Last month Defense Secretary Weinberger more or less killed DIVAD, the $4.5 billion automated antiaircraft gun that had come to symbolize rigged tests, unrealistic plans, undisciplined spending, and technology that doesn’t work. Weinberger said that production of DIVAD—suspended since last winter—would remain suspended and that only money for a final test next year would be left in the budget. Technically DIVAD is still alive, but all signs point to the project’s going out with a whimper rather than a bang. This would be fitting, since it was never able to produce a bang on the battlefield.

DIVAD is the first production-level weapons program to be shut down since the Army’s Cheyenne attack helicopter was canceled 15 years ago. The B1 was “canceled” in 1977 but brought back to life: the opposition had centered on political, arms control, and budgetary concerns, not its flight performance. DIVAD is the first weapon since Cheyenne—and one of the very few since World War II—to be stopped because it doesn’t work.

But what did it take to kill the program? Things had to go haywire to such an absurd degree that even institutional Washington couldn’t ignore DIVAD any longer. In order for DIVAD to fail, all of the following had to occur. The project had to: begin existence in the realm of fantasy, with nearly all real-world combat conditions, especially the need to hit maneuvering aircraft, assumed away; proceed to flunk its tests, even what around the Pentagon are known as “magic” tests, by huge margins; be condemned in reports by the General Accounting Office, the Congressional Budget Office, the Office of Management and Budget, the Heritage Foundation, the Department of Defense inspector general, and two military-run Pentagon offices; have its contractor’s performance officially declared “totally unacceptable” by the Army and have $84 million in contractor overcharges claimed by an official Defense Department report; have information on DIVAD’s failures withheld from the selection board that authorized production; have a former executive of the company that makes DIVAD sitting on that board; have the gun, which is supposed to be “interoperable” with NATO antiaircraft weapons, use ammunition that isn’t interoperable; have its costs, $6.8 million to $8 million per unit, rise to four times the cost of the tanks it is supposed to defend; and have other, proven antiaircraft weapons that cost a fraction as much be rejected.

And even this wasn’t enough to do DIVAD in. Had the Army not made two comical mistakes in rapid sequence, the weapon might still be in production. First, ordered by Weinberger to stage a “realistic” test that would prove DIVAD’s worth once and for all, the Army instead staged a phony test in which “targets” were equipped with radar amplifiers, roughly the equivalent of “testing” a bloodhound by strapping a beefsteak onto the quarry’s leg. Then it announced that DIVAD’s ineffectiveness would be cured by adding $1 billion-worth of Stinger antiaircraft missiles to the gun: in other words, the antiaircraft weapon would carry an antiaircraft weapon to protect it from aircraft. These two developments caused DIVAD, which until recently had lived the charmed life of the constituent spending program, to become a subject of ridicule in Congress, and after that to become a media event. Finally Weinberger was forced to act.

Usually weapons systems waft through Congress because they are difficult for the uninitiated to understand, difficult for the press to cover in single paragraphs, and difficult even for the openminded to assess. We hold our breath and hope that the generals know something we don’t. If it takes a total breakdown of every aspect of a weapons program to get action, what does that say about the rest of military spending, where choices are not so black-and-white? If Congress lacks the nerve to kill spending programs until problems reach the point of general ridicule, how can it be trusted to make public-spirited decisions on programs outside the public eye? If the secretary of defense won’t be honest with himself until he stands in danger of being seen as a buffoon—if money is being thrown indiscriminately at weapons whether they work or not—what legitimacy does the Reagan defense buildup have? And if the president is too timid about offending the Pentagon brass to tell them to clean up their act when it comes to obvious problems in peacetime, how will he command them in time of crisis? Will we learn anything from DIVAD?

These are essential questions in the wake of DIVAD’s semi-demise. After Cheyenne was killed in 1969, it merely was replaced by a new project designed to do the same thing at even higher costs—the AH64 Apache. Cheyenne’s purpose was never questioned, nor was the institutional framework that had produced a multibillion dollar weapon that didn’t work. Heads did not roll. More money was given to the same people to spend in the same way, with no new accountability imposed, and Congress went right back to sleep. Only the acronym changed. DIVAD can be different. It can be the source of lessons that show the way to greater national strength at a lower cost.

Defending the Defender

First, a look at the details. This will take a moment, but a grasp of the technical reasons behind DIVAD’s failure makes the magnitude of the political disaster more evident.

No one disputes the Army’s need for a new antiaircraft weapon. Its present antiaircraft gun, Vulcan, is outdated. Its large missiles (SAMs, or surface-to-air missiles), such as the Hawk, are effective, but too elaborate to move around during combat and thus useful only for protecting rear-area installations, not tanks and troops. Miniature SAMs, such as Stinger, which can be fired off a man’s shoulder like a bazooka, are carried easily to the front but are not as effective as large missiles.

