1	Retributive and Restorative Just Cultures

There are basically two ways to approach a just culture: inside your organization as well as outside of it. One is based on retribution, the other on restoration. When faced with actions that led to a (potentially) bad outcome

1.  Retribution imposes a deserved and proportional punishment.

2.  Restoration repairs the trust and relationships that were damaged.

Retributive and restorative processes ask very different questions in the wake of an incident. Consider this when you are trying to build a just culture of trust and accountability in your organization.

•  Retribution

•  Which rule has been broken?

•  Who did it?

•  How bad was the infraction, and so what does the person deserve?

•  Which manager, department, or authority gets to decide?

•  Restoration

•  Who has been (potentially) hurt?

•  What are their needs?

•  Whose obligation is it to meet those needs?

•  What role does the community play in learning from the event?

Let’s first look at retributive and restorative approaches to justice separately, to see how each tries to generate accountability, trust, and learning. Then we’ll compare and contrast the two. You will find that the two approaches don’t have to be mutually exclusive. In your own organization, you might well be able to create a mix of responses that imports the best of both worlds.

RETRIBUTIVE JUST CULTURE

Retributive justice is as old as humanity itself. It is easily recognizable in varieties of ancient law (e.g., the law of “measure for a measure”). Retribution considers punishment, if proportionate, to be a just response to a sanctionable action and to be an appropriate deterrent. Retributive justice wants the offender to pay something back, to forfeit something in return for what he or she did. Retribution is not vengeance. Retribution is not supposed to be directed at the person, but at the actions (the wrongs). It is supposed to come with inherent limits (i.e., there is only so much punishment that you can exact), it involves no pleasure at the suffering or pain of others, and it is constrained and governed by procedural standards.

SHADES OF RETRIBUTION

A popularized way to think about a just culture is based on shades of retribution. This type of just culture, introduced in hospitals, airlines, oil companies, and other organizations over the past years, understands that it is foolish to expect fallible people to be perfect. These organizations find it important to learn from mistakes, and their approach to just culture wants to hold people accountable not necessarily for the outcomes they create, but for the choices they (supposedly) make while doing their work. When things go wrong, the consequences should depend on the choices that people made in the lead-up. Such a just culture often distinguishes among a minimum of three types of actions:

•  An honest mistake is an inadvertent lapse, slip, or mistake. It was unintended and can happen to anyone in those circumstances.

•  At-risk behavior is a choice that increases a risk that is not recognized, or mistakenly believed to be justified.

•  Negligence or recklessness is a choice to consciously disregard or take a substantial and unjustified risk.

Those who have adopted such a “just culture” believe that different choices deserve different consequences—different shades of retribution. Honest mistakes require compassion and an investigation into the conditions that triggered them. At-risk behavior calls for coaching and warnings. Recklessness must lead to disciplinary action, including suspension, dismissal, or referral to other authorities. The idea is that employees who feel that they will be treated fairly—that their honest mistakes will not be punished—are more inclined to report mishaps and failures. But will they?

“You have nothing to fear if you have done nothing wrong.”

What if a prosecutor would say this to respond to concerns from the community of practitioners? It might be said to assuage concerns from the sector that human errors—normal, honest mistakes—are being converted into criminal behavior by the prosecutor’s office. This is not just hypothetical, of course. Practitioners have been fined or charged for rule infractions that were part and parcel of getting the job done. They and their colleagues may feel some anxiety as a result. Can they supply incident data in good faith, or are the data going to be used against them? Are there enough protections against the prying of a prosecutorial office? Don’t worry, a prosecutor might say in reply. Trust me. There is nothing to fear if you have done nothing wrong. I can judge right from wrong. I know a willful violation, or negligence, or a destructive act when I see it.

But does he? Does anybody?

Retributive just cultures draw a line somewhere between acceptable and unacceptable behavior. A willful violation is not acceptable. An honest mistake is. Risk-taking behavior probably isn’t. Negligence and recklessness are certainly not. And if what you have done is acceptable—if you have done nothing wrong—you have nothing to fear. A just culture definition in use in air traffic control, for example, promises and warns simultaneously that “Front-line operators or others are not punished for actions, omissions or decisions taken by them that are commensurate with their experience and training, but where gross negligence, willful violations, and destructive acts are not tolerated.”¹⁸

The idea of a line makes intuitive sense. If just cultures are to protect people against being sanctioned for honest mistakes (when they’ve done nothing wrong), then some space must be reserved for mistakes that are not “honest” (in case they have done something wrong). Consequently, all proposals for a just culture emphasize the establishment of, and consensus around, some kind of line between legitimate and illegitimate behavior: “In a just culture, staff can differentiate between acceptable and unacceptable acts.”¹⁹ There is not a single proposal for just cultures—indeed, not a single appeal to the need to learn from failure in aviation—that does not build in some kind of escape clause into the realm of essentially negligent, unwanted, illegitimate behavior. An environment of impunity, the argument continues, would neither move people to act prudently nor compel them to report errors or deviations. After all, if there is no line, then “anything goes.” So why report anything? This is not good for people’s morale, for the credibility of management, or for learning from mistakes and near misses.

So calls for some kind of border that separates tolerable from culpable behavior make intuitive sense. And ideas on just culture often center on its embrace and clarity: “A ‘no-blame’ culture is neither feasible nor desirable. Most people desire some level of accountability when a mishap occurs. In a Just Culture environment the culpability line is more clearly drawn.”¹³ Another argument for the line is that the public must be protected against intentional misbehavior or criminal acts, and that the application of justice is a prime vehicle for such protection.

A recent directive from the European Union (2003/42/EC) governs occurrence reporting in civil aviation. This directive has a qualification: a state must not institute legal proceedings against those who send in incident reports, apart from cases of gross negligence. But who decides what counts as “gross negligence”? The same state, of course. Via its prosecutors, investigating magistrates, and judges.

The directive, as does much guidance on just culture today, seems to assume that cases of “gross negligence” jump out by their very nature, that “willful violations” represent an obvious category, distinct from violations that are somehow not “willful.” It assumes that a manager, a prosecutor, a judge, or any other authority can recognize—objectively, unarguably—willful violations, negligence, or destructive acts.

If we want to draw a line, we have to be clear about what falls on either side of it. Otherwise there is no point in a line—then the distinction between acceptable and unacceptable behavior would be one big blur. Willful violations, say many people, clearly fall on the “unacceptable” side of the line. Negligence does too. But what is negligence then? To begin with, look at this definition:

Negligence is conduct that falls below the standard required as normal in the community. It applies to a person who fails to use the reasonable level of skill expected of a person engaged in that particular activity, whether by omitting to do something that a prudent and reasonable person would do in the circumstances or by doing something that no prudent or reasonable person would have done in the circumstances. To raise a question of negligence, there needs to be a duty of care on the person, and harm must be caused by the negligent action. In other words, where there is a duty to exercise care, reasonable care must be taken to avoid acts or omissions which can reasonably be foreseen to be likely to cause harm to persons or property. If, as a result of a failure to act in this reasonably skillful way, harm/injury/damage is caused to a person or property, the person whose action caused the harm is negligent.¹³

A few concerns come to mind. First, the definition is long. Second, it is actually not really a definition. It does not capture the essential or finite properties of “negligence.” It does not allow you to grab “negligent behavior” and put it on the unacceptable side of the line. Instead, it presents an array of questions and judgments that you need to make. Rather than this solving the problem of what is “negligence” for you, you now have to solve a larger number of perhaps equally intractable problems instead:

•  What is the “normal standard”?

•  How far is “below”?

•  What is “reasonably skillful”?

•  What is “reasonable care”?

•  What is “prudent”?

•  Was harm indeed “caused by the negligent action”?

So instead of clarifying which operational behavior is “negligent,” this shows just how complex the issue is. And how much of a judgment call it is. In fact, there is an amazing array of judgment calls to be made. Just see if you, for your own work, can (objectively, unarguably) define concepts such as “normal in the community,” “a reasonable level of skill,” “a prudent person,” or that you could achieve “a foresight that harm may likely result.” What, really, is normal (objectively, unarguably)? Or prudent, or reasonable (objectively, unarguably)? Don’t we all want to improve safety precisely because the activity we are engaged in can result in harm? And, of course, research showed us a long time ago that once we know the outcome, we overestimate the amount of foresight we, or someone else, could and should have had.²⁰

Just responses to bad events are not a matter of matching the inherent properties of undesirable behavior with appropriate pigeonholing and a fitting punitive level. It involves the hard work of deciding what story to tell and whether to see something as reckless, as at-risk, or as erroneous. Merely supplying the categories leaves this issue unresolved. It boils down to fairly empty guidance on how to create a just culture. It also creates an ethical issue: assigning an act to a category will forever be a judgment.

