The Money Is Already Gone: Escaping Sunk-Cost Entrapment

The correct question about a sunk cost is not whether to continue. That framing invites you to weigh the investment, which is exactly the error.

Arkes and Blumer's theater-ticket study, chapter 18, is the demonstration: randomly discounted season-ticket holders attended fewer plays than full-price buyers, though every ticket was equally paid for and equally unrecoverable. What changed behavior was not the value of the ticket but the size of the acknowledged loss.

The two drivers are loss aversion — abandoning the investment converts an expenditure into a realized loss — and self-justification, since the decision-maker is defending their earlier judgment as well as their money. Staw's escalation work shows the second is strongest where the person was publicly responsible for the original decision.

Manipulators use this deliberately and the phrasing is always the same. Think how far you've come. After everything you've put into this. You can't stop now. Every one of those sentences is an appeal to a cost that cannot be recovered by any action, and it is worth noticing that it is never accompanied by a claim about the future.

The detection signal in yourself is when your reasons for continuing are all historical. Write down why you are staying — in a job, a relationship, an investment, a project — and check the tense. If the list is about what has already been spent rather than about what is likely to happen next, the decision is being made by the past.

The protocol is the fresh-start question, and it has to be phrased carefully to keep the history out.

Not: is this worth continuing. But: if I were offered this position today — this job, this relationship, this investment, at this price, with these prospects, and with none of the history — would I take it?

Hafenbrack, Kinias and Barsade tested an addition to this in 2014 and found it worked. A brief focused-breathing induction of around fifteen minutes reduced sunk-cost bias, and their proposed mechanism is that it shifts attention away from past-oriented negative affect and toward present information. That is a modest effect from a modest intervention, and it is one of the few debiasing findings in this area with direct support.

The counter-response is to decide forward only, and to say so explicitly when someone invokes the past: that money is gone whatever I do next, so it cannot be a reason.

What the past can legitimately tell you is what it taught you about the future. That is the only claim on it you should accept.

This counters Law 16.