In the coming months, the Ford Administration must decide either to offer the Soviet Union compromises on the Vladivostok SALT Accord, permitting completion of the agreement as a permanent treaty, or to face the prospect of a prolonged period of strategic competition with the U.S.S.R., unconstrained by formal limits on strategic offensive forces. If the agreement is completed, the Congress must then decide on ratification or rejection. While this issue will occupy center stage in the strategic debate until it is resolved, the United States also faces a second major decision regarding its strategic program: whether to respond to the ongoing Soviet deployment of new, large, land-based missiles equipped with multiple independently targetable reentry vehicles (MIRVs). This Soviet deployment is not affected by the Vladivostok Accord. Thus, if it is important to respond by adjusting our strategic program, we will have to do so whether the agreement is completed or not.

In the January issue of Foreign Affairs, Paul Nitze recommends that, in addition to ongoing programs to improve accuracy, two specific major actions be taken by the United States to ensure strategic stability in the next decade: the deployment of a "multiple launch-point" land-based intercontinental ballistic missile system (more frequently called a mobile ICBM), and a greatly expanded civil defense program.1 He bases these recommendations on a series of considerations, including the nature of recent Soviet strategic force deployments, his view of Soviet motives concerning détente, and his view of the overriding importance of missile throw-weight in determining overall strategic force capability.

With regard to the Vladivostok Accord, Mr. Nitze points out that the agreement will not change the basic nature of the current Soviet force buildup. While apparently resigned to its completion, he nevertheless implies that the Accord is probably worse than no SALT agreement at all; he clearly believes it to be worse than the type of agreement he would have preferred-primarily because it fails to limit Soviet missile throw-weight-and finds "without adequate foundation" "the general belief that while strategic stability may not be assured by the SALT agreements, it is not and will not be substantially endangered." "In sum, the trends in relative military strength are such that, unless we move promptly to reverse them, the United States is moving toward a posture of minimum deterrence in which we would be conceding to the Soviet Union the potential for a military and political victory if deterrence failed."2

Clearly, the rate and scale of the present Soviet deployment of MIRVed land-based missiles are disappointing, if not outright suspicious. Furthermore, Soviet behavior in the Middle East, Africa, Southeast Asia and Southern Europe seems to most Americans far from consistent with a genuine relaxation of tensions. For their part, the Soviets argue that certain U.S. actions, such as interference in "internal" Soviet affairs, denial of normal trade relations, and continued heavy overseas U.S. force deployments are also inconsistent with such a relationship.

While this situation makes it clear that the process of establishing a better Soviet-American relationship is not an easy one, I believe that Mr. Nitze's concern that the Soviet Union might be close to obtaining strategic superiority is not justified. In my view, he has considerably overemphasized missile "throw-weight"3 as an indicator of aggregate strategic capability. Furthermore, Mr. Nitze's recommendations for deployment of land-based mobile ICBMs and large civil defense programs are, in my opinion, both ill-advised; neither should be pursued, either on its merits or in the light of other more modest measures that are available to meet the situation.

Finally, with regard to the Vladivostok Accord, while the agreement contains provisions which are clearly less constraining to the arms buildup of the two sides than many persons had hoped for, its terms are relatively favorable to the United States. If it is completed, it should provide a reasonably good basis for controlling the strategic arms competition between the two sides in the longer term.


Let us first review carefully the significance of missile throw-weight, both in general and in the specific terms stressed by Mr. Nitze, i.e., a comparison of the throw-weight on both sides that would survive after an assumed Soviet first strike against U.S. strategic forces. Since the early 1960s, when the United States abandoned the Titan ICBM program in favor of the Minuteman ICBM, the Soviets have, by and large, deployed considerably larger missiles than the United States. In the mid-to-late 1960s, the size of these missiles, particularly the SS-7, SS-8, and SS-9, permitted the Soviets to surpass the United States in the aggregate megatonnage carried by the strategic missile forces of each side. This situation generated charges that the United States was slipping into strategic inferiority. However, reasoned response to these charges eventually led to a widespread understanding that total megatonnage is simply not a very accurate measure, or even a meaningful measure, of strategic force capability. Throw-weight, while a considerably more meaningful measure of capability than megatonnage, can nonetheless be similarly misleading as an indicator of the overall U.S./U.S.S.R. strategic balance.

In evaluating the importance of a Soviet advantage in throw-weight, one must first ask: What, in a military sense, can the Soviets do with their potential throw-weight advantage? The bulk represented by more throw-weight does permit the deployment of more deliverable weapons, and, up to a point, having more weapons can complicate any effort the other side might make to deploy an anti-ballistic missile system (ABM). More weapons can also increase the flexibility and target coverage capability of one's force. However, both sides now have adequate throw-weight to deploy essentially all the weapons they could conceivably need-up to at least 10,000 for each side. Thus, the main concern over the large Soviet throw-weight is not that it will permit more weapons, but rather that it permits the deployment of larger-yield warheads on Soviet missiles. Having higher yields, it is argued, will give Soviet missiles the capability to attack U.S. missile silos effectively. Furthermore, additional throw-weight, when used to deploy larger-yield warheads, provides a hedge against uncertainties in achieving, in actual operation, high missile accuracies, the only other way in which one could develop a capability to attack missile silos-i.e., a hard-target counterforce capability.4