Stinger has been’ much in the news lately because of the upheaval caused by its sale to Saudi Arabia. What exactly Stinger does was skipped over for the most part during the debate, politicians and reporters alike taking for granted that because it is modern and expensive, it must be an awesome instrument of destruction. A heat-seeking missile similar to the highly effective Sidewinder, Stinger is sound in concept but has a severe practical limitation—in order to be light enough for a soldier to lift, Stinger contains hardly any fuel or explosives. The small fuel supply limits Stinger’s range to about four miles under ideal conditions and less under combat conditions, because if an enemy aircraft makes a sharp turn to “duck” a Stinger, the missile will rapidly use up its fuel trying to follow. Such sharp evasive maneuvers are a standard tactic of ground attack aircraft. Stinger’s warhead weighs only 2.5 pounds, about the size of a hand grenade, meaning that even a hit is no guarantee of destroying the target. None of these are flaws or defects, but simply the engineering limits of a toy-sized machine. Advocates of high-technology weaponry like to point out that the Soviet SA7, a miniature SAM, hit 30 Israeli planes during the 1973 Middle East war. They don’t like to add that only two of the planes went down.

Stinger (like all SAMs) is of limited value also to field commanders since it cannot be used against ground targets, notably tanks. Because high-performance aircraft are the most glamorous and sophisticated threat facing U.S. divisions, it is assumed that they are the most serious threat, but this is not so. “At the front lines, for every one time you fire at the sky you fire 50 times at ground targets,” says a former high Defense Department official. “Tanks, artillery, infantry, and ground guns are far more likely to kill you than aircraft.”

DIVAD (short for “division air defense”), conceived during the mid-1970s, was supposed to correct all these deficiencies—providing an antiaircraft weapon that could go wherever tanks go and participate in ground fighting when not firing at the sky. Ideally it would be as fearsome as the German 88, which was one of the most successful weapons of World War II because it was equally effective against aircraft or tanks.

The hundreds of aircraft shot down by German 88s and similar guns were flying approximately the same speeds and maneuver patterns as attack aircraft fly today. While the maximum speed and G-forces airplanes can achieve has increased substantially since World War II, high speeds cannot be used at tree-top level. Even the most “capable” supersonic fighter must drop to subsonic speeds of no more than 450 knots for a ground attack run. All the aircraft shot down during World War II were hit by guns aimed visually, without electronics. Ninety-one percent of the modern U.S. jets lost over North Vietnam were shot down by visually aimed guns, many of them World War II leftovers. From the ground gunner’s standpoint, modern aircraft are not much different from what they were 40 years ago.

Despite this, the Army insisted that DIVAD be an ultra-technology weapon that would find targets with radar and aim itself using a fully automated computer system. Soldiers would be required to do little more than switch DIVAD on and pull the trigger—when the computer instructed them to. This decision meant that even if DIVAD worked, it would be so expensive only a small number could be purchased; the added protection available to U.S. troops would be small.

The Army has tried to build DIVAD several times before. Following World War II it tried to win funds for a 75mm radar gun calledSkysweeper, which was aimed by analog computers and did poorly in tests. During the 1960s it experimented with a radar-computer 37mm weapon called Vigilante. Vigilante was canceled after tests showed it could not hit a maneuvering aircraft—the system could hit aircraft only if they flew straight lines. An internal Army evaluation of Vigilante, declassified last year, concluded that radar-computer directed antiaircraft fire was a technological impossibility and recommended that the Army instead build a large number of simple visually aided cannons. But rather than learning the lesson of its own attempts, the Army’s antiaircraft hierarchy became even more firmly dedicated to the idea of getting an “all singing, all dancing” weapon, as one former Pentagon official described the goal, to restore the prestige lost by the failures of Skysweeper and Vigilante.

In the early 1970s the West German military won money for a self-propelled radar-computer gun with the ominous name, Gepard Flakpanzer. The Flakpanzer eventually turned out to cost $5 million and be ineffective. Nonetheless the Army was envious. Then, during the 1973 war, Israeli troops captured a much-rumored Russian super gun, the ZSU23/4 Shilka. Shilka was everything Vigilante would have been—radars, computers, automation!!! Taken to MacGregor gunnery range near Fort Bliss, Texas, Shilka was evaluated during a series of tests code-named HITVAL. The “super weapon” proved so unreliable that engineers could barely keep it running for the tests to proceed; hours or days often separated firings. More importantly, Shilka was unable to hit maneuvering targets, only drones flying straight, predictable lines.

Even so, the Army’s envy grew worse. The Threat had something we didn’t have! All too often procurement decisions are driven by hyperbolic assessments of Soviet equipment, and so it was with the Shilka. Never mind that it didn’t work; just knowing that The Threat’s procurement bureaucracy had talked its politicians into coming across with funds where our boys had been refused was insult enough. Shortly after HITVAL was completed, the Army composed a “requirement” for DIVAD.

