i 442 Jul y 5, 1985 IN THE KEY BATTLE OF COMPARATIVE COSTS STRATEGIC DEFENSE IS A WINNER INTRODUCTION To its critics, Ronald Reagan's Strategic Defense Initiative is a dangerous and costly quicksand that will swallow ever greater U.S. resources while providing litt l e new protection. The price of maintaining and improving the so-called Star Wars defense, say these critics, always will be substantially greater than MOSCOW'S cost for countering or negating the defense. This would mean, in military argot, that the "cost -exchange ratio" always would be against the U.S. missile defense efforts and in favor of MOSCOW~S attacking missiles.! For strategic defense critics, this unfavor- able ratio is a major drawback of the Reagan proposal.

The critics would be right if their premise were valid. It possibly was, during the last national debate on missile defense in the late 1960s. But since then, technological leaps have been shifting the offense-defense cost-ratio to favor defense. These changes include declining costs of com puters, communications sensors, and the lift of strategic defense components into space.

At the same time, these advances are having far less impact on offensive ballistic missiles, because they depend on relatively mature technologies. There thus is evide nce that most of the presumed Soviet countermeasures to Reagan's Strategic Defense Perhaps the most extreme statement on countermeasure costs has come, not unexpectedly, from a Soviet source A 'highly efficient countermeasure system' would cost only 1 to 2 percent of the cost of the Star Wars sys tem I' Dusko Doder, "Soviets See U.S Deception The Washington Post, January 7, 1985, p. A-1 This is the tenth in a series of Heritage Backgrounders examining strategic It was preceded by "Strategic Defense and Ame rica's Allies," Back- defense. grounder No. 425, April 16, 1985. 2 Initiative (SDI) will be costlier, and probably less effective than alleged by SDI critics.

The cost-exchange ratio therefore no longer is a sound reason for opposing SDI. Technology and do ctrine, and hence costs, do not stand still. Assumptions that the offense will inevitably defeat the defense are based on old technologies and doctrine and need careful reexamination. Also needing reexamina tion is the much-cited trillion dollar cost? for SDI. If this figure, pulled out of the air, has any meaning, it is as the total system price tag over several decades and covers, among other things, an extensive system of several layers, including 21st century boost-phase defenses (the Itlong poles in t h e tent And there are many factors, moreover, such as increased Soviet emphasis on developing and deploying defenses rather than offensive countermeasures or an arms control agreement limiting offenses to relatively low levels, that could limit the require d components and costs of SDI NEW TECHNOLOGY Some SDI concepts were considered and abandoned in the 1960s It is the great technological progress since then that makes the current reconsideration very timely. Some of these developments may help the design o f cheap countermeasures; but many more will help reduce the costs of ballistic missile defense BMD compared to those of strategic offense and countermeasures.

General Technolouical DeveloDments The major areas of technology development and cost reduction h ave been Microchi s The cost of electronic devices may be dn dollars per "digital logic gate or basic computing unit. In the late 1950s, the best vacuum tubes were so expensive that each gate cost about 10. By 1963, early transistors had reduced this to l ess than a dollar. The subsequent revolution in solid state electronics and miniaturization has reduced the current cost per gate to two one-hundred thousandths of a dollar (or 0.00002 This is expected to be cut in half again by 19

87. The down trend probably has not yet stopped Com uters. A $3,000 personal computer of today has kly the capabilities of a 5 million IBM 360-40 ltmainframelt computer of the mid-1960s (conser 6 See, for example A trillion here, a trillion there," Oakland Tribune Aug u st 26, 1984, and R. Jeffrey Smith, "Schlesinger Attacks Star Wars Plan Science, November 9, 1984, p. 673 Data are based on private communication from W. Russell Young, SRI Inter national, Washington D.C Strategic Studies Center j 3 vatively 15 million in t oday's dollars). This is a reduction in cost by a factor of 5,000, or about 35 percent per year. Moreover, the 1960s machine required an air-conditioned room, plus a corps of trained operators and maintenance personnel; today's PC requires a desk top, one relatively briefly trained operator, and virtually no maintenance Software. Increasingly, the labor-intensive program ming requirements of computer systems are proving to be the bottlenecks and 'Icost drivers." This is I particularly true of the computing for ABM tracking l and battle management (as it was 15 to 20 years I ago for the NIKE-X SENTINEL and SAFEGUARD systems But increasing attention to cheaper "firmll or Ilhardll wiring of parts of the programming into the computer hardware, distributed compu t ing in sub systems, more efficient algorithms (rules for solving problems artificial intelligence, and computer processing and data management systems are contributing to eliminating software problems strategic defenses. that might impede further developm e nt of effective I Fiber 0 tics. In communication by light waves are transmitted by means of laser driven signals sent through glass fibers, or Ilightwave guides Optical computation tech nologies are also in the offing.) Detailed figures on the costs of fi ber optics are hard to come by because the technology has been moving so rapidly since its commercial introduction in 19

