Naval Historical Center home page Image of an anchorColloquium on Contemporary History Project
Flag banner

David Alan Rosenberg

The history of American attempts to establish constraints on nuclear weapons can be divided into two periods. Between 1945 and 1960, Western reliance on nuclear weapons as the foundation of defense against Soviet conventional and nuclear forces grew largely unconstrained. From 1961 on, American and NATO leaders moved to impose a variety of policy, planning, procurement, and physical checks on existing and planned nuclear forces.

A number of dynamics fed the growth of the nuclear genie through 1960. The two greatest external factors were the tremendous imbalance between Soviet and Western conventional forces in Europe, and, after 1949, the threat posed by Soviet nuclear weapons. As a counter, the United States relied on rapid expansion of its nuclear weapons capability and planned nuclear offensives. Advances in technology made possible reductions in the size and weight of atomic bombs, while thermonuclear technology reduced the amount of fissionable material needed in each bomb, thus greatly accelerating bomb production.

In 1948, an American thirty day air offensive against 70 Soviet cities with 133 atomic bombs was projected as causing a 30 to 40 percent reduction in Soviet industrial capacity and killing 2.7 million people. By 1955, when the U.S. stockpile numbered 2280, an American atomic campaign against the USSR and its Warsaw Pact allies was expected to result in the "total loss of 118 out of 134 major Soviet cities" and cause more than 77 million casualties, including 60 million dead. By 1961, full execution of the first Single Integrated Operational Plan (SIOP) was estimated to result in 285 million dead in the Soviet Union and the People's Republic of China.1

What planning constraints were established over the arsenal capable of wreaking such awesome destruction? Far fewer than might be supposed. From 1945 through 1974, the United States had no detailed national nuclear weapons employment policy. Indeed, National Security Council (NSC) policies relating to nuclear weapons were so general as to place virtually no limitations on their use. The first of these was approved in the fall of 1948, in the midst of the Berlin crisis. It was intended to resolve confusion over whether the military was authorized to plan for the use of atomic weapons. President Truman had rejected the first Joint Chiefs of Staff atomic war plan the previous spring and ordered a conventional alternative, a directive which was subsequently reversed by Secretary of Defense James Forrestal. The NSC originally intended to address the issue of when and how nuclear weapons might be used, but instead chose to endorse an Air Force originated paper that forcefully argued against limiting military options.

NSC 30, "United States Policy on Atomic Warfare," concluded that "in event of hostilities, the National Military Establishment must be ready to utilize promptly and effectively all appropriate means available, including atomic weapons, in the interest of national security and must plan accordingly." Nevertheless, "the decision as to the employment of atomic weapons in the event of war is to be made by the Chief Executive when he considers such decision to be required."2 NSC planners, including George F. Kennan, the head of the State Department Policy Planning Staff, endorsed these conclusions.

While placing final control over nuclear use in the hands of the President, NSC 30 gave the military complete freedom to plan and prepare for nuclear war according to their own best judgment. Such judgment was embodied in a special annex to what became the annual Joint Strategic Capabilities Plan (JSCP). Annex "C" to the JSCP was approved by the Joint Chiefs of Staff and was used by military planners overseas and in the newly established Strategic Air Command (SAC), to prepare operational plans for nuclear war.

Based on World War II experience, air power theory, and the very limited size of the American nuclear stockpile in the 1940s, military planners had initially identified the Soviet urban industrial base as the most lucrative attainable targets for attack. In 1949, the JCS approved an additional target category. Reflecting the recent commitment to Western Europe under the North Atlantic Treaty, attacks on "retardation" targets were given high priority. These were fixed targets that could be struck to slow the Red Army's invasion. They included additional petroleum refineries, electric power plants, submarine construction yards, and synthetic ammonia plants, including some in Eastern Europe. In August 1950, the JCS added a third target category and assigned it preeminence: the Soviet capability to deliver atomic bombs against the United States and its allies. The move recognized the need to defend the U.S. homeland by any means available. It also provided a powerful dynamic to increase the number of targets for a strategic attack. By 1955, this led to identification of some 960 airfields as potential targets, greatly increasing estimates of nuclear weapons requirements.3

Because of the need to get the most from limited resources (35 specially modified B-29 bombers in the spring of 1948 and 264 nuclear modified bombers in the summer of 1950), and lack of adequate intelligence on the Soviet Union, SAC planned to strike the full range of enemy targets in as large a simultaneous attack as feasible. Such an attack would overwhelm the presumably powerful Soviet air defense network and increase the chances that the bombers would get through to their targets and home again with minimal losses. This operational approach was maintained even after 1955, when SAC had increased to the point of having approximately 1200 operational bombers with trained crews available for war with a few hours warning. Since all targets were to be struck simultaneously, there was little or no meaningful discrimination between target categories prioritized by the JCS.4

In 1951-1952, the State Department developed a proposed set of procedures for making a decision on nuclear use. It recommended that the JCS would either initiate or immediately review any proposal that nuclear weapons be used, and that the president would then consult with the Secretaries of Defense and State and the Chairman of the Atomic Energy Commission before making his decision. Time permitting, he would also consult with congressional leaders, the American people, allied governments ("especially those whose consent is required before their bases can be used by the United States for atomic strikes"), and the United Nations. Although Truman did not approve or even comment on this proposal, the incoming Eisenhower administration came to view it as "established procedure."5

The question of consultation with allies is especially interesting. In 1943, while cooperating with Great Britain on atomic development, the United States had signed an agreement stating, among other things, the neither the U.S. nor U.K. would use atomic weapons without the other's prior knowledge and consent. This restriction, however, was removed by a new modus vivendi adopted in January 1948, which nullified existing atomic energy agreements between the U.S., Great Britain, and Canada.6

By 1950, the United States had deployed nuclear modified bombers and atomic weapons minus their cores of fissionable material to air bases in the United Kingdom, and was negotiating to do the same in Canada. In January 1952, President Truman and Prime Minister Winston Churchill agreed that no nuclear strike would be launched from U.K. bases without the consent of the British government. No other requirements for consultation were agreed to, although Truman assured the British he intended to keep communications open. Under Eisenhower, the NSC ultimately wrote into national policy that the president should consult with allies before using nuclear weapons, but only if time were available to do so.7

Except for the question of procedures for considering nuclear use and possible consultation with allies, the Eisenhower administration appears to have actively discouraged the development of constraints on nuclear war planning. During an October 1953 NSC discussion on the role of nuclear weapons in national policy, JCS chairman Admiral Arthur Radford sought to clarify exactly how the weapons might be used. This was to be a basis for U.S. force planning in both general war and limited conflicts, and for use in upcoming talks with the British. He proposed that a policy for the graduated use of nuclear weapons be adopted, one which assigned first priority to attacking "military forces operating against us or our allies," followed by their supporting structure. "Unrestricted atomic operations" would be launched only "in retaliation for such attacks on the west."8 President Eisenhower, however, was satisfied with existing language that "in the event of hostilities, the United States will consider nuclear weapons to be as available for use as other munitions." The only clarifying guidance was issued in January 1954. The document declared that while no advance decision to use nuclear weapons was implied by such language, "In certain cases," such as an atomic attack on the U.S. or Western Europe, "the use of nuclear weapons by the United States would be automatic." In other situations, such as limited hostilities, such immediate use would have to be weighed against whether it "would increase the danger of their strategic use by the enemy, lose the support of allies, expose them to devastation, or widen the Hostilities."9

By the mid-1950s, the question of whether the United States should plan for limited nuclear operations, in contrast to full-scale nuclear war, was practically moot. By end of 1954, after the French had failed to ratify the European Defense Community agreement and member states had not met the 1952 Lisbon conventional force goals. Moreover, after an intensive evaluation of the implications of the new nuclear technology by a multi-national New Approach Group of military officers, the North Atlantic Treaty Organization moved to adapt its future strategy so as to rely on the power inherent in both strategic and tactical American nuclear weapons. NATO strategic concepts approved between 1954 and 1957 mandated that in event of a Soviet attack, theater nuclear forces would be employed "from the outset" in conjunction with an "instant and devastating nuclear counter-offensive by all means available," particularly the U.S. Strategic Air Command, against the USSR itself. It was emphasized that "in no case is there a NATO concept of limited war with the Soviet Union."10 By 1956-1957, American army leaders were contemplating total requirements for nuclear weapons at 151,000, "104,000 for tactical battlefield use, 25,000 for air defense and 20,000 for support of our allies." A "typical field army" was described as using "a total of 423 weapons in one day of intense combat not including surface to air weapons." Between 1952 and 1958, the number of nuclear weapons allocated for war planning purposes to NATO's supreme allied commanders grew from 80 to more than 3500.11

In April 1956, NSC 30 was finally superseded as American policy on atomic warfare by a paragraph in NSC 5602/1, the current statement of Basic National Security Policy. It read:

It is the policy of the United States to integrate nuclear weapons with other weapons in the arsenal of the United States. Nuclear weapons will be used in general war and in military operations short of general war as authorized by the President. Such authorization as may be given in advance will be determined by the President.12

This remained national policy on the use of nuclear weapons through the end of the Eisenhower administration.

