The strategic culture of the Cold War combined great eagerness to accumulate weapons with great caution in their use. Fearing that any act of war might start a progression of moves and countermoves leading to catastrophe, the nuclear powers strenuously avoided any direct combat with each other. There were many wars, but the remarkably deliberate and controlled behavior that became a new norm for nations around the world deterred the thoughtless escalation of confrontation and the eruption of war through sheer miscalculation. With the end of the Cold War, the size of armed forces, military expenditures, fear of nuclear attack, and learned habits of restraint are all much diminished. Today, disputes over minor diplomatic aims or mere posturing for domestic constituencies are enough to provoke reckless displays of bellicosity or imprudence. Thus China advanced its claim of sovereignty over Taiwan in March with military exercises that were clearly meant to be as alarming as possible. For their part, Taiwan’s leaders, in their futile pursuit of U.N. membership, had earlier seen fit to score largely symbolic diplomatic points in defiance of emphatic Chinese warnings -- as if there were no danger of war, and the balance of power were nothing more than an academic construct.

Likewise, a long-standing territorial dispute between Greece and Turkey over an islet in the Aegean Sea inhabited only by about a dozen goats quickly escalated in January to a confrontation. Naval forces were deployed in dangerous proximity to each other, accompanied by much nationalist bombast from politicians on both sides and frenzied media coverage. A group of Turkish journalists deliberately exacerbated tensions by planting a flag on the ten acres of rock. When a mutual withdrawal of forces was eventually negotiated with U.S. help, Greek editorialists accused newly installed Prime Minister Costas Simitis of treason. The episode was reminiscent of the warmongering press campaigns of the last fin-de-siecle, being not merely prenuclear but pre-1914 in its raw truculence.

The Republic of Korea also asserted a claim earlier this year over small uninhabited islands, in the Sea of Japan. President Kim Young Sam, beleaguered by accusations of corruption, sent warships to the scene against a backdrop of foolhardy taunts and idle threats against Japan in the South Korean media. Pakistan, too, has of late seemed bent on starting a war with India over Kashmir by smuggling in weapons and making inflammatory declarations.

As in prenuclear times, when fighting breaks out there is no great concern by either protagonists or outside powers to limit its scope. Consequently, infantry clashes can escalate directly to aerial bombing, as in the border fighting last year between Ecuador and Peru, or to artillery barrages against cities, as in the diverse struggles within the former Yugoslavia and the former Soviet Union.

None of these quarrels is new. But while the Cold War induced caution, present circumstances evidently do not. A new season of bellicosity is upon us, and it is unlikely to long endure without consequences. Because wars have become less dangerous to fight, the danger that wars will be fought has increased.

When national leaders manufacture or amplify external crises for internal political purposes, as noted above, they rarely intend to start a war. But as each side’s gesticulations turn into the other side’s perceived threats, confrontations can easily get out of hand. That happened often enough before the Cold War and is now likely to happen again. Indeed, in the case of Ecuador and Peru it has already occurred, not coincidentally when each was in the midst of a presidential campaign.


A strategic environment so greatly transformed calls for an equally transformed U.S. military policy. Certainly there have already been important changes. The stock of long-range nuclear weapons has been greatly reduced, and both the defense budget and the number of personnel in the armed forces are considerably smaller than a decade ago. But at a more fundamental level, much remains unchanged and is therefore outdated.

That the end of the Cold War should leave disorientation in its wake was inevitable. Previous great wars entailed strenuous mobilizations, so that in 1918 and 1945 rapid demobilizations naturally ensued once hostilities ended. On both occasions the U.S. military went beyond the steep reduction or outright disbanding of war-inflated forces. It tried to resuscitate prewar routines and customs while attempting to recast doctrines and force structures in light of the presumed lessons of the last war. Thus battleships gave way to aircraft carriers, armored formations became the standard for ground forces, and jet aviation became the norm.

