Joel Hayward's Books and Articles

During the Gulf War, the A-10 Thunderbolt II "Warthog", usually employed in close air support missions, conducted extensive valuable interdiction, while one wing of F-111 s, usually employed as deep-interdic­tion aircraft, destroyed scores of tanks and armoured vehicles with precision weapons released at low altitude. This did not happen during Operation Allied Force. Information released so far reveals that few multi-role and fighter aircraft had permission to fly lower than 15,000 feet before mid-May, and almost no ground-attack and close air support aircraft gained permission to fly at all over large parts of Kosovo and Metohija.28 Even F-16C and F-16CJ Fighting Falcons, equipped with the HARM targeting system to suppress enemy air defences29, were employed cautiously and without the freedom of action and ability to fly low that would have given them greater effectiveness against a wide range of targets, including Serbian tanks and troops deployed in Kosovo.

The issue of flexibility relates to another of the fundamental principles of war: economy of force, which refers to the rational use of force by selecting the best mix of combat power. In other words, even though a force must be proportionate in scale and type to the overall objective, and be applied in such a manner as to ensure appropriate "mass", it must devote no more effort than is absolutely necessary to perform particular tasks. It ensures this through an appropriate balance of missions and the correct use of weap­ons and systems.

During Operation Allied Force, however, the internal inconsistencies of the overall NATO objective con­founded efforts to apply effective economy of force. Initially, when the primary aim was supposedly to coerce, the air commanders directed missions at the Serbian integrated air defence system and the nation's political and military C3 infrastructure.30 Even then they failed to apply sufficient force, conduct­ing a small number of sorties and missile attacks each day. Yet the selection of targets was reasonable and, had the mission rate been substantially higher, this may have proven more successful.

However, when Kosovar refugees began streaming into Macedonia and Albania on 26 March, horrified NATO leaders promptly ordered a shift of emphasis from coercion and prevention to punishment and destruction. They consequently began falsely claiming that their intention had never been a short, sharp coercive attack, but had always been (to quote Vice Admiral Scott A. Fry, Joint Staff Director of Opera­tions, United States Department of Defense) a longer campaign designed "to degrade the capability of both the special police and the military to conduct repressive actions against the Kosovar Albanians. That was the mission that we started with. That's the military mission that continues today."31

Thus, while missions against Serbia's air defence systems and C3 nerve-centre continued, they also began targeting "strategic infrastructure", (including oil refineries, storage facilities and spare parts ware­houses), fielded forces in Kosovo and Metohija (airfields, barracks, known field installations and any concentrations of armour) and their logistical network (fuel dumps, supply depots, roads, railways and bridges).

To illustrate the wide range of targets under attack a week after Operation Allied Force commenced, one only needs to note the targets mentioned by Vice Admiral Fry at a major press conference on 30 March (that is, one week into the operation). NATO was attacking Serbia's integrated air defence system, which was so "robust" that "we need to grind away at this system to set the conditions for operations both in progress and in the future." Serbia's command and control infrastructure was also under attack, he said, in order "to degrade the ability of the leadership to control their forces." Meanwhile, attacks against Serbian military and paramilitary forces were now underway, designed "to degrade their ability in the field to conduct repressive actions." Added to all this, Fry said, NATO was now attacking industrial targets "so that he [Milosevic] cannot repair or manufacture new weapons."

Unfortunately, NATO lacked the means to do all these things simultaneously -- that is, to conduct the type of parallel operation that had paralysed Iraq in 1991 -- having only 150 strike-capable aircraft at that time.32 Even for the "low-level contingency response" (to quote NATO sources33) that Operation Allied Force started as, this was a weak force. Yet when the Balkans campaign quickly transformed into a war of real substance -- and by late-April even General Michael Ryan, the USAF Chief of Staff, had admitted "the USAF is in a Major Theatre War" -- it was entirely incapable of undertaking all these multiple missions while still applying an appropriate economy of force.

Rather than employing its limited combat power in the most effective manner possible, using the right assets for specific jobs and wasting no effort on secondary objectives (the hallmarks of economy of force), the NATO political leadership directed its air force to attack a very wide range of targets. Because of its excessive fear of casualties, it then limited the air force's ability to use many of its air assets against targets they were specifically designed to hit. A-10 Warthogs, for instance, were used cautiously and apparently only in the relatively safe southern border areas during the final two weeks or so of the campaign. AH-64 Apache attack helicopters were not used at all. This failure to prioritise, and restriction on the use of certain assets resulted in a dissipation of strength and a consequent loss of effectiveness.

