Larry Carri/ll, UniU-d

Reducing U.S. and Sovietnuclear arsenals

A reduction of U.S. and Soviet nuclear arsenals to one-tenththeir present sizes would maintain deterrent capabilities

while exerting a stabilizing influence on the balance of terror.

by Harold A. Feiveson, Richard H. Vllman, andFrank von Hippel

THE ALMOST TOTAL absence of discussion of alter-native futures that has characterized the nuclear weap-

ons debate was first broken by the freeze movement andthen by President Reagan's 1983 Strategic Defense Initiative("Star Wars") proposal. Thus far six basically different alter-native nuclear futures have been discussed:

• The abolitionist vision would completely eliminatenuclear weapons.

• The president's vision would effect a transition to a"defense-dominated" world in which increasingly effectivedefenses result in offensive systems' withering away.

Harold Feiveson is a research political scientist at Princeton Uni-versity's Center for Energy and Environmental Studies. RichardUllnian, also a political scientist, and Frank von Hippel, atheoretical physicist, are both professors in Princeton's WoodrowWilson School of Public and International Affairs.

• An arms race unconstrained by the 1972 Anti-BallisticMissile (ABM) Treaty would maintain the mutual hostagerelationship of the United States and the Soviet Union byvirtue of the continuing dominance of offensive nuclearweapons systems.

• A constrained arms race would proceed more or lessalong current lines —constrained by the ABM Treaty andby modest SALT Il-type limitations on some categories ofoffensive nuclear weapons.

• A tightened arms-control regime or freeze would takeSALT II as its starting point to put stringent limits on thestrategic arms competition, but would leave both super-powers with nuclear forces not much reduced in quantityor variety from those they now possess.

• Finite deterrence would couple very deep reductionsin the superpower nuclear arsenals —but not enough to putin doubt their mutual hostage relationship —to severe con-straints on the development and deployment of first-strikeand ABM technologies.

August 1985144

The first two alternatives would represent radical depar-tures from the current "balance of terror." The third wouldrepresent the breakdown that many fear is imminent in thecurrent arms control regime. The lasi: three represent at-tempts to rationalize the current situation.

Both radical alternatives will remain infeasible for theforeseeable future. Although we hopt that abolition willultimately be feasible, so many difficult issues would haveto be dealt with that intermediate goals are required. Thepresident's notion that nuclear weapons can be made "im-potent and obsolete" by a unilateral U.S. technical fix istechnologically impossible, and the effort to move towarda defense-dominated world will only lead to the third alter-native: an all-out offense-defense arms race.

Ofthe proposals to rationalize the current situation, aninterim freeze on new nuclear weapons would probably bean essential prerequisite to a comprehensive scheme of re-ductions. Beyond that, however, only i:he finite-deterrencealternative provides a rationale for reducing the current scaleofthe superpower arsenals. These arsenals, which containtens of thousands of nuclear weapons, are completely outof correspondence with the reality of the world that nuclearweapons have created: they cannot be used without greatrisk of triggering the murder-suicide pact thai binds Eastand West together.

The adoption of finite deterrence would make possiblea 10-fo!d reduction of the superpower nuclear arsenals andthe elimination of their most destabilizing and dangerousweapons. Thus, it could transform the relationship betweenthe United States and the Soviet Union, reducing the dan-gerous fantasies and paranoia that feed and are fed by thearms race and making it much easier for them to build thefoundations for a more satisfactory modus vivendi.'

1 HE IDEA OF EINITE or minimum deterrence goesback at least to the later years of the Eisenhower Adminis-tration, when it was advocated by the U.S. Navy. Duringthe Kennedy Administration, Jerome Wiesner, the presi-dent's science adviser, a r^ed that the United States requiredonly a few hundred survivable nuclear weapons. And atabout the same time the Soviet Union v/as offering disarma-ment proposals v/hich were compatible with this approach.^

However, even as the Navy was arg jing that 232 survi-vable Polaris missiles would be "sufficient to destroy all ofRussia," the Strategic Air Command was putting on its tar-get list 645 airfields from which Soviet strategic bombersmight be launched, and thousands of tactical nuclear weap-ons were being deployed to Europe. By the mid-1960s, theU.S. arsenal contained approximately 30,000 nuclear weap-ons—slightly more than today's. The Soviet arsenal grewto comparable levels during the 1970s.

