The Reciprocity Gap: Half the People You Call Friends Don't Call You One

Try this. Write down the names of the people you would call your friends. Not acquaintances, not colleagues you like, friends. Now go down the list and ask a harder question about each one: if I handed them a blank sheet, would my name be on it?

Most people answer yes for nearly everyone. That confidence is the interesting part, because when researchers actually ran this exercise and checked both sheets, the answer was closer to a coin flip.

The study that nobody wants to be true

In 2016, a team led by Abdullah Almaatouq at MIT published a paper in PLOS ONE with a title that reads like a dare: Are You Your Friends' Friend? They asked 84 people in a university class to rate how close they felt to every other person in the class, on a scale from zero to five. Anything above two counted as calling that person a friend. That produced 775 friendship nominations.

Then they lined the nominations up. If A named B as a friend, did B name A?

Only 53 percent of the time. Nearly half of the friendships existed in one person's head and not the other's.

The second finding is the one I keep coming back to. The researchers also asked each person to predict how the other would rate them. In 94 percent of cases, people assumed the friendship was mutual. So the gap is not between people who know they are chasing a friendship and people who are not. It is between what almost everyone believes and what is actually there.

To check this was not a quirk of one classroom, the team re-ran the analysis on five other friendship datasets, some of them quite large. Reciprocity ranged from 34 percent to 53 percent. In no dataset did it come close to the near-universal mutuality people expected.

Why we get it so wrong

There is a lazy reading of this study, which is that people are deluded about their social lives. I do not think that is what is going on. Three more specific things are.

Friendship has no receipt. Romantic relationships get defined. Someone asks, someone says yes, there is a moment you can point to. Friendship never gets that moment. You drift into it, and because nobody ever declared it, nobody can tell when only one of you has arrived. The word "friend" is doing an enormous amount of quiet work, covering everything from "we have lunch most weeks" to "I would call you from a hospital," and two people can use it about each other while meaning completely different things.

Asking is socially impossible. You can ask a partner "what are we?" You cannot ask a friend "am I one of your friends, or am I someone you are pleasant to?" The question itself would damage the thing it is trying to measure. So we never get the data, and in the absence of data we assume the answer we prefer.

Non-reciprocal ties run uphill. This was the study's structural finding, and it is the part with real consequences. When friendship was one-directional, it tended to flow from the person with lower social standing toward the person with higher standing. The popular person in a group gets nominated by many people they would not nominate back. The person on the edge nominates upward and gets nothing in return. If you have ever felt like the one who is always slightly more invested, the study suggests that feeling was probably correct, and it was probably about position rather than about you.

What the gap does to a lonely person

Here is where I want to be careful, because this research can be read as bad news and I do not think it is.

If you are lonely, you are likely already suspicious that some of your friendships are thinner than they look. The reciprocity gap confirms that suspicion, which sounds bleak. But look at what it actually says. It says the gap is normal. Roughly half of everyone's friendships are lopsided, including the friendships of the confident, popular person you are comparing yourself to. They just have not noticed, because they are on the receiving end of the uphill ties.

The lonely person's mistake is not having one-sided friendships. Everyone has those. The mistake is treating a one-sided friendship as evidence of a personal defect rather than a common structural feature of how friendship works. We wrote about the related pattern of having plenty of people around and still feeling unaccompanied in You Have Friends but Still Feel Lonely? Here's Why, and the reciprocity gap is a big part of the explanation. You can have fifteen names on your list and be unsure which of them would list you back. Uncertainty about that is its own kind of loneliness, separate from being alone.

What we see on our own platform

I work with the data on YaraCircle, and one pattern shows up so consistently that I now think of it as the reciprocity gap in miniature.

When two people match and chat, either of them can send a friend request. What we notice is that the person who sends the request is often not the person who does more of the talking afterward. A request is a declaration: "I would like this to be a friendship." Whether it becomes one depends on what the other person does in the following days, and a large share of accepted requests go quiet within a week because acceptance was polite rather than intended. The request was reciprocated on paper. The friendship was not.

The pairs that hold are different in a way you can see early. Both people initiate. Not equally, nobody keeps score that precisely, but neither person is the only one who ever opens the conversation. When we look back at friendships that are still active months later, that early two-way initiation is the most reliable thing they have in common. It is a better predictor than how long the first conversation was, how much the two people had in common, or how quickly the request was sent.