It is not that making such judgments is impossible. In fact, we do this often. It is, however, important to remember that these are indeed judgments. They are not objective and not indisputable. To think that there comes a clear, uncontested point at which everybody says, “Yes, now the line has been crossed, this is negligence,” is an illusion. What is “normal” versus “negligence” in a community, or “a reasonable level of skill,” versus “recklessness” is infinitely negotiable. You can never really close the debate on this. As a result, there really is no line. There only are people who draw it.

What matters is not whether some acts are so essentially negligent as to warrant serious consequences. What matters is which authorities we in society (or you in your organization) rely on to decide whether acts should be seen as negligent or not. Who draws your line?

Assigning acts to categories becomes a matter of power. Who has the power to tell the story, to say that behavior is one thing and not the other? And who has the power to decide on the response? It is a power that can finesse and fudge a whole range of organizational, emotional, and personal issues. A conclusion of wrongdoing could, for instance, be underwritten by a hospital’s risk manager’s greatest fears (of liability, loss of reputation or political clout) or by how a manager is held accountable in turn for evidence of trouble in the managed unit.

Suppose nurses take to scanning the barcode label that one of their colleagues pasted on the wall behind the patient because it actually reads well, and is always easy to find as opposed to others. This may have become all but normal practice—everybody does it because everybody always has the next patient, and the next, and the medication barcode scanners are of such poor quality that they can’t read anything except flat and high-contrast labels (indeed, labels pasted on a wall).

Managers may want to call such behavior “at-risk” and mete out supposedly appropriate countermeasures. But nurses may no longer see their behavior as at-risk, if they ever did. In fact, it may be a sure way to get a good scan, and not doing that could create more risks. And, of course, barcode scanning is not their main job—taking care of patients is. This means that nurses may see any punitive responses to scanning a label stuck on a wall as pretty unjust. After all, such responses may show that the manager is not aware of the unrelenting pressures and ebbing and flowing demands of nursing work, or of the shortcomings of putatively labor-saving and safety-enhancing technology. Justice is a matter of perception.

But managers are under different pressures. Managers appropriate the power to call something “at-risk” not because of their putative insight into the risks and realities of nursing work, but because they can and because they have to relative to the pressures and expectations of their own positions in the hospital. From a manager’s point of view, operational behaviors that bypass instructions or protocol, for example, could end up eroding productivity and reputation and eventually impair the financial performance of his or her part of the organization. Or, for that matter, having to make structural or equipment changes (e.g., procuring new or better barcode scanners) involves sacrifices that are much larger than reminding people to be more careful and follow the rules.

What if we involve domain expertise in drawing the line? Or get the experts to do it altogether? Does that help? There is no research that suggests that domain experts automatically prevent the biases of hindsight slipping into their judgments of past performance. Hindsight is too pervasive a bias. And domain experts may have additional biases that work against their ability to judge the quality of another expert’s performance fairly. There is, for example, the issue of psychological defense: if experts were to affirm that the potential for failure is baked into their activity and not unique to the practitioner who happened to inherit that potential, then this makes them vulnerable too. Sometimes it can be more comforting to think that the errors made by a fellow practitioner would not happen to you, that they really are unique to that other person.

DIFFICULTIES AND FAIRNESS IN RETRIBUTION

Could retributive just culture approaches undermine honesty and reporting and lead to arbitrary judgments after all? To date, no research evidence has been offered that organizations that have implemented a retributive just culture program are better at learning (as indicated by a greater willingness to report safety events or incidents). That said, there is only limited evidence that retributive just culture approaches inhibit openness and learning.

1.  Systematic evidence comes from a 2006 survey study (n = 1984) into perceptions of just culture across disciplines in healthcare. In exploring the existence of retributive responses, it asked questions such as, “Are employees held accountable for their actions?” “Is there blame or favoritism?” “Does the organization recognize honest mistakes?” Accountability was perceived significantly differently. Physicians tended to have the most positive view of their culture as “just,” followed by management—over groups with less power in the medical competence hierarchy (e.g., nursing and nonclinical staff), who considered the same culture significantly less “just.”²¹

2.  Anecdotal evidence suggests that a program that “blames workers themselves for job injuries and illnesses, drives reporting underground. If injuries aren’t reported, the hazards … go unidentified and unaddressed.”²²

3.  Evidence by analogy. In a study of workplace complaint systems, formal options that led to retribution and top-down responses affected people’s willingness to disclose. If “the organizational culture is too hierarchical and oriented toward punishment, [this] may inhibit willingness to act or come forward.”²³

Retributive just culture approaches are organized around finding out who did something wrong and how to deal with him or her, rather than asking what was responsible for things going wrong and how to fix that. If the response focuses on the individual, the who, then underlying conditions that gave rise to the problem in the first place can be missed and left unaddressed. In one case, a nurse was blamed for the death of a 16-year-old patient, dismissed from her job, and then criminally prosecuted.²⁴ A later report, however, found how systemic issues at the hospital set the stage for the nurse’s fatal drug error.²⁵ Asking who was responsible downplayed the role played by other factors—local pressures and constraints; scheduling issues; fatigue; personnel shortages; systematic gaps in support; barcode scanner problems; issues of training, hierarchy, and information flow; and officially condoned workarounds. By blaming the nurse, nothing of value was learned from the patient’s death.

Let’s look at two important aspects of retributive justice here, and see how these play out in a retributive just culture.

1.  Substantive justice prescribes how regulations and rules that people are supposed to follow in their jobs must themselves be fair and legitimate.

2.  Procedural justice sets down legitimate processes for determining rule breaches, offers protections for the accused, and governs who should make such determinations.

SUBSTANTIVE JUSTICE

Substantive justice relates to the morality and legitimacy of a rule’s content. If rules themselves are unfair, illegitimate, or obviously contradicted by other workplace expectations, then there is little point in trying to have a “just culture” response to their violation. It won’t ever be just.

In the case of the 16-year-old patient and nurse mentioned previously, a workaround “rule” had been approved by the hospital in which anesthetists could give drug orders by phone from the floor where they worked and did not have to see patients in the obstetric ward at that time. Nurses could do all the actual work by the bedside. When things went fatally wrong this time, it wasn’t the workaround or unfairness of the rule that was put on trial: it was the nurse. You might say, of course, that workers like that nurse consent to organizational (and other) rules when they take up their occupations. In many cases, such consent is indeed an important consideration in assessing the rule’s legitimacy. But consent is not definitive. Workers can be placed in positions where they have little choice but to submit to preexisting rules or workarounds.

Rules may not be written with everybody’s legitimate interests in mind, or in ignorance of the goal conflicts and resource constraints that drive real work. Here is an example from Ref. 26.

BREAKING THE RULES TO GET MORE RECRUITS: SOME SAY CHEATING NEEDED TO FILL RANKS

It was late September when the 21-year-old man, fresh from a psychiatric ward, showed up at a US Army recruiting station. The two recruiters there quickly signed him up. Another recruiter said the incident hardly surprised him. He has been bending or breaking enlistment rules for months, he said, hiding police records and medical histories of potential recruits. His commanders have encouraged such deception, he said, because they know there is no other way to meet the Army’s recruitment quotas.

“The problem is that no one wants to join,” the recruiter said. “We have to play fast and loose with the rules just to get by.” Others spoke of concealing mental health histories and police records. They described falsified documents, wallet-size cheat sheets slipped to applicants before the military’s aptitude test, and commanding officers who look the other way. And they voiced doubts about the quality of troops destined for combat duty.

Recruiting has always been a difficult job, but the temptation to cut corners is particularly strong today, as deployments in Iraq and Afghanistan have created a desperate need for new soldiers, and as the Army has fallen short of its recruitment goals in recent months. Says one expert: “The more pressure you put on recruiters, the more likely you’ll be to find people seeking ways to beat the system.”

A retributive just culture presumes that existing rules enjoy a priori legitimacy. But what if they don’t, or when they are really trumped by other organizational pressures and expectations as in the example above? Listening to the voice from below—from those who have to get the job done and use the rules—is the best way to achieve substantive justice.