This lack of effectiveness was patently clear, particularly throughout April, when NATO's air strength was still weak. NATO failed to change Milosevic's behaviour. It did not slow or stop -- let alone prevent -- the forced expulsion of hundreds of thousands of Kosovars. As early as 29 March this caused an embar­rassed General Clark to claim, contrary to his earlier statements (and NATO's stated operation objec­tive), that "we never thought we could stop this. You can't conduct police actions from the air in any country." A month later a more embarrassed Clark had to admit that Serb military and paramilitary in Kosovo and Metohija were still enjoying great freedom of movement and, even worse, "you might actu­ally find out that he [Milosevic] has strengthened his forces in there."36

One cannot deny that NATO had substantially increased its overall air strength and mission rate by the beginning of June, but until at least early May its operation remained marred by a mismanaged economy of force. No wonder Colonel John Warden, principal architect of the Desert Storm air offensive, com­plained on 16 May: "the way the air war has been designed suggests it was a very bureaucratised, compartmentalised, and not very competent approach. The target list has clearly not been designed to have a systematic impact on the Serb forces ... This is very unprofessional on the part of the various political authorities."37

Warden's claim, although doubtless self-serving, is clearly substantiated by even the most cursory ob­servation. Yet the air campaign -- despite suffering from political idealism and unrealistic expectations, and consequently violating many of the basic principles of war -- did eventually lead to the Yugoslavian government's decision to accept a generally unfavourable compromise. This fact should not be downplayed. It means that, at least on some levels, John Keegan seems justified in claiming that "the capitulation of President Milosevic proved that a war can be won by air power alone."38 Those who doubt this, he added, "are wrong. This was a victory through air power."

This is a bold statement from someone with Keegan's grand reputation. It carries weight and will doubt­less relieve the many airpower theorists and practitioners who had worried, right up until Milosevic's sudden and unexpected capitulation on 6 June. that Operation Allied Force was not progressing well and might even tarnish the gleam of Gulf War accomplishments. But is Keegan right? Was this an effective and successful air campaign, and was it the only reason the Yugoslavian leadership capitu­lated? Answering these questions necessitates first a critique of the campaign, and then an analysis of the other factors that may have contributed to NATO's eventual "victory".

The Air Campaign Dissected

As noted, NATO commenced Operation Allied Force believing it would be a short, sharp coercive action similar in scale and nature to Operation Desert Fox. It consequently did not dedicate a war-size force to the operation. Instead it utilised the air units already stationed in Italy for air verification missions to support United Nations Security Council Resolution 1199, passed on 23 September 1998, which called for the immediate end of hostilities in Kosovo. In the months before launching Operation Allied Force, NATO reinforced those units, adding additional wings and positioning a fleet of warships in the Adriatic. That fleet possessed additional carrier-borne air assets. Even so, NATO commenced its attack on Yugo­slavia with only one-fifth of the total number of fixed-wing aircraft employed in Desert Storm's opening stages.39

The first wave of attacks against Yugoslavia (and at that stage they were not just against Serbia) bore little resemblance to the massive initial blows against Iraq in 1991, which had involved 1,300 combat sorties on the opening night alone.40 Because of Operation Allied Force's emphatic "no losses" stipulation, attacks in the first couple of nights involved mostly ship-launched Tomahawk cruise missiles and B-52H bombers launching their AGM-86C cruise missiles from up to 160 kilometres away from their targets.41 However, once NATO air commanders felt confident these missions had damaged Serbia's integrated air defence system enough, they began introducing increasingly more strike aircraft. By 27 March those aircraft were attacking a wide range of targets, mostly with PGMs, but because of their very small number their efforts were dissipated and their results only mediocre.

