LOOKING BACK: The Limits of Limited Nuclear War

William Burr

In recent years, the Bush administration has put forward policy documents that seem to suggest that nuclear weapons can be used as limited war-fighting tools, such as for bunker-busting missions. As policymakers grapple with whether such a strategy is possible or necessary, they might contemplate the Nixon administration’s internal policy discussions on nuclear weapons use more than three decades ago. While President Richard Nixon and his national security advisor, Henry Kissinger, presided over a quest for limited nuclear options, others raised searching questions about the value or relevance of that approach.

The Nixon administration began by considering a basic problem: the nuclear war plans that were the foundations of deterrence during the Cold War would have caused the deaths of hundreds of millions of people. The catastrophic nature of the U.S. nuclear war plan, the Single Integrated Operational Plan (SIOP), made Nixon and Kissinger wonder if there was a less suicidal, more credible way to make nuclear threats meaningful and reduce the danger of all-out nuclear war in the event of a superpower confrontation. In his quest for contingency plans for limited strategic options, Kissinger was successful to the extent that, in January 1974, Nixon signed a directive mandating such planning. Nevertheless, U.S. government officials, civilian and military alike, questioned whether limited nuclear options were any more plausible than the massive attack options. The arguments for limited strategic options rested on such shaky foundations that even Kissinger professed to have lost his faith in them, although the planning would continue.

Massive Attack Plans

The Nixon administration’s search for nuclear options began early in its history and flowed from a larger concern about the credibility of U.S. power, not least the plausibility of strategic nuclear threats to defend NATO forces in Europe. Only days after the inauguration, Nixon and Kissinger heard a briefing on the latest version of SIOP-4, which had been approved in 1964. Guiding the SIOP was the National Strategic Targeting and Attack Policy, which treated nuclear threat targets (e.g., missile silos, strategic bomber bases, nuclear weapons stockpiles, and command and control centers) as the highest priority targets for destruction. The SIOP gave the president five basic options for retaliatory or pre-emptive attacks. If the Pentagon received strategic warning of an imminent Soviet attack, the president had a choice of three different pre-emptive attacks: against Soviet nuclear threat targets (counterforce), against nuclear and conventional forces, or against those forces as well as Soviet urban-industrial centers. If the Pentagon received notice of an incoming missile attack, the president had a choice of two retaliatory attacks, either against Soviet military targets only or against all military and urban-industrial targets.

The scale of the attack options was massive. By late 1971, for example, a pre-emptive strike against Soviet bloc military and war-supporting urban-industrial targets would have involved some 4,200 nuclear weapons targeting 6,500 installations. Some installations were near each other, making it possible to use one weapon against several.

Nixon, according to one National Security Council (NSC) staffer, was “appalled” by the briefing. In messages to Congress during the following years, Nixon criticized the lack of alternatives to massive nuclear strikes. For his part, before taking office Kissinger had become well known for his advocacy of limited nuclear options (e.g., in his 1957 book, Nuclear Weapons and Foreign Policy.) Soon after the briefing, Kissinger asked the Pentagon to study an option that would give the president a choice for less than an all-out nuclear strike. He wanted to be able to use military force for signaling political intentions; if a superpower confrontation turned violent, he wanted “smaller packages” of limited nuclear options available to signal the greater danger of all-out war.

Kissinger’s proposals for “smaller packages” met bureaucratic resistance. Although some like-minded Air Force officers and RAND Corp. analysts prepared studies on the possibility of selective nuclear options, they received no high-level support at the Pentagon. Indeed, the Joint Chiefs of Staff thought that the idea of smaller nuclear options was dangerous. For the Joint Chiefs, creating smaller nuclear attacks would weaken the SIOP. Composed of “mutually supporting tasks and options,” the SIOP’s strike options were designed to be delivered in their entirety rather than in piecemeal form. For example, a small attack against selected Soviet missile bases could provoke a large counterattack with soft nuclear delivery systems launched to prevent their destruction by U.S. follow-on attacks. Soviet forces could then destroy U.S. forces assigned to the SIOP, thus making it difficult to execute the options.

The Joint Chiefs also argued that, if the Soviets realized that the United States was contemplating smaller nuclear strikes, it would reduce the “deterrent value” of the overall U.S. nuclear posture. Moreover, a small nuclear strike would encourage the Soviets to put their nuclear forces on a higher state of alert, creating disadvantageous circumstances if U.S. leaders anticipated additional nuclear strikes.

