(From Thinking About the Unthinkable in the 1980’s, Simon and Schuster, New York, 1984, pp. 23-35; © Hudson Institute)
January 1, 1984
by Herman Kahn
“today the only way to achieve genuine national defense for any nation is for all nations to give up violence together ... if we had begun with Gandhi's law of love we would have arrived at exactly the same arrangement. E. M. Forster told us, "Only connect!" Let us connect. Auden told us, "We must love one another or die." Let us love one another. . . . Christ said, "I come not to judge the world but to save the world." Let us, also, not judge the world but save the world”.
This concept has also been suggested in A Pastoral Letter on War and Peace by the National Conference of Catholic Bishops. Redemption may be an appropriate and correct concern for a church, but it has nothing to do with any policies that the government can—or should—carry out. If there is a "redemption of mankind," it will not occur as a result of a great debate on national security policy or defense. It is, then, the "nonissue" of least relevance to government policy on nuclear war.
2. The control of nuclear weapons should be pursued through the creation of an effective world parliamentary government and/or total worldwide disarmament. A world government of sorts already exists: the UN Security Council. It is definitive on almost any issue on which a majority, including the five great powers (the United States, Soviet Union, China, Great Britain, and France), can agree. History, however, has proven its basic ineffectiveness. The effort of trying to establish a more effective world government would itself involve major problems. In fact, it is very doubtful if the creation of the Security Council, or a successor body, could be negotiated today. The small nations intentionally turned over their power to the great nations of the world in 1945; today they fight fiercely to retain it by voting in blocks, changing allegiances and alliances as it suits their purposes. And there could be no chance of agreement today on who would have the veto power.
But even assuming a new world government were negotiable, how would it be structured? If it followed the "one man-one vote" principle, it would be dominated largely by Asian nations; under "one country-one vote," it would be run largely by the small states; and with "one dollar-one vote" it would result in domination by the United States, Japan, the Soviet Union, and Western Europe. All three options are unacceptable, and a compromise seems almost impossible. Under the circumstances, the Security Council is the best we have, and while a consensus by that body potentially could still have enormous impact (e.g., theoretically it is able to overrule national laws and possibly even the Constitution of the United States), it is unrealistic to expect a major and "binding" UN resolution on nuclear weapons to have much real-world significance.
As for total disarmament, there are almost 50,000 nuclear weapons in the world today; even if they were banned, not all would be destroyed. And even if they were destroyed, there is still a large amount of weapons-grade uranium and plutonium available, plus the knowledge of how to turn these materials into nuclear devices. It would not be acceptable to have a disarmament "solution" that allowed those with hidden weapons or weapons-grade material to gain an extraordinary advantage over the rest of the world.
3. Even if it cannot be total, the goal should be disarmament rather than arms control. The objective of nuclear-weapons policy should not be solely to decrease the number of weapons in the world, but to make the world safer—which is not necessarily the same thing. World War I broke out largely because of an arms race, and World War II because of the lack of an arms race. Similarly, many scenarios for the outbreak of nuclear war which are now "implausible" would become "not implausible" or possibly even "plausible" if the existing U.S. and Soviet nuclear arsenals were reduced to very small, less intimidating, and probably more vulnerable forces.
Developments that contribute to a safer world need not be the result of negotiations to reduce the number of nuclear weapons. For example, between 1967 and 1983, the number of weapons in the U.S. nuclear stockpile actually declined by 30 percent. The aggregate mega-tonnage in the stockpile declined by 75 percent between 1960 and 1983. These changes are attributable more to narrow military considerations (e.g., improvements in delivery accuracies diminished the utility of multi-megaton warheads in many missions) than to arms-control influences. Nonetheless, the drop in the number of weapons and total explosive yield did lessen some of the dangers associated with nuclear war. Appropriate arms control could increase the trend toward decreased megatonnage and even toward fewer weapons, but unwise disarmament could set it back.