80. However the cost per channel per kilometer, where demand is heavy, is already a small fraction of that of microwave transmission,4 and the rate of continuing cost reduction appears to be even greater than that of computing. Other advantages-=translating into cost savings--include the lightness and compactness of the material, the resistance to electromagnetic,pulse EMP) damage and in t erfer ence, the security of the system against Iltappingll or eavesdropping, and its suitability to digital transmis sion. Fiber optics are being used for satellite ground station systems and communications The cost of the fiber itself is declining at alm o st 30 percent per year while the number of channels per fiber and the distance between "repeaters for reamplification--a high-cost system element--are increasing rapidly and the rate of data transmission is rising even more rapidly as both laser and trans m ission advance. Executive Briefing Optical Communica tions and Mobilization for Policy, F.P. Hoeber, et al SRI International, Strategic Studies Center, May 1982 A Case Study for the Undersecretary of Defense 4 Satellite Communications. Boost-phase interce p t systems will still require two-way satellite-ground and inter-satellite communications. Satellite communi cation costs are coming down, albeit less rapidly than other SDI-related technologies. Increased satellite lifetimes (from 3 years or so around 197 0 to 10 years today multichannel and multifrequency transponders, single-side band AM transmission, and other technological advances are contributing to the dropping costs.

ImDlications of General Technolow DeveloDment The evidence suggests that the technologies cited already have advanced strategic offense capabilities as far as they can.

On the other hand, they can still contribute substantially to defense improvements. Example: while these technologies are the key to missile guidance and accuracy, 97 percent of what can be done in offensive missile accuracy already has been accomplished.

The average distance by which missiles miss their targets has declined from roughly five miles in the early 1960s to under 100 meters for the Pershing I1 today. Guidanc e packages are now such a small part of the throw-weight of missiles that further improve ment in miniaturization will be of marginal value in multiple warhead missiles ABM technologies, on the other hand, are still underdeveloped. I ABM systems will be h e avily dependent on computing and communica tions to discriminate real missiles and warheads from decoysf I determine trajectories, determine which targets to shoot at and I 1 I in what order, and "hand offtt data on the nature of the threat to other layer s of the system. The huge reductions in the cost and weight of these computing and communications technologies therefore will yield enormous savings for strategic defensive systems evolving from SDI research and development.

Savings will result not only fr om the reduced cost of the defensive system components but also from reduced weight, which lowers the cost of lifting the components into space. Effective battle management, command, control, and communications (BM-C3 also will allow more efficient use of weapons, potentially further reducing lift requirements. The cost savings from effective and lightweight BM-C3 will play a large role in shifting the offense defense cost-exchange ratios toward defense.

Strateqic Defense Technology Developments Other majo r technology developments include 0 Smaller, cheaper phased-arrayf solid-state, elec tronically steered radars for tracking rather than just early warning For a simple description of these mechanisms, see Ashton B. Carter and David N. Schwartz, eds Ballis t ic Missile Defense (Washington, D.C The Brookings Institution, 1984 pp. 68-69. 5 0 Directed-energy weapons (DEW from lasers to particle beams 9 Homin capabilities against missile warheads w ich were not feasible when the NIKE-ZEUS and NIKE-X ABM systems w e re developed in the 1960s 0 Kinetic energy weapons (KEW utilizing homing capability, which destroy the target missile or warhead by direct impact with fragments from an explosive warhead or with a small, self-guiding warhead. These may be rocket-launched, as in the successful HOE (Homing Overlay Experiment) of June 10, 1984, rocket-powered, or electromag netically launched (the ItEM rail gun 0 ifDistributed,lf or decentralized, ground-based systems, which remove the vulnerability of the system to a hit on a main radar and/or computer processing unit 0 Lonq-wave infrared (LWIR) detection and tracking systems, which can be an adjunct to or substitute for radar in the midcourse regime 0 Feasible improvements in heavy-lift launch vehicles for space-borne assets . The cost of Shuttle launches is falling as usage increases and turn around time is reduced. Costs of expendable launchers are also declining. Follow-on systems for heavier lift should be much cheaper per pound lifted, because of better technology and eco nomies of scale.

