Three hundred forty-five million dollars can, roughly speaking, buy one F-22 Raptor -- the U.S. military's new stealth fighter plane -- or pay the average annual cost of 3,000 soldiers (although it would cost far more to equip, maintain, and deploy either the fighter or the troops). The soldiers are a better investment. Yet U.S. military personnel, pundits, and policymakers have been downplaying the importance of ground forces since 1991. Even today, in the face of ongoing, manpower-intensive counterinsurgency campaigns in Iraq and Afghanistan, the Bush administration is emphasizing long-range strike capabilities over land forces. The recently released 2006 Quadrennial Defense Review and the president's budget proposal for fiscal year 2007 both reaffirm this priority.

The administration has maintained this emphasis despite the fact that the long-term neglect of U.S. ground forces has caused serious problems in the Iraqi and Afghan campaigns. If not corrected, moreover, this neglect will cause even worse problems in the future. War is fundamentally a human activity, and attempts to remove humans from its center -- as recent trends and current programs do -- are likely to lead to disaster.


The current manpower crisis in the U.S. military predates both the attacks of September 11, 2001, and the Iraq war. The problem started in the early 1990s, when George H. W. Bush began recklessly cutting military spending without paying enough attention to the foreseeable (and unforeseeable) uses to which the military would be put. Bill Clinton accelerated these cuts, even as the number of U.S. forces deployed abroad steadily grew. By the end of the decade, the U.S. military was overstretched and inadequately staffed for the missions it faced.

Calls for Washington to reverse some of the cuts began to proliferate. Just what critics were asking for, however, varied dramatically. Some recommended an increase in traditional military spending. But others demanded that more money go to research and development (R & D) in order to spur a "revolution in military affairs." These RMA enthusiasts viewed the 1990s as a "strategic pause": the United States faced no imminent threat, they argued, and so should use the time to gird itself for future challenges by developing new technology.

In the 2000 presidential campaign, George W. Bush promised to repair the damage done to the military during the previous decade. Even before winning the election, however, he made it clear that he planned to address the problem in a narrowly focused way. Bush was (and remains) a firm believer in the idea of an RMA; he had proclaimed it a priority as early as 1999, long before anyone imagined that Donald Rumsfeld would again become secretary of defense. After winning office, Bush began rapidly to translate his promise into Pentagon reality. In February 2001, Bush announced that he would increase military R & D funds over the next five years by $20 billion and that he would devote 20 percent of total R & D spending "to especially promising programs that propel America's Armed Forces generations ahead in military technology." The Pentagon announced that all Department of Defense programs would be evaluated by the degree to which they were "transformational." And the Office of Force Transformation was established to coordinate and oversee all of the department's "transformational" efforts.

The Bush administration also sought change in another area. The cuts of the late 1990s had made it difficult for the military to retain experienced and talented officers in all of the services. After taking office, Bush and Rumsfeld, following the advice of the Joint Chiefs of Staff, proposed to address this problem by investing in military quality-of-life initiatives: pay raises and improvements to health care, housing, military schools, and so on. The idea was to encourage skilled service members to delay retirement by closing the gap in quality of life between the military and civilian society.

Both of these initiatives -- the RMA and the quality-of-life improvements -- may have been well intentioned. But their combined effect was to exacerbate the central problem that the U.S. military currently confronts: the inadequacy of its ground forces. Consider the RMA. Its enthusiasts generally favor using long-range airpower instead of ground forces as a solution to most military problems. Proponents such as Rumsfeld base their case on the superior efficiency of airpower, its ability to reduce collateral damage and conduct pin-point strikes, and its cost-effectiveness. Some even argue that using airpower is morally superior to using ground forces, since attacking from above puts the smallest possible number of U.S. personnel at risk. Even before Bush got on board, RMA enthusiasts had regularly recommended cutting ground forces to pay for new weapons and communications systems. These efforts were not generally successful -- the size of the ground forces remained basically constant from 1993 until very recently -- but they did ensure that proposals for increasing the size of the ground forces gained no traction in Washington. Once Rumsfeld took office with his determination to make the RMA a reality, rumors began to leak out of plans to cut the strength of the ground forces to pay for RMA-style programs.

Bush's quality-of-life initiatives, meanwhile, made the prospect of increasing troop levels even more distant by dramatically raising the cost of such increases, and by raising the cost of even maintaining current ground-force levels, these initiatives made that less likely too. According to a recent report by the Government Accountability Office, the cost of compensation for the active-duty military has risen by 29 percent since 2000, driven in large measure by a 69 percent increase in military health-care expenses. It now costs an average of $112,000 to maintain one service member on active duty for one year, and the total annual compensation cost for the active-duty force, which was $123 billion in 2000, had grown to $158 billion by 2004. Such numbers make it all too easy to argue that personnel increases are just too expensive.

