Cass Sunstein has written a superb analysis of how cost benefit analysis can usefully be applied to extreme situations - worst case scenarios. He takes up two - mostly climate change, but then compares it to counterterrorism. My comments below focus on the application of cost benefit analysis in responses to terrorism; I admire Sunstein's analysis, but think it has limitations in its application to terrorism.



Regnant approaches to terrorism are driven not just by narrow cost benefit analysis, but by a still narrower focus on something we might call "event-specific catastrophism": preventing the next attack. This is as true of the Bush administration as of its leading opponents. What has the Bush administration focused upon, in speech after speech to the public? The imminence of the next attack, and the need to prevent it. One hopes this is mobilizing rhetoric for larger policies against jihadist terrorism, but in considerable part, the uncertain next attack is the focus of policy - a long-term strategy, if one can call it that, even after seven years, of just trying to make it to the next day unscathed.



Despite much discursive rhetoric about long-term policy and the war on terror, much US policy is what, in a strategically informed plan, might well be considered the very last defensive perimeters. Airport security, daily monitoring of cell phone traffic, internet analysis in hopes of seeing spikes that might indicate imminent terrorist action, watch lists, and many, many cement barriers. Presumably no one in Britain is reassured by the fact that the Glasgow attack was prevented not by perceptive police work, nimble intelligence agents, deep penetration of homegrown terrorist cells - but simply by a physical barrier at the airport. But perhaps people are comforted; the cement barrier worked, after all, effectively and cost-effectively, while the rest of the counterterrorism apparatus, at enormous absolute and relative cost, did not. Still, these are fundamentally defensive measures aimed at preventing the next attack, counterterrorism in a vital but stiflingly narrow sense. The cost benefit analysis underlying such planning, shaped toward event-specific catastrophism, is necessary and fruitful, but bears little resemblance to planning or conducting a "war" on terrorism or, really, any strategic conceptual response to jihad that goes beyond preventing particular events of uncertain probability and magnitude.



Yet the Bush administration's legion critics ridicule US counterterrorism policy in great measure within the same narrow framework that the administration has used. Sometimes the cost benefit analysis would scarcely past muster in an undergraduate economics class - political scientist John Mueller, in his bestselling Overblown: How Politicians and the Terrorism Industry Inflate National Security Threats and Why We Believe Them, or journalist James Fallows, each breezily announcing that the chances of getting killed in a terrorist attack are less than getting struck by lightning, or that 9-11 killed 3,000 people but 40,000 Americans die a year in automobile accidents and, ergo, well what? Cost benefit comparison of opportunity costs makes sense only if comparing genuinely apposite, available opportunities.



Serious cost benefit analyses are offered in criticism of US policy, and Sunstein's Worst Case Scenarios is the best of them. It calmly and restrainedly demolishes, for example, the so-called "One Percent Doctrine" - Vice-President Cheney's assertion that even a one percent chance of a catastrophic terrorist event requires a response as though it were a complete, 100% certainty.



Worst-Case Scenarios is not primarily about terrorism; it is mostly a discussion of cost benefit analysis applied to climate change and the limitations of the `precautionary principle' (of which the One Percent Doctrine is a variety). Sunstein uses terrorism as a striking, but adjunct, case study in order to illustrate, among other things, what the behavioral economists call `salience' - the contrast between the Bush administration's all-in commitment against terror, even on a one-percent threat basis, but its unwillingness to adopt similarly drastic principles with respect to the possibility of another kind of catastrophe, one that is not, however, as immediately visible and `salient' to the general public.



In using this splendid book here, in a discussion strictly limited to terrorism, I want to be clear that its arguments over the nature of cost benefit analysis, its critique of the precautionary principle, and its careful application of them to climate change bear close, repeated reading on their own terms and the intellectual ends to which Sunstein puts them. Solely with respect to countering terrorism, however, Worst-Case Scenarios correctly points out that not even `all the instruments of the national will' (what President Bush solemnly committed to the fight against terrorism after 9-11) are unlimited. They never are. Choices still have to be made and priorities established and, as Sunstein acutely observes, preventative actions bring risks of their own. Countering every one percent risk as though it were a 100% certainty creates many more one percent risks of their own, and countering all those one percent risks will preclude many, if not all, of the others. The One Percent Doctrine, and the precautionary principle which it instantiates, is finally "incoherent," as Sunstein says. It induces simultaneously a demand for actions of many contradictory kinds, and yet precludes them at the same time: paralysis in every direction, demands for action and inaction. It cannot serve as a universal basis for policy, and no one, to my knowledge, has made this argument better.



Nonetheless, even sophisticated cost benefit analysis of the kind found in Worst Case Scenarios tends to take the prevention of particular events as the fundamental analytic objective. There is, to be sure, an important political reason for this. The American public has been gradually downgrading terrorism as a political priority, even while continuing to say (in the fine democratic tradition of wanting your cake and eating it too) that it supports serious measures against it.



American elites, for their part, have been sliding to a dismissively contemptuous view that questions the whole idea of counterterrorism as a serious, large-scale necessity. The threat is downgraded, deploying cost benefit style arguments to call the administration's counterterrorism programs trumped up and exaggerated, and to suggest that the terrorist threat is quite capable of management without special military or intelligence measures. Leaving aside the frequent starting assumption that the Bush administration has illegitimately grabbed executive power and that this, rather than terrorism, is the primary thing against which to protect, the fundamental factual claim is that the probability of a successful attack has been seriously exaggerated.



How to interpret, in other words, the fact that the US has not been hit on its territory since 9-11: evidence of the effectiveness of the anti-terrorism efforts, or evidence that it was always more chimerical than real? Ultimately that is what Sunstein's very fine book, at least as applied to terrorism, seeks to answer and to say, in the absence of being able definitively to answer, how to respond. Still, this is ever a relentlessly tactical approach, and the most important criticism of it is that over the long term, we seek something more comprehensive. Even when not event-specific - even when it takes, for example, "Islamist terrorism" as the whole scenario for consideration - it is by its very nature reactive. Cost benefit analysis does not propose solutions; it evaluates proposed solutions offered by other processes. It is not finally a strategic form of thinking. And that is a problem.