Aroa:
And who dictates what 'safety' you should get ?
That has been amenable to mathematical calculation for at least Thirty years! Insurance companies do it all the time.
1. Calculate the cost of an aircraft accident - take a sample, work out the actual costs of a variety of accidents and average them out. You CAN put a price on human lives, when I was working in an oil terminal mid 70's the figure was $4 mllion.
2. Now that you have the cost of an accident say for a light aircraft, regional jet, jumbo, etc. Multiply by the probability per year. Say for a jumbo, once every ten years - annual probability of 10%.
That figure in dollars is the expectation cost associated with that event in todays dollars (NPV - net present value). This figure is negative, it is the notional cost of accidents each year.
3. Compare safety activities with their associated budgets and expected impact on the accident rate by calculating the expectation. Use discounted cash flow on the budget to arrive at a cost in todays dollars(NPV).
4. If the impact of what you propose has a positive NPV - in other words it reduces the overall cost of accidents, then the safety measure is worth doing. If the NPV is zero its a worthless idea - no value. If its negative, can it.
For example, you can use this tool to compare things like training, ADSB, CTAFs', rewriting regulations, etc.etc.etc.
Any actuary can show you how to do this. They do it every day to set insurance premiums. I could write an example if pushed.