To refute this view, let us examine some specific combinations of size and accuracy. The present Minuteman-III missile has three warheads of a yield of 170 kilotons each. If we assume its accuracy is approximately 0.15 nautical miles,5 it would have about a 40 percent chance of destroying a missile silo, assuming two warheads were targeted against each silo.6 If the new generation of Soviet MIRVs are assumed to have a yield of about six times this amount, or 1 megaton each, yet if their accuracy is only 0.25 nautical miles rather than 0.15, their ability to destroy silos would still be no greater than that of the present Minuteman-III. If by 1985 the Soviets were able to increase this yield to 1.5 megatons and improve accuracy to 0.15 nautical miles, they could increase the probability of destroying a silo with two warheads to about 85 percent. But this is no better than the effectiveness expected from the Minuteman-III once the improved Mark 12A warhead is deployed and certain low-cost accuracy improvements are made. Yet it would take at least three times as much throw-weight per warhead for the Soviets to deploy these 1.5-megaton warheads as it would for the United States to deploy the Mark 12A.

Moreover, when accuracies get better than 0.1 nautical miles, as essentially all experts predict they will by the late 1980s, throw-weight becomes almost completely irrelevant. A 5oo-kiloton weapon with a 0.06 nautical mile CEP has a 95 percent probability of destroying a missile silo, targeting only two weapons per silo-essentially as high a probability as one could ever obtain. Thus, in this situation additional throw-weight, even enough to permit the deployment of one- or two-megaton MIRVs, is of no advantage whatsoever. Even relatively small weapons, such as the 100-kiloton warhead projected for the new Trident-I submarine-launched missile, can have quite good anti-silo counterforce capabilities at these accuracies.

In sum, there is only a very narrow range of accuracies in which throw-weight is important to a hard-target counterforce capability. Unless accuracies are better than about 0.2 nautical miles, no reasonable MIRV system can have much of a counterforce capability; once accuracies are better than 0.1 nautical miles, essentially any size missile, even those of relatively low throw-weight, can destroy silos. Since the technology is clearly in hand to permit both sides to obtain accuracies better than 0.1 nautical miles in the late 1980s, limiting throw-weight would at best put off the theoretical vulnerability of land-based missile silos for a few years.7

There are other technical problems associated with using throw-weight as the primary measure of strategic force capability. First, throw-weight is difficult to define, especially if one includes an allowance for bombers.8 Second, one must deal with "range/payload" trade-offs. For example, our new Trident-I missile, which will replace the present Poseidon missiles, will be considerably more powerful in terms of its total lifting capacity than the Poseidon missile. However, all of this additional capability will be used to extend the range of the missile, not to increase its throw-weight. The United States could have just as well increased the throw-weight, had it been thought important to do so. Nonetheless, our military planners-correctly in my opinion-chose to emphasize the survivability of our submarines, and thus to use the additional capability of the missile to extend its range; this extra range gives our submarines a much greater operating area and reduces dramatically any possibility that the Soviets could determine their location and attack them.

Even if one accepts Mr. Nitze's specific criteria for measuring the relative surviving capacity of the two sides-i.e., the throw-weight ratio of the two forces on the one hand, and the net throw-weight differential on the other-his tables would be open to challenge because they do not presume that the two sides will act in a manner which will maximize their relative situation according to the very criteria being considered.9 And in any event, by using the criteria and scenarios he has chosen, he obscures the fact that the United States retains a very large force even after any conceivable full-scale Soviet attack.10 In particular, even if the Soviets destroyed all our land-based ICBMs, non-alert bombers and non-alert submarine-launched ballistic missiles (SLBMs), we would still retain our alert bomber force and our alert SLBM force. These forces alone would be capable of delivering something like 4-5,000 nuclear weapons against Soviet targets. I find it difficult to conceive of a military need for more than 4-5,000 nuclear weapons, albeit weapons much smaller than the relatively large Soviet weapons.


Apart from the potential military importance of a Soviet throw-weight advantage, there is the question of Soviet motives behind their throw-weight buildup. While it is possible that, as Mr. Nitze implies, the Soviets have emphasized large land-based missiles in their strategic force deployments in order to gain a measure of strategic superiority over the United States, one need not hypothesize such Soviet motives in order to explain their weapon systems choices. For the last ten years, the Soviets have clearly felt a great need to deploy a new generation of missiles equipped with MIRVs. Yet they lacked the technology to deploy MIRVs efficiently on relatively small land-based missiles or on submarine-launched ballistic missiles. Furthermore, the Soviets have always had great operational difficulties with their SLBM force. This has led them to put much heavier emphasis on land-based missiles relative to sea-based missiles, while the United States has done just the opposite.

The Soviets needed MIRVs for two reasons: first, to increase the target coverage and flexibility of their force; and second, to ensure that they could penetrate potential U.S. anti-ballistic missile systems. Since the 1972 signing of the SALT I treaty effectively prohibiting ABMs, we have tended to lose sight of the environment in which the generation of Soviet missiles now being deployed was designed. At that time, in the mid-1960s, not too long after the Cuban missile crisis, the United States was very far ahead in ABM technology, and appeared to be moving toward a large ABM deployment. The Soviets were probably aware of MIRVs at the time they first designed these missiles, and recognized that the most effective way to overcome an ABM defense is to launch a greater number of weapons against the defense than there are interceptors available to the defense, simply exhausting the defense. However, such a tactic is economically feasible only if one can use MIRVs. Thus, this alone would have been good reason for the Soviets to make a major effort to deploy MIRVs on their new missiles.11

The Soviets also needed MIRVs to increase the flexibility and target coverage of their force. With only about 2,000 total missiles in their force, considering reliability and the number of missiles that might be destroyed in a strategic exchange, the Soviets could probably count on no more than about 1,000 nuclear warheads available to deliver against U.S. targets. While 1,000 weapons are certainly more than adequate to destroy virtually all the cities of the United States, in more complicated scenarios it might be possible for the United States to use some of its MIRVs to attack Soviet military targets, keeping population damage relatively low, while retaining a clear capability to destroy all the Soviet cities. Without MIRVs, the Soviets would not be in a similar position. Thus, this need for more weapons undoubtedly added to the attraction of MIRVs.

