Judge Richard Posner, University of Chicago Law School:
"The U.S. invasion of Iraq, the U.S. decision not to invade Afghanistan before the 9/11 attacks, and concern with the apparent efforts of Iran and North Korea to obtain nuclear weapons raise acutely the question when if ever a preemptive or preventive war is justified. If "preemptive war" is defined narrowly enough, it merges into defensive war, which is uncontroversial; if you know with certainty that you are about to be attacked, you are justified in trying to get in the first blow. Indeed, the essence of self-defense is striking the first blow against your assailant.
But what if the danger of attack is remote rather than imminent? Should imminence be an absolute condition of going to war, and preventive war thus be deemed always and everywhere wrong? Analytically, the answer is no. A rational decision to go to war should be based on a comparison of the costs and benefits (in the largest sense of these terms) to the nation. The benefits are the costs that the enemy's attack, the attack that going to war now will thwart, will impose on the nation. The fact that the attack is not imminent is certaintly relevant to those costs. It is relevant in two respects. First, future costs may not have the same weight in our decisions as present costs. This is obvious when the costs are purely financial; if given a choice between $100 today and $100 in ten years, any rational person will take $100 now, if only because the money can be invested and through interest compounding grow to a much larger amount in ten years. But the appropriateness of thus discounting future costs is less clear when the issue is averting future costs that are largely nonpecuniary and have national or global impact.
Second, and more important, and well illustrated by the failure to find weapons of mass destruction in Iraq, if the threat of attack lies in the future it is difficult to gauge either its actual likelihood or its probable magnitude. But this is not a compelling argument against preventive war. What is true is that a defensive war is by definition waged only when the probability of an attack has become one; the attack has occurred. The probability of attack is always less than one if the putative victim wages a preventive war, because the attacker might have changed his mind before attacking...." [December 5, 2004; The Becker-Posner Blog has more]