Implications for Future U.S.
Forces of Operation Allied Force:
Is the Force Mix Right? Can we Accomplish The Mission?
Dr. Michael E. O’Hanlon; Senior Fellow, Brookings Institution
A wealth of questions can be raised , and modes of analysis employed, to assess the implications of Operation Allied Force for military planning. My own overall view is that the lessons are of generally limited value. Most of what we learned we already knew—or should have already known; along those lines, I begin this essay with my predictions of how Operation Allied Force would unfold as published in the New York Times the day before the war began. Even an armchair analyst at Brookings could understand the likely effects of weather, and the likely limits of airpower, in such a conflict. (I acknowledge being proven wrong, however, that Slobodan Milosevic would accept NATO demands prior to credible preparations for a ground invasion. But do we really want to count on such behavior from future adversaries, given the poor historical record of strategic bombing campaigns?) Those few things we did learn from this war apply primarily to a subset of conflicts that probably should not be the primary focus of attention in future U.S. military planning.
All that said, one would be remiss in not searching the record of NATO’s first true war for some lessons. Substantial wars do not happen particularly frequently, so we can undoubtedly learn a great deal—even if primarily at the level of detail and nuance—from the western alliance’s 78-day campaign against Serbia.
After revisiting my oped (which also appears on the Brookings website at www.brookings.edu), I take three different slices at the issue of the significance of Operation Allied Force for future U.S. military planning. One asks about the potential role of airpower in future humanitarian interventions, and concludes that we should not overrate its potential based on this conflict alone. Another asks whether we can even afford to conduct limited ground wars in places like Kosovo given the current two-war strategy and the size of the force structure. My own answer to the question is yes. Thus, these first two sections cast doubt on the proposition that NATO’s air war against Serbia should lead to any major restructuring in the U.S. force posture or defense budget. Finally, however, I do consider several specific cases where certain technologies would have helped NATO in any ground war it might have had to conduct against Serbia, and argue that some modifications to the defense budget are thus sensible as a result of the lessons of the war.
OPED FROM BEFORE THE WAR
Should Serbia Be Scared?; The New York Times, March 23, 1999, p. A31; By Michael O'Hanlon, Fellow, Foreign Policy Studies NATO's threat to conduct air strikes against Serbian military forces is not a sufficient strategy for ending the conflict in Kosovo. The United States and its allies need a backup plan in case this instinctive response does not work--and they need one fast.
Western leaders seem to think that if Slobodan Milosevic, the Yugoslav President, continues to resist pressure to end his attacks on the ethnic Albanian majority in Kosovo, a thorough drubbing by some of the world's best air forces might stop him. But it is more than possible that air strikes would fail--both at crippling the Serbian offensive in Kosovo and at convincing Mr. Milosevic that he should back down. For one thing, NATO--which does not want to risk killing innocent Serbs in the process of saving innocent Kosovo Albanians--has already publicly placed constraints on any bombing campaign. The attacks would begin by destroying as much of the Serbs' air defense network as possible and then strike Serbian heavy armor within Kosovo. Most targets in the heart of Serbia have been placed off limits, diluting NATO's overall threat.
In addition, Western leaders, still giddy from Desert Storm and Desert Fox, overestimate the effectiveness of air power. Kosovo is not Iraq, and there are several crucial factors that they have ignored.
First is the simple matter of the weather. It may seem strange that, more than half a century after the development of radar, we still have problems conducting military operations in rain or through clouds. But we do. While cruise missiles can negotiate inclement weather without difficulty, more precise weapons like laser-guided and infrared-guided bombs cannot. And these are the weapons we rely on to destroy mobile armored vehicles.
True, NATO aircraft could fly below many types of cloud formations to attack Serbian armor. But this would expose them to shoulder-launched surface-to-air missiles and and antiaircraft guns. (In the Persian Gulf war, American aircraft typically flew at two to three miles' altitude to stay out of harm's way.) So Serbia could continue its campaign of ethnic cleansing on cloudy days, hiding and camouflaging its armor on clear days.
