Ethics@Work: Assessing the value of a human life

Sometimes regulators must put a price tag on it – and it falls in the $5-10 million range.

What is the value of a human life? This is a question for philosophy and theology, to some extent for law (in wrongful death suits) but also a question for economics. We commonly assert that a human life has infinite value, and certainly few people would accept any amount of money in return for instant death (What would they do with the money anyway?). But the fact is that almost every person, almost every day, makes decisions that de facto place some dollar value on the risk of death or disability.
Did you save $3,000 by buying a slightly less safe car that gives you a one in a thousand greater chance of dying in an accident? Evidently, you value your life at around $3 million. Did you consider taking a job as a construction worker but decide that the extra $10,000 a year you could make was not enough to make it worth the 0.5% chance of a fatal accident? Evidently you think your life is worth at least $2 million. Every time you decide whether to drive faster and arrive earlier yet increase your accident risk, you are weighing the chance of death against the expense of delay.
Regulators also have to value a human life somehow.
Many regulations save lives and all of them cost money, so someone has to decide on the balance.
Concentrations of toxins in the environment are a good example. If we had zero air pollution, no one would die from air-borne toxins but we would also revert to a pre-stone-age existence. But we also don’t want to give polluters carte blanche to poison the atmosphere.
A recent article in The New York Times pointed out that agencies in the US use a range of values for human life, most of them in the area of $5-10 million. This is the value that economists have deduced from studying the risk behavior of Americans in the type of everyday situations we described above. However, a closer look shows that there are important differences in the way these valuations are used.
The regulatory decisions are of three kinds. One is where the regulator has no choice but to decide for the public. The Environmental Protection Agency decides on safe levels for various toxins, taking into consideration the cost of the limitations; no individual has any control over this process and so it is necessary and fair that the agency should make the calculation.
In other cases, regulators help people make their own decisions. For example, the Food and Drug Administration decides on labeling requirements.
Providing information on the dangers of a drug costs money and can save lives. But do these lives have the same value? Consumers have access to information on drug safety. It is plausible that those who take the effort to look up this information are the most risk averse; those who need the FDA label evidently are less interested in investing valuable effort to minimize risks.
In yet other cases, regulators make decisions for citizens who are perfectly capable of deciding on their own. The Transportation Department has regulations for auto safety. Since every person can choose which auto to buy, and since safety information is widely available, it is hard to justify using the same valuation. Anyone who wants a car with a safer roof can choose to buy one. There is certainly some value to having the agency weigh the information on behalf of consumers, but not to the tune of millions of dollars a life saved.
Interestingly, the actual values used by these agencies varies according to this exact gradation.
The EPA, which must decide for all citizens, uses the highest value – $9.1 million. Next in line is the FDA, which helps consumers make their own decisions. They use a figure of $7.9 million.
Finally, the Transportation Department, which decides for consumers who are mostly capable of deciding for themselves, uses the lowest figure – $6 million.
So it seems that not only is science being used in the service of politics, but politics is also serving science and producing policy guidelines that take into account more involved considerations than econometrics has tackled.