Throughout the first week, NATO aircraft flew an average of only 48 strike sorties per day, and even though their efforts benefited from a host of support aircraft -- including AWACS42, JSTARS43, and C­130s equipped for jamming and communications relay -- their impact was nowhere near strong enough to disrupt significantly, let alone disable, the Serbian political and military command and control systems. (During the first week of Desert Storm, to offer a comparison, allied air forces flew more than 1,000 combat missions per day.44)

NATO's failure to amass adequate air assets was certainly the main reason for the operation's slow initial progress, but it was not the only reason. Despite their sophisticated electronics, NATO aircraft proved unable to defeat airpower's greatest and most constant foe: inclement weather.45 On 30 March, Vice Admiral Fry explained to assembled journalists the impact of the typically bad Balkans weather on the operation: "On night one ... we were able to complete every mission without the impact of weather. But as the operation has continued, the weather has got more difficult with each succeeding night, and we have had instances where sorties were unable to complete their missions in their target areas be­cause they were weathered out."46

Bad weather actually plagued the entire first month of Operation Allied Force, inhibiting reconnaissance (both satellite and aircraft47) and reducing the number of strike missions with laser-guided weapons, which need acceptable visibility for lock-on. This prompted Secretary of Defense Cohen to complain (albeit with some exaggeration) on May 11: "We've had only, let's say, a half dozen days when we have not had to pull our forces, the air missions, back. We've had day after day or night after night when various flights of combat aircraft have had to turn around and come back because they couldn't pen­etrate the weather."48 Perhaps to justify the lack of results to date, Cohen then compared this situation to the Gulf War, when allied aircraft operated with "clear" visibility over flat terrain. Yet Cohen is wrong; bad weather had also seriously handicapped allied operations in the Gulf, prompting Richard Hallion, the USAF Historian, to later write: "about half of all sorties to Iraq were affected by weather, resulting in cancellations or diversions."49

Inclement weather may go at least part way to explaining why the daily mission average of strike aircraft throughout Operation Allied Force's entire first month totalled a mere 92 missions.50 The weather not only tested pilots' flying skills; it also made targeting very difficult. Even the much-touted Laser-Guided Bombs (LGBs) need fairly cloud-free conditions to work properly, a fact that contributed to the surpris­ingly low rate of target destruction in the first weeks.

Bad weather, coupled with a constant fear of Serbian SAMs and AAA fire, also explains why NATO initially relied so heavily on sea-launched Tomahawk cruise missiles, which were less affected by weather than even NATO's most sophisticated and supposedly all-weather piloted aircraft. In fact, by late-April, after a month of bombing, NATO began running low on both sea- and air-launched cruise missiles. Discussing "current shortages", General Richard Hawley of the USAF Air Combat Command noted on 29 April that NATO had to be "very judicious" in its use of Tomahawks because "there aren't many left."51 No official figures for the total number of Tomahawks used during Operation Allied Force have yet been released, but one can safely conclude the number was greater than either the 288 fired during Opera­tion Desert Storm or the staggering 415 used during Operation Desert Fox.53

For the entire first month, the excessively rigid rules of engagement that evolved from the "no casualties or collateral damage" dictum also inhibited the ability of aircrews to inflict truly punishing damage on many troublesome Serbian targets, especially the mobile components of the nation's air defence system.

Because of the nagging fear of Serbian SAMs and AAA, pilots were seldom permitted to fly lower than 15,000 feet, which prevented most "missions of opportunity" (attacks based on pilot discretion after the sighting of unexpected targets). Pilots were also expressly forbidden from releasing bombs on any targets that could not be confirmed 100 per cent or that risked even a remote chance of civilian targets.54 Some pilots ignored this limitation, occasionally with predictable and well-publicised disastrous results.

The consequent slow progress bothered many NATO airmen, who felt frustrated that, although they were flying the most modern aircraft and dropping the most precise bombs ever made, they were func­tioning far less effectively than they had anticipated. At one briefing session in Aviano, Italy, General Short found himself face-to-face with this frustration. After he opened the floor to suggestions and com­ments, a German Luftwaffe captain raised his hand and bluntly suggested a simple remedy: loosen the rules of engagement, or relax the limits on when and how aviators can strike. After a long pause, Short replied: "We're working on that."55

Short was right; they were working on it. Acutely aware that Operation Allied Force was making slow progress, NATO's political leaders succumbed to pressure from their military colleagues and, after the NATO Summit in Washington from 23 to 25 April, permitted significant changes to the way the war could be prosecuted.56 As a result, by the beginning of May NATO had not only increased the strength of its air force two-fold (to over 700 aircraft, 400 of them strike-capable) and its daily strike mission rate three-fold (to over 300, a figure matched by non-strike air missions), but had also relaxed the rules of engagement (with targets of opportunity reaching ten per cent of the daily total).57

The air campaign immediately "changed gear" and began to inflict heavy punishment on Serbia. John Keegan astutely observed that this change represented Operation Allied Force's separation into "two air wars, the first lasting a month [until late April/early May], the second six weeks."58 The first had virtually no impact, he said. The second, however, had a powerful impact, "thereby visiting a true blitz on the Serb homeland."

