely revise our ascriptions of agential responsibility.

2.2 Reflective endorsement and responsibility for past behavior

In addition to this identity-based critique of the responsibilisation of healthcare policies, one might also bring into question the very model of responsibility on which theorists base their defense of responsibility-based criteria for healthcare rationing. Many philosophers focus on the epistemic and control conditions for responsibility. That is, we must consider whether an agent possessed an adequate degree of control or freedom in performing the action, and also whether the agent’s epistemic or cognitive state was such that she can properly be held accountable for the action and its consequences. Yet some theorists argue that responsibility [End Page 62] transcends the immediate mental states of agents. Rather, responsibility could be said to concern an agent’s reflective endorsement of the relevant mental processes that led one to act in a particular way. This alternative view has received less attention in the literature, yet it has interesting and important implications for how we think about responsibility across time. Specifically, it would seem that someone’s responsibility for an action or set of actions could diminish or even be extinguished as an agent’s fundamental mental anatomy changes.

There are multiple ways of formulating the responsibility as reflective endorsement (RRE) view. A strong formulation of RRE is that responsibility requires a second-order endorsement of a first order desire to do something (regardless of whether this decision is made in the presence of alternative possibilities). What matters, on this account, is that an agent takes responsibility for the decision that is being made or the desire that is being followed. Frankfurt, for example, argues that the key feature of responsibility is that we form a higher-order attitude that endorses and elevates our lower-order desires such that they become part of our agency (Frankfurt 1971). The process of making a decision can be understood as a process in which we form or revise our higher-order attitudes. For instance, let us suppose that a friend wants to go to the movies. If she only has a first-order desire to go, then she would be wanton with respect to this desire, to use Frankfurt’s terminology. It would not matter to her if a stronger desire were to emerge and overpower her desire to go to the movies, sending her to a baseball game instead. She is in a state of indifference vis-à-vis the two options. If she decides to go to the movies, however, then she would not be indifferent because she will have formed a higher order desire that her conduct should be determined by this particular first-order desire. Her decision results in an internal state in which she not only wants to go to the movies, but she also wants her desire to go to the movies to win out in any conflict with present or future desires. It is in this way that a desire goes from being a simple reason for action to becoming a constituent part of our agency.

Yet this strong account of RRE seems too restrictive. We seem to act freely even in situations where we don’t consciously endorse our decisions. A soldier might, for example, rush to save a fellow officer who is wounded on a battlefield, and, in the process, put himself at risk of being wounded or killed. This action may be instinctive, or involve minimal forethought. In the heat of the moment, the soldier might not reflectively endorse their decision to put themselves in mortal peril. Yet despite the instinctive [End Page 63] character of the decision, one would not describe the soldier’s actions as forced or constrained in any way (let’s presume that no one coerced the soldier into attempting to save their fellow officer). In saving a fellow infantryman, the soldier has performed a beneficent action, and it is an action that is plausibly one that we can attribute to their agency. Arguably, responsibility in this case does not require reflexive endorsement of my own desires.9 The Frankfurtian account of RRE, therefore, seems to rule out instances of free choice which we would otherwise want to include in our account of agency.

What we need is an account of RRE that allows for at least some non-conscious decisions to count as free, while still holding that responsibility is fundamentally tied to reflective endorsement. An account of REE that allows for this is John Martin Fischer and Mark Ravizza’s theory of moral responsibility (Fischer 1998). Fischer and Ravizza argue that, in essence, responsibility requires “guidance control”. The key elements of guidance control are that the agent “takes ownership” of the mental mechanism that issues in the relevant behavior, and that the mechanism is “reasons-responsive”. By taking ownership of a mechanism, Fischer and Ravizza are referring to a reflective or non-reflective identification with the mental mechanisms that characterize one’s agency—practical reason, non-reflective habits, attitudes, and so on. At some point in their moral development, a child will begin to identify with and take responsibility for the decisions that issue from their own practical reason. The child, in this sense, “takes ownership” of their decisions. By reasons-responsiveness, Fischer and Ravizza are referring to the relationship between reasons and the decisions that issue from practical reason. Practical reason is reasons-responsive insofar as it responds in a rational and consistent manner to different reasons for action.

