Here, Posner turns to North Harvard Yard—and specifically, to Cogan Professor of Law and Economics W. Kip Viscusi ’71.

In a 2003 paper, Viscusi and economics grad student Joseph E. Aldy ask: how much income are we willing to forego in order to decrease our risk of mortality?

For example, how much of a pay cut would a police officer be willing to take for a less dangerous beat? How much are consumers willing to spend on smoke detectors or bicycle helmets?

By multiplying the cost of these precautions by the probability that they will avert an untimely death, Viscusi and Aldy seek to gauge “the value of a statistical life.”

The authors conclude that “the value of a statistical life for prime-aged workers has a median value of about $7 million in the United States.”

So is the price-tag for life on earth equal to the Viscusi and Aldy figure—$7 million—multiplied by the global population (roughly 6 billion)?

Not quite, Posner explains.

First, Viscusi and Aldy’s method would produce a much lower estimate for “the value of a statistical life” in a Third World country, where workers might—out of necessity—accept much greater risks for higher pay. (Many will find this observation so repugnant that it renders the entire “statistical life” approach illegitimate.)

Second—and this is perhaps Posner’s most perplexing point—we tend to value our lives less as the risk of mortality grows more remote.

Princeton philosopher Peter Singer argued in a recent New York Times review that Posner’s argument on this point is “bizarre.” But upon closer examination, it seems rather intuitive.

As Posner notes, it would take quite a large wad of cash—perhaps many billions of dollars—to convince most of us to engage in a round of Russian roulette, where the probability of death is one in six.

Yet at some point, we are unwilling to pay any more money for a slightly better safety belt. That doesn’t mean we place zero value on our lives. But ultimately, Posner writes, “the difficulty…that people have in grasping miniscule probabilities would cause them to ignore the risk completely and thus demand zero compensation for bearing it.”

According to Baird Professor of Science Gary J. Feldman, “the public is very poor at risk analysis, let alone risk-benefit analysis.”

“[R]isk-benefit analysis is only useful if the risk is non-zero,” Feldman wrote in an e-mail. “But in the strangelet scenario, “the risk is sufficiently close to zero that risk-benefit analysis is not helpful.”