From Parameters, Winter 1995, pp. 70-79.
Americans have gotten used to thinking about deterrence, but usually in terms of what it takes us to deter an adversary. Few Americans like to think about being deterred by others. It offends their sense of righteousness: good guys deter, bad guys are deterred. It offends their sense of honor: wimps or bullies are cowed by threats, not brave souls like we are. But the logic of deterrence has no moral content. Deterrence should work against anyone, anytime, and anywhere that the costs of the action being deterred outweigh the gains.
What will it take to deter the United States in the future? Short answer: Maybe less than we think, maybe more than it should. Less than we think if policymakers base their decisions on cost-benefit calculations of material national interests; more than it should if policymakers act on the basis of Cold War tradition and emotional concerns about honor and prestige.
Without a superpower adversary competing for global domination in the post-Cold War world, the stakes for the United States in most international conflicts and hence the benefits for fighting over them will be low. Logically, therefore, the price that we should be willing to pay for those benefits should also be low. So if an adversary can pose even modest costs as the price of involvement, he should be able to deter us.
American policymakers, however, do not always make decisions by carefully calibrated cost-benefit analysis. When emotions, feelings about national honor, or domestic political sensitivities become engaged, the imperative to demonstrate that we cannot be coerced may come to the fore, even if it requires accepting higher costs than the objective involved appears to be worth. Bravado may then lead decisionmakers to take risks that cool, detached analysis might suggest they should not.
Concepts and Context
A number of holdovers from Cold War modes of thought get in the way of dealing effectively with this question because most defense professionals lived with them so long that they became, as Eliot Cohen has written, hard-wired into our strategic mentality.  Virtually all of modern deterrence theory was grounded in the particular situation of the East-West conflict. These modes of thought can continue to channel thinking in directions that differ from probable future cases. For example:
What Is to Be Deterred
Whether and how the United States might be deterred from taking action depends on the international incentives the United States has to act, the disincentives posed by the enemy trying to deter us, and domestic political constraints on US action. As to the first, there are three main situations of interest:
Domestic political constraints matter because the efficacy of any adversary's deterrent threat is likely to vary with who is in power in the US government, and with the tenor of public opinion. There is no good systematic evidence so far about just when, how, and how well foreign governments can manipulate either US public opinion or officials. It is also not easy to predict which US political leaders would be most easily deterrable. A President who has been restrained in the international arena and has been criticized for insufficient backbone might be more inclined to lurch into action in a crisis than one with several successes under his belt.
We have no experience with one case, which has been only a science fiction scenario so far. It is far from implausible, however, and is the one in which public opinion could be crucial. That case is a public deterrent threat by adversaries to use weapons of mass destruction within the United States if Washington takes action against them: Don't attack us, or a nuclear weapon will go off in some East Coast city or Don't interfere with us, or we'll let anthrax loose in Chicago. This would of course be very risky for the would-be deterrers, since assuming they are a government with an address, rather than a shadowy subnational group they would be provoking Washington to make counter-threats to annihilate them. At the same time, however, how would the US public and Congress react: with unified outrage and staunchness, or with confusion and panic?
Finally, the political context of strategic choice would be complicated by confusion or disagreement about whether indeed the adversary was acting in a defensive and deterrent manner or not. Deterrence is in the eye of the beholder. Americans are not accustomed to thinking of themselves as a threat to anyone else, and therefore tend to think of deterrence as something that we do to someone else not something that our opponents ever have a reason to do to us. If the adversary believes his motives are defensive, but US officials see them as aggressive, then the logic of deterrence is shaken, and the disagreement can lead to posturing and escalation.
Strategies of Deterrence
For attempts to deter by either denial or punishment, there are three main levels of force that might be used against us. The analysis in this section is summarized in Figure 1.
Conventional military defense against attacking US forces. This is the least worrisome option that Washington has to face. After the collapse of Cold War bipolarity, the United States is utterly unmatched in either actual or potential capability GNP, technology, quality of standing forces, and cutting-edge doctrine and military systems integration. Most of the likely targets of US attack lack any allies of military consequence. Except for great powers like Russia or China acting in their border areas, it is not likely that any potential adversaries could deny entry and occupation by US forces.
