The Magazine

Deterrence and Prevention

Why a war against Saddam is crucial to the future of deterrence.

Feb 3, 2003, Vol. 8, No. 20 • By TOD LINDBERG
Widget tooltip
Single Page Print Larger Text Smaller Text Alerts

THE QUESTION of what to do about Iraq--and moving down the track, what to do about North Korea--typically gets described as a choice between deterrence and preemption (or perhaps better, "prevention"). If Saddam Hussein can be contained and deterred from using weapons of mass destruction, as some contend, then there is no need to go to war against him. If, on the other hand, we cannot be confident that he can be deterred, then preventive action is necessary. Reaching the latter conclusion is generally considered a doctrinal leap--a declaration of no confidence in the theory and practice of deterrence.

This idea of a radical break with past practice and past theory is embraced by both sides--by the advocates of deterrence and by the partisans of prevention. In the case of the former, the movement from deterrence to prevention represents a rejection of time-tested means of dealing with adversaries in favor of the always risky course of waging aggressive war--and losing in the bargain the justification of necessity, thus imperiling the moral legitimacy of our cause. For the advocates of prevention, it's good riddance to deterrence. Now that an alternative is available, who needs a doctrine that keeps the peace only at a level of utmost precariousness?

In practice, of course, U.S. policy has long been a blend of both deterrence and prevention. As Max Boot has noted, the United States has often chosen to act preemptively or preventively. At the same time, we have also fielded forces meant to deter parties from taking action we would find inimical, in many cases with apparent success. Of course, the contrast between messy reality and tidy theory is no refutation of a theory; it may simply represent a failure to apply the theory as systematically as it might have been. In the case of deterrence and prevention, however, I would suggest that the mess runs deep, and the theory is not tidy at all.

When people talk about deterrence, they usually assume that the unacceptable conduct to be deterred is both clear and consistent over time. In many cases, this may be true, the case of a nuclear first strike between rival superpowers being clearest of all. But it is not always true, and it is certainly not true in the case of Iraq.

Until and during the first Gulf War, the objective was to deter Iraq from the use of weapons of mass destruction against coalition forces or Israel. At the time of the cease-fire agreement, however, something important changed, and the change was little appreciated at the time. The United States insisted that Iraq disarm, and Iraq agreed. It is clear in retrospect that this amounted to a major shift in the U.S. view of deterrence, as least as applied to Iraq. The United States was seeking to deter not only Iraq's use of weapons of mass destruction but also its acquisition and possession of such weapons.

Iraq, it quickly became clear, was disinclined to abide by the new terms. The question, then, was what the United States (here, as leader of a coalition) would do to enforce them. How serious was Washington? Serious enough to insist on sanctions against Iraq until it fully complied with the terms of United Nations resolutions demanding disarmament. But this failed to impress Iraq overmuch. Serious enough to sever Saddam Hussein's sovereignty over parts of the country, the northern and southern no-fly zones, by maintaining a steady military presence. But again, evidently not impressive enough to persuade Saddam to disarm. Serious enough, in the wake of September 11, to begin assembling allies for military action to change the regime in Iraq and to return to the United Nations to seek what became Security Council Resolution 1441, declaring Saddam in "material breach" of his obligations and offering him a "final opportunity" to comply--which he remains unwilling to do. Serious enough, finally, to amass a huge force in the region with the evident intention of putting it to use.

All of this activity has aimed to persuade and if necessary coerce Iraq to accept a new standard of conduct (though now more than a decade old) and to deter Iraq from breaching it in the future. If at any point Iraq had relented and disarmed (or relents now and disarms), then the terms of deterrence would have been reestablished and that would have been the end of the matter.

But if the showdown comes to blows, in what will be called a preventive action on the part of the United States, it will be preventive in two senses: both in the immediate sense of thwarting Saddam Hussein, but also as an advertisement more generally of what is acceptable action by other parties.