Even if Serbian forces suffered significant losses, that might not stop their ethnic cleansing operations. NATO aircraft would primarily attack tanks, other armored fighting vehicles and large-bore artillery pieces--of which Serbia has roughly 1,000 each. But Serbia also owns nearly 3,000 mortars and 3,000 recoilless rifles. These are, unfortunately, also highly effective weapons for demolishing homes and otherwise carrying out ethnic cleansing. Many of them can be hidden inside trucks and buildings, making them very hard to target from the air.
If Serbia lost much of its armor and had to curtail its military operations in clear weather, the Kosovo Albanians might stand a better chance on the battlefield. But the rebels--armed with only automatic rifles, rocket-propelled grenades and similar small arms, and numbering perhaps 15,000 to 25,000 troops--would still be badly outnumbered. Serbian forces already number more than 30,000 in Kosovo, with another 75,000 or so available in other parts of the country.
Mr. Milosevic may yet cave in to NATO pressure. But he may also decide to ride out an air campaign. He may conclude, perhaps correctly, that doing so gives him a better chance of holding onto Kosovo than the Rambouillet peace plan would. In this situation, NATO needs to couple its threat of air strikes with two other things. First, it needs to make clear to Mr. Milosevic that it will use ground forces, if necessary, to prevent the wholesale ethnic cleansing of Kosovo. It also should offer Mr. Milosevic a carrot: rather than demand autonomy for the ethnic Albanians throughout Kosovo, NATO could demand autonomy for only part of the region. That way, if autonomy leads to partition, the Serbs will not lose everything.
This broader strategy would not be easy for President Clinton to sell at home or abroad. But threatening air strikes and hoping for the best is much more dangerous, especially for the ethnic Albanians whom NATO is trying to save.
©Copyright 1999 The New York Times
AIRPOWER IN FUTURE WARS
Nonetheless, even if my predictions were generally correct as far as they went, airpower in the end did surprise most analysts in its capacity to be the primary military instrument of victory against Slobodan Milosevic. Might airpower thus have become a primary instrument of humanitarian intervention? Ground units of the Kosovo Liberation Army (KLA) played a role in the military victory as well.1 But that does not change the facts that the outcome was reasonably good (as the outcomes of wars go), and that airpower deserves most of the credit for it.2
NATO’s war against Serbia was rather unusual for this type of conflict, however. Most civil wars occur in less developed countries in places such as Africa, where armored forces susceptible to attack from the air are few and far between; they also commonly involve roving militias that are difficult to influence through strategic attacks against fixed infrastructure. Finally, airpower worked against Serbia only when Russia tightened the diplomatic noose around Slobodan Milosevic by essentially accepting and promoting NATO’s demands, and when NATO sent increasingly clear signals that it was considering a ground invasion if necessary.3 So, on the whole, while airpower may have a certain role to play in some humanitarian interventions, infantry-style operations are generally much more central to them.
Should we try to hasten the arrival of new technology that will greatly improve future airpower? Unfortunately, technology offers only so much potential to change peace operations and humanitarian interventions. It cannot resolve the political challenges that outside countries always face in conducting these types of missions, such as whether to take sides in a conflict, simply create safe havens, impose a partition line, or forcibly disarm all combatants even at the risk of having to fight them to do so. In military terms, as argued previously, technology seems unlikely to spark radical changes in infantry combat over the next two decades. And as discussed later in regard to Kosovo, while airpower may improve significantly for certain kinds of warfare in the years ahead, its relevance to these types of missions is likely to remain limited.4
Technology might make a difference in other ways. For example, it could help make intervention forces lighter, thereby permitting them to be deployed very quickly to stop genocides or other severe forms of violence. Although many commentators overstated the difficulty that NATO would have had in deploying ground forces into Kosovo quickly, they were right that the alliance was not as capable in that regard as it should have been. 5 Trends in the technologies undergirding military vehicles, propulsion systems, and weapons, while short of revolutionary in promise, might permit enough improvement to make a major difference in such limited missions. Some of what is needed to make forces lighter and more mobile--lighter tanks, more mine-resistant infantry vehicles, smaller sealift ships for use in less developed harbors--is already technologically possible today (see below). Indeed, some key units and systems are already in the U.S. force structure, such as army air assault divisions and C-17 airlifters capable of carrying heavy equipment onto short runways. By taking advantage of these existing capabilities, as well as further improvements over time in the weight of armor and fuel efficiency of vehicles and lethality of munitions, it may be possible to rapidly deploy well-armed intervention forces in the future. Specific new technologies, such as a large transport blimp, might make an important difference as well.6
Still, once troops arrive in a combat theater, they may need to fight in difficult infantry combat settings, and technology offers only modest hope that an advanced military will be able to do so more effectively and safely in the decades ahead. On the whole, technology may offer some interesting improvements for peace operations and humanitarian interventions, but it is highly unlikely to change their basic character.
