Publications
Publications
china leadership monitor
policy review
defining ideas
hoover digest
education next
June 2, 2008

Expanded Deterrence

Broadening the threat of retaliation


We are facing a threat that is catastrophic in its scale.1 The damage that even a single attack with weapons of mass destruction would wreak could  run into the millions of lives, and do egregious damage to American economic, political, and social structures. There is no graver threat to the United States.

This threat is only going to get more serious. The progress of technology and the increasing interconnectedness of global systems are driving both productive and destructive power down, to lower and lower levels of agency, and outwards, to the fringes of society. Accelerating advances in computing, biotechnology, nanotechnology have democratized destructive power — up to the point at which a single individual may have the power to do enormous damage.2 Today we see this peril most plainly in the justified fears about the use of the first and greatest absolute weapon — the nuclear bomb. But the threat of biological and biotechnological weaponry, powered by the highly diffused and swiftly advancing progress of the life sciences, may be even graver. Similar dangers are growing in the fields of nanotechnology, computing, and the like.

The proliferation of massively destructive technologies can and should be retarded, but it cannot be prevented. We must accept both that the threat is very real and that it cannot be “solved,” only managed.

The United States has begun to respond to this grave threat through a “layered defense” that includes military, intelligence, diplomatic, political, public diplomacy, homeland defense, and humanitarian components. This policy commendably seeks to integrate all elements, hard and soft, of American and allied power to stave off disaster. And all elements of this layered defense are important in preventing attacks, including efforts to stem proliferation and “soft power” strategies designed to address real root causes of terror.

Proliferation of massively destructive technologies cannot be prevented. It cannot be “solved,” only managed.

Unfortunately, the current policy is insufficient. Prevention and defense, while clearly central, cannot alone address our problem. An attempt to prevent the proliferation of all potentially catastrophic technologies or to build effective defenses against them would be far too costly, both in terms of resources and political capital. More important, the advent of the nuclear weapon and now, in its wake, of comparably destructive technologies has finally decided the age-old struggle between offense and defense in favor of the former. It is not that defense cannot be effective; it is that, given the destructive power of the weapons in question, defense must be perfect. And no policy of prevention and defense can promise that.

Toleration is another critical element in our response. We balance the needs of a free society that can produce and distribute goods and ferry people efficiently against the costs of car accidents, plane and train crashes, and pollution. But, obviously, any significant incidence of catastrophic attacks cannot be tolerated, certainly not if we wish to maintain our society as we currently enjoy it.

Another pillar is the positive incentive structure associated with sociable behavior. The forces of habit, social conformity, ritual, and morality suffice to draw most people into the system. Unfortunately, terrorists are precisely those who are not susceptible to these positive incentive structures. While any effective counterterrorist or counterinsurgency strategy must include carrots, there is invariably a group of those resistant to any reasonable offers. Furthermore, some contests are zero-sum and therefore there may be an irresolvable tension between what some of our opponents want and what we can give.

Likewise, direct deterrence against terrorists is an important tool, and is the cornerstone of law and order both in the domestic and international contexts. But, as many have pointed out, terrorists are hard — and sometimes impossible — to deter directly. Clearly, people willing to kill themselves in order to conduct terrorist attacks are unlikely to be deterred by direct threats.

While these approaches are substantially effective, they leave a major gap in our ability to prevent catastrophic terror attacks. We can be neither omniscient nor omnipotent, and therefore prevention and defense are limited in a world where destructive power is spreading downwards and outwards. We cannot tolerate catastrophic attacks in a world of weapons of mass destruction. We cannot hope to convince all terrorists to behave through positive incentives when some are fanatics, cultists, or simply pursuing objectives the U.S. can never concede. And we cannot rely on direct deterrence when many are willing to sacrifice their very lives in pursuit of their aims. Our traditional approach to the problem is therefore insufficient to meet the danger.

Expanding deterrence

If there can be no escape from the presence of terror, then we must respect and adapt to it.3 The best approach to this threat would involve shifting the burden for preventing catastrophic attacks onto those who have the capability and moral responsibility to prevent them — effectively expanding deterrence. In doing so we would be engaging the assistance, through a combination of threats and incentives, of those whom a more narrow vision would consider unimplicated. This expansive but carefully and reasonably defined category would encompass not only those directly involved in a terror plot, but those individuals, governments, or other entities whose material support, cooperation, complicity, or gross negligence enabled an attack. (These terms, while they would need to be defined more specifically in a formal policy, have settled meanings in American law that could, with some modifications, be transferred to the international arena.4) This posture would strongly incentivize those with the capability to act to do so, since gross negligence or complicity would incur retaliation (not necessarily, it should be emphasized, violent in nature). And our demands would be reasonable, because all we would be asking for is active assistance in preventing catastrophic attacks from those who, despite their own involvement — active or passive — in such attacks, benefit from the restraint of our current, excessively narrow posture.