•  Involvement of those who will have to do the work enhances the legitimacy of the rules that apply to that work.

•  Taking part in the process of developing the rules increases the sense of ownership the workers feel toward the rules. The rules derive from their own insights, arguments, and experiences.

•  Developing the rules in connection with the workers ensures the rules connect with reality. The standards are not designed for an ideal environment, imagined without time pressures, complicating factors, and conflicting information. Instead, the written rules (and practices taught by educators) align with and support normal practice in the field.

PROCEDURAL JUSTICE

Procedural justice sets down legitimate processes for determining rule breaches, offers protections for the accused, and governs who should make such determinations. Let’s look at

•  The necessity for independent judges

•  The right to fair hearing and appeal

•  A differentiation between guilt-phase and penalty-phase deliberations

Independent judges have no personal stake or conflict of interest in the affair at hand. This aspect of procedural justice bars a person from deciding a case in which she or he has something to win or lose. Think about what this means for a line manager who applies a just culture process to one of his or her reports. Line managers can be fairly suspected (if not shown) to have reputational, career-related, economic, or other stakes in adjudicating an error or violation that happened on their watch. This gives them an interest in the outcome, which introduces the kind of actual or suspected bias that procedural justice and due process rights aim to prevent. At the same time, knowledge of the messy details or subtleties of what it takes to get a job done under goal conflicts and resource constraints is certainly as important. Research has shown that the legitimacy of being called to account is linked to how much the judging person knows about the process, profession, or practice in question.²⁷ Those who know the “messy details” of real (rather than imagined) work tend to enjoy greater credibility. The problem of course is that these are not likely to be independent. Finding a judge who is both independent and intimately knowledgeable about how work is actually done may require an organization to look outside or across different sites or parts of itself (such as employing another line manager from a separate site as an independent assessor).

Due process rights also include the right to a fair hearing. This normally involves

•  Prior notice of the case made against you and getting to see the evidence that will be used against you.

•  Knowing what is at stake (not just that there is “a case”).

•  A fair opportunity to answer the case and evidence against you.

•  The right to bring an advocate who might support you or argue on your behalf (this could be a colleague or union representative).

•  An opportunity to present your own case or angle.

•  Openness to scrutiny of the case and its proceedings by other parties. If things are done behind closed doors, then this opportunity may not exist. If colleagues or other stakeholders are present, then it does.

•  The right of appeal. This offers a process for requesting a formal change to an official decision. Appeal may be called for, among other reasons, because of a suspicion of abuse of power, because someone acted in excess of jurisdiction, because evidence was used that shouldn’t have been included, or because evidence was ignored that should have been considered.

Nothing in a typical just culture program itself offers such assurances, even though it is possible (and practically and morally desirable) to develop and offer them. If you have a retributive just culture process, then think what it would take in your own organization to give your employees assurance of procedural justice. When you see a list of procedural guarantees like the one above, you might realize that having a just culture is no longer so simple. It is certainly not as simple as buying an algorithm with three categories (honest mistake, risk-taking behavior, negligence) off the shelf.

If you are ready for even more nuances and complexities, then consider this. Even with procedural guarantees in place, a “just culture” program organized around categories of culpability does not distinguish between guilt-phase versus penalty-phase culpability. This divides the question about culpability in two:

1.  Did the person knowingly commit the act (guilt phase)?

2.  What penalty should be assigned once guilt is established (penalty phase)?

A retributive just culture program presumes guilt, or at least (some) responsibility for the outcome; otherwise the algorithm wouldn’t be applied. This might remove the presumption of innocence until proven otherwise, and thus create a short circuit to penalty deliberations. Research has shown that penalty-phase deliberations often focus on the transgression, the transgressor, and the outcome, rather than on mitigating factors.²⁸ A just culture process that makes no differentiation between guilt-phase and penalty-phase deliberations might automatically overlook mitigating factors (the-what-is-responsible question) in favor of asking who is responsible and holding that individual accountable.

These are all checks and balances that have made it into retributive justice throughout the ages. This has occurred under names like “natural justice,” “due process,” or “the duty to act fairly.” Lots of people have fought and died for these assurances. Yet many organizations today that have adopted a retributive just culture are proceeding without any of that in place. Without such checks and balances, your culture will not be seen as just. It will only be seen as an exercise of power, as a serious game that decides who wins and who loses.

SUMMARIZING AND MANAGING THE DIFFICULTIES WITH RETRIBUTIVE JUSTICE

Given the negotiability of the line between acceptable and unacceptable behavior, it is not surprising that a just culture based on retribution typically creates the following problems:

•  There is a lack of clarity, agreement, or perceived fairness about who draws the line between the shades of culpability. In many cases, the judge is not independent: he or she actually has a stake in the outcome.

•  Does the judge or “jury” know the nuances and messy details of the practitioner’s work? If not, how can they really know what constitutes risk or risk-taking in that world?

•  There is often no possibility for appealing a decision that is made. This type of retributive just culture is known by some employees “as a good way to get yourself fired” and by managers “as a good way to get rid of someone.”

•  There is actually no convincing evidence that organizations with a retributive just culture have higher reporting rates or that they learn more of value after an incident.

•  The more powerful people in an organizational hierarchy typically consider their organization’s culture to be more “just.”²¹

•  Retributive justice is not always known to promote honesty, openness, learning, and prevention.

So if you believe you want to institute a just culture based on retribution, be sure to at least assess these three questions for your own organization:

1.  Is the “judge,” the one who draws the line on the practitioner’s behavior, independent? A “judge” who has a stake in the outcome is not independent. For instance, a nurse manager who assesses the performance of one of his or her nurses in case of a medication adverse event is not independent. In fact, any manager is not independent. They always have a stake in the outcome of the judgment, as they and their decisions may be implicated in an incident too—unless blame can be put on the worker.

2.  Does the “judge” or “jury” know enough about the messy details of practice to know about the many unwritten rules, standards, and expectations about how work actually gets done? Are you allowing people’s choices and actions to be judged by their “peers”?

3.  Is there an opportunity for appeal? If you really acknowledge that everybody is inescapably fallible (as is one of the premises of a retributive just culture), then that goes for judges and juries too. Justice should therefore offer people a chance to be heard again by an unbiased party. Does that happen in your organization, and if not, how could you provide that assurance?

Here are some additional steps and considerations for your organization if you do wish to pursue a path of retributive just culture.

•  Design and advertise your just culture process clearly. Tell your people, for example, where and with whom it starts, what the following steps are, where and when a judgment is made on the employee’s behavior, and what the opportunities for appeal are.

Recognize that an adverse event review is not a performance review. Consistency of your review processes across professional groups and departments is difficult, but also important for achieving fairness and justice. Training your investigators to conduct the kind of learning review that asks what is responsible for an incident (rather than who is responsible) is an important part of this.²⁹

•  Decide who is involved in the just culture process in your organization. If the employee’s manager is in charge of the process, then the potential of career or reputational jeopardy may hinder an employee’s honest disclosure.²⁰ A set-up in which impartial staff take in the story and then funnel it to the manager for appropriate action can generate more opportunities for learning and less fear of retribution.^17,30 Yet how much domain expertise do you need to involve in the process? Understanding the messy details of practice (including the many gaps between rules or guidelines and actual daily work) is crucial for both credibility and a sense of justice.

•  Decide who is involved in deciding who is involved. There is something recursive about this, of course. But if decisions are made top-down about the previous points, then any just culture process will lack the buy-in, ownership, and constituency for employees to feel that it is something of their own creation—something to stand for and participate in to the benefit of the organization.

I recall how one safety-critical industry was under intense media scrutiny in a country where I once lived. The newly elected government had pledged to the public that it would let the industry continue to function if it were safe. Then reports started to leak out about operators drinking on the job, about an internal erosion in safety culture, about a lack of trust between management and employees. The regulator was under exceptional pressure to do something. To show that it, and the government, could be trusted.

So the regulator sent parts of the cases it had discovered to the prosecutor. The media loved it: now something was happening! Maybe crimes had been committed by people to whom the public had entrusted the running of this safety-critical technology! Now somebody was finally going to be held accountable.