What about the post-conflict phases of such missions? Might technology make it possible, for example, to conduct peace operations in a place like Bosnia or Kosovo with fewer troops in the future? Sensors that listen for vehicles, radars that see through foliage, infrared cameras that allow humans to be distinguished from animals, and robotics that can attach themselves to vehicles or mark people will all reduce the need for human patrols along certain types of borders and zones of separation. These types of technologies, perhaps most notably microrobotic networks, could also help in the search for weapons stocks and other militarily significant assets. Such capabilities will be welcome.
However, the missions of monitoring and patrolling do not generally determine troop requirements. Establishing and maintaining human intelligence networks typically require large numbers of troops, and will continue to do so. Responding to violent outbursts or incursions will still require enough troops to conduct infantry combat operations, which--as noted previously--are unlikely to change greatly. Among other things, they will continue to rely on line-of-sight gunfire, and on mechanized and helicopter transport technologies not notably faster than those that have been in use in recent decades. In fact, a survey of various types of stability operations since World War II shows no significant trend line in the numbers of troops that have been needed per 1,000 inhabitants. Required forces vary much more with the nature of the operation than with the state of communications, weapons, or transportation technologies.7 This trend line is unlikely to change direction anytime soon. Again, the promise of technology for aiding humanitarian interventions and peace operations, while hardly negligible, is rather limited.
Trends in airpower technology are likely to produce some important progress. Among the probable achievements, improved real-time communications systems will help a great deal. Although improvements have been occurring, today’s weapons still do not always obtain targeting data quickly enough to destroy moving enemy assets before they can hide or take shelter.8 NATO is also likely to be able to operate much more effectively in bad weather in the future.9 Operation Allied Force against Serbia witnessed the use of munitions with somewhat better all-weather capabilities than those available during Desert Storm. Notable was the joint direct attack munition (JDAM), particularly when dropped by an aircraft with a precise radar like the radar on the B-2. (Other aircraft could carry equally good radar, although they did not in 1999.) Even these munitions were too inaccurate to target moving tanks, however. Submunitions with their own infrared, acoustic, and (someday perhaps) even radar sensors to automatically search for enemy armor will soon be able to attack below cloud cover without requiring aircraft to drop to low altitudes. Increased numbers of JSTARS aircraft and UAVs will provide the initial targeting information so that munitions can be dropped in the vicinities of targets at the correct times. The munitions may be severely challenged by complex terrain and some decoys and countermeaures, but in the open--where an aggressor may sometimes need to mass troops, as Serbia did against the KLA--they are likely to be effective. 10
However, important limitations will remain in 2020. Enemy weapons hidden within buildings and forests are unlikely to be attacked successfully in most cases. The most important reason is that sensors will either not see targets in such places, or have a hard time identifying and tracking them. Enemy supply lines are likely to be hard to shut down as well. Easily identified choke points and vulnerabilities will of course be targetable, even in bad weather. But assets such as small trucks using secondary roads and pontoon bridges will still be able to carry individual barrels of fuel or small stores of ammunition by avoiding main roads, otherwise minimizing their exposure to enemy sensors, and interspersing themselves within civilian traffic. There is some chance that new technologies, such as radio-frequency weapons, may allow airpower to shut down virtually all traffic into a province like Kosovo without causing civilian casualties; however, much stands in the way. For example, in wooded terrain, it may not be possible to get radio-frequency weapons near enough their targets to destroy the electronics. And at river and stream crossings, nonmotorized boats and ferries could move supplies without being vulnerable to such weapons.11
