THE current debate on arms control and disarmament puts great stress on the problem of how to detect violations of whatever agreements may be reached. To this end inspection schemes and instruments for detection are developed, their capabilities and limitations discussed, and efforts made to test and improve them. Indeed, the technical question of detection dominates not only the domestic debate but also the international disarmament negotiations.

Yet detecting violations is not enough. What counts are the political and military consequences of a violation once it has been detected, since these alone will determine whether or not the violator stands to gain in the end. In entering into an arms-control agreement, we must know not only that we are technically capable of detecting a violation but also that we or the rest of the world will be politically, legally and militarily in a position to react effectively if a violation is discovered. If we focus all our attention on the technicalities of how to detect a violation, we are in danger of assuming that our reactions and sanctions will be adequate.

A potential violator of an arms-control agreement will not be deterred simply by the risk that his action may be discovered. What will deter him will be the fear that what he gains from the violation will be outweighed by the loss he may suffer from the victim's reaction to it. In other words, even if we can develop an inspection system that makes the probability of detection very high, a nation contemplating a violation will not be deterred if it thinks it can discourage, circumvent or absorb our reaction.

We have learned (almost too late, in the case of the nuclear test ban) that an opponent may thwart our detection techniques by evasive techniques of his own. We should also realize that he may thwart the consequences of detection--which we count on to deter violations--by military or political stratagems. We must study, therefore, not only what our opponent may do to avoid detection, but also what he may do to escape the penalty of being detected.

Let us discuss the question of what may happen when an evasion is detected under four general headings: (1) the reaction of world opinion; (2) the political reaction by the injured country; (3) various military measures that the injured country could undertake in an effort to restore the situation that would have existed without an arms-control agreement; and (4) military and political measures that would go beyond this "restoration."


World opinion, it is sometimes argued, will help to enforce disarmament agreements. World opinion supposedly will turn against the violator, provided he is discovered and "convicted" in an internationally accepted forum. He will lose prestige and influence in the uncommitted countries. In addition, various world-wide political reactions are expected to work to his disadvantage.

"World opinion" is such an amorphous concept that one finds it difficult to determine just how it can injure a violator of arms-control agreements. Speeches or resolutions in the United Nations, or critical editorials in the world press, are not likely to hurt him very much. One reason world opinion is so impotent is that its memory is so short. If the world's reaction cannot be translated immediately into substantive political or military changes damaging to the violator, it will lose all force.

The Soviet suppression of the Hungarian revolution illustrates the point. This gave an exceptionally violent shock to world opinion--in fact, more violent than many possible violations of arms-control agreements are likely to be. This is particularly true since evidence of a violation might often be equivocal and involve technicalities hard for the public to understand. Some of the most cherished beliefs of the West and also of the uncommitted countries were flouted in Hungary: a popular revolt against a dictatorial régime in a small nation was crushed from outside by a large power. Agreements were broken in the most flagrant fashion. One was the promise given by the Sovietinstalled Kadar Government to the Jugoslav Government not to take punitive action against Imré Nagy when he left sanctuary in the Jugoslav Embassy. Another was the invitation extended to General Maléter and other delegates of the legal Hungarian Government to negotiate the withdrawal of Soviet forces--a trap to catch and execute them. Yet if one tries to list the penalties that world opinion imposed on the Soviet Union and the Kadar Government for these violations of its most sacred norms and of several important articles of the United Nations Charter, there is very little to record. There was a slight loss in the strength of Communist Parties in Western Europe (confined mostly to intellectuals on the fringe of the Party), but the loss is no longer noticeable. The strain on President Tito's relations with Moscow and the strengthening of NATO ties (particularly with Iceland) were largely ephemeral. Kadar has not been officially recognized as Hungary's legal representative in the United Nations; but he has been sitting in the General Assembly.