Two conclusions follow. First, preemption or prevention cannot be said to have superseded deterrence. Rather, preemption is the violent reestablishment of the terms of deterrence. Second, insofar as war is intended not merely to reverse an unacceptable situation in the here-and-now but also to have a pedagogical effect (pour encourager les autres), it has in general a "preventive" component. To the extent that one says the Civil War "settled once and for all" the question of whether or not states may secede from the Union, one refers not only to a bloody conflict with a particular outcome but also to a successful exercise in conveying a message. One might say that, in a sense, war is the ultimate means of making your point.

THUS WHAT APPEARS to be a dispute over doctrine, deterrence vs. preemption/prevention, is actually a dispute over whether, hereinafter, the United States should seek to deter not only the use of weapons of mass destruction but also the acquisition of such weapons, at least by certain states. Should the United States, acting alone or (preferably) in concert with others, be willing to go to war in order to prevent a state from acquiring weapons of mass destruction, especially nuclear weapons, in the expectation that by establishing our willingness to do so we will deter other states from trying to acquire weapons of mass destruction? Or should the United States remain relatively indifferent to the acquisition of nuclear and other such weapons--offering merely declaratory support for nonproliferation--and remain satisfied with trying to deter their use?

Before we get to that question, however, there is underbrush that needs to be cleared away, the byproduct of mistaken notions about deterrence and prevention and the present conflict.

Some have objected that by embracing a doctrine of preemption the United States invites other states to assert a similar right, thereby serving as a general pretext for aggressive war. This misconstrues the current situation. In order to successfully act preventively, one must have the capacity to do so--which is to say, one must be able to prevail in teaching the lesson one wants learned. At the height of the Cold War, the United States was in no position to declare that it was no longer acceptable for the Soviet Union to have nuclear weapons. The result would have been catastrophic. Nowadays, in areas in which the United States takes an interest, generally speaking it is only the United States that has the capacity to revise the terms of acceptable international behavior. Other states' willy-nilly attempts to undertake revisions of their own run into the question of whether the United States would find their efforts acceptable.

Only by misconstruing what it means to act preventively or preemptively does one arrive at a difficulty here. If anything is possible in a world where deterrence has been abandoned in favor of unilateral wars of prevention, then we are indeed in a mess. But in truth, the capacity of any given state to set or revise the terms of acceptable international conduct neither increased nor decreased as a result of any declaration by the Bush administration. The United States, under current circumstances, can do things other states cannot, something well understood by other states and by Washington. The question of a substantial revision in standards of acceptable international conduct--in this case, the proposition that acquisition of weapons of mass destruction by some parties is unacceptable--is thus mainly a question raised by and answered by the United States (although there is no reason not to seek international participation). It is not a question posed generally to states.

Nor can it really be said that the United States, in trying to move from use-deterrence to acquisition-deterrence, is articulating a universal standard. Critics who raise this issue usually have in mind accusing the Bush administration of hypocrisy for a decision to act militarily against (say) Iraq but not (say) North Korea. Their aim is not really to see the supposed "doctrine" of preemption applied to North Korea but to discredit its application to Iraq. What do you do with your "doctrine" of preemptive action against states acquiring weapons of mass destruction if the state is, say, Australia? But Australia is precisely not the question. The point of the shift is to establish that states like Iraq may not acquire weapons of mass destruction without being met by force. Canberra need have no worries on that score; so unlike Iraq is Australia that the latter has no desire to acquire weapons of mass destruction, notwithstanding that Australia is a vigorous actor internationally. The United States has not set and is not setting the rules for Australia. The question for a head of state contemplating the acquisition of nuclear weapons in relation to the new terms of deterrence is as follows: Does the United States think I am more like Australia or more like Iraq, because if (and only if) it thinks I'm more like Iraq, it will treat me as Iraq was treated.

