Since the Nuclear Test Ban Treaty in 1963, American arms-control policy has been dominated by efforts to prevent the further spread of nuclear weapons capabilities. Proposals for a non-proliferation agreement-a treaty which would prohibit signatory states not already having nuclear weapons from acquiring them-have recently been put on the table by both East and West in the form of draft treaties, but beyond that little real progress has been made. Meanwhile, the growing nuclear capability of Communist China has led a number of non-nuclear countries to reassess their strategic security positions and to ponder whether they, too, do not want to manufacture nuclear weapons. To deter them, the nuclear powers have intensified their search for something to sweeten the non-proliferation pot.

The United States has held out the prospect of support against nuclear threats to countries which continue to refrain from acquiring nuclear weapons of their own. On October 18, 1964, before the dust from the first Chinese nuclear explosion had settled, President Johnson declared that the United States would continue to work against "nuclear spread." He then said: "The nations that do not seek national nuclear weapons can be sure that, if they need our strong support against some threat of nuclear blackmail, then they will have it."[i] The United Kingdom has also included in its negotiating position the possibility of some form of guarantee against nuclear attack for those nations willing to pledge themselves to a course of nuclear-weapons abstinence. To date, however, no nuclear power has been prepared to be specific about how such guarantees might work.

To guarantee the security of a nation without nuclear weapons against nuclear attack seems at first glance to be a reasonable, in fact minimal, price to pay to avoid proliferation. However, the deeper implications of nuclear guarantees need to be examined before we decide to move from the domain of the speechmaker, where the matter presently rests, to the kind of foreign policy upon which nations act.

First, we should clarify what we mean by a "national nuclear weapons capability." It rests on the possession of nuclear weapons, means of delivery and the ability of the national government to make its own decision on the use of the weapons. None of these three elements can be ignored. An effective delivery system in particular may be more difficult to acquire in some cases than nuclear warheads.

The present non-nuclear nations seem to fall into two categories: those with sufficient technological expertise, industrial competence and economic base to acquire a nuclear weapons capability, once they have made the decision, within a reasonable period of time-say five years; and those which are simply not close enough to the nuclear "take-off point" to be a primary concern. Admittedly, however, the take-off point depends a great deal on a nation's willingness to make the necessary economic sacrifices.

There remains the possibility, too, that states may obtain a nuclear weapons capability with substantial outside assistance, ranging from technical and material aid to the outright transfer of nuclear warheads or delivery vehicles. The two major nuclear powers seem fairly well agreed to avoid giving assistance in making warheads. The Soviet Union is not happy with the results of having helped Communist China; and the United States is not likely to transfer nuclear warheads to the control of any other single nation, even though it will not at present agree to a non-proliferation treaty which forecloses the option of developing a multilateral nuclear force within NATO. Control over the transfer of nuclear-capable delivery systems is, however, complicated by the fact that many types of aircraft can be adapted to carrying either conventional or nuclear payloads, and there appears to be less agreement between the two powers in this matter. The United States has transferred outright to other countries both ground- to-ground missiles and aircraft, and the Soviet Union has at least transferred aircraft.

Communist China does not adhere to the philosophy of non-proliferation, but its own requirements will probably prevent it from offering substantial nuclear aid abroad, either warheads or delivery systems, in the near future.


In discussing various possible types of guarantees for non-nuclear nations we may find it helpful to break the problem into three components: first, possible variations in the nature of the undertaking; next, possible parties to the undertaking; and finally, strategic postures necessary to support it.

In regard to the undertaking itself, both form and substance must be carefully considered. A guarantee might take a variety of forms, from a unilateral statement of national policy to a solemn treaty commitment. Even though a unilateral declaration would not be legally binding it could doubtless be regarded as a definite "commitment" by the guarantor, especially, in the case of the United States, once domestic public opinion had solidified in support of the policy. But the non-nuclear beneficiary would not be likely to regard such a simple statement of policy as a sufficient guarantee of its security.

Even a binding treaty obligation is likely to lose some of its credibility over time, especially if the good relations between the states prevailing at the start are later impaired for one reason or another. And in any case, any treaty obligation which restricts a nation's freedom of action in matters touching its own security comes into doubt when that nation feels its security in jeopardy.[ii]

In summary, the force of a unilateral policy expressed in sufficient detail to make clear that it will be a basis for future action should not be underestimated, nor should the binding character of a guarantee in treaty form be overestimated.