Once the Soviets decided to deploy MIRVs, it is not surprising that they chose to emphasize size and develop the relatively large missiles we now see. Whereas the United States made great use in its MIRV programs of its already-existing capability to develop small on-board computers and small, but relatively high-yield, nuclear warheads, in the mid-1960s the Soviets simply did not have such technologies. Of course, these technologies are now becoming available to them, but they were not when the systems now being deployed were designed.

Additionally, the Soviets' decision to emphasize land-based missiles undoubtedly led them to conclude that they should increase the hardness (i.e., the resistance to nuclear attack) of their land-based missile silos. In order to do this, they had to make major structural changes in these silos. Relatively small increases in the depth or diameter of the silos could be easily accomplished in conjunction with a silo-hardening program they probably planned to undertake in any event. Since the new Soviet missiles are not much larger in volume than the ones they replaced, these small increases in silo dimensions were adequate to permit deployment of the new missiles.12

In conclusion-given the Soviets' difficulty with SLBM technology, their lack of sophisticated miniaturization technology, and their strong military need for MIRVs, both in order to penetrate ABMs and to increase the target coverage flexibility of their force-it is not at all surprising that the current generation of Soviet land-based missiles have large throw-weight and MIRVs. Had the roles of the two countries been switched, with the United States in the position in which the Soviets found themselves, we probably would have made similar decisions, with no necessary intention of acquiring a position of strategic superiority.

This point, of course, applies principally to the Soviet choice of technologies. Similar considerations relate to the rate and scale of the Soviet deployment of big missiles. If one puts oneself in Soviet shoes, one can see that they have faced, and still face, such factors as a major inferiority in the number of deployed warheads (roughly 3,500 versus 9,000 at present), an inability to keep their SLBM force at sea for long periods, and what may be significant problems of reliability. The point, in both areas, is not to argue for complacency about the Soviet buildup; it is simply to note that the technological and numerical factors are not so one-sided as to indicate-to anything like the degree Mr. Nitze contends-an underlying motive of seeking strategic superiority. In my judgment it is simply impossible to tell from Soviet actions alone whether their "true" motives have been other than technological.


Turning now to the Vladivostok Accord, I believe it is first important to review precisely what was settled there and the negotiating history that led up to the agreement reached in November 1974.

The Accord calls for an equal numerical limit on two important quantities: the number of strategic delivery vehicles (missiles and strategic bombers), which are limited to 2,400, and the number of missiles equipped with MIRVs, limited to 1,320, with both numbers effective through 1985. As has been pointed out, 2,400 is only slightly below the Soviets' present force levels (about 2,500), and is above what the United States plans to deploy anyway (unless "mothballed" strategic bombers are counted in the total). Furthermore, 1,320 is dramatically above the present Soviet MIRV force level (zero at the time the Accord was reached and only 50-100 today), and also above the number of missiles with MIRVs the United States plans to deploy, even assuming all existing Poseidon submarines are retained in the force as the new Trident submarines come on line. Thus, it is clear that neither side gave up much, if any, of its present strategic programs.

We can only speculate about what future Soviet force levels would be without the agreement. But there is some indication that they had planned to expand further with new submarines, bombers and mobile missiles. Whether or not they would have gone above 1,320 in the number of missiles with MIRVs is also speculative, but the best guess is that they intended to deploy MIRVs on almost all their missiles-a total closer to 2,400 than to 1,320. Thus, the agreement probably would have a restraining effect on Soviet strategic planning.

More importantly, over and above the limited effect on the programs of the two sides, the Accord settled a number of issues of basic principle. Perhaps foremost was the issue of equal numerical limits. The Interim Agreement on strategic offensive arms now in force froze the missile levels deployed and under construction by both sides at the time the Agreement went into force (October 3, 1972). Since the Soviets had more missiles at that time than did the United States, this left them with a numerical advantage in missiles (about 2,350 to our 1,710). In support of the Agreement, the Nixon Administration argued that this numerical advantage was offset by several U.S. advantages: our large bomber force (500 B-52s and 72 FB-111s versus about 150 older Soviet bombers); our tremendous advantage in MIRV technology; and other qualitative advantages in our forces. Nevertheless, many in the Congress insisted that in all future agreements there should be numerical equality. As a result, this principle was somewhat ambiguously embodied in the "Jackson Amendment" (to the Joint Resolution approving the Interim Agreement), which provided that "the President seek a future treaty that, inter alia, would not limit the United States to levels of intercontinental strategic forces inferior to the limits provided for the Soviet Union."

As SALT negotiations resumed after the Interim Agreement went into effect, within the U.S. government both the Joint Chiefs of Staff and the Office of the Secretary of Defense argued strongly that any follow-on SALT agreement should provide for so-called "equal aggregates"-a term referring to the aggregate sum of strategic land-based missiles (ICBMs), submarine-launched ballistic missiles (SLBMs), and strategic bombers such as our B-52s. The Pentagon's argument was based not so much on the military importance of this measure, which was acknowledged to be minimal, as on its political significance. It was argued that to accept anything less than numerical equality would lead to "political perceptions" of U.S. strategic inferiority.