These and other new technologies offer real promise for reduced defense system costs as well as higher performance.

The most basic changes in technology may prove to be advances in ABM potential for space-borne boost-phase and early-midco urse intercept, and for ground-based terminal and late-midcourse defense. These should be compared to the relatively mature state of offensive ballistic missile (especially booster) technology.

In particular, the Soviets have invested billions of dollars in their current intercontinental ballistic missile (ICBM) systems.

New systems might have to be started from scratch and paid for in costly future outlays. At the same time, current Soviet systems would have to be maintained See Brian Green Strategic Def ense: The Technology That Makes It Possi ble Heritage Foundation Backgrounder No. 375, August 23, 1984.

SOVIET COUNTERMEASURES TO EACH DEFENSE LAYER SDI has envisioned a defense that can intercept missiles at different phases of their flights. These phase s include the boost phase, lasting three to five minutes, from the time the missile is launched until the missile burns out the post-boost phase, lasting up to 10 minutes, during which the post-boost vehicle (the llbusll sets the multiple warheads and pen e tration aids on their independent courses 0 the midcourse phase, lasting up to 20 minutes, during which the warheads and penetration aids glide along their flight trajectories the terminal phase, which lasts only a minute or so,'during which the warheads r eenter the atmosphere and arrive at their targets. In the context of a defense against ballistic missile attack, the terminal phase can be subdivided into Illow exoat mosphericll (just before the warheads reach the atmosphere; there is no sharp dividing l ine between the midcourse and terminal phases high endoat mosphericll (just after the warheads reach the atmosphere) and Illow endoatmospheric just before the warheads reach their targets).

Intercepting missiles and their warheads in each phase of flight b y deploying what amounts to vertical layers of defenses complicates the task of the attacker and should permit high attack attrition rates. It is not yet clear, however, how the layers should be proportioned or, when costs are considered, what the optimal number of layers would be.

A complete vertically layered defense must await development of weapons to intercept missiles in the boost- and post-boost layers. This probably will take a dozen years or more. Mean while, different layers using different technologies could be deployed a s they became feasible. Recent cost estimates are $60 to $100 billion for a terminal defense that protects the U.S. missile silos combined with a boost-phase intercept defense.

These could be ready by the early 1990s Soviet countermeasures would take consi derable time and rubles for Moscow to develop and would require knowledge of U.S I Zbigniew Brzezinski, Robert Jastrow, and Max Kampelman, "Defense in Space Is Not 'Star Wars New York Times Magazine, January 27, 1985, p. 28ff, gives a figure of $60 billio n through the early 199Os, and "Two Analysts Put Cost of Antimissile Program at $70 Billion by 1993 The New York Times, February 12, 1985, gives a range of $70 to $100 billion. 7 defense plans. While the Soviets would be expending considerable resources tr ying to counter the first phase of U.S. defenses, the U.S. could be working on counter-countermeasures.

Boost-phase Countermeasures Intercepting a missile shortly after it is launched (the boost-phase) is potentially very cost-effective. With one inter cep tion, all of the missile's multiple warheads and/or penetration aids (devices carried and released by the missile to fool the defense) can be destroyed.

One way that Moscow could try to foil U.S. boost-phase intercept is to develop a new generation of fas t-burning, high acceleration solid-fueled boosters. These would make intercep tion by the U.S. more difficult by shortening the time available for the space-based defenses to acquire, track, and attack the boosters. This timespan could be cut from the cur r ent 3 to 5 minutes to perhaps 1 to 2 minutes. These faster boosters would also burn out within the atmosphere. Thus the U.S. would be limited in the kinds of boost-phase technologies that it could utilize since X-ray lasers and some types of particle beam weapons cannot penetrate the atmosphere.

Such a countermeasure, however, would be costly, particularly in-terms of weight, since fast-burn boosters of a given size must carry more fuel and thus have less room for their weapons' payload.