As for the substance of the guarantee, a key consideration would be the circumstances in which it could be invoked. Assurances of general military assistance are incorporated into various existing alliances. In the NATO treaty, for instance, assistance is to be given if an "armed attack" occurs "in Europe or North America." A parallel obligation under a nuclear guarantee could be to take action in case of "nuclear attack" against the territory of designated non-nuclear states or regions. This would mean that the guarantor would not be compelled to do anything until a nuclear attack had actually occurred.

For a potential victim, that would be little comfort. What would matter would be the strength of the deterrent against nuclear attack created by the probability or possibility of retaliation by the guarantor. But the circumstances which create reliability in the eyes of allies on the one hand, and credibility in the minds of potential adversaries on the other, may not always be identical.

The nuclear aspect of the U.S. security guarantee under NATO is believable to many of our European allies partly because we have committed substantial numbers of troops in West Germany. It is rightly inferred that the United States would not expose a large number of its own men along the central European front without being willing to go all the way up the conflict escalator, if necessary, in defense of the region. At the same time, from our point of view, it is important that NATO be able to withstand a conventional assault by conventional means, in order to keep nuclear weapons in the last-resort category where they belong.

We should not over-generalize from NATO defense problems, but the NATO experience and others do suggest that nuclear guarantees cannot easily be divorced from the issues of conventional defense. To work as a non- proliferation inducement, a guarantee of security formally confined to cases of "nuclear attack" may well need to entail a further, implicit commitment.

The nature of that commitment is an important question. In the event of an armed attack in the NATO area, the United States is obligated to take "forthwith . . . such action as it deems necessary, including the use of armed force." It is apparent that the military commitment is imprecise and the nature of the response is left largely to unilateral determination. In the SEATO treaty and several bilateral defense treaties the language of the guarantee is even weaker-the obligation on a signatory being merely to "act to meet the common danger in accordance with its constitutional processes." The question arises whether, in light of these treaties, any potential guarantor power would be prepared to commit itself in advance to making a sufficiently well-defined response to justify a non-nuclear nation in placing full reliance on it.

A guarantee could range from unilateral to universal. One nuclear power could simply declare to the world, as the United States seems to have done, its intention to come to the aid of non-nuclear countries that are victims of nuclear attack. Or it might exclude certain countries-for example, the Communist countries-or be limited to a single designated country. A policy for the benefit of a particular non-nuclear nation may have some utility, especially if expressly addressed to an imminent nuclear threat. A unilateral declaration addressed to the world at large will probably not be taken too seriously by anyone. In both instances, the absence of a definite treaty obligation would limit credibility.

A joint declaration by two or more nuclear powers would be one way of strengthening the guarantee. An example would be for the United States and the Soviet Union to guarantee jointly the security of non-nuclear nations against nuclear attack. This degree of Soviet-American coöperation has so far been well beyond reach, and the Soviet Union's problems with the Chinese, as well as its deep-seated mistrust of the U.S. nuclear posture, seem to preclude such an approach now. But it might become more attractive, especially as applied to countries outside the European area, as the Chinese nuclear menace grows and if the Southeast Asian conflict is moderated.

In the case of a non-nuclear nation which already has mutual security arrangements with a nuclear power, a specific nuclear guarantee might perhaps be viewed as adding to the strength of the overall defense commitment. However, not all non-nuclear nations would find a simple bilateral arrangement agreeable. Those nations not already in alliance with either the United States or the Soviet Union would consider a guarantee from either country an unacceptable departure from a non-aligned position. Here a joint guarantee, or possibly a guarantee "through the United Nations," might ease the problem.

It may be useful to revert at this point to President Johnson's statement of October 18, 1964, still the most explicit official U.S. position on the subject of nuclear security guarantees.[iii] He described the circumstances which would invoke American support as "a threat of nuclear blackmail." The President added that "if they need our strong support . . . they will have it." A simple declaration of "strong support" may well be lost in the background noise generally pervading international diplomacy.

Since the American position is couched in sufficiently ambiguous terms not to disturb anyone, we should not expect it to have a measurable effect on the decisions of governments which possess the possibility of developing nuclear weapons.

A nuclear guarantee which involved the deployment of nuclear forces in a non-nuclear country might be more effective in dissuading it from developing a nuclear capability of its own. Even though the guarantor power retained full ownership and control of those forces, some muscle would obviously be added to its guarantee. But such a course might also, it is true, raise strategic and tactical problems similar to those that have plagued NATO for over a decade.