Willful inaction in the face of catastrophic threats against the United States is permitted to pass with impunity.

In effect, we would simply be bringing to bear the real correlation of factors, matching accountability with responsibility and capability. In today’s anachronistic and unbalanced system, willful inaction in the face of catastrophic threats against the United States is permitted to pass with impunity, despite our clear capability and moral grounds to increase the costs of such disastrously harmful behavior. An expanded framework of deterrence would rebalance the calculus of threats and duties to match obligation with responsibility, backed by the threat of credible retaliation.

This policy is necessary, moral, and hardly newfangled. The threat posed by a catastrophic attack is so great that we cannot allow any reasonable effort to go unpursued in the attempt to frustrate one. The people of the United States should not be the ones most concerned about a wmd attack on the homeland. Rather, the people most afraid of such an event should be those who have the ability and the moral responsibility to frustrate it. The policy would reset the conceptual parameters for accountability for frustrating such strikes and, in the event of an attack, broaden those understood to be appropriate targets of retaliation. Its ultimate object would be to inspire among the complicit and the negligent a considerably greater fear of American wrath, a fear that would outweigh whatever combination of fear and affection our terrorist enemies are able to inspire.

Such a policy would provide a fitting moral and strategic basis for dealing effectively with the catastrophic terror threat, rather than prescribing a certain set of actions. As Clausewitz pointed out, in war, formulating a suitable moral-strategic rationale is vital for setting appropriate objectives and pursuing them successfully. As he wrote, “Nothing is more important than finding the right standpoint for seeing and judging events, and then adhering to it. . . . Policy is the guiding intelligence and war only the instrument.”5 The expanded deterrence policy would set the appropriate moral and strategic paradigm for dealing with this grave threat.

Many elements of a deterrent policy against terrorists are in place already. Indeed, one might reasonably argue that, implicitly, an expanded deterrence policy already exists. It is taken as a matter of course that terrorist groups such as Hezbollah are the object of a deterrent strategy by the United States. The U.S. government’s response to al Qaeda’s attacks, while stymied by insufficient focus and resources, exhibits the marks of a latter-stage deterrent strategy: They attacked us, thus they will be destroyed. Operation Enduring Freedom and elements of the Administration’s National Security Strategy moved towards a deterrent framework. The U.S. has already expanded notions of culpability in terrorism with its campaigns against material supporters of terror, financial backers, direct action, threats against state sponsors of terror, and so forth. Among Democrats, Senator Barack Obama endorsed a deterrence approach against terrorists in his August 1, 2007 speech, threatening to strike terrorists within Pakistan to retaliate for or prevent attacks.6 More recently, in a little noticed speech on February 8, 2008 that represented a major step forward for U.S. counterterrorism policy, National Security Advisor Stephen Hadley announced that the U.S. had recently adopted “a new declaratory policy to help deter terrorists from using weapons of mass destruction against the United States, our friends, and allies.” This policy would threaten with retaliation “those states, organizations, or individuals who might enable or facilitate terrorists in obtaining or using weapons of mass destruction.” But the policy, while deserving of applause, remains inchoate and unpublicized. The argument presented here attempts to develop further Hadley’s commendable but incomplete proposal.7

The U.S. government’s response to al Qaeda’s attacks exhibits the marks of a latterstage deterrent strategy.

But it is important to be clear that a deterrent strategy confined within traditional boundaries will not be satisfactory. Certain terrorist groups will cross clear U.S. “red lines” and strike the U.S. homeland even in the face of credible threats against the group’s members from our government.8 This leads some observers to argue that deterrence is of no use against radical catastrophic terrorists — and there is a powerful logic to this critique, which is why a deterrent strategy that encompasses in its targeting only the terror network itself is insufficient. What this critique leaves out, however, is the fact that a deterrent strategy need not confine itself only to reacting against the core members of the network. Instead, we should not seek solely to deter those hardened terrorists who may be undeterrable (a debatable point), but also those others whose cooperation, complicity, material support, and gross negligence are vital for the core members to operate. Recognizing that terror networks themselves are difficult to deter, the policy would conceptually break entities down into subcomponents and seek to undermine them through threatening these constituent subgroups and/or members.9

Terrorist networks aren’t only bin Ladens. They are passport forgers, secretaries, financial supporters.

Indeed, this indirect approach may be especially effective against terror networks.10 Terrorist networks are not simply bin Ladens and al Zawahiris. They are also passport forgers, messengers, secretaries, financial supporters, recruiters, propagandists — a whole network of individuals cooperating together, at various levels of involvement, in an enterprise. And beyond the core membership itself, numerous individuals and organizations are specifically aware of the activities, location, and membership of a terrorist network; without the complicity of this broader group a network cannot survive, let alone thrive, as Mao aptly described. These people and organizations help a terrorist network to operate. Therefore they are, to varying degrees, culpable.