This is not to argue that quality-of-life improvements were not necessary to help the military retain its senior personnel; they were. But they were not sufficient to address the general manpower shortage. The quality of life enjoyed by U.S. troops is affected just as dramatically by the length of their combat tours as it is by their housing, salaries, and health care. Solving one problem at the expense of the other is unlikely to succeed -- indeed, it has not: high numbers of young officers have begun leaving the force despite the new incentives. Over the long term, it seems certain that the repeated and back-to-back deployments made necessary by the military's ground-troop shortage will erode morale (and therefore recruiting and retention rates) faster than quality-of-life improvements will restore it. Maintaining a skilled, motivated volunteer military requires addressing both aspects of the problem.

New weapons systems, meanwhile, are not necessarily any cheaper than new troops. As the cost of service members has increased in recent years, so has the price of new technology. The projected unit cost of the F-22, for example, has risen from $149 million in 1991 to $345 million today. It will cost approximately $63 billion -- or 40 percent of the annual compensation cost (based on 2004 figures) of all active-duty personnel -- to purchase the 178 F-22 aircraft now in the defense program. This price tag does not include the costs of R & D, modernizing the aircraft (the design of which began in 1986), arming it, or flying it. And all this is for a single weapons system. The current defense budget also includes plans to procure the F-35 Joint Strike Fighter and to accelerate the development of a long-range manned bomber as well.


In its five years in office, the Bush administration has avoided improving the human capabilities of the military -- and the crisis has grown steadily worse. The long-term deployment of U.S. soldiers to Iraq and Afghanistan has taken a severe toll on the ground forces. Combat tours, which lasted six months in the 1990s, have been extended to a full year for most army troops in Iraq and Afghanistan. Many soldiers in the active force (and in the National Guard and the Reserves) have already been deployed twice and are now facing their third tour. Although reenlistment rates have remained high, recruitment rates have fallen dangerously, morale has dropped in some units, and some experts, such as retired General Barry McCaffrey, warn that "the wheels are coming off" the army as it struggles to sustain a large deployment with insufficient personnel.

Unless the United States rapidly withdraws from Iraq, moreover, there is no sign of relief on the horizon. Although the administration has permitted the army to maintain nearly 30,000 extra soldiers in its ranks for the past several years, the president's budget for next year requires the army to shed those additional troops. And the ground forces proposed both in that budget and in the 2006 Quadrennial Defense Review would support a long-term deployment of only about 18 brigade combat teams (each comprising about 3,500 troops). At the height of the campaigns in Iraq and Afghanistan, by contrast, the United States had more than 20 brigade combat teams deployed to combat zones, and even these were not enough to pacify and rebuild those countries. It is hardly a secret that the U.S. Army and the Marine Corps are short on troops; senior officers and analysts regularly refer to the problem when discussing operations in Iraq and Afghanistan or when explaining U.S. options -- or the lack thereof. Lieutenant General John Vines, who stepped down as commander of U.S. ground forces in Iraq at the beginning of this year, has pointed out that many U.S. soldiers are now on their third or fourth tour of duty in Iraq. "The war has been going on nearly as long as the Second World War and we're asking a lot of the forces," he said in April. What is hard to understand is why Washington has steadfastly refused to address the issue.

The explanation seems to stem from two beliefs, deeply held by top members of the Bush administration, about how war works. The first is the notion, shared by most RMA enthusiasts, that war is fundamentally about killing people and destroying things. The second is the conviction that military preparations should be guided by the business principle of investing in success. The basic flaw in both beliefs is that they take parts of the problem to be the whole.

Consider the first concept. The major priority of the current "information revolution" in warfare is to enable the military to locate, identify, track, target, and destroy enemy weapons systems -- from aircraft and radar installations to individual soldiers. All of the services' "transformational" programs, including the army's Future Combat Systems program (a network of manned and unmanned systems), emphasize the sensors they will deploy. The aim, clearly articulated on a number of occasions, is to gain "near-perfect" intelligence -- that is, a nearly complete understanding of where the enemy's forces are so that they can be eliminated. In fact, with a few exceptions, all of the new sensor and intelligence systems the Pentagon is now developing are designed to locate and identify enemy systems -- not to interact with them.

