How To Welcome Someone Back After They Have Caused Harm

The history of how communities handle wrongdoing is mostly a history of choosing between two inadequate options, cycling between them, and wondering why things don't get better.

Option one is punitive exile — the offender is cast out, shunned, fired, excommunicated. The harm is named, the person is removed, and the community reconstitutes itself around the absence. This feels like justice because it produces consequences. But it accomplishes four things, all of them inadequate:

1. It protects the community's sense of moral cleanliness rather than actually repairing the harm 2. It removes the person from accountability rather than holding them within it 3. It leaves the conditions that enabled the harm entirely intact 4. It transmits the unprocessed person and their pattern to another community, which will eventually face the same problem

Option two is suppressive forgiveness — the harm is minimized, the person is quickly absolved, the community moves to "putting this behind us." This is sometimes dressed up in the language of grace and healing. It's actually a negotiation strategy, and the people who were harmed are the ones who pay the price. They are being asked to sacrifice their need for acknowledgment on the altar of the community's desire to feel like good people again.

Neither of these options is new. Neither of them works at scale. And yet these are the two tools most communities reach for, again and again, because they're fast and they're familiar.

The third option — genuine reintegration — has been practiced most consistently in Indigenous restorative justice traditions, in religious communities that take repair seriously (as distinct from communities that use forgiveness as a silencing mechanism), and in the modern restorative justice movement that emerged partly from Māori community practice in New Zealand and has since been adopted in schools, prisons, workplaces, and neighborhoods worldwide. The evidence on restorative processes is now substantial, and the finding is consistent: when restorative processes are implemented well, they produce higher victim satisfaction, lower recidivism, and stronger community cohesion than punitive alternatives.

The problem is that implementing them well is demanding. And communities mostly don't.

Before describing what genuine reintegration requires, it's worth being precise about what it is not, because these confusions destroy the process.

Reintegration is not premature forgiveness. Forgiveness is a personal act undertaken by the person who was harmed. It is not something a community can grant on their behalf. It is not something with a deadline. Rushing anyone toward it is a form of secondary harm. Reintegration can succeed even without the harmed person achieving forgiveness — these are separate processes and must not be conflated.

Reintegration is not performed redemption. The person who caused harm announcing their growth, posting about their journey, giving speeches about what they've learned — none of that is reintegration. Reintegration is evidenced by changed behavior over time, in the same contexts where harm occurred, visible to the same people who were affected. The announcement is not the thing. The thing is the thing.

Reintegration is not mandatory for the harmed person. If the person who was harmed does not want to be in community with the person who harmed them, that preference is not an obstacle to the community's process — it is a condition the community must work around. Reintegration cannot require the harmed person to disappear or absorb ongoing contact they haven't consented to.

Reintegration is not erasure of the record. What happened, happened. Reintegration does not require the community to stop knowing it, stop acknowledging it, or pretend the person has been cleansed of it. The past is part of the person's history in the community. What reintegration offers is: that history doesn't have to be the only thing.

Research on restorative practice — and centuries of Indigenous community law — converge on the same three demands. They must all be present, and they must all be genuine.

1. Full and accountable acknowledgment

This is harder than it sounds, because most acknowledgments are not acknowledgments — they are deflections in acknowledgment's clothing.

"I'm sorry you felt hurt" is not an acknowledgment. It locates the problem in the other person's feeling rather than in what happened.

"I made a mistake, but I was under a lot of pressure" is not an acknowledgment. It's an explanation that functions as a partial retraction of the apology.

"I know what I did was wrong, and I take full responsibility" followed by a detailed accounting of the mitigating circumstances is not an acknowledgment. It is a performative acknowledgment followed by erosion of that acknowledgment.

A genuine acknowledgment names what happened specifically, locates the harm in the actor's behavior rather than the victim's response, and makes no moves to diminish it. It is brief. It does not try to be comprehensive or therapeutic for the person making it. It is not primarily about the speaker's feelings.

Jun Tangney's research on shame and guilt is directly relevant here: genuine acknowledgment requires a person to sit in guilt — the behavioral kind, focused on what they did — rather than shame, which collapses into who they are and produces either self-destruction or defensiveness. A person processing shame cannot make a genuine acknowledgment because they are too busy managing their own survival to stay focused on the reality of the other person's experience.

