Why people ghost: 5 motivations the evidence measured

Five real reasons people stop replying without warning, what the evidence says about each (Freedman 2018, 2022), and one concrete step to start letting go.

15 min read
A phone screen with conversations half-closed, several messaging apps open at once, the feeling of a silent decision

26% of adults in recent relationships reported being ghosted in the last two years, per Freedman, Powell and Le (2018). It excludes people who left dating apps or drifted apart months ago; it counts only those who replied once and then stopped. If you are reading this because it happened to you, you are not in a minority: you are in the silent majority that does not talk about it.

If someone disappeared without warning, this is the other half of what you need. The first is the definition (what it is, what it is not, what separates it from a slow fade); this one is the why — the five motivations that research documented as the most frequent, what activates them, and what to do with each one from your side. It is not cheap empathy.

1. The data that unsettles — what happens when someone stops replying

The first data point that unsettles is the number: 26%. It is one in four. not a generational fashion, not a problem with dating apps. It is a relational pattern documented in large samples that cuts across ages, contexts and platforms. The second data point is that the frequency on the receiving side (you) and on the doing side (her) do not match. Most people who report being ghosted do not report having done it.

The third data point is the most useful here. Freedman, Powell and Le (2018, DOI: 10.1177/0265407517748791) found that people with what they call implicit destiny theories of relationships — the idea that partners are “made” for each other — ghost more than people with growth theories. The reason is counterintuitive: if relationships are destined, a closing conversation means admitting this one was not — which hurts more than leaving without a word. The easiest exit for the person doing it is the one that has no conversation.

This excuses nothing. It explains the mechanism. And understanding the mechanism changes the question your head is running.

2. The first motivation: not having the closing tool

The most frequent motivation — and the most misunderstood — is the absence of the closing tool. There are people who literally do not know how to say “I do not want to continue.” It is not malice. It is a learned skill, not a character trait, and many people never practised it.

What happens when someone with that absence has to end a relationship is: either they close clumsily (a long, confused, contradictory message, full of “it is not you, it is me”), or they do not close at all. The second option, which seems easier in the moment, is ghosting. The person knows she does not want to continue, knows she has to say something, but does not have the how — and chooses the silence.

What the research documented is that this modality is associated with low prior relational experience and with families where the closing was also not practised. If in someone’s home separations happened without conversation — parents who left, siblings who stopped talking without explanation — that person learned that this is how you close. They do not do it out of malice. They do it by imitation.

The approximate distribution of the motivations reported by the people who admitted to having ghosted at least once (Freedman et al. 2018):

████░░░░░░
███░░░░░░░
██░░░░░░░░
█░░░░░░░░░
░░░░░░░░░░

3. The second motivation: fear of confrontation

The second most documented motivation is more active: it is not an absence of tool, it is fear of using it. The person knows what to say, but saying it implies an uncomfortable conversation, a possible reply from you, a possible scene, a possible frustration. The calculation she makes — not always conscious — is: the cost of the conversation is higher than the cost of silence. And she chooses silence.

What the evidence shows is that this motivation is more frequent in people with avoidant attachment styles, and in people with low tolerance for interpersonal frustration. It is not a pattern that activates in all relationships — it activates in the ones the person perceives as “of higher emotional cost.”

The problem with this motivation, from your side, is that the other person’s fear of confrontation leaves you without closing, without warning, and without data. And the confusion it produces is disproportionate to the motivation. If the person who left you on seen did it out of fear, not out of cold decision, that does not change your pain — but it changes the kind of exit that makes sense.

4. The third motivation: the perceived cost of telling you no

The third motivation is less sympathetic but more common than it seems. There are people who weigh the cost of the closing conversation — your possible reaction, your possible insistence, your possible request for explanations — and conclude it is less costly to cut without saying anything. It is not active malice. It is a pragmatic calculation. The person prioritises their own emotional cost over your right to closing.

