Victor A. Utgoff
Don’t Rule It In and Don’t Rule It Out: The Role of Nuclear Weapons in Deterring Chemical and Biological Attacks

• Ambiguity of U.S. Policy on Nuclear Deterrence of Chemical and Biological Attacks
• Horrific Chemical and Biological Attacks Are a Real Possibility
• Should the United States Rule Out Nuclear Retaliation for Large-Scale Chemical and Biological Attacks?
• Is Nuclear Retaliation for Large-Scale Chemical and Biological Attacks Credible?
• Should the United States Unambiguously Promise Nuclear Retaliation for Large-Scale Chemical and Biological Attacks?
• Ambiguity in Practice
• Potential for a More Consistent Bias against Nuclear Retaliation for Large-Scale Chemical and Biological Attacks
• Defending against Chemical and Biological Attacks
• Potential Effectiveness of Chemical and Biological Defenses
• Corresponding Potential for Changes in Chemical and Biological Deterrent Policy
• Concluding Observations

[page-number of print edition: 107]

Victor A. Utgoff
Don’t Rule It In and Don’t Rule It Out: The Role of Nuclear Weapons in Deterring Chemical and Biological Attacks

Many new questions about nuclear deterrence have been raised since the end of the Cold War. Among the most important is the question of how nuclear deterrence relates to the threat or actual use of chemical and biological weapons (CBW). This paper addresses that relationship. In particular, it argues why U.S. policy on nuclear deterrence of CBW use should remain ambiguous for the foreseeable future. It also addresses the nature of that ambiguity. Finally, it discusses how improved defenses against CBW attacks for the United States and its allies could favor the adoption of a policy that, while still ambiguous, is more consistently biased against nuclear retaliation than is today's policy. [ Scott Sagan argues against a policy of nuclear deterrence of chemical and biological attacks in Scott D. Sagan, „Should the United States Use Nuclear Threats to Deter Chemical and Biological Use?" in Sidney Drell, Abraham D. Sofaer, and George Wilson (eds.), The New Terror: Facing the Threat of Biological and Chemical Weapons (Stanford: Hoover Institution Press, 1999).]

Clearly, this paper is very speculative, and the more so as it progresses toward the second and third of the above three arguments. Thus, it is all the more appropriate that I state that I am expressing my own opinions, and they are not necessarily endorsed by the U.S. government or any other organization or person.

Let me set the stage by considering a specific scenario. That scenario has the United States and its allies confronting some future regional aggressor that is armed with CBW. This aggressor is assumed to be capable of attacking many concentrations of allied forces and at least a dozen large cities with CBW, thus causing great devastation if these forces and civilians are not well protected. The question is how to deter such attacks in the course of an effort by the United States and its allies to roll back the aggressor’s military gains, or better yet, how to deter the aggressor from seeking such gains in the first place. This kind of scenario clearly raises the question of nuclear deterrence, in contrast, for example, to the use of CBW by terrorists that present no targets that could justify nuclear retaliation.

Ambiguity of U.S. Policy on Nuclear Deterrence of Chemical and Biological Attacks

Let me begin by addressing the nature of current U.S. policy on nuclear deterrence of chemical and biological weapons attacks. At first glance, this policy might seem quite clear, given the U.S. long-standing „negative security assurances" policy, reaffirmed in 1995, that promises not to use nuclear weapons against states that do not have nuclear weapons and are not fighting with the support of those that do. This policy implies that such states need not fear nuclear retaliation for chemical and biological (CB) attacks.

On the other hand, in recent years, high-level U.S. national security officials have stated that the U.S. response to chemical weapons (CW) and presumably biological weapons (BW) attacks would be „absolutely overwhelming and devastating". Furthermore, U.S. government officials have argued that any opponent's use of CBW to cause great harm to the United States and its allies would remove any policy restrictions on retaliation with nuclear weapons, at least for proportional direct responses against the aggressor that are judged necessary to end CBW use. Thus, the United States is on record with repeated high-level declarations that

[page-number of print edition: 108]

suggest no nuclear retaliation for CBW use as well as other declarations from similarly ranked officials that suggest the opposite and have not been repudiated.

