What is the Air Force really worried about: national security or job security?

Maybe the Pentagon wants the new stealth planes because they are technological marvels that will revolutionize air warfare. Or maybe it's just because without them, the pilots might lose their jobs.

"It is difficult at this remove to view dispassionately the short-sightedness of the prevailing official attitudes towards the new weapon," Peter Young, a British general has written of the decade following the invention of the airplane. "In 1912 the American colonel Isaac Newton Lewis fitted his famous aircooled machine gun to a Wright Biplane. Official reaction was tepid." That

attitude did not last. As tenaciously as military officialdom once opposed the notion that the airplane's time had come, it now resists the possibility that its time is about to pass. The Pentagon is about to commit itself to three new manned aircraft of record-setting expense: the B2 bomber, the Advanced Tactical Fighter (ATF), and the A12 Advanced Tactical Aircraft. (If Congress lets it; the House defense bill minimized B2 funds and cut the ATF altogether.) Though information about the B2 is slowly becoming available, hardly anything has been published about the ATF and A12, which combined will likely cost more than the stealth bomber program. ATF, intended to be the world's hottest "air superiority" fighter, is being designed by the Air Force. The A12, intended to attack land targets from carrier decks, is being designed by the Navy. Both will be stealth-that is, radar-elud-

ing-aircraft incorporating features similar to those of the B2. Because unusual hierarchies of secrecy cloak stealth programs, it is difficult to estimate whether these aircraft will succeed from a technical standpoint. For some time the military refused to confirm the existence of the B2 or the F117, a limited-edition stealth fighter built principally to determine whether radar-evading jets could be aerodynamically controllable. Today the Pentagon says precious little about ATF; even less about the A12. Both are black programs, meaning many references blacked out of public documents. (The Pentagon now shuns that term, preferring Special Access Required. Slang had mutated to the point that public-record programs were called white;" work areas referred to as "the white world" and "the black world.") Assessing whether the new aircraft can fly from the standpoint of cost effectiveness may be a different matter. The planes will incorporate so many costly features-some focused more on career security for the pilots' guild than on military necessity-that the world's richest country will be able to afford only a comparative handful.

Current plans call for 132 B2s and 750 ATFS. During World War II the U.S. fielded aircraft in quantities like these: 16,494 B24 bombers, 13,586 P51 fighters. The advent of the jet age did not immediately alter this equation: in 1951 the Air Force

acquired 6,300 aircraft, mostly jets, at an average of $1.3 million in current dollars. But by 1984 the service was spending an average of $40 million in current dollars for the mere 322 aircraft acquired that year. At about $530 million, each B2 will cost nearly 400 times as much in real terms as a B24. Each ATF will cost about 150 times as much, adjusted for inflation, as a P5 1.

If the only factor at play here were trading more expensive and glorious aircraft for steadily reducing numbers-Soviet military jet production has declined too, though not as rapidly as ours-the choice might be a toss-up. But while the number of combat aircraft declines, antiaircraft weapons are multiplying like mad. Recent advances in computers, sensors and manufacturing techniques have rendered surface-to-air missiles SAMS) increasingly effective. Improved air-to-air missiles are in the offing, and the miniature SAM exemplified by the U.S.-built Stinger has now been added to the quiver. Wielded by illiterate Afghan guerrillas, this $40,000 weapon overwhelmed Soviet high-performance jets. In the Pentagon's vision of the future a small force of sensationally expensive manned aircraft will do battle against a huge proliferation of such weapons.

Air-superiority fighters and superbombers have become as much symbols of American technical prestige as elements of military proficiency. Alternatives such as unmanned drones and affordable manned designs are spurned for their very worldliness. Yet a measure of subtle awareness in military circles that the end for traditional aircraft may be in sight is that already occupants of the cockpit are called not pilots or people but "humans." Airplanes, their primary role increasingly to serve as racks from which to launch missiles, are called "platforms." Around Air Force labs I heard many references to "the human on the platform."