The regulator saw how some of the media spotlight on it got dimmed. It could breathe a little easier now. But it was a bittersweet lull. The relationship with the industry was dramatically disturbed. Regulators have to rely on open disclosure by people in the industry they regulate; otherwise they have no accurate or truthful information to go and regulate on. Such disclosure was now going to be very unlikely. It would be, for years to come.

In addition, safety improvements, at least for the media (and thereby public opinion, and, by extension, the government’s stance on the issue) could now be largely collapsed into the pursuit of a few bad apples in the industry’s management. Now that these people would be held accountable, any other safety improvements could simply be assumed to be less important, or to follow automatically. Of course they would not. Publicly or legally reminding people of their responsibilities may have some effect in getting them or others to behave differently (though never for a long time). And the negative consequences of such accountability easily outweigh these effects.

RESTORATIVE JUST CULTURE

Restorative justice as we know it today began in response to relatively minor crimes—often property crimes, such as burglary in the 1970s. But its roots can be traced much further back: hundreds and even thousands of years. The ancient Sumerians, Babylonians, Hebrews, Romans, and Gauls—to name a few—all applied forms of restorative justice. They organized restitution for property crimes, relying on input from victims and offenders. First nations in North America, New Zealand, and other parts of the world have also long had values and practices consistent with restorative justice. Encouraged by the results and humane approach, stakeholder groups have advocated restorative justice particularly for juvenile offenses in a number of countries. And it has been spreading to schools, workplaces, and religious organizations. The restorative approach to justice has even been applied successfully on a massive scale, through the Truth and Reconciliation Commission in post–Apartheid South Africa. Restorative justice, broadly, is defined as

…a process where all stakeholders affected by an injustice have an opportunity to discuss how they have been affected by the injustice and to decide what should be done to repair the harm.³¹

Restorative justice does not have to be very difficult, actually. And it can even come intuitively, without policies or rules to guide it. Not long ago, I was speaking with hospital administrators in Asia who run a large, complex campus hospital. A delivery man, coming in with supplies on a cart, had knocked over a pram with a baby in it. The mother was of course quite upset, even though the baby was not hurt in the incident. The mother, in conversation with administrators, gave the hospital seven days to come up with an intervention that would prevent it from happening again. She took no other action, neither toward the hospital nor toward the delivery company or its employee.

Once they starting looking into the problem during those days, hospital administrators found that they had no guidance for their deliveries whatsoever: delivery contractors did what seemed right, and there was a wide variety of practices and quality. They developed standards in consultation with their contractors, and the mother was satisfied.

Justice was served without punishing anyone. The delivery company did not sanction or fire its employee. In fact, the company was a bit angry initially because it said the hospital had never told them how it would like deliveries to be done. So why did they suddenly come with this now?

But without knowing it, the parties had asked and answered the restorative questions. Who got hurt? The baby, the mother. What were their needs? Reassurance that it won’t happen again. Whose obligation was it to meet those needs? The hospital’s, in consultation with the delivery contractors. And in developing the solution, the community was involved: relevant people in the hospital, the mother, the contractors.

Then, two years later, another delivery contractor knocked over a woman, resulting in a leg fracture. Having learned from its spontaneous restorative processes, the hospital and delivery company apologized, agreed to manage and cover the costs of treatment, and made some additional adjustments to its delivery standards and practices. Again, nobody was sanctioned or fired, and no punitive damages were paid. The woman felt that justice was served.

Let’s look at the various steps needed to create restorative justice. This includes finding out who has been hurt, what the person(s) needs are, and whose obligation it is to meet those needs. Restorative practices are focused on keeping the “offender” in the community rather than separating or exiling him or her, and to have the community play a big role in the restorative practice.

RESTORATIVE JUSTICE STEPS

Who Was Hurt, and What Are His or Her Needs?

The first questions asked by restorative justice is who was hurt and what his or her needs are. An incident in a hospital, for instance, or an airline, oil company, or other organization can hurt various (groups of) people. Recognizing the ways in which they hurt and responding to their needs is necessary if you want your organizational culture to become truly just.

•  First victims: Patients, passengers, colleagues, or surrounding community who suffer the consequences. Their needs might center on information about the incident, access to the practitioners involved, and some type of restitution and reassurance of prevention.

•  Second victims: The practitioner(s) involved who feel(s) personally responsible and suffer(s) as a result. They might need anything from empathy and compassion, to opportunities to show remorse, to counseling and trauma care.

•  Your organizational community also has needs. This is likely information about the incident and the organizational response to it. But they may also want an opportunity to help first and second victims, contribute to restoring relationships and trust, and achieve a sense of joint problem ownership.

What does each group need? First victims typically need information about the incident. What happened, and why? First victims are quick to see through speculation or legally constrained information. Access to the practitioner(s) involved can be an important way to help get them some of the authentic information they might crave. It is well known that lawsuits in the wake of patient harm are significantly less likely when there is immediate, open, and honest disclosure of what happened.³² This suggests that many first victims need honesty and information and an acknowledgment of their humanity more than they want financial compensation.

The daughter of a woman who was injured after receiving a medication to which she had a documented allergy commented on her mother’s preserved trust in her physician: “The reason [the physician’s] apology felt genuine was because it was direct. He didn’t beat around the bush. He didn’t try to cover things up.” Rather than simply assigning blame, patients and families want both to understand their situation fully and to know what the event has taught caregivers and their institutions.³³

First victims also need an opportunity (or multiple opportunities) to tell their story. The incident likely disrupted their trust in the system or its practitioners. Telling and retelling their story can help them integrate the incident into their world-view, to give it a place with some boundaries around it so that it does not forever keep affecting everything they do and are.

First victims might also feel a need to regain control over their experiences and emotions. They are likely to feel upset, betrayed, angry, disillusioned, disappointed, and confused. One way to help them is empowerment, for instance by involving them in investigative and restorative justice processes. First victims can help define the kind of obligation that will be asked of (and agreed with) the “offender.” This gives first victims an active part in determining proportionality, through which they can attempt to express what the incident meant to them—while keeping in mind the humanity of the second victim, the practitioner(s) involved. First victims also might need some sort of restitution. If their actual loss cannot be compensated, first victims typically want to know that everything is going to be done to prevent recurrence. They don’t want others to suffer like they did. This might even lead to the first victim urging justice for the second victim, as in the case that follows.

Air traffic controllers in Yugoslavia were charged with murder and were jailed in the wake of a midair collision between two passenger aircraft. One hundred and seventy-six lives were lost. It was 1976, and Zagreb was one of the busiest air traffic control centers in Europe. Its navigation beacon formed a crossroads of airways heavily used by traffic to and from southeastern Europe, the Middle East, the Far East, and beyond. The center, however, had been structurally understaffed for years. At the time of the accident, the radar system was undergoing testing and the center’s radio transmitters often failed to work properly. Through a combination of different languages and flawed data presentation to the controller, one of the aircraft managed to level off exactly at the altitude of another. Three seconds later, its left wing smashed through the other’s cockpit and both aircraft plummeted to the ground. “Improper air traffic control operation,” the accident investigation concluded. One controller, however, was singled out and sentenced to a prison term of seven years, despite officials from the aviation authority offering testimony that the Zagreb center was understaffed by at least 30 controllers. Significantly, the father of one of the victims of the collision led an unsuccessful campaign to prevent the controller’s jailing. He then joined the efforts of other controllers to have him released after serving two years.³⁴ One of the major reasons for his efforts was that he, as a first victim, did not believe that jailing the controller was fair to either the first or the second victims. First victims got no assurance of any improvements and possible prevention of repetition, and second victims (practitioners from the sharp end) were unfairly singled out for what was the failure of an entire complex system and the organizations set up to manage it. It was not until the early 1990s that the air traffic control system around Zagreb was revamped.

What do second victims typically need? For most professionals, an error that leads to an incident or death is antithetical to their identities. They themselves can see it as a devastating failure to live up to their professional commitment. Having made an error in the execution of a job that involves error management and prevention is something that causes excessive stress, depression, anxiety and other psychological ill health. Particularly when the work involves considerable autonomy and presumptions of control over outcomes on the part of the actor (such as doctors, pilots, air traffic controllers), guilt and self-blame are common, with professionals often denying the role of the system or organization in the spawning of their error altogether and blaming themselves entirely. This sometimes includes hiding the error or its consequences from family and friends. Practitioners might distance themselves from any possible support, or attempt to make atonement themselves with those who were harmed by the error. The memory of error stays with professionals for many years.³⁵ All of these effects are visible, and can be present strongly even before your organization does anything, or before a manager or prosecutor might do anything.