Moreover, this ambiguity is no accident. As Secretary of Defense William Cohen stated in November 1998: „We think the ambiguity involved in the issue of nuclear weapons contributes to our own security, keeping any potential adversary who might use either chemical or biological (weapons) unsure of what our response would be." [ Dana Priest and Walter Pincus, „U.S. Rejects ‘No First Use’ Atomic Policy: NATO Needs Strategic Option, Germany Told", Washington Post , 24 November 1998, p. A-24.]

Taking into consideration its effects on deterrence of potential enemies with CBW, reassurance of allies, and global nonproliferation interests, is this ambiguity a good idea? I think so, at least for the time being. Let me outline why.

Horrific Chemical and Biological Attacks Are a Real Possibility

Over the last decade, the United States and its allies have discussed the potential for horrific CB attacks extensively. While proliferation of BW weapons in particular has been a rising concern for many years, a variety of CBW-related events during the last decade have been particularly disturbing. These include:

• discovery after the 1990-91 Gulf War that Iraq had a well-developed capability to make large-scale BW attacks;
• emerging evidence of the size and sophistication of the longstanding Russian BW program;
• CW attacks on the Tokyo subway by the Aum Shinrikyo, together with subsequent discoveries that it had also attempted a series of BW attacks; and
• evidence of CBW development programs in many states.

Some observers have expressed skepticism about how real the BW threat is. They note that very few states or organizations have attempted mass destruction with chemical or biological weapons during the twentieth century, and no one has succeeded. They argue that creating and effectively employing CBW is a complex task that can easily fail, and point to the long series of failures of the Aum Shinrikyo. They question why any state or terrorist organization would be motivated to initiate such attacks.

I am afraid these sentiments are wishful thinking. That no one has employed CBW to cause mass destruction in the last hundred years is not in itself a good argument that this horror will not be visited on us in the future, particularly as additional states and organizations obtain CBW capabilities.

Creation of CBW is not particularly difficult. It requires a modest technical education, some care to protect against unprotected exposure to the agent, a patient test-and-fix engineering approach, and modest resources. It is vastly less complex and expensive than creating nuclear weapons.

As for motivation, it is all too easy to imagine an aggressor state lashing out with CBW in an attempt to dissuade the international community from rolling back some ill-gotten gain. And, examples of organizations motivated by raw hatred or bizarre beliefs that Armageddon needs to be helped along are already a reality.

[page-number of print edition: 109]

Finally, the amount of potential damage that might be done with a well-planned and executed BW attack is enormous. Covert attacks against a dozen large unprotected cities over a few days could kill millions. Use of highly contagious agents could do even more damage.

Should the United States Rule Out Nuclear Retaliation for Large-Scale Chemical and Biological Attacks?

If the United States and its allies are to protect their vital interests against states that have the capability to make these kinds of large-scale CB attacks, either they must be able to defend themselves well enough to keep casualties reasonably low, or they must deter such attacks. The populations of the United States and most of its allies are not protected against large-scale CB attacks. Furthermore, allied military forces have much to do before they achieve a reliable capability to survive and operate for extended periods in the face of large-scale CB attacks against military forces or their supporting infrastructure.

Thus, for the foreseeable future, when the United States and its allies are called on to face aggression by states capable of large-scale CB attacks, they will have to resort to the prospect of retaliation to suppress whatever temptations the opponent may see to make such attacks. Those temptations may be very large in some situations.

What kind of retaliation should the allies be capable of? Timely, reliable, at least equally painful, and credible seem to be important characteristics. If retaliation cannot be done on a timely basis once the decision has been made, the opponent can hope that events might intervene to discourage or prevent it. Similarly, if retaliation depends upon uncertain factors such as knowledge of the location of the opponent's leadership or remaining CB capabilities, the opponent might hope to avoid it. If retaliation cannot be at least equally painful, the opponent can hope that the greater threat posed by its CB capabilities could lead the alliance to make important concessions or try to avoid confronting it altogether. And, of course, retaliation with nuclear weapons must be credible. The issue of credibility will be addressed below.

