A group of colleages and I have been thinking about the sunk cost fallacy a lot lately, contemplating at its prevalence and the possibility that it may be a features rather than a flaw.
So I wrote a cute story about it.
Assumption -> in caveman times, life expectancy for infant born with n guardians (parents/elders etc) is higher than life expectancy for infant born with n-1 guardians.
(written in my finest caveman dialect)
Grog and Urg are male and female cave-persons, respectively, who share cave. They are part of local tribe, where each member serves distinct role. Grog is hunter for tribe, Urg is gatherer. Grog and Urg are familiar with birds and the bees, and are expecting baby Grug in many moons.
Suppose, in waiting for Grug to arrive, things between Grog and Urg deteriorate. Grog begins long think about Grog, Urg, and baby Grug.
First, picture a version of Grog that “thinks at the margin”. His decision to leave the cave (and Urg) would be made by weighing the benefits of leaving Urg against the costs of leaving Urg. Among the benefits may be more disposable income, and possibly more leisure time. Among the costs would be the possible disutility of living alone, and the uncertainty regarding baby Grug’s survival. If Grog doesn’t care much about baby Grug’s life expectancy, and therefore doesn’t care about passing his genes to offspring, then his assessment may be that the benefits of leaving Urg exceed the costs of leaving Urg, and they separate. Grug is raised with one less guardian. If baby Grug really does have a lower probability of survival with one less guardian, then the probability of Grog’s genes surviving in the long-run is low if he thinks this rationally.
Let’s now imagine a Grog that falls for the sunk cost fallacy. This would be Grog looking back on all the time invested in the relationship. The time spent fending off wolves from cave to protect sleeping Urg, the times Grog went without food so that Urg could eat, etc. These sentiments make him overestimate the cost of leaving, at the margin. So Grog continues to invest in the relationship: continues to fend off wolves, give Urg his gruel, etc. The result of his continued investment in the relationship is that Grog and Urg do not separate and raise baby Grug together. So, it’s possible that Grog falling for this sunk cost is really his lizard brain trying to commit Grog to propagating his genes- the child has a better shot at making it past infancy with both Grog and Urg together. So he falls for the sunk cost fallacy and baby Grug has a better shot at survival. Furthermore, Grug has better odds at surviving long enough to have a child and pass the genes responsible for this behavior to the next generation. Probability of Grog’s genes surviving in the long-run is higher than the case above.
So what is this story really saying? Alexander McQuoid paraphrases my story in the following way:
Sunk cost fallacy helps overcome inefficiencies associated with externalities? If I only think about my private forward looking utility, I will systematically undervalue actions that impose externalities on the community, which leads to long-term inefficiencies that can lower my ability to pass on my genes? So systematic selection of genes based on those that make decisions thinking about sunk costs, not just on the margin.
This blog is a therapeutic outlet for me to write about life on the tenure track in economics.