The logic behind expanding deterrence is that a threat total in scale to us should be confronted as expansively as necessary to prevent grave damage. Deterrent threats and retaliation thus should not be confined solely to those whom the enemy elects to put forward as “warfighters,” but should extend as far as reason and effectiveness allow, including those materially supporting, cooperative, complicit, or grossly negligent. Our terrorist enemy deliberately uses asymmetric strategies and tactics designed to frustrate the advantages of the modern nation-state, including by collapsing the distinction between combatant and civilian. If the prize we are fighting over is not a vital interest, then we may elect to accept these disadvantages and try to achieve our objectives under restrictive rules of engagement. But if the fight involves the risk of the loss of thousands or even millions of American lives, then it would be both irresponsible and immoral to be so cavalier. If we are not to be supine in the face of catastrophic terror, we will have to broaden our targeting.

How it would work

Reports suggest that al Qaeda is plotting wmd attacks against the West from its hideouts in the northwestern regions of Pakistan. The Pakistani Army has made halting, ineffective efforts to track them down. Many in the area are likely hiding or at least complicit in hiding members of al Qaeda. Yet, perversely, these individuals most likely either fear or sympathize more with al Qaeda than the faraway Americans and our apparently unintimidating  sometime-allies in the Pakistani Army. How does it make sense that it is the American people rather than the relevant inhabitants of the frontier provinces of Pakistan who must bear the risk of an attack launched by al Qaeda elements sheltered in this very region?

Expanding deterrence would redress this imbalance. Instead of permitting individuals and entities engaged in protecting or abetting terrorists to be bystanders, this policy would force  them to bear the risk of retaliation in the event of a catastrophic attack. This would strongly incentivize those potentially implicated either to take action on our behalf or at least not cooperate in hostile activities. The same logic would apply to those in the Pakistani Inter-Services Intelligence agency and other elements in Pakistan who have protected and fostered al Quaeda.

In the event of a catastrophic al Qaeda attack against the U.S. homeland, the United States would, having set out this policy, therefore have the moral and strategic rationale to retaliate not only against al Qaeda members themselves, but also against those whose cooperation, material support, complicity, or gross negligence made the killing of thousands of Americans possible. Legitimate targets of American retaliation would now include supporters, facilitators, moneymen, back office workers, and onwards, as well as infrastructure, housing, food and other supplies, land, political control over territory, marks of prestige, and so forth. Depending upon need and the degree of culpability, American options would include detention, capture, confiscation, disabling, humiliation, pressuring, all the way up to (but not necessarily emphasizing or even including) military operations.

As a matter of policy, there are several questions about the expanded deterrence approach. First, is the threat to respond so broadly to a catastrophic terrorist attack credible? Second, even if credible, would such a policy actually serve to deter catastrophic terror attacks? Third, do the benefits that would flow from this policy outweigh the costs?

The credibility of a deterrent threat is vital to its success. Yet the threat to expand our retaliation beyond those directly responsible might strike our opponents and others as incredible. During the Cold War, many questioned whether the United States would actually resort to the use of nuclear weapons to defend Europe and other allies, but it was argued that even a nuclear threat that “leaves something to chance” would be sufficient to deter a state like the Soviet Union. Some might argue that the threat that “leaves something to chance” would be insufficient here.

Many questioned whether the U.S. would actually resort to the use of nuclear weapons to defend Europe.

While there is strength to this point, it is not decisive. Several factors tell in favor of the credibility of an expanded deterrence policy. First, the threat to retaliate vigorously to a murderous attack against our own (and potentially allied) citizens is inherently credible. Common sense as well as history suggest that threats to retaliate in response to serious attacks on one’s homeland are highly credible. Few questioned the far more severe U.S. threat to respond with nuclear weapons to attacks against the homeland. As to the expanded notion of culpability proposed here, there is a strong intuitive appeal that those who materially supported, cooperated in, were complicit in, or were grossly negligent in the effective murder of large numbers of Americans are legitimate targets for retaliation. Contrarily, the approach differs advantageously from policies that were of more dubious credibility, such as the counterintuitive Mutual Assured Destruction posture or the oft-questioned American extended deterrent commitments — policies that included threats of far more harrowing consequences in response to less cataclysmic attacks.