Like NSC 30, this statement placed no particular constraints on nuclear war planning, and seemed to mandate preparation for a variety of nuclear conflicts. In fact it was interpreted to mandate planning for primarily full-scale nuclear conflicts. In the spring of 1956, the JCS split over whether a war in Europe might be contained at the level of conventional or tactical nuclear weapons. With the apparent concurrence of the President, Secretary of Defense Charles Wilson ruled that JCS planning should be governed by the concept that "in a general war, regardless of the manner of initiation, atomic weapons will be used from the outset."13

NSC 5602/1 also introduced the important new concept that presidential authority over nuclear weapons could be transferred in part through pre-delegation. In 1956, even before the relevant systems were deployed, President Eisenhower approved advance authorization for the use of nuclear weapons in air defense. In May 1957, he issued an "authorization for the expenditure of nuclear weapons," which apparently gave U.S. theater commanders advance permission for nuclear response under a variety of emergency conditions. The instructions involved required extensive review and were not finally approved until 1959. In late 1959 and early 1960, Eisenhower authorized the secretary of defense to send to theater commanders and the commander of SAC additional, still-classified instructions pertaining to nuclear weapons expenditure. These authorizations appear to have been continued through at least 1964.14 One source indicates that these four-star commanders were permitted to pre-delegate their authority to their three-star subordinates in charge of corps, numbered fleets, and air forces, further weakening the constraint on nuclear weapons inherent in strict presidential control.15

Another possible avenue for placing constraints on nuclear war planning was through national policy pronouncements on war objectives. Between 1945 and 1960, three such statements were approved. The first was appended to NSC 20/4, the November 1948 statement of general U.S. objectives with "Respect to the USSR to Counter Soviet Threats to U.S. Security." It mandated reducing "the power and influence of the USSR to limits which no longer constitute a threat to world peace, national independence and stability in the world family of nations," but rejected "a predetermined requirement for unconditional surrender," occupation of the Soviet Union, or other specific expectations. The one exception was that any surviving "bolshevik" regime should not control enough military resources to threaten any other regime.16 The Truman administration's war plans were prepared under this statement.

In 1954, the Eisenhower administration approved a specific declaration of objectives for general war, beginning with the goal of achieving "a victory which will insure the survival of the United States," while preserving allies to the extent possible, and reducing the capabilities of the Soviet Union (and if necessary Communist China) to wage war. Among the other objectives was that of trying to enlist the support of the people in the Soviet Union and Soviet satellites in the American war effort, a goal that appears inconsistent with a nuclear offensive. During the NSC discussion of the paper, however, President Eisenhower noted that "in such a war the United States would be applying a force so terrible that one simply could not be meticulous as to the methods by which the force was brought to bear."17

In March 1959, in the midst of the Berlin crisis, the NSC approved a new statement of war aims that eliminated this modifying objective, and indeed removed any semblance of constraint war objectives might have placed on war planning. NSC 5904/1 reaffirmed that the United States should prevail and survive, and that it should bring the Soviet Union (and Communist China) "to the point where they have lost their will or ability to wage war against the United States and its allies." The only limitation proposed on the nuclear offensive was, if feasible and without jeopardizing other war aims, "to avoid unnecessary destruction and casualties in all countries not involved in the war."18 In 1960 Secretary of Defense Thomas Gates overruled a majority of the Joint Chiefs of Staff to reaffirm Wilson's earlier directive that JCS planning be guided by the principle that general war was any "war with the USSR." No possibility that such a conflict could be contained at the conventional level was to be considered.19

By this time it was apparent that the Eisenhower administration had no intention of imposing any meaningful policy constraints on planned nuclear use in general war. The president may have entertained notions of controlling the execution of U.S. war plans in circumstances other than a surprise attack, but the plans themselves had become exceedingly complex, requiring careful adherence to timing of attacks and bomber routes to targets to insure success. Further, between 1959 and 1961, U.S. nuclear capability had nearly doubled, growing from 12,305 weapons to more than 23,000 in barely two years.20 Nuclear delivery capability also grew, fueled by stockpile growth, and by improving intelligence which made it easier to identify additional Soviet targets.

The decision not to impose policy constraints on nuclear war planning clearly reflected the views of the commander in chief about the nature of nuclear war. Eisenhower desired to insure that the U.S. commitment to employ nuclear weapons in event of a Soviet attack on the U.S. or Europe was not subject to misperception by the Soviet leadership. Further, he believed that the more terrible the prospect of a nuclear war continued to appear, the more deterred the USSR would be from ever beginning one. Finally, Eisenhower was very much aware that any attempts to lessen the American response to a Soviet strike in this era when the U.S. had become vulnerable to not just attack, but a surprise nuclear "Pearl Harbor," for the first time in its history, might have grave domestic political ramifications. Explicitly limiting the nature of a U.S. nuclear attack, either by encouraging a retaliatory as opposed to preemptive strategy or imposing actual limits on the number or types of nuclear weapons procured would have strengthened the positions of congressional Democrats, who harped about American weakness in the face of a "bomber gap" in 1956 and a "missile gap" during 1957-1960.21

Army and Navy planners were concerned that lopsided emphasis on nuclear weapons and nuclear delivery capability was leaving the nation unprepared to fight the limited, conventional conflicts that they argued were the most likely future challenges under the emerging U.S.-USSR nuclear stalemate. While the arguments revolved around issues of forces and budgets, they offered potential for introducing constraints on war planning.

The first set of proposals looked to reducing the size of the planned SAC offensive. In August 1956, the Navy began to examine a possible "minimal target system, the threat of destruction of which would deter the USSR."22 By the next summer, Army and Navy planners jointly presented a critique of SAC's weapons requirements, demonstrating that more weapons than needed were being assigned to installations targeted for attack. In late 1957, an "alternative undertaking" target list was proposed. This effort sought to create an exclusively retaliatory target list of highest priority targets that would be struck if the United States had only twenty-five percent of its current nuclear strike force remaining after a Soviet first strike. The alternative list emphasized Soviet government controls and population centers and targeted residual Soviet nuclear capability only "to the extent profitable." Designed to supplement existing target lists, it also had the potential to replace them, if the decision were made to shift funds from nuclear systems into preparing to fight limited wars at the conventional level. Work on the alternative list continued into 1960.23

The "alternative undertaking" would not have altered the operational approach in existing plans for a comprehensive counterforce and urban/industrial attack. That issue was addressed by two other initiatives. The first was a Navy argument in favor of "finite deterrence, controlled retaliation" which built both on the targeting concepts underlying the alternative undertaking and on the anticipated deployment of the Polaris ballistic missile submarine. Polaris missiles, thanks to their secure basing in a relatively invulnerable submarine, could be withheld and used selectively against Soviet targets to "apply political coercion if we like to gain national objectives more advantageous than simple revenge."24 Although the primary targets for such attacks were urban industrial concentrations, overall casualties were expected to be considerably lower than in the combined attack on military and urban targets. Further, the invulnerability of the submarine-based missile force was expected to enhance deterrence by alleviating pressures to use or lose targetable land-based forces.