This suggests a further question: what effects would nuclear guarantees have on existing alliance systems? The security of some of our closest allies already rests on a U.S. nuclear guarantee, backed up by extensive and elaborate deployments of both nuclear and non-nuclear forces. To a somewhat lesser extent our collective defense treaties guarantee the security of 28 other nations outside NATO.

New nuclear guarantees would probably have to be more specific than the existing loosely-worded alliance obligations, or alternatively would have to provide physical evidence of the undertaking in the form of deployed nuclear forces. But the stronger the nuclear guarantees given to states outside the alliance structure, the more likely it is to demoralize our present allies and lead them to make increased demands on us, or possibly to pay increased attention to General de Gaulle's nuclear thesis.

It is worth noting that the Soviet Union would probably arouse less dissent within the Warsaw Pact if it offered a nuclear guarantee to a non-member than the United States would find among its allies if it proposed comparable action. But for important reasons growing out of the conflict with China, the Soviet Union is less likely than the United States to entertain the notion seriously.


It remains to consider the potential effect of offering nuclear guarantees to specific countries which have a nuclear option. India, Israel and Sweden deserve first attention.[iv] Each has the prerequisites, each has a serious security problem and none is under the collective security umbrella of either the United States or the Soviet Union, at least in any formal sense.

India, today a target for nuclear blackmail by Communist China, is of primary concern to those interested in keeping the nuclear club from expanding. It has a relatively advanced civil nuclear industry, including its own plant for separating plutonium from irradiated reactor fuels, and though its economy is in desperate shape it could probably stand up under the burden of a modest military nuclear program. India is clearly capable of providing itself with the warhead component.

India would not find it so easy to come by a delivery system. Moreover, although Communist China will be able to strike from its own territory at India's vitals with glorified V-2s, the reverse is not true for India. To make a nuclear strike at Peking or the industrial heartland in Manchuria, India will require a long-range delivery system-either long-range land- based missiles, bomber aircraft or medium-range sea-based missiles. Conversion of a portion of the Indian civil jet-transport fleet is, of course, a possibility, but any savings would be at the expense of an increase in vulnerability. With the added burden of the delivery system, the cost to India of exercising its nuclear option is substantially increased and should provide a powerful argument against a decision to "go nuclear."

The effect of India's decision on Pakistan also deserves consideration. Pakistan would hardly consider India's acquisition of nuclear armaments as a step toward peaceful relations. Pakistan shows little signs of giving up its ability to play both ends of the Sino-Indian conflict, and in the circumstances India cannot hope to obtain Pakistan's support for embarking on a nuclear-weapons program on behalf of all the nations south of the Himalayas. The potential snowball effect in South Asia of such a program is apparent.

Would a guarantee against nuclear blackmail help India decide to forego the chance of developing a nuclear capability? Since Communist China began setting off nuclear blasts, India has raised its price for making an unequivocal renunciation; it indicates that it wants considerably more than a simple guarantee. An undertaking "through the United Nations" to safeguard the security of non-nuclear nations is only part of its demand. Also included are an agreement by the nuclear powers not to transfer nuclear weapons to other powers, a comprehensive nuclear test ban and a freeze on further production of nuclear weapons and delivery systems, coupled with substantial reductions in existing inventories. Moreover, India has not excluded Communist China from the list of required subscribers to a non-proliferation agreement. It seems clear that a "paper" guarantee will not be enough for India. Furthermore, any guarantee other than one by the United Nations, or conceivably one given jointly by the United States and the Soviet Union, would be unacceptable as contrary to India's commitment to non-alignment.

Israel's situation is different from India's in almost every respect. The technological and industrial comparison between Israel and the state which constitutes its chief threat is to a large extent the reverse of India's. Israel is clearly ahead of the United Arab Republic in terms of the base upon which a nuclear weapons capability can be built. Moreover, one or two nuclear warheads dropped on the Aswan Dam could submerge the whole lower Nile Valley. Israel may, indeed, view nuclear weapons as necessary insurance against the gradual dwindling of its present superiority in conventional forces.

In addition to cost, the basic deterrent to Israel's adoption of a nuclear weapons program is uncertainty about the reaction of interested powers outside the Middle East. The U.A.R. could be expected to respond with an outburst of righteous indignation and a demand for nuclear help. Conceivably the Soviet Union in these circumstances would make an exception to its adamant non-proliferation policy. If it were to step in and redress the nuclear balance in the Middle East, could Israel be certain what the U.S. reaction would be?

Israel has not fixed a price for its pledge to abstain from nuclear weapons. For some time, though, it has been interested in obtaining an assurance of security from the United States. From the Israeli point of view, even an American guarantee confined to assistance against nuclear attack might hold distinct advantages. So far, however, the United States has refused to make a commitment to one side in the Arab-Israeli conflict.

