You should have dropped it in March. You knew in March.
What stopped you wasn't hope, exactly. It was the eighteen months already in it, and the sense that walking away now would make those eighteen months a waste — as though continuing could still redeem them.
It can't, and everyone knows it can't. The odd thing about sunk cost is that being told about it changes almost nothing.
What the fallacy actually says
The economics is simple enough that it's taught to teenagers. Money and time already spent are gone regardless of what you do next, so they shouldn't feature in the decision. The only question is whether the future cost is worth the future benefit, judged from here.
Applied to a goal: the eighteen months are gone whether you continue or stop. They're not a reason to continue, because continuing doesn't recover them. They're not a reason to stop either. They're simply not evidence about anything, and the decision is identical to the one you'd face if you'd started yesterday.
Almost nobody can act on this, which tells you the problem isn't a shortage of the argument.
Why knowing doesn't help
Because sunk cost isn't really about the money. It's about what dropping the thing would mean.
Continuing lets the eighteen months stay in a particular category — an investment, unfinished. Stopping recategorises them, in one move, into a mistake. The cost of quitting isn't the loss of the project; it's the retrospective rewriting of a year and a half of your life, and that's a genuinely large thing to be asked to do on a Tuesday evening.
There's a second layer where other people are involved. If anyone knows about it — a partner who covered for the evenings, a friend who asked how it was going — then stopping requires saying so, and the anticipated conversation is often heavier than the project.
And there's a version that's about who you are rather than what you've spent. Eighteen months in, the thing has become part of your description of yourself. Someone doing an MA is a person doing an MA. Stopping doesn't just end a project, it removes an identity, and there's usually nothing waiting to replace it.
That's why the economic argument bounces off. It answers a question about resources, and the resistance is about meaning.
It also explains the specific shape of what happens next, which isn't a decision to continue. Nobody sits down in March and resolves to press on. What they do is stop looking at it — defer the difficult module, put off choosing a topic, quietly stop mentioning it to people who might ask. The project isn't continued and isn't dropped. It's suspended in a state where no verdict has to be delivered, and it can stay there for years, because the only thing sustaining it is the absence of a moment where anyone says anything out loud.
What works instead
Not arguing yourself out of it. Three moves that sidestep the problem rather than winning it.
Name the number, exactly, in writing. Eighteen months and about £900. People expect this to hurt and it usually does the opposite — an unnamed cost operates on you as a vague enormous weight, and a named one turns out to be a specific finite thing you can look at. Vagueness is what gives it power.
Then ask the fresh-start question, which is the only version of the economic argument that actually lands: if I were offered this today, at this price, knowing what I now know, would I take it? Not would you have started it. Would you start it now. That question removes the history from the frame without requiring you to argue that the history doesn't matter.
Then find what's genuinely kept. This is the one that does most of the work, because it dissolves the mistake framing rather than accepting it. Eighteen months bought something, even if it didn't buy the qualification — a skill, a contact, a discovery that you dislike a field you'd otherwise have spent a decade in. That last one is enormously valuable and never feels like it. Write down what you keep, and the ledger stops reading as a pure loss, which is the thing that was making it unwriteable.
The argument for staying in
There's a real case on the other side, and sunk cost gets invoked too easily against it.
Long projects have a middle. It's flat, the novelty has gone, the end isn't visible, and the honest feeling during it is often indistinguishable from the feeling of a project you should abandon. Plenty of finished books, degrees and companies went through a year of that. Someone who applies the fresh-start question during the middle will quit, and will be able to construct a rigorous-sounding justification each time. Do that repeatedly and you have a life of eighteen-month attempts.
There's also a version where the cost genuinely does change the maths. Sunk cost says ignore what's spent — it doesn't say ignore what's built. If eighteen months of work means the remaining work is three months rather than twenty-one, that's a real fact about the future, and it belongs in the decision. The fallacy is counting the past as a reason. Counting progress isn't the same thing, though it's easy to dress one as the other.
The distinguishing question is whether you can describe, concretely, what the remaining work is. A person in a difficult middle can tell you what's left. A person in a dead project talks about it in general terms and changes the subject, and has done for months.
Eighteen months, £900, and a Sunday
He starts the MA in September, part-time alongside work. The first year is genuinely good. The second begins with a module he doesn't care about, and by March the essays have become something he does the night before.
He knows in March. What he does instead is defer the dissertation topic, twice, and stop mentioning it to people.
In September, two years in, he does three things in about twenty minutes.
He writes the number: eighteen months, £900, roughly six hours a week. Seeing it written is worse for a moment and better afterwards, because six hours a week is a fact rather than a fog.
He asks the fresh-start question: offered this today, at this price, knowing the second year? No. Not close. There's no hesitation at all, which is itself informative.
And he writes what he keeps: the research methods module changed how he works and has been useful at his job every month since. That's real, and it happened, and no decision he makes now can take it away.
Then one line — dropping the MA, it was my mother's idea and I've known that for two years — and it comes off the list. Written down rather than merely decided, because a decision that only exists in your head will be back in November wearing a slightly different name, and a written one won't.
The relief arrives about ten minutes later and is considerably larger than expected. That's normal, and it's also the confirmation. Nobody feels lightness at abandoning something they wanted.
What the list was doing in the meantime
For eighteen months the goal sat there. In the last six of them it was at the top, because it had the longest gap, which is what sorting by silence does to things nobody is touching.
That's not the app telling him to work on it. Quite often the right response to a goal at sixty days quiet is to admit it's finished and delete it, and that counts as the system working. The gap doesn't tell you which goals to complete. It tells you which ones you've stopped being honest about, and it keeps asking until you answer.
Related: Quitting is a skill · The anatomy of a stuck goal · Review by gap, not by list
Share: