The Sunk Cost Trap: Why You Stay in Relationships Past Their Expiry

Eleven years. Two moves, a mortgage, both sets of parents, a shared history nobody else has. When you imagine leaving, the argument that stops you is almost never about the future. It is about everything already spent. That reasoning has a name, it has been studied since the 1980s, and understanding it will not make the feeling go away, but it will let you see what the feeling is actually doing.

This is written for reflection rather than instruction. Nothing here tells you to stay or go, and no article could. What it can do is separate the reasons that are about tomorrow from the reasons that are about yesterday.

What Is the Sunk Cost Fallacy in Relationships?

A sunk cost is anything you have already spent that you cannot get back regardless of what you do next. Money, years, effort, hope. The fallacy is letting the size of that spend determine whether you continue.

The reason it is a fallacy and not just sentimentality is that the past investment carries no information about the future. Whether the next five years are good has nothing to do with whether the last eleven were expensive. But the mind treats the spend as though it were a deposit that continuing might somehow redeem.

The foundational research is Hal Arkes and Catherine Blumer's 1985 paper The Psychology of Sunk Cost in Organizational Behavior and Human Decision Processes, which established the effect experimentally.

One of their studies is worth describing because it is so ordinary. They arranged for a university theatre to sell season tickets at randomly assigned prices to the first sixty customers: some paid full price, some got a small discount, some a large one. Everyone received identical tickets to identical plays. The only difference was what they had paid.

People who paid full price attended significantly more performances in the first half of the season, around 4.11 plays against roughly 3.3 for the discounted groups. The plays were no better. The value of attending was identical. The only thing driving the difference was the size of the money already gone, which no amount of attending could bring back.

Arkes and Blumer traced the effect to waste aversion rather than to any rational reading of future value. People were not calculating that the plays would be good. They were avoiding the feeling of having wasted something.

Translate that into a relationship and the season ticket is a decade of your life. The plays are the years ahead. And the reason you keep attending is not that you expect them to be good.

What people usually mean when they invoke sunk costs, in order of how often it comes up: the time, the shared property, the children, the wedding, the friends who became joint friends, the version of the future you had already built in your head, and the sheer embarrassment of having been wrong in public for so long.

Why Does Your Brain Resist Leaving After Heavy Investment?

Several things stacking on top of each other, and it helps to see them separately because they need different answers.

Waste aversion. The mechanism Arkes and Blumer identified. Ending something feels like throwing away everything spent on it, and the aversion to waste is strong enough to override a clear-eyed reading of what comes next. The error hidden in this is a matter of timing: the waste, if there was any, already happened. Leaving does not create it. Leaving only stops adding to it.

Loss framing. Leaving is experienced as losing something, while staying is experienced as keeping it. Losses register more heavily than equivalent gains, so the scales are tilted before you begin. Nobody frames it as continuing to spend, though that is what it is.

Self-consistency. Admitting a relationship is over means accepting that a long stretch of your own judgement was wrong, in front of people who watched you make it. That is a genuinely painful thing to concede, and staying protects you from conceding it.

The stranger problem. There is a well-known argument in philosophy about how weakly we identify with our future selves, sometimes described as treating that person almost as a separate individual. It shows up here in a specific way. The person who paid the eleven years feels vividly real, and the person who has to live the next eleven feels theoretical. So the decision gets made on behalf of the one who already spent, rather than the one who has to live with the outcome.

Loss of the sense that acting works. Where effort has repeatedly failed to change anything, the capacity to act at all thins out. Maier and Seligman, revisiting fifty years of work on helplessness in Psychological Review in 2016, concluded that passivity under prolonged difficulty is the default state and that what actually has to be learned is the detection of control. If nothing you did shifted the relationship, the machinery that would carry you out of it has had nothing to feed on.

Escalation of commitment. The trap tightens as it goes. Each additional year raises the sunk cost, which strengthens the argument for staying, which produces another year. It is self-reinforcing, and it is why the reasoning feels stronger at year fifteen than it did at year six, even as the underlying case for staying has usually got weaker.

How Do You Know If Sunk Cost Is Keeping You Stuck?

The clearest tell is in the grammar of your own reasoning. Sunk cost arguments point backwards. Reasons to stay point forwards. Listen for the tense.