The anticipated development of ballistic missiles also gave rise to a very different initiative for controlling nuclear war, the "no cities counterforce" strategy advocated primarily by analysts at the RAND Corporation, an Air Force contractor. Like the Navy's proposal, it was based on the assumption that general war would begin with a Soviet attack, most likely against U.S. land-based strategic systems. Retaliation against population centers was perceived as a "confession of weakness."25 The appropriate response was for surviving U.S. nuclear forces to be targeted against unexpended Soviet nuclear capability, including mobile ICBMs, while withholding attacks on urban centers in expectation that the Soviets would do the same. Such a response would put pressure on the Soviet Union, while limiting further damage to the United States. This retaliatory damage-limiting concept, could also be used in preemption. If implemented, its emphasis on withholding attacks on cities would constrain SAC's total response by breaking it up into more discrete options.26

In the spring of 1958, President Eisenhower moved to investigate whether future requirements for strategic forces could be reduced without weakening the U.S. defense posture. He directed the NSC to undertake an analysis of what would constitute a minimum adequate deterrent and retaliatory capability, and subsequently of whether counterforce or urban industrial targets were best suited to deterrence. The so-called Hickey Committee of the NSC staff completed its report in February 1960. Consistent with the war aims of NSC 5904/1, the committee focused not on what targets would best deter, but what would allow the U.S. to prevail in war. They concluded that an "optimum mix" of counterforce and urban industrial targets, similar to but smaller than that already being targeted by SAC. When defense secretary Gates created the Joint Strategic Target Planning Staff (JSTPS) in August 1960 to consolidate the services' nuclear planning, the Hickey report was provided to it as guidance in the preparation of the first Single Integrated Operational Plan (SIOP).

The new JSTPS, headquartered at Omaha and directed by the commander of SAC, completed the first SIOP in December 1960. Designated SIOP-62, it would enter into force at the start of the 1962 Fiscal Year in July 1961. The plan governed the operations of all American strategic and many theater nuclear forces and was designed to be executed either in retaliation or as "a preemptive measure." It contained assurance of delivery factors of from 75 to 97 percent against each target category (nuclear targets, other military targets, and the urban industrial base), resulting in considerable redundancy in order to assure destruction of the specified targets. The target list included 3729 individual installations, many of which were co-located so as to be included in a single Designated Ground Zero (DGZ). The plan had 14 to 16 attack options, all based on the amount of alert time available. With 15 minutes warning, 1004 delivery systems, carrying 1685 weapons yielding some 2100 megatons, were to be launched against more than 650 DGZs in the Sino-Soviet Bloc. With strategic warning of 14 hours or more, 2244 bombers and missiles, carrying 3267 weapons yielding more than 7800 megatons, would attack a total of 1060 DGZs, including those in more than 150 urban areas. Eight hundred of the DGZs were defined as "military targets."27

SIOP-62 represented a technical triumph in the history of war planning. In less than fifteen years, the United States had mastered a variety of complex technologies and acquired the ability to destroy most of an enemy's military capability and much of the human habitation of a continent in a single day. Although mandated by high policy, the process of choices that had produced SIOP-62 was not arbitrary. It reflected analysis (not always based on adequate data) of the threat posed by the Soviet Union, and incorporated operational choices aimed at reducing the friction of war, to coordinate and protect bomber forces, and to integrate bomber and missile forces at the cusp of two eras in warfare. SIOP-62 was an American Schlieffen Plan, an ultimate strategy for war winning under all circumstances of war initiation, with an even less tenable basis in political and military realities than the German plan infamous for its inflexibility when executed in 1914.

Those responsible for creating the SIOP were not insensitive to the issue of constraints. SIOP-62, as the JCS briefing for President Kennedy emphasized, was not so inflexible as to preclude options. It was possible to withhold attacks against most targets in the satellite nations (radar sites, surface-to-air missile batteries, and other "defensive" targets would, however, have to be hit). It was also possible to withhold attacks against cities, if the President chose to do so. There were, however, strong arguments against making such a choice. The plan was designed to be executed "as a whole," and there would be no way to rapidly revise it to withhold attacks. In particular, forces withheld might not survive and thus would not be available for later use if needed: "Thus, withholding of a portion of the planned attack could degrade our plan and the forces committed to it to the point that the task essential to our national survival might not be fulfilled." Furthermore, "limiting attack to military targets" would have "little practical meaning as a humanitarian measure," since atomic weapons are "relatively non-discriminating, especially with respect to fall-out." With so many military targets located near population centers, it was impossible to substantially reduce civilian casualties through withholding, or to convince the Soviets that the U.S. was seeking to do so. It was far better to execute the plan in its entirety, if necessary, than to risk the nation's survival by seeking to impose constraints at the last minute.28

The incoming Kennedy administration could not simply reverse fifteen years of American war planning by high policy fiat. In addition to confronting the technical realities involved in attempting to change the SIOP, as noted in the JCS brief, there were political limitations as well. The Soviet Union was expanding its own nuclear forces, and the Democrats had been elected in part on their promise to close a perceived missile gap. The best the new administration could hope for was to gradually constrain nuclear war plans, bringing them under control by incremental actions.

Efforts to reshape and constrain nuclear planning took many forms over the next two decades. While John F. Kennedy explicitly declared in March 1961 that the nation's arms would not be used to strike first in the event of war, no such directives were sent to the Pentagon. Indeed, during the Kennedy and Johnson years, no specific NSC statements on national security policy, war aims, or national nuclear employment policy were issued. This was despite the fact that a number of such statements were drafted, and the JCS made clear that they would have found such guidance valuable. Rather, pointed National Security Action Memoranda (NSAMs) were issued by the President or National Security Adviser to prod the bureaucracy and improve responsiveness to presidential initiatives. Rarely did these memoranda become policy per se. As such, except in one significant instance relating to the Berlin Crisis of 1961, no attempt was made to mandate nuclear employment policy from the White House.

Instead, Secretary of Defense Robert McNamara used the enhanced powers of his office to regulate the procurement of both nuclear delivery systems and nuclear weapons, and also mandated changes in the JCS guidance relating to development of the SIOP. For example, whereas in the 1940s and 1950s the Joint Chiefs of Staff had sent requirements for nuclear weapons through a Military Liaison Committee directly to the Atomic Energy Commission, in the 1960s, Secretary McNamara's office provided a formal statement of nuclear requirements. This kept the 1961 stockpile of 23,000 weapons from doubling in size by the end of the decade, as 1961 JCS requirements dictated. Instead, the nuclear stockpile leveled off at some 32,500 weapons in the mid-1960s.29

In fact, McNamara came into office inclined toward a strategy of finite deterrence, and had little faith in first-strike counterforce. The briefing he received on SIOP-62 in Omaha, however, convinced McNamara that it would be extremely difficult to revise American war plans or cut forces back to a finite deterrent level. Seeking alternatives, he was very impressed with a subsequent briefing on the RAND "no-cities" counterforce concept, and had consultants draft a military strategy section of a basic national security policy statement to mandate such an approach.30

The Kennedy administration's attempt to reformulate nuclear strategy was accelerated and focused by the problem of Berlin. In fact, the continuing Berlin crisis, which began in 1948 and continued into the mid-1960s, provided many of the crystallizing moments in the history of U.S. nuclear strategy. The 1948 crisis had forced the American government to come to terms with nuclear weapons and acknowledge the necessity of incorporating them into war planning. Ten years later, the "Deadline Crisis" from November 1958 to March 1959 had prompted the Eisenhower administration to adopt a tough new declaration of war aims, and harden reliance on nuclear weapons. Also in 1959, the Live Oak planning cell, consisting of the U.S., Great Britain, and France, was started. Its purpose was to plan tripartite and NATO responses to Soviet belligerent actions toward Berlin. But proposals of extensive graduated action to determine Soviet resolve in Berlin met with skepticism from Eisenhower, who preferred a "one bite" strategy based on the theory that a single large probe, perhaps as much as a division moving down the Autobahn, was all that was needed to force the Soviets to back down or accept general war.31

For the Kennedy administration, Berlin was an equally important focal point. The intensive review of American military strategy in general and military options in Berlin began before the 1961 crisis. By the fall of 1961, it had led to directives to revamp American war planning. What was ultimately accomplished, however, was far less than the administration desired. The process illustrates the difficulty of imposing meaningful constraints on nuclear planning, even when the stakes are high and time is available to design reforms.

In October 1961, two directives were approved as the basis for reshaping American nuclear strategy. The first was NSAM 109, "U.S. Policy on Military Actions in a Berlin Crisis," which laid the groundwork for a series of graduated conventional and nuclear responses in a Berlin crisis, and for American efforts to promote a flexible response strategy for NATO. The second was the National Strategic Targeting and Attack Policy prepared by the Joint Chiefs of Staff and approved by Secretary McNamara as guidance for redesigning the SIOP.