Hitting a maneuvering target with an automated system hasn’t worked for Shilka or DIVAD because it cannot be done. Once again this is not a flaw or defect, but a limitation of technology. Sometimes military projects stand or fall not on questions of detail but on fundamental principles of physics, and this was such a case. While radar can tell with high precision where an airplane is, it cannot tell where the airplane is going—which is the vital part of aiming at a moving target. Computers, however powerful and fast, cannot compensate. “Humans,” as the Pentagon now often calls them, actually have more information at their disposal than computers when aiming a gun visually. Computers see targets as featureless blips; a gunner sees what type of aircraft he is aiming at, sees what position its nose and flaps are in (indicating which way it is about to turn), knows what type of tactics the enemy generally flies, and can use his intuition.

“Of course DIVAD will miss if the target jinks [maneuvers],” Lt. General James Maloney, head of the program, told me two years ago. “No computer can handle a jinking target .” There is no shame in failing to do that which cannot be done. But all organizations have difficulty admitting to themselves what they cannot do, and most of all the modern Pentagon, which has staked its prestige on the notion that high technology is omnipotent. So DIVAD went ahead even though its planners knew from the start that the weapon would be useless against maneuvering attack planes, the most serious aerial threat to troops.

The attack plane problem was assumed away; the Army, one of General Maloney’s aides told me, had determined that Russian aircraft would not use evasive maneuvers (let’s hope the Kremlin has been informed) and that shooting helicopters would become DIVAD’s primary assignment. This assuming away of reality has caught on to an astonishing degree. In congressional testimony Army officials say with a shrug that DIVAD’s inability to handle jinking targets is not an issue because everybody knows DIVAD can’t handle jinking targets, recalling the California gubernatorial Candidate who, when ties to the mob were revealed, responded by declaring, “I never said I was tough on crime?’ Most press accounts have also skimmed past DIVAD’s uselessness against maneuvering planes. So, freely admitted, the weapon’s most basic defect has in a weird way become noncontroversial.

Cooperative targets

By 1980 two competing DIVAD prototypes had been built, one by the Ford Aerospace division of Ford Motors and one by General Dynamics. A “shootoff,” officially called DT/OT, was held to judge the winner. Neither gun did particularly well considering their expense and the fact that they were not tested against maneuvering targets—only drones flying lazy, straight lines were employed. The guns were shooting, in the words of a retired Air Force fighter pilot, against “extremely cooperative targets.” Though neither prototype did well, Ford’s was least effective, recording only nine hits to 19 for the General Dynamics contender. Ford’s longest range hit was less than half as long as General Dynamics’, and the Ford prototype never managed to hit an aircraft drone, even one on a sitting-duck flight path. It could hit only helicopters, the easier target.

You’ve guessed by now that Ford won the contract. On Ford’s team were four recently retired three-star Army generals, all from the procurement and air defense branches. Every contractor hires retired procurement officers to smooth its path, but four recent three-stars is high even by Pentagon standards. In addition, as soon as it won the contract, Ford hired the colonel who supervised the DT/OT II test, who coincidentally chose that moment to retire.

One factor that worked, oddly, in Ford’s favor, was that just before the initial contracts, during the recession of spring 1981, Ford declared the largest corporate loss in U.S. history. In the White House there was fear that Ford would declare bankruptcy or demand a Chrysler-style bailout. In the Pentagon officials were inclined to throw Ford some business on the “industrial base” argument—championed by Undersecretary Richard DeLauer—that production lines must be kept open. In 1981 Ford had little defense business while General Dynamics had plenty. So Ford was chosen.

“Industrial base” preservation has merit. The defense industry is already too concentrated, and letting a major contractor drop out would only decrease competition further. But throwing a company some business only results in national gain when what is acquired works.

Lunenbergs running the asylum

Chosen as the DIVAD “source” in 1981, Ford was not yet out of the woods. Army inspectors found so many deficiencies in its prototype that a production award was put off until the next year. Fortune smiled on Ford in September 1981, when James Ambrose was named undersecretary of the Army: his last job had been vice president of Ford Aerospace, where he oversaw development of DIVAD. In May 1982, with Ambrose the highest-ranking Army official present, a review board approved DIVAD production.

The first DIVADs came off the line six months late and in such shoddy condition that the Army had to waive its contract requirement in order to accept them. During a test one DIVAD locked on to a latrine fan. Michael Duffy, a reporter for the industry publication Defense Week, who broke this aspect of the story, received a conference call in which Ford officials asked him to describe the target as a “building fan” or “exhaust fan” instead.

In February 1984 the Army formally notified Ford that its performance had been “totally unacceptable,” and activated a rarely used contract clause to suspend DIVAD production. In June Weinberger dressed down some generals at DARCOM (the Army procurement hierarchy) who had come to give him a briefing about how DIVAD was just dandy, and told them to show cause for the project not to be canceled. The Army responded by staging another rigged test.