DETERRENCE IS USUALLY thought to rest on a thing, the capacity to inflict violence--the deterrent--and on a logic that flows inexorably from the existence of the thing. This essentially materialist conception misses the mark. It is not so much a deterrent, the thing, that deters. What deters is the idea that the deterrent deters--the conclusion reached by a real human being who has thought about what lies ahead and decided, all in all, it's too risky. There is a substantial literature on the need to make a deterrent "credible" for it to be effective. But the discussion of credibility tends to end where it really should just be getting underway, namely, with the capacity of the deterring party to act and the clarity of its intention to act (or at least the clear possibility it will act) in the event of unacceptable conduct. Is there really so little that needs to be said about the party that decides whether or not to be deterred?

We have had about 50 years' experience with nuclear deterrence. It mainly took the form of two massive arsenals squared off against each other. Without question, the deterrent capacity of each superpower was unmistakable to the other, meaning both the capability of the hardware to produce Armageddon and the willingness of each to unleash it in the event of an attack by the other. But deterrence as such is not an invention of the nuclear age. Deterrence has been practiced for thousands of years. It is, quite simply, the attempt to get someone not to do something by making it more costly for him to try. The walls of the city might not be a guarantee that no conqueror will prevail. But a city without walls could be relatively assured of its swift conquest. And the walls can indeed be said to deter those hordes who might be willing to sweep through an unwalled city but who are unwilling to lay siege.

The problem quickly becomes apparent: If deterrence has been practiced for thousands of years, sometimes successfully, the history of warfare in those same thousands of years is nothing other than the record of the failure of deterrence. This is true both in a formal sense (if deterrence works, there is by definition no war) and in the sense that the need to deter potential enemies is universal insofar as politics has something to do with enemies, who after all cannot be welcomed.

Why didn't those deterrents deter? We can say that if you make war on someone who has tried to deter you, you have probably concluded, rightly or wrongly, that you can pay whatever price the deterrent adds. Again, conclusions with this result would seem to have a long pedigree among human beings. The point is that in thinking about deterrence, we must consider not only the deterrent itself but also what people think about the deterrent. And from the Bible forward, the historical record is replete with accounts of disagreements over whether to go to war among generals viewing the same deterrent before them.

So we have several millennia of attempted deterrence, sometimes successful and sometimes not, at the conclusion of which falls a 50-year period of successful nuclear deterrence (accompanied by various ongoing attempts to deter conventionally, again with mixed success). Surely, there is something in the character of the nuclear weapon as an "ultimate" weapon that changes things. The question, however, is whether it changes not some things but everything. The claim that it changes everything is quite radical: For millennia, the success or failure of deterrence has had a material component (the deterrent itself) and a concrete ideal component (its contemplation by someone, an actual human being, considering war). Now, however, we have empowered ourselves to deduce our conclusions from the thing itself.

More precisely, what we are doing is creating to our own specification an abstract idea of a human being--which is to say, "someone" who does not exist and has never existed, but whom we have imbued with certain characteristics of our choosing, first and foremost a desire for self-preservation. Then we confront this abstraction with something that, in the real world (though not in this exercise) is a distinctly non-abstract thing, a nuclear weapon. We proceed as if this confrontation were somehow occurring in actuality between a human being and a nuclear weapon, without our agency--when all along, it is we ourselves who have created the abstraction solely for the purpose of having it "contemplate" this thing--in this exercise, actually an idea of the thing. Whereupon (and what an astounding coincidence this is!) the abstraction "reaches" the same conclusion about the "thing" that we ourselves, the only real human beings in the process, have already reached.

We congratulate ourselves for thinking about the unthinkable. But what we have actually been doing is hiding the conclusion we reached at the outset from ourselves long enough to let the suspense mount until we bring it back in time for a happy ending. We have, in short, cheered ourselves up.

The most serious article I have read in opposition to going to war in Iraq appears in the January-February 2003 edition of Foreign Policy. It was written by John J. Mearsheimer and Stephen M. Walt, two leading international relations theorists of the "realist" school. Their chief conclusion is that "deterrence has worked well against Saddam in the past, and there is no reason to think it cannot work equally well in the future." The authors do, however, grant a certain novelty to the situation: "The real nightmare scenario is that Saddam would give nuclear weapons secretly to al Qaeda or some other terrorist group."