Keegan's description is accurate; NATO unleashed a veritable torrent of steel on Serbia, and did so in keeping with the principles of warfare it had previously ignored (even if its end-state remained as mud­dled as ever). Finding a correct balance between concentration of force and economy of effort, missions against Serbia's air defence system continued -- and with increased intensity -- as did those against the political and military command and control infrastructure. NATO aircraft also intensified attacks on key war industries (especially vehicle and weapons factories and repair shops) and the logistics sys­tems of the fielded forces. Most importantly, NATO greatly stepped up its attacks on the two "systems essentials" (to quote John Warden) that sustained Serbia (and almost all other modern nations): elec­tricity and petroleum.

Attacks on electric power transformers, petroleum refineries and storage facilities had occurred sporadi­cally since Operation Allied Force commenced on 24 March, but with mediocre results. Several trans­formers were destroyed, along with two petroleum refineries, and by mid-April General Clark was able to report that 70 percent of Serbia's oil storage facilities had been hit.59 But "hit" does not mean destroyed, or even disabled. Most remained functional. By early May, however, NATO's vastly strengthened air force, flying 24-hours per day with an impressive sortie rate, had effectively destroyed Serbia's petro­leum refining capabilities and was "working its way" through the storage facilities. This immediately began to cripple Serbia's logistical capabilities and left it reliant on oil imports. These were difficult to obtain after broken bridges and dangerous slicks of industrial pollution, caused by NATO attacks, closed Serbia's main waterway: the Danube.

Electricity generation and transformation suffered worse. On 2 May, for example, NATO aircraft con­ducted simultaneous attacks on five transformer stations and succeeded in blacking out over 70 percent of the population.60 They accomplished this using special "soft bombs" (carbon pellets or graphite strands that short-circuited electrical systems) designed to shut off power without excessive permanent damage to the transformers themselves. After this blackout, the first of many throughout the remaining weeks of Allied Force, Pentagon spokesman Kenneth Bacon explained that it would "confuse their [the Serbians'] command and control system; it diverts resources that might otherwise go into running the command and control system; it disorients and confuses their computers by shutting them off quickly."61 It would have a "major impact" on Serbian resistance, he concluded. NATO spokesman Jamie Shea agreed, boasting that the blackout sent the Milosevic regime a clear message that NATO "has its finger on the light switch. ... We can turn the power off whenever we want to."62

The destruction of Serbia's petroleum industry and the disabling of its electricity grid also delighted John Warden, the main architect of the Gulf War air campaign, who considered it "better late than never". Without electricity, he said, "production of civil and military goods, distribution of food and other essen­tials, civil and military communication, and life in general, become difficult to impossible."63 Unless the stakes were very high, he concluded, "most states will make concessions when their power-generation system is put under sufficient pressure or actually destroyed."

This had been the basic principle underpinning Operation Desert Storm. A nation is a "system", Warden had maintained, which has centres of gravity of varying importance.64 He depicted these centres of gravity as five concentric rings, with national command and control (which he also called "leadership") being the core. Moving outwards in terms of importance were these rings: key production ("systems essentials"); infrastructure; national will; and fielded military forces. One could ignore no centre of grav­ity, but should apply strongest force against the central core whilst applying successively less force to each of the other rings, so that fielded forces -- considered most important according to traditional doctrines -- receives least priority. Warden had also argued that all target rings should be struck simul­taneously (that is, in "parallel"), rather than one after another.

Warden's philosophy formed the conceptual basis of the Gulf War air campaign, a massive parallel operation that placed greatest priority on paralysing Iraq through strike attacks on national command and control systems, and then maintaining and extending the paralysis through attacks on systems essentials, infrastructure, national will and, finally, the fielded forces themselves. And, overall, this worked well, producing an air offensive far more effective than anything previously seen.