To be clear, this account of responsibility does not require that one reflectively endorse every decision. Rather, it only requires that one endorse the mechanisms that characterize one’s agency. It would be sufficient for an agent to have taken ownership of their own practical reason, non-reflective habits and attitudes for the decisions that issue from these things to be considered “free”. On an alternative interpretation, the requirement in question could be even weaker. It may suffice for reflective endorsement that, were an agent’s attention drawn to the mechanism that produced a behavior, they would endorse it (i.e., they need not actually endorse the mechanism, but need only be disposed to endorse it). [End Page 64]

Importantly, Fischer and Ravizza’s theory of RRE suggests that when an agent has changed to the extent that she is less prepared or able (“taking ownership” need not be a binary concept) to recognize the intellectual process that produced the action as her own, she is less responsible. An agent who has little capacity to identify with the intellectual processes that produced their past actions cannot be held responsible for these actions in the same way they might have when they carried these actions out. It also provides plausible answers with regard to real-world cases involving attributions of responsibility. Fischer and Ravizza’s “intellectual process” view connects responsibility with the agent’s attitude to the mental process that produced the action. Someone might regret her choice the morning after, but it is not plausible (in the absence of some major psychological event) that she no longer recognizes the intellectual process that produced the action as her own. In contrast, someone might not even recognize the intellectual process by which she arrived at a decision that took place 40 years ago. The patterns of reasoning that lead one to act in particular ways several decades ago may be utterly unintelligible now. In such a situation, it seems plausible to suggest that responsibility has been diminished if not extinguished.

While we have not provided an extended defense of the intellectual process view of moral responsibility, we can, nevertheless, gesture towards the implications of this view for responsibility-sensitive criteria for distribution. Simply put, RRE implies that an agent can be held responsible for an action only if they can identify with the intellectual process or mechanism that produced that action. People who have engaged in imprudent, health-affecting behavior may or may not be related to their past behavior in a manner relevant for responsibility. An agent, on this view, is less responsible for their actions to the extent that the constitution of their agency has changed. One may have, for example, aged or matured or acquired certain intellectual virtues that they lacked in the past. A 60-year-old who smoked heavily in their youth and then quit in early adulthood might today find the thought of smoking repulsive, and would now never accept an offer to smoke a cigarette. The 60-year-old might now live a very healthy lifestyle—one that is responsive to public health warnings about the health risks of lifestyle choices like smoking. This might be an example of someone who we might hold less accountable for past behavior.

This account of moral responsibility can, of course, be rejected. One might prefer a standard account of responsibility based on epistemic and [End Page 65] control conditions; alternatively, one might accept skeptical arguments against broadly Kantian, reflection-based accounts of agency, endorsing instead a broadly Humean account of agency.10 Two points can be made in reply. First, RRE need not be seen as standing in conflict with all alternative accounts. For instance, RRE is (arguably) compatible with the epistemic and control requirements for moral decision making. What RRE requires is that—in addition to epistemic and control constraints that we might place on attributions of responsibility—we should also consider whether one still reflectively endorses the processes that led to an action or set of actions. In this respect, RRE provides a plausible account of the sorts of conditions that must be met for retrospective responsibility (responsibility for an action in the past), as opposed to prospective responsibility (responsibility for an action that one intends to carry out). Second, it is worth re-iterating that our aim in this paper is not to defend an alternative account of identity or responsibility such as RRE over and against rival accounts of moral responsibility. Rather, our aim is to show that responsibility-sensitive proposals are more controversial than one may think. For this purpose, it suffices to motivate RRE as one plausible account of moral responsibility which opposes the assumptions underlying responsibility-sensitive proposals.

RRE, in summary, presents a problem for responsibility-sensitive criteria for rationing. RRE suggests that people who engaged in imprudent behavior in the past may not be suitably related to those behaviors so as to be held responsible.

2.3 The relationship of responsibility to blame

In this section, we argue that responsibility comes apart from blame, such that people cannot be blamed for imprudent behavior even though they may strictly speaking be responsible for it. We should reconsider, in other words, the way in which we define the terms responsibility and blame.

Some ethicists argue that moral responsibility and blameworthiness can come apart. Tannenbaum, for example, argues that there are cases where someone can be morally responsible for their actions even if the action was not wrong, not blameworthy, and not the result of blameworthy deliberation or bad motivation (Tannenbaum 2018). According to Tannenbaum, responsibility is nothing more than a failure to aid (or to avoid harming) someone, and this may or may not imply blameworthiness on the part of an agent. She argues that three conditions must be met for someone to be both responsible and blameworthy. These are: [End Page 66]

  1. 1. X has an obligation to help and not to harm.

