Yesterday was a day of widely shared feelings, and one of the most widely shared was the urge to retaliate. Man-in-the-street interviews revealed it in coarse form—"I want revenge," a woman told the Washington Post—and even some pundits didn't bother to put much varnish on it. "We must pulverize them," William Safire wrote in the New York Times this morning.
The first problem with retaliation is the practical one that many have noted: identifying and finding the attackers (though Safire doesn't find this hurdle as high as some: "Lashing out on the basis of inadequate information is wrong, but in terror-wartime, waiting for absolute proof is dangerous.") The larger problem with retaliation—at least, retaliation of the sort that most people envision—is that it will make terrorist attacks on the United States more likely in the future.
Yesterday someone asked me to discuss terrorism in game-theoretic terms, and I realized that, in this case, you almost can't. Game theory assumes that all players are amenable to positive and negative reinforcement. When you're dealing with people who don't mind death—who in a sense even welcome it—your arsenal of negative reinforcement shrinks considerably.
Indeed, killing Islamic fundamentalist terrorists (which the perpetrators almost certainly were) can be not just ineffective, but counterproductive. If death in a holy war is a ticket to the highest echelons of heaven, then the people we kill become not just martyrs, but role models. Or, at least, they become martyrs to many and role models to the small but consequential number of fundamentalist Muslims who themselves aspire to martyrdom.
Safire today derided President Clinton's response to the African embassy bombings—lobbing cruise missiles into a terrorist camp in Afghanistan—as an ineffectual and "demeaning pretense." But it was worse than ineffectual. Does anyone doubt that the incident gave Osama Bin Laden five eager recruits for every one of his soldiers who died? Now Safire and many others want to do the same thing on a larger scale.
Don't get me wrong. If Bin Laden is indeed behind this, then he should be either killed or put on trial. There is a difference between Islamic terrorists and major financiers of Islamic terrorism, and the latter are more amenable to game-theoretic logic. The number of people in the world who are in a position to fill Bin Laden's shoes is small, and I doubt that any of them welcomes death.
Still, how we go about bringing Bin Laden to justice (assuming, again, that this is his work) will massively influence how safe Americans are in the decades to come. One thing yesterday's attack did is give President Bush great but temporary influence in the shaping of international anti-terrorist norms and institutions. The NATO allies, and many other nations, will in the coming weeks show inordinate assent to his initiatives in this area. But if his first initiative is to launch a unilateral assault on Afghanistan, that political capital will have been mostly spent, with few if any good long-term effects and some clear-cut bad ones.
Consider an alternative. Bush declares that the Afghan government is morally obliged to turn Bin Laden over and that, if it doesn't, it will risk military attack and occupation—and its leaders will themselves risk being either killed or put on trial for complicity in murder. He asks for support from the international community—including military support from NATO in the event of a war with Afghanistan. And he puts all of this in the proper rhetorical context: He is not just retaliating, but rather setting the kind of precedent that the entire world needs to set as we approach an age when terrorists will have nuclear and biological weapons.
In all likelihood, Bush would get military backing from NATO, and, more generally, sufficient international support to help turn the entire exercise, however bloody, into a precedent of lasting value. (Note: If he wants to avoid the bloodshed, he could whisper this threat to the Taliban before announcing it to the world; once public, the threat becomes psychologically and politically harder for Afghan leaders to succumb to.) In fact, as Anthony Lewis suggested this morning (in a column that was a paragon of reason in the face of crisis and nicely counterbalanced Safire's column on the other side of the op-ed page), Bush should seek forceful support from the United Nations; Russia and China, as Lewis noted, have no interest in sustaining terrorism.
Yesterday Bush said there will be no distinction made between terrorists and the governments that harbor them. This is a valid principle. But the choice he faces is whether to make this an ephemeral talking point that accompanies unilateral American action or a lasting international norm.
Indeed, if you want to think really long term, you could imagine this norm further evolving into a principle of international law that is truly enforceable. But of course, this train of thought could lead to discussion of an international criminal court, and even after yesterday this administration probably isn't prepared to countenance such a thing. For now I'll settle for a simpler goal: Don't do anything rash, and don't do anything unilateral unless our friends desert us.
Any military action—including the one I've described—will have a big downside: fomenting Islamic radicalism, a radicalism that, at the grass-roots level, is simply not susceptible to normal deterrence. The object of the game is to outweigh that downside with an upside: 1) Deter the future financing of these radicals; 2) deter the future hosting of these radicals by state governments; 3) give the mechanism of deterrence the broadest possible base of geopolitical support, and hence the most enduring effect.