This motivation is the hardest one to accept, because it appears as selfishness. And in part it is. But the evidence describes it differently: it is a person who is optimising for their own wellbeing in the short term, without considering (or being able to consider) the cost they leave you with.

The reason it matters to understand it is that it frees you from personalising the behaviour. If the person made a calculation where your closing did not enter as a variable, it is not that you are unworthy of that closing — it is that it did not enter. That distinction does not remove the pain, but it changes the internal operation you have to do to let it go.

5. The fourth motivation: low emotional investment

The fourth motivation is the simplest, and the hardest to accept when one wants “what we had” to have meant something. There are people who, looking back at the relationship, conclude they did not invest enough to justify a closing conversation. It is not that what they had was not real — it is that it did not reach, for them, the “real” threshold that requires a closing.

This motivation is associated with short relationships (less than three months) and with relationships where one of the parties (her) had multiple parallel connections. In those contexts, the emotional energy to close is not justified — the person rationalises that “it was not anything serious” and applies that yardstick to the decision to cut. What she does not see is that the yardstick of “what was serious” is not the same for both sides.

For the one receiving it, this motivation is the most hurtful, because it puts into doubt something that for him was significant. The evidence does not have an easy exit for this. What it does say is that the asymmetry of investment is not evidence of anything wrong in the one who invested more — it is information about the other person, not about you.

What makes this motivation harder to process is that it puts into doubt the yardstick of “what it was.” If for her it was nothing serious, the conversation you need to close will not exist, because there is nothing she considers needs closing. The asymmetry of yardstick is the specific wound: you invested more, she less, and the closing is paid by the one who invested more.

One subtype deserves a name: the person who invested emotionally at first, then withdrew in the final weeks. There the low investment was not constant but a progressive decision: desire became habit, habit became silence. For the receiver, this subtype is especially confusing, because the early messages did look like someone who invested. The exit is the same: her yardstick is not yours; let go from that asymmetry.

6. The fifth motivation: direct malice

The fifth motivation — and the least frequent, according to the data — is direct malice. There are people who use ghosting as a tool of punishment, control or humiliation. They do it on purpose, knowing the silence harms, and choose it because they want to harm.

This motivation is associated with personality profiles with manipulative traits, and with relationships where there was explicit conflict before the disappearance. If what you had was a fight, a tense breakup and then silence, and the person has a known pattern of doing this in other relationships, this motivation enters the conversation.

The direct malice motivation changes the exit: it is not waiting for a closing, because it will not come. It is taking distance and, if the relationship has other accumulated costs, talking with someone trained in controlling relationships. The tools page has a bridge to a program that works this specific picture.

7. What the five motivations have in common

The five motivations share a common structure: in all of them, the person made a decision about your closing without including you in the decision. That is the constant. What changes is the motivation — the missing tool, the fear, the calculation, the low investment, the malice — but the result is the same: you without closing.

That common structure is what has to operate in your head, not the specific motivation. If you stay trying to diagnose which of the five it was, you will enter a rumination with no exit — because the answer is not hers, and even if it were, she will not give it to you.

What the evidence of Freedman, Powell and Le (2022, DOI: 10.1080/00224545.2022.2081528) measured as effect on the one receiving it — confusion, self-questioning, sense of invalidity — was the same across the five motivations. What changes is not the pain. It is the internal operation each one requires. The next section works that operation.

The common operation to the five motivations is learning to look at the other person’s behaviour without that look turning into judgement about you. That operation is not natural. The head that is ruminating prefers the condemnation to the information, because condemnation closes the loop and information leaves it open.

The way out is to realise that the open loop, when sustained with real information and not with stories, also closes. Only it closes slower. The minimal practice of the phrase tree works that exact distinction.

8. What the motivation does not tell you about you

It is worth saying this explicitly: none of the five motivations says anything about you. The motivation says something about the person who did it, her history, her tools, her calculation, her level of investment, her possible pattern. It says nothing about your value, your ability, what you deserve or what you lack.

The confusion between “what I did” and “what she did” is one of the most frequent mistakes for the one living it. And it is the one that feeds the self-evaluation rumination the evidence measured. The motivation does not involve you. The motivation is information about another.