In the best case, second victims seek to process and learn from the mistake, discussing details of their actions with colleagues or employers. Practitioners can punish themselves harshly in the wake of failure. You or your organization or society can hardly make such punishment any worse—other than confirming what the practitioner already feels. As told by one physician in 1984 about a couple whose pregnancy was lost:

…although I told them everything they wanted to know and described to them as completely as I could what had happened, I never shared with them the agony that I underwent trying to deal with the reality of events. I never did ask their forgiveness…. Somehow, I felt it was my responsibility to deal with my guilt alone.³⁶

What second victims typically need is an opportunity to tell their story, to not feel alone, singled out, vilified, or shunned. They probably do want to offer their account of what happened; they too want to find out how things could go so wrong and what they and others could do differently in the future. They may need anything from empathy and compassion to counseling and trauma care. These are among the things your organization might consider:

•  Second victims need to regain trust in their own competencies, and rebuild relationships with others who rely on them.

•  They also want reassurance that they, their organization, and their community have put things in place to prevent recurrence.

•  They may want to contribute to an investigation, and help by suggesting countermeasures or improvements. This gives them an opportunity to convert their feelings of guilt into social or practical action.

The role of the organization or peers in facilitating such coping is important. This can be done through peer or managerial support and appropriate structures and processes for learning from failure that might already be in place in your organization. Research on employee assistance programs stresses that employees must not get seen as the source of the problem, or treated as somehow “troubled” as opposed to “normal” employees. Social support, and particularly peer support, can then be a crucial moderator of the stress, anxiety, and depression that a second victim can experience in the aftermath of an incident. Such support is also a strong predictor of the second victim coming out psychologically healthy. Guidance on setting up effective peer support and stress management programs in the wake of incidents is available in separate work.³⁷ I have written much more about the experience of the second victim, and about trauma, guilt, resilience, and forgiveness in the book Second Victim.⁹

Criminalization, the topic of Chapter 4, makes things a lot worse for the second victim. Criminalization affirms feelings of guilt and self-blame and exacerbates their effects, which are linked to poor clinical outcomes in other settings.³⁸ It can lead to practitioners going on sick leave, divorcing, exiting the profession permanently, or even committing suicide. Another response, though much more rare, is an expression of anger and counterattack, for example through the filing of a defamation lawsuit. Criminalization can also have consequences for a person’s livelihood (and his or her family), as licenses to practice may be revoked. This in turn can generate a whole new layer of anxiety and stress. One pharmacist, whose medication error ended in the death of two patients, suffered from depression and anxiety to such an extent that he eventually stabbed his wife to death and injured his daughter with a knife.³⁹

Identifying the Obligations to Meet Needs

Wrongs or harms result in obligations. In a restorative just culture, these obligations are acknowledged and articulated. They get met by different stakeholders, preferably in collaboration with each other.

The practitioner (or second victim) can, for example, be obliged to

•  Honestly disclose his or her role in the incident and give an account to the others involved or affected

•  Recognize the needs of first victim(s), organization, and community

•  Show remorse and be open to various ways to put things right with first victim, organization, and community

•  Identify pathways to prevention in collaboration with first victim(s), organization, and community

The organization and surrounding community can embrace the obligation to

•  Offer support to first and second victims (through open disclosure or critical incident stress management programs)

•  Not fire or sanction people just because they were involved in an incident

•  Ask itself honestly what was responsible for the incident, not who

•  Perform an investigation on the premise that people did not come to work to do a bad job, and one that asks why it made sense for people to do what they did

•  Identify pathways to prevention, in collaboration with first and second victim(s)

The first victim typically has an obligation to

•  Respect the humanity of the practitioner involved in the incident

•  Be willing to be part of the solution, for example, by contributing ideas for possible prevention

I learned of an incident in an airline not long ago, to which the chief executive and safety manager had very different responses. On the back of significant expansion, the airline had hired many new cabin crew members. The ones in the back galley were typically the junior ones and had to work really hard on short flights (with short turnaround times at each end) to get everything done that they needed to do. Right after landing, a junior crew member in the back galley noticed that a catering cart slid out of its place and moved forward when the aircraft started braking after landing. Strapped into her jumpseat, she tried to grab it, but it was too late. The cart careened down the aisle, all the way to the front of the aircraft. On the way, it crushed a passenger’s foot and then caused significant damage in the front of the cabin where it came to rest.

The CEO of the airline, on hearing about the incident, decided to write an angry email message to all of his managers, impressing upon them the need to make sure their people followed the rules to the letter and were careful and vigilant in their work. A catering cart had not been properly secured! How could that even happen if people were simply doing the right thing? Stuff was going to get broken this way, and it was going to cost the airline a lot of money.

The safety manager had a very different response. Without knowing anything about systems of restorative justice, he asked the very questions that animate it. Who was hurt? What were the person(s) needs? Whose obligation was it to meet those needs? The passenger was clearly hurt. He needed medical care. He needed bills paid. He needed an apology, or various apologies. But the cabin crew member was hurt too. She was devastated about what had happened, and fearful of losing her job. Other passengers too, as a community, may have been frightful of what they saw happen. Trust in the airline might have been hurt. Carefully plotting his way through the tense political landscape in the aftermath of this incident, the safety manager was able to get people to talk to each other about the hurts, needs, and obligations that had resulted from it. He was successful. The passenger was satisfied, the cabin crew member kept her job, and even the CEO was content with the outcome.

Identifying and meeting obligations is ultimately about putting right what went wrong. It is about making amends. In restorative practices, this means promoting reparation and healing for all affected by the incident. This notion of reparation, of restitution or “paying back,” is central to retributive justice too, of course. In restorative practices, however, everything possible is done to reintegrate the practitioner into the community, and the “payment” typically is made in a different currency. Restorative practices ask you to

•  Address the harm done to first and second victims of the incident, as well as the surrounding community

•  Address the systemic issues that helped produce the incident by asking what was responsible for it, so that other practitioners and first victims are less likely to end up in a similar situation

For restorative practices to be meaningful and seen as just by all involved, you have to be collaborative and inclusive. Effective restoration relies on this engagement. An incident can affect many people, and its aftermath typically has many stakeholders. These might be given access to, and information about, each other. All can then be involved in deciding what justice requires in their case. This may mean an actual dialogue between parties (e.g., first and second victim), to share their accounts and arrive at an agreement on what should be done. How might the creation of restorative justice look in your organization? It will likely involve the following steps and people:

•  Encounters between stakeholders. The first one is likely to be between your organization and the practitioner(s) involved in the incident. Remember your organization’s obligations above!

•  An encounter between first and second victims, appropriately guided, may follow. Surrogates or representatives may need to be used in some situations.

•  Encourage all stakeholders to give their accounts, ask questions, express feelings, and work toward a mutually acceptable solution.

•  Acknowledge the harm, restore the balance, and address your future intentions.

To succeed, you will need to broaden out the conversation about a restorative just culture. Where necessary, include senior management, the board, your regulators, human resources, your safety department, unions or professional associations, customers, and other stakeholders. The more a solution is created and supported by the community from which both second and first victims stem, the more likely it is that such a solution is accepted by them and acted upon. This even goes for any sanctions you might want to come up with. Indeed, research shows that the severity of sanctions is a poor predictor of the effectiveness of the sort of social or managerial control you wish to exercise through them. Instead, the extent to which such sanctions are socially embedded in the community is a stronger predictor.³¹ In other words, if you want your practitioners to act differently, do not scare them by separating out one of them and turning him or her into an “example.” Rather, engage them with the community that will have to keep working together to ensure safe work, to ensure future successful outcomes. There is nothing wrong with having a practitioner feel temporarily out of favor—if indeed pathways for restoration are offered (see later under forgiveness). But your organization will gain nothing from a climate in which practitioners are constantly fearful and insecure in their relationship to you and to each other. That surely is a culture without trust. And such a culture can never be a just culture.

Ask yourself these questions to check how close to restorative just culture you and your organization might be:

1.  Does your just culture process address harms, needs, and causes?

2.  Is it adequately victim oriented (including both first and second victims)?

3.  Are practitioners encouraged to recognize their contribution to the (potential) harm caused, but also treated as potential second victims?