Held at Fort Bliss in July, the “Limited Operational Test” was confined to shooting at a helicopter drone equipped with radar amplifiers. The amplifiers (“Lunenberg lenses”) refocus radar beams to make them easier for the sending unit to detect. At first the drone was fitted with only one Lunenberg lens, but that wasn’t enough; DIVAD couldn’t find it. Eventually four were added, on the compass points below the mast. With four amplifiers DIVAD finally scored a hit, and the tests were pronounced a success. (Needless to say, not only was the target cooperating but it was not maneuvering, there were no radar jammers operating, and there were no other aircraft in the area as distractions. To continue the previous analogy, this one was like testing a bloodhound’s ability to track a man covered with beefsteaks standing alone and upright in the middle of a parking lot.)

Exact results of the LOT test are classified, but a Pentagon radar expert who has studied them told me, “The test was a total fraud. It bore no relationship whatsoever to any condition the DIVAD would encounter in wartime.” Army officers claim amplifiers were used to “enhance” the reflection because Russian attack helicopters are larger than the Huey-size drone; the Pentagon radar expert said the amount of amplification was many times more than required for this purpose.

DIVAD’s radar, which is adopted from the F16 fighter, needs help from amplifiers because it was found to possess a farcical drawback—it can track moving targets but not stationary ones, the greatest sitting ducks of all.

Again this is not a manufacturing defect but a boundary of technology. For reasons too complicated to explain here, an object must be moving if the DIVAD’s radar, which operates on the Doppler principle, is to locate it (as a train must be moving for the pitch of its whistle to change). What it boiled down to is that DIVAD’s radar could track maneuvering airplanes, but its computer could not aim at them. Meanwhile DIVAD’s computer could handle hovering helicopters, but these its radar could not track.

Nonconcurrent resolution

Building a $6.8 million radar-computer system to shoot at helicopters is an especially dramatic example of the contemporary Pentagon’s insistence that every objective be accomplished in the most complex and expensive manner possible. Almost anyone can aim at a hovering helicopter, and almost any kind of gun will bring a helicopter down. During the Vietnam war the Army—flying with total air superiority—lost 4,643 helicopters, or nearly three times the number of jet fighters lost in Korea. All these helicopters were destroyed by visually aimed fire from small arms; no radars or fancy SAM missiles were employed. In Vietnam the Russian AK47 assault rifle, which costs about $600, proved to be an effective anti-helicopter weapon.

Responding to criticism of LOT, the Army said it would stage a realistic test of DIVAD sometime in 1985. Why couldn’t realistic tests be staged until nearly a year after the secretary of defense ordered them? Because, the Army said, it will take that long to train an operating crew—for this, the completely automated system.

This fall Weinberger was briefed on the results of the LOT test by Brig. Gen. Michael Hall, an Air Force officer who is acting head of the Pentagon’s Operational Test Evaluation office. OTE had written what is believed to be a scathing critique of DIVAD. Simultaneously the Defense inspector general was finishing his report, which accused Ford of overbilling the Army by $84 million.

The report found what is fast becoming the customary spare-parts pricing schedule—a power control assembly jumping from $7,805 on the first quote to $151,500 on the third, for example.

Also it was found that in many cases, Ford had given the Army an estimated price for subcontracted components, then obtained the components for significantly less, pocketing the difference. According to the inspector general, Ford actually paid the Swedish firm Bofors, which makes DIVAD’s cannon, $16 million less than it told the Army in bid documents the cannon would cost. DARCOM had been warned during contract negotiations that DIVAD subcontracts could be obtained for less than Ford was predicting but chose to ignore the warnings. Ford, however, did not ignore them. It used the very same government “field pricing” reports that DARCOM dismissed as an instrument with which to bargain the subcontractors down. Responding to the inspector general’s charges, Assistant Secretary of the Army Jay Sculley wrote, “The Army nonconcurs.”

The key piece of information here is that DARCOM, which is supposed to ride herd over contractors, was disguising and apologizing for excess prices. Within institutional Washington, “clientism”—becoming an advocate for the interests one is charged with controlling—is common. Ambassadors often try to cover up what’s really going on in their client countries; cabinet secretaries come to their jobs threatening to wield the budget ax and six months later are up on Capitol Hill asking for dramatic budget increases. In military procurement agencies like DARCOM, generals measure their status by how much money they spend—putting them on the side of cost inflators. And as most envision a lucrative second career with a defense contractor, they know that the less disciplined the programs, the more money will be available to provide them handsome salaries and expense accounts when they retire.

It was at this low point—word of critical audit reports and the latest magic tests beginning to circulate—that the Army chose to announce its plans to add the Stinger to DIVAD. Just how these missiles were to be incorporated into a system already so complex that Ford technicians were having trouble keeping it running, the Army didn’t explain. (In addition to never shooting at maneuvering targets, DIVAD had never been subjected to a field reliability test. That was expected to be the next embarrassment.) There was another consideration: Stinger was in trouble too.

