Sunk cost vs opportunity cost: which should affect your decision?

Two questions about the same decision

Ask ‘Can I recover this?’ to identify a sunk cost. Ask ‘What is my best alternative from here?’ to identify an opportunity cost. Neither is simply another name for the price on a receipt.

Opportunity costs can involve time, enjoyment or learning as well as money. You do not add up every possible alternative: if you have one free evening, you cannot spend it at three different events. Compare the feasible choices and identify the best one you would forgo.

[1][2]

A worked example: finish the booth or take another job?

Imagine a maker has spent £120 designing a display booth. None of that payment can be recovered, and the design has no resale value. For this hypothetical decision, there are two mutually exclusive uses of the same weekend: finish the booth for a guaranteed £200 payment, spending another £50 on materials, or take a separate job paying £180 with no extra costs. Assume the time, effort and all other benefits are equal.

Finishing produces £200 minus £50, or £150, from this point forward. Taking the other job produces £180. The other job leaves the maker £30 better off. The £120 already spent is the same past loss whichever option is chosen.

Where is the opportunity cost in that example?

By finishing the booth, the maker forgoes the £180 net payment from the alternative job. Comparing that £180 with the booth's £150 net benefit shows the £30 disadvantage. This way of comparing options already accounts for the booth's £50 materials cost.

Do not subtract the alternative job twice. You can compare both net outcomes directly, or compare the booth's net outcome with the forgone alternative. Those are two ways of expressing the same decision, not two separate penalties.

Real choices rarely make every benefit certain or equal. If the booth creates a valuable client relationship, or the job might be cancelled, include those future consequences and uncertainties. The example isolates the cost distinction rather than pretending money is the only thing that matters.

What changes if some money can be recovered?

Suppose cancelling the booth would also return £40 of the earlier payment. That £40 is now available only if the maker cancels, so it belongs in the comparison. The amount originally paid is less useful than knowing what is recoverable today.

Similarly, an old sewing machine may have resale value even though you paid for it years ago. Using it can mean giving up a sale or another use. ‘Paid in the past’ does not automatically make the full purchase price irrelevant in every form: separate the unrecoverable expenditure from the asset's remaining value.

[2]

A common misconception: sunk-cost reasoning means always quitting

It does not. Change the booth payment to £260 while leaving the additional cost at £50. Finishing now yields £210, which exceeds the £180 alternative. Continuing makes sense under the example's assumptions, even though the original £120 remains sunk.

The useful question is ‘Which option has the better future consequences?’ That can favour finishing, switching or stopping. It cannot be answered simply by counting how much you have already invested.