In economic analysis, we define past costs as historical costs that have occurred for the item under consideration. Then, we define sunk costs as past costs that are unrecoverable.
The sunk cost fallacy is using sunk costs to influence decisions. Basically “throwing good money after bad” and refusing to “cut your losses”.
Jason Rock had this analogy: imagine one person had a 0 movie ticket. Both go watch a really bad movie. The person with the 100 ticket stays because they spent money on it, even though they could be doing something more productive with their time.
Analysis
The main important part about our analysis is that sunk costs shouldn’t influence our decisions, because they’re unrecoverable. This means that if we’re mid-project, the only thing that should influence how we assess alternatives is their future benefits and future costs.
Server example
For example, suppose we have a server purchased for 10k after 5 years (economic life).
- The estimated yearly depreciation charge is: .
- The price of the service provided by the server should take the depreciation charge in mind, so that over the economic life, it recovers the capital cost (depreciation) of the server.
Then, suppose the salvage value in reality turned out to be $2000.
- The original depreciation charge was 7600.
- The yearly difference can’t be recovered, and is therefore a sunk cost of .