The General Accounting Office was preparing to release a report saying purchases of the miniature missile should be cut because of cost (at $88,000 each, Stingers cost more than Sidewinders) and technical limits on its performance. Adding Stinger to DIVAD, the Army felt, would tie the fates of two troubled programs together, making each more difficult to defeat.

Additionally, Stinger is built by General Dynamics, the nation’s largest and best-connected defense contractor. It was General Dynamics from which the original DIVAD contract was stolen in 1981, and the company had been smarting ever since. General Dynamics is said by some Pentagon sources to have a “Pearl Harbor file” of details still unreleased on how the DIVAD selection was rigged, and to have alluded to that file at opportune moments. A billion in extra business might soothe its feelings; “Stinger was added to buy G.D:s silence,” an informed Pentagon official told me.

(Further public relations improvements were sought by a concerted effort on the Army’s part to call DIVAD by its formal name, the Sergeant York Gun, making it sound like something new. At a low point for the Ml the Army rallied by suddenly beginning to speak not of the MI but of the Abrams tank; Reagan tried, with less success, to rid himself of the MX by proclaiming it the Peacekeeper. The B1 has not yet been named, so watch for that day—probably the day before a cost overrun is revealed—that the Air Force begins talking of the LeMay bomber.)

The Stinger gambit worked in reverse, making DIVAD’s plight so obvious that Congress had to act. After a House committee voted to “fence” all DIVAD funds until completion of the scheduled realistic tests next year, and the Senate voted to deny fiscal 1985 production funds, Weinberger finally gave in, agreeing not to resume production and to await further results. While there could be some attempt next year to revive DIVAD, a quiet death is now expected. It’s important to note that even in this environment, with DIVAD becoming a laughing stock that would undermine all Pentagon budget requests, the secretary of defense still could not bring himself simply to be a man and cancel the program.

The Good, the Badham, and the Ugly

Congress had done its best to overlook DIVAD previously: In June 1983 Rep. Larry Smith of Florida introduced an amendment to kill the program. Five congressmen rose on the House floor to speak for continuation: Samuel Stratton of New York, Robert Badham of California, Marjorie Holt of Maryland, Bill Nichols of Alabama, and Ronald Coleman of Texas. Stratton is chairman of the procurement subcommittee of the Armed Services Committee, and he spoke briefly to object to intrusion on his turf. The other four spoke at length, although without giving any indication of the reasons for their interest in the program.

Badham, also on the Armed Services Committee, represents the district where DIVAD is assembled. Holt, another Armed Services committee member, is from the suburbs of Baltimore, home of the Westinghouse division that builds DIVAD’s radar. Nichols, of Armed Services, represents Anniston, Alabama, site of the factory where DIVAD’s chassis are made. Coleman is from El Paso, location of Fort Bliss, where DIVADs would be based. No congressmen without a personal stake in keeping the funds flowing spoke in DIVAD’s defense.

Badham declared that DIVAD “is on schedule, under cost and has no significant performance problems Coleman an found an innovative use for The Threat saying we shouldn’t worry that DIVAD’s radar will act as a beacon revealing the location of troops (while DIVAD may note be able to hit aircraft, ‘aircraft with radar-seeking missiles will have little trouble hitting DIVAD) because the Russians don’t worry about Shilka’s radars giving them away.

Smith’s amendment lost by a large margin, and a similar amendment he offered in May 1984 lost as well, with a similar cast of characters rising to speak without revealing their personal interests. Opposition to DIVAD was then taken over by Rep. Denny Smith (no relation) of Oregon, a conservative Republican and former fighter pilot who flew 180 combat missions during Vietnam. He has said that what pilots fear is a large number of guns setting up a crossfire, not a single gun, no matter how good. On the Senate side one of the most outspoken DIVAD opponents was John Glenn of Ohio—also a former fighter pilot.

Unfortunately some DIVAD critics fall into the trap of saying everything that reflects poorly on the weapon without regard for whether the objections are significant or internally consistent. For example, the Congressional Budget Office criticized DIVAD because its guns have a maximum range of four kilometers, while new Soviet helicopter-launched missiles are thought to have a maximum range of ten kilometers. But this range is strictly theoretical. In actual use Soviet helicopters will have to draw much closer to U.S. tanks just to locate them, let alone to target missiles. Basing procurement decisions on what weapons do in theory is like my saying it will take me an hour to get from Washington to Philadelphia because my car has a top speed of 120 miles per hour. I could get into serious trouble if I confused what my car is capable of doing with how I use it under real-world conditions: the same kind of trouble is now afflicting the Pentagon.

The exaggerated ten-kilometer objection gained just enough currency to backfire. When the Army announced its Stinger-DIVAD merger plan, the official explanation was that this would counter the new long-range helicopter threat. As alternatives to DIVAD begin to be discussed in the armed services committees next year, the range argument may well be used to justify some other very complex, expensive system.