• Your reasons are quantities of past. "Eleven years." "After everything." "We've been through so much." Notice these describe expenditure, not expectation.
• You cannot describe a good version of next year. Ask yourself what a good twelve months would look like here, specifically. If the honest answer is more of this but calmer, that is not a forecast, it is an absence of one.
• You find yourself defending the decision to friends. Particularly with history rather than with anything current.
• "Wasted" is the word that keeps appearing. If the dominant fear is that the years would have been wasted, you have identified the mechanism precisely. That is waste aversion, and it is a feeling about the past dressed as a plan for the future.
• The threshold keeps moving. You said you would leave if X happened. X happened. There is a new X now. If you have done this three or four times, the threshold is not a decision rule, it is a way of postponing one.
• You are waiting for something bad enough to justify going. This is worth sitting with. Needing permission in the form of a serious enough incident means you already know the answer and are waiting for evidence that would satisfy an imagined jury.
• The relief fantasy is specific and the fear is vague. When you picture leaving, if the imagined relief is detailed and the imagined disaster is a general dread, that asymmetry is informative.

The single cleanest test comes straight from the logic of sunk costs. Imagine you met this person today, with everything exactly as it is now, and none of the shared history. Would you begin this relationship? If the answer is a clear no, then the history is doing all the work, and the history is the one thing that cannot change.

What Is the Difference Between Commitment and Sunk Cost Thinking?

This distinction matters, because the sunk cost idea can be misused as an argument for abandoning anything difficult, and that is not what it says. Commitment is real, valuable, and not a fallacy.

Four tests separate them.

• What is the reason pointing at? Commitment is forward-looking: I am staying because I believe in what we can build, or because this rough patch has an identifiable cause we are addressing. Sunk cost is backward-looking: I am staying because of what leaving would waste. Same behaviour, opposite justification.
• Is there a shared project, and is the other person in it? Commitment involves two people working on something. If you are the only one making effort while invoking the history, that is not commitment, it is maintenance of a structure.
• Is there evidence of change over a reasonable window? Commitment through difficulty is reasonable when difficulty is moving. Look at eighteen months, not eighteen days. If nothing measurable has shifted across a long window despite genuine effort, staying is no longer a bet on improvement.
• Would you choose it again today? Someone committed can usually say yes without hesitation, and the years are a bonus rather than the argument. If the years are the argument, that is the diagnosis.

A fair caution against over-applying this. Long relationships have flat stretches. Years three to five of parenting small children are difficult for almost everyone, and treating an ordinary hard season as a sunk cost trap would be a mistake. The distinction is whether the difficulty is a phase with a cause or the permanent condition, and whether both people can name what would make it better.

How Does the Sunk Cost Trap Interact With Emotional Abuse?

Badly, and in a way worth understanding because the two patterns feed each other.

In an ordinary unhappy relationship, sunk cost is a bias you are running on your own. Where someone is actively controlling, the same reasoning gets supplied to you from outside, in language you then repeat as though you had arrived at it yourself.

How it typically shows up:

• The investment is quoted back at you. After everything I have given up for you. After all these years. The history becomes a debt with a named creditor.
• Your alternatives are made to look smaller. Nobody else would put up with you. You would not manage on your own. This works on sunk cost directly, since the fallacy weakens when leaving looks survivable and strengthens when it does not.
• Costs get added deliberately. Joint finances, a move away from friends, a pet, a pregnancy, a business in both names. Each addition is presented as commitment and functions as raised exit cost.
• Hope is topped up just as it runs out. This is the one that does the most damage. Right when you are close to leaving, something changes: an apology, a good month, a promise. That resets the clock and adds to the investment, and unpredictable reward binds harder than steady treatment does. Our guides on intermittent reinforcement and future faking cover both halves of that mechanism.
• Your read on the situation is undermined. If you no longer trust your own judgement, you cannot run the assessment that would tell you to go, and our guide on the signs of gaslighting covers how that erosion happens.

Why the combination is so sticky. Sunk cost gives you a respectable reason to stay that has nothing to do with fear, and that reason is much easier to say out loud. Telling a friend you are staying because of the years sounds sane. Telling them the truer version is much harder. So the fallacy ends up doing double duty as both a trap and a cover story.

Two things worth stating plainly here. Leaving a relationship where you have been steadily undermined is genuinely harder than leaving an ordinary unhappy one, and struggling with it is not weakness, as our guide on trauma bonding explains. And if there is any question of safety, the calculus in this article does not apply. Practical planning and outside help come first, and specialist domestic abuse services exist precisely because leaving can be the most dangerous moment.

