RESEARCH_FILE
The Science of Friendship Breakups
There is no ritual for the end of a friendship — no breakup talk, no paperwork, usually not even a fight.
SEE THE PRACTICE
Turn a thought this research explains into one clear move.
THE THOUGHT
“I must have done something wrong”
YOUR RECORDED RESPONSE
“Their silence is information. I do not need to turn it into a sentence about me.”
ONE PRIVATE MOVE
Open a blank note and write three short lines: what you know, what you are guessing, and one thing you will not keep checking tonight. Save it and put your phone face down for five minutes.
The research says that's the norm, not the exception: friendships mostly dissolve by drift and circumstance, about half of a personal network is replaced within seven years, and only about half of the friendships people name are even reciprocated. Yet the loss is real — grief researchers call it disenfranchised grief, mourning that gets no social permission. Here's what forty years of dissolution studies actually found.
How the science changed
- 1984
Suzanna Rose publishes 'How Friendships End', one of the first systematic studies of friendship dissolution: young adults' close friendships ended mainly through four routes — physical separation, replacement by new friends, growing dislike, and interference from dating or marriage — mostly circumstance, not confrontation. ↗
- 1989
Kenneth Doka names 'disenfranchised grief': the grief people feel after losses that society does not openly acknowledge, publicly mourn, or socially support. Losing a living friend — a relationship with no formal status — is a textbook case of a loss that gets no funeral and no condolences. ↗
- 2014
Mollenhorst, Völker and Flap publish a seven-year Dutch panel study in Social Networks: network size stayed stable, but the people in it changed massively — about seventy percent of confidants and practical helpers named in the first wave no longer held that position seven years later. Turnover is the baseline, not a failure. ↗
- 2016
Almaatouq and colleagues test friendship reciprocity in PLOS ONE: while the vast majority of people expected their named friendships to be mutual, only about half actually were — people are systematically poor at perceiving whether a friendship goes both ways. ↗
- 2021
Khullar, Kirmayer and Dirks map how emerging adults' friendships actually end: dissolution was overwhelmingly a gradual fading — declining contact and closeness — rather than an explicit break, and often went unannounced, leaving people to infer the ending from silence. ↗
- 2025
Telari, Pancani and Riva run multi-day experiments comparing ghosting with explicit rejection in Computers in Human Behavior: ghosting produced a slower, more prolonged negative response — the unresolved uncertainty hindered coping, so the silence hurt longer than the clear 'no'. ↗
What people believe vs. what the data shows
The belief“Real friendships last forever — if one ended, it was never real.”
The dataIn a seven-year panel study, personal networks kept their size but swapped much of their membership: about seventy percent of the confidants and helpers named at the start no longer held that position seven years later. Turnover is the documented norm for real, functioning friendships. ↗
The belief“If a friendship ended, someone must have done something wrong.”
The dataThe earliest dissolution research already found that close friendships end mainly through circumstance — physical separation, replacement by new friends, and interference from dating or marriage — with growing dislike only one of four routes. Most endings have no villain. ↗
The belief“It was just a friend — it's not a real loss you're allowed to grieve.”
The dataGrief research classifies exactly this as disenfranchised grief: a loss that is not openly acknowledged, publicly mourned, or socially supported. The lack of a ritual and of social permission doesn't make the grief smaller — it removes the usual supports for processing it. ↗
The belief“Ghosting is the kind way out — a clear 'this is over' would hurt them more.”
The dataMulti-day experiments comparing ghosting with explicit rejection found the opposite: ghosting elicited a slower, more prolonged negative response, because the unresolved uncertainty hinders coping. The clear ending hurt sharper but resolved faster. ↗
The belief“If I feel this close to them, they must feel the same about me.”
The dataIn reciprocity studies, the vast majority of people expected their named friendships to be mutual — but only about half actually were. Perceived closeness is a poor guide to whether the tie runs both ways, which is one reason quiet endings blindside people. ↗
TEST_YOURSELF · How well do you know this science?
01 According to dissolution research, how do most adult friendships actually end?
In Khullar, Kirmayer and Dirks' study of emerging adults, friendship dissolution was overwhelmingly a gradual fading — declining contact and closeness — rather than an explicit break, consistent with Rose's 1984 finding that circumstance, not confrontation, drives most endings. source ↗
02 In the seven-year Dutch panel study of personal networks (Mollenhorst et al.), what happened to people's networks over time?
Re-interviewing participants seven years apart, Mollenhorst and colleagues found network size was roughly unchanged while membership turned over heavily — only about half of network members remained, and about seventy percent of confidants and practical helpers no longer held that role. source ↗
03 What did multi-day experiments (Telari, Pancani & Riva, 2025) find when comparing ghosting with explicit rejection?
Trend analyses showed ghosting elicited a slower and more prolonged negative response than explicit rejection: the unresolved uncertainty and lack of closure appeared to hinder coping, prolonging the distress. source ↗
04 What does 'disenfranchised grief' (Doka, 1989) mean?
Doka defined disenfranchised grief as grief experienced after a loss that is not or cannot be openly acknowledged, publicly mourned, or socially supported — with losses of relationships that lack formal status, like a friendship, among his core examples. source ↗
05 In Almaatouq and colleagues' 2016 reciprocity study, what share of the friendships people named turned out to be mutual?
Only about half (53%) of the friendship ties examined were reciprocal, even though the vast majority of participants expected them to be — showing that people perceive the directionality of their friendships poorly. source ↗