Article Text

Download PDFPDF

What ‘Just Culture’ doesn’t understand about just punishment
  1. Samuel Reis-Dennis
  1. Correspondence to Dr Samuel Reis-Dennis, Berman Institute of Bioethics, Johns Hopkins University, Baltimore, MD 21205, USA; sreisde1{at}


Recent years have seen the rise of ‘Just Culture’ as an ideal in the patient safety movement, with numerous hospitals and professional organisations adopting a Just Culture response to incidents ranging from non-culpable human error to intentional misconduct. This paper argues that there is a deep problem with the Just Culture model, resulting from its impoverished understanding of the value of punitive, fundamentally backward-looking, practices of holding people accountable. I show that the kind of ‘accountability’ and ‘punishment’ contemporary Just Culture advocates endorse disrespects both patients and providers. I claim, first, that punishment is good because it respects participants in the healthcare system by restoring an equilibrium of social and moral status that wrongdoing disturbs, and, second, that it only does so when it communicates a backward-looking message of resentful blame.

  • medical error
  • ethics
  • philosophical ethics
  • persons
  • moral psychology

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.


Recent years have seen the rise of ‘Just Culture’ as an ideal in the patient safety movement, with numerous hospitals and professional organisations adopting a Just Culture response to incidents ranging from non-culpable human error to intentional misconduct.i

The Just Culture ambition emerges from a desire to find a middle ground between two extreme models for responding to adverse events. On the one hand, we might imagine a caricaturised version of a ‘blame and shame culture’, in which the medical provider who most directly caused the adverse outcome would be sanctioned, shamed and shunned. Adherents to this approach, perhaps fuelled by angst about the limits of human agential control, would show little interest in how systemic factors may have made the mishap more likely, or positioned the blamed agent to fail. It is easy to see how this sort of myopic scapegoating would reflect and reinforce an inaccurate, or at least incomplete, understanding of how errors occur, and how such a culture would be, not only cruel, but inimical to patient safety efforts.

On the other hand, we might imagine a version (again, an extreme one) of a ‘no-blame’ model. Adherents to this approach would not be concerned with accountability. Instead, they would favour total openness about errors, believing that such a policy would best promote the kind of learning required to make the hospital as safe as it could be. Even granting this assumption, however, we might have reservations about a system that made no attempt to hold malicious and reckless actors responsible for their behaviour.

Just Culture aims to offer a mean between these two extreme caricatures. This approach grows out of, and aims to be sensitive to, the systems-focused work of James Reason, Sidney Decker and others, who have argued that most medical errors are attributable not to individual agents but rather to imperfect systems. It aims to balance this insight, and the corollary thesis that it would be wrong to blame and punish practitioners for adverse events that are not products of their agency, with the intuitive need to hold at least some individuals personally accountable for their risky or malicious actions.

I agree that we should hold individuals accountable at least for reckless and malicious behaviour, and that we should not punish individual practitioners who are merely unlucky to be causally involved in catastrophic, but purely systemic, breakdowns. Nevertheless, I will argue that there is a deep problem in the Just Culture movement that stems from an impoverished understanding of the value of punitive, fundamentally backward-looking, practices of holding people accountable. In what follows, I show that the kind of ‘accountability’ and ‘punishment’ contemporary Just Culture advocates endorse disrespects both patients and providers. I claim, first, that punishment is good because it respects participants in the healthcare system by restoring an equilibrium of social and moral status that wrongdoing disturbs, and, second, that it only does so when it communicates a backward-looking message of blame.

Punishment and accountability in a ‘Just Culture’

The Just Culture movement values a form of accountability, at least for certain kinds of behaviour. But what, exactly, does such accountability come to? Two threads have emerged in the Just Culture literature. One proposal, developed by Virginia Sharpe and endorsed by Sidney Decker,1 recommends ‘forward-looking accountability’. The other, defended by David Marx and others,2 identifies accountability with ‘punitive’ sanctions that are justified entirely by their deterring effects. I will consider these suggestions in turn.

