Whatever the aspirations for developing a nuclear strategy geared directly towards a land battle in Europe the debate remained focused on large-scale nuclear exchanges between the super-powers. This was not surprising. Super-power strategic relations had become a highly charged political issue because of SALT. The concepts of first and second strikes had taken root and had provided the frame of reference for the public debate on strategic forces since the late 1950s. The fact that the US had accepted an obligation to consider employing its nuclear arsenal on behalf of its allies if they were suffering a conventional invasion was less well understood, and often only alluded to somewhat cryptically by senior US officials (other than at NATO gatherings). Even when this issue was understood, it was normally felt that the overall state of the strategic balance or the intensity of the land battle, would determine the US response to a crisis. As Schlesinger sought to justify developing new American options it was the growth in Soviet counter-force capabilities, embodied in their large force of large ICBMs, that commanded attention. This deteriorating balance in counter-force capabilities was addressed by Kissinger (when out of office) in a celebrated speech to an orthodox, and consequently somewhat distressed, NATO audience. He warned that it was no longer possible to offer the assurances which he, with his predecessors and successors, had regularly provided, of an ‘undiminished American military commitment’.1
One stark scenario came to dominate the debate. Developed by Albert Wohlstetter, though often associated with Paul Nitze, and publicized by like-minded colleagues on the Committee on the Present Danger, it involved a surprise attack by Soviet ICBMs directed against US missile silos, effectively eliminating the ICBM force. Colin Gray claimed that the question of what to do about this growing vulnerability of fixed-site, land-based missile forces, in the 1980s, would be ‘the question of the decade’. The loss of one leg of the strategic triad would be ‘an event so momentous that its anticipation should be the occasion for a fundamental review of strategic doctrine’.2 Of course, a change in strategic doctrine was needed to make the event ‘momentous’, for under established concepts the hypothetical ‘loss’ of this leg in a surprise attack was more than compensated for by the other two legs (bombers and SLBMs) which, even if somewhat depleted themselves, would still be capable of delivering a powerful retaliatory blow. When, in the mid-1960s, Pentagon planners had first begun to consider the possible vulnerability of the Minuteman ICBM force, the most natural reaction was to strengthen the ‘survivable’ submarine-based missile force. In 1969 the Nixon Administration came to focus on the ICBM vulnerability issue because it was casting around for a problem for which the Safeguard ABM system could be a solution. Defense Secretary Melvin Laird conceded that it would only really matter if the rest of the triad became vulnerable.3
The simple arithmetic of MIRVing, combined with the large size of Soviet warheads, meant that there was a good chance that US ICBMs could become vulnerable to a counterforce threat one day even if not as quickly as claimed by the Nixon Administration. Moreover liberal arms controllers had criticised the advanced US MIRVing programme as destabilizing because of its counter-force (taken to be synonymous with first strike) implications and so they could not so easily dismiss concerns about comparable Soviet capabilities. There was therefore a shared assumption that MIRVed ICBMs held in quantity by either side could be profoundly disturbing. The eventual Soviet development of MIRVs in 1974, after a number of false alarms, meant that, for the rest of the decade, the evolving vulnerability of Minuteman was charted and its nadir projected, usually at some point in the first half of the 1980s, in a manner similar to intelligence calculations of the first years of the cold war on the imminent ‘year of maximum peril’.4
Senator Jackson, as he berated the SALT I Treaty in mid-1972 for doing nothing to prevent this threat, acknowledged the US would still have a capacity to retaliate as required by assured destruction concepts, even after a Soviet attack against US ICBM and bomber bases. He identified those Soviet forces left in reserve as presenting an even more fearful threat. The US President would have only one option: to annihilate Russian cities, ‘knowing full well, however, that the Russians could strike back at our populated centers far more effectively, killing more people, virtually wiping out our population’. This possibility might undermine deterrence because the Russians ‘might conclude that no American President would order such a move’.5 Helping Jackson develop his ideas was Richard Perle, a student of Wohlstetter’s who shared his mentor’s unwavering focus on all developing American vulnerabilities.