Other recent events have aroused world opinion, such as the Peking régime's violent repressions in Tibet and its violations of the Indian border. Yet in February 1960, only a few months after indignation in India had reached its peak, the Communists increased their vote in Kerala from 35 to 43 percent. And many of Communist China's neighbors continued to favor her admission to the United Nations.

Perhaps significantly, when Khrushchev discussed the nuclear test ban before the Supreme Soviet in January 1960, he chose to emphasize the reaction of world opinion as a deterrent to disarmament violations. By arguing that it was a sufficient deterrent, he tried, in effect, to brush aside the problems of inspection and control. But even if one assumed that the reaction of world opinion constituted an adequate sanction--an assumption challenged above--inspection would still be essential. A violator who does not risk being detected obviously does not need to fear world opinion. In any case, the West has paid insufficient attention to the stratagems which a detected violator can pursue to avoid or mitigate whatever action an aroused world opinion might take.

Many devices are available for this purpose. Thus the violator can frustrate the international inspection system and prevent it from reaching an official finding (study of Communist obstruction of inspection in North Korea reveals a large bag of such tricks). Or he can blame the other side for having violated the agreement first, and thus confuse the issue, or even generate an adverse political reaction against the injured party. Or he can accuse the other side of fabricating the evidence as a pretext for breaking the agreement or for covering up some other misdeed. Or he can assert that the agreement is obsolete in view of what he claims are changed political or military conditions and denounce it unilaterally prior to the intended violation (this would be analogous to the Soviet declaration that the Four-Power Agreements on Berlin were no longer valid). Finally, if some unfavorable reaction in world opinion is unavoidable, it may turn out that the violators "will cover themselves with shame"--as Khrushchev argued when he spoke about the nuclear test ban: "If some side violates the assumed commitments, the initiators of this violation will cover themselves with shame, they will be branded by all the peoples of the world." Yet, six weeks before making this assertion that a nuclear test ban would be enforced by world opinion, Khrushchev had this to recommend: "International reactionary circles are still trying to discuss the so-called 'Hungarian question' in the United Nations. Let them keep it as a souvenir if this consoles them."

Not only may the violator be contemptuous of world opinion, but he also may justify his acts on the grounds that they are demanded by the welfare of "the people" or by History--History being his conception of a superior morality that takes precedence over world opinion. "Had we not helped you," Khrushchev told the Hungarian Communists, "we would have been called stupid, and History would not have forgiven us this stupidity."


To be effective, a sanction must be applied as a result of governmental decisions by the injured countries. In democratic countries, government decisions are influenced by active public opinion, or, more precisely, by the conception of public opinion held by the government leaders. In these circumstances, democratic governments might experience serious political difficulties in reacting effectively to a detected evasion:

(1) The injured government must acknowledge the fact that there has been a violation. If the violation is open and well-publicized, no difficulty exists. But if evidence of the violation is equivocal or based on secret intelligence, the government may be reluctant to acknowledge the evasion or feel unsure of its ability to convince public opinion. For example, an admission that the control agreement had failed might be exploited at home by the political opposition, particularly if the agreement had been made originally by the party in power. In such a situation some decision-makers may favor an interpretation which casts doubt on the intelligence data relating to evasion or which belittles the importance of the evasion. Responsible decision-makers seldom distort evidence deliberately.[i] But the interpretation of complicated information is often a matter of judgment and discretion; hence subtle biases may decide the issue. Responsible officials would be particularly disinclined to accept equivocal evidence about an evasion of a disarmament agreement if they had previously been forced to defend the agreement against partisan charges that it might be violated. Yet a democratic government could institute only minor penalties against an evasion without informing legislative bodies and the public about the exact situation and explaining the need for drastic retaliatory or corrective measures.