This means that before a community expects acknowledgment, it needs to create conditions where the person who caused harm is not being publicly destroyed. Shame produces either flight or performance. Neither produces accountability.

2. Concrete repair

After acknowledgment comes repair — and repair must be proportionate, specific, and real.

"What can I do to make this right?" is the right question. The wrong answer is letting the community decide what counts, because the community will tend toward either punitive excess or inadequate minimization. The people who were harmed should have primary input into what repair looks like.

Repair takes many forms: financial restitution where material damage occurred; labor and service where capacity was the resource taken; structural change where institutional harm occurred; social repair — the work of rebuilding relationship and trust, if that's what the harmed person wants — where the damage was primarily relational.

One of the most neglected forms of repair is changing the conditions that enabled the harm. If someone in a position of power abused that power, repair is not just apologizing to the specific person harmed — it's changing the power structure so it cannot happen again. If someone's harm was enabled by community silence or complicity, repair includes the community examining how it participated.

Without this level of repair, the community has acknowledged the harm but hasn't actually done anything about it. Which means the conditions for harm remain.

3. Changed behavior, sustained and visible

This is the long game, and most communities never get here because they declare the process finished too early.

The person who caused harm must be willing to demonstrate — not promise — that the behavior has changed. This means:

- Remaining in relationship with the community (not fleeing to a fresh start elsewhere) long enough for the change to be observable - Accepting that trust is rebuilt incrementally, not through a single dramatic gesture - Tolerating the discomfort of being known as someone who caused harm, rather than demanding that this be forgotten - In some cases, actively supporting whatever monitoring, mentorship, or accountability structure the community has established

James Gilligan's work on violence and transformation suggests that behavioral change without identity change is fragile. The person must not only change what they do but who they understand themselves to be — specifically, the part of their self-understanding that made the harm possible. This is deep work, not public performance. It happens in therapy, in mentorship, in the long private reckoning that communities rarely have visibility into.

Communities can support this by providing the time and relationship for that reckoning to occur, rather than demanding the visible outputs while the internal work is still happening.

The community is not a passive backdrop for the harm-doer's redemption arc. The community has its own work to do.

1. Clear and sustained accountability

The community must be capable of naming what happened, accurately and without minimization, for as long as is necessary — which is longer than feels comfortable. This is not about punishing the person who caused harm with permanent naming. It's about the community not reaching for the relief of forgetting too soon.

Premature forgetting is almost always primarily for the community's benefit. The community wants to feel like good people again. The community wants to return to the narrative of being a safe place where harm doesn't happen. Sustained accountability refuses that relief until it's been earned — until the repair is real and the behavior has actually changed.

For communities that have a strong commitment to positive identity (religious communities, progressive organizations, tightly-knit professional groups), this is acutely difficult. Sustained acknowledgment of harm disrupts the narrative. But the alternative — the community's narrative-preservation at the expense of the harmed person's reality — is a form of institutional gaslighting.

2. Genuine belief in the person's capacity to change

This is what distinguishes reintegration from exile with a waiting period. The community must actually believe — not pretend to believe, actually believe — that the person who caused harm is capable of becoming different.

Without this belief, everything that looks like reintegration is actually probation. The community is waiting for the person to fail, holding the exile in reserve. The person can feel this, which produces either performance (trying to pass the test) or resentment (refusing to submit to surveillance that feels terminal). Neither produces genuine change.

This does not mean naive belief. It means the kind of belief that can coexist with accountability — the belief that says: what you did was real, and I think you can be different, and these two things are not in conflict.

For communities traumatized by the harm, this belief may need to be carried by specific members while others remain in more protective stances. That's legitimate. The community does not need unanimous belief. It needs enough people holding the possibility that the possibility doesn't collapse.

3. Structural examination

Every individual harm occurs within a context. The person who caused harm made choices — but those choices were shaped by the structures, norms, hierarchies, and silences of the community itself.

Genuine reintegration requires the community to ask: how did we participate in the conditions that made this possible? Was there a power imbalance we knew about and didn't address? Was there a culture of silence around this type of behavior? Were there people who saw warning signs and said nothing?

This examination is uncomfortable because it distributes responsibility beyond the identified wrongdoer, and communities tend to resist this. But without it, the community has externalized the problem entirely onto one person, and the structural conditions remain available to produce the next harm.

The structural examination is not about excusing the person who caused harm by pointing to systemic causes. It's about the community doing its own accountability in parallel with the individual's.