Can conventional retaliation meet these needs? In some cases, it may be able to, but we cannot expect it to. The opponent will do its best to hide its leadership and reserve CB capabilities. Conventional weapons cannot be counted on quickly to defeat the opponent's military forces or to match the levels of damage that it can do with large-scale CBW attacks. [ These arguments are developed at greater length in Victor A. Utgoff, „Nuclear Weapons and the Deterrence of Biological and Chemical Warfare", Occasional Paper No. 36, Henry L. Stimson Center, October 1997.]

The prospect of nuclear retaliation thus seems necessary if the allies are to realize the full potential of deterrence through punishment to protect themselves from CB attack. Nuclear weapons also bring with them an extra measure of deterrence. In contrast to conventional bombing, history has no examples of a state that has fought on despite nuclear attacks. Furthermore, nuclear weapons share an awe-inspiring and fearful mystique with plagues and poisons. They promise at least commensurate punishment for CBW attacks in ways that conventional weapons cannot match.

Is Nuclear Retaliation for Large-Scale Chemical and Biological Attacks Credible?

This logic implies that nuclear retaliation for large-scale CBW attacks should not be ruled out. But is it credible? I expect that everyone is horrified at the idea of using nuclear weapons.

[page-number of print edition: 110]

But this is peacetime, and we are not grappling with reports surfacing hourly of thousands of sick and dying citizens in dozens of U.S. and allied cities or across some theater of allied military operations. We must ask ourselves what allied decision-makers and populations would think when actually immersed in such a horrible situation.

In such circumstances, nuclear retaliation will be a most salient option, and many arguments and powerful emotions will favor it. Thus, for example, in the aftermath of a large-scale CB attack, the U.S. and allied leaderships and populations will feel a white-hot anger at the carnage and suffering caused by the opponent's initiation of a forbidden form of warfare. Taking revenge with equally horrible forbidden weapons will likely seem just.

Disengaging will seem impossible. The opponent cannot be allowed to defeat an otherwise vastly more powerful alliance by means of CBW. That would make these weapons seem not just usable, but capable of underwriting successful aggression. In addition to losing the immediate conflict, North Atlantic Treaty Organization (NATO) and U.S. security guarantees would be sharply devalued. Further proliferation of nuclear, biological, and chemical (NBC) weapons would seem assured, both by would-be aggressors and by those who would see a stronger need to defend themselves rather than depend upon allies.

At the same time, allied leaders will find it most difficult to require their military forces to fight on if only the opponent were allowed to use nuclear, chemical or biological weapons. Nuclear retaliation will seem the best way to suppress further use of CBW in the conflict at hand.

These arguments suggest that nuclear retaliation in the aftermath of large-scale CB attacks is not just credible, but, in the event, could seem almost compelled. This, in turn, implies that the United States should not adopt a declaratory policy that rules out nuclear retaliation for large-scale CB attacks. The United States and its allies seem likely to decide on nuclear retaliation primarily on the basis of the kinds of arguments stated above, among which any prior commitment not to retaliate with nuclear weapons will be only one among many considerations and perhaps not even the most important one. In pointing to the theory of belligerent reprisal, U.S. officials have said as much.

Thus, the United States has a de facto dependence on nuclear retaliation for large-scale CB attacks, no matter what government officials might say about it. This has an important implication. If nuclear retaliation for large-scale CB attacks is a real possibility, we should not mislead anyone about it. It would be a double tragedy if some future opponent launched large-scale CB attacks because it took too seriously a U.S. pledge not to respond with nuclear weapons and then was surprised to find the United States responding with nuclear weapons to impose commensurate damage.

In sum then, it seems important not to rule out nuclear retaliation for large-scale CB attacks.

Should the United States Unambiguously Promise Nuclear Retaliation for Large-Scale Chemical and Biological Attacks?

Well, for the skeptics of nuclear retaliation for CBW attacks, let me note that I have reached what you should see as the low point of this paper. Let me turn then to the question of whether nuclear retaliation for large-scale CB attacks should be ruled in. This is also a relatively complex question, and again, I believe the answer is no.