How Do You Make a Clear Decision When You Are Inside the Trap?

By changing the question you are answering. You will not reason your way out using the frame that created the problem.

1. Ask the forward question only. Not "was this worth it" but "knowing what I know, do I want the next five years of this." The first question is unanswerable and endless. The second has an answer, and you probably already have it.
2. Run the stranger test. If a friend described this relationship to you in exactly these terms, with the same history and the same present, what would you think? People are consistently clearer about other people's situations, because they are not carrying the sunk cost.
3. Write the two years out. Two honest pages: the next two years if nothing changes, and the next two years if you go. Not best case, realistic case, including the hard parts of leaving. Getting them out of your head and onto paper stops the fear side from remaining conveniently vague.
4. Set a real threshold with a real date. Not "if things do not improve." Something specific and dated: by March, this particular thing will have changed, or I will accept it is not going to. Write it down. Private, and non-negotiable with yourself.
5. Separate the decision from the exit. These get fused, and the fusion is paralysing. Deciding you want to leave is one thing. Working out the logistics of housing, money and children is another. You can reach the first conclusion months before you act on it, and treating them as one question stops you reaching either.
6. Tell one person the true version. Not the version you say at dinner parties. Isolation is what lets the reasoning run unchallenged, and one accurate outside view is worth more than a year of private deliberation.
7. Expect the pull to persist after you decide. Understanding sunk cost does not switch it off, any more than knowing about optical illusions makes the lines look equal. The aim is not to stop feeling the pull. It is to stop treating the pull as though it were an argument.

One closing thought, and it is the reframe that seems to help most. The years are not a reason to stay. They are also not wasted, whichever way you choose. They contained real things that happened to a real person, and they made you someone who now knows something they did not know at the start. What they cannot do is buy you a future. That has to be worth choosing on its own terms.

Whatever you decide, rebuilding a sense of your own worth separate from the relationship tends to be the work that follows, and our guide on self worth is a place to begin.

What Else Do People Ask?

What is the sunk cost fallacy in a relationship?

It is staying because of what you have already invested rather than what you expect from the future. The years, the money, the shared history. Those costs cannot be recovered by continuing, and they carry no information about whether the next five years will be good. Arkes and Blumer established the effect experimentally in 1985 and traced it to waste aversion rather than to any rational reading of future value.

Why is it so hard to leave after many years?

Several forces stack up. Waste aversion makes ending feel like throwing away everything spent. Leaving is framed as a loss while staying is framed as keeping. Admitting it is over means conceding that a long stretch of your own judgement was wrong. And each additional year raises the sunk cost, which strengthens the argument for staying, which produces another year.

What is the difference between commitment and sunk cost thinking?

The direction the reason points. Commitment is forward-looking: staying because of what you believe can be built, or because a rough patch has an identifiable cause being addressed. Sunk cost is backward-looking: staying because of what leaving would waste. Same behaviour, opposite justification. If the years are the argument rather than a bonus, that is the answer.

How does sunk cost interact with emotional abuse?

The reasoning gets supplied from outside rather than generated by you. The investment is quoted back as a debt, alternatives are made to look smaller, exit costs are deliberately added, and hope is topped up just as it runs out. Sunk cost also serves as a respectable cover story, because saying you stay for the years sounds saner than the truer version.

How do I decide when I am inside the trap?

Change the question. Ask whether you want the next five years of this, not whether the past was worth it. Run the stranger test by imagining a friend described the same situation. Write out two honest years each way so the fear side cannot stay vague. Set a specific dated threshold. And separate deciding from exiting, since fusing them is what causes paralysis.

Sources: Arkes, H. R., and Blumer, C. (1985), "The Psychology of Sunk Cost," Organizational Behavior and Human Decision Processes, 35(1), 124-140, including the Ohio University Theater field study in which full-price ticket holders attended significantly more performances than randomly discounted ones. Seligman, M. E. P., and Maier, S. F. (1967), on uncontrollability and its transfer into new situations, relevant to why the decision feels unavailable from inside (ppc.sas.upenn.edu/sites/default/files/lhtheoryevidence.pdf).

For self-reflection and educational purposes only. Nothing here is a diagnosis or a substitute for professional mental health support. Only a qualified clinician can assess or diagnose any condition. If you feel unsafe, please contact a local domestic abuse or crisis helpline.