Conventional deterrence by punishment, on the other hand, is a different matter. The prospect of inflicting high costs on US military forces in the process of an inevitably successful US intervention, or just of costs disproportionate to the value of the target, is a reasonable strategy by which a weak state or group might hope to deter US action. US policies toward Iraq in 1991, Somalia in 1993, and Bosnia for several years illustrate different possibilities.
The war against Iraq showed that it may not even always be possible for a strong, well-armed middle-power to inflict high losses against a determined US air or ground campaign. The emphasis, however, should be on determined. The US government had decided that vital interests and high stakes were involved in the Iraqi seizure of Kuwait, and that a real mid-sized war was worth fighting over the issue. This was true in only two of the cases of US military intervention during the Cold War, Korea and Vietnam. It is unlikely to be true of many prospective situations in which the United States might wish to act militarily after the Cold War, when stakes in local conflicts will usually be limited. The Somalia case showed that vividly, when 18 fatalities were enough to shift US policy into reverse. Policy toward Bosnia has shown that the prospect of a large-scale imbroglio has been enough to preclude commitment to ground combat in support of diplomatic rhetoric.
The two most obvious cases where a deterrence-by-conventional punishment rationale would serve US adversaries well are Cuba and North Korea; the latter is now probably supplemented by a finite nuclear deterrent (see below). Nothing could keep US forces from subduing these two countries, but the cost probably would be high. Terrain differences severely limit opportunities for Desert Storm-style US blitzkrieg operations. Without the Cold War linkage to Soviet power, or a threat to deny access to vast portions of crucial resources like Middle Eastern oil, neither country is likely to pose such a clear threat to genuinely (as opposed to rhetorically) vital US interests. Consequently Washington could be deterred from a decisive military engagement with either country by the prospect of high casualties.
Subconventional military resistance or guerrilla warfare. It is more feasible to mount a threat that promised to hamstring a US intervention, making it prolonged and indecisive, than to deny occupation by the United States altogether. If an adversary can keep the credible capacity to sustain low-level insecurity and prevent the establishment of order that would survive US withdrawal, Washington may well be deterred from engagement. US policy toward Somalia, Bosnia, and Cuba may reflect the deterrent value of such a strategy.
This category focuses the issue of how effective a threat to inflict casualties on US forces will be in deterring the United States from undertaking an intervention. It has become axiomatic that Americans will not tolerate many body bags in the course of an intervention where vital interests are not at stake. There is no clear evidence for this conventional wisdom, however, and ample evidence to the contrary. What is crucial for maintaining public support is not casualties per se, but casualties in an inconclusive war, casualties that the public sees as being suffered indefinitely, for no clear, good, or achievable purpose. Public support for war varied inversely with US casualty levels in Korea and Vietnam (two protracted wars in which classic victory was not in sight), but not in World War II (when casualty rates were far higher, but coincided with evidence that US forces were winning). US casualties in Grenada in 1983 and in Somalia a decade later were identical, but the first operation was quick, successful, and popular, while the second was none of those.
Weapons of mass destruction. Two great powers with large numbers of nuclear weapons remain potential concerns for the United States: China and Russia. For Russia, however, the tables are turned from the situation that governed US-Soviet mutual deterrence in the Cold War. Then, the perception of Soviet conventional superiority in Europe fostered the NATO doctrine of deliberate escalation massive retaliation under Eisenhower, flexible response under his predecessor and successors, but in all cases the expectation that the West had more reason to rely on threats of first-use than did the East. Now, however, Russian conventional forces are much reduced from those of the Soviet Union. They are pushed back hundreds of miles from where they were positioned in the center of Germany, within striking distance of the English Channel, and their intervention in Chechnya has called into question their effectiveness. Unlike the United States, Russia has no significant allies indeed the former Warsaw Pact allies of the Soviet Union, and many of the newly independent states of the former Soviet Union itself, are now not even neutral but tacitly aligned with NATO. In this situation it is Moscow, not NATO, that may perceive an imperative to rely on nuclear first-use to redress a conventional imbalance.
This hardly matters if we assume that the United States and its allies will never have any interest in aggression against Russia, thus obviating any plausible reason that such a Russian deterrent might be brought into play. The critical exception, however, is the danger of a crisis over Russian attempts to reincorporate former parts of the USSR, such as Ukraine or the Baltics. It would be extraordinarily risky for the United States to commit itself to defense of those new states' independence, but political pressures to do so might be intense. In that case the logic of who was trying to deter or compel whom would be quite muddy both sides would be likely to see themselves as the party defending vital interests against aggression.