That matches the study almost exactly. The one-directional tie feels like a friendship to the person sending it. The two-directional tie is the only one that survives.

The one question that sorts your list

You cannot ask your friends to rate you. But you can look at behaviour, which is what friendship is made of anyway. The question I would put to every name on your list is this:

In the last three months, have they reached toward me without being prompted?

Not replied. Reached. A message that was not an answer to yours. An invitation you did not fish for. A "saw this and thought of you." A question about something you mentioned weeks ago. Any unprompted movement in your direction counts.

Go through the list honestly. You will probably find it splits into three groups.

1. People who reach. These are your reciprocal ties. The study says they are about half of your list. They may not be the ones you would have guessed.
2. People who only respond. They are warm when you write, they never write first. This is the uphill tie. It is not nothing, and it is not what you thought it was.
3. People who do neither. Names that are on the list out of history. You have not spoken in a year. This is not a friendship, it is a memory of one.

The exercise is not meant to make you cut the second and third groups. It is meant to stop you from spending your energy as if all three were the first. Lonely people often pour their limited social effort into group two, because those are the people who feel most important, and get nothing back, because those are precisely the ties that run uphill.

What to do with what you find

Invest where the reaching already exists. The single most effective change is boring: give more of your time to the people who already reach toward you, even if they feel less exciting than the people who do not. A reciprocal friendship at closeness level three will do more for you than an aspirational one at level five that only exists on your sheet. This is also where the deepening happens. Our piece on the five stages of adult friendship describes the wall most friendships stall at, and you can only get past it with someone who is climbing too.

Stop reading responsiveness as reciprocity. A quick, warm reply is nice. It is not evidence that you are on their list. Plenty of people are warm to everyone. If you want to know where you stand, watch for the unprompted move, and give it a full quarter before you draw conclusions.

Make one uphill tie explicit, once. If there is a person in group two you genuinely want, you are allowed one direct move. Not "are we friends," which is unanswerable, but a concrete, low-stakes offer that requires them to initiate something: "I'm free most Thursday evenings this month, pick one and I'll come to you." If they pick one, you have learned something. If they say "definitely, soon" and never do, you have also learned something, and you can stop wondering.

Notice when you are the higher-status node. The gap runs both ways. Somebody out there has you on their sheet and you do not have them on yours. If you are lonely, that person is a real opportunity you have been ignoring because they did not seem important enough. The study's finding about status is uncomfortable precisely because most of us are doing the same uphill nominating we resent in others.

Why this is good news, in the end

A 53 percent reciprocity rate sounds like a verdict on human friendship. I read it as a correction to a fantasy we all share, that the people we love automatically love us back in the same shape and size. They do not, and they never did, and the ones who do are rarer and more valuable than the length of your list suggests.

The practical upshot is a relief. You do not need fifteen friends. You need to know which four or five are actually mutual, and to stop mistaking the other ten for something they are not. Once you know, you can put your effort where it returns, and a person with four reciprocal friendships and clear eyes is far less lonely than a person with fifteen names and a sheet nobody else has signed.

People also ask

What percentage of friendships are mutual? In a 2016 MIT study published in PLOS ONE, only 53 percent of friendship nominations were reciprocated. Across five other datasets the researchers analysed, reciprocity ranged from 34 to 53 percent, while 94 percent of people assumed their friendships were mutual.

How do I know if someone considers me a friend? Look for unprompted contact: messages that are not replies, invitations you did not fish for, questions about things you mentioned before. Responsiveness alone is not evidence. Over about three months, a friend will reach toward you without being prompted at least once.

Why do I always feel like I care more in my friendships? The study found one-directional friendships tend to flow from people with lower social standing toward people with higher standing. If you regularly feel more invested, you may be nominating "uphill," which is common and structural rather than a personal flaw.

Should I end a one-sided friendship? Not necessarily. One-sided ties are roughly half of everyone's friendships. The useful move is to stop investing in them as if they were mutual, redirect your effort to people who reach back, and make at most one clear, low-stakes offer to anyone you genuinely want to keep.

Is it normal to have very few real friends? Yes. If roughly half of a typical friendship list is non-reciprocal, a person who names ten friends probably has four to six mutual ones. Knowing which ones they are matters more than the total count.