‘Forward-looking accountability’

‘Forward-looking accountability’ is explicitly disconnected from practices of praising and blaming, focusing instead on goal-setting and role responsibility (Sharpe, 2004, p13). When an adverse event occurs, members of an institution that embraces forward-looking accountability will not ‘hold people accountable’ by determining who is blameworthy or deserves to be punished. Instead, they will ask who should be ‘held accountable’ for taking on certain responsibilities in the future. Sharpe3 explains that on such an understanding, ‘Responsibility takes the form of preventative steps to design for safety, to improve on poor system design, to provide information about potential problems, to investigate causes, and to create an environment where it is safe to discuss and analyze error.’ Forward-looking accountability is entirely non-punitive and has nothing to do with blame; it aims only to make the healthcare system safer and more effective. ‘The point of forward-looking responsibility’, Sharpe writes, ‘is to specify the obligations entailed in achieving a safer health care environment’. Decker, in Just Culture (2012), endorses Sharpe’s understanding, explicitly rejecting a more traditional understanding of accountability that involves punishment and blame: ‘For you and your organization, such backward-looking accountability is pretty useless beyond getting someone’s hot breath out of your neck” (Decker, p. 102).1


The other dominant form of accountability in the Just Culture movement involves sanctions in response to malicious and reckless behaviour. This sort of Just Culture has a more punitive feel, and indeed its proponents refer to its sanctions as punishments. But I will argue that the kind of ‘punishment’ these proponents of deterrence have in mind is only superficially related to the sort of punishment we are intuitively familiar with. This is because, like ‘forward-looking accountability’, deprivations that aim solely at deterrence are not necessarily tied to condemnation or restoration. According to the deterrence-only model, ‘punitive’ sanctions are justifiable only insofar as they promote safety. Marx, for example, explains that we ought to think of ‘disciplinary action as harm dispensed by an authority to deter future undesirable conduct’ (Marx, p5), and that ‘in the context of event investigation, the important question is whether human factors learning from events outweighs the deterrent effect of punishment against … employees’ (Marx, p25) Most bluntly, he directly addresses healthcare executives, asking: ‘Is your current disciplinary policy supportive of or detrimental to your system safety efforts? Your job requires that you balance the interests of communication with those of deterrence.’ Similarly, the Global Aviation Safety Network’s ‘Roadmap to a Just Culture’ (2004)4 asks: ‘Is it more worthwhile to reduce accidents by learning from incidents (from incidents being reported openly and communicated back to the staff) or by punishing people for making mistakes to stop them from making mistakes in the future?’ In the same vein, Robert Wachter and Peter Pronovost5 write that ‘a unidimensional focus on creating a blame-free culture carries its own safety risks’ (Pronovost, p1401), and that ‘the main reason to find the right balance between “no blame” and individual accountability is that doing so will save lives’ (Pronovost, 1405).

Punishment and respect

I will now argue that these Just Culture approaches, which license only ‘forward-looking’ responses of one kind or another to bad behaviour, are disrespectful to all participants in the healthcare system. Drawing from the work of Herbert Morris, Joel Feinberg and Jeffrie Murphy,ii I will offer a brief but, I hope, suggestive sketch of an approach to accountability that essentially communicates resentful blame.iii With such an outline in mind, I will compare this system to a Just Culture to show what we lose by eschewing backward-looking condemnation in favour of merely forward-looking responses to adverse events that aim solely to promote safety.

A good healthcare system is designed for the mutual benefit of all parties. Patients receive care; practitioners make a livelihood providing it. But the system will not work well—that is, the parties will not be able to reap its benefits—unless everyone accepts certain burdens in the form of rules that constrain their conduct. Physicians, for example, take the Hippocratic Oath and agree to adhere to a range of professional and ethical standards. Nurses also accept rules of conduct and may take the Nightingale Pledge. Among other obligations, patients agree to contribute to the compensation of their caretakers. And everyone agrees to maintain a baseline standard of respect, decency and civility in their interpersonal conduct.

