How to put a price on a human life

Pause on a clifftop, fill your lungs with cold Atlantic air, and the breath can feel restorative, almost priceless. Cleaning the air the rest of us breathe is anything but free: factories retrofit scrubbers, dirty industries shrink, households swap coal for cleaner fuels. On the other side of the ledger sit fewer asthmatic children and fewer hospitalisations among the old and the already ill. Cost-benefit analysis demands a number on both sides; so how do you attach one to a death not yet suffered?

America's Environmental Protection Agency has decided it would rather not. Last month the EPA announced that, in assessing future rules, it will no longer assign a monetary value to clean-air health benefits, citing irreducible uncertainty in the figures. Sceptics read the move as a quiet attempt to water down regulation by stealth: as the adage goes, what is not counted does not count, and refusing to grapple with a number is itself a way of discounting it to zero. Squint at the announcement and a ghoulish irony emerges — the agency's reasoning now echoes the very objections once raised against pricing exercises altogether.

The pricing exercise dates back to a 1981 executive order signed by Ronald Reagan, whose aim was to cut red tape whose costs could not be clearly justified by its benefits. Murray Weidenbaum, then chairing his Council of Economic Advisers, argued that money lavished on rules saving only a handful of lives might fund better cardiac wards. Democratic critics, sensing a ploy to dress up deregulation as analytical rigour, retorted that no life could be reduced to a dollar figure. Four decades on, that protest reappears almost word for word in the mouth of an environmental regulator.

The EPA of the 1980s reached instead for the value of a statistical life — a figure quite distinct from any individual's worth. An earlier attempt, the human-capital method, drew on actuarial habits, multiplying expected wages by the working years still ahead. Wages, the reasoning ran, are the price at which someone sells their finite hours; scale them across a lifetime and a market value emerges. The approach fell out of favour once economists noticed its corollary: anyone without a wage — the elderly, the chronically sick, parents who eschew paid work to raise children — was implicitly priced at zero.

Thomas Schelling, who took the 2005 Nobel in economics, diagnosed the deeper problem: valuation depends on whose life is in view. The named child in a fundraising appeal commands almost limitless generosity — let a six-year-old need an operation by Christmas, he wrote in 1968, and the post office will be swamped with coins. Ask the same donors for a higher tax to lower mortality among an unnamed slice of the population, however, and few will open their pocketbooks. Yet it is precisely that anonymous statistical life — possibly their own — that regulatory policy must learn to price.

Schelling's resolution was to read the price off choices people already make: drivers weigh a safer car against its added cost, soldiers weigh hazardous postings against bigger paychecks. Observe what a population will pay to shave a known fraction off its risk of death, and the arithmetic does the rest. If 100,000 people each spend $100 to avoid a one-in-100,000 chance of dying, $10m has bought one expected life. The EPA's working figure — $7.4m in 2006 dollars, roughly $12m today — comes from such studies, including one tracing how US Army signing bonuses shifted with battlefield fatality rates.

Britain has been less squeamish still. Its National Health Service rations treatments using quality-adjusted life years, which borrow the same logic but reweight each remaining year by some measure of utility. A drug prolonging life in chronic pain may earn fewer such years than a less efficacious alternative leaving the patient comfortable; all else equal, shielding statistical children outranks shielding statistical pensioners.

None of these methods answers the unanswerable. Each, however, insists that the value of a life is plainly not nothing: people prefer to remain alive, and they will pay, in money and inconvenience, to stay so. Regulators need a benchmark — imperfect, contested, occasionally absurd — against which to weigh any decision's costs. Economists are routinely accused of an anti-social coldness for asking how much a life is worth; the EPA, in declining to ask, has chosen the warmer-sounding option, and also the worse one.