Does Deterrence Have a Future?

October 2000

By Lawrence Freedman

For some four decades, deterrence was at the center of U.S. defense policy. Every move was made with reference to its requirements—a centrality that was due to three important presentational features. First, it sounded robust without being reckless. Forces were not being used to compel a change in the status quo, only to contain an enemy. Second, it was hard to think of a better way to make sense of a nuclear inventory. Preparing to use nuclear weapons as if they were conventional seemed criminal, yet the prospect of nuclear war appeared to encourage a welcome caution all around. If, as it seemed, there was no way of getting out of the nuclear age, then deterrence made the best of a bad job. Third, it seemed to work. Exactly how it worked was often difficult to explain, and historians can point to some terrifying moments when catastrophe was just around the corner. But World War III did not happen, and the fact that the superpowers were scared of this war surely had something to do with its failure to materialize.

In the 1960s, the role of nuclear weapons in securing superpower restraint came to be recognized in an almost formulaic way as "mutual assured destruction." So long as each side was confident that it could inflict utter hell on the other and it was understood that both would need to buy new weapons in a precautionary way to sustain such confidence, then a wider political equilibrium was possible. There was, however, an awkward thought at the heart of this concept, which is why its critics seized on the acronym "MAD": if a nuclear war meant an inevitable slide into the ultimate catastrophe, then who would be irrational enough to set it in motion? If the argument was that circumstances could produce irrationality, carrying the risk of the ultimate madness, then how could we be confident that these dangerous circumstances would not arise over incidents that were comparatively trivial in their origins, conflicts marked by confusion rather than unremitting belligerence?

In practice, the political circumstances never quite arose for nuclear employment, although there were times when it seemed close, such as during the early 1960s and early 1980s. One conclusion that might be drawn from the Cold War experience is that deterrence worked because it was not asked to do too much. The East-West conflict became institutionalized and relatively stable over time.

In the post-Cold War world, the demands might have seemed to be even less severe, perhaps even to the point of insignificance. Certainly if deterrence was about nuclear weapons, then it was hard to see any conflicts around which any interests, or at least any Western interests, would be sufficiently at stake to warrant issuing nuclear threats. Even if deterrence was about trying to maintain a sort of stability, then the fluidity and uncertainty of recent international affairs argues against attempts to control matters through threats. Policy has had to be much more reactive and interventionist. Even with potential nuclear threats from relatively weak states that have yet to demonstrate the requisite technological capability, there is a marked lack of confidence in dealing with them through deterrence—hence the proposals for a national missile defense system. So, deterrence has moved from the center of American security policy to its margins.

The possibilities of deterrence should not be ignored. As an approach to security policy, deterrence still has a role to play, although not the role it was granted during the Cold War. Deterrence still helps explain why states, and even non-state actors, fail to act against the interests of others. Actors may be deterred because they have constructed a possible future in which they are worse off. These caution-inducing constructions are often developed without any external help, but occasionally they can be encouraged by various combinations of statements and military deployments, put together as deliberate acts of deterrence. In this way, we can understand deterrence as a feature and consequence of good strategy.

So, at one level deterrence never goes away. Certain options, whole categories of actions, are precluded because of the possible responses of others. Land may be coveted, but it is not grabbed; the unacceptable practices of foreign governments are denounced, but they are left untouched; ideological ambitions are shelved; inconveniences, disruptions, and outrages are tolerated; punches are pulled. Over time, after operations have been delayed and plans shelved, it is forgotten that these operations were ever proposed or that the plans were once taken seriously.

All this may happen without the potential victim of these proposed moves issuing threats or doing much by way of military deployments. It simply results from the sensible application of what should always be the first principle of strategy: anticipate the probable responses of the opponent. This sort of deterrence is far more regular than the sort that captures the most attention from policy-makers and academics, when a determined effort is made to persuade another party from taking action you judge harmful to your interests. Again, even when it takes this more engaged form, the act of deterrence may be no more than a hint here and a quiet word there, sufficient indication that the possibility of the offending act has been noted and preparations are being made. Only on occasion does it become necessary to move to the direct, explicit threats commonly associated with a deterrent posture.