The Soviets, on the other hand, continued to argue that "equal aggregates" were not reasonable so long as the U.S. maintained large "forward-based systems" (FBS), i.e., nuclear-armed aircraft on carriers and at foreign bases capable of striking the U.S.S.R. They also insisted that our British and French allies' nuclear forces be taken into account. This argument was particularly telling with respect to the 64 British SLBMs, since these were of American manufacture, carried on submarines built with American assistance, and essentially indistinguishable from the 160 U.S. Polaris missiles which we had already agreed to count. The argument concerning French forces and our own FBS had less intellectual validity, but nonetheless was not completely without merit.

During the period leading up to Vladivostok, the United States strongly pushed one other point-the throw-weight of Soviet missiles-stressing above all the combined effect of throw-weight and MIRVs. The principal argument was the one spelled out above, that the combination of a larger number of weapons and the relatively high yields possible with large missiles (up to two megatons for each warhead) would pose a threat to the survival of our land-based missile silos more quickly than had been anticipated. Thus, the United States insisted that if the Soviets put MIRVs on their large missiles, they agree to deploy a relatively smaller number of these missiles to compensate for their large size. If the number were low enough, the Soviets would simply not have enough weapons to attack Minuteman effectively, regardless of the accuracy of the weapons, and still retain an adequate "assured destruction" reserve. Only then, it was argued, could we avoid the instabilities inherent in retaining a significant component of our force potentially vulnerable to a surprise attack, possibly tempting such an attack in a crisis.

In reaching agreement at Vladivostok, the Soviets yielded on "compensation" for forward-based systems and allied forces, and the United States yielded on controlling the size of Soviet missiles. Many on the U.S. side had argued that both objectives-equal numbers and limits on missile size-were of key importance. Nevertheless, between the two, there was tremendous pressure to make equal numerical limits the sine qua non. This position was somewhat ironic, since within the U.S. government almost everyone acknowledged that equal numbers had little military or strategic effect, while controls on the size of Soviet missiles (especially those equipped with MIRVs) could have some potential military and strategic effect-although even this effect was acknowledged to be limited since, as already noted, improved technology would eventually permit higher accuracies, which would eliminate the benefit of big missiles and their large-yield warheads.

One can raise a legitimate question whether the United States should have agreed to this compromise arrangement. Clearly, had the Soviets been willing to agree to very low limits on the number of land-based missiles they would deploy with MIRVs (e.g., 400-500 such missiles), or to cut back the size of their MIRVed missiles, we could have put off the day when we have to be seriously concerned about the survivability of our Minuteman ICBMs. Nevertheless, it is difficult to see how any other compromise could have been made. As discussed in detail above, given their present technology the Soviets could not quickly deploy MIRVed missiles with the type of small warheads we have on our missiles. Thus, were they to accept significant constraints on the size of their missiles, they would remain dramatically behind the United States in the more important strategic measure of numbers of available warheads. It seems unlikely that they would have accepted this de facto superiority unless the United States made major concessions on FBS and allied nuclear forces-something we were simply not prepared to do.

It is worth reemphasizing that on its face the Vladivostok agreement is relatively favorable to the United States. It does not place great constraints on either side; however, it does require the Soviets to reduce their force levels in terms of aggregate strategic delivery vehicles somewhat, even with the present 2,400 limit. Based on what is known of the results of Secretary Kissinger's January 1976 trip to Moscow, it appears that it might even be possible to obtain a somewhat lower limit-perhaps as low as 2,200. As for the United States, the agreement would not prohibit any program which we now have under way, nor any program we are likely to undertake by 1985. We could increase the throw-weight of our missiles to a level essentially equal to that of the Soviet Union, should we decide it to be important to do so, and, as the agreement's terms now stand, we could deploy mobile ICBMs if we wished. (I shall return later to this point.) We can certainly continue with all our qualitative improvements, including improvements to our bomber force and SLBM force. While the final agreement on cruise missiles may impose some limits on our longer-range sea-based cruise missile force, we have limited strategic need for sea-based cruise missiles (although short-range cruise missiles, which would not be limited, are an important component of our conventional naval forces). What we do have a strong need for are medium-range air-launched cruise missiles to ensure the penetrability of our bomber force, and it seems unlikely that the final Vladivostok limit will compromise our ability to deploy a force of the required size. Thus, as was the case with SALT I, the agreement probably imposes some real, although modest, limits on the Soviet strategic program, but imposes no operational limits on the U.S. program when compared with what we would have done without the agreement.

The agreement also has the major advantage of eliminating "worst case" scenarios from the force planning of both sides. Before the Vladivostok Accord was reached, the intelligence community had projected a possible maximum Soviet deployment of 3,300 strategic delivery vehicles, of which 3,000 would be MIRVed missiles, by 1985. If the agreement is turned into a treaty, there will be no need to plan our actual deployment programs against the possibility of such large Soviet deployments, although we should continue to hedge against Soviet treaty abrogation by maintaining an active research and development program.

Some will argue that the Vladivostok Accord should not be completed because it will permit the Soviets to deploy their Backfire aircraft outside the limits of the agreement. I find this argument somewhat incredible. On the one hand, the same critics frequently argue that our massive strategic bomber force, which costs almost half of our total strategic budget, is not worth much compared to Soviet missiles. On the other hand, the Soviets' Backfire, an aircraft vastly less capable than our B-52s and even more dramatically less capable than our B-1, is alleged to represent a major threat, despite its short range, low payload, low alert rate, and second-strike character.