NSAM 109 built upon a variety of studies and proposals. These included an April 1961 policy statement, based on a study prepared by former Secretary of State Dean Acheson, which proposed reorienting NATO strategy for localized, conventional contingencies, such as "a clash over Berlin." Advocates considered a massive Soviet attack on Europe unlikely. It also built on JCS studies ordered by Secretary McNamara of how many divisions NATO would be required to hold out over specified periods against a Soviet invasion of Europe without resort to nuclear weapons. The structure of NSAM 109 was furnished by an OSD study, initiated in the Army staff toward the end of the Eisenhower administration, that examined the full range of possible Soviet actions and allied countermeasures that a Berlin crisis could precipitate.32

NSAM 109 described a four-phase, preferred sequence for military operations over Berlin, from dealing with localized harassment of Western access, to challenging significant blockage, to expanding conventional ground and air operations in Europe and maritime pressures worldwide in the face of continued obstruction of access. The final step was the employment of nuclear weapons. The fourth phase called for responding to continued Soviet intransigence in the face of "substantial non-nuclear" pressure with first, "selective nuclear attacks" to demonstrate a willingness to use such weapons, second, "limited tactical employment" to gain tactical advantage and/or to increase pressure on the Soviets, and finally, if all else failed, general nuclear war. This planning directive was provided to the JCS and General Lauris Norstad, the Supreme Allied Commander, Europe, as well as to NATO allies (some, like the French, were never shown phase four) for discussion and approval over the next year.33

One question that concerned the Kennedy White House and McNamara Pentagon, for which NATO was not responsible, was how the United States might wage general nuclear war if such a contingency became unavoidable. Over the summer of 1961, key deputies in the NSC and the Office of the Secretary of Defense (OSD) examined the implications of new intelligence indicating that the Soviet Union's intercontinental nuclear forces were much smaller than had been thought. By mid-August, they produced a critique of SIOP-62 and the outline of an alternative strategic concept. The ultimate distribution of this concept remains unclear, but apparently a copy was transmitted from the White House to the JCS in September.34

The only available documentary evidence of this effort indicates that it explored how to capitalize on gaps in Soviet air defenses to launch a limited surprise first strike aimed at neutralizing Soviet ICBM, IRBM, and bomber forces. Ten B-52 bombers, each carrying 4 H-bombs and two nuclear-tipped air-to-surface missiles, would infiltrate at night at an altitude of 500 feet to carry out a synchronized attack with U.S. ICBM and Polaris missiles against ICBM and bomber bases deep within the Soviet Union. This would be followed up by a relatively small number of strikes by airborne-alert B-47 bombers and theater land- and carrier-based aircraft against Soviet theater nuclear forces. Damage estimates projected that the Soviets would have two or fewer of their ICBMs remaining, as well as 30 IRBMs, 45 medium bombers, and 5 sea-launched ballistic missiles.35

Possible Soviet responses to such an attack were considered limited: if they attacked U.S. cities, a maximum of nine million American fatalities was anticipated; if the Soviets responded against European cities, 36 million might perish. The U.S. would retain the bulk of its nuclear forces to deter such city attacks, however. The circumstances under which such a strike might be launched were not predicted. Further analysis indicated that if Soviet nuclear forces were alerted at the time of a U.S. attack, many more of their forces would survive. Soviet counter-city strikes against the U.S. and Western Europe would cause more than 100 million deaths.36

Despite the transmission of this study to the JCS, it appears to have played no role in the development of guidance for revising SIOP-62. Some off-line contingency planning may have been undertaken in the Joint Staff, but its ultimate fate is shrouded in mystery. The new SIOP guidance, which had been debated by the JCS through the summer of 1961, was presented to Secretary McNamara on 23 October and approved two days later.37 Recently declassified material provides new insights into the changes mandated by this guidance.

SIOP-63, as completed in June 1962 to enter into force on 1 August, broke SIOP-62 into five attack options, some of which were designed for preemptive execution, others for retaliation. There were three tasks assigned involving destruction to specified damage levels of, respectively, Soviet nuclear capability, other military forces, and the urban-industrial base. The five attack options did not address each target category (much less any subsets) separately. Rather, the options were cumulative, each adding a target category to the previous one. All required the expenditure of thousands of nuclear weapons and were subsequently criticized as "five options for massive retaliation."38 They permitted the President to order strikes "against nuclear threat targets only, against nuclear threat plus other military targets, or against nuclear threat plus other military plus urban-industrial targets." Provisions were included for withholding attacks, by task, against satellite countries and the Peoples Republic of China, or against military and government controls in the Moscow or Peking area so as to keep open the possibility of a negotiated settlement. Finally, "selected forces of high survivability" were kept in reserve "to be employed as the last resort threat in any retaliation situation," apparently against the Soviet urban-industrial base.39

The new SIOP was prepared over the protests of the Strategic Air Command. SAC expressed "doubt that there were sufficient forces available to withhold a reserve for later use, doubt that forces held in reserve would survive, and doubt above all that the command and control system [which had been built with emphasis on surviving only long enough to get the Emergency War Order through to launch the forces] would operate long enough or effectively enough to launch and direct them."40 As the USSR deployed increasing numbers of strategic forces, such doubts would grow. Nevertheless, this basic design for the SIOP remained the foundation of U.S. nuclear employment strategy for more than a decade and a half.

On 17 October 1961, even before the Kennedy administration promulgated NSAM 109 or the JCS had sent the new SIOP guidance to Omaha, Soviet Premier Khruschev informed the 22nd Communist Party Congress that he would not insist on signing a peace treaty with the German Democratic Republic before the end of the year. The threat of an immediate confrontation abated, although tensions and harassment continued. The planning set in motion by NSAM 109, however, was carried forward into 1962. In January, President Kennedy and his advisors met with General Norstad and reviewed, and possibly amended, the command and control procedures within his command for the selective and limited use of nuclear weapons.41 On 20 June (after he had given his famous speeches at Athens, Greece, and Ann Arbor, Michigan, announcing the counterforce shift in U.S. nuclear targeting policy), Secretary McNamara was briefed on SIOP-63.42 Then, on 18 August, President Kennedy and his advisors were briefed on the status of Berlin contingency planning.

The briefing outlined the plans that had been developed over the past year. A wide array of options had been identified, for initiatives short of military force in Phase I and II, and for Tripartite (U.S., U.K. and France) or Quadripartite (plus Germany) military initiatives in Phase III, which would be turned over to NATO control if allied forces came under attack. Phase IV, which was defined by the U.S. as beginning with any use of nuclear weapons, whether in response to Soviet use, for demonstration purposes, or "to avoid defeat of major military operations," had been prepared for by nuclear annexes attached to the plans for each type of military operation to be used in Phase III, and a specific nuclear only plan for the demonstrative use option.

Phase IV posed the most complex problems of coordination and control, especially since the allies were not in agreement on the role nuclear weapons might play in a Berlin confrontation. The British agreed with the U.S. assessment that the Soviet Union was unlikely to deliberately launch a general nuclear war, given its relatively weak strategic position, but felt that more attention should be paid to the threat of general war begun through miscalculation. The French, on the other hand, were skeptical about conventional conflict scenarios, and believed that nuclear weapons should be used early to avoid a prolonged war in Europe.43 President Kennedy appears to have been equally doubtful about the prospects for success of the list of options mandated by NSAM-109. He interrupted the briefer's explanation of the subtleties of Phase IV to remark: "I suppose if we get involved in a war in Europe we will have no choice but to use nuclear weapons."44 This frustration may have fed the President's caution when faced with the Cuban Missile Crisis two months later.