The Cold War required a vast effort, but well short of an all-out mobilization. When it ended, therefore, there was no overwhelming pressure to drastically demobilize. Nor was there any urge to abandon its peculiar habits and institutions. On the contrary, the U.S. military has striven to preserve continuity in all things and has resisted as much as possible the effects of declining budgets.

Cold War military practices and priorities persist in all sorts of subtle ways because they are embodied in the millions of people who make up the armed forces, the Pentagon’s civilian bureaucracies, intelligence organizations, defense laboratories, research centers, and military industries large and small, but mainly because they are simply not recognized as products of that era. The Cold War lasted so long that nobody remembers any prewar normality to which the military should revert. Nor are there are any sharply defined threats to which the military can adjust. A deliberate refusal to break with the past is not the issue, nor is the habitual conservatism of military institutions; the issue is the persistence of unquestioned premises. Because all but the youngest officers and virtually all defense officials were educated throughout their professional lives in the strategic culture of the Cold War, its axioms remain embedded in their mentality.

Of these axioms, by far the most significant is the paradox of deterrence: military forces are most useful when they are not used at all. Despite this, the U.S. military constantly struggled throughout the Cold War to maintain high standards of combat readiness, essentially as a matter of principle. But this axiom had a powerful effect on all other core policy decisions -- on the sizes, roles, and budgets of the branches of the armed forces. The risk of combat casualties faced by different categories of the armed forces varies dramatically. The prospect of high casualties, which can rapidly undermine domestic support for any military operation, is the key political constraint when decisions must be made on which forces to deploy in a crisis, and at what levels. Yet the distinction between forces that are only theoretically available and those that are in fact politically usable in real-life confrontations is still largely disregarded. The composition of U.S. military forces reflects the lingering priorities of the Cold War, when U.S. forces served primarily for deterrence display and self-reassurance. Today, however, the former is largely ineffectual and the latter unnecessary. Calculations of the cost-effectiveness of conventional forces should therefore be governed by their usability. That is far from being the case at present, and the result is a contradictory military policy.


When senior Pentagon officials and military officers discuss how the United States might intervene in this or that outbreak of violence somewhere in the world, the likelihood of combat and the probable magnitude of U.S. casualties are invariably dominant in their deliberations. Hence the usual recommendation of such conclaves is to avoid armed intervention altogether, except for bloodless visits by American naval forces to nearby waters or perhaps reconnaissance flights, if possible by remotely piloted aircraft. If the situation is seen as requiring more forceful action, a strike by cruise missiles launched by ships or submarines from a safe distance might be suggested. Because of the considerable cost and limited supply of cruise missiles, however, only manned aircraft are suitable if more persistent bombardment is required. Yet the Pentagon is extremely reluctant to allow their use. The difficulty of finding and hitting specific targets, the possibility that stray bombs will kill innocent civilians, and the strength of enemy antiaircraft defenses are the standard objections, invariably coupled with pessimistic assessments of the likely impact on the enemy and his decisions.

In spite of the Pentagon’s huge annual expenditure on airpower, many officers and officials endorse the most critical estimates of its worth and discount claims of its efficacy as "the airpower myth." That was certainly the case during the prolonged controversy from 1992 to 1994 over the use of tactical bombing against Bosnian Serb forces. When the strikes were finally carried out last year, their swift technical and political success surprised the Pentagon, for the accomplishments of airpower in the 1991 Persian Gulf War had been attributed to uniquely favorable circumstances.

Pentagon decision-makers rarely propose the use of ground forces, for which the risk of casualties is inherently high because of their proximity to enemy fire. If the introduction of ground forces is nevertheless advocated by State Department or White House officials, military leaders almost always strongly resist it, arguing for demanding preconditions, beginning with the insistence that "vital" U.S. interests be at stake. With the United States now as secure as it can be in a nuclear world, and with its closest allies largely unthreatened, such a precondition is a virtual prohibition. The United States has all sorts of interests that are or might be threatened, but few can accurately be said to be vital -- that is, encompassing the country’s very survival.

