The Environmental Protection Agency is suddenly queasy about putting a price on life. Like other government agencies, the EPA uses an estimate of the value of Americans’ lives in order to assess the costs of regulations against their benefits.
The government estimates the value of life by observing what precautions Americans take. If a home loses 10% of its value for every mile it draws closer to a dump laced with carcinogenic waste, home prices can be used to estimate how much people value an increase in the odds of dying of cancer, and thus the price homeowners put on life. This estimate will help the EPA decide whether to clean up the site. It is worth it as long as the cost per life saved is no more than the price homeowners put on their own life.
This kind of calculation is inevitable in a society that must allocate resources between competing wants: do we build the extra firehouse or decontaminate the Superfund site or invest in a new maternity ward? If you were running Medicare you would want to know how far to go to add a year to the life of a patient. Is a $100,000 per year treatment worth it? Should you fund a $1 million therapy? What if it costs $10 million?
But despite the soundness of the principle, people get queasy about the idea that the government is putting a price tag on their life –especially since that price tag is substantially lower than infinity, their preferred valuation. In response, the EPA is proposing to tinker with the terminology, replacing “value of a statistical life”, the standard term used in these calculations, with the more antiseptic “value of mortality risk.”
Maybe the new term will set our mind to rest.




Some different term is appropriate. What we are talking about is not the VALUE of a human life — although economists habitually conflate value and price — but some kind of COST. Risk appropriately captures that, although I feel the term “risk” is overused.
I would say something like “The cost of preventing a probable highway fatality” and “The cost of preventing a probable death of a bystander shot by crossfire in a drug shootout” and “The cost of preventing a probable death by drowning in a pool” and so on. This doesn’t erase the implication that if the cost of preventing a probable pool-drowning is much less than the cost of preventing a probable plane-crash-death, then we should spend the money on prevention of pool-drowning because we could save more lives for a buck.
(Implicit is the assumption that all lives are equally valuable to protect — I won’t comment — as well as a unidimensional projection implicit in the word “cost” — as if all costs can be compared directly, something else that bears questioning.)
On the subject of value versus price–remember there are N values (and it’s dubious whether an individual’s _value_ of an item comes from a totally ordered set, as Theory says it should) — but only one Price at a time. (P_t) which does come from the incomplete field (cuerpo) of floating-point decimals.
As long as I’m meandering: here’s another thing to consider. Perhaps instead of insurance premia we should estimate “the value” of a human life by how much _other people_ would be willing to give up to keep someone alive. On this score Scrooge’s life would be worth less than — name escapes me right now — the othe rmain character in A Christmas Carol.
It doesn’t seem like you are distinguishing between the values of the different (prob dist of) lives that would be saved. Or do you distinguish?
The distribution of values doesn’t play into this post. But it clearly plays into the public policy debate. Remember when an IPCC analysis (can’t remember which report now) of the impact of climate change assumed lives in the developing world were worth 1/10 of lives in the developed world? It didn’t end well.