[page-number of print edition: 111]

By far the most important reason not to adopt an unambiguous declaratory policy that large-scale CB attacks will be met with nuclear retaliation is the bad effect such a policy would have on efforts to curb and roll back nuclear proliferation. Three effects seem particularly important.

First, because the infrastructure for developing and maintaining a capability for large-scale BW capability can be so small and innocuous, it may ultimately prove impossible to verify with adequate confidence that no state has such a capability. Thus, an unambiguous policy of linking nuclear deterrence with BW threats would imply that nuclear weapons must be retained indefinitely. This would unambiguously and unnecessarily repudiate eventual U.S. compliance with Article VI of the Nuclear Non-Proliferation Treaty.

Second, linking nuclear deterrence unambiguously to CBW capabilities would elevate the importance of all three types of weapons in the minds of proliferators, actual and potential. If the United States thus were to highlight an unambiguous need for nuclear weapons as a deterrent to CB threats and attacks, that could suggest that other nations need nuclear weapons for the same purposes. It could also elevate the value of CB weapons by suggesting they are more or less equivalent to nuclear weapons.

Third, a real and immediate demand for unambiguous nuclear deterrence of large-scale CB attacks will not likely arise very often. If so, why should the United States adopt an unambiguous policy of nuclear deterrence in peacetime - a constant drag on global nuclear nonproliferation efforts - in order to maximize deterrence of a CB threat that may only rarely become a real possibility?

Finally, an alternative to nuclear retaliation seems possible, albeit difficult. [ Ibid.] If the United States and its allies can develop and field effective defenses against CB attacks, the damage that an opponent might do with such attacks could be limited to low enough levels to make nuclear retaliation seem unnecessary. Under such conditions, conventional bombing should be able to impose commensurate punishment quickly and reliably. Alternatively, effective defenses against large-scale CB attacks could allow the alliance to push for total defeat of the opponent - and to do so at acceptable cost despite the opponent’s use of CB weapons. We will consider the potential value of CB defenses further in the next section.

For all these reasons, and some other lesser ones, it seems wiser for the United States and its allies to forgo a policy of unambiguously promising nuclear deterrence for large-scale CB use. This conclusion, when added to our previous one, seems a reasonable rationale for U.S. policy regarding nuclear deterrence for CB attacks: Don't rule it in and don't rule it out.

Ambiguity in Practice

Not ruled in, but not ruled out allows a great deal of latitude, however. Within such broad bounds, the United States can suggest a bias in favor of nuclear responses or against them. A bias in favor might be articulated by saying, for example, that the American people would demand the strongest possible response to the use of chemical or biological weapons. Alternatively, a bias against might be articulated by suggesting that the United States and its allies find it difficult to foresee circumstances in which the use of nuclear weapons would be required.

[page-number of print edition: 112]

An ambiguous policy on nuclear response to large-scale CB attacks can also be dynamic, moving to a greater bias toward nuclear responses in times of crisis or war, and then a lesser bias when circumstances change. Thus, for example, we have the Bush administration's letter to Saddam Hussein during the opening stages of the 1990-91 Persian Gulf War. That letter seems to have led the Iraqi leadership to believe that nuclear retaliation for their use of CBW was likely, although the word „nuclear" was not used. After the war was over, American leaders stated in various memoirs that they had decided that they would not respond to Iraqi use of CB weapons with nuclear weapons.

Of course, the fact that the American leadership had decided not to respond with nuclear weapons suggests the possibility that they could have decided the opposite. And, of course, they might have changed their minds for some of the reasons articulated earlier, had the Iraqis actually made devastating use of CBW.

Making ferocious threats that the American people would demand the strongest response (Do the American people believe that we have stronger responses than nuclear retaliation?) and admitting afterwards that the strongest response had already been ruled out is one way to exploit the latitude in the U.S. policy of ambiguity. In the future, when the United States and its allies again find themselves confronting an aggressor armed with CBW, it seems likely that the United States will again make statements that will make the strongest possible response seem pretty much assured. And, when that confrontation is over, it will again find ways to retreat into versions of ambiguity that sound less committed to nuclear retaliation.