[I]nstead of the simple model of aggressive Soviet first strike and US retaliation, we may face a situation where the Soviets could strike first and still retain their own assured destruction retaliatory force, leaving the United States in a position of being the initiator of nuclear war against a civilian populace and the Soviet Union being in the position of retaliator.6
On current trends in the strategic balance, an American President should, prudently, be deterred from initiating strategic nuclear employment; should he proceed nonetheless, the war would very likely terminate after an almost wholly counter-military exchange (which the Soviet Union should win unequivocally) because the United States could not possibly secure an improved war outcome by initiating attacks against Soviet industry and (through re-location) population.7
The scenario involved an analysis of the military balance that was described as dynamic rather than static, by considering the consequences of successive strikes. It involved a number of steps. First, only ICBMs were really suitable for counter-force attacks on enemy ICBMs. Second, using only about one-third of its force to destroy all US ICBMs, the USSR would deprive the President of the option of responding in kind. Third, the President’s only option then would be to order attacks on softer targets, destructive of life and property, but of less military significance. Fourth, this would invite retaliation against American cities. Fifth, even after the original attack on US ICBMs the USSR would still be left with a far greater capacity for engaging in counter-value strikes, thereby making it an even more unattractive option for the US. Should large numbers of US submarines be lingering at port or bombers immobile at their base they might also be caught in the attack. If (very) worst-case assessments of Soviet civil and missile defences were included, the scenario approximated to a straightforward first strike.8
The realism of this scenario fell short on several levels. In the first place, questions were raised about the Soviets’ technical ability to carry out a strike that would destroy all US ICBMs. And even if technically possible, such an attack would necessitate killing between 2 and 20 million Americans within the first 30 days. Moreover, focusing only on casualties ignored the longer-term disruption of societies with impaired fuel supply, communications and agriculture, and, presumably, somewhat traumatized by their ‘limited’ experience of nuclear war. A Soviet attack on the US in mid-summer might cause less casualties than one in winter, because the winds would be calmer and so the distribution of fallout would be moderate; however, if the US response, at comparable numbers of megatons, made it impossible to gather the harvest in the USSR the consequences for Soviet society could be devastating.9
The idea that attacks of this sort would be experienced as limited could not be justified by comparison with any event that had gone before but only by the knowledge that it could be worse. Up to 1975, all the wars of America’s two centuries had resulted in deaths of less than 1,200,000 (including the Civil War). An attack on US ICBMs therefore would be an immoderate act that could not deserve a moderate reaction. The President, in response, might be paralysed with fear and horror, or stirred into a ferocious retaliation. That fallible and unpredictable human beings might be faced with such a choice were always one of the fundamental uncertainties upon which nuclear deterrence depended.
Secondly, the same loss of perspective on the human meaning of nuclear attacks affected the concern about a comparative Soviet advantage in counter-value capabilities after this first Soviet volley, when the US would still be in a position to destroy the USSR as a viable society.10 In a reply Nitze argued that this would only be true if such retaliation as the US could muster was directed deliberately against counter-value rather than counter-force targets ‘despite the desperate consequences to us and the world of doing so’.11
Echoing the argument used by Wohlstetter some two decades earlier, Gray suggested that, in the 1980s, the USSR ‘might possibly be able to hold down its civilian casualties to a level below that suffered in the Great Patriotic War of 1941–1945—even if the United States should proceed all the way up the escalation ladder’. As before this assumed that something considered intolerable as it happened could be a benchmark for what might be tolerated in the future.12 There was little inherent value in taking out US ICBMs, representing only 20 percent of America’s nuclear arsenal, whilst leaving untouched US bombers and submarines, accounting for the remaining 80 percent. Hitting different targets with different systems created problems of choreography in mounting an attack. To eliminate US ICBMs using Soviet ICBMs would take 30 minutes, but it would take only 15 minutes to scramble the bomber force. By contrast, Soviet SLBMs might be able to destroy US bomber bases in 15 minutes but this then left the ICBMs still available to attack the Soviet Union as it became known that there had been a massive nuclear attack on the US. Consequently, should the Soviets decide to strike all the US ICBMs to be followed by a demand to leave Europe or face the destruction of American cities, this did not account for the fact that the US still had ample nuclear weapons to destroy Soviet cities, major military targets and most of the population, even if they lacked the hard target kill capability to target any remaining Soviet ICBMs.13
