- How many people would be injured or killed by the event.
- How much financial damage the event would cause.
For example, if nobody ever dies, and 10 million in damage/problems occurs each time the event does, then it's almost certainly not worth spending 1 billion dollars to to fix the problem.
Obviously, those numbers are ridiculous, but the point is there. The numbers need to be considered.
And yes, human life is a number, too. It's just harder to quantify because there's a lot of opinion involved. But if it the expectation is that such an event will kill 1 person each time, and it would cost 1 Trillion dollars to fix/prevent... it's almost certainly not worth fixing. It's harsh, but spending a Trillion dollars for a single life would be unreasonable.