The distinction is important. Under the old system (still in use in Iraq today), U.S. units seeking intelligence about the enemy did not confine themselves to halting nearby and observing the enemy's disposition. Armored or motorized cavalry formations would actually attack the enemy before them (as the Second Armored Cavalry Regiment did during the Gulf War, in 1991, and as the cavalry of the Third Infantry Division did repeatedly in 2003). The goal of such attacks was threefold: to locate and identify the enemy, to determine its intentions, and to set the terms of the ensuing battle.

The second of these tasks -- determining the enemy's intentions -- is particularly hard for the new sensor-shooter systems being promoted by RMA enthusiasts to do. As military strategists have long understood, simply bombing an enemy force (as the new technology would do) does not necessarily force it to reveal its intended course of action. Troops under bombardment take cover in various ways; they rarely respond by attempting to execute the attacks, defenses, or movements they had initially been preparing for. Bombing enemy troops thus yields far less information than actually engaging them on the ground.

Some RMA advocates might argue that it does not matter; if the enemy's troops are all dead, what difference does it make what they were planning? That view would be valid if the aim of war were the literal annihilation of the enemy's military, or if one could be sure that using airpower to destroy a certain proportion of the enemy's armed forces would force the enemy to surrender.

The problem is that actual enemies rarely cooperate. The record of attempts to compel governments to surrender by such methods is not a good one. The Germans tried it against the British in 1918 and again in 1940-41 and failed. The United States tried it against North Vietnam (most intensely in 1972) and also failed. It is true that on several occasions the United States has seemed to win a conflict through airpower alone -- but in such cases, it has generally either had the support of indigenous allied forces or backed up the bombing with the threat of a ground invasion. Thus, although U.S. airpower achieved victory in Bosnia in 1995, it did so only with the support of a Croatian land offensive; in Kosovo in 1999, the Serbs only capitulated when rumors began to circulate of an impending U.S. ground attack; and the Afghan campaign of 2001 required the support of the Afghan Northern Alliance and Pashtun tribes in the south. Moreover, 39 days of bombing in 1991 did not make Saddam Hussein surrender, nor did the "shock and awe" campaign of 2003, which struck several thousand targets with precision-guided munitions within a few days. Large-scale ground invasions were needed in both cases.

Why has airpower alone almost always failed to force enemies to surrender? The reason is actually fairly simple. The destruction or partial destruction of a military, by itself, places a state's ability to perform its core functions at risk, but it does not destroy that ability permanently. Militaries can be rebuilt. Shattered infrastructure can be repaired. Even losses in population can be made good over time. State leaders canny enough to think in such terms -- as many U.S. opponents have been -- are not readily persuaded to surrender simply by the partial destruction of their militaries.

The occupation of an enemy's land with ground forces is a different story entirely. A state under aerial bombardment need only survive until the bombardment stops. A state under occupation, however, risks never regaining control of its territory. Worse still, an occupying force can usurp the basic functions of a state by, for example, governing territory and reorganizing the mobilization of resources to achieve different objectives. Occupied or partially occupied states may not be able to reverse such reorganizations. This is especially true of authoritarian governments, which cannot survive without the physical control of their populations.

Occupying an enemy's territory thus places the core purposes of that enemy's state at much greater and more imminent risk than mere bombing. Many of the United States' foes, moreover, can comfort themselves with the knowledge that any U.S. bombing will likely be precise and calculated to do minimal collateral damage.

The decision to develop a method of war that relies for success primarily on identifying and destroying targets rather than occupying territory therefore reveals a fundamental misunderstanding of the nature of war itself. Such a method focuses too much effort on what should be a subordinate part of war -- destruction -- and ignores other critical aspects of war, without which that destruction itself has little meaning.


The Bush administration's preference for military solutions that emphasize identifying and destroying targets stems from its second mistaken conviction about the nature of war: its faith in the applicability of the business model to military matters.

In business, profitable companies reinforce success rather than failure. If one product line is working well while another is not, businesses often shut down the latter and devote more resources to the former. One recent example is Dell's announcement that it will move away from selling personal computers, even though this was the original core of the company's business model.

The Bush administration enshrined this principle as a central element of its military-transformation agenda early on by adopting the concept of "network-centric warfare." Developed by a team of retired military officers, the idea was to apply the lessons of the "information revolution" in business to war. In books and articles written in the late 1990s (before the stock-market crash), advocates of network-centric warfare, such as Admiral Arthur Cebrowski, cited companies such as Dell, American Airlines, Cisco Systems, and others to illustrate the competitive advantages that the extensive application of information technology gave to business -- and could give to the military.