Neither the United States nor the Soviet Union seriouslypursued a finite deterrence posture —on the U.S. side, prin-cipally on the grounds that nuclear weapons must be avail-able for counterforce targeting, that is, for striking at mili-tary targets as a deterrent to Soviet aggression. Each sidehas also sought the ability to mount preemptive strikes to

reduce the nuclear threat from the other.^ Finite deterrencehas also been criticized for resting upon "incredible" and"immoral" threats to destroy cities in order to deter attackson other targets. Critics have also raised concerns that ifnuclear arsenals were much smaller than those of today,they might be more vulnerable to neutralization by surpriseattack or technological breakthrough. These concerns maybest be addressed by discussing a concrete example.

T H E ACCOMPANYING table shows an illustrativefinite deterrence force and compares it with current super-power nuclear arsenals." This may not be the best possiblefinite deterrence force. A strong ai^ument can be made, forexample, that the United States should take advantage ofthe relative invulnerability of its submarine-based forces andshift all of its ballistic missiles to sea. Furthermore, thesuperpowers might—for organizational or other reasons-choose very different mixes of nuclear weapons withinoverall arsenals of approximately equal size. (Such ques-tions will be addressed in future studies by the PrincetonProject on Finite Deterrence.)

The key changes in the transition from the current nu-clear arsenals to the finite-deterrence force in the table are:

• Strategic warheads have been reduced by about 80 per-cent (from about 10,000 to 2,000), in large part by replac-ing multiple-warhead with single-warhead missiles.

• Intermediate-range nuclear weapons have been largelyeliminated, although some land-based missiles might belocated in Europe.

• Tactical nuclear weapons have been eliminated.The resulting force is therefore quite similar to one thatwould be obtained by stripping the current force of its mostdestabilizing elements.

The destructive capacity of the finite-deterrence force isfixed by assuming that each ofthe warheads in the finite-deterrence arsenal has a yield of 100 kilotons. That yieldis at the low end of the range of warhead yields in the cur-rent strategic arsenals of the superpowers, but it is approxi-mately eight times larger than the yield of the bomb thatdestroyed Hiroshima. Such a warhead could destroy, byblast and fire, an area of about 50 square kilometers (20square miles), containing, in a typical large urban area,about 100,000 people. Several such warheads in the illustra-tive arsenal could be tai^eted against every U.S. and Sovietcity with a population of over 50,000.

Eigure 1 shows the results of calculations done in 1967for Secretary of Defense Robert McNamara of the percent-age of the estimated 1972 Soviet urban population thatcould be killed and industrial capacity destroyed as a func-tion ofthe "equivalent megatonnage" used.' The fatalitiesshown at a given level of equivalent megatonnage are signi-ficantly lower than could occur. Many effects —includingthose of radioactive fallout and the impacts of the destruc-tion ofthe economy on the rural and surviving urban popu-lations—appear to have been neglected. Only about 50equivalent megatons would be required to destroy by blastand fire about one half of the urban area of the Soviet

Bulletin of the Atomic Scientists145

U.S. and Soviet nuclear arsenals in 1985 and in an illustrative finite-deterrence (FD) regime

Missiles or bombers WarheadsUnited States Soviet Union FD (each side) United States Soviet Union

Long- and intermediate-rangeICBMs 1,023

Intermediate-range missiles 104(land-based ballistic and cruise)Submarine-launched ballistic 690Long-range bombers 297Subtotals 2,114

Other warheadsArtillery shellsAntisubmarine warheadsAntiship cruise missile warheadsBattlefield ballistic missile warheadsAnti-aircraft missile warheadsAnti-ballistic-missile warheadsAtomic demolition minesNonstrategic bombsOverall total warheads^