  2. 2. X fails to fulfil their obligations.

  3. 3. X’s failure is due to a blameworthy condition of the agent (e.g., overconfidence, an inappropriate orientation in their will, etc.)

Tannenbaum provides an example of a student who is texting a joke to a friend while driving in a highly populated area and doesn’t see a pedestrian on a zebra crossing. The student in question hits and injures the pedestrian. In this situation, the agent should have known that texting while driving could lead to the kind of accident that eventuated. In light of this culpable oversight, the agent can be held both responsible and blameworthy for the harm done to the pedestrian.

Yet there are also cases where someone can be morally responsible but not blameworthy. Tannenbaum suggests that conditions 1) and 2) are still met in these cases, but not condition 3). Rather, condition 3) should be substituted for an alternative condition (which we will call 3*):

3*) X’s failure to fulfil their obligations is a ‘mere failure’ -- one cannot criticize them for negligence or malice.

By ‘mere failure’, Tannenbaum has in mind an action that involves a failure to discharge one’s duties, but that does not involve culpable wrongdoing on the part of the agent. Tannenbaum provides an example of a man looking after a friend’s child at a restaurant:

“A friend is holding my child as he walks into an unfamiliar restaurant. He notices that the floor is wet and takes care to step where it is dry, so as to avoid slipping and dropping my child. However, he forgets to look up, does not notice a low hanging lamp, and so bangs my child’s head into the lamp as he moves forward.”

According to Tannenbaum, the friend is responsible for the child’s injury, yet he is not blameworthy. As Tannnenbaum suggests, reasonable people sometimes suffer from lapses of judgement, and this is true of the man holding the child. He was trying to avoid injuring himself and the child by slipping, yet in the process he ironically encountered another obstacle, namely, a light.11

We wish to make an analogous argument with respect to people who engage in imprudent health-related behaviors. Granted, public health scenarios are very different to the examples provided by Tannenbaum. Reasonable people, however, can have lapses of judgement with respect to their lifestyle choices, such that they can be held responsible but [End Page 67] not blameworthy for their behavior. Many times people who engage in imprudent behaviors are acting in circumstances such that it is entirely understandable that they would behave in the way that they have. While their behavior may still be deemed imprudent, and they may still be strictly speaking responsible (they may have failed to fulfil some obligation viz-à-viz themselves or society), they should not be blamed for what they have done.

Our claim is based on the view that health is in part determined by social and genetic factors Pampel et al. 2015). So-called lifestyle choices such as smoking or an unhealthy diet are to some extent the product of one’s socio-cultural milieu. That is, these behaviors follow a social gradient, and tend to be impacted by socio-economic factors (locational factors, financial considerations, health literacy, and so on). Reasonable people can adopt unhealthy dietary or smoking habits because of where they live, their level of health literacy, the behaviors of their peers, and even factors such as stress. This is to the extent that it becomes difficult to determine whether and in what way an individual bears personal responsibility for their imprudent behavior.

A person may, furthermore, have a genetic predisposition to become addicted to a particular kind of destructive behavior. Empirical evidence suggests, for example, that people with certain genetic profiles are predisposed to develop alcohol addiction (Mayfield, Harris and Schuckit 2008). We also know that some people are genetically predisposed to develop conditions such as heart disease or liver disease, and this predisposition, combined with an addiction, can render someone far more likely to develop a lifestyle disease than other people. One need only think of a person with a genetic predisposition for both alcoholism and liver failure. Such a person is clearly at a genetic disadvantage when it comes to conditions such as cirrhosis.

Indeed, recent literature suggests that the etiology of disease includes a complex matrix of genetic, social, cultural and economic factors; this fact to some extent lessens the force of any attempt to ascribe personal responsibility to persons with lifestyle-related diseases. Michael Marmot, for example, has written extensively about the relationship between a person’s position in a social hierarchy and the decisions they make that affect their health. Marmot challenges the idea that bad lifestyle choices (such as unhealthy eating, smoking, and lack of exercise) are free and rational decisions that people make. He argues that social factors—in particular, a person’s social class—have a profound influence on whether someone chooses to engage in unhealthy behaviors. Marmot writes: [End Page 68]

“If people choose freely, why does smoking follow a social gradient? It cannot be a coincidence that you are more likely to choose to smoke if you are low status than if you are high… smoking does not follow a social gradient because of ignorance... it is almost as if people know what is in the health warnings, but the degree of attention they pay to these warnings increases as they go up the social scale”

It is not just people’s health, then, that is affected by social factors. Rather, the very choices that lead to bad health are conditioned by a person’s position in society.