To launch the sam e number of warheads and penetration aids therefore, would require the launching of more missiles. The cost of these new missiles, each burdened by the added weight for fast-burn boosters, could well be greater than that of some boost-phase defensive syst ems. Further, their deployment would be at least ten years after Moscow learned that the U.S. had selected this defense system for development, which would mean a significant lag between the deployment of a given defense and specific countermeasures.

Shiel ding, Spinning, and Maneuver Another means suaaested for neutralizina anti-missile weax>ons I is to add shielding tG-offensive missiles, wkch would reflect lasers and absorb other types of beam weapons. It is very uncer tain, however, whether reflective c oatings could defeat lasers.

Shielding against particle beams also would add substantially to I missile weight against kinetic-energy-weapons, which are propelled at speeds of 15,000 or more miles per hour and shatter almost anything they hit.

Another pos sible way to defeat an ABM system, it is said, is to spin the booster of the offensive missile. Spinning minimizes the amount of energy absorbed at any given spot, and thus the lasers and other beams aimed at the missiles would fail to disable them. Yet s p inning should not be hard to counter. The U.S could design beam weapons to deliver huge energy bursts in very short pulses, destroying their targets by impulse or shock rather than with heat I And such shielding would offer no protection 8 It further is s u ggested that Soviet missiles could be designed to maneuver to evade U.S. anti-missile devices. But this would require completely redesigned, throttled rocket engines, as well as more elaborate guidance systems. Not only does this appear impractical in ter ms of cost, it very likely would reduce the Soviet missile's accuracy. Maneuver also would be ineffective against directed-energy, llspeed-of-lightll weapons which reach their targets very quickly, before they can maneuver evasively.

Direct Attack Direct a ttack on U.S. BMD satellite battle stations is also mentioned as a potential Soviet countermeasure. However, because Soviet anti-satellite missile boosters would be similar to ICBMs in boost phase, and since U.S. BMD satellites would be designed to interc ept missiles in boost phase, the U.S. system could intercept anti-satellite boosters as easily as ICBM boosters.

Blinding or destroying the BMD satellites with Soviet ground based lasers would be a more formidable threat. However, the U.S. satellites could be shielded, and MOSCOW'S ground-based lasers might be counterattacked. The technological uncertainties are such that the cost-exchange ratios in this case probably cannot yet be estimated. One long-range. proposal for defense has been the mining of aste roids for materials to shield U.S. BMD satellites, a process potentially much cheaper than lifting the mass required from the earth; however, the technological and arms control implications of this, too, cannot yet be evaluated.

None of these Soviet countermeasures thus has any certainty of success. The only certainty is that they are all very costly.

Most current Soviet ICBMs could not be modified to accommodate the countermeasures effectively. For example, the liquid-fueled SS-18 burns relatively slowly and cannot burn out at low altitudes.

Consequently, the Soviets would have to redesign their ICBM force to a significant degree and deploy an almost brand-new ICBM force in order to obtain any countermeasure benefits available from fast-burn boosters. Their ICBM production lines would have to be rebuilt or extensively modified-all at substantial cost. This new missile force almost surely would have to be larger than the current Soviet force, since the added weight of shielding would mean that each missile could carry fewer warheads.

The enormo us cost of redesigning and rebuilding the Soviet missile force, combined with the lack of certain success, makes the cost ratio very promising for defense would seem to favor the anti-ICBM systems,, the marginal cost of ICBM kills by U.S. satellites could be expected to be below the marginal cost to Moscow of adding new weapons to attack the satellites Since the cost ratios Post-boost phase This phase is important for defense against missiles carrying a number of warheads, or multiple-independently targete d reentry 9 vehicles (MIRVs). Such missiles take up to ten minutes to dispense their reentry vehicles or RVs A U.S. strike against the llbuslf that carries the MIRVed warheads midway through the post-boost phase could potentially knock out half of the miss i les' reentry vehicles (RVs) and penetration aids. The difficulties for the U.S. in attacking a Soviet weapon in the post-boost period are the brief duration of this phase and the bus's smaller "signa ture" (the specific radiation characteristics--heat, li ght, or reflected radar waves-=that can be detected by defensive sensors).