1,398

534

967

300

500

200

2,126

104

5,728

3,334

6,420

1,362

2,887

600

FD (each side)

500a

500

1,000

3.199 1.200 11.292 11.269 2.000

2,4002,000

0

300

200

0

600

4,000

900600

1,0001,600

300^32

some4,000

0

0

0

0

0

0

0

0

20.792 19,701 2,000

•'Some of the 500 land-based missiles in the finite-deterrence arsenal might be intermediate-range ballistic or ground-launched cruise missiles.' 'Not including reloads

Sources: For strategic weapons: U.S. Department of Defense, Soviet Military Power, 1984 {Washington, D . C : Government Printing Office, 1984).pp. 24, 26; U.S. Department of Defense, Report of the Secretary of Defense, Caspar W. Weinberger, to the. Congress, 1986 (Washington, D . C :Government Printing Office, 1985), Chart III.E.4 and Appendix C For U.S. forces, we assume 10 warheads per Poseidon, eight per Trident I, eightbombs and short-range attack missiles on all 241 B-52G/Hs and six on each ofthe 56 FB-Uls , and 12 air-launched cruise missiles on each ofthe90 B-,')2G bombers. For Soviet forces, we assume four warheads per SS-i7, 10 per SS-18, six per SS-19, seven per SS-N-18, nine per SS-N-20. Wedo not include Soviet bombers assigned to naval aviation, and we assume an average of two bombs and/or attack missiles per bomber, based onSenate Committee on Armed Forces, Department of Defense Authorizations for Appropriations for FY 1985: Hearings, 98th Cong., 2d sess., Feb.1, 1984, p. 123.

For intermediate-range missiles: Neu/ York Times, April 14, 1985, p. E l .For other nuclear weapons: Nuclear Weapons Databook Staff, in World Armaments and Disarmament: SIPRI Yearbook, 198S (London: Taylor

and Francis, 1985).

Union or a comparable area in the United States.^ Sincea 100-kiloton warhead has approximately 0.2 equivalentmegatons* destructive power, if each warhead in the 2,000-warhead illustrative finite-deterrence arsenal had a 100-kilo-ton yield, the arsenal's total destructive power would bealmost 400 equivalent megatons.

This article addresses a series of questions about thisfinite-deterrence force: Would a commitment to finite deter-rence brake the arms race dynamic? Is it moral? Would itdeter? Would it be stable? Would it be adequately verifi-able? And is it realistically achievable?

T H E N U C L E A R A R M S race is driven largely by at-tempts to make nuclear weapons more "usable" and todevelop combinations of first-strike and defensive capabili-

*The area that could be subjected to a certain level of blast overpressurevaries as the two-thirds power of the yield (Y) of a nuclear weapon. Thisfaa is captured by measuring the potential area-destructiveness of a nuclearwarhead by its "equivalent megatonnage," Y-". The equivalent megaton-nage of a nuclear warhead with less than one-megacon yield is larger thanits megatonnage. Above one megaton, the situation is reversed.

ties that would make possible escape from the mutual host-age relationship. But no matter how technically sophisti-cated nuclear weapons systems have become, the mutualhostage relationship has made them unusable and that rela-tionship itself has proved to be very robust.

Figure 2 illustrates this situation dramatically. Despitethe recent U.S. scare about a "window of vulnerability," dur-ing a crisis neither superpower could reduce the other's stra-tegic arsenal by more than about a half— far from the hun-dred-fold reduction required even to begin to loosen the gripof the mutual hostage relationship. Efforts to escape fromhostage through defense appear similarly hopeless —somuch so that the superpowers agreed in the 1972 ABMTreaty not even to try. And few independent analysts seeany escape through new generations of counterforce weap-ons or the proposed Star Wars defenses.

Adoption of finite deterrence would require acceptanceof the implications of the mutual hostage relationship andtherefore a surrender of the illusions that drive the armsrace. As Admiral Arleigh Burke, then chief of naval opera-

August 1985146

Figure 1. Potential consequences from blast alonein an all-out attack against Soviet cities

McNAMARA'S-"ASSUFIEOOESTRUCTION

LEVEL

•"Megatonnage estimate for fatalitiesare several times tno high.