If we accept this, then it seems difficult to blame people who engage in irresponsible health-related behaviors. Rather, it seems that these behaviors are heavily influenced by social and genetic factors, and that reasonable persons could be expected to act in similar ways when put in these social situations. This is not to say that these people were not responsible for their behavior. It may very well be that they meet the epistemic and control conditions for moral responsibility. But responsibility does not in and of itself imply blameworthiness.

An interlocutor might reject our attempt to distinguish blameworthiness from responsibility. While in some cases it may be understandable that someone has done the wrong thing—indeed, one’s actions may still be within the pale of reasonableness—this does not excuse the person in question from wrongdoing. A man may assault another man, for example, for having an affair with his wife. The motive behind the assault in this case is something we can comprehend (i.e., it is intelligible). Yet this does not excuse the individual in question from wrongdoing. They have intentionally committed an offence, and they are culpable for the harm caused to the injured party.

We agree that an intelligible motive does not ipso facto exculpate an agent from all wrongdoing. This is not our claim. Our claim is, rather, that in some situations there are circumstantial factors that give us reason to hold someone responsible but not blameworthy for their actions. The examples that we have focused on are concerned with the social-determinants of health, and the role that these determinants play in influencing someone’s health-related behaviors. The impact of social factors seems to be so profound in some cases that, while someone may not go as far as to say that an agent didn’t have a choice, they can’t really be held accountable for their actions. [End Page 69]


This paper has critically evaluated responsibility-sensitive proposals for healthcare rationing. We have argued that attributions of responsibility for illness are difficult to make. In lieu of responsibility-sensitive rationing policies, we here offer a brief defense of an alternative set of policies that are pre-emptive rather than retrospective in character, and that avoid problematic ascriptions of responsibility. We do not pretend to be offering a comprehensive assessment of the ethics and efficacy of these policies, but rather merely wish to gesture towards alternatives to responsibility-sensitive healthcare rationing policies.

Rather than making controversial assumptions about the moral responsibility that people bear for their health-related decisions, we would argue that the State should instead seek to prevent people from engaging in reckless behaviors by heavily taxing products that could be harmful to a person’s health. Specifically, we would argue that the State should heavily tax products such as cigarettes, alcohol, fast food, sugary foods and high-sugar soft-drinks. These products are the sorts of products that will likely produce health complications for people in the future, and, as such, should be of significant concern to healthcare policy makers. These products are already taxed in many countries, though we would advocate increased taxation proportionate to the scarcity of healthcare resources available for treating the health problems that these products cause. Indeed, it may be that the revenue from taxation is directed to improving public healthcare infrastructure.

This approach is a more rational approach to decision-making under uncertainty, as it does not run the risk of wrongly blaming someone for a behavior that they should not be blamed for. Rather, it preempts imprudent behavior by making it more difficult for people to obtain the sorts of products that negatively affect one’s health. This is a less ethically controversial means of addressing the problem of imprudent, health-related lifestyles.

So-called sin taxes have, of course, been subject to heavy criticism.12 Some commentators argue that they unfairly disadvantage the poor, or that they merely lead people to switch to other unhealthy products. Yet these criticisms rest on a misunderstanding of the real impacts of taxes on unhealthy lifestyle products. Importantly, poorer households may in fact get the biggest health benefits and long-term health care savings from taxes on unhealthy products (Backholer et al. 2016). Furthermore, there is evidence to suggest that people switch to healthy options in response [End Page 70] to sin taxes. There is evidence that people switch to water, for example, in response to taxes on sugary drinks (Colchero, Molina and Guerrero-Lópex 2017).

A critic might argue, however, that the introduction of sin taxes would be inconsistent with government policy on the sale and purchase of other risky products. For example, it might be considered unfair to tax tobacco alcohol and sugar, while not taxing goods such as motorcycles, off road vehicles and guns. This may be true, but it does not necessarily mean that we should not tax food products, tobacco and so on. First, all taxes, not just sin taxes, face fairness objections. It would be wrong to assume that this objection is somehow a unique problem for sin taxes. Second, one may agree with the criticism and argue that we should also tax other products that give rise to similar health risks for a population. That is, rather than abandoning taxes on foodstuffs and tobacco, we should broaden the taxation scheme to include other products that surpass a risk threshold equivalent to the risks entailed in unhealthy eating or smoking.