Should that most stringent prerequisite not suffice, however, the military has others. One is that the goal of any military action be precisely and permanently defined in advance, with no "mission creep" allowed -- a demand almost impossible to satisfy in this world of human ambiguities and inconstancies. Another is that a swift and decisive resolution to the conflict be assured, a highly desirable aim to be sure, but one that will always depend on what both sides do and how well they do it. Nor can any results be truly decisive for long. With Iraqi President Saddam Hussein still in power and recurrently threatening Kuwait, was the Persian Gulf War truly decisive?


Yet when senior Pentagon officials and military officers convene to discuss the overall composition of the armed forces rather than their use in combat, the wide variance in the risk of casualties of the different kinds of military forces does not seem to matter very much. To cite the largest differential, Army and Marine Corps troops are the most likely to suffer combat casualties, while long-range missiles are of course immune, and so-called strategic bombers expose very few people to enemy attack. If the entire b-2 force now planned were to be deployed in a conflict, the operation would risk the lives of only 32 crew members. Yet under current plans, in 2000 the United States would have the equivalent of 29 divisions of ground forces but only modest numbers of cruise missiles, very few stando missiles for tactical aircraft, and only six wings of long-range bombers, with 130 aircraft in all. Compare the long-term prorated costs: the annual bill for an active-duty armored or mechanized division is almost $3 billion, and a much smaller light-infantry division requires almost $2 billion, while the average annual cost of the costliest bomber wings, comprised of 16 b-2 bombers, is less than $1 billion. Likewise, for the annual cost of one armored division, the United States could buy more than 2,000 cruise missiles -- a truly formidable force, seven times larger than the number launched during the Persian Gulf War.

Ground forces are undoubtedly more versatile than bombers or cruise missiles. Through well-publicized redeployments, bombers can be used as diplomatic instruments during crises, as well as weapons against valuable targets. And cruise missiles have the virtue of being usable in isolation, for one-time strikes. However, none of that compares with the versatility of ground troops across the entire spectrum of conflict, from mere observation for peacekeeping to high-intensity warfare.

Versatility, however, presumes that the ground forces be usable in combat. In reality, political constraints greatly restrict their availability. The best evidence of that, paradoxically, is last year’s decision to send 16,000 troops to Bosnia to oversee the Dayton peace accords. Only after five years of intense national debate, amid countless reports of widespread atrocities and exceptionally destructive warfare, was the Pentagon’s opposition to the deployment overcome. Even then, very restrictive conditions were imposed: a rigid one-year time limit and an equally rigid troop limit.

No adversary, should any emerge, need fear that attacking U.S. troops would only provoke the arrival of more U.S. troops. Thus the deterrent effect normally inherent in a great power expedition -- the high risk that attacks against it may well backfire by provoking an expansion of the force and its role -- was renounced at the start. With that risk removed, the danger to the troops in Bosnia would be correspondingly increased were it not for the time limit, which invites prospective troublemakers to wait a while before acting. That constraint makes it more likely that the expedition will have a violent aftermath, and will thus be seen as futile. Both restrictions may yet prove unfortunate, but under the circumstances they were politically inevitable. They accurately reflect the current underlying reality that U.S. ground forces are not available as instruments of U.S. foreign policy, except under very unusual conditions.

In theory, the lack of obvious vital U.S. interests in Bosnia’s vicissitudes conditioned the debate and resulted in the ambivalent decision, although the advocates of intervention did make a rather strong case for a wider U.S. interest in world order. In practice, however, the decisive consideration for opponents of intervention was not the presence or absence of U.S. interests but the fear that too many U.S. troops would be killed. The most senior U.S. military officers stated that explicitly in 1992. Remarkably, General Colin Powell, then chairman of the Joint Chiefs of Staff, not only gave background press briefings to that effect, but even published an article in the daily press to argue against any form of U.S. military intervention in Bosnia.