All participants, then, at least in theory, have an investment in certain rules and principles, as adherence to them allows the system to function to their benefit. Sometimes, however, actors within the system knowingly violate these rules and principles without a good excuse.iv We may imagine, for example, a doctor who leaves a surgery she’s supervising at a crucial time in order to take a personal phone call, or a nurse who doesn’t wash his hands in certain examination rooms because he doesn’t like the smell of the soap.v

A central premise of the punitive scheme I will outline is that when such violations occur the wrongdoer creates a social and moral imbalance. Three features of wrongdoing conspire to give rise to this inequity: first, in knowingly breaking the rules, the offender gains an unfair advantage over others who do not break them. Others might wish to allow their inclinations and desires to express themselves unchecked, but they do not do so. Instead, they reign them in, dutifully doing their part to promote the system’s functioning. The wrongdoer, however, enjoys the system’s benefits without having to shoulder her fair share of the burdens that make those benefits possible. Second, in breaking the rules, wrongdoers express an attitude of contempt for the rules themselves and for those who follow them. Such violations communicate a message that one is above the law, that the rules apply to others but not to oneself. In other words, the violation is a kind of insult to one’s upright peers, and especially to those whom one puts at risk. In expressing such contempt, the rule breaker disrespects others by implying that they are not her moral equals. Third, when such violations lead to harm, the social imbalance is exacerbated. Injury is added to insult, and the violation ‘puts the victim down’ in a more practical sense: the wrongdoer has unjustly deprived him of valued goods.

A just system of punishment aims to restore the equilibrium wrongdoing disturbs. It endeavours to fairly redistribute benefits and burdens, assuring that participants feel their investment in the organisation is not naive, and that they are not being taken advantage of. These levelling ambitions are expressions of moralised resentment, a desire to bring offenders down (and/or raise up the diminished) to an even plane. When moralised in this way, these desires are not only morally permissible, but laudable. An unfair distribution of benefits and burdens is fundamentally at odds with a conception of community members as moral equals. In punishing wrongdoers, we insist that everyone deserves the respect of fair treatment.

Merely confiscating any ill-gotten gains that arose as a result of wrongdoing (eg, by fining the doctor who fielded a phone call during surgery only the amount of money that the hospital had paid for work not in fact done) is insufficient to level the playing field. Returning the wrongdoer to her starting position would not be enough to restore moral and social balance because it would allow those who had insulted and disrespected others to remain above their law-abiding peers. After all, they would enjoy equal access to the system’s benefits without having shared equally in its burdens of self-restraint. As Morris notes, wrongdoers, in a sense, owe a ‘debt to society’, and punishment is a way of extracting payment (Morris, p. 478).6

This extraction, imprecise though it may sometimes be, proceeds by depriving the offender of something he or she values, and the seriousness of the deprivation roughly corresponds to the severity of the social and moral imbalance generated by the prohibited behaviour. (This explains why the harshness of punishment is at least partially responsive to the consequences of the transgression, even when those consequences are not entirely under the offender’s control.) Committing to a system of punishment, then, promotes respect for persons as moral equals in two ways. First, rules of punishment communicate the message that unfairness is unacceptable. Second, the execution of a punishment attempts to restore equality by (at least symbolically) erasing the unfairness that necessitated it.

I should be clear that while such a system of punishment does imply that wrongdoers deserve to feel the pain of sanctions, I am not defending the fanciful metaphysical thesis that all evil merits compensatory suffering. The system of punishment I have in mind is deeply human. It is constructed and maintained by us, and it is embedded in our culture and practice. It aims to rectify social rather than cosmic imbalances. My claim is that extracting a debt from wrongdoers is a way of restoring all parties to equal moral footing.

