----------
In article , Jim Carriere
wrote:

Another term for that is "sunk costs" (or writeoff), which is not a good
analytical justification for continuing with any project. Political
justification, certainly, but it is similar psychology to holding on to
a bad investment (hoping it will eventually turn around). If a project
is a dud, there really is no sense going on with it no matter how much
money has been spent or how close it is to being complete.

That's not always true. Sunk costs can be an important consideration if the
replacement is going to cost more to develop than the existing program will
cost to fix. Put it this way--suppose you've sunk $8 billion into a program
that will require $4 billion more to fix. But if you determine that any
replacement is going to cost you at least $8 billion, then you are better
off continuing with the existing program--assuming that you have a
reasonable chance of succeeding.

Sunk costs can also represent assets, because they bought _something._ Most
of that might be blown on salaries, but at least some of it bought hardware.
Starting from scratch means junking those assets.

Yeah, it's not precise, but sunk costs serve as a crude benchmark for
comparison to the costs of starting over.




D