Second, the United States has demonstrated a willingness to take severe measures when necessary to protect its interests. American credibility in this respect should not be underestimated. Even in the Pakistan-al Qaeda context, Deputy Secretary of State Richard Armitage’s (considerably less tailored) threat to “bomb” Pakistan “back to the Stone Age” reportedly was taken as very credible by the Pakistani leadership, which swiftly fell into line.11 Third, under this policy the U.S. would be justified in taking severe retaliatory measures, but would not be required to do so. Indeed, our actual response might be relatively limited. Thus our threats do not require that our opponents believe that our retaliation would be total, only that they know that it would be grave and fear that it might, indeed, be extremely severe. Fourth, even if credibility could not be sufficiently established to ward off a catastrophic attack, such a tragedy would give the United States the unfortunate opportunity to demonstrate the profound unwisdom of any kind of involvement, cooperation, complicity, or criminal passivity in the face of a catastrophic attack against the United States.

Another concern about an expanded deterrent policy is that such an approach, even if credible, would be no more effective in deterring catastrophic terror attacks than our current strategy. It is a truism that no policy solves every problem, and the policy here proposed may fail for a number of reasons. There is no single, neat concept that can “solve” the problem of catastrophic terror — the best we can do is try to minimize the risk. But security policy is about probabilities, and including expanded deterrence in our counterterrorism strategy would significantly lower the chances of a catastrophic terror attack through raising the costs and decreasing the likelihood of the successful execution of steps critical to conducting such a strike. First, as discussed above, terror networks survive and thrive in an environment that facilitates or at least permits them to operate, plot, and execute strikes. Indeed, it is well known that key terrorist groups, including al Qaeda, are hiding virtually unmolested in sanctuaries. The expanded deterrence strategy would seek through plausibly effective means to transform that physical, social, and moral environment from benevolent to toxic. Holding those complicit at risk would radically change the calculus for many in the network in our favor. The previously nonaligned would be incenvitized to become hostile to terror groups plotting massive strikes against Americans, and the previously benevolent would become at the very least more fearful and hesitant in their support. Of course, such efforts should be accompanied by efforts to improve the security situation the deterred will face, such that it is easier to resist threats or blandishments from terrorist groups.12 This “providing security to enable cooperation with U.S. interests and allies” insight has yielded successes in Iraq and can be replicated in this context.  

Holding those complicit at risk would radically change the calculus for many in the network in our favor.

Second, the vast majority of terrorists, even those contemplating catastrophic attacks against us, have some kind of rationale in mind, a strategy, a rational calculus that we can affect.13 And it is clear that even as the suicide bombers of al Qaeda and like-minded terrorist groups do not care about their own lives, they do care about something, including their political aims and their personal connections.14 Broadening our deterrent threat will let us seize more levers on these groups’ behavior.

Third, while the policy would place a premium on intelligence, it would not be an excessive burden. Preventive or warning intelligence must be correct, relevant, timely, and appropriately understood by decision-makers; errors on any of these counts court failure. Retaliatory intelligence, while still of course requiring accuracy, has the advantages of time and focus. Since it is reacting to an event, it operates under no inherent time constraints and so may be deliberate and thorough. Its object is clear, since the act in question has already occurred, and so the U.S. and its partners may focus their resources and efforts. And decision-maker interest in the wake of a terrorist attack can be presumed to be keen. Thus while the intelligence requirements for an expanded deterrence policy are significant, requiring ample investment in forensic and other identification tools, they are substantially more modest than the requirements for a preventive/defense approach.

A third concern with the expanded deterrence policy is that its benefits would not outweigh its costs. Critics might point to the policy’s potential to radicalize fence-sitting populations, alienate public opinion (reasonably or not), ease al Qaeda’s effort to establish common cause with other groups, generate anti-Americanism, weaken friendly states, and so forth. Could not similar American requirements be met through a more narrow and muted deterrent policy?

This pragmatic objection touches on a central point: This policy must be integrated into a broader policy of a “layered defense” against catastrophic terror and should only be emphasized if it seems to offer a net benefit to the United States. But a number of factors suggest that the benefits of an expanded deterrence policy outweigh its costs. First, the policy as such is likely to do far more to dampen hostile behavior than to catalyze it. The demands are reasonable, the target population is susceptible, the threat is credible, the capability to retaliate well-known. The upshot would be that behavior that had previously been undercounted in the scales of culpability would cost a lot more. Surely this would contribute to a reduction in such behavior by normally rational human beings. In addition, the U.S. should consider publicly scaling back other policy demands, in tandem with an expanded policy of deterrence, to meet reasonable expectations for balance in our strategy. Our expanded deterrence policy should only aim at deterring catastrophic attacks, not at lesser foreign policy objectives, as important as they might be.

The strategy would offer extensive benefits for cooperation while preserving a firm and real threat of retaliation.

Second, the policy, while it would need to be publicized, could be tailored to minimize its alienating aspects, especially when coupled with “soft power” approaches. “Tough” policies such as this one are not zero-sum with “friendlier” ones. And since the expanded deterrence stance would serve more as a strategic-moral framework than a requirement for specific actions, there would be ample discretion in setting out precisely how to execute both the approach’s declaratory and retaliatory aspects. Ideally, our overall strategy would offer extensive benefits for cooperation while preserving in the background a firm and real threat of retaliation in case of attack. The United States managed this kind of “carrots and sticks” approach during the Cold War, when cooperation with and even aid to the Soviets was coupled with an underlying nuclear threat in the event of war.

