Skip to content
MyFreud

Passive Aggression: Anger That Cannot Be Said Directly

Passive aggression is not a personality type and it stopped being a diagnosis in 2013. It is what anger does when saying it out loud feels unavailable.

6 min read

Pop-art illustration of two people sitting at opposite ends of a sofa, one looking down at a tablet, the other looking away with their hands in their lap.

Key takeaways

  • Passive aggression is a behaviour, not a personality. It appeared as a diagnosis in earlier manuals, was moved to an appendix for further study in 1994, and was dropped from the DSM-5 entirely in 2013.
  • It is what anger does when direct expression feels unavailable. The usual reason is a history in which saying you were angry produced something worse, so the anger goes out sideways instead.
  • The recognisable shape in a relationship is one person pressing and the other withdrawing. A meta-analysis of 74 studies covering more than 14,000 people found that pattern is reliably associated with worse relationship outcomes.
  • Both positions feel like the reasonable one from inside. The person pressing feels ignored, the person withdrawing feels attacked, and each is responding to the other rather than starting it.
  • The exit is not learning to express anger more forcefully. It is saying the ordinary thing early, at a volume that does not require a build-up first.

Passive aggression is anger going out sideways because the direct route feels closed. The agreement that is never acted on, the “fine” that is not fine, the job done just badly enough to make a point.

It is worth being clear at the start that this is a behaviour rather than a type of person, and that the difference changes what you can do about it.

It stopped being a diagnosis

Passive-aggressive personality disorder was a diagnosable condition in earlier editions of the American manual. In 1994 it was renamed negativistic personality disorder and moved to an appendix for further study, partly because clinicians could not reliably separate it from chronic depression and dependent personality. In 2013 it was removed from the DSM-5 altogether.

That history matters for a practical reason. “He is passive-aggressive” gets used as a description of somebody’s character, as though it were a fixed trait they carry between situations. It is not, and the people who do it most tend to do it in specific relationships rather than in all of them, which is a strong hint that it is a response to a situation rather than an attribute of a person.

Where it comes from

The short version: direct anger was not available.

For most people the pattern traces to somewhere that saying “I am angry with you” reliably made things worse. A parent who escalated. A parent who went cold for two days. A household where the person who raised the objection became the problem. In that environment, indirect expression is not a character flaw, it is the strategy that worked.

It also has a real payoff now. You register the objection, you keep deniability, and no confrontation happens today. The bill arrives later, in the form of a grievance nobody ever addressed because nobody was ever allowed to hear it stated.

The thing being avoided is usually not the anger. It is the conversation.

The shape it takes with another person

Passive aggression rarely happens alone. In a relationship it settles into a recognisable pattern, and this is the part with actual evidence behind it.

One person presses for a response. The other withdraws. The pressing intensifies because nothing is coming back, the withdrawal deepens because the pressing feels like an attack, and both of them are now responding to each other rather than to whatever started it.

A meta-analysis of 74 studies covering more than 14,000 people found this pattern reliably associated with worse outcomes, including lower relationship satisfaction and poorer communication. [schrodt-2014-demand-withdraw]

How the same conversation looks from each side Illustrative
0 25 50 75 100 How strongly felt 88 Feels ignored 84 Feels reasonable 76 Thinks they started it 30 Believes they are pursuing
0 25 50 75 100 How strongly felt 86 Feels attacked 84 Feels reasonable 78 Thinks they started it 22 Believes they are punishing

A schematic of the two positions described in this article and in the research cited on demand and withdraw. Not measured data.

The two middle bars are the difficult finding. Both people feel reasonable and both believe the other one began it, which is why the argument about who started it never concludes.

The bottom bar is the one worth sitting with. Almost nobody withdrawing experiences themselves as punishing anyone. From inside it feels like self-protection, or like the only way to end a conversation that is going nowhere.

The forms it actually takes

  • Agreeing, then not doing it. The commonest by a distance, because it avoids the disagreement entirely.
  • Doing it, badly. Late, half-finished, or technically complete in a way that makes the point.
  • “Fine.” Delivered in a tone that makes clear it is not, which is what makes it deniable and what makes it work.
  • The joke. True content, comic frame. If it lands badly, it was only a joke.
  • Going quiet. Not stated as a decision, so it cannot be discussed, and lasting until somebody else repairs it.
  • The oblique complaint. Raised as a general observation, or about somebody else, so it cannot be answered directly.
  • Forgetting. Genuinely forgetting the things that matter to one particular person, and only those things.

Where does yours go?

Tick anything you recognise in yourself. This is a reflection prompt rather than a test, and it produces no diagnosis.

0 of 8 ticked

No screener on this site measures anger, which is a genuine gap rather than a decision. The hub lists the seven we do cover, and the anxiety and depression screeners are often the relevant ones, because irritability is a common feature of both.

What actually changes it

Say the ordinary version early. Almost every passive-aggressive act is a large statement made too late, in disguise. The alternative is a small statement made on time and in plain words: “I do not want to do that”, “that annoyed me”. It feels disproportionate to say something that minor. Saying it while it is still minor is the entire mechanism.

Notice the yes you do not mean, at the moment you say it. That is the single highest-yield intervention available, because it is the point where the whole sequence starts and the only point where it is cheap to stop.