They say, however, that "the likelihood of clandestine transfer by Iraq is extremely small." In the first place, a "history of enmity" between Saddam and al Qaeda makes "the Iraqi dictator . . . unlikely to give al Qaeda nuclear weapons, which it might use in ways he could not control." U.S. pressure "might eventually force these unlikely allies together," but even so, "Saddam would still be unlikely to share his most valuable weaponry with al Qaeda," and besides, he "could hardly be confident that the transfer would go undetected." But "even if Saddam thought he could covertly smuggle nuclear weapons to bin Laden, he would still be unlikely to do so." Why would he give away what he has worked so long to acquire? Besides, in the event of a nuclear attack on the United States, he "could never be sure the United States would not incinerate him anyway if it merely suspected" he had been involved.

I will return to the question of the United States incinerating a perpetrator, which I regard as far more problematic than Mearsheimer and Walt do. For now, I would like to draw attention to the authors' assertions about what is "unlikely," "unlikely," and "unlikely" again.

First of all, they give us no assurances. "Unlikely" is not the same as impossible. Second, note the way in which the unlikelihood of Saddam and al Qaeda cooperating, given their vast differences, gives way (in the event that this is not so unlikely after all) only to a second unlikelihood, that Saddam would share his weaponry. But surely if that "unlikely" reconciliation does come to pass, the possibility of a transfer is not merely "unlikely" but more likely, which our authors do not note. That is because their concern here is not really with Saddam or al Qaeda. They are interested in applying the proposition that if you fear incineration, you will not use or allow others to use your nuclear weapon against a power that can incinerate you. They apply it well. And it is entirely persuasive that if Mearsheimer and Walt were running Iraq, we would have little to fear. But it is no test of whether Saddam Hussein is deterred to apply a test that in actuality asks whether Mearsheimer and Walt would be, especially when the conclusion is clear from the outset.

I think, further, that the al Qaeda problem is worth more discussion in the context of deterrence than it has hitherto received. Everybody seems to agree on this point: that if al Qaeda could get its hands on a nuclear weapon, it would try to use it. All right, but why would al Qaeda not be deterred? The usual answer is that al Qaeda is a terrorist organization, not a state. Therefore, deterrence doesn't apply to it. But is that all there is to say? In the first place, al Qaeda was intimately entwined with a state, Afghanistan, at the time of the September 11 attack. In response, the United States toppled the Taliban government and began a worldwide manhunt for its leaders and for members of al Qaeda. They are hunted men, and they will be until they are dead--a death that might take the form of a drone-guided Hellfire missile descending at any moment without notice on a vehicle in which they are traveling. I would think that the prospect of having your government toppled and spending life on the lam, under an implicit sentence of death, might make a certain kind of person chary about provoking such a response or reluctant to tolerate certain activities on the part of folks operating on his soil. Unfortunately, I do not think Osama bin Laden and Mullah Omar are of that certain kind, and neither does anyone else.

There may be truth to the proposition that you cannot deter a terrorist non-governmental organization because only states can be deterred. The theory is in this sense safe, all praise to it. But we have saved it at the price of losing a complete account of the phenomenon before us. In saying that bin Laden and associates would use a nuclear weapon if they had one, aren't we also really saying that they don't share the concerns of self-preservation typical of heads of state? That they don't necessarily fear violent death? That they have an end that is perhaps higher than self-preservation? Clearly, we could not solve this problem by giving them a state of their own in the expectation that they would not risk losing it.

We thus arrive at the "madman" exception to deterrence theory: You can't deter a madman, of course. But "madman" is a loaded term. What we are actually talking about here is the full human range of tolerance for risk as well as the full human range of disparity in the ability to weigh risks, and this in the context of vastly varying expectations about what the future holds. Saddam Hussein is often said to be cautious (and therefore not a madman). I think some tyrants may in some sense be cautious, but only in the context of one who has proved willing to assume the considerable risks of tyranny. It is at any rate not the caution of a European Union bureaucrat.