There is clear evidence that the much-cited calculus of U.S. interests was in fact irrelevant to the Bosnia debate. At the very time the Pentagon was successfully resisting any use of force in the former Yugoslavia in the closing months of 1992, it proffered a U.S. expedition to Somalia. Senior military officers were willing to send U.S. troops there because they mistakenly believed that no casualties would ensue in a mission labeled as humanitarian. They were unwilling to send troops or even use tactical airpower in the former Yugoslavia because they feared there would be fighting and casualties. As always, talk of U.S. interests, present or absent, vital or not, was merely part of the rhetorical carapace of policy decisions driven by more compelling motives.


In a world of wars that are less dangerous but much less likely to be prevented, U.S. military forces must once again be kept primarily for combat rather than for deterrent display. Thus the current composition of U.S. forces as a whole, however versatile in theory, is no longer appropriate. What this or that type of force could do in combat -- if only the United States had high birthrates, large families, and a hypernationalistic population that sanctioned with equanimity thousands of casualties in any minor military affray -- is or should be irrelevant when deciding the composition of the armed forces. What counts is not cost-effectiveness in general, but the cost-effectiveness of the military forces and materiel that can actually be used during a conflict. The two criteria yield widely divergent results because it is invariably expensive to reduce casualties by developing and deploying ways to move personnel and materiel farther from harm’s way on the battlefield while maintaining their effectiveness.

U.S. military policy must also prepare for possible drastic changes in the strategic environment, including the emergence of situations that would require the large-scale use of ground forces and the toleration of high casualties. A full range of military capabilities is needed, but that can be assured by preserving capacity for remobilization and by maintaining lower-cost reserve forces. Once the necessary precautions against sudden and powerful new threats are in place, the United States should acquire the military forces it can actually use as instruments of policy. That would be a prosaic prescription -- indeed a mere banality -- were it not for America’s half-century of immersion in Cold War deterrence, during which weapons and forces were most useful precisely insofar as they were not used.

A wholesale recalibration of U.S. military policy is needed to shift money and resources to the forces actually usable, at the expense of the less usable. There is no way of measuring how usable different kinds of military forces are, even in theory. Such a quality is not fixed and inherent, but depends on the perceived political importance of the task at hand, which can change very quickly indeed. On August 3, 1990, President George Bush evoked scarcely any criticism when he ruled out the use of American military strength to expel Iraqi invasion forces from Kuwait. Yet that became the imperative of his administration very soon thereafter. No U.S. forces were usable under the first decision; all nonnuclear forces were equally usable under the second.

What can be measured as a surrogate for political usability is the inherent exposure to casualties of different categories of forces. As a first approximation of what might be called a casualty exposure index, it would be enough to count the personnel in the close-combat echelons of each kind of force. By that measure, for example, armored brigades with their 400 or so tank crewmen and 300 or so riflemen would have a significantly lower exposure index than light infantry or Marine brigades, which have several times as many riflemen and few heavy weapons. Therefore, under current strategic conditions, the use of armored forces would be advantageous, even in terrain and for tasks that would call for infantry forces according to traditional tactical criteria. One lesson of the Somalia debacle is that armored forces nowadays are preferable to light infantry forces in almost any environment, in spite of their tactical limitations. Inserting forces that are theoretically more suitable only to have to withdraw them abruptly because of politically intolerable casualties accomplishes nothing.

Such rudimentary casualty exposure indexes could be made somewhat more realistic by taking into account the proximity of U.S. combat echelons, broadly defined to include the crews of ships and aircraft, to a combat locale, under the reasonable assumption that the closer the proximity, the greater the likelihood of casualties. With that adjustment, for example, the Navy’s amphibian and gunfire-support warships, which must come fairly close to shore, would be assigned higher casualty exposure indexes than Navy ships used primarily to launch cruise missiles. Admittedly, complications intrude when a given type of force can be employed for different purposes; most combat aircraft, for example, can carry out both close-support missions against antiaircraft fire and remote attacks with stando missiles.