Finally, a critic could argue that sin taxes constitute a responsibility sensitive proposal themselves. That is, sin taxes (as the name suggests) are at least in some cases a means of penalizing those who willingly engage in irresponsible health behaviors. In response, we would say two things. First, sin taxes need not be interpreted as constituting a punishment for people who make imprudent lifestyle choices. Rather, they can—and often are—seen functioning as a deterrent for imprudent behavior/consumption. That is, they deter people from buying products that may increase one’s risk of illness or injury. There is no overt claim about responsibility implicit in this deterrent. Indeed, the term sin tax is an unfortunate one, as it implies that these taxes constitute a punishment for people who engage in sinful behaviors such as unhealthy eating or tobacco consumption. Yet this distorts the intention of policy makers, who in many cases introduce these measures to deter people from making risky lifestyle choices, rather than to punish people who persist with imprudent consumption habits.

Second, it would not be fatal for our argument to concede that there is a responsibility-sensitive dimension to some sin taxes.13 One could argue that sin taxes constitute a punitive measure for reckless behavior, while at the same time avoiding the more problematic implications of responsibility sensitive criteria. Critics of sin taxes argue that these health measures create stigma around the activities that they are trying to discourage (Hirono and Smith 2018). Taxes on unhealthy and dangerous products, however, do not foster questionable assumptions of responsibility in the same way [End Page 71] as responsibility-sensitive ranking systems for healthcare rationing. The social signaling about responsibility is arguably weaker in the case of sin taxes (though changing the name associated with such taxes may also be appropriate lessen the overtones of condemnation).

Authorities may also have reason to maintain a prohibition on the sale of illicit drugs, if only to prevent people from engaging in protracted substance abuse such that could result in health complications in the future. Illicit drugs can have a variety of harmful physical and psychological health impacts, and it would behoove the State to preclude an increase in the number of persons needing treatment for mental health issues or other drug-related health problems.14 Responsibility-sensitive criteria, we submit, would be particularly problematic in the context of drug-related health problems, and, as such, we would advocate an alternative, preemptive approach to addressing the pressures that community substance abuse places on the healthcare system.

Finally, we believe that health authorities would do well to seek to ameliorate the broader impacts of the social determinants of health. One of the leading factors in developing an addiction to smoking or illicit drugs appears to be a person’s location and socioeconomic status.15 In light of this, health authorities should attempt to lessen the impacts of factors such as locational disadvantage on the health of at risk populations. One aspect of this might be improving education programs surrounding substance abuse and how to live a healthy lifestyle. Another factor could be subsidizing activities in disadvantaged areas that would positively influence the lifestyle that people in at risk groups. These activities might range from sporting events that promote regular exercise to youth events that in some way direct adolescents away from recreational drug use.

These proposals may sound naive and may ultimately prove to be of limited efficacy. We do, however, believe that these preemptive approaches to healthcare are less ethically problematic than responsibility-sensitive approaches to healthcare rationing. It would be appropriate for healthcare authorities to explore these preemptive measures first before opting for a more radical, retrospective approach to addressing the impacts of lifestyle related illnesses.16


This paper has considered whether, for the purposes of healthcare rationing, agents should be held responsible for their imprudent actions and behaviors. After providing a brief survey of the literature, we presented [End Page 72] two challenges to the use of responsibility sensitive criteria in healthcare rationing. These were, namely, that responsibility for past behavior can diminish as an agent changes, and that blame can come apart from responsibility. These challenges cast serious doubt on the defensibility of attributions of responsibility for imprudent behavior. We closed by very briefly discussing alternative policies for healthcare rationing that do not involve making contentious judgements about one’s moral responsibility for one’s lifestyle choices.

There are many issues in the responsibility and healthcare debate that this paper has not discussed. One such issue is the question of whether reckless lifestyle choices are even moral choices in the first place, or whether they would be best described as prudential choices (Brown and Savalescu 2019). Depending on one’s answer to this question, the very basis on which the debate about moral responsibility for illness may be rendered void. Yet presuming that health-related behaviors do have a moral dimension, we hope to have shown why we should still think twice about holding people accountable for imprudent past behavior.