Quality is job number 378

Of course, congressmen are loath to attack each other’s boondoggles for fear that their own boondoggles will be attacked in turn. Ford, the Army, and Weinberger all had institutional dilemmas of their own.

DIVAD’s condition is not entirely Ford’s fault, since the company was sent on a technological snipe hunt. Public scorn for DIVAD tarnishes the company’s otherwise admirable image—besides its rapidly improving cars, Ford builds the Sidewinder missile, perhaps the most costeffective weapon in the modern arsenal. To top it off, even if Ford keeps the $84 million, it may end up losing money on the deal, as the costs of building a radar-computer system turned out so much higher than predicted. But the company had no way to escape gracefully. If Ford withdrew from the project it would be admitting to shareholders that company investments in DIVAD tooling were a waste, and the Pentagon would be incensed. A Ford withdrawal would be tantamount to a formal admission that the project was a flop, and Ford’s prospects for future defense work would plummet.

The Army could not drop DIVAD, either, without admitting it was wrong in the first place. Chief of Staff General John Wickham has lately been reduced to pleading that “we’re stuck with it,” that stopping now will mean the $1.5 billion already spent will have been wasted. True enough, but the $1.5 billion spent is already wasted no matter what we do next; pouring several billion more down the same drain will not cause the original $1.5 billion to float back up. (Though a sure sign of desperation, the “we’re stuck with it” argument is among the most persuasive to Congress.) Wickham will be chief of staff for a few years at most, then retire. If he blunders along denying that there are problems, his tenure will will not have an official mark against it—and the effect of his actions, fielding a weapon that’s no good, will not be felt until someone else is on the hot seat.

Not all the military’s problems are of its own making. Congress likes to hear about wonder weapons that can do everything everywhere automatically. The political sex appeal quotient of the best alternative to DIVAD—visually aimed simple guns—is roughly zero. DIVAD was also designed in the environment of the mid-1970s, when enlistment and the “quality” of troops were down. Congress expected Army planners to operate with as few people as possible and to build idiot-proof weapons that required little skill; “putting the brains into the machine” became the slogan, and it accounts for much of what is emerging today, a decade later, in great-sounding but ineffective weaponry. Finally, both the Pentagon and Congress have a fondness for wonder weapons because they offer the hope of fewer deaths on our side. One perfect DIVAD with three soldiers reasonably safe inside an armored shell might accomplish the work of five scattered gun crews less well protected. This motive is a noble one, and well worth any expense. But to serve the ideal the wonder weapons must work: if they don’t the death toll may ultimately be greater.

The symbolism of spending rules Washington debate. Conservatives, who so often opine that the welfare system cannot be trusted because it is run by self-interested bureaucrats and that in welfare throwing money at a problem can actually do more harm than good, turn around and say that when it comes to the Pentagon the experts must not be questioned; throwing money is the only solution. They tend to see money itself as a form of defense and the details of how that money is used as a nagging side issue. Would giving Chrysler a bigger budget—without requiring it to improve its cars—have saved the company? Many leftists, for their part, tend to fall for the theoretical performance claims of high-technology weaponry because the idea that those menacing multinational conglomerates can’t build a machine that hits a stationary target does not fit with their preferred world view. Both sides try to avoid the low-glamour, non-geopolitical question of whether the weapons work. The result is that, in order to stave off an admission of error that would be awkward today but improve matters over the long run, the Pentagon ends up looking foolish day in and day out.

The pack pounces

I have a personal interest in this subject, having done the first DIVAD story more than two years ago in The Atlantic. For a while it was my word versus the Army and Ford, and my word wasn’t doing too well. DIVAD was all but unknown then, most of my facts came from off-the-record sources and all were being vigorously denied. The networks and the big papers wouldn’t touch the story.

Fortunately for me, the Project on Military Procurement took up the cause. The Project is a sort of clearinghouse for whistleblowers. After my story was done, one of my sources, a man with lifelong Pentagon connections who later lost his job because he helped me, delivered copies of some DIVAD documents to the Project. Paul Hoven, a former Vietnam helicopter pilot who is assistant to Dina Rasor, the Project director, then began a tireless effort to get other reporters to write about DIVAD. The Project’s involvement allowed reports to be hooked to a semi-official source. Should something go wrong, a paper could quickly wash its hands of the DIVAD story by saying it was merely reporting the allegations of a nonprofit foundation.

As public admission of DIVAD’s problems began to occur this year, the frequency of stories increased. DIVAD became a media star, even though the substance of what was being reported had been known for two years. (At one point in September 1984 The New York Times ran as a front-page story something we’d had in the Atlantic in October 1982. But am I bitter? No-o-o-o.) Partly this delayed response owes to the structure of Washington life. Government affairs produce such an information overload that, even considering the number of reporters per square inch in Washington, time often passes between when news reaches print and when it sinks in. Billygate broke nearly a year after the substance of the case—that a president’s brother had registered as a paid agent of a hostile power—had been published without reaction. David Stockman mentioned Reagan’s possession of the Carter briefing book during his Elkhart, Indiana speech two years before Debategate; the mining of Nicaraguan harbors was reported roughly four months before it became a national headline issue.