Finally and most fundamentally, even if NATO dominated the air in 2020 the way it did in 1999, units conducting ethnic cleansing with small and medium weapons could still operate effectively unless challenged by a ground force of comparable strength. In previous guerrilla conflicts, a small rebel force immersed in a friendly population could resist much larger government forces. But the tactic of ethnic cleansing, as applied by Serbia against the Kosovar Albanians, puts the onus on the rebel force, which must try to defend its populations rather than engage in hit-and-run warfare. In these conditions, a force like Serbia’s--armed primarily with machine guns, mortars, and antiaircraft guns--can wreak havoc on a population like that of the Kosovar Albanians without possessing a huge overall force advantage against local insurgents. Since these types of weapons are easy to hide inside buildings and vehicles, they will remain hard for airpower to target in the future. 12
In conclusion, for wars motivated by humanitarian concerns, technology is unlikely to change the basic situation NATO confronted in 1999: while an advanced military can use airpower to punish an aggressor, and limit its ability to conduct mechanized warfare, it cannot defeat small arms from long distances or stop the violence perpetrated by them.
EVEN A LIMITED GROUND CAMPAIGN IN KOSOVO WOULD NOT HAVE BADLY HARMED U.S. GLOBAL MILITARY PREPAREDNESS ELSEWHERE
Could the Pentagon handle a ground war like the one it might have waged against Serbia without letting down its guard in other parts of the world? Some critics of U.S. military involvement in the Balkans raised this question in 1999. They base their concerns in part on the idea that, under the Pentagon’s 1997 Quadrennial Defense Review, the nation is supposed to have the capacity to fight two regional wars beginning nearly simultaneously and then overlapping chronologically. War against Serbia could in theory count as one of these two wars, but in practical terms the two-war strategy is focused largely on Iraq and North Korea. It is based on the real fear that they could both stir up trouble at the same time--much as in 1994.
The concerns of critics on this point are valid. But they are often overstated. Global readiness concerns would become significant if the United States deployed 150,000 troops as part of a NATO force of some 200,000, as envisioned by those who contemplate a full-scale invasion of Serbia. (Europe would be hard pressed to deploy more than 50,000 troops quickly, so the remainder of the 200,000 would have to be American.) However, worries about North Korea and Iraq need not rule out a limited invasion of Kosovo involving some 50,000 to 75,000 troops.
Today’s U.S. armed forces are one-third smaller than in the 1980s. They are almost 15 percent smaller than the Bush Administration had envisioned when drawing up the first blueprints for the post-Cold War U.S. military. Meanwhile, they are working hard from the Persian Gulf to the Western Pacific to the Balkans. As a result of doing more with less, the U.S. military’s readiness--indicated by metrics such as spare parts inventories, the condition of major weapons, training hours, and military personnel shortfalls--is undeniably declining. If readiness deserved a grade of C during the 1970s, an A- in the 1980s, and an A in the early 1990s, it has now declined to perhaps a B/B+.13
Consider how three types of possible military operations against Serbia would affect the military readiness and global deterrent posture of the United States:
o A prolonged air war (estimated total U.S. cost: $5 billion14). The readiness costs of the current air war are modest, and will remain that way even if the United States sends an additional 300 aircraft to aid in the campaign as now envisioned. Most aircraft deployed to Europe could just as easily and quickly be redeployed to fight Iraq or North Korea if necessary. Although air-launched cruise missile stocks have been significantly depleted, other weapons can substitute for them--albeit at a somewhat greater risk of American casualties.
There is a readiness cost associated with flying 10 to 20 percent of all U.S. military aircraft at two to three times their normal pace. Consumption of fuel is not a major problem, since Congress will soon give the Pentagon money to buy more. But spare parts will be used up more quickly than anticipated, and it generally takes many months to replace them. In the interim, more airplanes will become "hangar queens," unable to fly for want of basic equipment like engines and functional radars. Perhaps 2 to 5 percent of the country’s combat aircraft fleet will be adversely affected.