4. There should be a total nuclear freeze. A total nuclear freeze is counterproductive—especially now, when technology is rapidly changing and the Soviets have some important strategic advantages. An effective and verifiable freeze would worsen the U.S. position by "institutionalizing" apparent Soviet military superiority, and no freeze would be reliably verifiable, despite many claims to the contrary. More important, a freeze would prevent agreement on better arms-control measures by eliminating the incentive for useful negotiations and for the development of beneficial technology.
In fact, the most important reason for rejecting a freeze is that much of the weapons technology ahead is, relatively speaking, beneficial. Many people seem to believe that any change in weapons technology has to be for the worse, but that is demonstrably untrue. For example, almost all military analysts agree that the change from bombers to missiles in the early and mid-1960s made the world safer, since at that time missiles were much less vulnerable and accident-prone than the bombers. In addition, a freeze would preclude many adjustments and refinements that could make existing systems and forces safer. Nuclear-arms reductions or trade-offs are unlikely to take place under a freeze, but if our objective is to make the world safer, we must have the option of increasing, decreasing, or changing forces to achieve this goal.
(…) As a political maneuver, the call for a nuclear freeze has turned out to be an effective way of telling national leaders in the West that something must be done to allay public fears of nuclear war. But it has also been counterproductive by sending (…) the message that they can gain more by manipulating Western public opinion than by making genuine attempts at arms control. As a national policy, however, there is simply no case for a freeze.
5. Deterrence must be made 100 percent reliable. This is a nonissue simply because there is no way to make any complex human (let alone technical) system 100 percent effective.
6. Deterrence must fail eventually, and probably will fail totally. There is no way we can ignore this possibility. But many doomsayers argue that since it might fail, it eventually has to fail. This is technically incorrect, but not entirely unreasonable (unless it assumes that the failure must be total). My guess is that nuclear weapons will be used sometime in the next hundred years, but that their use is much more likely to be small and limited than widespread and unconstrained. Deterrence would then have failed—but not totally. This is why the following points are so important.
7. Useful "damage limitation" in a nuclear war is infeasible; or
8. One can achieve totally reliable damage limitation. If counterforce attacks and strategic defenses could reduce damage in a nuclear war to, perhaps, 10 million to 40 million deaths instead of the 50 million to 100 million that might occur in their absence, and if various other measures could improve the effectiveness and rapidity of recovery from a nuclear war, then they would be worth the effort. ("Counterforce" attacks are strikes directed against the enemy's military assets (e.g., land-based missiles, bombers, command-and-control facilities). "Strategic defenses" are systems providing protection against nuclear attack. Ballistic missile defense and air defense systems are "active" defenses; civil defense measures are "passive" defenses.) Some people believe that the development of such capabilities would be reckless because they would encourage U.S. complacency about nuclear war and U.S. risk taking in crises, and/or excessive Soviet fears which would lead them to preempt a U.S. strike. Others believe systems and operations intended for damage limitation must be made foolproof to be worthwhile. There is no such thing here as total reliability, but if appropriate efforts could mitigate the effects of a nuclear war, then perhaps even an unprecedented catastrophe would not be a total disaster. There are many contexts in which even marginally effective damage-limitation programs might be very effective. The usual discussion about counterforce wars and strategic defenses does not recognize this in-between position, i.e., that some reduction is better than none and that efforts to limit damage will not produce a material increase in the likelihood of nuclear war. I would argue that it is immoral for a nation not to take at least the relatively simple and inexpensive measures it can to achieve an "improved war outcome" in the event that deterrence fails.
9. A nuclear war can be reliably limited; or 10. There is no possibility of a limited war. No one can guarantee that either of these predictions would accurately describe an actual nuclear war. In some circumstances, some kinds of limited nuclear war are clearly possible. There are very large and very clear "firebreaks" between nuclear and conventional war. In nuclear war, the primary firebreak might be "no attacks on the homeland," the next might be "no attacks on cities," and so on. Both sides may choose to observe these firebreaks, though no one can absolutely guarantee they will. But one also cannot be sure they will be totally flouted.