The United States might accept a guarantee generalized to include all the nations of the Middle East. Such a guarantee might be given in coöperation with the Soviet Union. In such an arrangement Israel and the Arab nations alike would agree not to acquire nuclear weapons and the United States and the Soviet Union jointly would guarantee all of them against nuclear attack. Steps to create such a "nuclear free zone" in the Middle East could be taken, if necessary, outside the framework of current negotiations on other non-proliferation measures. Political obstacles for both the Israelis and the Arabs are formidable; moreover, open coöperation with the United States in the Middle East is probably impossible for the Soviet Union at present. But the all-round gain in stability in the Middle East that would result from avoiding a nuclear arms race there justifies at least an effort to secure a breakthrough.

The issue for Sweden is whether its neutrality will be better served with or without nuclear weapons in its arsenal. Sweden would not attempt to mount a major nuclear force against the Soviet Union; its position relative to the Soviet Union is even more hopeless in a quantitative sense than India's to China. But unlike India, Sweden's medium-range land-based missiles could reach some vital targets in the Soviet Union. Sweden may also wish to keep its nuclear option open until it sees how West Germany's nuclear ambitions are to be satisfied or denied.

Not surprisingly, given Sweden's exposed position, it asks something in return for its signature to an agreement not to acquire nuclear weapons. Generally, Swedish officials have talked in terms of a comprehensive nuclear test ban. In view of Sweden's traditional adherence to neutrality it is unlikely to welcome a security guarantee from the United States, and a guarantee from the Soviet Union would of course be rejected. A nuclear guarantee "through the United Nations" might be acceptable, but by itself it would hardly influence a Swedish decision whether or not to develop nuclear weapons.


The initial U.S. declaration of "strong support" for non-nuclear nations against threats of "nuclear blackmail" was intended to help stabilize the political atmosphere in the wake of Communist China's first nuclear achievement in 1964. Its carefully guarded phrases, reiterated on several occasions and most recently after China's third nuclear test, have not disturbed existing U.S. security arrangements.

Efforts to develop and implement a more definite policy of nuclear containment continue to occupy a place of top priority in the Administration's arms-control policy. As the urgency and dimensions of the problem increase, the United States is trying hard to shift from talk to action.

The particular problem of guarantees to non-nuclear nations is inseparable from the broad issues of collective security in the nuclear age. Neither purely nationalist solutions, alliances for mutual defense nor world-wide international organizations offer satisfactory solutions. And a policy of nuclear containment, to the extent it is successful, simply buys time.

How the present members of the nuclear club conduct themselves will substantially affect the club's ultimate size. But the final decision on whether to expand the membership rests primarily with the non-nuclear nations. As they approach their decisions, the offer of nuclear security guarantees of the kind talked about at present would seem to have at most a marginal bearing. Hopefully, non-nuclear nations will be influenced more by considerations of cost-effectiveness than national prestige. Perhaps a rational assessment of the net gain in security to be derived from acquiring a national nuclear-weapons capability will be the strongest deterrent to proliferation. In the long run, the world will avoid the multiplication of centers of nuclear decision-making and the creation of many-sided balances of nuclear terror only if there is a profound change in attitudes toward the role of force in pursuit of national interests, and if there is a redefinition of those interests in relation to a larger community of nations.

[i] Department of State Bulletin, November 2, 1964, p. 610-614.

[ii] Thus both U.S. and Soviet proposals for a non-proliferation treaty permit any party to withdraw "if it decides that extraordinary events, related to the subject matter of this Treaty, have jeopardized the supreme interest of its country/' This provision also appears in Article IV of the Nuclear Test Ban Treaty.

[iii] In a message to the Eighteen Nation Disarmament Conference on January 27, 1966, the President restated his policy in this respect, along with a general plea "to strengthen United Nations and other international security arrangements." Department of State Bulletin, February 21, 1966, p. 263. The pledge was again reiterated by the State Department in a statement of May 9, 1966, following Communist China's third nuclear blast.

[iv] There are, of course, other countries which have an even greater potential nuclear-weapons capability than those listed, in particular the Federal Republic of Germany, Canada and Japan. However, each of these countries is already under the U.S. security umbrella. Therefore, the relationship between a specific nuclear security guarantee and the incentives at work in these countries to acquire nuclear weapons will not be considered in this article.

You are reading a free article.

Subscribe to Foreign Affairs to get unlimited access.

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