Whether adjusted for proximity or not, meaningful casualty exposure indexes could readily be calculated for each type of force. As proxies for political usability, they would define which forces are most suited for current post-Cold War and post-heroic conditions, in which the invariable limiting factor for U.S. military operations is Americans’ low tolerance of casualties.

In descending order, a rough ranking of usability would begin with unmanned, long-range weapons (notably ballistic and cruise missiles) and proceed through remote forces with small combat echelons, such as air crews that launch stando weapons, to the least usable forces, the Army and Marine infantry forces whose combat echelons are useful only in close proximity to the fighting. Such a ranking system should become a central criterion of U.S. military policy, as it would further the acquisition of the forces that are less exposed to casualties and therefore more usable in present conditions.

The United States requires reasonably substantial forces in all categories, from aircraft carriers to infantry divisions. But the current inventory of combat forces contains too many manpower-intensive ground units that are highly exposed to casualties and simply unavailable as instruments of foreign policy.

The overall effect of spending too much money on the upkeep of unusable forces instead of merely maintaining a remobilization capacity for them is that Congress now deems an exceptionally rich menu of important technological innovations unaffordable. The list includes the first practical directed-energy, or laser, weapons capable of intercepting rockets and ballistic missiles, which are a scourge from Sarajevo to Israel; stealth aircraft, which are substantially immune to radar detection in most situations and therefore capable of doing much with few fliers exposed to enemy attack and even fewer losses; remotely piloted aerial vehicles, now available only in sample numbers and not yet for strike roles; more and better cruise and stando missiles; versatile advanced munitions; and all the diverse applications of information technology that only await funding to be implemented, not to mention those that could be developed with more money.

The promise of these innovations is not a stronger United States -- it is strong enough -- but rather a greater ability to use force remotely, yet accurately and with discrimination. Tactical bombing, guided and protected by a host of electronic ancillaries and aimed with now-routine precision, proved very effective in Bosnia notwithstanding the seemingly decisive reasons why it should have failed. So far, however, only fixed or large, easily recognizable targets are vulnerable to remote attack. That may be good enough in classic war situations, as in the Persian Gulf War, but the lower the intensity of conflict, the less can be achieved by attacking airfields, depots, headquarters, or other major targets. The promise of the potential innovations is a greater ability to contend with more elusive enemies in more ambiguous situations.

It is unwise to retard these innovations in the armed forces, but not because dangerous threats to vital U.S. interests are imminent. On the contrary, capital-intensive forces that have the lowest possible exposure to casualties are needed precisely because at present there are no threats that might justify substantial American casualties. The use of military forces will likely continue to be discretionary, as it has been in Bosnia, Haiti, and Somalia, rather than imperative. Some continue to argue against all such discretionary interventions, not coincidentally with the vehement support of senior officers who represent the most manpower-intensive forces. Abstention from all interventions would certainly allow an indefinite delay in the refocusing of U.S. military policy at the expense of those very forces.

There are always strong arguments against any particular U.S. military intervention, and all discretionary interventions can be dismissed as social work, i.e., feel-good endeavors disconnected from core U.S. interests. But what may be true of any single intervention -- and was certainly true of the Somalia expedition -- is not true of all. If the possessor of much of the world’s military power refuses to use it, greater world disorder is only the most immediate consequence. At a time when Cold War restraint has given way to adventurism, if U.S. military power is withheld in one crisis after another, it is bound to stimulate the growth of other military powers. A vacuum will have been created that other countries will fill.

A policy of nonintervention would yield a world not only less stable, but also more militarized. That would certainly affect core U.S. interests, at the very least by damaging economic development and undermining civilian rule, even if no direct threats were to emerge or increase. Hence the urgency of redefining the priorities of U.S. military policy, to yield forces that can more readily be deployed for discretionary interventions. To fight only when necessity compels is the attribute of small powers. The United States fortunately lacks the classic expansionist proclivities of the great powers of history, but it does have widespread responsibilities that call for appropriate military means.

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