But for little or no additional cost to the U.S. anti missile system, weapons designed for boost-phase and midcourse intercept also could generally attack the bus. Specific possible Soviet countermeasures are limited. Since each RV must be placed on a very precise, preplanned trajectory, the only maneuvering possible for the offensive vehicle is aiming at the target, and it thus could not overcome U.S. defenses. Hardening or shieldi ng RVs would add weight and thus cost. If the Soviets design the bus to maneuver, the extra fuel needed adds even more weight to the bus.

Of course, the bus's vulnerability could be avoided if the missile dispensed its multiple warheads on their flight pat hs while still within the atmosphere. This would be very difficult however, involving additional cost, some additional weight penalty and potential degradation of accuracy.

Midcourse The midcourse phase begins after all RVs and potential penetration aids are separated from the ICBM. The cost ratios in this phase also appear to favor the anti-missile system.

One possible Soviet counter to midcourse interceptions by the U.S. would be to use large numbers of penetration aids to overwhelm, or Ilsaturate the U .S. systems. Such penetration aids could be balloons, heated to simulate RVs, or possibly other lightweight devices (since light and heavy objects travel at the same speed in the vacuum of space). The problems and costs of inflating and dispensing balloon s in space, however, are signifi cant. The balloons must be equipped for delayed inflation heating, and altitude control with small jet motors.

There also could be I1natural1l penetration aids, or "space junk1I--shrouds, buses, and miscellaneous parts destined to follow similar paths until they burn up upon reentry into the atmosphere.

Reverse, or anti-simulation, in which RVs are made to look like decoys or junk is also possible.

Midcourse interception of an ICBM has long been regarded as the most. diffi cult phase for anti-missile defenses On the other hand, it is the longest phase,.giving the U.S. systems some 15 to 20 minutes to distinguish the RV from decoys and other objects, a process known as Ildiscrimination 10 Techniques for such discrimination a re progressing rapidly.

Long-wave infrared (LWIR) sensors, for example, which can detect the very faint heat of a warhead against the cold background of space (and, very likely, the distinct heat Ifsignatureif of a warhead, in contrast to those of decoys o r space junk) were used in the successful June 1984 U.S. Homing Overlay Experiment inter cept of a dummy warhead. Sensors to lldiscriminatell warheads from decoys are in development.

Maneuver in midcourse does not appear to be a cost-effective Soviet coun termeasure.' For one thing, it has the same disadvan tages and added costs as maneuver in the boost-phase. And more important, maneuvering RVs would distinguish themselves from decoys and space junk by virtue of their movement.

I1Activei1 U.S. measures fo r discriminating or destroying decoys and space junk are also possible. There is a real possi bility of If jinking, If or displacing, the trajectories of decoys and even space junk with a sweeping ground air or space-based laser. This would be too weak to disturb the heavier RVs per ceptibly, but would affect the decoys, which then would be spotted by radar and/or optical detection If the number of false targets were large, such active measures might be well worth the cost.?

If the RVs were sorted out from the decoys, they might be killed from satellite battle stations or from the ground. The cost of the kill itself would likely be low enough to justify hitting some of the decoys as well as the real targets. With the several responses available for a U.S. anti-missile system, the Soviets might even be deterred from paying the cost of deploying decoys and hardening their RVs.

Terminal If terminal defense is exoatmospheric (interception taking place outside the atmosphere it overlaps with midcourse defense in fact, it may be called "late midcourse.Il considerations for exoatmospheric terminal interception are thus very similar t o those for midcourse interception. The principal difficulty remains how to discriminate real RVs from decoys chaff, and space I1junk.lf With a layered defense, however, there would be maximum time to accomplish this during the midcourse phase The cost ra t io 5 It could prove cheaper to use nuclear weapons to clear decoys, though this is currently ruled out by the non-nuclear kill policy of the SDI Nuclear weapons launched by submarine (far enough away from U.S. borders to minimize damage to U.S. military e q uipment from electromagnetic pulse could destroy lightweight balloon decoys through a 60-mile swath, and could require the offense to harden all of its warheads to blast and EMP effects as an offensive warhead, the cost exchange ratio might well be quite f avorable to the defense If one defending warhead could destroy several decoys as well 11 Meanwhile, the U.S. is transforming a Boeing 767 jetliner into what is called an airborne optical adjunct that carries longwave infrared sensors plus computing and co m munications equipment. This would aid in discriminating decoys from genuine warheads. Initially, terminal phase interceptors surely will be of the kinetic energy type, already demonstrated in the Homing Overlay Experiment. Later, ground-based lasers may b e available.

