Nobody finishes a cold, bitter coffee because they want it. They finish it because they paid for it. Watch a table at any cafe long enough and you will see it: the cup pushed aside an hour ago, the surface gone matte, the drinker returning to it with the grim resolve of someone settling a debt. The pleasure left the building before the second sip. What remains is accounting.
This small daily surrender has a proper name. The sunk-cost fallacy is our reluctance to abandon something we have already invested in — money, time, effort, the part of ourselves we cannot itemise — even when walking away is plainly the smarter move. Economists drew a clean line decades ago: a cost already incurred and impossible to recover should carry no weight in what you do next. The future does not refund the past. Yet the mind keeps the receipt and quotes it back at us, loudly, at exactly the wrong moment.
Coffee makes the bias visible because the stakes are low enough to be honest. Nobody lies to themselves about a two-dollar cup; the self-deception only kicks in around the things that matter. But that lukewarm mug is a faithful model of the larger holdings we refuse to liquidate. The job that stopped teaching you anything three promotions ago. The relationship running on the fumes of a shared history. The startup, the degree, the renovation, the manuscript — each one a cup we keep returning to not for what it still offers, but for what it already cost. We mistake the size of the bill for the quality of the drink.
The trouble is that the bill and the drink have nothing to do with each other. Whatever you have already spent is gone in identical measure whether you continue or quit — that is precisely what makes it sunk. A rational decision weighs only what lies ahead: the value still on offer, the cost still to come, the better cup you could be drinking instead. The past investment is not a reason to stay; it is simply a fact, and a closed one. The psychologists Hal Arkes and Catherine Blumer pinned this down in their 1985 study on the psychology of sunk cost, showing how readily people throw fresh resources after spent ones — and later work on the sunk-time effect found the trap snaps shut hardest when we chose the commitment ourselves, because then quitting feels like admitting we were wrong.
And that admission is the real toll. The fallacy is not stupidity; it is dignity, misfiled. Continuing lets us pretend the original choice is still being vindicated, that the money and the months are an investment maturing rather than a loss already booked. Quitting forces the loss into the open, makes it final and legible. So we keep sipping, converting a small mistake into a slow one, because a cold cup we are still drinking does not yet feel like a cup we got wrong. This is also why the bias scales so neatly into escalating commitment — boardrooms pour billions into failing projects for the same reason you finish the bitter dregs. The logic does not change with the zeroes; only the audience does.
The cure, mercifully, is a single question, and it has nothing to do with what is already in the cup. The money is spent. The coffee is cold. Neither fact will improve by your continued attention. So the only thing worth asking — about the mug, the job, the whole portfolio of things you are loyally finishing out of obligation — is whether you would order this exact cup again, right now, knowing what it tastes like and what it costs from here. If the answer is no, you are not honouring an investment. You are just drinking it cold.
Which is the quiet joke of the whole thing. We tell ourselves persistence is a virtue, and often it is — but persistence and the sunk-cost fallacy wear the same coat, and only one of them is keeping you warm. The trick is learning to tell them apart before the next round, while the kettle is still on and the choice is still yours to make.
