Consider that lovely phrase cost/benefit. We're talking *perceived* cost/*perceived* benefit.
As far as TEPCO executives were concerned, the cost of protecting Fukushima Daichi
was enormous, while they could pooh-pooh the possibility of an earthquake which might
need such protection.
Such costs can be reasonably estimated, so perceived cost closely equals actual cost.
However, earthquake probabilities are much easier to dismiss, so it is easy to have
perceived benefit MUCH lower than actual benefit when the earthquake shows up.
Security costs have much the same problem. You can't say for certainty that someone
WILL find a way in if there is one,, so...
"Son, the guards we hire for our caravans look like a loss on the books. But the books
don't show the losses we'll take if we're hit by bandits."