The Soviets appear willing to make on-the-record assurances that the Backfire is not capable of being used against the United States, will not be used against the United States, and will not be deployed with a companion tanker force which would permit midair refueling. Without such refueling, the aircraft would be limited to one-way, high-altitude, subsonic missions against the United States. Such assurances are consistent with a reasonable interpretation of the aircraft's mission. If the Soviets had intended to develop a new strategic aircraft for use against the United States, they certainly wasted a lot of money building the Backfire. They undoubtedly had the technology to develop at little additional cost an aircraft which would be vastly more capable against the United States. Furthermore, they clearly had a need for an aircraft like the Backfire to replace their aging Badger and Blinder medium-bomber force deployed for use against targets in Europe and China.

Thus, it makes little sense to attribute to the Soviets the sinister motive of attempting to use the Backfire to get around the Vladivostok limits. Arguments over the Backfire should not hold up a new SALT agreement.


Based on the analysis presented earlier, I am considerably less concerned with the Soviet throw-weight advantage than is Paul Nitze. Nevertheless, no one can deny that Soviet capabilities are increasing and that Minuteman is becoming vulnerable, albeit more because of improvements in accuracy than because of increases in throw-weight. Thus, the United States must decide whether to react to the Soviet buildup by adjusting its own strategic program; and Mr. Nitze's suggestions for how we should respond deserve consideration even if one disagrees with his reasons for recommending a response.

Mr. Nitze's primary recommendation is that the United States employ a shelter-based land-mobile ICBM system to offset the Soviet throw-weight advantage. He argues that such a system would "absorb" the Soviet throw-weight advantage; if the Soviets were to attempt a first strike against our strategic forces, they would be required to devote a large proportion of their own force to attacking the mobile missiles, reducing their surviving throw-weight.

Again, this argument is open to question on Mr. Nitze's own assumptions concerning Soviet attack strategy. He apparently does not assume that the Soviets would carry out futile attacks against U.S. submarines at sea; why, then, should one assume that they would "waste" their throw-weight in an equally futile attempt to destroy dispersed land-mobile missiles? Yet, if they should not do so, the net Soviet advantage after the assumed attack-Mr. Nitze's main criterion-would not be greatly affected.

On the other hand, the deployment of land-mobile missiles would substantially increase the absolute amount of surviving U.S. retaliatory capability. Thus, for reasons somewhat different from Mr. Nitze's, I agree that a mobile land-based system, by reducing the vulnerability of U.S. strategic forces to a counterforce attack, could help to prevent the Soviets from ever attaining strategic superiority.

But even on this basis Mr. Nitze's proposed program would be extravagant and untested, going beyond the degree of threat that now appears likely, and far more costly than other measures, at least as feasible, to improve survivability. Furthermore, it might well set off a new competition in which the Soviets would have all the advantages. Let me address these points in order.

First, there remain serious cost and technical problems associated with such systems. They require either very hard movable capsules to protect the missiles, or very cheap hard shelters, neither of which has yet been proved technically feasible. With regard to cost, it is very unlikely that any feasible technical concept could cost as little as Mr. Nitze's prediction of $12-15 billion over a ten-year period. A much more reasonable estimate, when all costs are considered, would be about $30 billion. Since initial cost estimates of new weapons systems inevitably turn out to be somewhat lower than final costs, this final ten-year cost could easily run as high as $50 billion.

Second, we should not forget that potential Minuteman vulnerabilities are strictly theoretical; these projected vulnerabilities are based on force exchange calculations which are probably extremely conservative. Mounting an operational attack against Minuteman would not be a simple task. No one has ever tested weapons against land-based ICBM silos in an operational environment, so the Soviets would face considerable uncertainties in mounting such an attack. Furthermore, there are many technical problems associated with attacking silos: two examples are the problem of "fratricide" mentioned earlier, which necessitates very precise attack timing, and the possibility that the side being attacked will launch its missiles as soon as it sees the incoming attack.

Third, the United States can make potential force improvements which might significantly reduce Minuteman vulnerability, especially through the use of defense concepts oriented exclusively toward defending silos. While most of these concepts are far from proven technically, the Soviets cannot presume we will be unable to protect the Minuteman force if they are seriously considering striving for a capability to destroy it.13 While I share some of the concerns raised in the Congress about the present direction of our missile defense research and development programs, clearly we should not short-change this crucial area.14

But in any event the issue is not confined to land-based missiles. The United States has other ways of conserving or increasing its survivable second-strike capability. We maintain large and effective submarine-launched ballistic missile and bomber forces, the alert portions of which are essentially invulnerable to potential Soviet attack. We have active programs to modernize and improve both our SLBM and strategic bomber forces. Perhaps the most important of these programs is the new Trident-I missile. This missile is designed for deployment on both our new Trident submarine and our older Poseidon submarines. It will carry up to eight MIRVs and, with over twice the range of the present Poseidon MIRVed missile, will increase the operating area of our SLBM force by a factor of 10, tremendously reducing its vulnerability.

