Thoughts About Virtual Nuclear Arsenals

Thoughts About Virtual Nuclear Arsenals

Home•IRChina • IR Scholars • Foreign Scholars • Waltz, Kenneth. Thoughts About Virtual Nuclear Arsenals Source: Washington Quarterly, Summer97, Vol. 20 Issue 3, p153, 9p

When I read Jonathan Schell’s 1984 New Yorker essay, "The Abolition," I wondered why anyone would prefer a world in which no one had more than a near-nuclear military capability to the world we had then. More than a decade later, I still wonder. The three main advantages of a system of virtual arsenals are said to be these: They would relegate nuclear weapons to the margins of international politics. They would remove the danger of nuclear weapons being fired accidentally or without authorization. They would strengthen efforts to halt the spread of nuclear weapons and help to solve regional problems. Seemingly, a system of virtual arsenals would bring substantial advantages. Moreover, it is thought that in such a system not much could easily go wrong. Any country would know that if it began to assemble weapons others would too. Attempts to break out of the system would be seen to be self-defeating. If several countries nevertheless reactivated parts of their arsenals, they would, as now, deter one another from using them.[1] A system of virtual arsenals promises large benefits at low or no cost. On close inspection, however, the benefits begin to dwindle and the costs appear to grow. It is said that substituting virtual for actual arsenals would move nuclear weapons to the peripheries of international politics, but in an important sense that would leave them right where they are now. Nuclear weapons are useless for fighting wars and even for threatening blackmail.[2] Nuclear weapons have always formed part of the scenery of international politics, which is the appropriate place for weapons suited for deterring rather than for fighting. One may hope, and virtual weaponeers do in fact hope, that nuclear weapons will continue to cast their shadow over international affairs, thus providing a considerable assurance of peace among states enjoying their protection. If deterrence would work as well with virtual as with actual weapons, then nations should surely agree to eliminate weapons-in-being. Low though the dangers have proved to be, weapons that are ready to fire may go off accidentally, may be exploded without authorization, and may be acquired by "rogue" states or terrorists. One must, however, wonder whether virtual arsenals will deter as readily as actual arsenals have. The argument that virtual arsenals are sufficient for deterrence is especially odd when it comes from those who earlier thought that the requirements of deterrence were endlessly demanding and inordinately expensive to meet, that only a seamless web of capabilities, able to dominate an opponent whatever the opponent might do up and down the escalation ladder, would deter. Belief in the difficulty of deterrence was summed up in the widely accepted thought that a strategy inadequate for the fighting of wars cannot deter.[3] Virtual weaponeers nevertheless argue that deterrence will work without weapons-in-being as long as some states retain or develop the ability to assemble them quickly. In the first step toward a system of virtual arsenals, the number of nuclear warheads in the hands of the major nuclear states would be drastically reduced. This would be a good thing in itself. Small numbers of warheads are easy to safeguard and control. Some people have long claimed that only small numbers are needed for deterrence, and recently weighty voices have buttressed the claim.[4] If they are wrong, then a system of virtual arsenals can be realized only by moving the world through a forbiddingly dangerous condition on the way to establishing the system. If they are right, then unilateral reduction of weapons to small numbers is possible without further ado. Once this is accepted, warheads numbering in, say, the low hundreds can be deployed and guarded in ways that are highly proof against threat and against accidental and unauthorized firing. The first two virtues of a system of virtual arsenals can be gained without eliminating second-strike forces-in-being. The minimal deterrence argument--that not much is required to deter--has never been widely accepted. If the leaders of states cannot be persuaded that small numbers are sufficient for deterrence, then surely they cannot be persuaded to go all the way to having no actual weapons at all. If we make the big assumption that finally they can be, can we then move on to a system of virtual arsenals and safely reap the additional advantages that its advocates promise? For two closely connected reasons a system of virtual arsenals is untenable. First, deterrence without second-strike forces will not work. Second, a system of virtual arsenals would be unstable. To deter means to dissuade someone from doing something for fear of the consequences. One country’s

http://www.irchina.org/en/xueren/foreign/view.asp?id=205 (1 de 5)30/8/2007 20:28:49