The same sorts of dilemmas, contradictions, and confusion between material incentives for restraint and emotional incentives to act could erupt in the event of a Chinese attempt to settle the Taiwan issue by force. Beijing could claim in a crisis, as it always has, that Taiwan is an internal security matter (a long-standing Chinese Chechnya, as it were), and remind Washington that the United States has accepted the de jure unity of China since the Shanghai Communiqué. If at the same time the United States was considering moves to support Taipei militarily against invasion, would a Chinese threat to interdict US naval support with nuclear weapons not be credible? And would Washington have a perfectly obvious reason to risk nuclear war over Taiwan, after it so long ago abrogated the mutual defense treaty with Taipei and recognized the PRC?
Apart from the disturbing scenarios involving Ukraine, the Baltics, and Taiwan, the interesting questions are about how weak adversaries might use weapons of mass destruction to deter US action against them. For example, what if Iraq had possessed a handful of nuclear weapons in 1990, declared that it had no quarrel with the United States, and threatened to set a few of them off in US cities if Washington attempted aggression against Baghdad and the new order in Kuwait? Would that have made a difference in the US response? One might cite the failure to use chemical weapons against attacking US forces, or the lack of conventional means of delivery against the continental United States, as reason to denigrate the potency of a putative nuclear threat. But such arguments are limited comforts at best.
In regard to the first point, it was clear that the US and its coalition allies had imperfect yet quite substantial defenses against chemical attack. The Iraqis' use of such weapons might have done little to the military force arrayed against them, while provoking retaliation with even stronger (nuclear) weapons that Baghdad could not match. Second, it would be foolish to take comfort from an argument based on the distance of US territory from an adversary state. The task of smuggling components and assembling a few devices in the United States (a country unable even to restrict the inflow of illegal drugs to levels low enough to cramp consumption) cannot be considered beyond the capacity of these governments. North Korea, Iraq, Iran, or Libya do not need ICBMs to threaten American territory; assuming that they do reflects residual Cold War thinking.
In the case of North Korea a finite deterrent threat against one or two US cities also would have great credibility because the regime in Pyongyang gave regular and impressive demonstrations over the course of 40 years that it was willing to take high risks. Moreover, with the exception of the initial attack in 1950, the North Koreans got away with all of these extreme provocations not one led to forcible US retaliation. Most recently, the North Koreans found that digging in their heels and refusing to rectify their violations of the Nuclear Nonproliferation Treaty led not to sanctions, but to the reverse material aid from outside powers. The agreement to freeze North Korean nuclear facilities entailed the promise of $4 billion in assistance for new reactors and oil in the interim. All this experience offers leaders in Pyongyang scant reason to believe that making nuclear threats in a crisis would not be to their benefit.
The small number of nuclear weapons likely to be available to proliferators in coming years makes it unlikely that the weapons would be committed to tactical rather than strategic purposes. Thus the common expectation that such proliferation represents more a threat to US conventional forces on the battlefield than to the American home front is probably misplaced. Thinking of a small enemy nuclear force as a tactical problem or as a threat to civilian society only in allied countries closer to the scene, such as Japan is too comforting. In contrast to the early 1950s, when concern with clandestine insertion of nuclear weapons was substantial, we became accustomed to debunking the idea as we focused on the intercontinental reach of the highly developed Soviet nuclear force. In terms of good old Cold War deterrence theory, the logic of using such a limited threat to deter US intervention makes perfect sense. Whatever the reason for a small rogue regime to confront the United States, the stakes could not plausibly outweigh the loss of one or two American cities.
In this context the North Korean nuclear threat becomes more interesting if they have five or more weapons rather than the one or two they may have already. Five would give them one to detonate for demonstration purposes, to make their threat credible, and one or two for each of their main adversaries South Korea, Japan, and the United States. If the recent agreement to freeze North Korean facilities succeeds, therefore, it may produce a worthwhile restraint on the country's strategic options even if the agreement remains undesirable in other respects.
