By Graham Peterson
For the uninitiated, which I don’t blame you for being considering the levels of excitement in economics education, a sunk cost is essentially the idea that once a resource has been expended, its importance is through — the gig is up — the cost is sunk. All that matters in a strictly economic calculus are future expectations of costs and benefits — agents (and firms) pursue an activity when its benefits are larger than it’s costs. Forget the past.
A sunk cost fallacy then, is the recognition that people persistently and ubiquitously think to the past to justify future-going investments. “Chasing bad money with good,” is a very 1950s colloquial demonstration of the idea. The constant investment in NASA, because we’ve already spent such enormous gabs of money, is another classic economist-as-political-curmudgeon example.
But it goes much deeper. An astute microeconomics instructor pointed out to me once that regret itself is in the economic view a blank mystery because it is a sunk cost fallacy — what’s done is done; water under the bridge. Buyer’s remorse in this view doesn’t exist. Nor does the very straightforward idea that people make mistakes in their decisions at all (that is that their preferences at t_1 can and often do conflict with those those they had at t_0).
But a potential explanation for sunk cost reasoning has been missed. In the economic view, there is no ego and ID, no me versus myself, no internal dialogue with one’s conscience or consciousness. Agents with a given portfolio of preferences look straight out through their eyeballs at the future and decide how to act — they don’t look back within themselves and reflect.
Now this view of self reflection, an internal dialogue, and an inherent ego-alter social relationship within the individual herself is relatively noncontroversial in most other human scientific theory, going back to at least George Herbert Mead and William James. And assuming this position starts to make sunk cost reasoning look a lot more reasonable.
Imagine that in this internally-social relationship you have with your self, you form the same kind of social commitments, and abide the same kind of reciprocity rules you do with other people. Now we’ll reintroduce the economics.
Imagine that your self indeed does rationally weigh expected costs and benefits, and that when your self commits hence to an action, he, a principle, signs a contract with you, an agent (your conscience or executive function). Given this contract, you, as a contract-bound agent, carry out the contract.
But of course as time goes on into t_1, you, as an agent of the principle, continue to gather information and weigh expected costs and benefits as well. You realize that the contract you are carrying out failed to account for contingencies that you are now aware of because your self didn’t have complete information in t_0. You thus enter into negotiations with your self in an attempt to decide whether to nullify the contact or not.
Fallacious sunk cost reasoning, in this scenario, would seem to arise from your self protesting that you had good reason at t_0 to embark on the action it contracted with you to undertake: “I put a lot of time into making this gravy because I expected to get sufficient long run benefits from making so much of it.” “Yeah, but bro, it’s starting to spoil and now you’ve got to invest more time into reheating it so it doesn’t.” (yup – I thought of all this while making breakfast – at 3pm)