Nonetheless, most of the planes in Kosovo as well as those now flying at bases in the United States would be quickly usable elsewhere if necessary. The readiness costs of this operation are real, but in the end seem likely to remain modest. Unfortunately, the same can be said of the air war’s likely military, political, and humanitarian benefits.
o A limited ground war (estimated total U.S. military cost: $20 billion15). If NATO sent a force into Kosovo to liberate some or most of the province, it might need 100,000 to 125,000 troops for the job. That would provide NATO with rough parity against the entire Serbian military and an advantage of two to one over Serbian forces there now. Indeed, as Brookings scholar Joshua Epstein, as well as the late Trevor Dupuy have shown, attacking forces do not require a large numerical advantage to win, especially when they control the skies and possess superior troops, equipment, and mobility.16 In Desert Storm, for example, the U.S.-led coalition had about 700,000 troops to Iraq’s forces in the Kuwaiti theater that numbered somewhere between 400,000 and 550,000 at the start of the air war.17 By deploying 100,000 to 125,000 troops to Kosovo, NATO would have a comparable advantage over Serbia even if the latter managed to substantially reinforce its units there now. (The total size of Serbia’s active-duty armed forces is 115,000, but it is highly doubtful that most of the troops could get to Kosovo with their equipment in the face of allied mastery of the air.)
Europe could provide around 50,000 troops to any such invasion effort. But our allies lack the type of transport and mobile logistics assets needed to send more. The remainder of the NATO deployment--roughly two divisions--would have to be American. That makes two divisions out of a total of 13 we possess in the active-duty force (counting Army and Marines).
Deploying this much force to the Balkans would affect the U.S. ability to respond to problems in other regions, but the costs would be tolerable. Our transport ships and airplanes would be in short supply for other crises while ferrying initial forces to the Balkans, but this period of relative vulnerability would only last a few weeks. As for combat forces, the remaining 11 Army and Marine divisions, backed up by perhaps another division of reservists that might be quickly called to active status, could handle likely requirements elsewhere. In reality, out of those 12 divisions we might really have the equivalent of only 10, since a number are undermanned at present and a good chunk of another is supporting the NATO operation in Bosnia. But even if major war broke out again in Korea or the Gulf, we would not need more than 6 to 8 divisions to win it decisively. That is because, in light of the deterioration of the Iraqi and North Korean threats and the growing strength of our South Korean ally, we could get by in either place with a slightly smaller deployment than in Desert Storm.
That would leave two to four U.S. combat divisions for use elsewhere. A modest amount of airpower to support them would remain available in the continental United States; additional combat jets could be diverted from the war in the Balkans if necessary. (Attack helicopters could provide much of the close-air support for the ground invasion in Kosovo.) This remaining U.S. force would not be large enough to go on the offensive and march to Pyongyang or Baghdad. Nor would it be able to call on many specialized assets like imaging satellites, Stealth fighters, JSTARS reconnaissance planes, and unmanned aerial vehicles (UAVs)--virtually all of which would have been devoted to the two ongoing wars. (We have many more JSTARS and UAVs than in Desert Storm, but virtually all would be required to prosecute two wars aggressively.)
Still, this remaining force would be more than adequate to set up a robust defensive perimeter and hit enemy forces hard with airpower. In all likelihood, such a capability would provide a compelling deterrent to Saddam or Kim Jong-Il.
An all-out ground war against Serbia (estimated total U.S. military cost: $40 billion18). Deploying a NATO force large enough to threaten Belgrade and occupy Serbia would probably require 200,000 troops, of which 150,000 would have to be American. If a major war then did break out somewhere else, most likely against Iraq or North Korea, we could require almost all remaining U.S. force structure to win it--leaving virtually nothing for whichever of our two longstanding nemeses, Saddam Hussein or the DPRK regime of Kim Jong-Il, we were not already at war with.19 We do not want to take that kind of risk unless absolutely necessary.