The most renowned flying humans are those who sit in fighters. Unlike bomber or transport pilots, fighter jocks are glory hounds and individualists. The weapons they use pose little threat to noncombatants. Their quarry are other virtuosos who know the risks and even have a sporting chance to bail out if hit: more like the competition than the enemy. Now there's a lifestyle.

Thus the romance of the fighter jock dies hard. Of equal importance, while the case for $530 million for any bomber is difficult to make, the ATF and A12 projects invoke shades of gray, with many points in favor of such aircraft. Are there enough? The scenario pilots love

It's the year 2000. Fighting breaks out along the

Central Front; rationality or terror prevents the adversaries from resorting to nuclear malice. Some ATF squadrons prepare to lead a raid, the objective being a Warsaw Pact airfield. A group of F15s and F16s, frontline fighters from the 1980s, crosses into Warsaw Pact territory flying "terrain masked," down low where ground radars

cannot see them. these planes have their own radars and electronic jammers cranked to full blast. During the first phase of the strike, ground guns and SAMs pose the main hazard. But new Soviet interceptors operating in the sanctity of Warsaw Pact airspace can use improved radars to spy attackers in the otherwise elusive low position. So even if the F15s and F16s make it past the ground weapons they still face jeopardy.

Concurrently ATFs cross enemy lines at a majestic altitude, high above the range of battlefield weapons. Because ATF has stealth, it is reasonably secure against radar detection at altitudes conventional aircraft would find suicidal. While the B2 is a subsonic aircraft, in order to be capable of intercontinental travel, the shorter-range ATF cruises at around mach 1.5. Unlike contemporary fighters that have impressive top speeds but rarely use them because then fuel consumption goes off the scale, new "dry supersonic" motors enable ATF to sustain mach velocity without activating an afterburner. Thus these fighters transit the lethal envelopes of defenses in less time than current aircraft. The ATFs are operating with radars and jammers off, so as not to betray their presence. Each is receiving a rich stream of information from U.S. sensors in space and at distant locations, while employing

onboard passive technology such as infrared scanners and electronic eavesdroppers, that leave no clues of the ATF's presence.

In ATF cockpits, "data fusion" computers put the streams of digital clues together and predict where the hot Soviet interceptors can be found. ATF's role is to eliminate them. "ATF isn't necessarily going to shoot everything it sees," says Colonel James Fain, a manager of the ATF development program. "The Threat is beginning to build aircraft with true lookdown, shoot-down capability, and the purpose of this airplane is to remove that threat." In Pentagon usage, threat means a specific tactical obstacle, whereas The Threat means the Russians.

Aboard ATF is the new Amraam missile, a "beyond visual range" (BVR) weapon for shooting at planes as much as 50 miles away. The plan is that long before Soviet fighters can draw close enough to target an ATF visually, it will unleash a hail of Amraams. Ideally not only does the enemy never see ATF, ATF never sees the enemy. Whenever the ATF launches a missile, its bay door opens-weapons will be stowed internally to reduce radar echoes-and the irregular shapes thus exposed render the plane momentarily visible on electronic scopes. But after the missiles are away, these doors shut and the fighter changes course sharply, maneuvering at random. Any missile launched toward ATF's fleeting image fizzles off harmlessly.

If a Soviet interceptor jumps an unsuspecting ATF, the pilot defends himself with heat-seeking missiles and cannon fire, the same dogfight weapons employed by existing U.S. fighters. If a superior force of Soviet aircraft arrives, the pilot exercises discretion and bolts, using sustained supersonic speed to outrun pursuers who can go into afterburner only momentarily.