There was also the possibility of the Americans launching their ICBMs on warning of an attack. There were strong indications, as we have noted, that this is what Soviet planners believed that they would do, faced with a similar attack. It became conventional wisdom in the US to caution against relying on warning systems and making retaliatory decisions in twenty minutes, preferring to ‘ride out’ attacks so that their extent and significance could be properly evaluated. However, the option was never explicitly ruled out and, as anxiety over ICBM vulnerability increased, so did the Administration’s ambiguity on this matter. ‘As an operational firing tactic, it would be monumental folly’, observed Gray, ‘but as a veiled suggestion of the “we refuse to rule it out” variety, it should not be despised’.14
But such a development could bring into question our ability to respond to attacks in a controlled, selective, and deliberate fashion. It could also give the Soviets a capability that we ourselves would lack, and it could bring into question the sense of equality that the principles of Vladivostok so explicitly endorse. Worst of all, it could arouse precisely the fears and suspicions that our arms control efforts are designed to dispel.15
The loss of command and control facilities would be the only sort of truly limited attack relevant to the first of these concerns. If only a portion of Minuteman ICBMs were destroyed then sufficient would remain for a suitably selected response; if all Minuteman ICBMs were destroyed then this would not necessarily create conditions conducive to a controlled and selected response. These conditions in fact could be varied by moving relevant bases closer to or further from centres of population. The question was whether one wanted to discourage this sort of escalation dominance by deliberately blurring (even more) the distinctions between different stages in the process of escalation, or whether one wanted to increase the enemy’s pure counter-force options on the grounds that if there were to be a nuclear war it would be preferable if it could be fought in this sort of manner!
Equally, the response could move beyond reducing vulnerability (for example by a move out to sea) to increasing counter-force capabilities to match those of the adversary. If the aim was to increase the possibilities for both sides to engage in pure counter-force attacks, then over-eagerness to protect forces could be as destabilizing. It would not be possible for both sides to have invulnerable counter-force capabilities. A counter-force race could tend to this, although both would probably end up with a ‘modern’ missile force on each side, mobile, MIRVed with excellent accuracy against hard targets, and a vulnerable fixed-site ICBM force. But if the counter-force attacks could not eliminate the most modern segment of the enemy’s force then presumably they would be only of marginal value. As already noted, given the plans for the Trident SLBM even if this was a problem there was a ‘solution’ in sight.
The anxiety exhibited by the Schlesinger quotation above was derived more from notions of essential equivalence than escalation dominance. If major asymmetry was perceived in counter-force capabilities of a sort that could be used to the Soviet advantage, then, however implausible the actual scenario when examined closely, it might influence attitudes. Repetition of concern that such perceptions might develop could become self-fulfilling. These perceptions did not develop independently but were shaped largely by the pronouncements of authoritative Americans. Most Europeans, for example, relied on US sources for their information and ideas on the state of the strategic balance.
As the 1970s progressed, the conventional wisdom reflected growing alarm over the increasing vulnerability of ICBMs. This was considered a matter of enormous significance, such that the Americans would only feel comfortable in offering, and its Western Europe and Asian allies in accepting, a nuclear guarantee if the situation were remedied at once. By the end of the decade large amounts of money were being allocated to the best available remedy, the MX ICBM, which would, at the same time, enhance both counter-force options and survive the worst the Russians could offer—once the most appropriate form of mobility had been chosen from a number of singularly unpromising schemes. The attention devoted to this matter reflected as much a crisis in American self-confidence and uncertainty over what made deterrence ‘work’ than any serious movement in the military balance. In 1980, the fact that ICBM vulnerability provided the greatest cause for concern could be taken as a symptom of the underlying stability of the strategic balance.
