The fallacy of sunk costs occyrs when people take past expenditure into account when they make a decision, even though it cannot affect their future and cannot be changed. It is often regarded as a canonical example of a cognitive bias in humans because almost everybody does it. Of course, if it’s so suboptimal to take sunk costs into account then one has to wonder why humans have evolved to systematically make that error. Sandeep Baliga and Jeff Ely have a paper out where they suggest that it’s actually on optimal response to limited memory capacity:
We offer a theory of the sunk cost fallacy as an optimal response to limited memory. As new information arrives, a decision-maker may not remember all the reasons he began a project. The sunk cost gives additional information about future profits and informs subsequent decisions.
Essentially they say that, if you don’t remember all the reasons why you invested in the first place, then the level of past investment in a project tells you something about the conclusions you initially reached. You can then use that information to inform your decisions about whether to continue, without having to rethink everything from scratch.
Now I’m late to this because they also write the excellent Cheap Talk blog and discussed their working paper a year ago! So if you want to know the details of the ‘Concorde effect’ and ‘pro rata effect’ then head over and have a read.