This is the principle scenario the Air Force has in mind for the new ATF. Underpinning it are two critical assumptions. The first is that ATF will operate mainly in the other guy's skies, not ours. NATO doctrines with names like Follow-on Force Attack and Win Early, or the generic name interdiction, emphasize immediate counterattack behind the lines of a Warsaw Pact advance. Such ideas are parcel of the competitive strategies" doctrine George Bush praised during the 1988 campaign. Some inconvenient facts Interdiction is an important element of Western strategy in part because the idea of carrying the fight to enemy turf is more palatable to NATO residents than other alternatives. Partly, too, it is because tacticians are more comfortable with counterattack than

defense. Though military theory says the defender holds the cards, in recent wars aggressive tactics have often prevailed: for example, Germany and Japan overran numerous entrenched positions early in World War II, later to be overrun in turn after they became defenders. Conventional interdiction of targets such as command centers holds out the hope of what might be called constructive combat: dealing a few decisive blows that bring hostilities to a prompt conclusion, rather than slugging it out at the line of scrimmage, which cannot help but be immensely destructive to both sides. On the down side, deep counterattack engages the military imagination because it is a maxtech assignment. Only the fanciest and most expensive airplanes will do.

The second driving assumption of ATF is that beyond-visual-range weaponry will work. The U.S. has possessed BVR weapons since the 1950s, principally an air-to-air missile known as Sparrow. In Vietnam, its performance was a major disappointment. More recent use of updated Sparrows suggest this weapon still falls well short of its sales pitch; during the recent encounter in which Navy F14s shot down two Libyan fighters, three Sparrows were fired, under conditions suited to the missile. Only one hit. Existing airborne radar missiles work by homing in on echoes from the nose radar of the launching fighter. This means any pilot firing a Sparrow must keep his radar on, announcing his position; he must further train his concentration on the target and restrict his aircraft's flight path as long as the Sparrow is flying, rendering himself vulnerable to attack from other aircraft or missiles in the area, and often ending up having closed to within visual range anyway.

Amraam, by contrast, is a "fire and forget" device with a full radar set of its own. As soon as a pilot launches one he can peel away and let the missile worry about itself. Amraam will cost much more than Sparrow, as much as $635,000 per missile. But in this case, a Pentagon price increase appears justified. Assuming Amraam works (it's in testing), half the BVR problem will be solved. Whether the other half can be solved is something else again. Since by definition a pilot firing from beyond visual range cannot see what he's aiming at, in a swirling air battle, there is enormous risk of shooting down the good guys. "This has been a somewhat sensitive issue," Colonel Fain allows. Researchers have attempted to build gadgets that can identify friend from foe using zoom-lens cameras; highly accurate radar bursts; even by analyzing the harmonics of jet exhaust to determine who built the engine. No system has been sufficiently effective to enable pilots to trust their BVR weaponry. "I wish

I had a penny for every time I've taken off a carrier deck loaded down with [BVR missiles] knowing darn well I'll never get clearance to fire one because I

won't know what I'm shooting at," Capt. Stewart Schmidt, an F14 pilot, said.

The U.S.S. Vincennes shot down an Iranian Airbus because the $600 million Aegis radar system on the ship was unable-not through failure, just technological limitation-to make as simple a distinction as whether the BVR blip being tracked was a fighter plane or a commercial airliner. An Aegis radar is far more powerful than any sensor ATF will be able to carry; and Aegis's inability to determine the nature of the target held true even after the Airbus had closed to within 20 miles. ATF plans call for pilots to make judgments about blips from considerably greater distances.

Many aircraft now carry transponders that in theory permit electronic "interrogation" separating ours from theirs. But such devices are notoriously finicky; a contributing factor in the Airbus tragedy was confusion aboard the Vincennes regarding what type of code the interrogator system was reading. Such confusion, in the ideal conditions of tracking a single target, bodes ill for blip identification during the madness of general combat. Above the Central Front will be aircraft of NATO allies using several different, incompatible interrogation systems. Long-running negotiations between the Pentagon and NATO, aimed at a common electronic protocol for identification transponders, recently broke down in bickering over whose standard would be adapted.

ATF will have a new integrated fire control system able to combine readings from many sources to provide improved friend-or-foe cues to pilots. An official developing this system, James Kocher, told me that positive identification cannot be guaranteed. "It will raise the level of pilot information [about BVR targets] from uncertainty to educated guess," Kocher said.

