I’ve heard that the Transportation Department, and possibly other agencies such as DHS or FDA, calibrate their safety efforts by considering what it will take to limit deaths to a set number, e.g., 1,000 highway fatalities per year nationwide. Agencies thus assign a monetary value to each life. Is this true - does the government use specific, predetermined numbers in making the inevitable tradeoff between safety and cost? If so, what are those numbers (expressed either as the monetary value of a life or the number of “acceptable” deaths) for various agencies and how are they determined?
Here (.pdf) is an old example of a US Department of Transportation number and its explanation.
Generally it works the other way, or is supposed to. You come up with your best guess for the VSL (value of statistical life) then put that into your benefit/cost calculation and fund those programmes or implement those regulations that pass the test.
There is considerable debate about various things in these matters. The last time I looked, the appropriate value is thought to lie between US$2m and $9m for the US.
To the best of my knowledge, not exactly. Rather, there are formulae for determining degree of risk. In other words, if you have $10 billion to spend on highways, and $25 billion in proposed highway construction nationwide, you calculate which projects are most needed to decide which to allocate the $10 billion to. Inevitably, that means putting dollar values to things like human life and suffering, not in a callous sense, but more in the sense in which a jury decides a damages award in a personal injury lawsuit.
Similarly, probability figures are used in connection with human life. If there’s a limit to what the Army Corps of Engineers can do, the projects that have the greatest potential to save lives are the ones that get funded – knowing that there’s a likelihood that 10 people will die and $50 million in damages will result from floods on river X project that doesn’t get funded, but the 35 which will probably die and the $120 million in probable damages on river Y project make it the better choice to fund.
Cost-benefit ratios sound terribly callous, until you realize that in practical terms you need to compare apples, not merely with oranges, but with the literary value of Ulysses (to pick something at random to suggest how truly incomparable these things can get). Another New Madrid earthquake or an asteroid impact would cause incalculable death and damage, but are relatively low probability, compared to a San Andreas earthquake or a Gulf Coast hurricane. You need to put your money where it will do the most good – and in doing so, you may be forced to put dollar values on human lives in order to make valid quantifiable comparisons.