If this distinction is hard to hold in the head — and it usually is — the facts exercise that is in the tools page helps to separate. Neutral writing separates the fact from the story. The story always involves you; the fact does not.

9. Five signals to know what type of person she was

It helps to have an idea — not a diagnosis — of the type of person who did what she did to you. Not to classify, but to choose well the exit.

MotivationWhat made it visibleWhat it suggests for you
No toolRepeated pattern in other relationshipsWaiting for the closing will not work
Fear of confrontationThere were prior fights and later silencesA brief message from you may or may not work
Cost perceived as highShe knows you well and knows you insistSending another message raises the cost, does not lower it
Low investmentShort connection, parallel bondsDo not personalise the closing, let go from the asymmetry
Direct maliceThere was explicit conflict and a pattern in other relationshipsDistance and, if more, professional accompaniment

None of the five signals is proof of anything on its own. They are data that, looked at together, help decide what to expect — and what not to expect — from the other side.

9.1 · What the signals do not say

The five signals serve to classify, not to predict. Knowing which motivation it was does not give you power over the outcome. It does not bring her back. It does not change what she did.

The most frequent trap when using a table like this is to treat the signals as proof of what you already suspect. If you want the motivation to be malice, you will read the signals as proof of malice. If you want it to be fear, you will read them as proof of fear.

10. Exercise: the “what I tell myself” tree — a five-minute minimal step

It is not therapy. It is a five-minute step you can do tonight. Numbered steps, in order. Do it with a piece of paper next to you and with the decision not to send anything to anyone when you finish.

  1. Write the phrase you have said to yourself most in the last few weeks. Literally, the phrase. "I must have done something wrong." "If only I had written differently." "What does she have that I do not." No filter.
  2. Mark whether that phrase is about her or about you. The phrases that are about her are of the style "she did not tell me anything because…". The phrases that are about you are of the style "I am…". Most phrases are of the second type, even though they seem to be of the first.
  3. For each phrase that is about you, write down the fact that contradicts it. If the phrase is "I am intense", the fact that contradicts it is "during four months she replied every day". If the phrase is "she never cared", the fact that contradicts it is "on Tuesday we had dinner".
  4. Read the list of facts as if it were from someone else. What sounds like interpretation, discard it. What sounds like fact, keep it.
  5. Fold the paper. This is your backup list for the next few days. Your head will go back to the phrases. When it does, look at the facts.
  6. If in two weeks your head keeps going to the phrases, the next step is to talk with someone trained. It is not exaggerating — it is what the evidence says when the tree does not dismantle on its own.

The phrase tree has an advantage that is not obvious: it separates what the head confuses. The head that is ruminating treats “what she did” and “what I am” as if they were the same problem. They are not. One is data. The other is interpretation. Writing separates them.

10.1 · How you notice the tree is starting to loosen

The first movement is that the phrases stop appearing alone. They begin to appear only when something triggers them (a song, a place, an anniversary), and they dissipate faster. The second is that the facts begin to weigh more than the phrases. The third is that you stop needing the paper: what you wrote is already integrated.

11. The difference between understanding the motivation and personalising it

Understanding the motivation is a clinical operation. Personalising it is an emotional operation. The two look alike, but the first frees you and the second locks you in.

Understanding the motivation of someone who ghosted is to take the behaviour as data: “this person prioritised their emotional cost over my closing.” Personalising it is to take the behaviour as verdict: “this person did not give me closing because I did not deserve it.” The first uses the motivation as information.

The trap is that personalisation feels more logical. The head that is ruminating prefers condemnation to uncertainty, because condemnation, although it hurts, is at least an answer. Information, on the other hand, requires continuing to process. That is why the head chooses badly at this point.

What the evidence showed is that people who personalised the motivation reported more pain at 8 weeks than those who understood it as information. Personalisation becomes chronic. Information processes. The choice between the two is yours, and is trained with exercises like the phrase tree.