Rationales for the high cost of ATF hinge on the ability to fight from beyond visual range, for if ATFs end up in dogfights, money will start pouring down the drain. The two engines powering ATF will have thrust-vector nozzles resembling those on the Harrier jump-jet. By manipulating exhaust flow, these devices will afford pilots maneuverability compensating for the inherent clumsiness of stealth aerodynamics. ATF is expected to have the same dogfight handling power as the F16. But the same maneuverability as current fighters, plus the same close-in weaponry, means ATF won't be notably better at traditional combat than what the U.S. already has. What it will be is outnumbered. Nobody can know how many of any weapon the

United States requires, nor whether current talks on European conventional-anns reduction will result in treaties reducing the importance of this whole issue. But based on the information available now, the numbers for ATF don't look good. Depending on who does the calculating, Warsaw Pact tactical aircraft outnumber NATO's two or three to one.

Focusing U.S. investment on ATF will further widen the gulf. Yet most commentators believe that the use of modern jets in Vietnam, the Middle East and the Falkland Islands shows pilot skill, tactics, and numbers to rank with or exceed technology as determinants of victory. Even Colonel Fain agrees. "The best fighter in the world, drawn into a dogfight against numerically superior forces, will eventually lose," he says.

Costs: flyaway vs. runaway The goal of 750 ATFs was arrived at via Air Force computer simulations of future aerial combat. Officials concede the number is low even by standards of high-performance jets: for instance, the U.S. built 5,300 F4 Phantoms. And by the time ATF appropriations wend their way through Congress, the total may be lower still.

Recognizing the smallness of the planned purchase, the Air Force buzzword about ATF is Ieverage." Though there won't be many ATFS, planners concede, their ability to take on the best Soviet fighters will leverage existing F15s and F16s into the next century. At one point an Air Force press release declared that the new plane would achieve an ,exchange ratio" of 10 Soviet fighters destroyed for each ATF lost. During the Vietnam War, U.S. pilots, working with a pronounced technological edge, could achieve no better than an exchange ratio of 2.6 to I over North Vietnamese fighters.

Significantly increasing the quantity of ATFs is not an option; the aircraft will be too expensive for that. The Air Force has established a $35 million cap on ATF flyaway" cost, and by all accounts Tac Air, the project's sponsor, is struggling earnestly to honor that objective. But the true price of ATF is certain to be higher. The $35 million is quoted in 1985 dollars, which equates to $44 million today. Flyaway cost is a stripped-down figure. In 1987 the Air Force claimed a flyaway cost of about $25 million for an F15; payment per F15 actually delivered was about $37 million. Most important, the Air Force already concedes that the $35 million figure will be arrived at by assuming unspecified savings in future years. During early production even flyaway costs will exceed the ostensible cap; the Air Force will warrant to

Congress that such high prices will be offset by savings down the road. Programs are often low-balled in this fashion, and somehow the huge future savings have a habit of failing to materialize. For these reasons some Pentagon documents list an ATF acquisition cost of $50 million per plane; the final figure may hit $60 million. Logically this would almost have to be so, because ATF will be produced in quantities too small for economies of scale.

These ATF price estimates include a proration for research and development, a calculation that drives the military to distraction. Air Force officials currently are heatedly complaining on Capitol Hill that the troublesome $530 million sticker price for the B2 would decline by approximately a third if R&D expenditures were not rolled in. But somebody's got to pay for R&D. Virtually all commercial aircraft are priced to reflect what the company spent on development; the $125 million price of the new Boeing 747- 400 certainly includes R&D. Military officials can't opt for a project like a stealth jet, for which they know R&D expenses cannot help but be high, then later claim these costs somehow do not count. Taking high R&D into consideration is essential to making rational choices about a project's worth relative to the defense resources consumed.