This is a quite different view from that prevalent during the Cold War, when deterrence referred to the whole gamut of policies designed to maintain a degree of stability in U.S.-Soviet relations. The starting point, back in the late 1940s, when the doctrine of containment was first adumbrated, was the need to prevent further westward expansion into Europe by the Soviet Union. Because their power seemed sufficient to stop any aggressor in its tracks, nuclear weapons appeared as the means by which this could be achieved without too much exertion on behalf of the allies.

This rather simple formulation then became complicated by two developments. First, East-West relations became more complex, and in some ways more stable, so the strategic problem came to be posed in terms of managing crises rather than simply blocking aggression. Second, the Soviet Union acquired its own nuclear weapons, so the issue of when and how to initiate nuclear strikes was complicated by the probability of Soviet retaliation in kind. This situation became more complicated still with the arrival of thermonuclear weapons and their introduction in a great variety of forms, from gross city-busters to more subtle munitions designed for battlefield use, capable of being delivered by missiles of ever longer range and improving accuracy.

Strategists began to work out ever fancier forms of targeting designed to disarm an opponent before retaliation was possible or to at least knock him sufficiently off balance so that an advantageous diplomatic settlement could be reached. Their scenarios often managed to combine the most sophisticated technical analysis with the crudest psychological and political presumptions, so their influence on actual policy-making is still hard to discern. The durability of the Cold War, and in particular the ideological divide at its heart, ensured that there was a relatively tight framework within which all questions of foreign policy and force planning were viewed. This framework was institutionalized in the major security organizations so that it became almost beyond reflection. It was confirmed by every NATO communiqué, Pentagon report, and foreign minister's speech, and it was inculcated into generations of officers, diplomats, and politicians.

The sense of the dynamic interaction between the political context and the instruments of power that is at the heart of strategy seeped away because it was only experienced spasmodically and at the margins of the Cold War. Statesmen and generals spent their time speculating and arguing about the implications for deterrence of new technologies or arms control agreements rather than worrying about whether the right military instruments were available for a developing crisis, choosing at times of tension and urgency from a range of imperfect options. Strategic debate became lazy, denied because of the success of deterrence and the challenge of hard decisions. The concept of deterrence was at the heart of these lazy formulas, and as a consequence its meaning came to be casually expanded. It became self-contradictory and confused. The same set of policies and activities, all undertaken in the name of deterrence, were required to contain Soviet power and remove some of the hazards from the superpower relationship, particularly those caused by over-reliance on nuclear threats, while keeping alliances in good order.

For all these reasons, deterrence came to refer not so much to a strategic move that might fit a variety of scenarios but more to the range of policies for managing the military dimension of the Cold War. Given the prominence of nuclear weapons in this dimension, it was assumed that these policies must largely be about nuclear weapons. This covered not only how they might be used in anger, but also how they might be configured to send appropriate diplomatic messages of reassurance to allies and resolve to adversaries, as well as how to keep them under careful control so they would not be launched through technical or political mishaps. Attention had to be given to whether there were ways by which international agreements might bring extra clarity, order, and stability to the process. In short, and apart from occasional flurries of anxiety, deterrence became not so much geared to the urgent avoidance of war as to the preservation of a sort of stability based on the fact that the nuclear age generated sufficient wariness in the breasts of policy-makers. The risks of utter calamity if a drive for military victory went awry were so evident that Cold War leaders did not find it that difficult to avoid a hot war.

Mutual assured destruction made its appearance as previous hopes for the ultimate breakthrough were dashed. A nuclear victory would depend on a surprise attack, catching enemy bombers and missiles in their bases before they could be launched in retaliation. If any did escape, then missile defenses would have to intercept them before they reached their targets. Preventing retaliation soon appeared a hopeless proposition: too many weapons were protected in silos smothered with hardened concrete or out of sight in submarines. Any defenses were likely to be overwhelmed by decoys and multiple warheads. In the final analysis, MAD seemed a safer proposition: if both sides could assure the destruction of the other, neither had any incentive to initiate a nuclear exchange.