4.  Are all relevant stakeholders involved?

5.  Is it based on dialogue, participation, and collaborative decision making?

6.  Does it identify and address deeper, systemic issues that gave rise to the incident in the first place?

7.  Is it respectful to all parties?

RESTORATION AND FORGIVENESS

There is of course a link between restorative practices and forgiveness. But forgiveness is not necessarily the goal of restorative justice. It can be one of the outcomes, but one party can never be obligated to forgive another—that would be a meaningless, hollow act. Forgiveness is fundamentally relational.¹⁷ Forgiveness, as well as the processes of disclosure (or confession) and apology (or repentance) that necessarily precede it, has religious connotations. But that is no reason to dismiss them as irrelevant to a just culture in your organization. In fact, Nancy Berlinger, a colleague at the Hastings Center in New York, suggests that there is a lot that you can do in your organization to put conditions in place that make disclosure, apology, and forgiveness possible, often at little or no monetary cost. Such conditions include

•  Promptly acknowledging an error and offering the first victim(s) an authentic account of what happened.

•  Being sympathetic to calls for accountability even when the system has contributed considerably to producing the failure. Some first victims may have trouble understanding system failure, seeing only personal shortcomings. Others may be keen to see the organization move beyond its focus on the individual and address system-level issues instead.

•  Providing opportunities for second victims to process incidents and receive support in an environment that is neither punitive nor demeaning.

•  Nurturing a commitment that withholding the truth violates the humanity and autonomy of the first victim, and has a corrupting effect on second victims and their colleagues.

•  Avoiding the sort of scapegoating of subordinates that would diminish your own responsibility.

•  Avoiding assertions that the first victim was somehow to blame (e.g., non-compliant patient, obese passenger who could not evacuate quickly enough, wayward pilot who did not follow air traffic control instructions).⁴⁰

In one case, the first victim actually made public disclosure a condition for forgiveness. For her to desist from legal action, the second victim had to write about the surgical error he had made and publish it in a top journal in the field. He did, and it appeared in Surgical Endoscopy in 1995. The first victim kept her word, and the second received many positive responses.⁴¹

Disclosure is incomplete without apology. And both are preconditions (though not guarantees) for receiving forgiveness. Apology (or, more correctly, repentance) means allowing yourself to feel, and express, sincere sorrow and regret for what you have done, and for what has happened as a result. Both disclosure and apology contribute to accountability: the giving of an account that includes narrative, explanation, expressions of regret, and even admissions of guilt and responsibility.

Disclosure and apology may still be an act of healing in cases where the relationship between the first and second victims itself is harder to heal. Writing in the New England Journal of Medicine, a surgeon told of performing a wrong operation on a 65-year-old woman’s left hand, something he managed to correct not long after.⁴² Yet in this case, a relationship with the first victim was no longer really possible. The patient’s son received apologies, fee waivers, and offers of follow-up care for the mother. But she had lost faith in the doctor and would not return to have her sutures removed, or receive any other care, or ask for an apology or explanation. There was nothing the surgeon could do to demand contact or urge forgiveness.

Organizational practices that could support repentance include

•  Not forcing the first victim to interact with the second (i.e., the practitioner involved in their injury or loss) if there is no wish to do so—and vice versa

•  Appreciating the difference between appropriate feelings of guilt (“I made a mistake”) and destructive feelings of shame (“I am a mistake”) that might be felt by second victims

•  Offering first victims and their families access to care or other services should they need them

•  Meeting the obligations that result from the incident, injuries, or loss on the part of the first victim(s)

•  Recognizing that asking a first victim for “forgiveness” may be obtrusive or culturally inappropriate, while at the same time working to create conditions that may allow both victims, in their own time, to detach from the incident as a continuing source of pain, anger, and injustice⁴⁰

Organizational practices can come very close to forgiveness, or at least create appropriate conditions, without demanding it. The relational and restorative nature of these practices can easily be recognized:

•  Inviting first and second victims to be part of your organization’s or industry’s quality and safety improvement processes, though not making it their responsibility

•  Offering safe places or rituals for second victims to explore their reactions and responsibilities concerning errors and incidents

•  Identifying and changing aspects of the professional culture in your organization that deny the fallibility, and therefore the humanity, of your practitioners

•  Identifying and changing features of your organization that work against truth-telling, accountability, compassion, and justice in dealing with incidents⁴⁰

COMPARING AND CONTRASTING RETRIBUTIVE AND RESTORATIVE APPROACHES

Remember the different sets of questions asked by retributive versus restorative approaches to justice (see Table 1.1). Notice the different ways of thinking about accountability in retributive versus restorative just cultures:

•  In retributive justice, an account is something you pay. Retributive justice holds people accountable by making them repay the debt they morally owe. People have to settle their account with their organization, victims, community, and society.

•  In restorative justice, an account is something you tell. In restorative justice, people together figure out how to offset the harm. It holds people accountable by having different parties tell their often contradictory and only partially overlapping accounts.

The line between retribution and restoration is not entirely clear. You could actually do some of both, at the same time. Both retribution and restoration can be seen or treated as processes that help reintegrate a person into a community. In some cases, retribution might even be a precondition for restoration. As in: First pay up as a sign that you are taking responsibility. Once that is settled, we’ll happily have you back—even if you are a bit poorer, or with less status. I have seen pilots who had been demoted in the wake of an incident in the crew room, behaving as if nothing had happened. But something had happened, and they had one fewer stripe on their shoulders to show for it. Yet there they were, happily laughing and chatting away with their colleagues, part of the “tribe” just like before.

TABLE 1.1
Contrasting Retributive and Restorative Questions to Ask

Retributive	Restorative
Which rule has been broken?	Who has been hurt?
Who did it?	What are the person(s) needs?
How bad is the infraction, and so what does the person deserve?	Whose obligation is it to meet those needs?

So we shouldn’t overstate the contrast between retribution and restoration. For instance, whether it is done under regimes of restoration or retribution, some form of “penance” is fundamental to processes of forgiveness. This involves telling your story, expressing that you are sorry, and in some way accepting (part of the) responsibility for the outcome. Of course, retributive and restorative practices have different ways of getting people to tell that story and accepting responsibility. And they may get different stories as a result. However,

•  Both retributive and restorative forms of justice and accountability acknowledge that a “balance” has been thrown off by the act and its consequences.

•  Both also understand that there needs to be a proportional relationship between the act, the consequences, and the response to it.

•  Thus, both are designed around some sense of reciprocity, of “evening the score.” They differ, however, on the “currency” that is used to rebalance the situation, to even that score, to fulfill the obligations.⁴³

NEITHER RETRIBUTIVE NOR RESTORATIVE JUSTICE “LETS PEOPLE OFF THE HOOK”

Neither form of just culture gets “people off the hook.” Both hold people accountable. In both, people are expected to engage with, and respond to, the community of which they are, or were supposed to be, part. Both forms of just culture impose accountability. But they go about it in different ways. Retributive justice achieves accountability by looking back on the harm done, or potentially done, by the person.

•  It asks what the person must do to compensate for his or her actions or consequences.

•  Justice is created by meeting hurt with hurt. (Potential) hurt is compensated by imposing more hurt—deserved and proportional.

•  People can feel that the person is held accountable by not letting him or her off the hook.

•  The community can demonstrate that it does not accept what the person did (it would not accept such actions from any of its members) and demonstrates that it makes the person pay.

The focus in restoration, in contrast, is not chiefly on what some specific “offender” deserves, like it is in retributive justice. Restorative justice achieves accountability by looking more ahead at what must be done to repair the trust and relationships that were harmed by the person’s actions.

•  This makes it important for others to understand why it made sense for the person to do what he or she did, and how others could perhaps be put in the same situation.

•  Rather than seeing the “offenders” as causes of trouble, restorative practice will tend to see “offenders” as inheritors of organizational, operational, or design issues that could set up others for failure as well.

•  Restorative practices are thus likely to get to the systemic issues that triggered the incident, to identify the deeper conditions that allowed an incident to happen.

•  For this to work, of course, the people involved need to tell their account, their story. This also gives them the opportunity to express remorse for what they did or for what happened, should that be appropriate.

•  The people affected by the incident, as well as other community members, can explore and agree on what needs to be done to restore trust and relationships.

•  The community demonstrates that it expects people to be accountable by getting them to reflect on their behavior and sharing the insights.