Then, in October 2001, Rumsfeld created the Office of Force Transformation to coordinate all aspects of military transformation and put Cebrowski in charge. Since Cebrowski's appointment as director of this new office, the Bush administration has steadily applied this business model to warfare.

Already, the change has borne fruit. The U.S. armed forces are now extremely good -- and far better than the competition -- at locating and destroying targets from thousands of miles away. At the level of soldiers fighting other soldiers, the advantage is less pronounced: in Iraq, the enemy has been able to kill a number of U.S. soldiers, although it has been virtually incapable of preventing the U.S. military from either striking any target it chooses or retaliating in kind. From a business perspective, then, it seems to make sense to reduce investment in soldiers and increase investment in target-and-strike systems, since these yield what looks like a higher marginal return. This is precisely the logic that a number of supporters of network-centric warfare have used to argue their case.

The problem with this approach is that unlike a corporation a military cannot safely decide that it will not compete in certain "markets," such as land warfare. Nor can it necessarily rely on "profits" in the airpower "market" to offset "losses" in ground combat. And given the fact that victory in most wars requires the occupation of the enemy's territory, or at least a convincing threat of occupation, the U.S. military must continue to compete in the land-power "market," however poor the "marginal returns" of land combat might be compared to those of combat using airpower. "Reinforcing success" by reallocating resources away from ground forces (and from those elements of the ground forces that provide capabilities different from airpower) will only create vulnerabilities that enemies will exploit.

This is especially so given the very different nature of competition in war versus in business. Businesses compete against one another but do not attempt to infiltrate and physically, psychologically, or organizationally destroy one another. Success is measured in profit, and profit can frequently be better raised by increasing internal efficiency rather than by harming a competitor. In business, efficiency directly translates into success.

The same is not so in war. After all, military organizations are designed to destroy one another as a prerequisite to achieving some larger purpose. Efficiencies within a military organization do not contribute to the attainment of this goal directly. They do so only indirectly, by freeing up resources that may or may not be used to achieve the objective. But how those freed-up resources are used, not the efficiency of that use, is the only measure that will actually tell in conflict. The competition between military organizations is therefore central to war in a way that competition between companies is not central to business. If one military opts out of a "market," it simply creates a vulnerability that another military can -- and inevitably will -- use to harm it.


Ground forces perform a wide variety of tasks. It is the ability to control territory and populations, however, that is land power's unique contribution to war in this high-tech age. Only soldiers are discriminating enough, in terms of both judgment and the capabilities of their weapons, to mix with an enemy's population, identify the combatants intermingled with that population, and accomplish the critical tasks of governance and reorganization that are so essential in persuading an enemy government to surrender. These are not functions that can be usurped by airpower, by computerization, or by mechanization in any way -- at least not until robots with real cognitive abilities can be fielded.

In the meantime, military occupation and population control will remain human endeavors and will be less amenable to technological enhancement than any other aspect of war. It has long been true that one soldier with a radio (and access to artillery or air support) can kill a large crowd. If the aim is to control that crowd without killing it, however, hundreds of soldiers are required, no matter how good their technology. The size of the ground force needed to control conquered territory is determined by the size of that territory, the density of its population, and the nature and size of the resistance, not by the nature of the soldiers' weapons. When it comes to reorganizing or building political, economic, and social institutions, there is no substitute for human beings in large numbers.

The idea that technological improvements in the U.S. ground forces, such as the army's Future Combat Systems, will be able to reduce dramatically the number of soldiers necessary for missions similar to those in Iraq or Afghanistan is therefore illusory and unrealistic. As long as war remains a process of human beings interacting with one another -- as all irregular warfare is -- the land-power "market" will require a heavy investment in people.

This need has been clearly borne out in the struggle in Iraq, where coalition success has rested entirely on the interaction between coalition troops and Iraqis. Airpower and long-range land-based firepower have been helpful in killing insurgents quickly and with minimal collateral damage, but they have played an entirely supporting role. The speed with which Iraqi soldiers can be trained; the number of villages in which the coalition can conduct its strategy of "clear, hold, build"; and the ability of coalition troops to restore and defend Iraqi infrastructure, polling places, and borders have been directly proportional to the number of coalition soldiers in Iraq, not to the quality of their equipment. And there is no reason to imagine that this situation will change in any future counterinsurgency or stability operation. The recently released Quadrennial Defense Review insisted that the U.S. military should remain able to conduct such operations in the future on a large scale and for prolonged periods. Making that possible, however -- not to mention ensuring U.S. preeminence in conventional warfare -- means maintaining large ground forces. Indeed, Washington will need a large pool of trained and ready soldiers for all sorts of conflicts at every point along the spectrum for decades to come.