As the Cold War drew to a close, the thought that general war was quite easy to deter was captured by the notion of "existential deterrence." This was based on the proposition that deterrence flowed not so much from specific preparations for weapons use and doctrinal pronouncements as from an overall sense that once any superpower war began there could be no knowing what might happen. Deterrence became dependent largely upon the risk perceptions of the deterred. As there was sufficient chance that the outcome of a war would be absolutely awful, it was best not to risk finding out. Deterrence worked because nuclear weapons existed and could be primed for action. The responsibilities of the deterror when it decided on force structure and made policy pronouncements were largely irrelevant— so long as they did not stray too far into the realms of recklessness and foolishness. But they could so stray, and we now know that when the ideas of existential deterrence were being promoted in the United States in the early 1980s by opponents of President Ronald Reagan's nuclear buildup and hawkish strategic assertions, the Russians were becoming sufficiently alarmed to believe that deterrence was breaking down and that a pre-emptive strike might even be imminent.

But there was a growing weariness surrounding even this attenuated form of nuclear deterrence, reflecting moral unease about such dependence upon threats of mass destruction and the nagging fear that something could go terribly wrong, even in the absence of active belligerence by either superpower. The reaction against deterrence argued either for radical disarmament (although unless it was complete, safety would be difficult to guarantee), a technical fix to reduce the nuclear danger, or a combination of the two. President Reagan's Strategic Defense Initiative (SDI), which was the most notorious attempt at a technical fix, was based on the idea that it was better to "protect than avenge." It sought to exploit new technology to catch missiles in their "boost phase"—that is, just as they are launched—or at any point on their trajectory. However, the desirability of the principle behind the initiative was always easier to explain than the feasibility of the technology.

After the Cold War, this weariness with the problems of deterrence was combined with optimism suggesting that while a technical solution had proved elusive, a political fix had been found in the end of the superpower conflict. But the political fix was not total, and so the whole edifice of deterrence could not be scrapped. There was still a need to think about deterrence.

The Warsaw Pact might have disintegrated, the Soviet Union might have collapsed, and the rump Russia might have been facing real economic and social problems, but the Kremlin still presided over a large nuclear arsenal. Moreover, during the 1990s aspects of the political "fix" appeared less firm. Whatever its leaders may have said about still being a great power, Russia was clearly no longer capable of exerting substantial influence beyond its borders. But around its borders, especially in the "near abroad" of the former Soviet Union, there was plenty of trouble.

In addition, Russia was experiencing a growing sense of insecurity as its illusions of great power were stripped away. Somehow, it seemed to Moscow, the West had not played fair: NATO had not only stayed in business while the Warsaw Pact had disbanded but, contrary to what Moscow had been led to believe, it had enlarged to take in Poland, Hungary, and the Czech Republic with the promise of more to come, edging ever closer to the Russian heartland. Many Russians were further shocked by the ruthlessness with which the United States, Britain, and other Western countries went after Iraq and Yugoslavia, even when they allowed that Presidents Saddam Hussein and Slobodan Milosevic had been the architects of their countries' misfortunes.

The relationship with Russia therefore remained uneasy at best and argued for a deterrent posture to be sustained on a sort of care-and-maintenance basis, while at the same time allowing for the nuclear inventory to be cut and reorganized. The more "destabilizing" elements of the strategic nuclear arsenal, in particular MIRVed ICBMs, could be cut back, with submarine-launched missiles, both ballistic and cruise, retaining a central role. Short-range "tactical" weapons appeared obsolete and so were extracted from general purpose forces and soon no longer carried at sea. It no longer seemed appropriate to have a nuclear equivalent to all conventional munitions.

In doctrinal terms, the shift to a lower-key nuclear posture was made easier by the evident conventional superiority of Western forces, a complete reversal of the perceived state of affairs during the Cold War, although Western military technology had been advancing far ahead of the Soviet bloc's for some time. With conventional superiority, there was no need to devise elaborate rationales for a doctrine of nuclear first use or to worry about how it could be credible. NATO declared nuclear use to be a "last resort."