11. There can be no victory in nuclear war (i.e., "nobody wins a suicide pact"), or
12. Either the United States or the Soviet Union could rely on victory.
These issues have become a morality test: to say that "a nuclear war might be limited" or that "there might be a victor in a nuclear war" is to label oneself as a nuclear war hawk (one who seeks to start a nuclear war) or a defender of similarly monstrous positions. To me, it is outrageous to make a morality test out of a realistic and important observation—and both of the quoted remarks are realistic and important. It is incorrect to say that victory in nuclear war is impossible. It is especially possible if either side, or both, have low levels of nuclear forces that are vulnerable to destruction through creative or clever enemy tactics. Unfortunately, it is even quite conceivable at current levels of nuclear armament—but neither nation is likely to choose to go to war just because it has developed some ingenious war plan that might work; the risk is too frightening, and the present governments on both sides are too prudent and cautious. However, that differs from saying that victory could not happen. Indeed, some reasonable facsimile of victory could be achieved even by a nation forced into war. The Soviets, for example, won the war started by Nazi Germany despite suffering 20 million deaths and losing a quarter of their capital stock. It is unwise to judge realistic analyses and preparations as a lust for war. There are no two sides to the nuclear debate: no one is "for" war; everyone is against it—some categorically so and others only to the degree that it does not result in an even less desirable alternative. These twelve "nonissues" are important because many of them are deeply held beliefs and reflect genuine concerns. But they offer no substantive guidance in dealing with nuclear dangers, cannot be translated into constructive programs, and often stand in the way of serious discussions of useful and necessary programs.
Twelve Almost Nonissues
One step removed from these "nonissues" are twelve propositions that are equally popular, widely held, and emotionally defended, but offer almost as little practical guidance on how to make the world safer from the threat of nuclear war. The difference is that while we are comfortable in our belief that the first twelve are truly irrelevant, we are not as sure about these:
1. Nuclear war would result in the destruction of the created order, and/or
2. Nuclear war would result in the destruction of all human life.
There are no respectable objective analyses or calculations to indicate that either of these is likely. But the data and theory are so lacking one cannot be absolutely certain. From a scientific perspective there is some indication that a nuclear war could deplete the earth's ozone layer or, less likely, could bring on a new Ice Age—but there is no suggestion that either the created order or mankind would be destroyed in the process. From a religious perspective these assertions are almost heretical, since only God can fashion or destroy the universe (but some believe that He may choose nuclear war as His means of doing so). As a practical matter, however, these concerns do not offer any useful policy guidelines.
3. The threat of a nuclear war would mean "everybody Red, dead, or neutral."
In Europe, when the only alternatives were presented as "Red" or "dead," the obvious choice was "Red." However, sophisticated Europeans are now formulating the choice to include "everybody neutral." A further revision of the slogan—"everybody Red, dead, neutral, or NATO"—is much more reasonable. Indeed, the purpose of the Atlantic Alliance is to make the last option the most attractive one.
4. Nuclear weapons are intrinsically immoral.
Nuclear war is such an emotional subject that many people see the weapons themselves as the common enemy of humanity. Nuclear weapons are intrinsically neither moral nor immoral, though they are more prone to immoral use than most weapons. But they can be used to accomplish moral objectives and can do this in ways that are morally acceptable. The most obvious and important way is to use them or their availability to deter others from using nuclear weapons. The second—of much lower, but still significant priority—is to use them to help limit the damage (human, social, political, economic, and military) that could occur if deterrence fails. Anything that reduces war-related destruction should not be considered altogether immoral.
On the other hand, the position that nuclear weapons are "just another" weapon and therefore as moral as any other is not accurate either. I would judge them as moral when used solely to balance, deter, or correct for the possession or use of nuclear weapons by others, and immoral when deliberately used against civilians, for positive gains, or to save money and effort on nonnuclear military alternatives. This rule precludes the first use of nuclear weapons to defend Western Europe (current NATO policy), simply to avoid the more complicated capabilities, plans, and costs required for improved conventional defenses. It likewise stamps as immoral the targeting of enemy cities simply to avoid the problems of counterforce weapons and attack planning (countercity targeting is endorsed by many supporters of the nuclear freeze). It is unacceptable, in terms of national security, to make nonuse of nuclear weapons the highest national priority to which all other considerations must be subordinated. It is immoral from almost any point of view to refuse to defend yourself and others from very grave and terrible threats, even as there are limits to the means that can be used in such defense.