The Cuban Missile Crisis, for all its significance in Cold War history as the preeminent nuclear crisis, appears to have had little impact on the efforts to change American nuclear planning. Guidance issued for SIOP-64 on 14 November 1962 "had very few substantive changes from previous guidance."45 Barely ten months later, however, on 12 September 1963, American nuclear planning reached a significant but until now previously unheralded milestone. The Net Evaluation Subcommittee of the National Security Council reported its findings to President Kennedy and the full NSC on the results to be expected in a nuclear war from 1964 through 1968. After the briefing, the President asked the head of the subcommittee, Air Force General Leon Johnson, "whether, even if we attack the USSR first, the loss to the U.S. would be unacceptable to political leaders." Johnson responded "that it would be, i.e., even if we preempt, surviving Soviet capability is sufficient to produce an unacceptable loss in the U.S." Secretary of Defense McNamara further emphasized that "there was no way of launching a no-alert attack against the USSR which would be acceptable," and "thus, preemption, today or in 1968 is not an acceptable course of action."46

Subsequent studies indicated that even with a 1950 Minuteman force (in contrast to the 1100-1300 then approved by McNamara for procurement through 1968), the least the U.S. could hope to suffer in retaliation for a preemptive attack on a smaller than expected future Soviet missile force was 28 million American and 60 million Western European dead. Against a "medium" Soviet force and with a nation-wide civil defense program, the U.S. could expect 58 to 95 million dead and 75 to 100 million allied fatalities, depending on whether 950 to 1950 Minutemen were deployed.47 Further analysis by the Office of Secretary of Defense raised even more questions about whether the U.S. could use offensive or even existing air and planned missile defense weapons systems to limit damage and reduce casualties in the U.S.48

As a result, McNamara seized on the concept of "Assured Destruction" (the ability to have a secure second strike force, which under any conditions could destroy half of Soviet industry and 20 to 25 percent of the population in large cities) as a basic measure for the size of the American programmed force. While the attack options in the SIOP did not change to reflect this priority, guidance issued through 1974 gave highest priority in the assignment of weapons to urban industrial targets first, then to nuclear threat and other military forces. "The capability of destroying urban-industrial targets was to be assured even with inadequate warning of an attack" on U.S. strategic forces.49

The end of counterforce preemption as a serious option, however, did not represent an end to the counterforce mission for SIOP-assigned U.S. nuclear forces. First, it is not clear whether President Lyndon Johnson was made aware of the NESC's findings upon taking office after Kennedy's death. Johnson was not at the September NSC meeting. McNamara declared in 1983 that he had warned both presidents against initiating nuclear use, but he has not provided dates or details of those discussions. The defense secretary did tell Johnson that "neither side can blunt an attack by the other" at a December 5, 1963 NSC meeting on Soviet military capabilities, and Johnson received a brief on executing the SIOP from JCS Chairman Maxwell Taylor on December 9, and a full JCS SIOP brief in August 1964. No records are available to indicate what Johnson understood or might have decided about executing the war plans available to him.50

Second, the United States did not abandon "damage limitation" as a task for strategic offensive forces. Rather, the emphasis was placed on "second-strike counterforce," which involved targeting residual Soviet nuclear forces after a first strike on the West, or launching U.S. forces on less than perfect warning in hopes of destroying as many enemy missiles and bombers on the ground as possible. This had become an extremely difficult challenge. Attacking Soviet follow-on forces was now less feasible than in the early to mid-1960s, when many of those forces had to undergo a laborious fueling and arming sequence prior to launch, including mating missiles to their nuclear warheads that were stored separately.51 In addition, new calculations involving heretofore unconsidered results of nuclear explosions-electromagnetic pulse, atmospheric ionization, and the dust created by near-surface explosions-suggested that the problem was far more complex than had previously been supposed. Taking into account the rapid growth of the Soviet ICBM force, such complexities appeared to have "pushed the counterforce problem beyond coherent calculation."52 The only way to carry out the assigned task appeared to be to place even greater emphasis on redundant cross-targeting of available systems, thereby increasing the likelihood of unintended damage to non-military targets.

In less than a decade from 1961, American nuclear force structure changed dramatically. The bomber force was significantly reduced. One thousand fifty-four hardened and dispersed land-based intercontinental ballistic missiles were deployed, as were 41 submarines armed with 656 ballistic missiles. The SIOP, however, remained basically unchanged. SIOP-64 was replaced by SIOP-4 in 1966 (the annual designation of the plan having been dropped). It continued the patterns begun by the first "flexible response" plan, SIOP-63, designed in 1961. While withholding of attacks on cities was provided for, large-yield warheads, redundant cross-targeting of warheads to achieve high damage expectancy, and the maintenance of damage criteria based only on the primary effects of nuclear explosions (blast and over-pressure) rather than such secondary effects as fire and radiation meant that even the military-targets-only attacks would produce extremely high civilian casualties. SIOP-4 remained the basic U.S. nuclear war plan through the mid- 1970s.53

McNamara had wanted to introduce greater flexibility and discrimination into nuclear war planning. When he left office in 1968, he described the lack of more controlled and deliberate response options as "one of the main weaknesses in our posture today."54 Nevertheless, his embracing of the Assured Destruction programming requirement meant that he put little effort into pressuring the JCS in Washington or the JSTPS in Omaha to further refine the SIOP to make it more discriminate, while the demands of the continuing counterforce requirement also tended to work against discrimination.

Two other factors complicated the push for greater flexibility and restraint. The first was the escalating requirement for MIRV warheads on SLBMs and ICBMs to counter the anticipated deployment of a Soviet anti-ballistic missile system. McNamara left office calling for a force of nearly 6000 independently targeted missile warheads by 1976. With several warheads on each missile, a decision to use even one meant the destruction of multiple targets. The single warhead missiles retained were less accurate and had a greater yield per warhead than the multiple warhead missiles that were deployed.55

The second problem was the enormous technical difficulty involved in perfecting and deploying systems that would permit greater flexibility and selectivity in nuclear force employment. In particular, limited and discriminate nuclear options required effective "tactical warning" sensors and "attack assessment" networks (TW/AA) to provide information on enemy nuclear operations, including the size and targeting of incoming attacks; the capability for rapid retargeting of ballistic missiles to accommodate changed circumstances; and secure and enduring command, control and communications (C3) systems to allow the President (or his successor) to transmit orders for a strike or a series of strikes over a protracted period of nuclear conflict with confidence that they would be received.

The capability to "classify the attack, as small or large, for example, accidental or deliberate, selective or indiscriminate, against cities or not, against high command or not" was critical in supporting "a decision as to an 'appropriate' retaliatory response."56 It was not until 1971, with the launch of the first operational Defense Support Program (DSP) orbital infrared detection satellite designed to detect missiles in the powered-launch phase, that potential warning time could be extended from the 15 minutes afforded by the Ballistic Missile Early Warning System (BMEWS) radar to "perhaps" 27 minutes. DSP-type systems improved "the capability to assess an attack and even evaluate the likely intentions of the attacker, and [did] so by a wide margin over other warning and surveillance systems." They were critical in providing leaders with extra time, particularly "to perform essential retaliatory command functions, including more opportunity to ascertain the situation and consider desirable alternatives," and "promised, for the first time-nearly a decade after programs were initiated to develop sufficiently flexible strategic forces and sufficiently flexible command and control systems, and a sufficiently flexible SIOP war plan-to make flexible response options more than a remote possibility."57

Likewise, the capability to retarget missiles rapidly was not available until nearly a decade and a half after SIOP-63's guidance was approved. Through the mid-1970s, it took as long as 90 days to retarget the Minuteman force. On an emergency basis, the target tapes of a single missile might be changed in 9 hours. The command data buffer system, which became operational in 1975-1976, enabled the re-targeting of a Minuteman III missile with 3 MIRV warheads in 36 minutes. In addition, guidance software changes increased the number of execution plans pre-stored in missile launch control centers from 100 to 200. With respect to submarine launched ballistic missiles, even if the submarines had had a rudimentary capability for rapid re-targeting, the SLBM's lack of accuracy and the fact that a single missile launch might give away the submarine's location served to diminish their potential utility in limited strikes.58

Creating a survivable command and control system was the most difficult problem. Such a system "would require a considerable capability to assemble, evaluate, and utilize information, almost certainly in an extremely fast-moving situation, probably with normal information systems, communications, and command operating under great stress, overload or other abnormal conditions, and perhaps with impaired or degraded capabilities."59

From the 1960s on, major improvements were made to the American strategic C3 system, including redundant communications links to the bombers, ICBMs and submarines, airborne and other mobile command posts for the President and major commanders, improved warning and attack assessment systems, and improved intelligence gathering capabilities against the Soviet Union. One major study of the results of this vast and expensive modernization program concluded in the mid-1980s, however, that even in response to a "limited" attack against U.S. nuclear forces, the vulnerabilities of the command system would "encourage comprehensive retaliation." Further,

a limited Soviet counterforce attack would not be unambiguously limited, and it would trigger preplanned operations that, in conjunction with extensive collateral damage to the command system, would create strong pressures for organizing retaliation around a single plan that releases most of the SIOP retaliatory forces. Regardless of what calculations political leaders might make at the time, they would have to select the plan that would achieve everything they would expect to accomplish in retaliation for the duration of the conflict.60

Despite this dismal prognosis, efforts continued in the 1970s and 1980s to increase options available to the Western alliance in general, and the President of the United States in particular, in using nuclear weapons. The main thrust of these efforts was to create punctuation points or pauses in a nuclear confrontation, to provide the opportunity for reevaluation before entering into a general nuclear exchange.