This section is short because the answer is simple: nothing is required from the person who was harmed.

They do not owe participation in the process. They do not owe the person who harmed them their presence, their story, or their emotional labor. They do not owe the community a forgiveness timeline. They do not owe anyone a demonstration of their healing.

What they deserve: a community that names what happened accurately, that does not pressure them to minimize it or perform recovery, that gives them real choice about their involvement in any reintegration process, and that protects their interests when those interests might conflict with the community's desire for smooth resolution.

In restorative justice practice, this is called centering the harmed. It means that the process is not structured around what's best for the person who caused harm, or around what's best for the community's functioning, but around what actually addresses the harm experienced by the person who experienced it.

This is violated constantly in practice. The harmed person is often the one who ends up leaving, who is pressured to be quiet about their experience, who is told that their ongoing difficulty is now the problem. When this happens, the community has not engaged in reintegration. It has engaged in harm management — managing the person who was harmed rather than the harm itself.

Reintegration takes longer than communities want it to.

The social pressure to "move on" accumulates quickly. People who were not directly harmed want to return to normal. The person who caused harm wants the process to be finished. The community's functioning is interrupted by the ongoing presence of the unresolved situation.

All of this pressure compresses the timeline in ways that undermine the process. The acknowledgment is rushed so the community can get to the repair. The repair is declared complete before it's actually addressed the harm. The changed behavior is asserted rather than demonstrated. The community calls it done.

And then, sometimes years later, the same harm happens again. Or the harmed person leaves, quietly, and the community never understands why. Or the person who caused harm causes harm in the same way in a new community, because they were never actually required to change.

The timeline is set by the people who were harmed and by the evidence of genuine change — not by the community's discomfort with unresolution. Communities that can tolerate that discomfort — that can hold "this is not finished yet" without collapsing into either denial or exile — are communities capable of actual repair.

A community that has successfully navigated genuine reintegration — slowly, imperfectly, with all the discomfort involved — has learned something about itself.

It has learned that it can hold the complexity of human fallibility without destroying everyone inside the category of "person who did something wrong." That is not a small thing, because that category eventually includes almost everyone.

It has learned that accountability and belonging are not opposites. That you can be held responsible and still be held. That the relationship survives the reckoning.

And it has built something in its culture: a collective understanding that when you cause harm here, you will have to face it fully — and that if you face it fully, you don't have to lose everything. That combination — accountability with possibility — is what makes honesty affordable. And in a community where honesty is affordable, harm gets named sooner. Patterns get interrupted earlier. The conditions that produce harm get examined rather than ignored.

This is the community analog to the individual work of guilt versus shame. The community that can hold accountable without destroying is a community that, like the individual who can sit in guilt without drowning in shame, can stay in the room with hard things and remain functional.

A community full of people who can do this — who can cause harm, face it, repair it, change, and remain — is a community capable of solving the things that are actually hard. Resource allocation. Historical injustice. Conflicting needs. The things that most communities never actually deal with because they cannot tolerate the discomfort of sustained difficult conversation.

If every community on the planet developed this capacity, the quality of governance would transform. Conflict between groups would not have to end in exile or annihilation — there would be a third path. The people who have the most power to cause harm would face real accountability. And the people who have been harmed throughout history would have something they almost never receive: an honest reckoning.

That is not a small vision for a hard process. But the hard process is what makes the vision possible.

---

- Zehr, H. (2002). The Little Book of Restorative Justice. Good Books. - Braithwaite, J. (2002). Restorative Justice and Responsive Regulation. Oxford University Press. - Maxwell, G. & Morris, A. (1993). Family, Victims and Culture: Youth Justice in New Zealand. Victoria University Press. - Tangney, J.P. & Dearing, R.L. (2002). Shame and Guilt. Guilford Press. - Gilligan, J. (1996). Violence: Our Deadliest Epidemic and Its Causes. Grossman. - Herman, J. (1997). Trauma and Recovery. Basic Books. - Umbreit, M.S. et al. (2006). "Restorative Justice in the 21st Century." Marquette Law Review, 89(2), 251-304. - Pranis, K. (2005). The Little Book of Circle Processes. Good Books. - Ross, R. (1996). Returning to the Teachings: Exploring Aboriginal Justice. Penguin Canada. - Lederach, J.P. (2005). The Moral Imagination: The Art and Soul of Building Peace. Oxford University Press.