Back in the ’80s, I worked for a small software company that was purchased by Xerox. As part of that deal, Xerox agreed to fund a vacation for our company (now a Xerox Division of about 30 employees) to Hawaii. As plans were being made, the word came down from the corporate office: we’d have to fly there in two separate groups. It was Xerox policy: the idea was to avoid wiping out the entire division in the event of a plane crash. By splitting us up, at least half the division would survive.
Yes, but, there’s a downside. By requiring two flights, you’ve just doubled the chance of a crash. Basically Xerox was choosing between two scenarios:
- Very small chance of utter disaster (all 30 employees killed), or
- Double that chance of a not-quite-as-bad disaster (15 employees killed).
From a corporate point of view, does choice #2 make sense? To judge this, you’d have to make a couple of estimations:
A. What would the division would be ‘worth’ if all 30 employees died?
B. What it would be worth if half of them died?
Since scenario #2 is twice as likely as #1, #2 is the right choice if B is at least twice as much as A. How do you estimate A and B? One line of thinking is, “With everyone dead, the division is worthless, so A = 0. With half the employees alive, the division has some value, so B is more than two times A”. Or maybe they broke down the value of the division with no employees (mostly, software assets), versus the value of the employees, and the likelihood that other Xerox employees could fill the roles of dearly departed, etc. It could be quite a task to work it all out.
My guess is that this is not how this policy was decided. Instead, someone in the corporate office said “let’s not put all our eggs in one basket” and everyone nodded.
Of course, we haven’t even considered the case where both planes crash…