The airwar over Kosovo showed how NATO's aircraft could hit with impunity almost any targets they chose (admittedly also some not chosen) without being caught by enemy air defenses. There was a residual case for deterring other nuclear powers and possibly countries tempted to try chemical and biological warfare, but in all cases deterrence would not necessarily have to take a nuclear form. For example, the real threat in 1991 to the Iraqi leadership was that they would be toppled if chemical weapons were used, though Baghdad later claimed it had taken into account the possibility of U.S. (and Israeli) nuclear use. In public, Bush administration officials had been ambiguous about their nuclear intentions although there was never any actual intention to go nuclear.

The political fix that had substantially reduced the West's dependence upon nuclear deterrence, however, had increased that of the Russian Federation. With their armies in disarray, the conscription system breaking down, and equipment in need of repair and replacement, Russian generals could not but despair at how their once-mighty and much-feared forces had been brought low, so that having failed to subdue Afghanistan, they could barely cope with an internal rebellion in Chechnya. If they looked to the West with anxiety, they could not avoid the conclusion that if their worst fears about NATO's intentions were realized, their best hope was to rely on nuclear deterrence. Presidents Boris Yeltsin and Vladimir Putin accepted security doctrines that appear to reject the "no-first-use" restriction that Soviet doctrine had once embraced. While there may be reasonable doubt as to whether Leonid Brezhnev would have felt bound by promises not to be the first to use nuclear weapons, the position was at least credible because of the strength of his non-nuclear options. With Moscow's non-nuclear options now much reduced, it is hard to dismiss the current Russian posture—that an assault on the most vital security interests of the state might well trigger a nuclear riposte—as just bluster.

At one level, it might be argued that this posture does not really matter. NATO has no intention of launching aggression against Russia. It does not need to be deterred. If deterrence serves as a form of therapy, helping Moscow cope with its phantoms, then why object? Others might argue that it does no harm for those Western politicians who habitually say, "We cannot allow Russia to have a veto," when it comes to pushing ahead with policies that obviously affect Russian interests to be reminded of the dangers of pushing Russia too far. Either way, there is not much that can be done about it. Under current circumstances Russia is not going to abandon its nuclear capability.

There are, however, reasons against too sanguine an assessment of the situation. The first relates to the state of the Russian arsenal. While there is no evidence to suggest that the Russian political and military leadership have neglected their responsibilities in ensuring that the risks of accidents, malfunctions, and security breaches are kept as low as possible, the system is under strain as a result of low resources and poor morale. To an extent, this reinforces the Russian deterrent by providing compelling reasons for the West to show additional restraint in any crisis so as to prevent a disaster from arising out of inadvertence rather than deliberation. The concern is obviously that the disaster could still be triggered by false alarms or some local irrationality at a key facility. Western countries are doing what they can to help improve the reliability of Russian arrangements, but when taken together with the problems of old power stations and decommissioned nuclear submarines, it is hard to be confident that the coming years will not see a nuclear incident of some sort.

The second set of concerns relates to the problem of modernization. It is clear that Russia will have difficulty replacing the more obsolescent parts of its arsenal, and this helps explain its interest in strategic arms reductions. At some point, unless the Americans agree to play along, the Russians may find that they have a larger interest in deterrence but a smaller arsenal to preserve it. At the same time, the need for Moscow to introduce new systems will provide arguments for those in the United States wishing to develop new weapons.

The situation therefore retains some resemblance to the Cold War era, but the activity is on a smaller scale, so we are not likely to see a reversion to the arms racing that took place at the height of U.S.-Soviet confrontation. Nor is there anything approaching the old superpower symmetry. Against any possible strategic measure, there is a deep imbalance in favor of the United States and its allies. Mutual assured destruction may still describe a possible outcome of an all-out war, but at every other level there is no mutuality. At each stage of a conflict, the onus would be on Moscow to up the ante in desperate attempts to avoid defeat.

One area in which there is continuity is that moves made in the name of deterrence are still part and parcel of a wider political relationship. This relationship requires opportunities to be conciliatory, to show understanding, and to engage in active cooperation, as well as opportunities to signal suspicions, insecurities, and a lack of confidence in the future. A good example of this has been the fate of START II in the Russian Duma, where the long, drawn-out process of ratification was regularly delayed as Russian parliamentarians took umbrage at some aspect of Western behavior that frequently had no relevance to the nuclear relationship. Putin's moves earlier this year to ratify START II and the Comprehensive Test Ban Treaty can be seen not only as a statement of the new man's readiness to honor past arms control commitments, but also as a means of putting political pressure on the United States.