Stop trying to be understood without speaking. A great deal of this is a hope that somebody will notice and ask. They generally do not, and the hope is what converts a passing annoyance into a grievance.

Do not vent it instead. The obvious alternative to going quiet is letting it out, and the evidence does not support that route. A meta-analysis of anger management activities found that approaches which raise physiological arousal do not reduce anger, while those that lower it do. [kjaervik-2024-anger-arousal-pa] Our guide to the venting myth covers this properly.

If you are the one pressing, press less. This is counterintuitive and it is what the demand and withdraw research implies. Asking once and then genuinely stopping produces more response than asking six times, because the sixth ask is what the withdrawal is now about.

Our guide to why you are so irritable covers the case where the anger is closer to the surface and comes out directly instead.

If it is the other person

Naming it does not work. “You are being passive-aggressive” is unanswerable by design, and the reliable response is denial followed by more of it.

What works better is treating the concrete thing as the subject. Agree what will happen and by when, out loud. Treat “fine” as missing information rather than as consent, and ask a specific question instead of a general one. And if you recognise yourself as the one who keeps pressing, the research above is describing you as much as them.

What treatment involves

Anger responds to treatment better than its reputation suggests. A meta-analytic review of anger treatments found meaningful improvement across specific anger problems. [delvecchio-2004-anger-treatments] Cognitive behavioural approaches are the usual route, and for the indirect version the work tends to be less about calming down than about the belief underneath: that saying the thing directly would cost more than it does.

Our guide to anger management therapy covers what the sessions actually involve.

When to seek help

See a GP if resentment is a permanent feature of your closest relationships, if people repeatedly tell you they cannot tell what you are thinking, if you find yourself unable to say a straightforward objection out loud, or if the pattern is costing you a relationship or a job.

It is also worth raising if the irritability underneath has become constant, because persistent irritability is a common presentation of both depression and anxiety, and it is treatable.

Go urgently if you have thoughts of harming yourself or anyone else. In the UK, Samaritans is free on 116 123 at any hour.

How MyFreud can help

The complaint that comes out sideways on Thursday was usually formed on Monday, and by Thursday nobody can remember Monday. MyFreud gives you daily mood tracking that takes seconds, so the gap between when something annoyed you and when it finally came out is something you can see rather than reconstruct.

Download MyFreud and start today: App Store or Google Play.

Frequently asked questions

What is passive-aggressive behaviour?

It is expressing anger or opposition indirectly rather than saying it. The usual forms are agreeing to something and then not doing it, doing it badly or late, going quiet, answering "fine" when it is not fine, making the point as a joke, or bringing up the complaint at an angle so it cannot be answered directly. What links them is that the anger is real and communicated, but never in a form the other person can respond to without appearing to overreact.

Is passive-aggressive personality disorder a real diagnosis?

Not any more. It was a diagnosable personality disorder through earlier editions of the DSM, was renamed negativistic personality disorder and moved to an appendix for further study in 1994 partly because it was hard to distinguish from chronic depression and dependent personality, and was removed from the DSM-5 in 2013. Calling somebody passive-aggressive today describes a behaviour, not a condition they have.

Why am I passive-aggressive?

Most often because direct anger was not safe or not permitted somewhere it mattered. If saying you were angry as a child produced escalation, withdrawal of affection, or being told you were the problem, indirect expression is the strategy that worked. It also has a short-term payoff in adulthood: you register the objection while keeping deniability, so no confrontation happens today. The cost arrives later, when nothing gets resolved and the resentment accumulates.

What is the difference between passive aggression and setting a boundary?

A boundary is stated. "I am not doing that" is a boundary, even said badly. Agreeing to do it and then quietly not doing it is passive aggression. The test is whether the other person can tell what has happened and respond to it. Withdrawing on purpose and saying so is legitimate. Withdrawing while denying that anything is going on is the version that damages the relationship, because it removes the possibility of a conversation.

How do I deal with a passive-aggressive person?

Not by naming it, which reliably escalates, because the whole structure depends on deniability and the accusation cannot be answered. What works better is responding to the concrete thing rather than the tone: agree what will happen and by when, in words, and treat "fine" as insufficient information rather than as agreement. If you are the one who keeps pressing for a response, notice that pressing harder produces more withdrawal, which is the pattern the research describes.

References

  1. 1.Schrodt P, Witt PL, Shimkowski JR ( 2014). A meta-analytical review of the demand/withdraw pattern of interaction and its associations with individual, relational, and communicative outcomes. Communication Monographs, 81(1), 28-58. doi:10.1080/03637751.2013.813632
  2. 2.Del Vecchio T, O'Leary KD ( 2004). Effectiveness of anger treatments for specific anger problems: a meta-analytic review. Clinical Psychology Review. doi:10.1016/j.cpr.2003.09.006
  3. 3.Kjærvik SL, Bushman BJ ( 2024). A meta-analytic review of anger management activities that increase or decrease arousal: what fuels or douses rage?. Clinical Psychology Review. doi:10.1016/j.cpr.2024.102414
Somewhere to talk about this

Our Discord has a channel per topic and you pick the ones you want when you join, so you are not dropped into fourteen rooms at once. Reading without posting is completely fine. There are no clinicians in there, so it is other people rather than treatment.

Open the Discord