But DIVAD suddenly hit the news for another reason—it had become official. There was no risk in reporting the official release of an Army contract ruling or an official statement made by Weinberger at a press conference. Nor any need for the reporter to make judgments on his own authority—as, inevitably, reporters must when evaluating contradictory claims that cannot be definitely proven. (Even reporters using the pure objective “‘he said she said” form must make some judgments about who’s closer to the truth or their stories will be incomprehensible.)

Everything worked out in the sense that DIVAD’s problems did become public. But what if they hadn’t? What if the Army had kept on denying? And how many billions could have been saved if the general press had been willing to tackle this story earlier, before mistakes were cast in stone? The press is thought of as “biased .” Where does bias occur? Generally, in loading value judgments into accounts of official events— presidential statements, economic indicators, results of summit conferences, and so oft. When the picture is less clear, most reporters and news organizations are fearful of making judgments on their own authority; stories must always hinge on an indisputable fact or a “study has found .” In one sense this is “just the facts, ma’am,” but in another and equally important sense it absolves the press from having to do its own thinking and take stands for which it might later be criticized. The press should take stands when mistakes are going uncorrected or wrongs unrighted; this is especially important in military procurement, considering that the press is the only major actor without a self-interest stake in unlimited budget increases.

If reporters are smart enough to say what is wrong—as nearly every newspaper and television station does on a daily basis—they are smart enough to suggest what would be right. Yet as proper coverage of complex issues like defense inevitably entails evaluative risks, most news organizations back away. Spending $7,600 for a coffee pot on the C5 cargo plane is obviously wrong, so the press can hammer away at it; hammering at this story is “objective.” At the same time the press has been timid about covering the revival of the C5 itself – a far more important issue, an $11 billion story affecting the deficit, the way the Pentagon does business, and whether the Air Force can get a vital job done.

Most of the weapons horror stories that have made the news in recent years have come either from low-visibility papers like The Arizona Daily Star, the Fort Worth Star-Telegram, or the Washington Times, or from the Pentagon itself: release of official audit reports is frequently treated in news stories as breathtaking revelations the reporter dug up. If it weren’t for Pentagon auditors most of the spare-parts stories would not have broken, a point Defense Department spokesmen have tried to make with increasing frustration. The Project on Military Procurement has been the source of some other spare parts news (like the $7,600 coffee pot) and many of the troubled weapons stories that have appeared in major papers in recent years. This is both a tribute to the Project and an indictment of bigdeal news organizations that are either unable or unwilling to do the real work on their own. NBC News’s recent segments on fraud in shipyard contracts were excellent, but the story was done in conjunction with the Better Government Association, another organization that, like the Project, does low-glamour work and learns enough to trust its own judgment. When the big-deal press does get Pentagon stories on its own, the stories are often approached with a trepidation Hildy Johnson would find unbelievable. For example, Charles Mohr’s important New York Times story on unrealistic testing of the Assault Breaker system, a potential multibillion-dollar program that is today where DIVAD was five years ago in terms of building bureaucratic momentum, was buried on page A21.

In a strange way, the national press corps actually prefers to focus on stories everybody has, which seems to me the reverse of what reporters are supposed to do. Press conferences; plane crashes; confirmation hearings; the campaign trail; ABC gets the senator’s statement 17 seconds before CBS. When the marines were killed in Beirut every newspaper and television station in America wanted to shove a microphone into General Kelley’s face and demand a statement— but where were they when the institutional problems that caused the fiasco were developing? Between the barracks bombing in 1983 and the embassy bombing in 1984, where was the reporting on whether changes were being made to prevent the fiasco from being repeated?

Of course, newspapers and television stations must devote the bulk of their attention to the breaking events of the day and treat risky enterprise reporting as their second concern. And of course, covering breaking events can be difficult and sometimes dangerous work. But most of the time it isn’t—most of the time it’s a breeze. In the upside-down world of the national press the best reporters are often given the least challenging assignments, such as press conferences. Amazing amounts of attention are paid to stories everybody has (like Ferraro’s finances), while stories only one paper or network has—which may be the biggest news—are played down. And though no news organization can be everywhere, it’s not lack of resources that prevents them from taking on the Pentagon. Anyone who has recently walked through newsrooms of the major networks, newsmagazines or the top five papers knows they are not exactly understaffed.

Yet the Department of Defense, with three million employees the free world’s largest single institution and for several reasons one of the most important stories of the 1980s, gets less press attention than the Iowa caucuses. And matters may get worse. When the postmortems on the CBS Westmoreland case are written, CBS’s ultimate sin may well be judged to have been reaching its own conclusion. Perhaps CBS was wrong in accusing Westmoreland of “conspiracy”; but instead of seeing this event as a sign that journalists working in complicated subject areas should redouble their efforts to insure their conclusions are correct, most of the big-deal press will probably interpret it as a sign that journalists shouldn’t take chances. Would CBS have been assailed if it had merely quoted “experts” alleging conspiracy?