There is a test the evidence recommends and that works: if your phrase from the tree says “I am intense” and the fact that contradicts it is “she replied every day for four months”, then the phrase is not a fact. It is a story your head is telling you.

If after the tree you still doubt, the safest thing is to go back to the main article of the hub and read it — because the distinction between understanding and personalising holds better when read with the full definition next to it.

12. What comes after understanding the motivation

Understanding the motivation is necessary, but not sufficient. What comes next is the internal operation of letting go — which is different from the operation of understanding. Understanding requires information. Letting go requires practice.

The minimal practice is the phrase tree that appears above. The intermediate practice is talking with someone you trust about what happened — not to ask for advice, but so the story stops being only in your head.

None of the three is instantaneous. All three are sustained over time, not in the moment. If you expect understanding the motivation to free you today, you will be frustrated.

There is a test the head can use to know whether it is processing or ruminating: if the phrase you say to yourself has changed, you are processing; if the phrase is the same after three weeks, you are ruminating.

The difference between the two is not in the amount of time, but in the movement. Rumination is static: the head returns to the same point.

"What if the motivation is malice and I do not want to admit it?"

It is a valid question. The way to know is not analysis — it is pattern.

"What if she did not know what she was doing?"

It can happen, especially without tools: some people cut without realising what they left behind.

"Why did she not tell me anything instead of doing it?"

Because the easiest behaviour is always the one that does not require conversation. The person who ghosted optimised for her own cost, not for yours.

"Does the person who ghosted ever come back?"

Sometimes: the longitudinal study shows 12-18% of ghosters send some contact message weeks or months later. Most are brief, do not seek to resume, and rarely lead anywhere.

13. Caja de crisis — if you are in crisis NOW

If right now your head overwhelms you, you feel you cannot keep looking for the motivation, or thoughts of harming yourself appear:

CountryPhoneType
Colombia123Emergency line (free, 24/7)
Colombia106Psychosocial orientation line (free, 24/7)
United States988Suicide & Crisis Lifeline (free, 24/7)
United Kingdom116 123Samaritans (free, 24/7)
Mexico55 5259 8121SAPTEL (free, 24/7)
Argentina135Centro de Asistencia al Suicida
Spain024Teléfono de la Esperanza
Internationalfindahelpline.comGlobal directory

If you are in immediate danger, call your country’s emergency number (911, 112, 066, etc.)

Ask for professional help

Via rdkterapia · No spam · Unsubscribe whenever you want

Frequently Asked Questions

Why does someone ghost instead of saying they do not want to continue?

Because they do not always have the tool to do it. Freedman, Powell and Le (2018) found that people with implicit destiny theories of relationships (believing that partners are 'made' or 'not made' for each other) tend more towards ghosting — because for them, a closing conversation would mean admitting the relationship was not destined. The easiest exit is the one that has no conversation.

Is ghosting always malice?

No. The evidence describes it as a behaviour with five main motivations, and several of them are not malice — they are lack of tools, fear of confrontation, or exhaustion. What it is, in all cases, is an active decision to cut communication knowing the other person is waiting for a reply. The motivation does not erase the effect, but it changes the way out for the person receiving it.

Are there people who ghost more often?

Yes. Freedman et al. (2018) identified that the pattern is more frequent in people with avoidant attachment and in people with low situational empathy. It is not a closed personality trait — it is a pattern that activates in specific contexts (short relationship, fear of conflict, low perceived investment).

Can you stop ghosting?

Yes, but not with willpower. The documented way out is to learn the alternative behaviour: the brief closing conversation. People who report having stopped doing it usually describe a specific moment when they learned that tolerating an uncomfortable conversation is less costly in the medium term than cutting without saying anything.

What if the person who ghosted me was not like that before?

It can happen. Ghosting is a situational behaviour, not necessarily a trait. If the person did it once in a context of exhaustion or fear, it does not mean they will do it again. The useful question for you is not 'what kind of person is she', but 'what is good for me to do with what I have today'.