Sunk costs

A few months ago I wrote about opportunity cost, the economic term for the value of the best alternative forgone when you make some decision. Opportunity costs aren’t always monetary: mutually exclusive alternatives are hiding in many decisions, and are especially difficult to spot and reason about in decisions involving time and effort. When you take on one project or job, you’re giving up anything else you could be spending that time on.

The concept of sunk cost is another piece of mental machinery I feel lucky to understand. “Sunk cost” is an economic term for a cost that you’ve already paid, irrecoverable regardless of what actions you take in the future. It’s a cost you can’t take back, unlike a prospective opportunity cost about which you have yet to decide. Sunk costs can be difficult to spot in casual decision-making, and much harder emotionally to avoid.

When you’ve already paid a cost, say putting down a deposit on an exotic trip, that money is gone whether you take the trip or not. But we fallible humans are often tempted to take sunk costs into account when making future decisions. If, after paying your deposit, you learn of an even better trip you could take instead, that’s also much cheaper than the first, it can be difficult to ignore the money you would “lose” on the deposit were you to switch. But the money doesn’t care which trip you take. A better option is a better option, and the difficult but rational thing to do is to ignore the sunk deposit (except as far as it makes the first trip a little cheaper) and choose the best option now available. To take the original, inferior trip anyway is throwing good money after bad.

A recent sunk cost I succumbed to: I bought a return train ticket to a social event. On my way home afterwards a friend offered me a lift, but I declined as I’d already paid for my return journey. What insanity! I’d faced a choice between a free train journey and a free, comfortable lift to my door, with bonus conversation, but had allowed a few dollars, already beyond my reach, to propel me to the train station. I realised my error as I plodded to the platform, too late to change my mind.

This hints at a good way of avoiding the emotional trap: re-frame the dilemma so that it doesn’t involve any costs. Treat whatever you bought for those irrecoverable dollars or hours, if anything, as something you can now have for free, should you want it. If this free start makes the associated option better than any alternative, great, but otherwise you shouldn’t feel so bad about giving up something free that you didn’t really want anyway.

Like opportunity costs, sunk costs are often non-monetary. As someone who designs and makes things (in software), it’s not uncommon for me to discover, a little way into a project, a different and much better approach. Perhaps the new technique will result in a superior end product, or the subsequent effort will be halved. I face a difficult decision of whether to continue with my initial, comparatively crude attempt, or throw it away and start fresh with a better way. The hours spent slaving away so far seem deeply relevant to this decision; in fact, the more painful the work has been the more valuable, too, surely! But they’re not. If I were to ask a dispassionate observer they could make a decision without regard to the hours spent so far, only the existence of a partially-complete attempt down a poor path. Reframing the dilemma to imagine that someone else had done the work that I have (perhaps I stumbled across it as an open-source project) helps me evaluate it only for the value it really carries, instead of the sunk cost to obtain it.

Just as with opportunity costs, I feel a little bit super-human every time I spot a sunk cost in my or a colleagues reasoning and can steer us towards a better decision. Take that, puny human brain!

2 Thoughts on “Sunk costs

  1. Good post – and hits close to home. This is otherwise summed up as the sunk costs fallacy.

    Short video on it here, similar vein to what you wrote.

  2. That said, you could argue the greater sunk cost was your investment of time into making that friendship. Though not sure how I feel about friendship value quantification. I’m sure there’s an xkcd comic about that.

Post Navigation