Third, including expanded deterrence in our policy best deals with the unpredictability and ambiguity of the catastrophic terror threat. An overweening focus on prevention is both unrealistic and dangerous, the former since a cunning and resourceful enemy in the modern world is likely to find weapons and use them, the latter because the logic of prevention can lead, if unbalanced, in a totalizing direction. Furthermore, prevention is of less value against “unknown unknowns.” The threat of catastrophic terror is not confined to extremist Muslims. Expanded deterrence is a general policy that targets those who might be complicit in a catastrophic attack, and thus does not rely on anticipatory knowledge. Expanded deterrence might deter them or their abettors. The policy, integrated into a layered defense also emphasizing prevention and other tools, thus best deals with the contingency and unpredictability inherent in the threat.

It is important to emphasize again that this policy would only be part of the overall American counterterrorism strategy and that it is certainly not a call for unrestrained retaliation, collective punishment, or violence against the innocent. It is instead a reasonable response to a catastrophic threat that shifts the burden of risk to those who bear some responsibility for and can materially and significantly contribute to stopping such attacks. Furthermore, this expanded deterrence approach should apply only to catastrophic terror attacks. It is justified by the awful scale of the catastrophic terror threat. In adopting the policy, the United States would be seeking to take certain forms of terrorist attack off the table by establishing red lines — such as wmd attacks against the homeland — that, if transgressed, would result in devastating and overwhelming retaliation. The point would be to make it in and of itself manifestly irrational to employ catastrophic terrorism against us. Limited terrorist attacks, however, would not warrant such an expansive response, but retaliation would take more limited forms of the variety to which we are accustomed — police, intelligence, and, if necessary, military measures. Of course we would seek vigorously to deter, defeat, and, if necessary, destroy terrorists, but we would reserve the starkest and most severe retaliation for catastrophic attacks.

A moral framework

An expanded deterrence approach raises a host of moral issues as well. But it is important to underline at the outset that the policy is not one of collective punishment or indiscriminate violence against civilians. It is a policy that carefully and reasonably expands the definition of guilt — it is not a policy that targets the innocent.

Indeed, despite auguring in a shift in policy, an expanded deterrence policy would fit well with traditional conceptions and practices of security responsibility. Notions of the culpability of complicity and even gross negligence, the responsibility of those who can materially frustrate harm to do so, shifting the burden of risk, and other concepts elemental to an expanded deterrent posture are deeply rooted in Western philosophical, legal, and moral systems.

The basic moral justification for the policy rests on the principle that the legitimate security needs of society may rightly demand greater levels of exertion from individuals and groups. In order to preserve itself from grave harm, society may reasonably call upon those within and without its borders not only to avoid doing it damage, but also to take active measures towards the same end. This is simply the principle of self-defense extended to the social scale. But its claim is not an unrestricted one. Instead, its chief criterion is a reasonable and strict connection between the scale of the threat and the demands placed. When threats reach grave levels, society may legitimately require actions that, in less perilous circumstances, would not be morally justified. Given that the United States faces the very real threat of catastrophic attack by terrorist groups, our society has the moral right and obligation to demand more of those who have it in their power to frustrate or at least materially impede such attacks.

Various fields of law and morality evince principles similar to expanded deterrence. In wartime, states hold both citizens and foreigners (both combatant and noncombatant, including diplomats) to much higher standards of behavior than during peacetime. Noncombatants and neutrals must scrupulously observe accepted restrictions in order to maintain their status. Indeed, a rather precise parallel concept under the law of war is that of “reprisal,” which provides that a nation may take otherwise unlawful measures in order to enforce or encourage compliance with the law of armed conflict. This concept is a part of the law of war under which U.S. military forces operate.15 Likewise, during epidemics, society may legitimately take drastic steps that in the absence of a public health emergency would be unacceptable in order to quarantine and defeat the disease.

Today we live in a thicket of regulations that require our active knowledge and participation to prevent harm.