(2) The injured government must be willing to increase military expenditures and to offend pacifist feelings. Now the reaction to a localized or minor violation need not disturb the defense budget appreciably (the new military equipment needed to counteract the North Korean violations of the rearmament clause was not a heavy burden); but the breaking of a major disarmament agreement will almost certainly require new military measures, perhaps a full-scale program of rearmament. The decision to react firmly and regardless of expense will be a hard one. Public opinion may not approve, especially if the evasion occurred gradually or if it merely consists of a resumption of some activity that had been discontinued--such as testing. If knowledge of the evasion is based exclusively on clandestine intelligence sources that cannot be revealed, the opponent's denial may find receptive ears among domestic opposition groups.

We have already questioned the effectiveness of world opinion as a sanction against arms-control evasions. It is ironic that it may be domestic public opinion--or rather the government's conception of it--that actually prevents effective sanctions being taken. The classic instance of this, and one that may have been a contributing cause of World War II, was England's reluctance to rearm in response to Hitler's violations of the Versailles rearmament restrictions. With what he called "an appalling frankness," Prime Minister Baldwin explained in 1936 why his own government had been unable to react:

You will remember at that time [1932-33] the Disarmament Conference was sitting in Geneva. You will remember at that time there was probably a stronger pacifist feeling running through this country than at any time since the war. You will remember the election at Fulham in the autumn of 1933, when a seat which the National Government held was lost by about 7,000 votes on no issue but the pacifist. . . . I asked myself what chance was there . . . within the next year or two of that feeling being so changed that the country would give a mandate for rearmament? Supposing I had gone to the country and said that Germany was rearming and that we must rearm, does anybody think that this pacific democracy would have rallied to that cry at that moment? I cannot think of anything that would have made the loss of the election from my point of view more certain.[ii]

(3) The injured government must accept the new risks created by its reaction to the violation. It may see more than the domestic difficulties involved. For example, it may have embarked on long-range policies which seem more promising and important than counteracting an accomplished evasion, and it may hesitate to jeopardize them.[iii]

It has been argued that all countries will be deterred from violating a major arms-control agreement in present circumstances because to do so would set off an unrestricted arms race that would eventually lead to disaster for the guilty as well as the innocent.[iv] But this is an assumption which may not be shared by a country set on violating the agreement. Its leaders may reason that the very prospect of an unrestricted arms race might itself inhibit the injured party from reacting to the violation. And in fact the injured party might feel it safer to write off the violation as a loss rather than risk new dangers by a policy of rearmament--especially if it now finds itself in a weaker military position as a result of having complied with the agreement.

This dilemma is most serious. For example, the nuclear testban conference adopted an article on March 19, 1959, upon the insistence of the United States and the United Kingdom, affirming a country's "inherent right" to withdraw from the treaty if its provisions, "including those providing for the timely installation and effective operation of the control system," are not being fulfilled. This article might be of cardinal importance in connection with China's accession to the test-ban treaty, because part of the control system would have to be installed in China. But would it give the Western powers much leverage against Chinese obstructionism? In the absence of a known instance of illegal testing, would the West be willing to withdraw from a test-ban treaty with the Soviet Union, resume testing and risk accelerating the arms race merely because the "timely installation" of the control system was being prevented by China?

(4) The injured government may have to reach agreement with allies before it can react. All disarmament agreements of current interest involve the United States with one or more of its allies. It is usually a difficult task to prepare a joint negotiating position vis-à-vis a Communist opponent. Agreeing on a Western response to a violation will raise anew the problem of allied coördination. The stronger and more explicit the reaction proposed, the more difficult it will be to achieve agreement. And all the problems of domestic public opinion and partisan politics discussed above will be evident in the allied nations whose coöperation is required.


The military sanctions against evasion of an arms-control agreement can either be confined to measures that restore the situation that would have existed without the agreement or they can go further. Let us call the former "restorative measures." If the violator resumes testing, the injured country will do likewise; if the violator reoccupies his part of a neutralized zone, the other will move back into his; and if the violator rearms, his opponent will rearm to the same extent.

