tolerance of catastrophic failure, when groups of people work on complex problems

Comment on Arnold Kling's blog Econlog: "My Ideas on Health Care Delivery"

> The autonomous, self-directed doctors produced by our medical schools are not suited to treating complex patients. Instead, what we need are team players, implementing consistent corporate policies.

This is a _hard_ problem. I am management for a engineering design and manufacturing company. Even with the authority to fire for insubordination, people enjoy being "autonomous, self-directed", and do so against rational self-interest.

Complexity, handled by groups of people, is regularly punctuated with catastrophic failure. Catastrophic failure is only avoided by systematically eliminating _EVERY_ excuse for _ANY_ failure (to the extent you can). This kind of discipline is in very short supply, because it is not usually rewarded in human society.

Human being are _very_ tolerant of catastrophic failure, when it happens to others. And it usually happens to others, and, so, they are usually _very_ tolerant.