We also have an active program to ensure the effectiveness of our strategic bomber force. This includes the planned deployment of the B-1 bomber, which I believe should go forward in small numbers, although I see little reason to deploy all 240 B-1s now planned. A better use of our funds would be to cut back the B-1 program somewhat and use the funds freed to modernize our older B-52s, equipping them with a moderate number of medium-range strategic cruise missiles to increase their capability of penetrating advanced Soviet air defenses.15

Indeed, given these alternatives to the deployment of mobile missiles, and the small chance that a feasible U.S. program can be developed from a technical and financial standpoint, serious consideration should be given to banning such systems in a SALT II treaty. It seems clear from the negotiating record to date that the Soviets would accept an outright ban on all mobile ICBMs as part of a final SALT II agreement. The United States apparently has yet to accept such an approach, primarily because of Pentagon pressure to keep the option for a land-based mobile missile open. While the United States has not yet come up with a feasible land-mobile ICBM program, it is a serious technical option for the Soviets. The Soviets have large uninhabited land areas and no political constraints that would confine the deployment of nuclear weapons to military reservations, as is the case in the United States. In fact, they have already deployed significant numbers of intermediate-range mobile missiles for use against China and Europe, and their newest solid-propellant ICBM, the SS-16, could probably be deployed in a mobile mode almost immediately.

Banning land-based mobiles would close off one additional area for potential arms competition between the two countries, as well as eliminating the potential SALT verification uncertainties presented by mobile missiles. While (like Mr. Nitze) I believe these verification uncertainties are not as crucial as some have made them out to be, they could be serious in some circumstances. In particular, if the Soviet Union chose to deploy a large shelter-based land-mobile system similar to the system Mr. Nitze proposes for the United States, our planners would have to assume that each shelter contains a missile. Since there could be thousands of shelters, such a situation could easily force a complete breakdown in the SALT process and lead to a new level of arms competition.

But a shelter-based mobile ICBM deployment is not one which the Soviets are likely to undertake in the near future, so I see no reason why the Vladivostok agreement should be abandoned if land-mobile ICBMs are permitted within its ceilings. Nevertheless the United States should actively seek to obtain mutual agreement that they be banned.

Mr. Nitze's second major program suggestion is that the United States undertake a large civil defense program. He bases this primarily on his belief that the Soviets have a large and active civil defense program, designed to give them the possibility of absorbing an American retaliatory attack with an acceptable degree of Soviet fatalities. Mr. Nitze believes we should develop the same capabilities.

There is some question about the actual size and extent of the present Soviet program. However, their present program is not likely to be able to significantly reduce overall fatalities in a U.S. retaliatory attack. Only a small percentage of the Soviet population has been adequately trained in evacuation procedures; fallout shelters are in many cases inadequately stocked and of inadequate size; and many would die from secondary nuclear effects even if they reached the shelters before the attack. Thus, while the U.S.S.R. spends more on civil defense than does the United States, it is not clear that the Soviet Union is anywhere near being able to eliminate our deterrent capabilities through such a program.

An American program would face similar technical difficulties. It would be almost impossible to disperse most of the American population now residing in large metropolitan areas to shelters which could not be attacked, given the very large number of Soviet weapons which will be available by the mid-1980s. Furthermore, protecting these shelters against radiation from fallout, adequately stocking them, and adequately training the population, even if all these were politically possible, represent an almost impossible task from a practical standpoint.

If we continue with the strategic programs I have discussed above, it seems highly unlikely that the Soviets could mount an effective threat to our deterrent force. The Soviets can build larger missiles, deploy more throw-weight, increase their megatonnage, and perhaps even come close to deploying as many weapons as the United States has deployed. However, our submarines and alert bombers should remain invulnerable to Soviet counterforce attacks, and we can seriously complicate any potential Soviet effort to develop a counterforce capability against our land-based missiles. Thus, any advantage they might have in throw-weight or other static capability measures will have essentially no military benefits; its only possible benefit might be to create adverse political perceptions in the rest of the world concerning the relative strengths of the United States and the Soviet Union.

But we must not forget that our own rhetoric largely determines these political perceptions. To the extent that we emphasize measures in which the Soviets have an advantage, such as missile throw-weight, we ensure that others will perceive us to be at a disadvantage. On the other hand, if we pursue sensible programs, designed to protect our deterrent capability, and explain carefully why we have decided to forego a "throw-weight race" or a "megatonnage race," I see no reason why we should be the subject of adverse political perceptions.


Two other suggestions have been made for a U.S. response to the Soviet buildup: further improvements in the accuracy of American missiles, and the development of new, larger throw-weight, silo-based missiles to replace Minuteman. Both of these programs are being pursued at an active, though not terribly vigorous, pace by the Pentagon. Critics charge that they are destabilizing to the strategic balance, in that they would increase the American threat to the Soviet retaliatory force, giving the Soviets more incentive to strike first.

My judgment differs. While we have no particular need for greater overall throw-weight, I see little reason for us to impose upon ourselves a unilateral restraint on the accuracy of our missiles. First, no one has yet developed a workable scheme that would permit controlling accuracy through verifiable arms control agreements. Thus, it seems inevitable that the Soviets will continue to improve the accuracy of their force. Second, since the Soviet Union, like the United States, maintains large submarine-based forces which would survive an attack regardless of how accurate our ICBMs become, the decrease in "crisis stability" brought about by U.S. improvements in accuracy is likely to be quite small. Third, as many (including Mr. Nitze) have pointed out, the Soviets have a large investment in their land-based ICBM force, and U.S. development of accurate missiles may well motivate them to reduce their emphasis on large land-based ICBMs-moving more of their force to sea, or perhaps even agreeing to reduce the size of the force through arms control agreements. In the absence of American improvements in accuracy, there seems to be little incentive for the Soviets to control the size of their land-based missile force.