100 200 400

The fatality levels shown could result from a much luwer level ofequivalent megatonnage.

Sources: Robert S. McNamara, The Fiscal Year 1969-73 Defense Pro-gram and the 1969 Defense Budget (Washington, D . C : Department ofDefense, 1969). The fatality levels shown could result from a much lowerlevel of equivalent megatonnage: see Frank von Hippel, "The Effects ofNuclear War," in David W. Hafemeister and Dietrich Schrocer, eds..Physics, Technology and the Nuclear Arms Race (New York: AmericanInstitute of Physics, 1983).

tions, argued almost three decades ago: if the superpowersabandoned the false hopes of "winning" through new coun-terforce or defensive systems, the rationale for new weaponswould be greatly weakened.'

M U T U A L DETERRENCE depends fundamentally onthe possibility that any large-scale direct confrontation be-tween the United States and the Soviet Union could leadto untold destruction. This is true both for current nucleararsenals and for i:he finite-deterrence arsenal proposed inthe table. In terms of this immense destruction, the threatsimplicit in both postures must be viewed as immoral. Never-theless, some have argued that a finite deterrence posturewould be particularly immoral because its smaller size im-plicitly emphasizes the threat to cities.

To acknowledge that threat, however, is not to insist thatcities be targeted in the event of a nuclear war. Nothingin the configuration of the illustrative finite deterrencearsenal would require the targeting of population centersrather than, for example, military installations. Nuclearstrategists currently can fantasize that thousands of nuclearweapons could be so used, but by the time even hundredsof them had been used against, for example, military targetsin Central Europe, civilian fatalities would number in themillions, command and control networks would be collaps-ing, and the chances of limiting the war would be rapidlyvanishing.* Under these circumstances., the moral distinc-tion between targeting military facilities and tai^eting citieswould have become nearly irrelevant.

Therefore, adoption of a finite-deterrence posture wouldin no way reduce the superpowers' abilities—if they sowished —to avoid mass slaughter in a nuclear war. It wouldsimply strip away the dangerous self-deception that a warcould be fought with thousands of nuclear warheads with-out destroying civilization. This realization is a moral ad-

vantage of the finite-deterrence posture. Moreover, to theextent that a finite-deterrence posture would reduce the pro-bability of accidental nuclear war and, in the event of all-out nuclear war, would inflict less overall destruction —especially on noncombatant nations —it also has a moraladvantage. The superpower allies would be attacked bymany fewer warheads, noncombatant nations would receivemuch less radioactive fallout, and the global environmentwould be less severely altered by the effects of ozone des-truction and smoke. (Such advantages, however, must notbe offset by increasing the average yield of the smallernumber of warheads.)

Political leaders understand that the mere possibility ofcatastrophe inherent in the mutual hostage relationship —not the details of the arsenals or the plans for targeting

Figure 2. The futility of counterforce: calculated results ofstrategic counterforce exchanges, 1985 forces

These calculations assume that in a first strike the SovietUnion assigns two ICBM warheads to each U.S. Minute-man silo, that the United States assigns two MinutemanIII warheads to each Soviet silo containing a MIRVedICBM, and that 80 percent ofthe missiles so attacked aredestroyed. It is also assumed that both sides are on gen-erated alert with as many bombers on alert and ballistic-missile submarines at sea as possible. This figure is an up-date of one whose derivation is explained in greater detailin Harold A. Feiveson and Frank von Hippel, "The Freezeand the Counterforce Race," Physics Today (Jan. 1983), p.36.