Defense contractor behavior reinforces the upward cycle of cost. During the 1950s the U.S. designed, tested and built no less than 18 fighters, most of them supersonic jets: in the three decades since just seven new U.S. fighters have gone into production. During the 1950s unit costs were lower, profit was in volume. Then military contractors began to merge. Fewer companies meant less competition; which led to higher price per unit; which caused the Air Force to reduced buys; which encouraged planners to think in terms of loading every conceivable device into those aircraft it got. As a corollary, today the profit is in aviation electronics: such "avionics" now account for 40 percent of Defense Department spending on aircraft. Unlike airframes and engines, these microchip concoctions have high markup. One defense executive describes current flying machines thus: "Not so much airplanes as self-propelled avionics suites." Kids on the ground Contractors love avionics because they know that while Congress will wring its hands interminably about building an airplane, approval for black boxes is nearly involuntary: if only because hardly anyone understands what the boxes do, let alone which ones are justified and which aren't. Black boxes can also be retrofitted to existing planes, providing a steady

flow of business. At $5 million each, a new fighter avionics package called Lantirn costs more than an entire F4 Phantom did about a decade ago. Whether maximum technology makes for the best weapon is the subject of unceasing debate in military circles. Technology proponents point out, for instance, that although the ATF's new system of practical supersonic speed won't be cheap, ferrying back and forth to base at blazing speed will allow ATF to generate more sorties than current fighters. Other new features such as "fail soft" circuitry should enhance ATF value. Fail-soft means that when components incur battle damage they will automatically reset for limited operation, instead of just winking out as current systems do.

New technology should also make ATF more reliable than current fighters. It's an encouraging sign that after years of cannibalizing some high-tech aircraft to keep others running, the Pentagon has gotten religion on the subject of reliability. The goal for ATF is double the dependability of an F15. "For ATF reliability and maintenance have been equal to per- formance as design goals," Colonel Michael Walton, director of an Air Force design bureau, says. "Five years ago no military laboratory could have said that."

Proponents of maximum aircraft technology further note that stupendous exchange ratios have been recorded by Israeli pilots using modem U.S. fighters. And they cite the most powerful argument of all: if we're going to be outnumbered, can we have anything less than the best possible aircraft? Don't we owe it to our pilots?

This last argument is circular, however. The reason our pilots are outnumbered is that we insist on superplanes. Such a choice may be fine for Israel, which has only a tiny land area to defend and need not care about the cost of its U.S.-subsidized equipment. Israeli pilots also do not fight outnumbered: Israel has more combat aircraft than Syria and Jordan combined.) Whether the approach is best for the United States, which must defend thousands of miles of borders the world over, is another matter. The principle objection air power advocates express to intermediate fighters, of which the $15 million Air Force F16 is the leading example, is that they are fundamentally defensive: best in interceptor roles near their own turf, not so hot for missions deep into enemy territory. Though defense-not-conquest is every U.S. officer's objective, defensive strategy just doesn't sit well with the military culture. But suppose the Soviets deep-strike our side? In that case several sneaky little planes may be of greater value than one gloriously powerful ATF. Pentagon planners tend to downplay this concern,

suggesting that few Soviet fighters could survive the excursion across our SAM belts. "The great problem for pilots in the year 2000 will be weapons from the ground," Lt. Colonel Ted Wierzbanowski, an F15 pilot, says. "You're going to have thousands of kids on both sides shooting cheap missiles at you. Other fighters may be the least of your problems." This in turn leads to a subject the Air Force really doesn't like to talk about: the modern antiaircraft missile.

Son of SAMs Guided missiles are a newer science than aircraft: the first attempts to build them did not come till the 1950s. These missiles fall into two basic categories, radar-homing and heat-seeking. During Vietnam, radar SAMs crude by today's standards inflicted consistent losses on U.S. forces over North Vietnam, though bombers were accompanied by scores of suppression" aircraft: jammer aircraft and strike fighters bearing special weapons to attack SAM installations. Some 1960s-vintage SAMs in the possession of Muammar Qaddafi caused such concern to U.S. commanders planning the 1986 Libya raid that the Navy had an estimated four suppression aircraft in the air for each plane that actually delivered a bomb on Libyan soil. Today the U.S. Army has thousands of Hawks, a radar SAM, and is building several thousand new long-range radar SAMs called Patriot. The great U.S. weapon fiasco of the 1980s, the Divad automated antiaircraft gun, is being replaced by three types of small SAMs that incorporate clever technological variations. For their part the Soviets have nearly 5,000 SAM launchers near Europe, plus another 9,000 ringing their home borders. According to the Defense Department, in the last decade the Soviets have built 140,000 missiles for these installations. Heat-seeking SAMs and air-launched heat-seekers like the Sidewinder have proven close to unstoppable under some circumstances: their only drawback is a shorter range than radar missiles. During the 1973 Middle East war, some 93 percent of the Sidewinders launched by Israeli pilots struck Arab aircraft, perhaps the best performance ever by a precision-guided weapon. The Soviet Union now builds similar missiles. Pilots can sometimes duck a radar missile by evasive action; heat-seekers don't give up as easily. Radar missiles depend on transmissions that can be jammed, and announce their approach to an aircraft's "fuzzbuster." But heat-seekers are passive. The only way a pilot even knows one is coming is if he sees it fired or his wingman spots a smoke trail and cries "Tallyho!"