5. Expenditures for strategic nuclear forces are bankrupting the United States and the Soviet Union.
U.S. strategic expenditures are now less than one-tenth of 1 percent of the gross national product (and have been less than 2 percent for almost two decades), while Soviet expenditures are probably about 2 percent of the Soviet GNP. The cost of U.S. strategic forces represents only about 10 percent of total U.S. military expenditures; roughly 15-20 percent of the Soviets' defense budget is allocated to strategic forces. So while reducing the cost of arms is a desirable goal, it is simply not an overriding priority in terms of bringing about a dramatic turnaround in U.S. fiscal solvency.
6. Defense expenditures should be reallocated to the poor.
A healthy and fully functioning society must allocate its resources among a variety of competing interests, all of which are more or less valid but none of which should take precedence over national security. This is not an argument for papering over instances of wasteful or excessive spending in the Department of Defense, but simply a recognition that national security programs have a legitimate and fundamental claim on the nation's resources. Furthermore, even if defense expenditures were cut, the savings would be divided along the lines of current federal fiscal allocations. Only those who are ideologically opposed to military programs think of the defense budget as the first and best place to get resources for social welfare needs.
7. "War-fighting" measures are simultaneously too ineffective and too effective, and
8. "Deterrence only" is the least undesirable policy; any "war-fighting" policy is fatally flawed.
As used by most strategists, "deterrence only" implies an all-or-nothing strategy: first, a very strong belief that deterrence can and must be made to work for the foreseeable future, and second, that if it fails we are all doomed—because the cities on both sides would be deliberately and automatically destroyed at the outset of a war. The policy of "mutual assured destruction" (MAD) is the clearest form of "deterrence only."
It is often believed that any attempt to mitigate the damage a nuclear war would cause, or even to reinforce deterrence with "war-fighting" capabilities, is "destabilizing" or a waste of resources. For its supporters, a "deterrence-only" policy offers a simple, clear and, by itself, unaggressive "solution" to the threat of nuclear war. But the term "war fighting" does not mean one wants to fight a war; the position simply recognizes that deterrence can fail and says that it is prudent to have programs both to reinforce deterrence and to alleviate such failure should it occur. As I will argue later in more detail, these efforts are likely to be effective enough to be worthwhile (i.e., they could have a positive effect on the course and outcome of a war), but would not be likely to cause or contribute to an appreciable increase in the risk of war.
9. No significant weakening of deterrence is acceptable.
In most situations, some small "destabilizing" or weakening of deterrence is not likely to be significant. In some instances, an intentional weakening might be an acceptable part of a trade-off for other gains. For example, a minor and relatively insignificant decrease in deterrence would be justified if it would bring about an enormous reduction in the damage done if deterrence failed. Deterrence itself is not a preeminent value; the primary values are safety and morality.
10. If retention of nuclear weapons is unavoidable, then "simplistic stability" is preferable to "multistability."
If we recognize nuclear deterrence as a means toward attaining a safer overall security environment, then simplistic stability (stability only against a first strike) should not be the sole objective of strategic forces. In fact, it is not the only mission of U.S. forces: their purpose is multistability—i.e., to deter serious provocations against the United States (and its allies), as well as to prevent an actual first strike. Multistable deterrence imposes much more stringent—and necessary—requirements on our strategic forces than simple deterrence. While the U.S. can no longer have the kind of extended deterrence that covers all areas where provocation is possible, it still needs a "not-completely-incredible" ability to punish an opponent for a variety of (extreme) provocations. (Multistable deterrence is described at greater length in chapter 5.)