Nuclear deterrence theory pretends to give us reassurance about Saddam, but in the end merely assumes he is the kind of man about whom we may feel reassured. In this context, one wonders if the nuclear peace of the Cold War was kept not so much by nuclear arsenals and the logic flowing from them as by the thoroughly bourgeois character of American society as well as the Soviet nomenklatura.

To return to the question of the moment, properly framed: Should we, then, act preemptively or preventively in Iraq in an attempt to shift the ground of deterrence from the use of weapons of mass destruction, especially nuclear weapons, to the acquisition of weapons of mass destruction, again especially nuclear, at least in the case of regimes run by people who give us serious cause for worry? That my answer is "yes" will come as no surprise, on the obvious grounds flowing from the above: The surest way to prevent something is to remove any one of the things it presupposes. We are safer if we never get to the question of whether Saddam Hussein would risk incineration by contriving to participate in a plot leading to the detonation of a nuclear weapon in the United States or Israel or Europe. We will be safer still if we can hasten the anchorage of bourgeois modern life in the Middle East, and so will the people who live there. But in the meantime, we can secure ourselves against all the unanswered and perhaps unanswerable questions related to the use of such weapons by stopping matters prior to their acquisition.

There is, however, a further reason, and it may be the ultimate reason. Nuclear deterrence theory, on its own terms, requires the capability and willingness to inflict massive retaliatory damage on an aggressor. I am not so sure about the certainty of such "retaliation" today.

For one, we suffered a devastating attack on the American homeland, and the decision the Bush administration made in response was to topple the Taliban government and to hunt down members of al Qaeda by conventional (if novel) means. It was not, for example, to use tactical nuclear weapons to ensure that no one hiding in the caves of Tora Bora would escape alive, let alone to use nuclear weapons more broadly. No one ever seriously proposed such a thing.

For another, there was an attack on the United States in October 2001 with a biological weapon, anthrax. Granted, the authorities suspect a homegrown terrorist. If, by chance, they are wrong, then those who were truly responsible have had the experience of getting away with something that would surely have called for some kind of retaliatory action.

Third, relatedly, the question of whom to retaliate against is substantially harder than Mearsheimer and Walt, among others, make it out to be. A nuclear blast has destroyed a major metropolitan area; the number of dead exceeds 100,000; there is near-certainty about which terrorist organization was involved, but no convincing proof of where it got the weapon. Grant the impulse to lash back in righteous anger: Would a U.S. administration be prepared to act on suspicion alone? I don't think so. At a minimum, one would want to know more, which would take time. The nexus between strike and counterstrike would be attenuated. In addition, as the intelligence services worked overtime, the option of launching a retaliatory nuclear strike would have to contend with and defeat in the policy-making process other potential retaliatory measures.

Which in turn takes us to the final problem: What do we gain from a nuclear attack on Baghdad? We reestablish our posture of use-deterrence, I suppose. And it is possible that things might reach a point at which domestic political opinion would settle for no less. But the price would be awful. A hundred thousand dead Iraqis? A million? What most of these people have in common would be the experience of living under the repression of Saddam Hussein. One could blame them for this fact, the epic kind of premodern blame that demands collective punishment for an erring or sinful people--in this case, calling them to account for the perfidy of their leader. But this would represent a drastic change from the distinctly modern idiom we take for granted, in which we present such people as victims of their rulers, as people with whom we have no quarrel, as people whose liberation from misrule we would like to see and have perhaps encouraged. "People" first of all, not "enemies."

Who would want to kill them? I would not. Perhaps I could persuade myself of the necessity to do so. Deterrence theory holds that the mere possibility that the national command authority might be persuaded to retaliate is enough for the mechanisms of deterrence to kick in. But if it isn't, and we can't be sure it is, the choice that results is essentially tragic.

It is accordingly to be avoided. And that is why attempting to deter the use of nuclear weapons is not enough when we have the capacity to deter their acquisition.

Contributing editor Tod Lindberg is a research fellow at the Hoover Institution, Stanford University, and editor of Policy Review.