The astonishing failure by the Senate to ratify the test ban in 1999—a failure that can only be fully explained by reference to U.S. domestic politics and in particular the prevailing antagonism between the executive and the legislature—was nonetheless taken by the outside world as a sign of American exceptionalism in practice. Congress refused to be bound by the constraints of a treaty that, strictly speaking, appears to work to U.S. advantage and that will in any case be honored in practice as testing is unlikely to resume. At the same time, the rejection indicated that among the American political class nuclear weapons still play a large role in strategic calculations.

This issue was soon overshadowed by a further debate that has been viewed outside of the United States in a similar way—the debate over national missile defense (NMD). It is possible to chart the shifting American attitudes toward deterrence by examining the three great debates in the United States on ballistic missile defense: that of 1965-1972, which ended with the signature of the Anti-Ballistic Missile (ABM) Treaty; that of 1983-1988, which was prompted by President Reagan's Strategic Defense Initiative; and the current debate over NMD, which is still ongoing.

The first great debate was couched throughout in terms of deterrence, with the original Democratic-proposed system (Sentinel) being justified as being inadequate to support a first-strike capability and, ironically in terms of current debates, geared to neutralizing a Chinese missile capability that had yet to materialize. It was replaced by the Republican system (Safeguard), which was justified in terms of protecting the American second-strike capability.

SDI's rationale was anti-deterrence, with Reagan explaining why it was better to protect than avenge. It was not the justification that undermined Reagan's defensive proposals but rather their confused and futuristic quality, especially when it came to demonstrating that the system would actually protect. SDI required an enemy sufficiently threatening to warrant the effort but not substantial enough to overwhelm the proposed system or clever enough to circumvent it. As such an enemy could not be guaranteed, the non-deterrence, anti-offensive weapon rationale led naturally, including in Reagan's mind, to a case for abolishing all offensive weapons. The disarmament solution made more sense than a hardware solution.

NMD has a much more modest objective. The claimed focus is no longer Russia or China but other hostile states that are unable to take on the United States and its allies in regular battle and are therefore seeking their own minimalist form of nuclear deterrence. NMD is posed as a challenge to the growing dependence of weaker states on nuclear deterrence as a counter to the overwhelming conventional strength of the West. There is no claim that it sustains Western deterrent strategies.

Attempts by states such as Iran, Iraq, and North Korea to obtain weapons of mass destruction can be seen as attempts to stop Washington from interfering when they throw their weight about in their neighborhoods. The NMD proposals betray a lack of confidence in Western deterrence of these weak states, which could expect to be completely eliminated if they dared to mount an attack on the United States. There is only a need for NMD if it is assumed that the responsible officials in these countries are crazy—Hitler-in-his-bunker sort of scenarios. The favorite rogue is North Korea, a hopelessly poor country worth far less than the cost of the proposed defensive system, with a regime that is likely to collapse before any shield is completed.

In practice, NMD is likely to be an irrelevance—if it happens at all. By and large, most states with a possible interest in developing weapons of mass destruction and the means of delivering them are prompted by regional concerns rather than some heroic desire to take on the United States directly. This is not to deny that a country may be very grateful for an investment in missile defenses if it stops just one stray or accidental missile or if it convinces a prospective attacker that a missile strike carries too high a risk of failure. Rather than argue that there is any merit in staying vulnerable for the sake of reassuring prospective enemies, opponents of NMD have concentrated on the technical difficulties the program has faced and the implausibility of the scenarios that inform it.

The problem with NMD is that in failing to solve one type of strategic problem (calming the behavior of states formerly known as "rogue" and now just "of concern"), it is likely to aggravate other problems, in particular the already tense relations with Russia and China. Worse, it could provide an illusion of security that, if ever tested, might come tragically apart. As far as Russia goes, NMD causes problems if it leads to the abrogation of the ABM Treaty and possibly also encourages Moscow to fear the coming together of American offensive and defensive systems in a first-strike capability. This in turn might lead to reliance on a launch-on-warning posture, with Russia's shaky command systems being put on a hair trigger. Few would argue that the NMD system as currently proposed, even assuming success, would actually generate that sort of capability. If it stays within definite limits, then it would not challenge the spirit of the treaty (as it would if it offered a nation-wide defense against a Russian retaliatory attack), but it would violate the letter.