Miniature SAMs of the 1960s and 1970s worked poorly, so air power advocates tended to write them off. A new Stinger version perfected in the early 1980s changed that. These missiles were devastating in Afghanistan; both armored helicopter gunships and modern strike fighters proved vulnerable to them. In the final years of the Afghan occupation Soviet planes scattered flares everywhere they flew, hoping to confuse Stinger sensors, or operated above 10,000 feet to stay beyond Stinger range. Ordinance released from such altitude almost always misses, so the Stinger gunners prevailed without even shooting. Because their passive sensors employ few moving parts, heat-seekers are cheap: Sidewinders cost about $55,000 and Stingers about $40,000. And as years pass, the scales may tip further in the direction of SAMs. Putting each new technological gimmick onto a small, unmanned missile is comparatively cheap, while adding the appropriate countermeasure to a large manned target usually means serious money. Needless to say the thought of $40,000 missiles shooting down $60 million ATFs or $530 million B2s is not reassuring. Backers of these projects say their aircraft will withstand missile onslaughts by foiling radar guidance with their stealth and by flying high above the limited range of heat-seekers. Stealth aircraft incorporate engine exhaust systems that reduce heat signature; devices broadcasting scrambler pulses on infrared frequencies may also help. Meanwhile the Air Force riposte about North Vietnam losses is that though many U.S. planes fell, huge numbers of SAMs were expended in the process; only about three percent of enemy missiles struck home. But one cannot help fearing the rising curve in antiaircraft technology represents a threat air power advocates are simply trying to wish away. Here's the bottom line: If about three percent of the antiaircraft missiles in the world today find their mark, the skies will be swept clean.

A future air superiority force based primarily on SAMs might be more fearsome and cost-effective than one based on manned aircraft. Don't expect the Air Force to advocate such a shield, however. Not only would that deprive the service of seats for pilots. Ground-based antiaircraft weapons fall under the jurisdiction of the the Air Force's second-most feared foe, the U.S. Army.

Diving drones Pentagon planners who normally drool over any new technological prospect have been dragging their feet against a flying machine recently made practical by advances in microelectronics, composite materials

and small turbine engines: the miniature drone. Drones are another annoying threat to the pilots' franchise. Their existence is such a touchy subject there is even a dispute regarding what to call them. Proponents say "remotely piloted vehicle," emphasizing that microcomputers and digital transmission now make it possible to control RPVs from ground stations, not just send them on preprogrammed courses. Air power believers can't stand the implications of "remotely piloted." They insist on saying "drones," connoting the plodding dim-wittedness of pilotless flight.

Israel has had relatively cheap drones in operation for a decade, using them for reconnaissance and electronic warfare. Until recently, of the U.S. services, only the Army showed interest in similar equipment, though its first drone, Aquila, was a spectacular dud. The Navy, acutely conscious that manned aircraft are what justify its beloved supercarriers, resisted drones until it lost two fighters during a 1982 strike against the very Bekka Valley positions that Israeli warplanes, assisted by drones, hit without losses. Today the Defense Advanced Research Projects Agency (DARPA), a secretive Pentagon offshoot that occasionally takes pleasure in twitting the powers that be, is tinkering with a low-cost drone called Amber which has demonstrated an ability to stay aloft for 30 hours on one tank of fuel. A defense contractor with the incredible name General Atomics has shown prototypes of a low-cost propeller-driven drone that could fly 300 miles and deliver a 300- pound warhead. And after considerable hemming and hawing the Air Force is pursuing drones that could destroy SAM guidance equipment by circling over a battle waiting for a radar to be turned on, then diving into the transmitter and exploding. Note that the drone the Air Force likes is one that would counter a threat to manned aircraft.