Sure to be fixed in 14 years

Lately Army officials have begun to say that DIVAD got screwed up because it was an “accelerated development” program given only seven years from proposal to production, instead of “the normal 14” Rep. Coleman dutifully echoed this claim on the House floor.

How much further removed from reality can an argument be? The Apple computer and the IBM PC; the Honda Accord and the Pontiac Fiero; the Boeing 747 and the nuclear magnetic resonance medical scanner; none of these, complex technology built to real world requirements, took 14 years. The space shuttle took less time; the highly successful F16 fighter was put together in two years by a bunch of guys at General Dynamics who had their own private hangar and a temporary exemption from Air Force R&D regulations. Recently Ford suspended production of its Tempo and Topaz models because a manufacturing defect was discovered. Nobody had to tell Ford the situation was “totally unacceptable”; the company acted on its own, because the cars were on their way to the real world where realistic testers (customers) would quickly find and complain about the problems. In DIVAD’s case the customer has done its best to cover up rather than correct the problems.

The notion that it takes 14 years to determine if a mechanical device works is ideally suited to institutional Washington. Washington has an attention span of about one week. Each week some issue pops up to become the most important question in the history of civilization, and each following week the issue is forgotten and the world created anew. Remember the week in 1983 when every headline was the shocking, astounding events in Chad? The week this fall when minesweepers being sent to the Persian Gulf dominated the news, only to vanish as surely as though they had sunk? Washington institutions, including the press corps, are oriented toward crisis-management issues that explode from nowhere, build rapidly, and then peak with some distinct conclusion. Weapons programs like DIVAD which lurch forward foot by foot, refusing to go away, generally overcome the system by boring it to death.

And it’s obvious where the 14-years-complaint is leading. The Army will use DIVAD’s failure as an argument for more wasted time, more spending, more insulation from reality. After the 1981 “shootoff” provided the first bad news about DIVAD, DARCOM commissioned a study of weapons selection techniques. The study recommended that actual tests like the shootoff be foresaken: it would be more efficient, the study said, to “test” strictly by computer simulation.

My last defense story

But even though these stories are important, you’re sick of reading them. I don’t blame you; I’m sick of writing them. I feel I have recently fulfilled my civic and journalistic obligations in regards to national defense, and promise not to write on the military again. So there’s a small consolation for your having to read 10,000 words about antiaircraft guns.

Allow me to close by first using up an anecdote I have been saving for years: a Pentagon official once told me, “I’m not authorized to say ‘no comment.’ You’ll have to get your ‘no comment’ through channels.”

One of the frustrations of doing stories like DIVAD is that the principals inevitably come off sounding like nitwits, which of course they are not. It’s difficult to draw the distinction between what people do and what the system does: journalistic shorthand usually makes people sound like the guilty parties, when in fact it is the system which is guilty.

The military men and women I have met in five years of defense reporting generally have been bright, dedicated and diligent. Naturally the military has its share of scoundrels—they tend to collect at the top—but very few in the armed forces are corrupt in the money sense or dull in the mental sense. In an important way they are idealists, having offered to lay down their lives if so required. Some in the military don’t take this offer seriously—but many more do. How many congressmen, reporters or experts would make a similar offer?

But put even the bravest and smartest soldiers into a committee, and they turn into the cast of Saturday Night Live. The system is the problem, not the personnel; the system is so slanted towards interest groups and ass-covering that bad drives out good. The solution is leadership from the president and the secretary of defense; not salesmanship for budget requests, but leadership. I feel strongly that if a president were to challenge the Pentagon to be better instead of bigger, more effective instead of more expensive, he would find a receptive audience both with the public—which, I think, senses how far out of control the Pentagon is—and within the military itself.

An important national reserve, pride, is being squandered by the present slapstick system. Real officers don’t want to see their men slaughtered in Beirut because they were paying more attention to politics than tactics; don’t want to be sent up to Congress to lie through their teeth; don’t want to prostitute their honor defending pork barrel. Real officers want to get up in the morning and know they’re doing their best and everything else be damned. Real soldiers feel the same way. Even military contractors would be happier and gain self-respect if what they were building was useful instead of wasteful.

But the present system does not reward true believers; it rewards game players. This can be changed, but not until we have a president and a secretary of defense who are interested in something beyond the easy way out. A president more interested in the country than the polls; who knows that strength is different from spending; who understands that patriotism is different from nodding yes.

Our ideas can save democracy... But we need your help! Donate Now!

Gregg Easterbrook has published three novels and eight nonfiction books, mostly recently It’s Better Than It Looks: Reasons for Optimism in an Age of Fear. He was an editor at the Washington Monthly from 1979 to 1981.