Thoughts About Virtual Nuclear Arsenals

weapons may deter other countries from using their weapons to score gains at the expense of one’s vital interests. Nuclear weapons, because they dominate other weapons, are especially good for the purpose. Virtual weaponeers would substitute "factory deterrence," or "weaponless deterrence," for deterrence with second-strike forces. As Schell put it: Weapons would no longer deter weapons; instead, "factory would deter factory." The ability to make nuclear weapons "would make abolition possible, because it would keep deterrence in force."s Would it? Factories cannot deter factories from producing their goods. Only the products of factories can serve as instruments of deterrence. Factory deterrence is deterrence one step removed. In a system of virtual arsenals, all countries would know that if they secretly made weapons with the factories they were allowed to have, or secretly assembled weapons from parts legitimately on hand, other countries, becoming suspicious, would soon do the same. Presumably, no country would have reason to break with the system. The penalty for doing so would, however, be a mild one: namely, that others might well follow suit. Such a penalty would be easily borne. Countries feeling insecure, and doubting that other countries were strictly following the rules, could safely risk incurring it. The inability of factories to deter would drive states out of the system, that is, away from reliance on it and toward relying on themselves by acquiring second-strike forces. Virtual weaponeers emphasize that their system will be viable only if deterrence continues to work. Indeed, some of them stipulate that the United States must retain its ability to extend its deterrent to cover U.S. interests abroad. Yet even when the United States had more than 10,000 strategic warheads, its allies wondered whether America’s deterrent would cover them. A latent nuclear force is at best a shaky deterrent at home; it will find no credit abroad. Virtual weaponeers also stress that a regime for the management of nuclear weapons must promise stability. A system of second-strike forces is highly stable. Indeed, stability is part of the meaning of the term "second strike." Second-strike forces are nearly impervious to the efforts of others to negate them, whether by seeking first-strike capability or by building defenses. As Harold Brown put it when he was secretary of defense, purely deterrent forces "can be relatively modest, and their size can perhaps be made substantially, though not completely, insensitive to changes in the posture of an opponent."[6] Because second-strike forces reduce worries about others’ military capabilities, surveillance of their forces by intrusive means of inspection is unnecessary. With virtual arsenals, countries would have to worry incessantly lest their capability for rapid production and deployment fall behind the similar ability of others to do so. The possession of nuclear weapons has brought not only security to states individually but also peace among nuclear states collectively. Weapons bringing such benefits are rarely found. States that believe their security endangered will want to keep or to get them. To thwart them will require heroic efforts. States will resist conforming their policies to internationally imposed stipulations that can promise satisfactory levels of security only if all states follow the rules. Believing that heroic efforts will sometimes fail, states will hedge their bets and bend the rules, or simply cheat. Virtual weaponeers argue that, because states know that if some cheat then others will too, all will have reason not to. Temptations to cheat grow, however, as the gains from doing so rise and as the chances of being caught fall. Since the security of nations will be at stake, the reasoning of states will reverse the reasoning of virtual weaponeers: Some states will feel that they had better cheat if only because others may be doing so. It is hard to hide battleships, but easy to hide warheads. No one can know about weapons that may bc hidden to the view of even well-equipped and free-ranging inspectors. Because this is so, counsels of prudence would require countries to cheat a little by hiding some ready weapons as a hedge against the possible cheating of others. If states are limited to very low numbers of weapons, and even more so if the number is zero, cheating is both worthwhile and easy. Michael Mazarr is right in saying that a system of virtual arsenals would require "extremely intrusive" inspection.[7] Inspectors would have to certify not only that proscribed weapons do not exist but also that nuclear facilities are incapable of making or assembling weapons in less than a specified time and in more than a specified quantity. Virtual weaponeers understandably emphasize the importance of intrusive and rigorous inspection. Unstable systems depend on it; stable systems do not. Yet by both common sense and recent experience, we know that the reports of inspectors will be thought unreliable. We can hardly expect states to agree to inspections as intensive and thorough as those to which Iraq has been subjected since its 1991 defeat. Yet five years later doubts persist about Iraq’s ability to produce, or to resume production of, chemical, biological, and nuclear weapons. In June 1995, the United Nations official overseeing the destruction of weapons told the Security Council that Iraq’s biological warfare program is larger than previously thought and that some of its materials are not accounted for. In November 1995, two experts on the control of nuclear weapons expressed fear that Iraq may be perfecting and testing components of a bomb, while lacking only the plutonium and bomb-grade uranium needed to assemble nuclear weapons rapidly.[8] Because nuclear warheads are small and light they can easily be hidden and moved. Inspection is at best an uncertain business. Weaponless deterrence would multiply the uncertainties. Uncertainties by each country about how other countries are doing would breed distrust all around, increase insecurities, and provide strong incentives to strengthen military forces. Some virtual weaponeers realize this. Having pointed out that "extremely intrusive" inspection would be required to maintain the system, Mazarr adds that some cheating would be tolerable, since the "state which covertly reassembled twenty or fifty nuclear weapons would have achieved little." Only if it had gained a first-strike capability could it prevent "other states from redeploying their own arsenals."[9] Logically, he may be right. What could a first strike accomplish? And how much would the country that struck first risk losing if the country struck http://www.irchina.org/en/xueren/foreign/view.asp?id=205 (2 de 5)30/8/2007 20:28:49