1 David Alan Rosenberg, "American Atomic Strategy and the Hydrogen Bomb Decision" Journal of American History 66 (June 1979) 70-73; Briefing of WSEG Report No. 12, "Evaluation of An Atomic Offensive in Support of the Joint Strategic Capabilities Plan," April 8, 1955, Document Two in David Alan Rosenberg, "'A Smoking Radiating Ruin at the End of Two Hours': Documents on American Plans for Nuclear War with the Soviet Union, 1954-1955" International Security 6 (Winter 1981/1982) 29-38; Fred Kaplan, The Wizards of Armageddon (New York: Simon and Schuster, 1983) 268-270.

2 NSC 30, United States Policy on Atomic Warfare, September 10, 1948, in FRUS, Vol. I, Part 2, General; The United Nations (Washington, D.C.: Government Printing Office, 1976) 624-628.

3 David Alan Rosenberg, "The Origins of Overkill, Nuclear Weapons and American Strategy, 1945-1960" International Security 7 (Spring 1983) 16-39.

4 Ibid. 18-39. On the growth of nuclear deliver capability, see David Alan Rosenberg, "U.S. Nuclear Stockpile, 1945-1950" The Bulletin of the Atomic Scientists (May 1982) 25-30; and Philip A. Karber and Michael D. Yaffe, "U.S. Strategic Nuclear Forces, 1946-1990, Volume I: Land-Based Bombers" Unpublished study prepared for the Nuclear History Program Data Base Project, August 13, 1992.

5 Draft Memorandum by N. Gordon Arneson, Special Assistant to the Secretary of State, April 24, 1951, Memorandum by Mr. Carleton Savage, Member of the Policy Planning Staff, May 23, 1951, Possibilities of War With The Soviet Union: Use of Nuclear Weapons, and Paper Prepared by the Joint Strategic Survey Committee and Representatives of the Department of State, August 3, 1951, United States Position On Considerations Under Which The United States Will Accept War And On Atomic Warfare, FRUS, 1951, Vol. I, National Security Affairs; Foreign Economic Policy (Washington, D.C.: Government Printing Office, 1979) 820-826, 834-840, 866-874; Staff Study prepared by Representatives of the Special Committee of the National Security Council on Atomic Energy, June 11, 1952, and Paper prepared by R.R. Bowie, the Director of the Policy Planning Staff to the Secretary of State, FRUS, 1952-1954, Volume II, National Security Affairs (Washington, D.C.: Government Printing Office, 1984) 973-979, 565-576. See also Leonard Wainstein, C.D. Cremeans, J.K. Moriarty and J. Ponturo, The Evolution of U.S. Strategic Command and Control and Warning, 1945-1972 (Arlington, Virginia Institute for Defense Analyses Study S-467, June 1975; Top Secret study, declassified 15 September 1992) 40-43.

6 Minutes of the Meeting of the Combined Policy Committee, at Blair House, Washington, D.C., January 7, 1948, with Draft Agreement Between the Governments of the United States, the United Kingdom, and Canada, Annexes A, B, and C appended, FRUS, 1948, Vol.I 679-687.

7 Wainstein, Cremeans, Moriarty and Ponturo, Evolution of Strategic Command and Control and Warning 40-43; Staff Study Prepared by Representatives of the Special Committee of the National Security Council on Atomic Energy, June 11, 1952 op.cit.; NSC 162/2, Report to the National Security Council on Basic National Security Policy, October 30, 1953, FRUS, 1952-1954, Vol. II, 593.

8 Admiral Arthur W. Radford, Memorandum for Mr. John Foster Dulles, 13 October 1953, CJCS 040 Atomic Energy Commission, Records of the United States Joint Chiefs of Staff, Record Group 218, U.S. National Archives (hereafter JCS).

9 Memorandum of Discussion at the 166th Meeting of the National Security Council, Tuesday, October 13, 1953, and NSC 162/2, October 30, 1953, FRUS, 1952-1954, Vol.II, 546-547, 593; James S. Lay, Jr., Memorandum for the Secretary of State, Secretary of Defense, and Chairman, Atomic Energy Commission, Subject: Policy Regarding Use of Atomic Weapons, March 14, 1955, reaffirming (and reproducing) policy statement of January 4, 1954, Atomic Weapons, Correspondence and Background for Presidential Approval and Instructions for Use, Box 1, NSC Series, Subject Subseries, Dwight D. Eisenhower Library (hereafter DDEL).

10 Quotations from NATO Strategic Concept MC-14/2, April 1957 taken from Robert A. Wampler, NATO Strategic Planning and Nuclear Weapons, 1950-1957 (College Park, Maryland: Nuclear History Program and University of Maryland Center for International Studies Occasional Paper 6, 1990), 40-41; see also Robert A. Wampler, "Ambiguous Legacy: The United States, Great Britain and the Foundations of NATO Strategy, 1948-1957" (Unpublished Ph.D. Dissertation, Department of History, Harvard University, 1991).

11 Office of the Assistant to the Secretary of Defense (Atomic Energy), "History of the Custody and Deployment of Nuclear Weapons, July 1945 through September 1977" (February 1978 Top Secret Study declassified with many deletions 1990), Copy Courtesy Dr. Michael Yaffe, 50; Wampler, NATO Strategic Planning and Nuclear Weapons 2, 52.

12 Paragraph 11, NSC 5602/1, Basic National Security Policy, March 15, 1956, FRUS, 1955-1957, Vol.XIX, 246; The fact that NSC 30 was superseded by paragraph 11 of NSC 5602/1 is noted in James S. Lay, Memorandum for the National Security Council, February 17, 1959, enclosing Annotated List of Serially-Numbered National Security Council Documents (updated to December 30, 1960), Subject Filing of NSC Papers Folder, Box 20, Subject Series, Alphabetical Subseries, White House Office, Staff Secretary Papers, DDEL.

13 Kenneth W. Condit, History of the Joint Chiefs of Staff, The Joint Chiefs of Staff and National Policy, Vol. VI, 1955-1956 (Washington, DC.: Historical Office, Joint Staff, 1992) 12-21, 30-38.

14 Rosenberg, "The Origins of Overkill," 48-49, especially note 163; Daniel Schuchman, "Nuclear Strategy and the Problem of Command and Control," Survival, 29 (July-August 1987) 336-359. Such authorization was recently reconfirmed with the April 1993 declassification of Brigadier General Andrew Goodpaster, Memorandum of Conference with the President, 16 January 1961, dated January 18. 1961, Staff Notes, January 1961, Box 55, Dwight D. Eisenhower Diaries, DDEL. It describes General Lyman Lemnitzer's discussion with President Eisenhower of his briefing of incoming Defense Secretary McNamara "on the emergency action plan, as well as on the subject of authorizations by the President for employment of nuclear weapons." Extension of these authorizations in a "Furtherance" paper is noted in Major General C.V. Clifton, Memorandum for the Record, March 14, 1964, Meetings with the President, Vol. 1, National Security Files, C.V. Clifton Papers, Lyndon B. Johnson Library (hereafter LBJL).

15 Daniel Ellsberg, "Stranger Than Strangelove" WIN Magazine (A short-lived anti-nuclear activist magazine) November 17, 1977, 4-8, 15, 21; Schuchman, "Nuclear Strategy and the Problem of Command and Control" 344-350.

16 NSC 20/4, Note by the Executive Secretary on U.S. Objectives With Respect to the USSR To Counter Soviet Threats to U.S. Security, November 23, 1948, FRUS, 1948, Vol. I, 662-669.

17 NSC 5410/1, Note by the Executive Secretary to the National Security Council on U.S. Objectives in the Event of General War With the Soviet Bloc, March 29, 1954, and Memorandum of Discussion at the 190th Meeting of the National Security Council, Thursday, March 25, 1954, FRUS, 1952-1954, Vol.II 637-646.

