The psychology of holding on to "bad" friendships

I spend most of my time here making the case for friendship. How much it gives back, how it's worth the effort, how it's one of the best investments you'll ever make.

So let me say the other part out loud: not every friendship is a good one.

No two relationships are the same, and some are meaningfully better for you than others. A few will leave you anxious, uncertain, and completely drained. Valuing friendship doesn't mean accepting whatever's on offer. It doesn't mean letting a relationship take up space in your life while it quietly costs you something. The good ones add. They carry you on the bad weeks and light you up on the good ones and give you the feeling of belonging somewhere.

"If a friendship doesn't serve me, I let it go." Sounds simple. It never is. We hold on. We make excuses, we call it loyalty, we defend the person to other people and to ourselves.

I read an article on Psychology Today called The Cut-the-Line Rule: When to Hold On and When to Let Go, which is about decision-making generally, and I could not stop applying it to friendship. Laura Berger asks a question I've been chewing on since: is this actually commitment, or is it just the fear of admitting that everything you already put in isn't going to pay off?

The best thing about that article is the reassurance. Holding on doesn't make you irrational. It makes you extremely normal. You're running into patterns so common they have names and decades of research behind them.

Here's what's happening under the hood.

Commitment bias

Also known as escalation of commitment. The Decision Lab describes it as our tendency to keep backing our past decisions even when the evidence says we shouldn't.

A big reason is that backing out feels like admitting we were wrong to begin with. We're wired to stay consistent with what we've already said and done, and when our actions and beliefs stop lining up we get cognitive dissonance — a specific, uncomfortable itch that we resolve by changing the belief instead of the behavior.

In a friendship it sounds like: this has been bad for a while, and the reasonable move is to step back. But stepping back means the last four years were a mistake. So instead of leaving, we re-commit. We double down and call it loyalty.

Sunk cost

This one's more straightforward. The Decision Lab defines the sunk cost fallacy as following through on something because of what we've already put in — time, money, effort, emotional energy — even when quitting is obviously the better call.

It happens because walking away triggers guilt and a sense of loss, and we are famously bad at accepting losses. So the past investment starts to feel more valuable than it is, and bigger than whatever continuing is going to cost us going forward.

In a friendship it sounds like: we've had so many hard conversations trying to fix this. We've known each other since we were nineteen. She was there when my dad got sick. Throwing that away would be a waste.

Maybe it would be. But sunk cost only asks what you've already spent. It never asks what the next three years will cost you.

And then there's the fact that we're human

This one isn't a bias or a fallacy. It's just how we're built.

We're wired to form bonds almost anywhere, with almost anyone, given enough time and proximity. And we don't lose those bonds gracefully. We grieve them, hard. There's a review in Nature showing that social pain — rejection, exclusion, loss — runs on the same brain machinery as physical pain. Not as a metaphor. The same regions.

So of course we hold on to friendships that stopped feeling good. Letting go genuinely hurts, in a way your brain does not clearly distinguish from injury.

Sometimes we stay out of real love and real hope that it can be repaired, and sometimes that's exactly right. And sometimes we stay because we're avoiding a confrontation, or a loss, or the feeling of having done all of it for nothing. So we keep giving.

A friendship is supposed to add to your life. Sometimes the healthiest thing you can do is stop pouring into bad friends that don't, and redirect it toward the people who lift you up. Including yourself.

Where to start with bad friends

Not with a decision. With noticing.

Look at your relationships and ask which ones leave you feeling supported and which ones leave you flattened. Which ones come with a sense of ease, of being liked as you are. None of them will be perfect, because no one is. The question isn't whether it's always balanced or whether you're always the one giving. It's whether, on the whole, this person adds.

Most of us have never actually looked. We just have a vague feeling.

Get re-centered

Centr can help with the noticing part. Track your plans, then track how you felt afterward — energized, drained, glad you went, wishing you hadn't. Over a few weeks you stop guessing and start seeing the pattern.

That's the whole point. A real view of your social life, so you can get re-centered.