Thoughts About Virtual Nuclear Arsenals

had a secret store of even a few nuclear warheads? Sensible answers to these questions make a first strike all but unthinkable. But some states will think of reasons why other states may do the unthinkable. Paul Nitze imagined in the mid-1970s that the Soviet Union might launch a first strike at us believing that somehow it had "deterred our deterrent" even though at the time the United States and the Soviet Union had more than 2,000 strategic delivery systems.[10] Less fertile imaginations will be able to invent ways in which states might exploit a "zero-some" gap, and some states will begin to worry that others may do so. Mazarr underestimates how much states worry about "gaps" in capability that may be thought to offer "windows of opportunity" to others. States would hasten to equip themselves with nuclear weapons, lest a newly rearmed state somehow gain an advantage from its moment of superiority. If some countries fear that other countries can move more quickly from virtual to actual weapons than they can, everyone will work on worst-case assumptions. Under such circumstances, crises easily develop from uncertainties the system spawns, aside from substantive matters. The parties to crises then find it risky to ride through them without furthering their military preparations, even though doing so predictably heightens tensions all around. The temptation to move first, eliminated by second-strike forces, would be re-introduced by a system of virtual arsenals. By placing a premium on speed of action in crises, virtual arsenals would increase the chances of accidents occurring, including the most important accident of all--the misjudgment of others’ acts and intentions. Especially in crises, one’s precautionary measures are easily taken by others as preparations to strike. One can scarcely believe that in an unstable system accidents are unlikely to happen. Competition over the ability to move first with decisive force is dangerous, a lesson well learned from the prehistory of World War I. The lesson was taken to heart by nuclear strategists of the 1950s and ’60s. We should not forget it. The first big problem of a system of virtual arsenals is that deterrence would be problematic; the second, that the system would be unstable. A third problem is a more subtle one. With nuclear forces-in-being, a state that has been attacked can choose the time for retaliation. Without nuclear forces-in-being, the attacker can make use of the pause imposed between aggression and retaliation. People would worry again, as they did in the 1950s and ’60s, about the possibility of a "Hamburg grab." (The worry was that the Soviet Union might pounce, secure, and hold an exposed territory and in effect ask whether its limited gain merited a military response.) A country could seize coveted ground, dig in to raise the cost of a counterattack, and announce that it had clandestinely assembled a small nuclear arsenal. The aggressor could then use its deterrent force to protect its illicit gains. Worrying about a Hamburg grab seems fanciful, yet states have often hoped or feared that time could be used to gain military advantage. Some states may well think of additional ways to exploit an enforced interval between the ability to produce and the ability to use nuclear weapons, and others will fear their doing so. The scope of Egypt’s and Syria’s combined attack on Israel in 1973 was apparently limited by fear of Israel’s nuclear deterrent.[11] If Egypt and Syria had believed that their forces could have joined hands by slicing Israel in two before a retaliatory strike could be mounted, they might have been tempted to try. If rigorously enforced, a system of virtual arsenals would invite aggressors to use time to secure a military advantage. At this point one may wonder how the virtual arsenals that some states have--the undeclared nuclear forces of Israel, India, and Pakistan--work their deterrent effects. The answer is that everyone believes that those states do, or at least may, have warheads ready to use; and that is enough to deter. In the same way, cheating may save a system of virtual arsenals. The possibility of cheating would preserve the deterrent effect against major aggression that only weapons-in-being can reliably provide. Knowing that states would still worry, virtual weaponeers add that states could always construct strategic defenses. With deterrence in abeyance and with the security of states uneasily dependent on a delicate mixture of obedience to rules and efforts to elude them, states will find strategic defenses attractive. A perfect defense against large nuclear arsenals is hard to imagine, but a defense capable of denying entry to warheads fired in small numbers may seem worth building. A virtual-arsenal system would be a rickety one. It would presumably work this way: Knowing that others would follow suit, no state would be likely to rearm heavily. That would be of some comfort, but not much. The nuclear deterrent would, as now, be the reliable one: fear of retaliation should rebuilt arsenals be used. And if a country clandestinely readied weapons before other countries could ready their own, missile defenses would provide protection. Might one then hope that there the system, if it can be called such, would come to rest, with small offensive forces posed against small defensive forces? Such a system would still be unstable with everyone worrying about the comparative quality of offensive and defensive forces and with each tempted to improve its capabilities if only to make sure that the offensive/defensive balance be kept quite even. Moreover, with the level of insecurity high, all states would fear that one of them might try to exploit a momentary offensive or defensive advantage by striking first. Under the circumstances, far from strengthening efforts to halt the spread of nuclear weapons, virtual arsenals would set the stage for the proliferation of nuclear weapons that the world has long feared but never seen. An unstable system with deterrence problematic would increase the insecurity of states. Insecure states would want at least to have the virtual arsenals that the new system would permit. The system would stimulate states to construct virtual arsenals and thus put themselves in a position to take the easy step from having the facilities for making the weapons to making the weapons themselves. Instead of halting the spread of nuclear weapons,