This, it seems, was clearly what General Short and his air planners wanted to achieve against Serbia after political constraints relaxed and permitted a "new" air war to commence in late April and early May. And by mid-May they were indeed achieving a "Warden's rings" type of parallel air operation against Serbia's leadership, systems essentials (especially petroleum and electricity), roads, bridges, railways and other infrastructure, television and radio stations and other targets involved in the maintenance of national will, and, within the boundaries of "reasonable risk" (SAMs and AAA were still a deadly threat), Serbian military and paramilitary forces.

By 4 June, General Michael Ryan, the USAF Chief of Staff, was able to claim: "Serbia's air force is essentially useless and its air defences are dangerous but ineffective. Military armament production is destroyed. Military supply areas are under siege. Oil refinement has ceased and petroleum storage is systematically being destroyed. Electricity is sporadic, at best. Major transportation routes are cut. NATO aircraft are attacking with impunity throughout the country. With the continued build-up of our aircraft and better weather, the attacks are intensifying and the effects are mounting."65 Victory was near, he added, and "inevitable".

The air campaign worked. By late May, after weeks of heavy, simultaneous attacks on virtually every essential component of national function, Slobodan Milosevic realised his country could take no further punishment. It was being bombed, to use a cliche, "back into the stone age." Milosevic began looking for a way out. Viktor Chernomyrdin, special Russian envoy to the Balkans, noticed the change as early as 19 May, when Milosevic seemed unusually attentive, lacking in bellicosity and defiance, and willing to consider options.66 By early June Milosevic felt unable to delay the inevitable and signalled that he was ready to accept the G8's terms of peace that Chernomyrdin and Finnish President Martti Ahtisaari were firmly insisting on. With Russia no longer "on his side," Milosevic folded. On 6 June, he and his parliament agreed to the terms. It took three more days before a "military-technical" agreement on the withdrawal of Serbian forces from Kosovo could be reached, but by 11 June NATO bombs were no longer raining down on Serbia. After 78 days, Operation Allied Force was over. A type of "peace" existed.

A Victory for Airpower Alone?

So, then, Operation Allied Force started poorly, but, after a major shift of focus and intensity at about its halfway point, ended well; as a large parallel campaign against Serbia's centres of gravity. But does even that second period represent a victory for airpower? Has airpower won its first war by itself? The answer can only come after two other factors are analysed: the role of naval and ground forces, and the impact of diplomacy on Milosevic's decision to accept terms of peace.

Airpower certainly dominated Operation Allied Force, and, even given the pathetic first month and the series of dreadful accidents, some of which almost prompted international escalation, the air offensive did turn out to be (to quote William Cohen's speech of 11 June) the "most precise application of air power in history".67 During 40,000 strike, reconnaissance, electronic warfare and support sorties throughout the 78 days, NATO hit Serbia with no fewer than 23,000 bombs and missiles, yet only 20 went astray and caused unfortunate accidents. More than one third of the munitions used were precision-guided, said Cohen, and the rest were "precisely dropped". Even more importantly, he added, "We flew for 78 days with no fatalities and only two planes lost."

Yet Operation Allied Force, despite its airpower dominance, was not actually an airpower-only effort. It also involved American and allied naval and ground forces, and these contributed to the eventual "vic­tory". And even if we are generous we have to call it a pyrrhic victory for the American leaders and other NATO politicians. They hoped for a short, coercive air campaign and found themselves in a destructive war that marked a significant departure from classic international legality, cost far too much, frequently risked escalation, seemed accident-prone, certainly went on far too long, and never achieved all the initial objectives. Even President Clinton acknowledged the contribution of naval and ground forces, noting on 11 June: "I know the Air Force is grateful for the radar jamming provided by Navy and Marine aircraft, the Navy [Tomahawk] cruise missiles fired from ships in the Mediterranean ... the Army and the Marine units taking care of the refugees."68

Clinton was right to mention the importance of naval forces, whose contribution has been overlooked in all the excitement over airpower. During the first night alone naval vessels had fired 55 cruise missiles and when poor weather grounded aircraft during the following weeks, naval forces continued firing Tomahawks.69 Vice Admiral Murphy, commander of both the US Sixth Fleet and Joint Task Force Anvil, recalled with some pride that, early in the war, bad weather made "his" Tomahawks "the only viable strike weapon available to NATO."70

Seapower assets also served as critically important platforms for airpower assets, noted Murphy. His Carrier Air Wing Eight (CVW-8) was "central" to NATO's ability to identify and destroy Serbian air de­fence installations. Navy F-1 8s and EA-6Bs "fired 147 HARM [anti-radar] missiles, or 47 percent of all HARMs fired by US forces. CVW-8 F-14s provided laser guidance to targets for US Air Force A-1 Os." His F-14s also served as forward air controllers and were airborne for approximately 50 percent of all close air support missions. More importantly, Murphy said, his carrier-borne aircraft had conducted over 3,100 strike missions, or more than 15 percent of NATO's total.