Many justifications for advanced aircraft involve their use as "platforms" for smart munitions that, launched far behind enemy lines, will score direct hits on high-value targets. So far smart munitions need help from Stanley Kaplan, as their mediocre test scores in actual uses such as the Libya raid attests. Beyond smart lies the brilliant munition: ones able to self-guide so that a strike aircraft could fire and forget them, preferably from standoff" range. No brilliant munition has yet been fielded, though several are being researched. If any is perfected it might, like Amraam, finally make real a technology whose practical value thus far has been vastly inflated. Now here's the dilemma. Suppose brilliant munitions do turn out to work. Then why employ megabucks manned airplanes to baul them to the launch point? Instead you might put the brilliant

seeker head in a General Atomics drone, and put the pilot out of business. From long johns to astronauts Just when things really start to get good they become obsolete. Consider what the military aviation subculture has been through. First, general mockery of its ideas. Next, death-defying performance in the rickety crates of World War 1. Following this sacrifice a thankless phase of internal exile culminating in court martial for the air power visionary Billy Mitchell. Next, tremendous achievements in World War 11, but at precious cost in battles such as Ploesti. Even then planes were still built like flying canoes. World War 11 pilots wore long underwear and silk scarves to keep from shivering. But in the postwar period things began to look up. A new service, the Air Force, was formed, and within it, an interest group controlled by pilots. Planes became powerful, glamorous and also safer-deaths per flying hour declined. Unlimited appropriations seemed to flow the pilot's way-merely training an F15 pilot, an easily overlooked factor in aircraft funding calculations, costs $7 million. And the pilot class became exclusionary. Today almost everyone in a military jet is an officer, with the fancy bombers and fighters flown by men ranking as high as Air Force colonel or Navy captain. Today pilots receive decent compensation (about $40,000 a year for a major in his early thirties), cash bonuses and flight pay, half-pay retirement after 20 years, plus great stature for their wives and families within military society. A key cultural difference between the jet-dominated services and the Army is that in the latter, even the costliest helicopters are flown by warrant officers," noncommissioned personnel who enjoy little institutional prestige and few lifestyle perks. Concomitant with the improvements in military aviation, the civilian airline industry developed. Although early generations of military aviators tended to go into non-flying work or became crop dusters and barnstormers, today a jet jockey who sours on military life can start a lucrative second career flying for the airlines. In short, what was once a crummy way to make a living has become a sweet deal. Now along comes the chance to build ultimate flying machines like the B2 and ATF. These are the airplanes a World War I ace would have given his left arm just to see. Modem electronics will make flying them a dream come true. Only a select few-astronauts practically-will qualify for the privilege.

It's easy to understand why the desire to preserve this exerts an emotional pull on the Pentagon, Congress and the public. There ought to be nostalgic lore attached: American citizens owe a great deal to the culture of military aviation.

But lore is one thing, logic another. Unless there's some very basic fact about the B2 that the Air Force is withholding to the detriment of its own funding prospects, or unless the Pentagon is willing to trade B2 funding for termination of another new strategic weapon, this program seems a deserving candidate for cancellation.

And unless ATF really is, as its price suggests, four times better than an F16, the money to be lavished on acquiring a small contingent of these planes might better be spent on a larger force of less magnificent aircraft. The additional development funds scheduled to be invested in ATF, about $6 billion, could readily be shifted to designing a more modest fighter that takes into account recent technical advances but does not try to be the most marvelous flying machine of all time. Higher numbers of a different fighter would even ensure jobs for pilots.