12000

8000

4000

BEFORE COUNTERFORCE AFTER SOVIETWEAPON EXCHANGE FIRST STRIKE

AFTER U SFIRST STRIKE

SUBMARINELAUNCHEDBflLLISTtcfMISSILES) I

U.S. SOVIETS

8000

4000

0

2 0 0 EMT

Bulletin of the Atomic Scientists147

them —is what deters each superpower from threatening thevital interests of the other. McGeorge Bundy has termedthis "existential deterrence."^ As Bundy wrote during thegreat debate over anti-ballistic-missile systems in the late1960s:

I

Think-tank analysts can set levels of acceptable damagewell up in the tens of millions of lives. They can assumethat the loss of dozens of great cities is somehow a realchoice for sane men. They are in an unreal world. Inthe real world of real political leaders—whether here orin the Soviet Union —a decision that would hring evenone hydrogen homb on one city of one's own countrywould be recognized in advance as a catastrophic blun-der; ten bombs on ten cities would be a disaster beyondhistory; and a hundred bombs on a hundred cities are un-thinkable.'^

The destructive capacity in the illustrative superpower finitedeterrence arsenal, although one-tenth the size of today's,is, nevertheless, many times greater than that required toaccomplish even an "unthinkable" level of destruction.

Of course, existential deterrence would also exist forwhat has come to be termed "extended deterrence"—thatis, not merely of nuclear attacks against the United States,but also of non-nuclear attacks against U.S. allies, parti-cularly in Europe. For its entire history, NATO has reliedupon the threat posed by U.S. nuclear weapons to makeup for what has always appeared to be an imbalance ofconventional forces in favor of the Warsaw Pact. Indeed,most of the U.S. nuclear arsenal is justified ultimately notby the need to protect the United States itself against nu-clear attack, but as a deterrent to Soviet aggression in areasof U.S. vital interest.

Yet, ever since the Soviet Union achieved a secure second-strike capability in the early 1970s, extended deterrence hasbeen largely a matter of doctrine and faith. Indeed, thedesire to make plausible the U.S. willingness to riskAmerican cities for the sake of the European allies has beena powerful motive in the constant search for additional cre-dible "nuclear options" and more "usable" nuclear forces.

Despite the elaboration of nuclear options, however, ex-tended deterrence seems no more (and, indeed, no less)plausible today than it did, say, two decades ago. Now, asthen, extended deterrence depends not upon any imbalancein nuclear capabilities but upon perceptions of relative will-ingness to risk nuclear war. If Moscow is now deterred fromlaunching a conventional war in Europe because of its in-herent uncertainty about whether the West would attemptto stem the tide with nuclear weapons, there is demonstra-ble reason why the same deterrence would not apply if eachside possessed 2,000 warheads.

The character and size of the illustrative superpowerforces have been largely determined by the design require-ment that the current degree of stability should exist afterdeep reductions. Despite the 90 percent reduction in thetotal number of warheads shown in the table, the numberof U.S. "delivery vehicles" has only been reduced by about

one-third. Assuming that Soviet nuclear forces would bereduced similarly, the U.S. finite-deterrence arsenal wouldbe less vulnerable than the current arsenal because theSoviet Union would have available many fewer warheadsper target for counterforce attacks. In addition, because ofthe deMIRVing, more than one ballistic missile warheadwould be required to destroy one ICBM warhead on theother side.

Calculations such as those done for Eigure 2 show thatabout half of the 2,000 warheads in the finite deterrencearsenal would survive a first strike. This result dependsprimarily on assumptions made about the percentages ofbombers that would be on alert during a crisis and ofballistic missile submarines that would be at sea, not onthe number of warheads used in the attack. Thus, even thegreat reductions envisioned here are not enough to destabil-ize the superpower strategic balance. That would occur iffurther reductions reached the point where such details as,for example, which side struck first or had more capablenon-nuclear forces once again began to matter.

The survivability ofthe illustrative finite-deterrence ar-senal could be further enhanced by making the single-war-head, land-based missiles mobile (if this could be donewithout making their numbers inadequately verifiable) andby distributing the single-warhead, submarine-launchedmissiles among a larger number of smaller submarines.

To discourage new threats to the stability of this situa-tion, the establishment of a finite-deterrence regime shouldbe accompanied by verifiable bans on the development ofnew types of weapons such as reentry vehicles that could"home in" on bombers in flight. Strict limitations on ballis-tic missile flight tests would severely hamper the develop-ment of such weapons and the pursuit of counterforce stra-tegies more generally. Placing restrictions on the deploymentof antisubmarine-warfare technologies would also be valu-able.