The Reagan Administration came to power in 1981 partly on the back of fears that America’s international standing had been diminished under Carter and that the Soviet threat had not been taken seriously enough. The issue of ICBM vulnerability was an important part of the charge sheet along with bold Soviet moves in Africa and the Middle East, and then in Afghanistan, along with the Iran hostage crisis. At the same time opponents of the Administration challenged it from the start for a reckless approach to nuclear issues. An unfortunate interview granted by a relatively junior official in the Pentagon concerned with civil defence conveyed the impression that the Administration was unusually relaxed in its contemplation of nuclear war. T. K. Jones was quoted as saying that in the event of nuclear attacks: ‘Dig a hole, cover it with a couple of doors and then throw three feet of dirt on top … It’s the dirt that does it … if there are enough shovels to go around, everybody’s going to make it’.16 Senior Pentagon figures were accused of zealous anti-communism unconstrained by any sense of risk. As the Administration backed away from a robust espousal of a radical nuclear doctrine it lost intellectual coherence. Thus it was deeply distrustful of Soviet intentions, considered past arms control agreements to be ‘fatally flawed’ and still violated by the Soviet Union, and yet it continued to negotiate with Moscow. It made preparations to fight a protracted and tactically complex nuclear war yet the President was obliged to insist that he was personally convinced that there could be no winners in a nuclear war. Most seriously, while the Administration found few problems in justifying their policies as an effort to deny a range of aggressive nuclear options to the Soviet Union they equivocated in terms of how much they would like comparable options for themselves.
The fundamental critique of mutual assured destruction was that it provided no guidance as to what to do if deterrence failed. The new Reagan strategy had to provide convincing nuclear answers to this question or fail according to its own standards. In the event it fared little better than earlier Administrations when it came to demonstrating how such a war could be brought to a tolerable conclusion.
The most serious political trouble came with the MX. This missile was required to meet two criteria which were not necessarily complementary: it was to have an ability to attack vital, protected Soviet military targets but also able to survive a comparable attack from Soviet ICBMs. The SALT II Treaty, agreed by President Carter with Soviet President Brezhnev in 1978, was not ratified in part because of arguments about ICBM vulnerability. The hawks of the Committee on the Present Danger had raised the profile of the problem of survivability. The Air Force also emphasized the survivability problem, highlighting advances in the accuracy of Soviet missiles rather than simply arguing in favour of the need to modernize the aging Minuteman system. This turned out to be an own goal when it came to the development of the MX because it meant that if a new ICBM system could not be based in the old vulnerable silos, then some new survivable basing system was needed to ensure ‘preservation of locational uncertainty’, otherwise the new system would add little value and should not be pursued.
The Air Force wanted the MX due its long-standing interest in developing an ICBM that could threaten Soviet hard targets. Essentially, their argument boiled down to the need to maintain deterrence without loopholes. If defence of a Soviet attack against Western Europe required a credible option of first use, then it was essential to have nuclear weapons that could attack Soviet military targets rather than Soviet cities. Thus, if the Soviets ever decided to escalate in a war, the US would have systems available that could limit retaliation to military targets, with the aim of preventing an all-out exchange that would destroy each other’s societies. It would therefore help to stabilize crisis situations. Put another way, it was argued that not to have the MX—to rely instead on less accurate ICBMs—might result in a situation where the Soviets would calculate the US would be unwilling to use them, and this would undermine deterrence.
The additional deterrence benefits derived from the MX quickly became a secondary consideration compared to the vulnerability concern. The Carter Administration had sought to solve this problem by means of a ‘racetrack’ that would allow individual missiles 23 possible hiding places, each of which the Soviet Union would need to target. But to hide 200 MX missiles in 4600 holes in the ground meant creating more targets—a highly unattractive option for those states where the missiles were to be based. There was also concern the Soviets would be able to locate the dummy shelters in the absence of a complex deception programme. Also, the Soviets had the option of overwhelming the system with more warheads. To have survivable basing was therefore dependent on the success of arms control to limit the number of deployable warheads, but especially after the collapse of SALT II this could not be guaranteed and at any rate created the surreal situation whereby a US hedge against a reckless Soviet move only became plausible if the Soviets cooperated.