Ground forces are expensive. Soldiers must be recruited, equipped, housed, fed, trained, and transported -- whether they are fighting or not. The expense is protracted, since soldiers receive retirement and health-care benefits even after they leave the service and for the rest of their lives (as do their dependents). As noted above, these factors have been exacerbated by the well-intentioned efforts of the Bush administration to improve the quality of life in the military in order to bolster troop retention. The cost of an individual soldier is now so high that some analysts claim it is impossible to field large ground forces anymore. These critics claim that economies must be sought and that economies can only come from the increased use of technology.

But arguments for the efficiency of technology frequently obscure the real costs of relying on it. Like troops, aircraft and other weapons systems come with significant life-span costs that far exceed their sticker prices. There is the R & D required to develop the technology, which usually takes years, sometimes decades, to produce a single usable product (the F-22 program, for example, was begun in 1986). There is the need to maintain large stores of spare parts, ammunition, and fuel. There are modernization and upgrade costs, since most major weapons systems remain in circulation for decades. The truth is that buying an advanced new weapons system, such as the F-22, involves the same sort of long-term costs as a brigade of soldiers. Indeed, the United States will probably have to continue spending money on the F-22 well after all of today's soldiers have retired.

This is not necessarily a bad thing. Whatever the flaws and the merits of the F-22, the United States certainly needs to develop advanced aircraft to ensure its future air superiority and its ability to deliver precision ordnance. If the F-22 or the F-35 Joint Strike Fighter suffers from major defects, as some critics have suggested, that does not mean the program should be scrapped; the defects should be corrected.

But new aircraft, although necessary, are not sufficient to meet U.S. defense needs. As the current strains imposed by the campaign in Iraq show, the United States also needs ground forces that are larger than is currently planned. And ground forces, like new fighter planes, cannot be created overnight. It takes at least a year to train a soldier in the basic skills of the incredibly advanced U.S. military. It takes a generation to produce generals to command those soldiers. Creating specialists able to interact with local populations in their own languages and cultures -- another need highlighted by recent events -- takes decades.

These time frames are fairly inflexible. Technology and improved training methods can help, but it is not possible to cut dramatically the time it takes a human brain to master certain kinds of skills. All of this means that ground forces must be established years before they are needed and must be maintained even at moments when they do not seem necessary. Although there is no army in the world now that could expect to defeat the U.S. military, certain tasks remain (such as conducting large-scale, long-term peacekeeping or counterinsurgency operations) for which the army is even now unprepared. Fixing that problem requires starting today.

In practical terms, this means increasing the size of U.S. ground forces by at least 100,000 and possibly by as many as 200,000 active and reserve soldiers and marines, combat and support forces both. No smaller increase can make up the deficit that now hinders U.S. operations in Iraq and elsewhere. Such an increase will be expensive and may require some reorganization of the defense budget. Although maintaining air superiority is important, the fact is that the United States is far less likely to be challenged in that arena in the near future than it is on the ground. War, in this respect, is the opposite of business: Washington must focus its attention first not on what it does well, but on the areas of greatest danger. Right now, the greatest danger is on land, not in the air, in space, or on or under the sea.

But simply killing the transformational programs would be unwise. Since it takes years and sometimes decades to field new weapons systems, it is time to start planning how to reequip the military. The best approach would be to increase the defense budget to a level adequate to support both efforts. Although this increase, which might amount to one or two percent of GDP, will inevitably inspire howls of protest in certain quarters, the absolute size of the defense budget, its percentage of the total U.S. budget, and its share of GDP are ultimately irrelevant. What matters is whether or not the money is sufficient to pay for the kind of military the United States needs in order to prevail in current and future conflicts.

Throughout the 1990s, this dictum was ignored. For political reasons, Republican and Democratic administrations set arbitrary limits on defense budgets, and they declared themselves satisfied with whatever sort of military those arbitrary figures provided. Since 2001, Americans have been confronted with the consequences of those decisions. The country must now not only spend money on current conflicts and future programs but also repair the damage done by a decade of underspending.

This problem must not be left to the next generation of Americans, who will face unforeseeable challenges of their own. Even in times of peace, it is wise to maintain adequate armed forces, since peace always come to an end (usually unexpectedly). The need is that much greater in this period of protracted war.

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
  • FREDERICK W. KAGAN is a Resident Scholar at the American Enterprise Institute
    and the author of the forthcoming finding the Target: The Transformation of American Military
  • More By Frederick W. Kagan