Murphy added that NATO's ability to conclude hostilities was only possible because it had a ground force capable of entering Kosovo as peace-keepers. Many of the troops, including marines of the 26th Marine Expeditionary Unit, came by naval transport. And naval forces would also bring in some of their urgently needed reinforcements, as well as their vast logistical requirements.

Ground forces had also played a key role throughout Operation Allied Force. In particular, NATO troops in Macedonia and Albania provided invaluable -- indeed, life-sustaining -- support for the hundreds of thousands of refugees streaming out of the combat theatre. They also closely guarded Macedonian and Albanian borders to discourage Serbian military units from attacking those states in order to root out the Kosovo Liberation Army or threaten NATO airfields.

It is still too early to detail the activities of special forces, although their presence in the Balkans was confirmed by the United States Special Operations Command (SOCOM) as early as mid-April."71 "Armed" with laser designators, they were almost certainly deployed in Kosovo to help NATO pilots select targets and direct laser-guided bombs through thick clouds (as they did in Bosnia in 1995).

One also cannot ignore the importance of diplomacy when analysing the cessation of hostilities, even though it is currently difficult, until further sources appear in coming months, to reveal the precise go­ings-on at meetings between Milosevic, Chernomyrdin and Ahtisaari.72 What does seem clear, how­ever, is that Milosevic, with his nation's infrastructure wrecked almost to the point where a humanitar­ian catastrophe might cause chaos and widespread misery, also felt deserted by his closest political allies: the Russians. By late May Chernomyrdin was essentially presenting him with an ultimatum: em­brace the G8 plan without delay, or forget any further support from Russia. Milosevic would be entirely isolated, and at NATO's mercy (which was clearly lacking; he was already indicted to stand trial as a war criminal).

Milosevic, the master of Realpolitik, saw the writing on the wall. Considering the G8 plan far more acceptable than NATO's five-point plan (based on the "accursed" Rambouillet accords), he decided not to prolong the war. He accepted the G8 plan, and was thus able to claim -- accurately -- that Serbia had not capitulated to NATO. On the contrary, his small nation had withstood massive attack by the world's only superpower and its allies for 78 days before accepting a G8 plan. That plan went nowhere near as far as the Rambouillet accords, and also took responsibility for Kosovo out of NATO's hands and en­trusted it to the United Nations.

Conclusions

To offer some concluding remarks, then, it seems clear, even from the fragmentary evidence available to date, that one can form some accurate early impressions of NATO's failures and successes. Operation Allied Force started very poorly indeed, mainly due to the Clinton administration's careless miscalcula­tion of Milosevic's resolve. It did not break after a day or two of bombing, as expected, which left NATO having to prosecute a lengthy air campaign without the means to do so. In the meantime, the flood of refugees from Kosovo hardened NATO's own resolve, but prompted a premature "phase 2" widening of targets. NATO's air force consequently dissipated its embarrassingly weak strength trying to fulfil all its new missions, all the while being hamstrung by excessive political constraints.

It was clear after a month that Operation Allied Force was failing, and something had to be done. And it was; at the NATO Summit in Washington the alliance's military leaders insisted to their political masters on a radical change in the war's prosecution. They gained their wish, and were finally able to intensify their offensive. They benefited from far more air assets, less constrained rules of engagement, and a more consistent and carefully constructed target prioritisation list. This package ushered in a "new" offensive; a parallel operation a la Warden's rings. It worked. Throughout the final five or six weeks, NATO aircraft systematically targeted Serbia's centres of gravity, effectively paralysing the nation. Un­able to tolerate further destruction, and suddenly aware of his total political isolation, Milosevic agreed to compromise. War ended (at least it has for now).