The problem of deterring violations has often been oversimplified by assuming that a detected evasion would automatically be taken care of by the cancellation of the agreement and the application of such "restorative measures." But three conditions have to be met if "restorative measures" by themselves are to be an adequate deterrent:

(1) The potential violator must fear the risk of being detected.

(2) He must also fear that a detected violation will cause an unwanted response by the injured country.

(3) He must not expect a violation to bring him an irrevocable advantage that would outweigh whatever gain he derives from abiding by the agreement.

The importance of the first condition is fully recognized. The second condition depends on the political factors we have just discussed. Both these conditions are needed for deterring an evasion by any type of sanctions, whether "restorative" or "punitive." Here we are interested in the third condition, because if it is not met, "restorative measures" alone are inadequate.

This third condition is not met, for example, if an agreement comprises several arms-control measures in such a way that the separate measures, taken individually, favor either one side or the other. The agreement remains in the interest of both parties only if all measures are observed. Violation of a part of it cannot be deterred by the threat of "restorative measures" confined only to this particular part. Additional sanctions are required. Otherwise the violator can break just those control measures that are not to his advantage. He will stand to gain if his violation remains undiscovered or ignored; and he will also gain if the violated part of the agreement is cancelled, because the residual agreement will then be more to his advantage.

This is precisely what happened with the Korean armistice. The clause prohibiting the introduction of new military equipment was violated by the Communists from the first day, but cancellation of this clause by the United Nations Command did not come until four years later. So the Communists gained on the first count. They also gained on the second count (after the United Nations eventually instituted "restorative measures"), because the residual armistice agreement was more favorable to them than the original agreement. (It was they who had been primarily constrained by the cancelled rearmament clause.)

It might be argued that an arms-control measure can survive only if all its separable components are equally in the interest of both parties. If this argument is true, the future for disarmament agreements is bleak. It is hard enough to arrive at over-all agreements that will not, over time, seem disadvantageous to one side or the other. But individual components of an agreement are inevitably of unequal value to opposing nations. For example, in addition to the Korean armistice, several of the current proposals for disengagement zones are composed of very unequal provisions.

There are other situations where the threat of "restorative measures" would be insufficient to deter an evasion. The violator may gain an irrevocable technological lead or an irreversible strategic advantage. As has often been pointed out, if American and Soviet troops were withdrawn from Western and Eastern Europe, the United States might find it difficult or impossible to return in the event that Soviet troops moved back in. Western alliance arrangements might have lapsed, the American troops might have been demobilized, and in any case they would have to be transported a greater distance--not to mention the American public's unwillingness to send "the boys" back overseas, particularly under a threat of nuclear war.

To sum up, "restorative measures" will not deter a nation contemplating a violation of a disarmament agreement in those situations where our third condition is not met, namely, when the violator expects to gain less from abiding by the agreement than from abandoning it. Indeed, a potential violator might enter into agreements solely in order to seek gains by violating them. He would calculate that there would always be a chance of his escaping detection or that "restorative measures" might be delayed or frustrated for political reasons. And if he lost out on these chances, a mere return to the status quo would leave him no worse off than before he entered into the agreement. The violator, in fact, would be playing a profitable game: "Heads you lose, tails we're even."


Where the threat of "restorative measures" is not enough to deter evasions, additional penalties are required. But to deter a would-be violator effectively they must be credible.

By far the most important and practical penalty would be a general increase in the military effort, going beyond what would be required to restore the pre-agreement situation. (A threat to start a war would not be equally credible and would therefore be less effective.) Suppose the aggrieved nation increases its defense budget by $20 billion. (As a result of the North Korean aggression, the United States increased its national security expenditures from $13 billion to $52 billion.) If the violator does not follow suit, he will become relatively weaker than he was before breaking the disarmament agreement. If he does follow suit, he would, in effect, be "fined" the equivalent of $20 billion, though of course both sides would bear this burden.