Because the number of ballistic missile reentry vehiclesthat defenses would have to deal with would be greatly re-duced, the importance of restraining defensive technologieswould be increased. Therefore, the ABM Treaty should bestrengthened in the gray areas where anti-tacticai-ballisticmissile and anti-aircraft defense capabilities overlap withanti-strategic-ballistic missile capabilities.

More worrisome than the vulnerability of nuciear weap-ons is the vulnerability of the superpower nuclear com-mand-and-control systems. Even after completion of thecurrent ambitious upgrade ofthe U.S.-command-and-con-trol system, its designers believe that it could, at best, with-stand an attack involving "a few hundred" nuclear warheadsbefore losing positive control over the U.S. arsenal." Thismismatch between the number of weapons in the super-power nuclear arsenals and the survivability ofthe systemsthat direct them could raise pressures for preemptive useof the weapons before centralized control was lost, andwould also encourage excessive decentralization of controlduring a crisis. "Decapitation" of either superpower's nu-clear weapons system could well result m a globally catas-

August 1985148

trophic reflexive "spasm" attack. Deep reductions wouldnot eliminate the vulnerability of command-and-controlsystems, but they would limit the number of warheads boththat are avaiiabie to attack them and that these fragilesystems currently control.

Eiimination oFtacticai nuclear weapons woiiid also con-tribute to stability. The superpower armies and navies arenow so thoroughly equipped with nuciear weapons forevery purpose (about one nuclear warhead per one hun-dred military personnel on average) chat there would beenormous risk of crossing the nuclear threshold in the eventof any iarge-scaie confrontation between the two militaries.This would derive from the myriad ambiguities and com-plexities inherent in the intermingling of conventional andnuclear forces on both sides, from the temptation to pre-emptively attack concentrations of nuclear arms, and fromappeals from units about to be overcome to higher authori-ties to authorize use of their nuclear weapons.

The above discussion can be turned around to argue thatthe fragility of command and control and the nucleariza-tion of tacticai forces enhance deterrence by increasing thedanger that any military confrontation between the UnitedStates and the Soviet Union might result in nuclear war.Such a prospect., no doubt, does help to instill caution onboth sides. But beyond a certain point rationalizing brittle-ness in this way becomes the irresponsible advocacy ofdeterrence by an all-purpose doomsday machine.

/ V N O T H E R K E Y criterion for the illustrative finite-deterrence force was that no credible level of undetectedcheating could allow either superpower to remove itselffrom hostage. If the forces are as large as those suggestedand are adequately survivable, then even the secret doublingof the strategic weapons available to one side would notsignificantly alter the mutual hostage relationship.

It appears that the most critical changes in the transitionto a finite-deterrence regime —reductions in the numbersof long-range bombers and replacement of large multiple-warhead missiles by smaller single-warhead missiles—couldbe verified by nonintrusive means such as satellites. A banon testing MIRVed missiles would be verifiable by long-range monitoring techniques and would, over time, erodeconfidence in the usability of any hidden stockpile ofMIRVed missiles, especially for a first strike.

Although a 100-kiloton limit on the yield of nuclear war-heads may not be verifiable (in this range, each additionalkilogram of warhead weight can result in an additional yieldof about one kiloton'^), a limit in the range of a few hun-dred kiiotons —a typical yield for the individual warheadson current multiple-warhead ICBMs— ought to be enforce-able. This could be done by limiting the throw-weights ofthe new singie-warhead bailistic missiles and the sizes ofcruise missiles.