18 NSC 5904/1, dated March 17, 1959 remains classified. This information is based on the draft paper, NSC 5904, U.S. Policy in the Event of War, February 19, 1959, and S. Everett Gleason, Memoranda of Discussion at the 398th and 399th Meetings of the National Security Council, Thursday, March 5, and Thursday, March 12, 1959, in 398th and 399th Meeting Folders, Box 11, NSC Papers, Ann Whitman File, DDEL.

19 Brig. Gen. H.L. Hillyard, Memorandum for General Twining, Admiral Burke, General White, General Lemnitzer, General Shoup, Subject: Definitions for Use in U.S. Military Planning, SM-185-60, 26 February 1960, and CSAFM 84-60, 29 February 1960, both in CCS 3120, JSCP (24 August 1959) JCS. JCS 1844/321, 22 September 1960, CCS 3120, JSCP, (24 August 1960), Sec. 1, ibid. notes Secretary Gates decision on 15 April 1960.

20 Nuclear stockpile growth is documented in Wainstein, Cremeans, Moriarty and Ponturo, Evolution of U.S. Strategic Command and Control and Warning 34, and "Nuclear Notebook" The Bulletin of the Atomic Scientists November 1989, 53.

21 This view is based on discussions with Professor Robert R. Bowie, former-Director of the State Department Policy Planning Staff, 1953-1957, and Professor Richard H. Immerman, who are collaborating on a book on national security strategy-making in the first year of the Eisenhower administration. See also Richard H. Immerman, "Confessions of an Eisenhower Revisionist: An Agonizing Reappraisal" Diplomatic History 14 (Summer 1990) 319-342, and John Lewis Gaddis, "The Origins of Self-Deterrence" in The Long Peace, Inquiries into the History of the Cold War (New York: Oxford University Press, 1987) 104-146.

22 Vice Admiral R.E. Libby, Memorandum for Op-00 (Chief of Naval Operations), Subject: Proposals relative to Atomic Operation Concept, Serial BM00043-57, 1 May 1957, File A16-10, Atomic Warfare Operations, Box 8, Op-00 Files, 1957, Operational Archives, Naval Historical Center, Washington, D.C.

23 Rosenberg, "The Origins of Overkill" 50-54.

24 Admiral Arleigh Burke, CNO Personal Letter No. 5 to Retired Flag Officers, Subject: Pertinent Information, Summary of Major Strategic Considerations for the 1960-1970 Era, 30 July 1958, Navy Folder, Box 28, General Thomas D. White Papers, Library of Congress.

25 Herbert Goldhamer and Andrew W. Marshall, with the Assistance of Nathan Leites, The Deterrence and Strategy of Total War, 1959-1961: A Method for Analysis, RM 2301 (Santa Monica, California, RAND Corporation, U.S. Air Force Project RAND, April 30, 1959) 191.

26 Kaplan, The Wizards of Armageddon 201-219, 238-247; Rosenberg, "The Origins of Overkill" 57-60.

27 Rosenberg, "The Origins of Overkill" 4-8, 61-65; Scott D. Sagan, "SIOP-62: The Nuclear War Plan Briefing to President Kennedy" International Security 12 (Summer 1987) 22-51; Kaplan, The Wizards of Armageddon 266-272; May, Steinbruner and Wolfe, History of the Strategic Arms Competition 592-597.

28 Sagan, "SIOP-62" 49-51.

29 Information on the levelling off of the nuclear weapons stockpile and the JCS requirements is contained in "Nuclear Notebook," The Bulletin of the Atomic Scientists, November 1898, 53, and Deputy Secretary of Defense Roswell Gilpatric, Memorandum for Mr. McGeorge Bundy, Subject: Nuclear Weapons, 3 March 1961, Nuclear Test Ban, April 1959-June 1961 Folder, Box 49-56, Theodore Sorenson Papers, John F. Kennedy Library (hereafter JFKL). McNamara's role in addressing the stockpile is indicated in Robert S. McNamara, Memorandum for the Secretaries of the Military Departments et. al., 1 March 1961, enclosing list of 96 projects, Defense, January-March 1961 Folder, President's Official File, Box 77, JFKL; the JCS reply on Project 5, General L.L. Lemnitzer, Memorandum for the Secretary of Defense, Subject: Reexamination of the Current and Long-Range Department of Defense Requirements for Production of Nuclear Weapons and Fissionable Material, JCSM-241-61, 18 April 1961, Directorate of Freedom of Information and Security Review, Office of the Secretary of Defense (hereafter OSDFOI); and General Maxwell Taylor, Memorandum for the Secretary of Defense, Subject: Nuclear Weapons Production and Study Guidelines, JCSM-1028-62, 27 December 1962, Tactical Nuclear Weapons Study Folder, Box 33, Maxwell Taylor Papers, National Defense University (hereafter MDT-NDU). According to a list of Draft Presidential and Tentative Record of Decision Memos at the Office of Program Analysis and Evaluation, Office of the Secretary of Defense, McNamara provided the President with annual statements outlining Nuclear Materials Production Schedules though Fiscal Year 1973 (1964), Ten Year Nuclear Weapons Program (1965), Nuclear Weapons Stockpile and Fissionable Materials Requirements (1966, 1967 and 1968); these memoranda remain classified. On Kennedy's White House National Security organization and McNamara's style of management, see Deborah Shapley, Promise and Power, The Life and Times of Robert McNamara (Boston: Little, Brown, 1993) 94-146, W.W. Rostow, The Diffusion of Power (New York: MacMillan, 1972) 160-184, and Rosenberg, "Reality and Responsibility" 35-38, 43-44.

30 Kaplan, The Wizards of Armageddon 252-283; Shapley, Promise and Power 104-111.

31 Historical Studies Division, Historical Office, Bureau of Public Affairs, U.S. Department of State, Crisis Over Berlin, American Policy Concerning the Soviet Threats to Berlin, November 1958-December 1962 (Unpublished Top Secret Research Project No. 614-A, October 1966; Declassified with deletions, July 1987) Parts I-IV, November 1959-January 1961. See also Marc Trachtenberg, History and Strategy (Princeton: Princeton University Press, 1991) 169-234.

32 Jane Stromseth, The Origins of Flexible Response (New York: St. Martin's Press, 1988) 30-41; McGeorge Bundy, National Security Action Memorandum No. 40, for Members of the NSC, enclosing Policy Directive on NATO and the Atlantic Nations, 24 April 1961, NSC-1961 Folder, Box 4, Vice Presidential Security File, LBJL; Admiral Arleigh Burke, Memorandum for the Secretary of Defense, Subject: NATO Force Requirements Study (Project 106C), JCSM-306-61, 5 May 1961, CCS 9050/3410 NATO (29 April 1961) Sec.3, JCS; and Transcript of Nuclear History Program Oral History Interview on Contingency Planning in the Kennedy Administration with Brig. Gen DeWitt Armstrong, USA (Ret.), Vice Admiral John M. Lee, USN (Ret.), Colonel Lawrence Legere USA (Ret.) and Ambassador Seymour Weiss, 7 June 1991. It was then-Colonel Armstrong's "Horse Blanket" paper that was boiled down to a "Pony Blanket" and then "Poodle Blanket" of preferred options as the basis of NSAM 109.

33 McGeorge Bundy, National Security Action Memorandum 109 to the Secretaries of State and Defense, Subject: U.S. Policy on Military Actions in a Berlin Conflict, October 23, 1961, and letter, President John F. Kennedy to General Lauris Norstad, with Enclosure, October 20, 1961, declassified in August and October 1992, respectively, National Security Archive, Washington, D.C. (hereafter NSA).

34 The above is based on an interviews with Professor Carl Kaysen, Deputy National Security Adviser to President Kennedy in 1961, (conducted with Marc Trachtenberg) Cambridge, Massachusetts, August 3, 1988, and with Dr. Henry Rowen, Deputy Assistant Secretary of Defense for International Security Affairs in 1961, Washington, D.C., March 20, 1991; General Maxwell Taylor, Memo for General Lemnitzer, Subject: Attached Memorandum on "Strategic Air Planning and Berlin," 6 September 1961, Berlin Planning Folder, Box 34, MDT-NDU; Trachtenberg, History and Strategy, 224-225; and Kaplan, The Wizards of Armageddon 296-303.