Any war in the Balkans is certain to take a toll on U.S. military readiness around the world. But the military costs of a limited ground invasion would be bearable--much more so than the costs to U.S. and NATO credibility, and to the ethnic Albanian people, that will result if we continue to lose the air war in the Balkans.
TECHNOLOGIES THAT WE SHOULD BUY—AND SOON
When it became obvious that NATO air power wasn't going to stop Slobodan Milosevic's murderous campaign of ethnic cleansing in Kosovo, the call went up for ground troops. Some military analysts then countered that it could take three months to assemble an invasion force. Together with the reluctance of President Clinton and many members of Congress to risk American casualties in the Balkans, that appears to have been enough to preclude the ground option--at least so far.
In fact, the United States and key allies could move enough force to win such a war in half that time. Serbian forces in Kosovo were only one-tenth the size and strength of the Iraqi units that took Kuwait. Even though Serbian troops are better fighters than Iraqis, and even though they would enjoy the natural benefits of Kosovo’s complex terrain, NATO could have overwhelmed them with a force roughly 100,000 strong that emphasized helicopters and infantry over heavy armor.
But it is nonetheless true that today’s U.S. military, while quite ready for certain kinds of wars in places like the Persian Gulf, is too unwieldy for others. If it had wound up fighting in Kosovo, it would have taken longer to get there than it should have--and probably suffered more casualties than should have been necessary as well.
How can we fix this problem? Many Republicans in Congress have responded to the unfortunate turn of events in Kosovo by declaring a readiness crisis and clamoring for multi-billion dollar increases in defense spending. Congress has done well to focus attention, and more resources, on matters such as military pay and additional purchases of spare parts. But the real problems we would face getting effective forces to a place like Kosovo quickly have little to do with readiness as Congress defines it. Nor would they be fixed by the specific spending increases that Congress and the Pentagon advocate for next-generation weapons such as Virginia-class new attack submarines and F-22 "Raptor" fighter jets.
We should learn from the Kosovo experience and be better prepared next time. Following are three inexpensive technologies that would help the U.S. military deploy effective ground forces more quickly than is possible today.
Smaller Tanks. The 70-ton Abrams of Desert Storm fame is an awesome fighting machine. But it's difficult to airlift and would have trouble maneuvering on the narrow mountain roads and treacherous terrain of the Balkans. Something like the 25-ton ''armored gun system,'' which the Army canceled in 1996, would be far better adapted to these conditions. It's no Abrams, but it could provide lead forces with important protection in places where they might be ambushed or heavily opposed. It would cost only about $1.5 billion to develop the weapon and buy more than 200 of them. By contrast, each Abrams costs more than $5 million just to upgrade.
Mine-resistant Combat Vehicles. The Army and Marine Corps rely on specialized equipment to clear minefields before main force moves in. As a result, tanks and armored personnel carriers are not built to withstand antitank land mines; Humvee's and military trucks are even more vulnerable. Both Australia and South Africa make fairly small and inexpensive vehicles--roughly $100,000 to $200,000 apiece--that have good bottom armor, specialized shock absorbers and independently-suspended wheels designed to keep going forward even if a single tire is blown out by a mine. We should buy some.
Smaller "Roll On/Roll Off" Transport Ships. Several years ago the Clinton Administration decided to purchase more ships to carry military equipment. Unfortunately, these ''large medium-speed roll-on/roll-off'' ships (or LMSRs) were designed with the wide, deep ports of South Korea and Saudi Arabia in mind. They are far too cumbersome for ports like Albania's. Smaller ships, actually a less expensive way to carry cargo and only slightly slower than LMSRs, would be very nice to have for conflicts near undeveloped ports in which heavy forces are required.
These types of capabilities would not make possible casualty-free ground wars. But they might make the difference between losing, say, 200 American lives and losing 100 in a war in a place like Kosovo. They might also cut the time needed to deploy adequate forces, now perhaps 6 weeks, to less than a month. In wars like the one in Kosovo, where thousands of people can be slaughtered or starved every day, that potential time savings could be of enormous importance.