11. Nuclear war would be fought mainly to achieve positive gains. Both the Soviet Union and the United States are essentially very prudent and cautious—the Soviets probably even more than the Americans—and both are unlikely to risk a nuclear war for positive gains (i.e., to fulfill or advance national or personal ambitions), even if they think they can do it successfully. Neither the United States nor the Soviet Union believes in "war by calculation" (i.e., planned and carried out as scheduled), but rather by miscalculation (plans always go awry). Calculated wars have not worked in either country's history, leading both to the realization that even theoretically sound plans are almost certain to go astray. Consequently, only desperation could persuade the leadership of either country to initiate a nuclear war. The belief that war has become virtually obsolete as an instrument for advancing positive ambitions (as opposed to averting disaster) is not completely the result of nuclear deterrence. It is also because, increasingly, the most effective and reliable way to achieve affluence, power, influence, status, and prestige is through economic development, and not through territorial or political gains. While war between two major powers may still be an important way for one or both of them to avoid what appear to be more severe alternatives (e.g., for the Soviet Union, the loss of political control), primarily it has become largely dysfunctional. And yet, maintaining a credible (or "not incredible") capability to resort to nuclear war remains basic to an adequate defense and foreign policy.
12. Normally, there is an automatic and increasingly dangerous "arms race." This concept is plausible and has sometimes been accurate—for example, the arms race preceding the outbreak of World War I. Post-World War II history has witnessed some automatic increases in armaments. New developments in weapon systems during the 1950s and early 1960s created a situation that was most dangerous, and even conducive to accidental war. In retrospect, we were fortunate that none occurred. However, it appears that in both the United States and the Soviet Union the radical development of military technology since World War II has been a greater engine of weapons development than has any sense of detailed measure/countermeasure competition. A careful study of the development of Soviet and American strategic weapons suggests that in almost all cases weapons were developed in conformity with strategic and technological thinking that took remarkably little account of the other side's real or likely programs and other strengths and weaknesses. If each side must more or less commit itself to a weapon system as much as ten to fifteen years before it enters service in any quantity, then there is little chance that the concepts being developed in one country will serve as a guide for the other.
The world became much less dangerous in the 1960s because of improvements in equipment (e.g., submarine-launched ballistic missiles—SLBMs), tactics (e.g., alert procedures), and thinking (e.g., planning for limited counterforce campaigns). But even with advances in U.S. and Soviet weaponry and defense planning, there was no "arms race." For example, from 1963 to 1980, the U.S. defense budget was more or less constant in constant dollars (except for the operational expenses of the Vietnam War), while the Soviet budget increased by 4 or 5 percent a year—about the rate of increase for the Soviet GNP.
More accurate than the "race" metaphor is the observation that if it was a contest at all, the Americans walked while the Soviets trotted. There was no race—but to the extent that there was an arms competition, it was almost entirely on the Soviet side, first to catch up and then to surpass the Americans. The United States barely competed: except for some retrofitting (e.g., equipping ICBMs and SLBMs with multiple warheads), the U.S. defense establishment languished. The present controversy over the expense and morality of "rearming" the United States is a result of two decades of a very lax U.S. defense effort; the controversy could largely have been avoided if a consistent pace and pattern of defense preparedness had been maintained all along. The United States and the world would be much safer today, and the current anxiety-provoking defense program would not be necessary. But even this recent attempt to redress the balance is very different from being involved in an inevitable and increasingly dangerous "arms race." In fact, as discussed later, many of the innovations in defense that we expect over the next decade or two are as likely as not to make the world safer.
The accuracy of these twelve oft-cited assertions is, then, dubious at best. Those who uncritically accept their validity are less likely to help make the world safer from the threat of nuclear war than those who question them. So while we doubt that these "givens" are relevant to government policy, they are less irrelevant than the first set of commonly held assumptions.
Herman Kahn (1922-1983), was the founder of Hudson Institute. View his bio here.
Home | Learn About Hudson | Hudson Scholars | Find an Expert | Support Hudson | Contact Information | Site Map
Policy Centers | Research Areas | Publications & Op-Eds | Hudson Bookstore
Hudson Institute, Inc. 1015 15th Street, N.W. 6th Floor Washington, DC 20005
Phone: 202.974.2400 Fax: 202.974.2410 Email the Webmaster
© Copyright 2013 Hudson Institute, Inc.