As for other forms of weapons of mass destruction, chemical weapons are probably overrated and biological weapons underrated. The tactical utility of chemical weapons is limited by the existence of moderately effective defenses for military forces, while their utility for threatening mass casualties among civilian populations is limited by the difficulty of delivering massive quantities at long range. Only artillery, which would need a successful ground force advance to get within range of a city, would be able to deliver vast amounts of chemical weapons over a large area both cheaply and effectively.
Biological weapons, on the other hand, more accurately represent the poor man's nuclear weapon. They are relatively easy to manufacture, and the facilities for manufacturing or storing them should be easier to disperse and hide from identification and attack than bulkier nuclear installations. More easily than chemical weapons, biological weapons could be devastating on a huge scale. It might also be hard to trace who was responsible for an attack with biological weapons, a point that makes them especially frightening as an instrument of coercion by stateless terrorists without an address. As a deterrent against US intervention, biological weapons could be as effective as the threat of clandestine insertion of nuclear weapons. And an attack on the United States would pose less danger of blowback on the population of the attacking country, something that would be a bigger problem for contiguous adversaries as, for example, if one of its neighbors were to use biological weapons against Israel. In this respect the traditional security advantage provided to the United States by two huge oceans could become more of a liability than an asset.
Two questions stand out when evaluating what it will take for a foreign power to deter American action:
What does it take to make the costs of US action exceed the benefits? In many cases of conventional or subconventional deterrence the answer will be not much. This is most likely when US stakes in the issue or region are low and the deterring country (or group) has enough political strength or social mobilization to pose even modest resistance or the prospect of prolonged instability. Somalia underlined the point, and Haiti does not invalidate it; in the latter instance, the decision to invade was made with the expectation that resistance would be negligible. For countries with the ability to package and deliver nuclear and biological materials, it should also take little to deter us where the stakes for the United States are limited. This means that a very small arsenal of weapons of mass destruction could have great value, since the possessor would have to threaten only a few US targets to make the US costs of opposing the adversary greater than those of refraining.
What does it take to make US leaders realize that costs would exceed gains, and that they should refrain from acting? Threats intended to deter sometimes provoke instead. American Presidents are not accustomed to being intimidated, and they worry about their credibility, about the diplomatic and domestic political costs of appearing weak. They sometimes make rhetorical commitments from which it becomes hard to back away. If the issue comes up in an unanticipated form, in a fast-developing crisis, it is not clear that US leaders would react coolly and analytically rather than emotionally, or would decide on the basis of economic calculation rather than vague grounds of national and personal honor. This prospect is especially worrisome: in some cases the only outcome worse than being deterred would be not being deterred.
The present article is derived from work prepared by the author for the Institute for National Security Studies at the National Defense University, Washington, D.C.
1. Eliot Cohen, The Future of Force, The National Interest, No. 21 (Fall 1990), 9.
2. As this article goes to press, a peace agreement in Bosnia raises the possibility that up to 25,000 US troops might be dispatched to assist in policing the accord. President Clinton's refusal to send troops before the agreement reflects the questionable assumption that the implementing force will not be involved in major combat.
3. John Mueller, War, Presidents, and Public Opinion (New York: Wiley, 1973).
4. The North Korean propensity to do wild and crazy things includes attacking South Korea in 1950, constantly infiltrating paramilitary forces into the Republic of Korea in subsequent years, raiding the Blue House and seizing the USS Pueblo in 1968, tunneling under the DMZ, downing a US EC-121 in 1969, killing President Park's wife in 1974, hacking two American officers to death in the 1976 DMZ tree-cutting incident, blowing up half the South Korean cabinet with a bomb in Rangoon in 1983, downing the KAL airliner in 1987, and occasionally kidnapping and imprisoning South Korean citizens.
Dr. Richard K. Betts is a professor of political science, Director of the International Security Policy Program, and a member of the Institute of War and Peace Studies at Columbia University. He received the B.A., M.A., and Ph.D. degrees in government from Harvard University. He has taught at Harvard and at Johns Hopkins University, and is a former senior fellow at the Brookings Institution. The two most recent of his several books are Military Readiness: Concepts, Choices, Consequences (Brookings, 1995), and Conflict After the Cold War (Macmillan, 1994).
Reviewed 25 November 1996. Please send comments or corrections to firstname.lastname@example.org.