Even in the more mundane context of modern tort law, the principles are similar. As with the traditional approach to terrorism, in torts there was originally an emphasis on individual fault, partially encapsulated in the mantra that “harms should lie where they fall.” But in an industrializing world, advanced societies progressively adopted tort systems that demand some level of active individual responsibility for the successful and smooth operation of the collective. Today we live in a thicket of regulations that require our active knowledge and participation to prevent harm or facilitate the smooth operation of complex systems. Particularly when we are engaged in activities that might harm others, we are held to much higher standards of accountability and culpability than in the criminal field. For instance, persons conducting what the law terms “ultrahazardous activities” — blasting, possessing dangerous animals, transporting dangerous chemicals — are usually absolutely liable for whatever misfortune ensues from their activities. Respondeat superior, the doctrine that employers are liable for the damages of their employees during the normal course of their duties, is another such an example. The United States regularly and effectively applies “third-party liability” laws, for instance holding hosts liable for the drunk driving accidents of their guests. In recent years countries have adopted “Good Samaritan” laws, regulations that require that citizens positively help people in distress, whether by notifying the authorities or themselves rendering assistance. When confronted with activities in which the danger is great, especially when of little social benefit, the law places a heavy legal burden on parties to minimize the risk of damage or provide restitution should accidents happen. The policies have both moral and pragmatic purposes, to provide recompense for those harmed and to minimize costs by placing burdens upon those best positioned to avert harm, rather than only on those who are at fault.

The American criminal justice system also offers ample precedents for an expanded deterrence policy. The Anglo-American criminal conspiracy doctrine holds all members of a conspiracy criminally liable for the crimes by any member within the scope of the conspiracy, even if individuals did not actively participate in or even approve of the crime (provided the crime was foreseeable). More directly, the U.S. Code prohibits the provision of “material support” to terrorists, an expansive term that includes “any property . . . or service.”16 Criminal negligence standards, meanwhile, are calibrated based upon the “reasonable person’s” standard of care and the damage involved.

We threatened to hold Communist leaderships accountable for the activities of their allies and agents, a form of extended responsibility.

Further, an expanded deterrent policy actually hearkens back beyond modern torts, to premodern laws adopted by governments that, like ours, could not guarantee order and security. English history offers a few examples. The ancient doctrine of misprision, for instance, deemed it a felony if subjects of the Crown failed to report to the authorities a treason or felony, even if the individual played no part in the actual crime. The graver offense of misprision of treason consisted in the simple knowledge and concealment of treason, which, without any assent necessary, rendered the individual a principal traitor.17 Similarly, Blackstone described that the historical law of England mandated that the community and its chief officers were responsible “for all robberies committed therein . . . for having kept negligent guard.”18 Both these doctrines reflected the reality, predominant before the nineteenth century, that the state could not pretend to keep order on its own and had to enlist, including through serious penalties, the cooperation of its subjects in the preservation of order.

Taken together, the law looks at guilt and innocence relatively, depending on context. Acts that are condoned in one context may be penalized in another. When the stakes are high, society may reasonably see complicity or even passivity as culpable. Knowing inaction in the face of petty theft is a qualitatively different matter than inaction in the face of the execution of a catastrophic attack.

An expanded deterrence policy would also follow in a long and successful tradition of American strategy. The federal strategy during the Civil War and U.S. and Allied policies during the Second World War, for instance, involved principles of expanded responsibility and the correlation of grave threats with vigorous responses. A more recent example was the U.S. and nato position during the Cold War, when the Allied powers credibly threatened to retaliate with devastating nuclear strikes against Soviet and Eastern European targets in the event of a Warsaw Pact attack against Western Europe — even a conventional one. The U.S. adopted a similar posture to defend allies in East Asia. The strategy involved issuing severe threats in order to deter our opponents from taking steps we thought likely to lead to disaster. It also involved threats to hold communist leaderships accountable for the activities of their allies and agents, a form of extended responsibility. Indeed, the form of expanded deterrence proposed here, limited as it is to those complicit in catastrophic attacks, is considerably less complicated morally than the Cold War’s balance of terror, which involved the implicit “exchange of hostages” and the threat to use massively destructive nuclear arms against population centers.

Suggestions for policy

Building on the government’s declaratory policy announced in February 2008, the United States should refine, implement, publicize, and, if necessary, follow through on an expanded deterrent posture against catastrophic terror attacks. In his speech, Hadley commendably retained ambiguity about the precise nature of our retaliation while making clear that the U.S. response would be “overwhelming.” He included those who “enable” and “support” a wmd attack, but the policy must be clear that it encompasses all those who materially supported, cooperated in, were complicit with, or were grossly negligent in such attacks. The policy should be clarified and then publicly and widely disseminated in order not only to deter but also to change the strategic framework for thinking about catastrophic attacks. From a traditional focus on individual fault, we should begin to shift the goalposts such that a higher level of responsibility would be expected in frustrating such strikes.

The demands of credibility and the practical necessities of conducting operations in line with this policy naturally require the conformity in relevant part of military, intelligence, and security service doctrine, procurement, and policy. Announcement of such a policy backed only by the inelastic support of high-yield nuclear weapons on the one hand and conventional main-force elements on the other would immediately call into question the policy’s real power. Instead, such a policy would require a capabilities mix that would allow the fullest range of responses compatible with budgetary considerations and international obligations — including, necessarily, the capabilities to meet our treaty obligations not to target civilians who are in no way complicit in a terrorist attack.

