DARVO at Work: When Raising a Problem Makes You the Problem

You raise a specific problem at work. Within a day, the conversation is about your tone, your attitude or your motives, and the problem you raised has vanished. That sequence has a name in psychology: DARVO, which stands for Deny, Attack, and Reverse Victim and Offender.

This post shows how it can look at work, what the research supports, and what it does not. The studies come from interpersonal violence, not offices, so the workplace use here is an analogy. DARVO is also a pattern of response, not a diagnosis, and it is not proof that anyone is lying.

The three moves, with an example of each

Jennifer Freyd, a psychologist, introduced the term in a 1997 paper in Feminism & Psychology. On her website she describes it as a reaction that people who did wrong may show when held accountable: they deny the behavior, attack the person confronting them, and reverse the victim and offender roles. She also describes “institutional DARVO,” when an institution does the same or is complicit in it. The guide’s DARVO entry lays out the mechanics. Here are three short scenes, each a (composite scenario) and none a real person or company.

Deny. You email your manager that a vendor invoice went out at the old rate. In a one-on-one, your manager says calmly, “I never approved that.” The approval is in the thread. You feel a jolt of “am I wrong?” and go and check. Denying something documented in a flat voice is close to gaslighting, and it is the first move.

Attack. A different manager, a different problem: you raise a missed handoff. The reply is, “Honestly, this is what people mean when they say you come at things aggressively.” Your mind starts to sort through your last three emails. The handoff is no longer the topic, and your manner is.

Reverse. A colleague used your figures in their report without credit. You mention it to them. By the afternoon, they have told the team lead they feel “targeted and blindsided,” and you get a call about “how we communicate.” The person who took the figures now needs support. The person who noticed is asked to adjust.

Run as one scene, the three moves look like this.

What the research shows, and what it does not

Three studies from Freyd’s lab are worth knowing.

In a 2017 study in the Journal of Aggression, Maltreatment & Trauma, Harsey, Zurbriggen and Freyd asked 138 undergraduates to recall a time they had confronted someone over wrongdoing. DARVO was commonly reported, and the three moves tended to occur together. More exposure to DARVO went with more self-blame among the people who had confronted.

In 2020, Harsey and Freyd ran two experiments in the same journal, with 316 and 360 university students. Participants read accounts of interpersonal violence. Those who saw a DARVO response rated the victim as less believable, more responsible and more abusive, and rated the perpetrator as less abusive and less responsible. In the second experiment, people who had first learned about DARVO rated the victim as more believable and the perpetrator as less believable.

A 2026 study by Durland, Harsey and Freyd in the Journal of Trauma & Dissociation used two samples (319 online participants and 261 students). It found DARVO exposure was higher in confrontations about emotional and psychological mistreatment, and that it predicted trauma symptoms even after accounting for trauma history.

The limits matter. These were mostly student samples, the vignettes were about violence, and none was designed as a workplace study. The findings suggest that the reversal works on observers, and on the person confronting, but they do not prove it works the same way in your office.

Why it works on the room, and on you

Here is our reading of why it lands, not a research finding. Many of the people who hear the reversal were not there for the original event. They meet the story in the middle: one person who is visibly upset, and another who has calmly raised a concern. It is often easier for listeners to file that as a personality clash than to work out who did what.

It also works on you. In the 2017 study, exposure to DARVO went with more self-blame. If you leave a conversation asking whether you were the problem all along, that fits the link the 2017 study found between DARVO exposure and self-blame, and it is not proof that you were.

The tells: fast, complete and off the facts

A fair caution first. A person who is wrongly accused will deny, and may be angry. Freyd’s definition centers on someone who actually did it and takes the role of the wrongly accused, and from the outside you often cannot know. So do not judge the person. Watch the sequence. These four checks are this site’s rules of thumb, not research findings.

• Speed: the counter-story appears within the same conversation or within a day.
• Completeness: it is a total denial, not "part of that is fair."
• Drift: your original fact is never answered on its merits, and the topic becomes you.
• Audience: the first people to hear the other version are people with power over you.

Several of the guide’s named patterns often travel with it. A demand that you prove your case on the spot is burden reversal. “Just to be clear, I only passed it along,” said before anyone has accused anyone, is preemptive innocence. A claim that you did something you did not do, lodged before you raise your own concern, is a false accusation. A remark like “some people have found you difficult,” with no names, is an implied accusation. If you see two or three of these together, the picture is clearer. One alone may just be a bad day.

It is worth turning the test on yourself, too. When someone raises a concern about your work, do you answer the facts first, even if you are stung? Being accused unfairly is painful, and a fair defense is not DARVO. The difference is whether the original point still gets an honest answer. If you can say “you are right about the deadline, and here is what I think happened with the rest,” you are doing the opposite of a reversal.

How to respond without feeding it

• Write the problem down before you raise it: date, what happened, the evidence, no adjectives. The [document contemporaneously](/guide/document-contemporaneously/) entry shows how.
• Raise it in writing when you can, or follow a conversation with a neutral recap email.
• When the topic drifts, return in one calm sentence: "I would like to stay with the invoice rate." Repeat it if you need to. Do not answer each attack.
• Do not counterattack. Two upset people is the picture that helps DARVO. Calm, specific and slightly boring is the opposite. If it happens live in a meeting, you can say once, "I would rather take this offline so we can look at the documents," and write it up afterward. Moving a heated exchange out of the room stops the audience from becoming the judge.
• Bring in a neutral third party or a formal process early, before the counter-story sets.
• Describe the sequence to whoever decides, in plain facts: "I raised X on Monday. On Tuesday the discussion was about my tone." You do not need to say "this is DARVO."

The 2020 experiments hint that helping listeners understand the pattern can reduce its pull, but that was in a lab. Expect a second round at the organizational level, since Freyd notes institutions can do it too. The guide’s know the HR reality and when to exit entry helps you plan for that.

The sibling posts on how to document workplace bullying and whether to go to HR go deeper on the record and the process. If you want to see the same theme at the level of memory, is my boss gaslighting me covers it.

Being turned into the problem after you did the right thing is exhausting, and it is common to feel self-doubt afterward. That is a normal response to a hard situation. It is not evidence that you did something wrong.

Sources