The issue here becomes one of whether U.S.-Russian political relations can be sufficiently friendly for the two sides to agree to amendments to the ABM Treaty that would permit NMD but ensure that it stayed within limits that would reassure Russian planners. As Russia actually has much more to fear from the proliferation of missile capabilities than has the United States, there is also the possibility that there could be an agreement to cooperate in establishing defenses. A further incentive to reach some understanding would be in dealing with potential proliferators by other means—political and economic inducements and sanctions. It is very clear, for example, that one of the reasons why the United Nations Special Commission charged with overseeing the disarmament of Iraq ran into the ground was that it had lost the support of key members of the Security Council, which was in itself a consequence of the general souring of relations between the United States and Britain on the one hand, and Russia and China on the other.

China is not a signatory to the ABM Treaty and has no legal grounds for complaint if the Americans amend the treaty with Russian connivance. On the other hand, whereas any conceivable ballistic missile defense might have little chance against a Russian attack, the much smaller Chinese long-range missile force could face difficulties. Furthermore, China has a specific scenario in mind: a clash over Taiwan. If at some stage China felt that Taiwan was slipping away, it might try to coerce Taipei into meeting its demands through a blockade or it might try to settle the matter once and for all by invading. In such a situation, it would be hard to avoid American involvement. It is doubtful that the Chinese would expect to deter U.S. involvement by nuclear threats, but they might at least hope that their nuclear capability would prevent the United States from issuing nuclear threats of its own. A successful NMD program could undermine this hope.

On this basis, there is concern that if the United States pushes ahead, then China will feel that it must accelerate the deployment of new ICBMs. This may happen anyway and would probably not in itself make much difference to American plans and policy. (Many in the United States would also argue that it does no harm at all if China's campaign against Taiwan were to be frustrated by a national missile defense.) The most likely effect of NMD deployment would be to convince Beijing that Washington is indifferent to its security concerns and thus add another reason for maintaining a wary stance when dealing with the West. So long as Russia feels the same way, then the more America acts like a hegemonic power, the more Russia and China are going to find themselves making common cause.

Missile defense programs come and go and the latest scheme may meet the fate of those of the past, but it illustrates a wider point. The traditional deterrence framework now provides an unreliable guide when it comes to assessing nuclear plans. The strategic context is much more complicated and so the impact and meaning of particular developments and deployments will be different from those that might have been suggested by the formalized models of the Cold War period. At the same time, there is substantial continuity. Even during the Cold War, the focus on these models missed the point about so much that was done in the name of nuclear deterrence. So dominant was this concept that everything had to be rationalized in its terms. There were many actions justified in terms of preparations for a grand superpower confrontation that were actually about the management of current alliance and East-West relations.

Now that the concept is no longer dominant, individual diplomatic or military moves have to be justified in their own terms. There is reluctance to give a high salience to nuclear weapons or to claim that only these weapons stand between the United States and some sort of strategic defeat. For most of the time, deterrence is marginal, tangential, or speculative. Yet even so, deterrence remains one feature of any activist state's view of the world, setting boundaries to its aspirations. As a strategy, it provides one option among many, possibly appropriate in particular circumstances. Deterrence helps explain why certain strategic moves are assessed to be foreclosed by frustrated states and why others are deemed essential by fearful states.

It may be wise to use deterrent threats only sparingly, but it can hardly make sense never to use them at all. In a developing confrontation there often comes a point when it might just make a difference if the opponent is made aware that if a particular line is crossed, then there will be a forceful response . Critics of deterrence can point to many instances when such conditional threats have failed to work. The implication of this is not that such threats should never be issued but that they should not be issued casually, without some readiness to implement them if necessary. Deterrence is one way for states to attempt to limit each other's behavior. The cause of a more orderly world is as likely to be helped by a readiness to provide explicit warnings about the risks associated with a breach of a crucial limit as by matters being left ambiguous and uncertain, so that the limits only become apparent once they have been exceeded.