In some ways, this kind of dynamic allows the best of both worlds. The United States can almost guarantee nuclear retaliation, when it sees the need for maximum deterrence of CB attack, and it can almost guarantee against nuclear retaliation when that serves its interests. On the other hand, this kind of dynamic admittedly has a price. U.S. deterrence and nonproliferation policies both suffer some loss in credibility when alternatively emphasized and de-emphasized, as circumstances seem to demand.

Potential for a More Consistent Bias against Nuclear Retaliation for Large-Scale Chemical and Biological Attacks

While some ambiguity in the U.S. policy regarding nuclear deterrence of large-scale CB attacks seems inevitable for the foreseeable future, there are things that can be done to allow it to be more consistently biased against nuclear retaliation. The most effective way to achieve such a bias is to alter the underlying realities that appear to require nuclear retaliation. Of course, the best way to do that would be to implement a universal and verifiable ban on CB weapons. This is a long way off, however.

Another way to reduce future needs to bias the ambiguity toward nuclear retaliation is to create conventional retaliatory capabilities that can substitute better for nuclear deterrence. This may prove possible in some cases. As noted above, intelligence coups and special conventional weapons may allow conventional strikes that can quickly destroy an opponent's leadership, or its capability to employ CB weapons. But these possibilities cannot be counted upon.

Defending against Chemical and Biological Attacks

Yet another path to reduced de facto reliance on nuclear retaliation is to create defenses that can reduce the damage from large-scale CB attacks to a small fraction of that which would

[page-number of print edition: 113]

otherwise be possible. The means for protecting military forces from CB attack are well known. They center on passive defenses, including masks and protective suits that filter out the intake of CB agents. They also include sensors to warn when CB agents are present and collective defenses provided with clean air for some high-priority functions and occasional relief for those who must generally operate outdoors. Additionally, vaccines and antidotes are available to guard against some of the most important BW agents, and new vaccines and antidotes are being developed to guard against other possible agents.

Improvements are being made in these defenses to increase their effectiveness and lighten the burden of employing them. Providing forces with effective CB defenses and the training to employ them is a challenge, but it is feasible. A reasonable marginal cost estimate for protecting forces from CB attack with the types of equipment that are available now or shortly to become available might be $2,000-$3,000 per person. [ By marginal costs, I mean the costs of additional defenses beyond those that would likely be spent for conventional war. Thus, in contrast to the cost analysis presented in „Nuclear Weapons and the Deterrence of Biological and Chemical Warfare", pp. 18-21, I am not including here the costs of systems such as tactical missile defenses. I am also assuming that inexpensive, broad-spectrum, reliable BW detectors will soon prove possible. ]

Concepts for protecting urban populations in the United States from CB attack are less well developed, but are getting a great deal of attention now. The general methods of protection will be a subset of those employed for military forces. Still, several general differences between the problems of protecting civilians and of protecting military forces suggest that civilians should be easier to protect on a per capita basis. First, civilians will generally have access to buildings that can be equipped with supplies of clean air. Second, most civilians can cease their normal activities and stay in a shelter for the hours during which a CBW attack might be a particularly high hazard. Third, most civilians can wear masks that need not accommodate particularly strenuous activity or provide special features such as undistorted vision for sighting weapons. Finally, since civilians would not generally be at risk to being hit by liquid CW agents, and would generally stay indoors until the major dangers from such an attack had largely dissipated, they generally need not be provided with suits to protect against skin exposure to CW agents.

The challenge presented by the need to protect civilians is primarily one of organizing, settling on an overall protection concept, and winning the necessary level of political acceptance. The technologies needed are generally similar to those required by the military. The one possible exception is CB detectors. Protection of the population in a large number of cities and suburban areas will place an even greater premium on developing inexpensive, highly reliable, fast-acting, broad-spectrum detectors. All things considered, a reasonable marginal cost estimate for protecting civilians might be $200-$300 per person.