Xavier Symons

Dr. Xavier Symons is a Postdoctoral Research Fellow at the Plunkett Centre for Ethics (Australian Catholic University) and an Honorary Research Associate at the Institute for Ethics and Society (University of Notre Dame Australia). His research interests include ethical issues at the beginning and end of life, conscientious objection, ethical issues in aged care, and pandemic ethics. His recently completed PhD thesis focused on the allocation of lifesaving healthcare resources. In 2020, Xavier was awarded a Fulbright Future Postdoctoral Fellowship, and he will be a scholar in residence at Georgetown University’s Kennedy Institute of Ethics from June to December 2021.

Professor Justin Bernstein (PhD) is an Assistant Professor in the Department of Philosophy at Florida Atlantic University. His research focuses on topics in ethics, political philosophy, and bioethics.

Reginald Chua

Reginald Mary Chua is a priest of the Order of Preachers (the Dominicans) and chaplain at the Sydney campus of the University of Notre Dame Australia. He holds a B.A. from Campion College Australia, an M.Phil in philosophy from the University of Sydney, and an M.Theol in theology from the University of Divinity.


1. Admittedly, the concept of a lifestyle disease is problematic, as it is rare for diseases to be caused solely by someone’s lifestyle choices. We can, however, identify certain illnesses, such as Alcohol Related Liver Disease and Chronic Obstructive Pulmonary Disease, that have a strong link to lifestyle, even if there may also be other factors -- such as genetic factors, and non-lifestyle environmental factors -- involved.

2. We are here assuming that the decision-makers are part of a socialized medical system like the National Health Service or the Australian Medicare public health system. In a system where care is free at the point of delivery, and also where resources are scarce, there is at least a prima facie case for placing limits on the care available for people with illnesses related to reckless behavior.

3. Someone might wonder whether alcohol abuse is linked to genetic factors. This may indeed be true in some cases, but for argument’s sake, I will presume that there is a subset of persons who engage in alcohol abuse by virtue of a free decision rather than having their behavior determined by genetics.

5. For a view which rejects personal identity over time due to a rejection of the coherence of the concept of selfhood, see Olson 1998.

6. Parfit 1987, op. cit. note 9, 206.

9. This assessment of the scenario is a controversial one: one might argue, against such an assessment, that the heat of battle substantially mitigates responsibility. In response, however, we would note that it is difficult to make sense of awards for military bravery (or the prosecution of soldiers for misconduct on the battlefield) if we have no concept of free and intentional action during battle. We do not wish to argue the point at length, though, since our purpose is simply to highlight the motivation for an alternative account of RRE.

10. For one recent skeptical discussion of reflection-based accounts, cf. Doris 2015. For a recent defense of a Humean account of agency, cf. Schroeder 2007.

11. One could argue that the distinction between responsible but not blameworthy behavior and blameworthy behavior is related to moral luck. It could be argued that (in the non-blameworthy case) the agent’s behavior only looks imprudent in hindsight; at the time, given evidence available, they were actually behaving prudently. If so this is not really a lapse in judgement; the agent acted in the right way but a statistically unlikely event happened. What we have here, then, is a case of bad moral luck. Yet even if we adopted this alternative view, the person in question would still bear responsibility for banging the child’s head. The implication is the same -- they are responsible but not blameworthy (at least insofar as we hold that bad moral luck relieves someone of blame).

12. Cf. Marmot 2004 op. cit. note 4.

13. Likewise, we are willing to concede that there can be a responsibility insensitive dimension to ranking systems for healthcare resources.

14. On the effects of recreational cannabis use, see, for example, Di Forti 2019.

15. Marmot 2004 op. cit. note 44. For a thorough discussion of the influence of social factors on drug use, see Spooner and Hetherington 2004.

16. Admittedly, responsibility sensitive policies for healthcare rationing are not common. Our article is instead framed as a response to suggestions by several ethicists. Though readers may find it illuminating to read so and so’s discussion of responsibility policies for healthcare rationing.


