How do you put a price on a human life?

How much is fresh air really worth? Standing on a mountain top, breathing deeply, you might say the answer is “everything”. Yet cleaning up dirty air costs real money. Factories must buy new equipment, some industries may need to shrink, and households need cleaner ways of cooking and heating. The benefit, on the other side, is that fewer people get sick from breathing polluted air, especially children, the old, and those already ill. Deciding how much to spend on clean air is therefore a classic problem of cost-benefit analysis, in which one side of the argument must be weighed against the other. But how do you measure a benefit that means avoiding sickness or early death? Pricing fresh air becomes an even harder question: how do you put a value on a human life?

This is a question that America’s Environmental Protection Agency (EPA) has now decided to stop asking. Last month it announced that, when judging new pollution rules, it would no longer try to put a price on the health benefits of cleaner air. Officials argued there was simply too much uncertainty about what those benefits are worth. Critics, however, see this as a quiet way to water down regulation. As the old saying goes, what is not counted does not count.

The American government has required this kind of analysis since 1981, when President Ronald Reagan ordered federal agencies to weigh the costs of new rules against their benefits. His aim was to cut red tape that, in his view, cost businesses more than it was worth. If a pollution rule saved only a few lives but spent a great deal of money, the argument went, the same money might save more lives elsewhere — funding hospitals, for example. Critics replied that no honest number could be placed on a human life. Oddly, today’s EPA sounds rather like those critics from forty years ago.

Earlier economists had tried to solve the problem with what became known as the human-capital method. A person’s life, they suggested, could be valued at the wages they would have earned if they had lived. Wages, after all, roughly show what people are willing to accept for the hours of their working lives. The approach was eventually rejected, because it gave a value of zero to anyone without a paid job — retired people, the sick, and parents caring for children at home. Lifetime earnings, it turned out, are not the same as a lifetime’s value.

A better idea came from Thomas Schelling, who later won a Nobel prize in economics. Schelling pointed out that we all take small risks of dying every day. Driving a car, for example, carries some chance of a fatal crash, and spending a little more on safety features lowers that chance. By measuring how much people are willing to pay to reduce their risk of death by a tiny amount, economists can work out a figure called the value of a statistical life. If 100,000 people each pay $100 to avoid a one-in-100,000 chance of dying, then $10 million has effectively been spent to prevent one expected death. The figure does not say what a life is worth; it shows what people will pay to protect it. The EPA itself has long used a figure of around twelve million dollars per statistical life.

None of these methods give a perfect answer, and perhaps no method ever will. Yet the value of a life is clearly not zero. People generally want to stay alive, and they are willing to pay something to do so. Regulators need a benchmark, even an imperfect one, against which the costs of their decisions can be measured. Economists are sometimes accused of cold-heartedness for asking how much a life is worth. Refusing to ask the question, however, would be far worse.