It was not only the Joint Chiefs that raised objections. CIA and civilian Pentagon analysts were skeptical as well. For example, during an interagency meeting, Department of Defense official Ivan Selin observed that the Soviets were “not apt to make a discriminating attack.” Kissinger, however, “wondered how one rationally could make a decision to kill 80 million people” and “whether if we make limited use of nuclear weapons, the Soviets would make an all-out response.” CIA Deputy Director for Intelligence R. Jack Smith opined that that indeed was the case: “Once nuclear weapons start landing, the response is likely to be irrational.”

Neither dissuaded by the bureaucratic resistance nor persuaded by the counterarguments, Kissinger continued to rail against the “horror strategy” for several years, but by late 1971, developments were starting to move in a direction more favorable to his cause. A major study by the NSC Defense Program Review Committee (DPRC) probed the pros and cons of limited nuclear options and suggested the need for limited strategic options. Like Nixon and Kissinger, the analysts believed that U.S. presidents should have “appropriate strike options” short of all-out strategic attacks, not least to make the Soviet Union more uncertain about U.S. responses to attacks on allies. Moreover, a policy of “restraint and resolve” during a crisis might make it possible to limit a nuclear exchange without sacrificing “vital” foreign policy interests.

The committee nevertheless saw serious risks that cast doubt on the whole enterprise of limited nuclear options. A serious problem was whether greater flexibility for the use of strategic weapons would lower the threshold for nuclear use. Having a capability for limited nuclear strikes could make them “more tempting” to use in a crisis. More options for the president could increase the danger of a nuclear catastrophe.

Another dangerous risk was the possibility that limited U.S. nuclear use in a crisis “could trigger uncontrolled escalation to general nuclear war.” There was no “reliable evidence” that Moscow had any plans for responding to a limited U.S. attack with limited strikes. Not only had the Soviets maintained that a conflict with the United States would escalate quickly, but military theorists had argued that “a limited conflict is both unlikely and inherently unstable.” If Washington ordered limited nuclear strikes on Soviet targets, there would be no “sound way” to predict the Soviet response, which could range from an attempt to negotiate to escalation with large nuclear strikes. How Moscow would have responded to a limited U.S. strategic nuclear attack on the Soviet homeland was necessarily an imponderable.

Given the risks and uncertainties, the January 1972 DPRC report drew out two implications. Washington should consider limited strategic strikes “only if faced with a challenge to vital national interests” and only after having taken steps to avoid escalation through such means as diplomatic channels, public announcements, use of the hot line, and strict avoidance of attacks that could weaken Soviet central control over nuclear weapons.

Just as the DPRC was completing its study on strategic nuclear forces, Secretary of Defense Melvin Laird directed senior civilian and military officials to undertake a review of U.S. nuclear targeting policy. First, Laird had to take into account pressure from the White House: Every year, in his annual foreign policy statement, Nixon had called for alternatives to massive nuclear attacks. Second, Pentagon officials realized that the U.S. war plan would have to be adjusted soon, as the United States and the Soviet Union were on the verge of signing a Strategic Arms Limitations Talks (SALT) agreement more or less banning anti-ballistic missile (ABM) systems. Once a SALT agreement was in place, thousands of U.S. multiple independently targetable re-entry vehicles (MIRVs), which were being deployed as a hedge against a Soviet ABM system, would have to be retargeted. To deal with these two issues, Laird created a special committee headed by Director of Defense Research and Engineering John Foster to review targeting policies and plans.

New Concepts for Nuclear Use

By mid-1972, the Foster panel developed concepts for massive and much less than massive nuclear options designed for controlling the escalation of nuclear war in the name of deterring all-out war. For example, limited attack options could be used when policymakers wanted to restrict the scale of violence. They would involve small numbers of tactical or strategic weapons that could be used for “response in kind” to gain “local advantage” in a military encounter or for “signaling” (to demonstrate the great risk of general nuclear war). Other options could involve regional attacks or selective attacks that could, for example, target Soviet naval or conventional forces threatening NATO. Major attack options were essentially the same as the SIOP retaliatory and pre-emptive options, just a repackaging. The panel, however, proposed a new focus for targeting priorities: the destruction of Soviet capabilities to recover from a nuclear exchange. This focus on political, economic, and military “recovery” targets may have been intended to give more scope for the MIRVs that were entering the nuclear stockpile.

Kissinger liked the Foster panel’s general approach and ordered an interagency review to look more closely at the implications of the Foster panel. The National Security Study Memorandum 169 report, completed by mid-1973, gave Kissinger a general endorsement of the Foster panel’s strategy proposals. It also pointed to uncertainties in the new strategy, for example, the possibility that limited options could weaken deterrence, impair NATO relations, and weaken the effectiveness of U.S. nuclear forces (e.g., withholding command and control targets could “facilitate enemy retaliation”). In addition, there was “no guarantee that escalation can be controlled,” for example, whether an adversary would read a limited attack as a signal to stop fighting or to escalate.