Nor am I advocating the imposition of draconian penalties for bad behaviour. After all, unnecessarily harsh or cruel punishments create moral and social imbalances of their own. Indeed, a backward-looking model like the one I’ve sketched is well-positioned to explain why punishment must, at least roughly, ‘fit the crime’.vi On an entirely deterrence-based justification of punishment, there is no necessary connection between blameworthiness and sanction: if brutal treatment of certain practitioners turned out to best promote patient safety goals, a pure deterrence theorist would be compelled to endorse it.

Finally, and significantly, the system of punishment I have in mind respects wrongdoers. This is because it treats them as persons who can, through free expressions of their values, become proper objects of the distinctively human reaction of resentful blame. With its logic of disrespect and apology, debt and restitution, punishment is a response that only makes sense when directed towards responsible agents who can stand in what P.F. Strawson7 termed ‘participant’ relationships with others. This is why Morris wrote that, ‘when what we do is met with resentment, we are indirectly paid something of a compliment’ (Morris, p. 487)6: to punish someone is to recognise his humanity. Indeed, Morris plausibly argues that morally responsible offenders have a right to be punished that derives from a more fundamental right to be treated as persons (Morris, p. 490).6 Unpleasant as it may be, one might reasonably prefer, and even demand, to be punished rather than managed, as one would manage a child, an animal or a machine.

The disrespect of ‘Just Culture’

Neither a system that exclusively featured ‘forward looking accountability’ nor an organisation that understood the deprivations and penalties it meted out as valuable purely as deterrents would respect its participants as persons. Let us first consider a system of ‘forward-looking accountability’, recalling that this notion of accountability has nothing to do with blame and resentment. Rather than requiring restoration of the moral order in the name of fairness and equality, it calls for participants to ‘take responsibility’ for playing their respective roles in the way that best promotes safety.

As a result, it leaves the moral field tilted towards wrongdoers and suggests that the dignity and self-respect of the dutiful are not significant concerns. This message is an affront to all participants in the system, but is especially insulting to those harmed by the offending conduct. A system of forward-looking accountability cannot even acknowledge when feelings of resentment and reprobation are reasonable. It certainly has no mechanism for attempting to restore participants' moral and social status as equals, on a par with their peers and their assailants. Victims would be justified in withdrawing from such a system: they may lose faith in it, adopt a purely cynical attitude towards it or attempt to distance themselves from it to the extent they are able. But choosing to remain deeply invested in an organisation that denied one acknowledgement and restitution would be to compromise one’s dignity and self-respect.

Forward-looking accountability does no better for wrongdoers. In distancing itself from ‘useless’ backward-looking reactive attitudes such as resentment, it treats practitioners as mere pieces to be managed, cogs in a (safe) machine. In such a paradigm, bad behaviour is not understood as a communicative, expressive act of a person that might call for condemnation and blame, but rather as an unfortunate outcome to process and learn from. If, in the wake of such bad behaviour, it became clear that a different distribution of role responsibilities would best promote safety, the reckless practitioner would be efficiently shuttled off to wherever he would be most useful (or least harmful). He would have no institutional outlet for his natural feelings of guilt, nor would he be able to pay his ‘debt to society’, apologise or be forgiven, for these steps belong to a logic of resentment, blame and punishment.

A system of penalties and deprivations justified purely as deterrents would be similarly disrespectful. Again, in divorcing itself from the backward-looking aspects of punishment that express resentment, it could not meaningfully stand for fairness and equality. If it turned out, for example, that the proposed deprivations such a system might visit on reckless practitioners did not make the hospital as safe as a no-penalty system, then the penalties would be abandoned without regret. And even if data showed that deterrents were necessary to promote safety, the deprivations involved in a Just Culture-style deterrent system would have nothing to do with levelling the playing field in an effort to eliminate moral inequality. In fact, the penalties would need not be communicative or restorative at all. It may turn out that the most effective scheme of deterrence would not treat wrongdoers as persons, but instead involve Pavlovian conditioning: the kind of management one might employ to get an animal to change its behaviour.