RETRIBUTIVE AND RESTORATIVE FORMS OF JUSTICE DEAL DIFFERENTLY WITH TRUST

The two forms of just culture also approach trust differently. Retribution builds trust by reinforcing rules and the authority of certain parties or persons to police and enforce them. It says that where people work to get things done, there are lines that should not be crossed. And if they are, there are consequences. Think about it like this: if you find that people “get away” with breaking rules or doing sloppy work, you don’t have much trust in the system, or in your community’s ability to demand accountability. Your trust can be restored if you see an appropriate and assertive response to such behavior. You can once again rest assured that the system, or your community, does not accept such behavior and responds in ways that make that clear—to everyone.

Restoration, on the other hand, builds trust by repairing fiduciary relationships. Fiduciary relationships are relationships of trust between people who depend on each other to make something work. Consider the work done in your own organization. People in your organization depend on each other. Every day, perhaps every minute, they have to trust each other that certain things get done, and get done in a timely, appropriate, and safe manner. They might not do these things themselves because they are not in the right place, or because they lack the expertise or authority to do them. So they depend on others. This creates a fiduciary relationship: a relationship of trust. It is this relationship that is hurt or broken when things go wrong. And it is this relationship that needs restoring.

TABLE 1.2
The Different Ways in Which Retributive and Restorative Processes Try to Create Justice

Retributive	Restorative
Believes that wrongdoing creates guilt, and demands punishment that compensates it	Believes wrongdoing creates needs, and obligations to meet those needs
Believes an account is something the offender pays or settles	Believes an account is something the offender tells and listens to
Asks who is responsible for the incident	Asks what is responsible for the incident
Learns and prevents by setting an example	Learns and prevents by asking why it made sense for people to do what they did
Focuses on what people involved in the incident deserve	Focuses on what people involved in, and affected by, the incident need
Creates justice by imposing proportional and deserved punishment	Creates justice by deciding who meets the needs arising from the incident
Meets hurt with more hurt	Meets hurt with healing
Looks back on harm done, and assigns consequences	Looks ahead at trust to repair, and invests in relationships
Builds trust by reinforcing rules and the authority to impose and police them	Builds trust by repairing relationships between people who depend on each other to accomplish their work

Both kinds of trust can be important for your organization or community (see Table 1.2).

CAN SOMEONE OR SOMETHING BE BEYOND RESTORATIVE JUSTICE?

Retributive theory believes that pain will vindicate.⁴³ That is, responding to hurt with more hurt will somehow equalize or even eliminate the injustice that has been inflicted. Restorative theory, instead, believes that pain requires healing. But are there cases where those who have inflicted the pain are beyond healing, beyond the reach of restorative justice? Advocates of restorative approaches might like to believe that nothing or nobody is beyond the reach of restoration or reintegration. Yet many others can point to cases in which they feel a retributive response is the only appropriate one. Some cases may call for a process that gives attention to societal needs and obligations above all others—particularly above the needs of any immediate stakeholders (e.g., first and second victims).

It isn’t the case itself that determines whether it is beyond restoration. Rather, it is our judgment about the case; it is about what we find important, what we find “just” or the morally right thing to do. So always ask the question: who gets to decide whether a case is beyond the reach of restorative approaches, and what are their stakes (if any) in saying it is so?

If the decision to forego restoration is made and generally agreed with, there are a few of things to remember.

•  Retributive justice is often criticized for not being sufficiently victim oriented. First-victim oriented, that is. You will find some examples of this in Chapter 4. First victims may feel left out or sidelined. They have a stake in the creation of justice, but are often given no voice to contribute toward this end.

•  Retributive justice is also criticized for not involving the community enough, for not embedding sanctions socially, for doing its own thing essentially between two parties (the party doing the judging and the party being judged). Retributive justice engages these parties in a process that is mostly removed from the rest of the community, often conducted in a language that is alien to that used in practice in the community, and away from the time and the place where the incident happened.

•  Openness to different accounts of what happened can get sacrificed in an adversarial setting where one account wins and one account loses. As a result, not much of value might be learned; not many systemic improvements may follow from retributive justice.

•  Retributive approaches can encourage “offenders” to look out for themselves and discourage them from acknowledging their responsibility in any concrete ways because it might self-incriminate them even further.

If you can find ways to mitigate these negative aspects of retributive justice, it will likely help others see your responses as more “just.” So whatever you do, ask who is hurt. Give a voice to the different stakeholders. Identify responsibilities and obligations that various parties need to meet—not just the “offender” or second victim. Try to socially embed your responses, so that the community feels part of the solution.

CASE STUDY

ARE ALL MISTAKES EQUAL?

Not all rule-breaking can be seen the same way. Or can it? Some actions of your people must be considered to be worse than others. Not all breaches of trust are the same. Not all actions, or even mistakes might be equally “forgivable.” Let’s look at one basic distinction that can be made in many professions: that between technical and normative errors. These are not types of errors that exist “out there” in the world. Rather, they are ways of constructing, or looking at and talking about, your people’s actions. Sometimes professions themselves have ways of making such distinctions, because they do useful work for them, their colleagues, their training and selection, and their organizations. Again, it does not mean that these categories are ready formed and should be hunted down and exposed. Instead, it is a way of looking at the way in which other people talk about error, and how they might end up with a judgment of whether the error is forgivable or not, whether the mistake is less equal than others.

When studying the way surgeons treat errors that can hurt (or have hurt) patients, Charles Bosk, a sociologist, saw a remarkable pattern. Surgeons and other physicians made a distinction between what he began to call technical and normative errors.⁴⁴ To be sure, it was not the error that is either technical or normative. It became technical or normative because of the way people looked at the error, because of what they saw in it, talked about it, and how they responded to it. The distinction can have powerful consequences for how your organization (or surrounding society) is prepared to deal with an error that occurred. Whether you construct an error as normative or technical has far-reaching consequences for exacting accountability and encouraging learning.

After this, we need to consider yet another really important factor in our judgment of whether a mistake is forgivable or not: knowing the outcome. Hindsight plays a huge role in how we handle the aftermath of a mistake. We assume that if an outcome is good, then the decisions leading up to it must have been good too—people did a good job. And it works in the opposite too. If the outcome of a mistake is really bad, or could have been really bad, we are more likely to see the mistake as culpable. There is more to account for. I will present a case that shows that knowing how things turned out determines whether we see actions as culpable or not.

TECHNICAL ERRORS: ERRORS IN A ROLE

When a practitioner makes a technical mistake, she or he is performing her or his role diligently, but the present skills fall short of what the task requires.

For example, when a pilot makes a hard landing, this could likely be the effect of skills that have yet to be developed or refined. Technical errors will be seen as opportunities for instructors or colleagues to pass on “tricks of the trade” (e.g., “start shifting your gaze ahead when flaring”).

People can be quite forgiving of even serious lapses in technique, as they see these as a natural by-product of learning-by-doing. Technical errors do not just have to be connected to the physical handling of a process or its systems; they can also involve interaction with others in the system, for example, air traffic control, nurses, night staff, physicians. The person in question, for example, may have seen the need to coordinate (and may even be doing just that), but does not have the experience or finely developed skills to recognize how to be sensitive to the constraints or opportunities of the other members of the system.

For an error to be constructed as technical, however, it has to meet two conditions:

•  One is obviously that the frequency or seriousness should decrease as a person gains more experience. When someone keeps making the same mistakes over and over, it may be difficult to keep seeing them as purely technical. However, as long as the person making the errors shows a dedication to learning and to his or her part in creating safety, that person is still conscientiously filling his or her role.

•  The other condition for a technical error is that it should not be denied, by the pilot involved, as an opportunity for learning and improvement. If a practitioner is not prepared to align discrepant outcomes and expectations by looking at him- or herself, but rather turns onto the one who revealed the discrepancy, trainers or supervisor or managers (or courts) will no longer see the error as purely technical.

A flight instructor reports: We had been cleared inbound to a diversion airport due to weather. We were on downwind when an airliner came on the frequency and was cleared for the instrument landing system (ILS) toward the opposite runway. The student proceeded to extend his downwind to the entry point he had chosen, even though the field was now fully visible. He was entirely oblivious to hints from air traffic control to turn us onto the base so we could make it in before the airliner from the opposite side. When the student still did not respond, I took control and steered our aircraft onto the base. We completed the landing without incident before the airliner came in. On debriefing, however, the student berated me for taking control, and refused to accept the event as an opportunity to learn about “fitting in” with other traffic at a dynamic, busy airport. He felt violated that I had taken control.