Despite the risks, the group concluded that controlling escalation in order to encourage an early termination of conflict provided the “most promising means” of limiting damage to the United States and its allies. Moreover, the panel speculated that various evidence, including Warsaw Pact planning documents, presumably acquired through espionage, and Soviet command and control exercises, suggested that Moscow, despite its public dismissal of limited nuclear war, might be developing its own strategies for fighting limited nuclear war. Yet, the language on Soviet options was carefully hedged; no one could be sure how Moscow would respond.

Panel members debated the impact of the new policy on deterrence, with some holding that, if an adversary believed that Washington would not “employ its central strategic systems…in response to local aggression,” conflict would be more likely. Others argued that the focus on post-war recovery targets would enhance deterrence by establishing a “direct threat to each of the three main power blocs within the Soviet Union and the People’s Republic of China, namely, the political regime, the technocrats, and the military.” Still others observed that political leaders or war-supporting industry, much less Soviet conventional forces, could not be targeted without damaging population centers. That observation suggested that the new policy was not much different from “assured destruction”; critics would see it as a “cynical attempt to rationalize what we would actually do in a massive attack.”

While NSC staffers were working on a directive for Nixon to sign, Winston Lord, one of Kissinger’s closest advisers, expressed a dissent. Lord, the director of the State Department’s Policy Planning Council questioned whether the concept of “controlled nuclear escalation” had been sufficiently thought through. Worried about lowering the nuclear threshold, he argued that the production of more and more plans for limited nuclear use implied an “over-reliance on nuclear forces.” Indeed, the NSC report “comes very close to regarding the use of nuclear weapons as routine.” The limited use of strategic weapons against Soviet targets posed “incalculable risks” of escalation. Finally, the new strategy could produce U.S. “overconfidence in the applicability of nuclear escalation,” a problem that could have unpredictable consequences. If the Soviets accepted the new U.S. approach, “might not [they] have some advantages in playing a tit-for-tat nuclear game?” Then, the danger was that Washington could be “more exploitable.” Kissinger thought Lord had written a “good paper,” but he was not persuaded.

Action on the presidential directive did not come quickly. The October 1973 Middle East war, the ongoing Watergate crisis, and debate over some extraneous issues delayed action on nuclear options until January 17, 1974, when Nixon signed National Security Decision Memorandum 242. For the first time, a U.S. president tried to impose specific guidelines for nuclear war planning.

A week earlier, Secretary of Defense James Schlesinger spontaneously announced some elements of the new policy in a famous press conference. Kissinger may have been furious when the new approach became known as the “Schlesinger doctrine” because Kissinger had been one of the prime movers. In any event, by early April, Schlesinger had signed a directive, “Nuclear Weapons Employment Policy,” that gave military planners detailed instructions for developing new targeting plans that were responsive to the White House’s search for limited options. Yet, the civilian-military consensus for limited options remained relatively shallow. For example, nuclear planners at the Pentagon dragged their feet, even questioning whether civilians should be involved in such sensitive issues as nuclear targeting or whether limited strategic options even made sense.

By mid-1975, Kissinger was getting fed up with the whole exercise. He told some arms control experts that he had “been here for seven [sic] years asking them to produce a war plan…. They can’t do it.” “The SIOP is still the same. All they have done is [to] give Omaha some theoretical planning capability.” When the work on SIOP-5 was finished in early 1976, it showed the same problems that had troubled Kissinger in the past. The emphasis on recovery targets had made SIOP planning more complex, but the thrust of it was familiar. It mainly consisted of large nuclear options that were useless for the kind of signaling that Kissinger and Schlesinger had sought. In the following years, however, U.S. war plans began to provide for truly small limited options; they now range between 2 and 120 nuclear weapons, close to what Nixon-era policymakers were seeking.

The turn toward limited nuclear options was an understandable development, but it could not reduce the great danger of nuclear deterrence. Not only was the course of nuclear conflict wholly unpredictable, command control and communications systems were vulnerable to disruption, making escalation control illusory.

With the end of the Cold War, the hazard of a catastrophic nuclear exchange has receded, but arguments about a Chinese nuclear danger and attempts to produce and deploy nuclear bunker-busters demonstrate that senior U.S. government officials still see nuclear threats and limited nuclear use concepts as policy relevant. How much support in the bureaucracy such concepts may enjoy is unclear. It is possible that mid-level officials see serious dangers in any use of nuclear weapons just as their forerunners did during the 1970s.

Corrected online August 29, 2008. See explanation.



William Burr is a senior analyst at the nongovernmental National Security Archive and directs its nuclear history documentation project. This article draws on a longer essay by the author published in the Journal of Cold War Studies.