As with space-borne, directed-energy weapons, there may be no effective limit on how often they can fire, and they thus may be able to afford wasting shots at decoys defense is the final stage of active defense (passive defense includes, among other things, making the defended target more resistant to the effects of a nuclear explosion defense generates no trouble in discriminating decoys, for light objects will burn up or slow down as they enter the upper atmos phere, and there is little like l ihood of the successful development of endoatmospheric dec0ys.Y As such, a terminal defense can be designed to deal solely with genuine warheads. Such defenses are relatively inexpensive. For one thing, their interceptor booster and kill mechanism is rela tively small. For another, the cost per intercept of the formerly expensive phased-array, solid-state electronically steered radars is declining. These factors should make the cost-exchange ratio highly favorable to the U.S. anti missile defense systems.

M aneuvering their attacking ICBMs within the atmosphere is unlikely to be an adequate Soviet countermeasure. The world's the Pershing 11, maneuvers in the terminal phase to improve accuracy. But evasion requires much greater maneuver capability than the Pe r shing I1 MaRV has. Because more capable maneuvering warheads would be heavier, fewer could be carried in each missile The result: Moscow would have to build extra costly missiles to carry the same overall number of warheads. These warheads also would be m ore expensive than non-MaRV warheads. Because they would be technologically more complicated, they would be less reliable or certain of success.

There is a special lllow-endoatmosphericll defense for "hard points,Il mainly missile silos. There are various schemes for such a system. These include firing clouds of pellets or steel flechettes and small rockets (llSwarmjetsll) at the approaching Soviet missile. Deployment by the U.S. of such systems might Endoatmospheric (interception within the atmosphere) te r minal Terminal I i first operational maneuvering reentry vehicle (MaRV), carried on I I y It is true that the first deployable U.S. ABM system, the NIKE-ZEUS, was canceled by Secretary of Defense Robert McNamara in 1961 on the announced ground that it wou l d be easy for the Soviets to design decoys that would fool the defensive radar, thus saturating the defense by forcing the defense to track and shoot at all the decoys. The Air Force, however spent billions in the 1960s and never designed a decoy that cou ld success fully fool U.S. test radars that the best "decoy" was an armed warhead.

Research in the 1960s all but concluded 12 require Moscow to expend more warheads per target than the current two. The cost estimates for point defenses are still preliminar y but one estimate for a system proposed by Sandia Laboratory suggests that, for a cost of about 3 to $5 billion, a terminal defense for the entire U.S. Minuteman I11 force could be deployed that would require the Soviets to use four warheads per target t o overcome the defense. Additional penalties could be imposed on the Soviets at modest cost by proliferating terminal defense system components.lU The Soviet cost of overcoming such a defense would likely be far greater than the U.S. cost of building one I t is possible, suggest some critics of SDI, that Moscow could counter a U.S. terminal layer defense by using "precursor warheads that would arrive on target before other warheads and destroy the terminal defense system with heat and blast or deflect the de f ensive U.S. weapons after they are launched by creating hurricane-force winds However, the U.S. could counter such an attempt by deploying in large numbers relatively inexpensive and expendable system components such as radars; further, in a layered defen se, even with only a high endoatmospheric (just after the warheads reach the atmosphere) layer, the Soviet precursors would be subject to interception at earlier points.

PREFERENTIAL DEFENSE During the boost- and post-boost phases, it is not feasible to pr oject Soviet missile trajectory and calculate the intended target. This is possible, however, in the midcourse and terminal phases. This gives the U.S. the option of preferential defense or deciding which targets to defend. By this means, the defense can p revent the attacker from achieving its objectives, fully or perhaps at all, while at the same time economizing its use of defense resources and improving its cost-exchange ratio. This tactic is particularly effective when combined, synergistically with su ch passive defense measures as using decoys, which in effect are fake missile silos.

The offense can also, of course, preferentially attack by concentrating its forces on particular targets. But it must then expend extra weapons per target to have high con fidence in destroying the targets it llprefers.Il This in turn limits the objectives it can seek to fulfill, or it would require the expen diture of considerable funds in order to overcome the defense.