Of course many such systems have been or are now being developed. Attribution technologies should be developed, constantly improved, and deployed. Political, military, and intelligence assets should be prepared to target not only the core members of a terror group but also the complicit. And, in order to shore up the credibility of our retaliation, our policy should indicate that key elements of our response would be initiated shortly after the attack in order to put potential attackers on notice that retaliation would not be whittled down during a “cooling off” period of political and bureaucratic deliberation. This would increase the strength of the deterrent by making elements of our response appear automatic, even though, in fact, decision-makers would retain ample discretion.19

The world needs to be put on notice about what a catastrophic terror attack against U.S. vital interests would mean.

Absolutely central to a deterrent posture against terrorism would be the widespread and effective dissemination of the message. The world needs to be put on notice about what a catastrophic terror attack against U.S. vital interests would mean, how responsibility would be apportioned, and what kind of retaliation could be expected. Such an anticipatory announcement would be necessary from both a pragmatic and moral viewpoint, the former in order for the threat actually to have a chance of working and the latter in order to comport with the moral principle that people must have reasonable warning if they are to be held liable in new ways. Demarches and official releases, however, are insufficient for these purposes. The message must make its way downwards and outwards, and therefore requires dissemination through the media, internet, and other outlets. The silence following Hadley’s excellent February 2008 speech leaves much to be desired.

This done, the United States would present the world with a relatively simple and eminently reasonable expanded deterrent stance against catastrophic attacks. We would hope that the very fear the credible threat inspired would entail that it would never have to be realized, as was our fortunate experience during the Cold War. But we would have to be prepared to follow through on our pledges if we did suffer a catastrophic attack. Beyond real considerations of justice, the continuing peril of a catastrophic terrorist attack would demand that we show decisively that such strikes would meet with a severe response, a response that would demonstrate that a catastrophic attack against the United States can never be consistent with any rational strategy, whatever its aims. Our response following a catastrophic terror attack should prove that all concerned would be far better off doing all in their power to frustrate any future such attacks then to let a single additional one go through.


Elbridge A. Colby served in the Office of the Director of National Intelligence and on the Commission on the Intelligence Capabilities of the United States Regarding Weapons of Mass Destruction.


1 The word “catastrophic” is used advisedly here. It includes catastrophic weapons of mass destruction attacks but also other strikes of equivalent or greater magnitude that do not use nbc weapons. The emphasis is not on the means but on the damage done. Smaller-scale terrorist attacks that do not threaten the functioning of an advanced society are more susceptible to more traditional modes of response, though an aggregated campaign of such attacks might constitute a “catastrophic” attack. For one sobering study of the possible consequences of a catastrophic attack, see Charles Meade and Roger C. Molander, “Considering the Effects of a Catastrophic Terrorist Attack,” rand Technical Report (2006).

2 For an excellent discussion of these issues, see John Robb, Brave New War: The Next Stage of Terrorism and the End of Globalization (John Wiley & Sons, 2007). See also Fred Ikle, Annihilation From Within: The Ultimate Threat to Nations (Columbia University Press, 2006) and Martin Reese, Our Final Hour (Basic Books, 2003). For a recent report on these possibilities, see David Ignatius, “Taking Stock of the War on Terror,”.Washington Post (May 22, 2008).

3 This is a paraphrase of Lawrence Freedman’s summation of the experience of the early Cold War: “If there could be no escape from the presence of terror, then terror had to be respected.” Lawrence Freedman, The Evolution of Nuclear Strategy (St. Martin’s Press, 1983), 207.

4 “Material support” is defined in the U.S. Code under 18 U.S.C. 2339. Complicity, or acting as an accomplice, is defined as “[o}ne who knowingly, voluntarily, and with a common interest with others participates in the commission of a crime as a principal, accessory, or aider and abettor.” Ballantine's Law Dictionary (1969). The Texas Code defines criminal negligence as: “A person acts with criminal negligence, or is criminally negligent, with respect to circumstances surrounding his conduct or the result of his conduct when he ought to be aware of a substantial and unjustifiable risk that the circumstances exist or the result will occur. The risk must be of such a nature and degree that the failure to perceive it constitutes a gross deviation from the standard of care that an ordinary person would exercise under all the circumstances as viewed from the actor's standpoint.” Texas Penal Code, Title 2, Section 6.03, http://tlo2.tlc.state.tx.us/statutes/docs/PE/content/htm/pe.002.00.000006.00.htm (accessed May 22, 2008). Clearly negligence with respect to a catastrophic terror attack would constitute criminal negligence.