Suggesting that airpower did not win this war would be more than ungenerous; it would be wrong. Airpower did win NATO's war in the Balkans. After all, the crushing weight of airpower applied through­out May and early June forced Milosevic to abandon hopes of holding out indefinitely, or even for several more months. Once he realised in early June that Russia had lost patience, and would not intervene to stop the nation-wrecking bombing, he threw in the towel, accepting what he considered a compromise (and what NATO called a capitulation).

However, one should resist sharing the euphoria of uncritical airpower advocates. Airpower certainly won the war, but not alone. Naval and ground forces contributed significantly both to Operation Allied Force and to NATO's ability to implement a cease-fire settlement. This must not be overlooked. And the role of Chernomyrdin's diplomacy, which certainly shortened the war and made that cease-fire possible, should not be downplayed. Milosevic may well have held out longer had he thought the Russians would eventually intervene.

One should also note that airpower still has major limitations. Even the world's most technologically sophisticated aircraft and weapons systems frequently proved no match for Mother Nature. The Gulf War had revealed the decisive impact of weather on air (and even space) power, prompting frantic efforts to devise all-weather technology. This war in the Balkans reveals that scientists and technicians still have far to go.

The technocrats will also doubtless be busy trying to improve weapon precision. Even though Operation Allied Force was, as Cohen claimed, the most precise application of airpower in history, it still involved a series of dreadful accidents that included wiping out a civilian refugee convoy, scattering cluster bombs in a busy market place, wrecking a full passenger train, and destroying the Chinese embassy. At times, these horrific events threatened not only to undermine the much-proclaimed morality underpinning NATO's intervention in Kosovo, but also to escalate the conflict. This reveals the inherent problem with "smart" weapon technology. It raises public expectation that air attacks will cause no civilian casualties. In fact, as weapons becoming increasingly precise, public disgust at any accidents becomes correspondingly stronger.

And what of tactical airpower? Throughout Operation Allied Force very few aircraft and no helicopters engaged in air-to-air combat. More importantly, they conducted relatively few attacks on Serbian troops, armour, vehicles and guns. Terrified of the inevitable casualties, which are no longer politically accept­able, NATO air planners kept all air assets that needed to go low and slow (especiallyA-10s and helicop­ters) safely out of the combat zone. They feared them becoming "MANPAD magnets", a reference to the acronym for the Man-Portable Air Defense Systems that Serbia had scores of. As a consequence, NATO inflicted relatively little damage on Serbian fielded forces. Even if NATO's statistics prove reliable, and there are already reasonable grounds to consider them exaggerated, Serbian forces, armour, vehi­cles and guns were not "degraded" much during Allied Force's 78 days.

One can argue, of course, that NATO airpower had pinned down, cut off and immobilised those forces, rendering their existence irrelevant and their destruction unnecessary. That view is consistent with much of the theorising in today's attrition-shy military environment, which insists the Clausewitzian emphasis on Vernichtung (physical destruction) is anachronistic. The new logic says: if one successfully disrupts and dislocates an opponent, one need not destroy him.

That might be acceptable logic in a context where large civilian populations are not involved, but when enemy forces are continuing to terrorise and drive out hundreds of thousands of civilians, the destruc­tion of the enemy should take on more of a priority. Yet in this case NATO's gripping fear of aircrew casualties from low altitude flying left its airpower unable to prevent, stop or even slow the terrorisation of the Kosovar population. This has doubtless been a harsh lesson for many NATO leaders to learn.

Another bitter pill for NATO leaders to swallow relates to airpower's supposedly ideal attributes as a means of coercion. Operation Allied Force failed hopelessly as a coercive effort, ending Clinton's run of good luck with these types of things. He will doubtless think twice before threatening another misbehav­ing nation with air strikes, at least before he puts in place the Powell-ish policy that the use of military force (if threats fail) should be overwhelming and decisive.

This does not mean that Clinton or his successors will shy away from using airpower for coercive pur­poses. That's highly unlikely, given the United States' self-perception as the world's policeman as well as the remarkable fact that NATO won its "victory" without the loss of a single life. This, however, sets a dangerous precedent and will certainly result in a frightening misconception about the nature of modern warfare. No, the mixed successes of this campaign will mean that American leaders will simply try to do it "better" next time, avoiding the creeping goals, gradualism and political micro-management that al­most ruined and certainly prolonged this unfortunate war in the Balkans.