There is, moreover, something to the argument that the United States should not permit the Soviets to obtain a significant military capability which we do not have. This would be the case if they developed very high accuracy, permitting "surgical" strikes against single military targets, while we imposed a unilateral restraint on our own accuracy program. It is extremely difficult to conceive of a situation in which the Soviets might use such a capability, but there would be even less reason for the Soviets to attempt such a strike if they knew the United States had a comparable capability and could respond in kind.

Finally, perhaps the most compelling argument against a unilateral American effort to limit the accuracy of its missiles is a technical one. As mentioned earlier, extremely high accuracies are not likely to be obtained using the missile guidance technology now being pursued, i.e., purely inertial guidance systems. However, very high accuracies are likely not only to be feasible, but to be relatively easy to achieve once sophisticated satellite navigation systems are perfected. If these systems permit the determination of location as precisely as they promise, it will be possible to obtain accuracies of a few hundred feet. There are feasible and not terribly expensive ways of ensuring that guidance systems based on these satellite navigation systems are both highly survivable and essentially impossible to jam electronically. Since these systems are being developed for a host of navigation purposes unrelated to warhead accuracy, they will go forward in any event. Thus, as much as we may believe we can restrain our own programs for improving accuracy, the basic technological developments which will eventually lead to high accuracy are sure to come.


To conclude, the United States has reasonable strategic force alternatives which can ensure our retaliatory capability, deny the Soviets any potential military advantage, and yet which do not require a return to a massive civil defense program or deployment of a costly and perhaps technically infeasible land-based mobile ICBM system. We need not match the Soviets in missile throw-weight, nor, within quite broad limits, need we even worry about their advantage. At the same time, our government must adopt complementary rhetoric, making it clear why we have chosen not to place great emphasis on throw-weight, lest the growing Soviet advantage in this one respect create an adverse perception of the strategic balance throughout the world.

With respect to the Vladivostok SALT Accord, I believe that it should be completed, even if this means compromising our positions somewhat on the Backfire and cruise missile issues. The Vladivostok Accord may well have been disappointing relative to what many of us had expected as a follow-on SALT agreement: it does not limit the threat to the survivability of the land-based missile forces of either side, nor does it provide for significant reductions in the size of existing strategic forces. Rather, it will undoubtedly permit a significant increase in the overall destructive capability of the forces of the two sides. These facts make it easy to criticize the Vladivostok Accord, since one can always compare it to the kind of agreement most of us would have preferred. Furthermore, it is easy to place the blame on the Soviets, ascribing to them a strong effort to obtain strategic superiority.

What is much more difficult is to work out a viable alternative to the Vladivostok Accord. Most of the specific suggestions critics have made for alternative agreements were tried vigorously in one form or another during the three-year period leading up to the Accord. Low limits on Soviet MIRVs were unacceptable to the Soviet Union for the obvious reason that this would have left the United States with a militarily significant superiority in this area. Limits on throw-weight were unacceptable because such limits would have effectively prohibited significant MIRV deployments, given existing Soviet technology. Significant reductions were not of interest to either side, perhaps more than any other reason because neither the Soviet leadership nor the U.S. Administration was willing to undertake the internal bureaucratic struggles necessary to decide which component would have to reduce. Finally, the Soviets even today remain quite concerned about U.S. forward-based systems, as we remain concerned about the size of their missiles. Thus, the Vladivostok Accord itself represented a fairly major breakthrough from what appeared to be a completely deadlocked situation.

Given this background, I believe the only serious alternative to completing the Vladivostok Accord is to abandon the SALT process for the time being, at least with respect to offensive forces, hoping that the two sides will trust each other enough to keep their force deployments under control by mutual example, without the benefit of formal agreements.

Such a hope is likely to be disappointed. With no agreements, the probability is that the Soviets will continue to increase their strategic spending, leading to ever-widening disparities in force size. While the thrust of this article has been that many of these disparities are much less meaningful than they are made out to be, at some point the disparities could become large enough to have political, and perhaps even military, consequences. We would then have to increase our force size to keep up, probably beyond the ceilings contemplated at Vladivostok, certainly beyond our present programs.

Some might argue that we need not match the Soviet buildup even roughly. But unless we did so we could hardly expect the Soviets, in renewed negotiations, to accept even Vladivostok-level limits, since they would then be at a far higher level and probably at a comparative advantage at least in numbers.

In short, abandonment of the SALT process, even for a time, would mean unnecessary extension of an arms competition that has gone on too long. Such an extension could only increase Soviet-American tensions and, in my judgment, decrease rather than improve our security.


3 Throw-weight is, roughly, a measure of the total weight of missile warheads which one side could deliver against the other side. Throw-weight also includes the weight of missile guidance systems, MIRV-dispensing systems, and certain other equipment which is carried out of the atmosphere by a strategic missile. For strategic bombers, the term "throw-weight" is technically inappropriate; the correct analogous term is "payload," and when "throw-weight" is used in reference to bombers, it is usually taken to mean payload. See footnote 8 for comment on Mr. Nitze's slightly different formula for bombers.

4 Some also argue that a throw-weight advantage could be militarily significant for attacks against large area targets, such as bomber escape areas and submarine patrol areas. However, "barrage" attacks against bombers and submarines are probably not feasible, even with very large-yield weapons. The Soviets would have to locate our submarines, and there is no indication that they will have such an ability by 1985. Our bombers could probably survive barrage attacks, even today. In any event, attacks against our bomber bases would cause such widespread civilian fatalities that the Soviets would have to assume an all-out U.S. response.