Professionals with limited experience may not be very sensitive to the unfolding context in which they work. They have simply not yet learned which cues to pick up on and how to respond to them. People will see such insensitivity as a technical issue consistent with the role of student, and a due opening for enlightenment. Sticking to the plan, or behaving strictly in the box, even though a situation has unfolded differently, has been known to lead to problems and even accidents. So valuable lessons are those that demonstrate how textbook principles or dogged elegance sometimes have to be compromised to accommodate a changing array of goals. Professionals can otherwise end up in a corner. Surgery has a corollary here: “Excellent surgery makes dead patients.”

The benefits of technical errors almost always outweigh the benefits. Of course, this is so in part because the division of labor between senior and junior practitioners in most operating worlds (or between instructors and students) is staggered so that no one advances to more complex tasks until he or she has demonstrated his or her proficiency at basic ones.

Bosk tells how Carl, a surgical intern, was closing an incision, while Mark, the chief resident, was assisting. Carl was ill at ease. He turned to Mark and said “I can’t do it.” Mark said, “What do you mean, you can’t? Don’t ever say you can’t. Of course you can.” “No, I just can’t seem to get it right.” Carl had been forced to put in and remove stitches a number of times, unable to draw the skin closed with the proper tension. Mark replied, “Really, there is nothing to it,” and, taking Carl’s hand in his own, said, “The trick is to keep the needle at this angle and put the stitch through like this,” all the while leading Carl through the task. “Now, go on.” Mark then let Carl struggle through the rest of the closure on his own.

If aid is necessary, there are almost always only two responses:

•  Verbal guidance is offered, with hints and pointers.

•  Or the superordinate takes over altogether.

The latter option is taken when time constraints demand quick performance, or when the task turns out to be more complex than people initially assumed. This division of labor can also mean that subordinates feel held back, with not enough opportunity to exercise their own technical judgment. The preceding example could be an instance of this, where the division of labor is seen by the student as stacked in favor of the instructor. For instructors, supervisors, managers, and others, the challenge is always to judge whether the learning return from letting the practitioner make the mistake is larger than from helping her or him avert it and clearly demonstrating how to do so.

In another example, Bosk tells of the difficulty of performing a myelogram (a diagnostic procedure involving the removal of spinal fluid and the injection of dye in the spinal column) that had been ordered for a patient named Mr. Eckhardt. A senior student was to instruct a junior student in the procedure. They tried without any success to get the needle in the proper space. After some fumbling and a few sticks at Eckhardt, the senior student instructed the junior student to go “get Paul” (a second-year resident). Paul came in and surveyed the situation. After examining Eckhardt’s back he told the students, who were profusely apologizing for their failure, not to worry; that the problem was in Mr. Eckhardt’s anatomy and not in their skills. He then proceeded with some difficulty to complete the procedure, instructing the students all the while.

As for professionals, they should not be afraid to make mistakes. They should be afraid of not learning from the ones that they do make. Bosk’s study showed how self-criticism is strongly encouraged and expected of surgeons in the learning role (which is to say, almost every surgeon). Everybody can make mistakes, and they can generally be managed.

Denial or defensive posturing instead discourages such learning. It allows the trainee or subordinate to delegitimize mistake by turning it into something shameful that should be brushed aside, or into something irrelevant that should be ignored. Denying that a technical error has occurred is not only inconsistent with the idea that they are the inevitable by-product of training; it also truncates an opportunity for learning. Work that gets learned-by-doing lives by this pact: technical errors and their consequences are to be acknowledged and transformed into an occasion for positive experience, learning, and improvement. To not go along with that implicit pact is no longer a technical error; it is a normative one.

NORMATIVE ERRORS: ERRORS OF A ROLE

Technical errors say something about the professional’s level of training or experience. Normative errors say something about the professional him- or herself relative to the profession. Normative errors are about professionals not discharging their role obligations diligently.

•  Technical errors create extra work, both for superordinate and subordinate. That, however, is seen as legitimate: it is part of the trade, the inevitable part of learning by doing, of continuous improvement.

•  The extra work of normative errors, however, is considered unnecessary.

In some cases, it shows up when a crewmember asserts more than his or her role allows:

A senior airline captain told me about one case that he constructed as a normative error. It was my turn to go rest, he said, and, as I always do, I told the first and second officers, “If anything happens, I want to know about it. Don’t act on your own, don’t try to be a hero. Just freeze the situation and call me. Even if it’s in the middle of my break, and I’m asleep, call me. Most likely I’ll tell you it’s nothing and I’ll go right back to sleep. I may even forget you called. But call me.” When I came back from my break, it turned out that a mechanical problem had developed. The first officer, in my seat, was quite comfortable that he had handled the situation well. But I was a bit upset. Why hadn’t he called me? How can I trust him next time? I am ultimately responsible, so I have to know what’s going on.

The situation was left less resilient than it could (and, in the eyes of the captain, should) have been: leaving only two more junior crewmembers, with no formal responsibility, in charge of managing a developing problem. Of course, there are potential losses associated with calling:

•  The superordinate could think the call was superfluous and foolish, and get cranky because of it (which the first officer in this example may have expected and, as it turned out, misjudged).

•  The subordinate foregoes the learning opportunity and gratification of solving a problem her- or himself.

But the safe option when in doubt is always to call, despite the pressures not to. That is, in many cases, how a subordinate crewmember is expected to discharge her or his role obligations. In other cases, fulfilling those obligations is possible only by breaking out of the subordinate role, as a chief pilot once told me:

My problem is with first officers who do not take over when the situation calls for it. Why do we have so many unstabilized approaches to runways in (a particular area of our network)? If the captain is flying, first officers should first point out to him or her that he or she is out of bounds, and if that does not work, they should take over. Why don’t they, what makes it so difficult?

The chief pilot here flagged the absence of what may turn out critical for the creation of safety in complex systems: the breaking-out of roles and power structures that were formally designed into the system. Roles and power structures often go hand-in-glove (e.g., captain and first officer, doctor and nurse), and various programs (e.g., crew resource management training in aviation and healthcare, team resource management in air traffic control, bridge resource management in shipping, and so forth) aim to soften role boundaries and flatten hierarchies. These programs want to increase opportunities for coordinating viewpoints and sharing information. Where people do not do this, they fail to discharge their role obligations too—in this case by not acknowledging and deploying the flexibility inherent in any role.

Perhaps the fact that others see this as a normative breach is not so strange. In the kinds of operating worlds where we believe a just culture is important, it is very difficult to know and anticipate all the problems that may occur during a lifetime of practice. There will always be things that practitioners remain inexperienced with, simply because that kind of problem, in that kind of way, has not appeared before. Indeed, in complex and dynamic work, where resource limitations and uncertainty reign, failure is going to be a lasting statistical reality.

The possibility of suffering technical errors will consequently never go away entirely. In such worlds, where the knowledge base on how to create safety is inherently and permanently incomplete, many believe firmly in the importance of disclosing, discussing, and learning from error. When that does not happen, even an honest, technical error can become seen as a dishonest normative one.

“Covering up is never really excusable,” Bosk quotes an attending physician as saying. You have to remember that each time a resident hides information, he is affecting someone’s life. Now in this business it takes a lot of self-confidence, a lot of maturity, to admit errors. But that’s not the issue. No mistakes are minor. All have a mortality and a morbidity. Say I have a patient who comes back from the operating room and he doesn’t urinate. And say my intern doesn’t notice or decides it’s nothing serious and doesn’t catheterize the guy and doesn’t tell me. Well, this guy’s bladder fills up. There’s a foreign body and foreign bodies can cause infections; infection can become sepsis; sepsis can cause death. So the intern’s mistake here can cost this guy hundreds of dollars in extra hospitalization and possibly his life. All mistakes have costs attached to them. Now a certain amount is inevitable. But it is the obligation of everyone involved in patient care to minimize mistakes. The way to do that is by full and total disclosure.⁴⁴

The obligation to report or disclose, discuss, and learn seems to be a critical hinge in how we believe a just culture should work. But honest and open accounting can seem dangerous to many practitioners. How an error might be interpreted after-the-fact is sometimes entirely up for grabs. A technical one (missing an approach, or supplying the wrong drug because of inexperience with that particular drug or procedure or kind of patient) can easily be converted into a normative error—with much more serious consequences for accountability (such as a criminal trial).