5 See Carl von Clausewitz, On War (Princeton University Press, 1976), 733.

6 See http://www.barackobama.com/2007/08/01/the_war_we_need_to_win.php (accessed May 19, 2008).

7 The author emphasizes that he does not pretend to U.S. government or other institutional authorization. For the author’s closer analysis of Hadley’s speech, see “The New Deterrence,” Weekly Standard online (April 10, 2008). http://www.weeklystandard.com/Content/Public/Articles/000/000/014/959tnykn.asp (accessed May 20, 2008).

8 Al Qaeda may not have been effectively deterred, however, in the sense of not understanding that its attacks would incur such a serious response. See Jonathan Schachter, The Eye of the Believer: Psychological Influences on Counter-terrorism Policy-Making (rand, 2002), Chapter 4.

9 For more on this point, see Daniel Byman and Matthew Waxman, The Dynamics of Coercion: American Foreign Policy and the Limits of Military Might (Cambridge, 2002).

10 For an exemplary analysis of this idea, see Paul K. Davis and Brian Michael Jenkins, Deterrence and Influence in Counterterrorism: A Component in the War on al Qaeda (rand 2002), esp. Chapter III and 47–49. The key, they note, is to understand that even such terrorists as al Qaeda can be deterred from doing specific things. As they point out, “everyone can be influenced sometimes” (22). The National Security Strategy of 2006 also takes promising steps in this direction, particularly in Section III, Part C, which discusses deterring terrorist networks. http://www.whitehouse.gov/nsc/nss/2006/sectionIII.html (accessed May 19, 2008).

11 http://news.bbc.co.uk/2/hi/south_asia/5369198.stm (accessed May 19, 2008).

12 Thanks to Robert Zarate for this point. For further elaboration on this issue, see Mark Smith, Janine Davidson, and Peter Brooks, “Micro-Foundations of Insurgent Violence: Implications for Iraq,” unpublished working draft, presentation prepared for the Pentagon’s Irregular Warfare Forum (Center for Adaptive Strategies & Threats, Hicks & Associates, Inc., October 5, 2005). Zarate points out that decisions the deterred would make fall along a spectrum, and that it pays to ease the ability to make decisions in our favor as well as threatening retaliation if they do not. Efforts in this direction include the expansion of lawful government control — or sovereignty — to areas such as the Northwest Frontier Provinces.

13 It is often said that groups such as al Qaeda cannot be deterred because they are not rational, meaning presumably that they do not have plausible objectives they are pursuing through a strategy of employing means to achieve ends. This is a highly questionable conclusion. The November 29, 2007 attempt by Osama bin Laden to break off European allies from the United States was only one of the more recent examples rebutting this argument. This followed a similar offer in April 2004. More broadly, see, e.g., Christopher M. Blanchard, “Al Qaeda: Statements and Evolving Ideology,” crs Report for Congress. (July 9, 2007); Robert Pape, Dying to Win: The Strategic Logic of Suicide Terrorism (Random House, 2005), esp. Chapter 7; Marc Sageman, Understanding Terror Networks (University of Pennsylvania Press, 2004); Scott Atran, “Mishandling Suicide Terrorism,” Washington Quarterly (Summer 2004) and “The Moral Logic and Growth of Suicide Terrorism,” Washington Quarterly (Spring 2006); Jonathan Schachter, The Eye of the Believer: Psychological Influences on Counter-terrorism Policy-Making (rand, 2002), 94–96; Robert Trager and Dessislava Zagorcheva, “Deterring Terrorism: It Can Be Done,” International Security 30:3, (Winter 2005/06); Martha Crenshaw, “The Causes of Terrorism,” in Charles W. Kegley, ed., International Terrorism: Characteristics, Causes, Controls (St. Martin’s, 1990), 117–120; and Andrew Kydd and Barbara Walter, “The Politics of Extremist Violence,” International Organization 56:2 (Spring 2002); Anonymous, Through Our Enemies’ Eyes: Osama bin Laden, Radical Islam, and the Future of America. (Brassey’s, 2002).

14 See Sageman, Understanding Terror Networks, and Pape, Dying to Win, esp. Part I.

15 See, for instance, The Commander’s Handbook on the Law of Naval Operations, Department of the Navy (July 2007). The handbook states: “A belligerent reprisal is an enforcement measure under the law of armed conflict consisting of an act that would otherwise be unlawful but which is justified as a response to the previous unlawful acts of an enemy. . . . Reprisals may be taken against enemy armed forces, enemy civilians other than those in occupied territory, and enemy property” (6–4).

16 18 U.S.C. 2339.

17 William Blackstone, Commentaries on the Laws of England, Book IV, Chapter 9.

18 Blackstone, Commentaries, Book I, Chapter 9.

19 Thanks to Paul K. Davis for this point. Thomas Schelling emphasizes the importance of automaticity, impetuosity, and even (from certain standpoints) irrationality in an appropriate retaliatory response. See Thomas Schelling, Arms and Influence (Yale University Press, 1966), esp. Chapter 2, “The Art of Commitment.”