The injured country may be able to step up its defense effort in ways that do not require a large increase in the budget and still impose significant penalties on the violator: for example, by changing the deployment and readiness of weapons, or by resuming military activities that were voluntarily limited beforehand. However, in doing this the injured party must be prepared to run the risk that such a "punitive" increase in its defense effort will renew or accelerate the arms race. Actually, the violator may wish to avoid an arms race with so determined an opponent; he may be unwilling or unable to pay his full "fine" and have to accept a loss in relative military strength.

Those who wish to prevent the violation of arms-control agreements must deter potential violators by their evident determination to make a double sacrifice. In the event a violation occurs they must be ready to assume a greater economic burden for defense, and they must risk a step-up in military competition. The willingness to make such sacrifices involves less, however, than would be required to deter limited aggression. To do that successfully a country must be willing not only to accept increased defense costs if deterrence fails, but also to suffer casualties and face the risk that the limited conflict may expand.

Political sanctions are likely to be less effective than an increased defense effort, although they may play an important complementary role. What they might be is difficult to predict in the abstract. If the potential violator is cautious, this uncertainty may help to deter him; if he is adventurous, like Hitler, he will gamble on his ability to meet and overcome the political reaction.

The remaining question is how to make the penalties of evasion seem more inevitable and severe and the gains more dubious. Parliamentary governments are more likely to take strong action against a violation if they are supported by public opinion. The evidence of violation must therefore be such as to impress the public as authoritative and impartial. A finding by an international organization will be influential in this regard, especially with public opinion outside the countries directly affected. An international body, however, has many weaknesses that can be exploited by a violator. Ideally, one would want the best of both worlds: the greater authenticity and dramatic impact that an international inspectorate provides, and the flexibility and versatility of national intelligence systems. One should at any rate avoid entering into arms-control arrangements that are administratively closed to intelligence information. The current draft treaty for the nuclear test ban sets up a rigidly confined scheme from which intelligence information is essentially excluded.[v] The Antarctica treaty, on the other hand, provides for complete freedom of inspection by anyone without any international mechanism (except suggestions for arbitration in the event of "disputes").

The deterrence of evasions could also be strengthened if parliamentary governments took steps to simplify and speed up their decision-making procedures. The United States Government, for example, has sometimes adopted enabling legislation to facilitate quick Presidential action, in order to disabuse a potential aggressor of the idea that partisan conflict or public quarreling about constitutional limitations or the issue at hand might inhibit an effective response. The Formosa resolution of 1955 is an example of Congressional authorization for the President to take action on the basis of his finding alone. The United Nations Participation Act of 1945 authorizes the President to act upon a decision by an international body, the United Nations Security Council. To strengthen a disarmament agreement both types of authorizations might be useful, and a good time to enact the appropriate enabling legislation would be when Congress ratifies an arms-control treaty.

The power and influence of the legislative branch of the government might be brought to bear in other ways so as to increase the likelihood that the reaction to an evasion would be prompt and strong. Special parliamentary committees might assume an explicit responsibility for all arms-control agreements, and stand ready to mobilize legislative support for any necessary response to some breach of a treaty. The Joint Committee on Atomic Energy, which has privileged access to classified information and is on intimate terms with the executive, offers appropriate administrative precedents. Thus Congress might create a "Joint Committee on the Observance of Arms Controls" to demonstrate its determination to make arms-control agreements succeed.

An effective response will often require coördination and agreement among allies. The difficulties which this involves might be lessened by making arrangements in advance for joint action. First, in order to ensure agreement as to the fact of evasion, all evidence could be evaluated by an inter-allied agency permanently set up for this purpose. To minimize considerations of domestic politics, it should not have the responsibility for recommending any action. The next step by the allied governments might be a relatively minor one, on which agreement could easily be reached, namely to give publicity to the committee's findings; for example, they might forward a report on the evasion to the United Nations or to an international control system provided by the disarmament treaty. From this point on, the allied governments having jointly held and publicized their interpretation of the violation would feel under more compulsion to reach agreement on the effective sanctions needed.