5 This customary measure of accuracy is called the CEP, or circular error probable. In lay language, a warhead is estimated to have a 50 percent chance of landing within this distance from the designated target.

6 For complicated technical reasons related to a phenomenon referred to as "fratricide," force planners usually assume that no more than two warheads can be targeted effectively against any one silo.

7 Personally, I do not believe that accuracies better than 0.1 nautical miles are likely to be obtained by the purely inertial guidance systems now used by both sides. However, using satellite navigational systems, or perhaps terminal homing systems, it seems quite feasible to develop these accuracies over the next ten years, or perhaps even sooner.

8 For example, Paul Nitze's calculations (pp. 224-25) assigned our B-52 bombers an equivalent throw-weight of 10,000 pounds and our B-1 bombers 19,000 pounds, although these bombers actually carry significantly more payload than this. Furthermore, he degraded the payload of the U.S. bomber force both for its "alert rate" and for "penetration factors."

While I would agree that some adjustment of gross bomber payload might be appropriate, Mr. Nitze's assumptions go too far in favoring the Soviets. If one were to use gross U.S. bomber payload, the Soviet projected throw-weight advantage in 1984 would disappear. While this is going to the other extreme, it illustrates that by picking a different yet still not absurd definition of aggregate throw-weight, one can obtain radically different views of the relative balance.

9 The tables (pp. 224-25) assume "an exchange in which the Soviet Union has attacked U.S. forces, and the United States has retaliated by trying to reduce Soviet strategic throw-weight to the greatest extent possible." But if what matters is net residual throw-weight, the United States should not try "to reduce Soviet strategic throw-weight to the greatest extent possible," but only to the extent that a pound of U.S. throw-weight knocks out more than a pound of Soviet throw-weight. A similar situation is true with regard to the throw-weight ratio criterion. For example, while it is not possible to tell exactly how Mr. Nitze's calculations were done since the assumptions are not presented, it appears that most U.S. SLBMs were assumed to be used to attack Soviet silos in a retaliatory attack. If the United States simply held in reserve most of these weapons, as well as alert bombers, the "after-attack" residual throw-weight advantage of the Soviets might be considerably less.

10 The potential importance of residual force levels is mentioned in general terms by Mr. Nitze on p. 226, but without spelling out their significance.

11 Indeed, there is an additional difficulty here with Mr. Nitze's argument about the objectives and strategy behind U.S. nuclear programs in the 1960s. While our actual deployments were indeed wholly consistent with the stated aim of a situation of assured second-strike capabilities on both sides, our research programs, notably on ABM and MIRV, were almost bound to convey a different future signal to Soviet planners, i.e., that we might be seriously contemplating a return to counterforce strategies and our own version of strategic superiority.

12 This additional volume of the missiles accounts for some of the large increase in throw-weight we see with the new missiles; however, most of the throw-weight increase is a result of significant improvements made in Soviet missile propellants and other missile technology.

13 For an example of two of the most intriguing possibilities for reducing Minuteman's vulnerability, see Richard L. Garwin, "How Real is the Soviet Threat to Minuteman," presented to the Senate Committee on Foreign Relations in answer to staff questions addressed to Mr. Garwin's testimony of September 18, 1975. Mr. Garwin suggests two possibilities: a "bed of nails" defense in which vertical steel rods and radar fuse jamming would disable, respectively, both the ground burst and radar fusing on incoming warheads, and a "pellet" defense in which incoming reentry vehicles would be destroyed while passing through a curtain of steel pellets thrown up by simple explosives and set off upon detection of the reentry vehicle by simple radars.

14 One counterargument often made is that the Anti-Ballistic Missile Treaty prohibits deployment of such local defense systems, so why should we waste our money on research and development? But, if an effective technology emerges for a defense exclusively devoted to ICBMs, and if arms control measures cannot be found which will improve the stability of the strategic balance in the longer run, we should seriously consider insisting upon modifications to the ABM treaty which would permit deployment of ICBM defenses which would be capable of defending only hardened military sites, and not capable of defending soft area targets such as cities. In pursuing ICBM defense technology, we should actively investigate new concepts, such as the inexpensive but promising technologies mentioned above. One of the major difficulties with our ABM research and development has been that it remains closely tied to the essentially abandoned Safeguard program, although our ABM objectives have changed completely from those which motivated the design of Safeguard.

15 Some critics charge that for the United States to develop strategic cruise missiles will simply fuel the arms race and further destabilize the balance between the United States and the U.S.S.R. In my view, this criticism is completely unfounded. These weapons will reduce the total destructive capability of our bombers, since the bomber must carry the weight of the missile in addition to the weight of the bomb carried on the missile. With free-fall gravity bombs, the bomber need not carry the weight of the missile, and can therefore carry a more destructive payload to any given range. Furthermore, cruise missiles should represent a stabilizing influence, since they increase our second-strike retaliatory capability, but add essentially nothing to our first-strike counterforce capability. See Alton H. Quanbeck and Archie L. Wood, Modernizing the Strategic Bomber Force-Why and How, Washington, D.C.: The Brookings Institution, 1976.

You are reading a free article.

Subscribe to Foreign Affairs to get unlimited access.

  • Paywall-free reading of new articles and a century of archives
  • Unlock access to iOS/Android apps to save editions for offline reading
  • Six issues a year in print, online, and audio editions
Subscribe Now