http://www.irchina.org/en/xueren/foreign/view.asp?id=205 (3 de 5)30/8/2007 20:28:49

Thoughts About Virtual Nuclear Arsenals

a system of virtual arsenals would promote it. States that feel insecure try to protect themselves. Some states believe, understandably, that nuclear weapons are the cheapest, safest, and surest way of doing so. Heroic efforts are required to keep states from getting the weapons they believe essential to their security. Virtual weaponeers admit the truth of this statement when they accept the likelihood of cheating within a system of virtual arsenals. Cheating would restore the effectiveness of deterrence by bringing it into closer correspondence with the system we now have. One wonders, then, why we should not leave well enough alone and why, with its obvious pitfalls, advocates of a system of virtual arsenals should be taken seriously. The notion that virtual arsenals can be substituted for weapons-in-being is symptomatic of an idea now gaining popularity in the United States, the idea that nuclear weapons can be abolished. The vogue of abolition owes more, however, to the recent shift in world-political forces than to the merits of an idea. Reflecting on the rich experience of half a century at the center of international affairs, Paul Nitze has recently noted that while we once used nuclear weapons to offset the Soviet Union’s conventional strength, others now use nuclear weapons to offset ours.[12] Les Aspin, when he was chairman of the House Armed Services Committee, put the same thought in the following words: "A world without nuclear weapons would not be disadvantageous to the United States. In fact, a world without nuclear weapons would actually be better. Nuclear weapons are still the big equalizer, but now the United States is not the equalizer but the equalizee."[13] A country’s nuclear weapons deter other countries from using force against them much more surely than its conventional weapons can. Against countries that have nuclear weapons, the United States loses much of the advantage of its superiority in conventional weapons. Pointing this out, however, rather gives the game away. Nuclear weapons deter with an effectiveness that conventional weapons do not approach. That is the most important reason for wanting to have them. For illustration, we need look no farther than America’s recent war in Iraq. If the United States had thought that Iraq might have had a few bombs, we would have had to manage the Iraq-Kuwait crisis differently, say by employing only an embargo. Invasion might have prompted Iraq to dump a couple of warheads on Haifa and Tel Aviv. We would not have wanted to run the risk, and Israel surely would not have complained about our unwillingness to use force in a headlong attack. A big reason for U.S. resistance to the spread of nuclear weapons is that if weak countries have some they will cramp our style. Militarily punishing small countries for behavior we dislike would become much more perilous. Nuclear weapons in the hands of the weak limit what the strong can do to them. That is why the spread of nuclear weapons is so hard