These marginal costs can add up to very substantial numbers for states with large forces and large urban populations. A large state with armed forces totaling one million and an urban population of 200 million could spend on the order of $40-$60 billion to create such defenses.

Potential Effectiveness of Chemical and Biological Defenses

No defense is perfect, of course. The one envisioned here would still allow a small percentage of the forces targeted by CBW to become casualties. Failure to use protective equipment effectively seems likely to prove the largest source of casualties. Industrial data and military

[page-number of print edition: 114]

experience with failures to use protective equipment effectively suggest a failure rate of perhaps a few percent of the population exposed to a contaminated environment.

Even with this low casualty rate, the total number of casualties from a large-scale CB attack could be very substantial. Thus, for example, an attack against a dozen large CB protected cities with populations averaging a million each might cause a hundred thousand casualties. While this would be a horrible loss, it is far smaller than the millions of casualties that could be expected without such protection.

Corresponding Potential for Changes in Chemical and Biological Deterrent Policy

Conventional bombing by forces of the size that are likely to remain available to the United States and its allies could readily cause commensurate and even substantially greater damage to an aggressor that made such CB attacks. This damage could be accomplished in less than a day if the forces had been prepared to carry it out. [ „Nuclear Weapons and the Deterrence of Biological and Chemical Warfare", p. 26.] Furthermore, given strong defenses against CB attack, U.S. and allied forces could consider fighting on to accomplish the total defeat of a state that had tried such attacks. Moreover, it would seem far less likely that an aggressor would attempt such CB attacks if it expected them to be so much less effective.

Given such defenses, it would seem reasonable to bias U.S. policy on retaliation for large-scale CB attacks strongly against nuclear retaliation. Even virtual nuclear threats would seem unnecessary. Totally ruling out nuclear retaliation could still prove unwise, however. Defenses against CBW attacks may work substantially less well than expected. It also seems unlikely that the United States and its allies would find it acceptable to engage in repeated exchanges of conventional retaliation for CBW attacks without threatening escalation to something even more devastating.

Concluding Observations

This paper has presented three main arguments. The first is that U.S. policy on nuclear retaliation for large-scale CBW attacks ought to remain ambiguous. The second is that this ambiguity has been and seems likely to remain dynamic in nature, shifting between a lesser and a greater bias in favor of nuclear retaliation, depending on circumstances. However necessary these shifts may be for the sake of deterrence in circumstances threatening CBW use against us, they come at some cost to longer-term U.S. nonproliferation interests. Third, U.S. and allied capabilities to defend against large-scale CB attacks can provide the extra benefit of allowing a policy that, while still somewhat ambiguous, can be more consistently biased against nuclear retaliation for such attacks.

The possibilities for dealing with the potential large-scale CB attacks all present serious challenges. Elimination of CB weapons is still a long way off at best. It will require much more work and greater and more consistent support from the highest levels of governments. Building effective military and civil defense capabilities against large-scale CB attacks will require substantial spending over at least five to ten years. Such civil defense capabilities will require public acceptance of the need, which at this point simply does not exist. In comparison, dependence on nuclear deterrence seems the simplest option. With continued proliferation of CBW, however, the de facto dependence of the United States and its allies on nuclear deterrence of large-scale CB attacks will grow. Perhaps this is inevitable, but perhaps not. Surely everything possible should be done to slow the growth and to reduce this

[page-number of print edition: 115]

dependence. The value of achieving effective universal elimination of CB weapons is clearly very high.

Finally, let me note a dilemma that could be posed for the United States, if chemical and biological weapons continue to proliferate and few states other than the United States develop strong defenses against them. In this case, aggression by a regional state with the capability to make large-scale CB attacks would pose yet another major problem for the United States. Asking unprotected allies to join with it in facing the aggressor would re-create the dependence on nuclear deterrence that strong U.S. CB defenses could otherwise allow the United States to avoid. Alternatively, not asking them would forgo the legitimacy that comes with acting in concert with a broadly based international coalition. Thus, it is important that the United States not be alone in making the preparations that are necessary if CB-armed aggression is to be confronted with minimal risk to all parties involved.

[page-number of print edition: 116 = blank page]