Some aspects of a finite-deterrence regime, however,wouid be more difficult to verify and would probably re-quire cooperative verification arrangements. For exampie,on-site monitoring will be necessary to verify the dismant-

Important Books by

Robert Jay LiftotiHome from the WarVietnam Veterans: NeitherVictims nor Executioners

Here is a timely reissue ofthe moving andcompassionale book about Vietnam vete-rans, their extremes of pain and aliena-tion, and their search for connection andcontinuity, written by the noted psychia-trist and anti-war activist."Lifton was early and he was right. Torover a decade, this tx>ok has helped tensof thousands of veterans and theirfamilies and counselors."—ARTHUR S. BLA !̂̂ JR., M.D., Director,Readjustment Counseling Service, Vete-rans Administration"Dr. Lifton is a distinguished psychiatristwho is deeply concemed with the effectsof aggression, war, and revolution onman's mental stability. . . . This volumeis a brilliant study."—Chicago Tribune Book Worldpaper, $9.95

Death in LifeSurvivors of Hiroshima

'The primary thrust of Ltfton's bookconies from his study of hlbakusha, thetainted survivors of Hiroshima.. . . Deathin Ufe is a magnificent book on severalgrounds. Its implications for future psy-chosocial research are significant. Dr. Lif-tons literary skill and his psychologicalacumen blend harmoniously, as he for-mulates what is clearly beyond formula-tion."^^VERYWEISHAri, M.D., fiychiatiyand Social Science Reviewdoth, $19.50 paper, $10.50

Indefensible WeaponsThe Politicai and Psychological

Case Against nuclearismwith Richard Falk

In this hard-hitting and closely reasonedbooK the authors provide a psychologicaland political antidote to the nuciear armsrace."This tXK)k effectively shatters the systemof illusions that sustains the nuciear armsrace. . . . It must be read by all those whowish to involve themselves in the struggleto prevent a nuclear holocaust, who seekto rid the world of the disease of nuc-learism, and to wrest control of our fatefrom the hands of those whom Lifton andTalk call the 'nuclear scenarists.'"^JOtin MACK, H.D., Harvard Universitycloth, $15.50 paper, $6.95

Basic Books, Inc.10 East 53rd St.

Mew York n.Y. 10022To order call 800-636-3030

Bulletin of the Atomic Scientists149

ling of nuclear weapons and the "burning" of the recoveredfissile material in nuclear power reactors and to ensure thatnuclear power installations are not being used to producefissile material for new warheads.

Smail, mobile missiies could present serious verificationproblems since they would be much more difficult to countthan current missiies which are relatively large and fixedin massive silos. This tradeoff has been left unresoived inthe iilustrative force. Sea-launched cruise missiles (SLCMs)pose a simiiar dilemma. Putting them on attack submarineswould greatly increase the number of submarines armedwith iong-range nuclear delivery systems that would haveto be destroyed in a disarming first strike. Their locationand smail size, however, would make them virtually impos-sible to count. As iong as SLCMs are deployed, ail attack-submarines and major surface ships will have to be assumedto be nuclear cruise-missile carriers.

The "denuclearization" of short-range systems such asfighter-bombers, sea-based and ground-launched cruisemissiles, short-range ballistic missiles, and artillery wouldaiso be relatively difficult to verify. The systems themselveswould stili exist to fire conventional munitions, and the nu-clear warheads, which are quite small, could be quicklydelivered from secret stockpiles. Successful concealment ofsome nuclear warheads for short-range delivery systemswould not, however, threaten the mutual hostage relation-ship.

J\ SUPERPOWER transition to a finite-deterrenceregime would affect other nations. Indeed, it would becritical to make the transition in a way that did not dis-rupt international relationships. For example, withdrawalof nuclear weapons from Europe would have to be donein a way that wouid give maximum reassurance to the Euro-peans. In fact, the accompanying table allows for the pos-sibility that some intermediate-range land-based missilesmight be based in Europe.

Complications would also arise in dealing with the"medium" nuclear powers —France, the United Kingdom,and China. Although the superpower arsenals —measuredby numbers of delivery vehicles—would still be an orderof magnitude larger than those of the medium powers, ifFrance and the United Kingdom completed MIRVing theirsubmarine-launched bailistic missiles (SLBMs), the Sovietswouid find arrayed against them a number of warheads farexceeding their own. Even if 2,000 warheads are enoughto pose an effective deterrent against ali these forces, thepoliticai appearance of such an imbalance might be unac-ceptable. For that reason it would almost certainly be nec-essary to induce these medium nuclear powers to limit thesize of their forces. The Chinese government has stated thatit would consider constraints on its nuciear forces if theUnited States and the Soviet Union cut back their nuclearforces by 50 percent."