In 1980, Congress commissioned a study by the Office of Technology Assessment on alternative basing modes for the MX. This comprehensive study took into account many earlier studies from the mid-1960s that had dealt with the problem of silo vulnerability, as well as many later proposals. Five of these proposals were given serious consideration: the Carter Administration’s multiple protective shelter (racetrack) scheme, anti-ballistic missile defence, launch under attack, basing on submarines, and basing on large aircraft. Each of these proposals contained inherent flaws. The option of adopting a launch under attack posture, despite being the cheapest of the alternatives, carried evident risks. Placing the system on a hair trigger alert increased the possibility of a catastrophic accident. A related problem was that launch under attack presupposed that the US was under a massive attack in which case there would be little point targeting empty Soviet silos; instead ICBMs were likely to be targeted at other military and industrial targets. Thus, there was the added danger that a small Soviet attack would be mistaken for a large one leading to an unnecessary massive retaliation. The very nature of the MX missile, designed to target Soviet silos, was also likely to raise Soviet anxiety about the survivability of their own forces and increase their incentive to strike first. Basing the new missile on aircraft did not offer any new guarantee of survivability as bomber bases were already vulnerable to surprise attack from Soviet submarine launched missiles. As for the alternative of basing the weapon on a submarine, this was objected to by both the Air Force and the Navy; the former as it would mean a loss of control, the latter because it would compromise the future of its preferred Trident system.
Reagan rejected the Carter administration’s racetrack scheme. His objection, and those of his advisers, was that in the absence of an arms control agreement, the Soviets could build more warheads to overcome the system. But as the Administration accepted ICBM vulnerability as a problem they had to come up with their own solution. One idea was referred to as ‘dense pack’, in which missile silos were grouped together, with survivability the result of a fratricide effect on incoming Soviet missiles. This option, supported by Defense Secretary Caspar Weinberger, had the advantage of being cheap, as there would be no need for multiple shelters, as well as requiring less land. It had the crucial disadvantage that the scientific and technical basis underlying the concept of missile fratricide was controversial. In addition, the idea of placing missiles closer together overturned the longstanding ‘common sense’ of placing them further apart.
The existence of several components of our strategic forces permits each to function as a hedge against possible Soviet successes in endangering any of the others…. Although the survivability of our ICBMs is today a matter of concern (especially when that problem is viewed in isolation) it would be far more serious if we did not have a force of ballistic missiles, submarines at sea and a bomber force.18
To some extent the Scowcroft Commission went along with the ‘Minuteman vulnerability’ scenario in identifying the process of MIRVing as the most unsettling feature of the strategic balance. It sought remedy in a move away from large launchers with many warheads to many small launchers with single warheads (soon dubbed Midgetman) and in a new approach to strategic arms control, known as ‘build-down’, which allowed the modernization of strategic forces only so long as more old warheads were removed than entered service. Congress found this idea attractive and for a while an extraordinarily complex version of build-down was incorporated into the American negotiating position at the Geneva arms control talks.19 Increasingly the MX was itself presented as a bargaining chip, a means of obtaining concessions from the Soviet Union. It had become dependent upon the arms control process and was no longer an alternative. The way to deal with what had been presented as the most pressing strategic problem of the decade was to decide that it was not really so much of a problem after all. The readiness of President Reagan to accept the recommendations of the Scowcroft Commission meant acknowledging implicitly that the attempt to escape from mutual assured destruction through the proliferation of nuclear options was effectively over. In practical terms options would accumulate as new weapons systems became available and command and control systems were refined, but the political argument was lost. If the Administration was still serious about ending the condition of mutual assured destruction then it now had to find another way of achieving this.