35 No copy of the actual limited counterforce strike paper has yet been found in government archives, although it is likely that a copy exists in Carl Kaysen's still unprocessed files at the Kennedy Library. This description is based on a Draft Memorandum for the President, Subject: Consequences of Thermonuclear War Under Various Conditions of Outbreak, October 28, 1961, unclassified paper in the mailroom rotary historical file in the Office of Program Analysis and Evaluation, Office of the Secretary of Defense. This document was discussed in person with McGeorge Bundy, Robert McNamara, Carl Kaysen, Henry Rowen (who in his capacity in March 1991 as Assistant Secretary Defense for International Security Affairs cleared the document for public discussion with this explanation), Marcus Raskin of the 1961 NSC Staff, and in telephone interviews with Frank Trinkl and W.W. Kaufmann, both consultants to OSD, and Fred Kaplan. None of them could identify the document itself. Kaufmann could not reliably recall the work in question. Rowen noted that while it looked like a "competent calculation," it was "not in any way, shape or form a planning document" he recalled being associated with, and further emphasized that it was a "non-paper" that "never went anywhere." Kaysen and Trinkl agreed that it represented something of the analyses they had been engaged in, but the draft memorandum was not written the way they would have written it. Kaplan recalled that General David Burchinal, then director of the Joint Staff, indicated in an interview with him that work on the limited strike plan went forward in a very secret cell under Burchinal's direct control, progressing to the point where actual units had been identified for participation, but its ultimate fate was uncertain. This document leaves no doubt that some "off-line" nuclear planning was conducted in the summer and fall of 1961. In the absence of any more concrete data, its contents are offered here as representing the constraints being considered at that time.

36 Draft Memorandum, October 28, 1961.

37 History and Research Division, Headquarters, Strategic Air Command, History of the Joint Strategic Planning Staff, Preparation of SIOP-63 (Prepared January 1964; Partially Declassified under Appeal, 1992) 7-13; Rear Admiral F.J. Blouin, Note to Control Division, Subject: Guidance for the Preparation of SIOP-63, 23 October 1961, and Rear Admiral Blouin, Decision on JCS 2056/285 (Guidance for Preparation of SIOP-63), 25 October 1961, CCS 3105 (Joint Planning) (8 March 1961) (3) Sec. 2, JCS.

38 Wainstein, Cremeans, Moriarty and Ponturo, Evolution of U.S. Strategic Command and Control and Warning 357, n.13.; Henry S. Rowen, "Formulating Strategic Doctrine," Part III of Vol. 4, Appendix K to The Report of the Commission on the Organization of the Government for the Conduct of Foreign Policy (Washington, D.C.: Government Printing Office, 1975) 220-234.

39 SIOP Decision Handbook, ca. 1971, quoted in Wainstein, Ibid., 346-347, 290-291.

40 Ibid., 290.

41 McGeorge Bundy, National Security Action Memorandum No. 122 to the Secretaries of State and Defense, January 16, 1962, and Acting Secretary of State George Ball, Memorandum for the President, Subject: SACEUR's Command and Control Procedures for Use of Nuclear Weapons (NSAM 122), January 24, 1962, (Both documents declassified April 1993) Folder 110, Box 37, MDT-NDU.

42 McNamara was briefed on SIOP-63 on June 20, 1962. He gave his famous Top Secret Athens speech to NATO on counterforce-oriented nuclear strategy on May 5, 1962, and an unclassified speech on the same subject to the commencement at the University of Michigan in Ann Arbor on June 16. History of JSTPS, Preparation of SIOP-62, 25-26; David N. Schwartz, NATO's Nuclear Dilemmas (Washington, D.C.: The Brookings Institution, 1983) 156-165. Whether the defense secretary was satisfied that what he had set in motion in October 1961, and spoken about so forcefully in May and June 1962, was in fact accomplished by the completion of SIOP-63 is not known.

43 John C. Ausland, Memorandum for Mr. Hillenbrand, Subject: Briefing for President on Berlin, August 2, 1962, with enclosed Top Secret briefing, Berlin Contingency Planning, declassified January 29, 1993, National Security Archive. See also DEFE 4/142, COS(62) 7th Meeting, 23 January 1962, Minute 4, Confidential Annex, attachment: JP(62)6(Final), "Berlin Contingency Planning--Phasing of Military Operations, Report by the Joint Planning Staff," 19th January 1962, Top Secret planning paper released to the public January 1993, Public Record Office, London.

44 John C. Ausland, "A Nuclear War to Keep Berlin Open?" International Herald Tribune, June 19, 1991; Transcript of Nuclear History Program Oral History Interview with John C. Ausland, October 8, 1991.

45 History and Research Division, Headquarters Strategic Command, History of the Joint Strategic Target Planning Staff: Preparation of SIOP-64 Top Secret History prepared August 1964, Declassified with deletions following appeal, 1992), 13-15.

46 Bromley Smith, Summary Record of National Security Council Meeting, September 12, 1963, 11:00 AM--Report of Net Evaluation Subcommittee, in NSC Meetings 1963, No. 517, 9/12/63, Box 314, National Security File: Meetings and Memoranda, JFKL. See also "Resume of Discussion During NESC Meeting of 12 September 1963," same file.

47 Secretary of Defense Draft Memorandum for the President, Subject: Recommended FY1965-FY1969 Strategic Retaliatory Forces, December 6, 1963, OSDFOI, 10, 14, 19, 21.

48 Kaplan, Wizards of Armageddon 315-327; Shapley, Promise and Power 191-201.

49 Rowen, "Formulating Strategic Doctrine" 220-232.

50 Summary Record of National Security Council Meeting, December 5, 1963, 3:00 PM--Subject: Soviet Military Capabilities, Vol.1, 12/5/63 Meeting, NSC Meetings File, Box 1, National Security File; Major General C.V. Clifton, Memorandum for Record, December 9, 1963, Meetings with the President, Volume 1, National Security Files, Box 2; and James V. Cornell for C.V. Clifton, Memorandum for Mr. Bromley Smith, August 17, 1964, SIOP Folder, Box 2, both C.V. Clifton Papers, all LBJL. Robert S. McNamara, "The Military Role of Nuclear Weapons: Perceptions and Misperceptions," Foreign Affairs 62 (October 1983) 59-80.

51 Sagan, Moving Targets 31-37; Rowen, "Formulating Strategic Doctrine" 229-233; Berman and Baker, Soviet Strategic Forces 18, 90-93.

52 May, Steinbruner and Wolfe, History of the Strategic Arms Competition 602-604.

53 Ibid. 596-608. See also Bruce G. Blair, Strategic Nuclear Command and Control, Redefining the Nuclear Threat (Washington, D.C.: The Brookings Institution, 1985) 79-181.

54 Draft Presidential Memorandum, Subject: Fiscal Year 1969-1973 Strategic Offensive and Defensive Forces, January 5, 1968, OSDFOI.

55 McNamara's last Draft Presidential Memorandum called for 5,943 independently targeted missile warheads. DPM, January 5, 1968. The conclusion about the capabilities of U.S. forces in the late 1960s is based on an analysis of warhead yields discussed in Cochran, Arkin and Hoenig, U.S. Nuclear Forces and Capabilities, 10-11, 39, and the estimated U.S. missile accuracies provided in Donald MacKenzie, Inventing Accuracy, A Historical Sociology of Nuclear Missile Guidance (Cambridge, Massachusetts: MIT Press, 1990) 427-435.

56 Wainstein, Cremeans, Moriarty and Ponturo, The Evolution of U.S. Strategic Command and Control and Warning 292.

57 Ibid. 345-346.

58 "Targeting Flexibility Emphasized by SAC" Aviation Week and Space Technology May 10,1976, 29-34; David A. Anderton, Strategic Air Command (New York: Charles Scribner's Sons, 1977) 138-139.

59 Wainstein, Cremeans, Moriarty and Ponturo, The Evolution of U.S. Strategic Command and Control and Warning 292-294.

60 Blair, Strategic Command and Control, 239-240. See also Desmond Ball, Can Nuclear War Be Controlled (London: International Institute for Strategic Studies, Adelphi Paper 169, 1981) 35-38, and Paul Bracken, The Command and Control of Nuclear Forces (New Haven, Connecticut: Yale University Press, 1983).

23 September 2003