to stop, and why some leading U.S. military experts have become abolitionists. Nuclear weapons in the hands of other states depreciate the value of our conventional forces. So long as U.S. conventional superiority lasts, devaluing nuclear weapons would seem to serve our interests. Enhancing the value of conventional weapons by depreciating the value of nuclear ones would, however, stimulate conventional arms races and make wars easier to start. These results would serve neither our nor the world’s interests. Nuclear weapons have helped to preserve the peace where it has been most endangered and prevented wars from getting out of hand in some of the most troublesome areas, as between the United States and the Soviet Union, between India and Pakistan, and in the Middle East. We should be wary of the false hopes held out by advocates of virtual arsenals. This article is reprinted from Jonathan Schell, The Abolition, copyright (C) 1984, by permission of Alfred A. Knopf, Inc. The author is grateful to Karen Ruth Adams and Radha Pathak for their assistance on this essay. Notes 1. Michael J. Mazarr, "Virtual Nuclear Arsenals," Survival 37, no. 3 (Autumn 1995), p. 22. Mazarr’s essay provides an excellent summary of the case for a system of virtual arsenals. 2. See Kenneth N. Waltz, "Nuclear Myths and Political Realities," American Political Science Review 84, no. 3 (September 1990). 3. Linton E Brooks, "Naval Power and National Security: The Case for the Maritime Strategy," International Security 11, no. 2 (Fall 1986), p. 71. John P. Rose, The Evolution of U.S. Army Nuclear Doctrine, 1945-1980 (Boulder, Colo.: Westview, 1980), pp. 102-106. Michael Howard, "On Fighting a Nuclear War," International Security 5, no. 4 (Spring 1981), pp. 3-17. 4. See, for example, Bernard Brodie, Robert S. McNamara, and Herbert York, cited in Scott D. Sagan and Kenneth N. Waltz, The Spread of Nuclear Weapons: A Debate (New York: W. W. Norton, 1995), pp. 108-109. 5. Jonathan Schell, The Abolition (New York: Alfred A. Knopf, 1984), pp. 119-120. 6. Harold Brown, Annual Report, Fiscal Year 1980 (Washington: U.S. Department of Defense, 1979), pp. 75-76. 7. Mazarr, "Virtual Nuclear Arsenals," p. 18. 8. Christopher S. Wren, "U.N. Expert Raises Estimates of Iraq’s Biological Arsenal," New York Times, June 21,

http://www.irchina.org/en/xueren/foreign/view.asp?id=205 (4 de 5)30/8/2007 20:28:49

Thoughts About Virtual Nuclear Arsenals

1995, p. A6. Paul L. Leventhal and Edwin S. Lyman, "Who Says Iraq Isn’t Making a Bomb?" International Herald Tribune, November 2, 1995, p. 8. 9. Mazarr, "Virtual Nuclear Arsenals," p. 19. 10. Paul H. Nitze, "Deterring our Deterrent," Foreign Policy 25 (Winter 1976-77). 11. Gerald M. Steinberg, "After the NPT Extension: Israeli Policy Options," IGCC Newsletter 11, no. 2 (Fall 1995), p. 6. 12. Nitze, "Deterring our Deterrent." 13. Les Aspin, House Armed Services Committee Hearings.

[Close] Copyright © 2005 www.IRChina.Org All Rights Reserved ••••••••••••••••••••••• ••••••••••••••• Email•[email protected]

http://www.irchina.org/en/xueren/foreign/view.asp?id=205 (5 de 5)30/8/2007 20:28:49