To summarise, punishment can only stand for equality and respect if it communicates resentful blame. This feeling, only appropriately felt and expressed towards agents, aims to level the moral playing field and restore a fair balance of benefits and burdens so that everyone may participate in the system without compromising his or her dignity. A scheme of deprivations justified entirely by their power to maximise patient safety has little to do with these goals and values. In fact, it works against them by treating people as mere means to (perhaps noble) ends.


At this point, one may be tempted to claim that the results of the punitive scheme I have outlined here are themselves merely good outcomes to be outweighed by the increased safety a Just Culture would provide. But to understand equality and respect in this way would be a mistake, for at least two related reasons. First, we cannot punish and blame wrongdoers in order to produce the consequence that we achieve the value of respect, for to respect someone is to engage in a genuinely backward-looking process of blame and punishment, restitution and restoration. Second, respect cannot simply be assimilated into a totalising calculus of welfare: the fact that treating people disrespectfully would lead to better consequences (eg, increased happiness, safer hospitals) would not erase their claims to be treated respectfully.

It seems conceivable, indeed likely, that the system that most efficiently promoted safety would be one that failed to treat participants as persons. This is a possibility we should be prepared to face. To what extent medicine should embrace the sort of punishment I have sketched is a question of great moral and practical significance, and I cannot answer it here. It may be that our institutions are so hopelessly corrupt, or that we are so incorrigibly biased, that we could never hope to enact a just punitive system that facilitated anything approaching a fair distribution of benefits and burdens. My own view tends not to be so pessimistic, but evaluating the practical feasibility of a just punishment scheme would require interdisciplinary research and dialogue I do not pretend to have undertaken in this paper. What I do hope to have shown, however, is that choosing to banish resentful punishment in an effort to promote safety would come with significant moral costs.


The author is grateful to Gail Geller, Travis Rieder, and Vida Yao for their helpful comments and suggestions. The author has also benefited from discussion with Thomas Hill, and an audience at Johns Hopkins University’s Berman Institute of Bioethics. The authors also thanks Matthew Dennis, Elizabeth Reis, and Pamela Reis for their feedback on an earlier draft.



  • i Among the most notable of these organisations is the American Nurses Association,8 which, in a 2010 position paper, endorsed ‘the collaboration of state boards of nursing, professional nursing associations, hospital associations, patient safety centers and individual health care organizations in developing regional and state-wide Just Culture initiatives.’

  • ii My discussion of punishment, particularly my comments about punishment’s ability to restore a just balance of benefits and burdens, and about the right to be punished, is heavily influenced by the work of Herbert Morris. See his ‘Persons and Punishment’ (1968)6 for more on these themes. My analysis of punishment’s expressive force owes much to Joel Feinberg’s ‘The Expressive Function of Punishment’ (1965).9 Finally, my understanding of the deep connection between resentment and self-respect has been shaped by the work of Jeffrie Murphy, particularly his ‘Forgiveness and Resentment’ (1982).10

  • iii By ‘resentment’, I mean a moral feeling that is essentially backward-looking, one that arises in response to perceived disrespect or unfairness and aims to restore a moral and social equilibrium wrongdoing disturbs.

  • iv A just punishment structure would leave room for excuses, and an appeal to systemic factors beyond one’s control might rightfully succeed in exempting one from punishment. I do not address this sort of appeal and its limitations in this paper; here, I confine my discussion to clear cases of recklessness and malice.

  • v For simplicity, I’ve imagined examples that reflect obvious wrongdoing, but my analysis will apply to a wide range of bad behaviours.

  • vi My thoughts here are inspired by Morris’s (see especially pages 483-4).6

  • Funding The authors have not declared a specific grant for this research from any funding agency in the public, commercial or not-for-profit sectors.

  • Competing interests None declared.

  • Patient consent Not required.

  • Provenance and peer review Not commissioned; externally peer reviewed.

Other content recommended for you