Such difficulties and the tremendous inertia of the armsrace would have to be overcome if drastic reductions ofthesuperpower nuclear arsenals are to be achieved. As a result

of the debates over the nuclear weapons freeze and StarWars proposals, however, the political conditions for aradical change in the current postures may be more favor-able than they have ever been. The finite-deterrence pro-posal would also represent a solution to the problems thatare stalemating current U.S.-Soviet arms control negotia-tions. The United States has been insisting upon reductionsin the numbers of Soviet MIRVed land-based missiles, andthe Soviet Union has been insisting that the United Statesnot proceed with its Star Wars program. The finite-deter-rence proposal, by eliminating MIRVed missiles and main-taining stringent limitations on anti-ballistic-missile sys-tems, would meet both of these concerns.

The arsenal described here should be about as effectiveand survivable a deterrent as the current superpower arsen-als. It should therefore be technically possible for eithersuperpower to adopt a finite-deterrence position unilateral-ly. This would seem unrealistic politically, but, given thesuperpowers' vast excess of avaiiabie nuclear forces, manyofthe steps toward a finite-deterrence regime could be takenindependently. For example, NATO couid unilaterally denu-clearize a large part of its artillery and short-range missiles.Since the military value of these area-destruction weaponsis increasingly being seen as marginal in an era of precision-guided munitions, there is already broad support for sucha move.

Therefore while a transition to a finite-deterrence regimewould be difficult, it should not be impossible. The resultwould still be a balance of terror with the same caution-inducing characteristics as the current regime —but withsome of its overkill and its dangerous and mind-twistingcomplexity stripped away. D

1. Richard H. Ullman, "Denuclearizing Intemational Politics," Ethics,95 (198.5), pp. 567-88.

2. U.N. General Assembly, "Revised Draft Treaty on General and Com-plete Disarmament under Strict International Control," in Documents onDisarmament, 1962. p. 913.

3. Desmond Ball, Targeting for Strategic Deterrence, Adelphi Paper 185(Lxindon: International Institute for Strategic Studies, 1983); DavidHolloway, The Soviet Union and the Arms Race (New Haven, Conn.;Yale University Press, 1983).

4. In a number of his talks, Richard L. Garwin has described finite-deterrence forces similar to the one discussed here.

5. Robert S. McNamara, The Fiscal Year 1969-73 Defense Programand the 1969 Defense Budget (Washington, D.C: Department of Defense,1968), pp. 50, 57.

6. Frank von Hippel, "The Effects of Nuclear War," in David W, Hafe-meister and Dietrich Schroeer, eds.. Physics, Technology and the NuclearArms Race (New York: American Institute of Physics, 1983), p. 1.

7. See, for example, David Alan Rosenberg, "The Origins of Overkill,"International Security (Spring 1983), p. 3.

8. William M. Arkin, Frank von Hippel, and Barbara G. I^vi, "TheConsequences of a 'Limited' Nuclear War in East and Wesi Germany"Ambio (June 1982}, p. 163.

9. McGeorge Bundy, "Existential Deterrence and Its Consequences,"in Douglas Maclean, ed.. The Security Gamble: Deterrence Dilemmasin the Nuclear Age (Totowa, N.J.: Rowman and Allanheld, 1984), pp. 3-13.

10. McGeorge Bundy, "To Cap the Volcano," Foreign Affairs (Oct.1969), p. 2.

11. Charles A. Zraket, "Strategic Command, Control, Communications,and Intelligence," Science, 224 (1984), p. 1306.

12. Thomas B. Cochran, William M. Arkin, and Milton M. Hoenig,U.S. Nuclear Forces and Capabilities (Cambridge, Mass.: Ballinger, 1984).

13. New York Times, June 22, 1983, p. 2.

August 1985150