MULTIPLE-ATTACK IICheapIl Soviet offensive countermea sures against U. S. anti missile systems thus look problematic and.risky at best. Their 1 For a description of the Sandia concept, see "Low Cost ABM Radar Given Emphasis uy, March 1, 1982 pp. 74-75 11 Ballistic Missile Defense, op. cit p. 395 13 probable e ffectiveness would appear too low to give the Kremlin adequate confidence in the success of its attack. The alterna tive is for Moscow to multiply the number of RVs in its arsenal and attack. The challenge will be to devise ballistic missile defenses that cost sufficiently less than offensive efforts to overcome defenses such that Moscow would be discouraged from efforts to continue its massive offensive arms,build-up--a goal that even the present state-of-the-art may achieve.'Z If a U.S. defense system th u s had a 90 percent probability with one or more layers) of intercepting an attacking Soviet weapon, Moscow would need eight to ten RVs to feel highly confi dent that the target would be destroyed. This suggests a cost to the attacker of 15 to 20 times tha t of the defender. Even a defense intercept probability of only 50 percent--surely cheaper to the defense--would require trebling the attack (around three RVs per target which would multiply the cost to the offense to at least six times that of the defense OTHER CONSIDERATIONS Such cost-ratio estimates are based on the latest and best estimates for systems already on U.S. drawing boards. If future defensive systems prove costlier, they will be deployed only as thin layersi1 to complicate the task of the att a cker, or they will not be built at all Calculations of the numbers of RVs the Soviets would need to offset the defenses are based on 'Iexpected values.Il These esti mates themselves are highly uncertain A "90 percent defense may mean "between 60 and 95 pe r cent and 50 percent may be between 25 and 70," especially given the confusion and unexpected developments of war. The uncertainty introduced by defenses thus reinforces deterrence;. an attack planner would have to make ltoffense-conservativell assumptions about the probability of success for its mission. Moscow at great cost would have to add considerably more weapons and/or countermeasures than described above--and would still be uncertain about accomplishing the mission.

CONCLUSION Cost-exchange ratios, of course, are only one factor in decision making ness to spend on defenses even it is not llcost-effective.lt'G The Soviets have shown in the past a willing For discussion, see Edgar Ulsamer The Battle for SDI Air Force Maga zine, February 1985, p. 48 T] h e Soviets have probably spent $100 billion in an air defense system for their continental territory have not made, because we do not think it is necessary [in the absence of 9 This is an expenditure that we 14 They may do so simply to increase the uncerta i nty for the United States, to reassure their own people, to satisfy bureaucratic pressures, or for other reasons. The U.S. in turn might want to do so to reduce or eliminate the ballistic missiles threat, with the possibility of eventually reducing the nu mbers and cost of strategic offensive missiles.

The purported relative inexpensiveness of MOSCOW'S developing countermeasures to U.S. missile defenses has not been proved by SDI critics, nor is it at this time probable that it can be. The clear advantage o f offense over defense, demonstrated dramatically at Hiroshima and Nagasaki, now seems to be yielding to a defense advantage. For this reason, arguments using cost-exchange ratios no longer are credible in opposing long-term research and develop ment, and intermediate deployments of a U.S. strategic defense system. Those aspects of the system that prove cost-ineffective of course, need not be deployed If arms control should solve the problems of strategic instability, the Soviet first-strike capability, an d U.S. vulner ability so much the better. If it does, it may be because the U.S. did persist, as indeed the Soviets are doing, in research and development of anti-missile weapons.

Since it can never be certain that U.S. nuclear retaliation will be carried out or be effective in the face of a Soviet first-attack, the best way to enhance deterrence--of nuclear war conventional war, or.nuclear coercion=-is by increasing the uncertainties of the.costs and outcome for the Kremlin. The best means of achieving th is would appear to be with a strategic defense.

Prepared for The Heritage Foundation by Francis P. Hoeber defense against ballistic missiles And we also think that the Soviet massive expenditure for this purpose has not been well advised, looking at the si tuation from their point of view because we still believe that with] our upcoming cruise missiles, our present bomber capability could penetrate this air defense system with relative impunity."

Carter at a Question and Answer Session, April 11, 1980, as reported in Presidential Documents, Monday, April 21, 1980, p. 660 Francis P. Hoeber is president of the Hoeber Corporation, an Arlington President Virginia-based defense consulting firm.