Albertsen, Andreas. 2016. “Drinking in the last chance saloon: luck egalitarianism, alcohol consumption, and the organ transplant waiting list.” Medicine, Health Care and Philosophy 19 (2): 325–338.
Anonymous. 2013. “Do “sin taxes” work? And are they fair?” The Economist. Sept 5.
Arneson, Richard J. 2004. “Luck egalitarianism interpreted and defended.” Philosophical Topics 32 (1): 1–20.
Backholer, Kathryn, Danja Sarink, Alison Beauchamp, et al. 2016. “The Impact of a Tax on Sugar-Sweetened Beverages According to Socio-Economic Position: a Systematic Review of the Evidence.” Public Health Nutrition 19 (17): 3070–84.
Briggs, Adam. 2019. ““Sin taxes”—the language is wrong, but the evidence is clear.” British Medical Journal 366: 14616.
Brown R.C.H., Julian Savulescu. 2019. “Responsibility in healthcare across time and agents.” Journal of Medical Ethics 45:636–644; 640.
Colchero, M. Arantxa, Mariana Molina, Carlos M Guerrero-López. 2017. “After Mexico implemented a tax, purchases of sugar- sweetened beverages decreased and water Increased: difference by place of residence, household composition, and income level.” Journal of Nutrition 147(8): 1552–1557.
Denier, Yvonne. 2005. “On personal responsibility and the human right to health-care”. Cambridge Quarterly of Healthcare Ethics 14(2): 224–234.
Di Forti, Marta, Diego Quattrone, Tom P Freeman, et al. 2019. “The contribution of cannabis use to variation in the incidence of psychotic disorder across Europe (EU-GEI): a multicentre case-control study.” The Lancet 6(5): 427–436.
Doris, John M. 2015. Talking to Our Selves: Reflection, Ignorance, and Agency. Oxford: Oxford University Press.
Frankfurt, Harry G. 1971. “Freedom of the will and the concept of a person.” Journal of Philosophy. 68 (1): 5–20.
Fischer, John Martin, Mark Ravizza. 1998. Responsibility and Control. Cambridge: Cambridge University Press.
Gheaus Anca. 2017. “Solidarity, justice and unconditional access to healthcare”. Journal of Medical Ethics 43 (3): 177–181.
Hirono, Katherine T., Katherine Smith. 2018. ”Australia’s $40 per pack cigarette tax plans: the need to consider equity.” Tobacco Control 27: 229–233.
Joyner, Michael J., and David O. Warner. 2013. “The syntax of “sin taxes”: putting it together to improve physical, social, and fiscal health.” Mayo Clinic Proceedings 88(6): 536–539.
Littmann, Jasper. 2014. “How high is a high risk? Prioritising high-risk individuals in an influenza pandemic”. Vaccine 32: 7167–7170.
Marmot, Michael. 2004.The Status Syndrome. London: MacMillan.
Mayfield R. Dayne, Robert A. Harris, and Marc A. Schuckit. 2008. “Genetic factors influencing alcohol dependence.” British Journal of Pharmacology 154(3): 275–287.
Nielsen, Lasse. 2013. “Taking health needs seriously: against a luck egalitarian approach to justice in health.” Medicine, Health Care and Philosophy 16(3): 407–416.
Olson, Eric T. 1998. “There Is No Problem of the Self.” Journal of Consciousness Studies 5 (5–6): 645–657.
Pampel, Fred C., Jason D. Boardman, Jonathan Daw, et al. 2015. “Life events, genetic susceptibility, and smoking among adolescents.” Social Science Research 54: 221–232.
Parfit, Derek. 1987. Reasons and Persons. Oxford: Oxford University Press.
Pillutla, Virimchi, Hannah Maslen, H, Julian Savulescu. 2018. “Rationing elective surgery for smokers or obese patients: responsibility or prognosis?” BMC Medical Ethics 19: 28.
Resnik, David B. 2013.“Charging smokers higher health insurance rates: is it ethical?” The Hastings Centre Bioethics Forum Sept 19.
Savulescu, Julian. 2018. “Golden opportunity, reasonable risk and personal responsibility for health.” Journal of Medical Ethics 44(1): 59–61.
Schroeder, Mark. 2007. Slaves of the Passions. Oxford: Oxford University Press.
Segall, Shlomi. 2016. Why Inequality Matters: Luck Egalitarianism, its Meaning and Value. London: Cambridge University Press.
Spooner, Catherine., Kate Hetherington. 20014. Social Determinants of Drug Use. Sydney: National Drug and Alcohol Research Centre.
Tannenbaum, Julie. 2018. “Moral responsibility without wrongdoing or blame.” Oxford Studies in Normative Ethics (vol.8). London: Oxford University Press.
Tomlin, Patrick. 2013. “Choices, chance and change: luck egalitarianism over time.” Ethical Theory and Moral Practice 16(2): 393–407.

Additional Information

Print ISSN
Launched on MUSE
Open Access
Back To Top

This website uses cookies to ensure you get the best experience on our website. Without cookies your experience may not be seamless.