It may be argued that allied governments cannot be "tricked" into such a procedure, because if they are opposed to or afraid of vigorous action they will avoid taking the first step: the evaluation of evidence. This argument would have force if the procedure were to be determined after the violation had been charged. But it is proposed here that by a preparatory agreement the Allies shall firmly establish the procedure at the time they conclude the arms-control treaty, when they are still fairly confident that the other side will adhere to it, hence are less opposed to a firm commitment for joint action against what seems a remote contingency.

All these political measures must be planned to the accompaniment of whatever military preparations will be necessary to deal with violations of the agreement. For instance, under an agreement that prohibits only the testing of certain weapons, both sides will remain free to continue research and development. The country that is determined to abide by the agreement cannot afford to neglect this research without opening the way for a potential violator to gain and then exploit a technological lead. Unless the public understands this fact, parliamentary governments will be handicapped in maintaining a research effort for weapons whose testing has been prohibited. The same problem would also arise under an agreement which does not prohibit the development of a weapon but does prevent the deployment of it--for example, a ban on placing weapons of mass destruction in orbit.

A program to deter evasion of arms-control agreements, like the one suggested here, does raise some additional problems for which an analogy can be found in the strategy of deterrence against nuclear attack. First, there is the problem of carrying out a threat if deterrence fails, that is, of imposing sanctions in the event of evasion or of retaliating in the event of attack. An advance commitment to carry out the threat is rational and necessary for a policy aimed at deterrence; but carrying out the threat after deterrence has failed may be undesirable or even irrational. Second, a policy of deterrence has to cope with accidental violations of the agreement, just as a policy of deterrence against nuclear attack has to control the risk of accidental war. In the former case, both sides will wish to correct the unintended violation and preserve the agreement; in the latter, both will want to avoid or correct an "accident" before it leads to full exchanges of violence. Third, there is some resemblance between the advantage of a first strike in mutual deterrence against nuclear attack and the advantage of gaining time through an evasion in certain arms-control agreements. None of these analogies is exact, of course. But they do suggest that ideas in the literature on deterrence can be as relevant to the prevention of violations in arms-control agreements as they are to the prevention of war.

[i] When Germany violated rearmament restrictions in the 1930s, Winston Churchill suspected that "somewhere between the Intelligence Service and the ministerial chief there has been some watering down or whittling down of the facts." (Speech of May 22, 1935, in "While England Slept." New York: Putnam, 1938, p. 190.) Prime Minister Baldwin's later admission suggested that there might have been something deliberate about this "watering down," and perhaps at the highest level.

[ii] Baldwin's reply to Churchill as quoted in Churchill, "While England Slept." New York: Putnam, 1938, p. 333.

[iii] Churchill was aware of this when he was pleading for a more effective response to Hitler's treaty violations: "Then it is said--and I must give this explanation of the extraordinary fact--that 'we were laboring for disarmament,' and it would have spoiled the disarmament hopes if any overt steps to raise our Air Force had been taken." Speech of May 22, 1935, op. cit., p. 190.

[iv] For example, the Committee on Science and Technology of the Democratic Advisory Council wrote on March 14, 1960, about the nuclear test ban: "A nation which violates such an agreement automatically sets into motion an arms race from which there may never be an end."

[v] Both the Soviet and the Western drafts allow the parties to select their annual quota of on-site inspections freely from among seismic disturbances that meet certain criteria. However, should a violation be unaccompanied by such a disturbance (e.g. a muffled underground test), the violator would not have to permit an inspection and the international control system would remain closed to all other evidence.

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
  • FRED CHARLES IKLE, of the RAND Corporation; author of "The Growth of China's Scientific and Technical Manpower" and other works
  • More By Fred Charles Iklé