You already know, somewhere under the noise, that it isn't working. Maybe it's the degree you stopped caring about two semesters ago. Maybe it's the business that eats your weekends, or the relationship where you keep waiting for the person you met in the first year. You know. And you stay, because leaving would mean admitting that all the time you gave it was wasted.

That sentence, all that time would be wasted, is the trap. It sounds like loyalty or grit. It is usually neither. It is an old and well-documented error in how minds handle the past.

The meal you finish because you paid for it

Economists call the thing you can't get back a sunk cost. The rule is simple: money, time, and effort already spent should not affect what you do next, because no choice you make now can recover them. The only question that matters is what each path offers from here forward.

People rarely follow that rule. In a classic 1985 study, Hal Arkes and Catherine Blumer sold theater season tickets at randomly assigned prices: full price, or one of two discounts. Those who paid full price attended more plays in the first half of the season than those who paid less. The seats were identical and so were the plays. Only the money already spent differed, and it pulled people through the door.

An earlier line of work by Barry Staw, published in 1976 under the title "Knee-Deep in the Big Muddy," described escalation of commitment. People who had made an investment decision and then got bad news often put in more resources instead of fewer, especially when they felt personally responsible for the original choice.

The second detail matters most. The pull isn't mainly about money. It's about me. Quitting feels like a verdict on the person who started.

Why walking away feels like losing twice

Two mechanisms are doing the work, and naming them helps.

The first is loss aversion, from Kahneman and Tversky's prospect theory. Losses weigh more heavily than equivalent gains. Stopping makes the loss official, while continuing keeps it open, a loss that hasn't been "realized" yet. So the mind prefers the open wound.

The second is self-justification. Once you've committed, stopping creates dissonance between "I'm a sensible person" and "I spent two years on something that didn't work." The easiest way to resolve that is to keep going and revise your opinion of the project upward. You aren't lying. You are quietly recruiting evidence.

This gets worse the more it hurts. The bigger the investment, the stronger the pull to protect it. The people most committed to a bad path are often the ones who gave it the most.

Who escapes it, and what they have

There is a hopeful finding here. Arkes and Peter Ayton argued in 1999 that young children and some animals seem less prone to the sunk cost effect than adults. They suggested that the habit of "don't waste" gets learned, and that it becomes a rule applied without checking whether it fits. Later research by Jennifer Strough and colleagues found that older adults were, on average, less susceptible than younger ones, which hints that the pull can weaken with experience.

Nobody is born stuck, then. It's a habit, and habits can be interrupted.

The most useful interruption is distance. Andy Grove, the former Intel CEO, described a moment in the mid-1980s when he and Gordon Moore faced a failing memory-chip business. He asked Moore what a new CEO would do if the board replaced them. The answer was that the new person would get Intel out of memory chips. Then, Grove wrote, they walked out the door and did it themselves. The question changes who is deciding. A stranger has no ego invested in the past, so the stranger sees the present clearly.

The problem with deciding in the middle

Here is the practical difficulty. When you're mid-project, your judgment is the compromised one. You are the person with the most reason to keep going and the strongest story about why. Willpower in the moment is a poor defense.

The better defense is to decide before the fog arrives. That means writing down, at the start, what would count as a signal to stop. Some people in decision-making circles call these kill criteria. They are the same idea as a premortem's tripwires, but pointed at a specific question: under what conditions will I walk away?

A kill criterion works because it moves the quitting decision to a time when you are calm and neutral. Later, when you feel the pull to rationalize, you aren't arguing with a vague feeling. You are checking a sentence in your own handwriting.

What a good kill criterion looks like

Vague criteria fail, because vagueness is where self-justification lives. "If it stops being fun" can be argued away forever. Compare:

  • "If I haven't had a paying client by the end of June, I stop and reassess."
  • "If I've dreaded the Sunday-night feeling for six weeks in a row, I have a real conversation with my manager."
  • "If I'm still the only one initiating plans by March, I tell her what I've noticed."

Each has a date or a count, and each names a next action rather than a dramatic ending. The aim isn't to quit more. It's to make sure that when you continue, you do it on purpose.

Your next moves

You don't need a crisis to start. Pick one thing you suspect has become a trap and do this today:

  • Name the one commitment you keep protecting. Write it at the top of a page, then write the single sentence you tell yourself about why you can't stop. Read it back as if a friend had said it.
  • Split the ledger in two columns. Put "Already spent" on the left (time, money, effort) and "From here on" on the right (what it will cost and what it could realistically give). Then cross out the left column. It's information about the past, not a vote about the future.
  • Ask the Grove question in writing. Complete this sentence: "If I were advising a stranger who had just inherited this situation with none of my history, I would tell them to..." Write the honest answer, even if you don't act on it yet.
  • Set one kill criterion with a date. Make it concrete: a number, a deadline, a named action. Put the date on your calendar as a review, not a verdict.
  • For anything new, write the exit conditions first. Before your next big commitment, a course, a project, a lease, spend two minutes on what would make you stop. Future you will be grateful.

A record that outlasts the mood

The reason this works is that memory is on the side of the trap. Over time you forget how it felt when you first doubted. You forget the date you set. You remember only that you've come this far. A written note doesn't forget, and it doesn't care how much you've spent.

That is where a fast, plain notes app earns its place. Pagebox opens in under a second and keeps everything local-first, so a kill criterion and a review date can go from thought to saved in the moment you have it, without turning it into a project. A list titled "Things I'm still deciding about," with a date beside each line, quietly becomes a record of the version of you who saw clearly.

If you'd like a place to keep it, you can try Pagebox at https://pagebox.lumenlabs.works. But the idea works on a napkin. What matters is that the decision to stop, or to keep going, gets made by the calm person and not the one who has already paid too much.