We all know the old saw: you can't put a price on human life. Like so much conventional wisdom, it is nonsense.
Say that an all-powerful being walks up and offers you a trade. You have two options: you can either (1) permanently satisfy the basic needs of everyone on the planet or (2) save one life. Which do you choose? Obviously (1). Yet, contrary to the old adage, your decision effectively puts a price on life. It's an extraordinarily high price, involving wealth that may outstrip the current production of our planet, but it's a price nevertheless. Although I have no idea what the equivalent price in dollars would be (it depends on how you define "basic needs," for one), $1 quadrillion is probably a decent estimate.
Now the being offers you a different trade. You can either (1) pay one dollar or (2) save one life. Here the moral imperative is equally clear: you should save a life. One dollar can't even buy a 20-ounce bottle of Coke at the new vending machines near my room; certainly it's a bargain for saving anything as precious as a human life.
But note that these two opposing answers imply an inevitable conclusion. Somewhere between $1 and $1 quadrillion, you will no longer be willing to spend the money. This is the price you place on human life.
Admittedly, I'm a little ambiguous about what a "human life" is. There is a world of difference between saving a healthy 10 year-old child and extending the life of a comatose 120 year-old by a day or two. But even if you place different levels of importance on preserving different lives, this thought experiment is still valid for any individual life. For that 10 year-old child, you'll be willing to pay $1 but not $1 quadrillion; again, somewhere in between, your decision-making will implicitly set a price on life.
Just a theoretical curiosity? Not at all. By refusing to acknowledge that some price on life is inevitable, we use inconsistent values in policymaking and thus cause unnecessary deaths. If we neglect to pursue a $2 million per life venture while doggedly maintaining a program that costs $10 million per life, we are wasting money that could have been used to save more people. Professor John Graham -- now working with Bush's OMB -- refers to this practice as "statistical murder." While I do not agree with all Graham's work, I cannot imagine a more appropriate term.
Meanwhile, this lesson is only a reflection of a broader, critical principle: all tradeoffs are quantitative. If having better X makes Y worse, any rigorous analysis of the proper balance between X and Y must rest, at some level, on a numerical comparison. After all, if you can get a massive improvement in X for a slight decline in Y, you are likely to opt for X; if the situation is reversed, you'll choose Y. Again, somewhere in between, there is a balance at which you're indifferent between improving X and Y. To find this balance, you must combine your best estimate of the tradeoff between X and Y with your preferences about the relative merits of X and Y. This will involve a comparison between two magnitudes; the only question is whether it is fuzzy or explicit.
Now, am I saying that there is a clear, mathematical way of resolving all policy disputes? Of course not. We will still place different values on different objectives. While I'd probably say that freedom and material prosperity are the most important policy goals, an Islamic fundamentalist might claim that widespread reverence of the Qur'an is vastly more important than anything else. Math cannot resolve our disagreement. It provides the best way to make decisions about policy after our objectives are set.