Uncle Sam's not-so-friendly skies: the FAA has a good safety record, but it could do much better.

It was Halloween afternoon. The American Airlines commuter turbo-prop was circling at 10,000 feet waiting for permission to land at Chicago's O'Hare airport. In the cockpit, the pilots joked with each other while the autopilot held them on course. Outside, a cold rain streaked by.

After more than half an hour in the holding pattern, the airport radioed permission to descend to 8,000 feet in preparation for landing. The pilots adjusted for descent, retracting the wing's flaps.

That's when everything went wrong. The autopilot suddenly switched off. The plane tilted violently to the right, the wings almost perpendicular to the horizon. The pilots wrestled for control, trying to bring the craft back to level flight. But the plane seemed to have a mind of its own: It turned over on its back and rushed for the ground.

"Mellow it out. ... Mellow it out ... nice and easy," the pilot said; then came a crunching sound as the plane fell apart at upwards of 400 miles per hour, moments before it smashed into a soybean field outside Roselawn, Illinois. Everyone--64 passengers and four crew--was dead.

Because of accidents like the one in Roselawn, the Federal Aviation Administration (FAA) is one of the few agencies within the executive branch that draws much attention from the press. When a plane falls from the sky, the story is compelling, albeit morbidly so: the pictures of twisted metal, luggage hanging from trees, the screaming mother at the airport where the flight never arrives. And because Roselawn was only one of a series of serious air accidents in 1994, Americans were becoming wary of flying. The press, as a result, took an interest in finding out what was happening.

The string of reports on the FAA proves that the media can do a far better job than it has been doing to explain what actually goes on in government. Indeed, when reporters probe the executive branch with the right questions, the answers can amaze, infuriate--and even reassure. It's not hard--just by putting some of these stories together and doing a little digging, I've been able to answer at least some of the public's questions: How well is the FAA, on balance, doing its job to prevent any air accidents? And how can it do better?

Setting Course

The United States could not function as a modem economy without the air links that the FAA oversees. If moving one ton of freight one mile is a "ton mile," U.S. airlines move more than 14 billion ton miles annually, as well as another 2 billion ton miles of maid. The industry also moves an enormous number of people--upwards of 550 million last year alone. Preventing accidents is also a high-stakes venture: Just one accident, the 1989 crash of a DC-10 in Sioux City, Iowa, killed 112 people and cost some $300 million.

With a $4.6 billion annual budget and 48,000 employees, it is the charge of the FAA to make sure that the system functions efficiently and safely. Yet this is a difficult time for the agency. Like much of government in this era of "dowrisizing" and "reinvention," the FAA faces ever-smaller budgets and increasing responsibilities. By 2006, the number of airline passengers in this country is projected to hit 930 million a year, an increase of 70 percent. At the same time, the FAA has faced sharp attack from critics who contend it's neither as effective nor as efficient as it should be; several bills to reform the agency wholesale are snaking their way through Congress--and some have even suggested privatizing the entire agency. In response, David Hinson, the agency's tough new administrator, is trying to get a handle on billions of dollars of waste and, above all, to aggressively pursue safety. The agency, in other words, is in the midst of a mini-revolution.

While Congress debates seemingly arcane matters like changing the FAA's procurement rules, and as Hinson promises to change the "culture" of the FAA, journalists need to keep an eye on the ball. "It may seem amazing to say, but like many big organizations, [the government] is primarily dominated by considerations of input--how much money do we spend on a program, how many people do you have on staff, what kind of regulations rules are going to govern it; and much less on output--does it work, is it changing people's lives for the better?" For the press, this means focusing on how well the FAA does its three most important tasks: developing new safety rules, making sure that the rules are actually being followed, and directing the nation's air traffic.

It's ironic that, at a time when Washington is filled with outcry over excessive regulation, one of the FAA's problems is its reluctance to issue new--and possibly life-saving--regulations.

Consider the example of the Ground Proximity Warning System (GPWS), recently recounted by Jeff Brazil in the Los Angeles Times. The GPWS is a relatively simple device that warns the pilot if the plane is close to the ground. The FAA began requiring the GPWS in all of the larger passenger planes after two particularly bad accidents--an Eastern Airlines L-1011 that sloped into the Florida Everglades and a TWA Boeing 727 that slammed into a mountain while approaching Dulles--that together killed 191 people.

But while common sense might suggest that smaller commuter planes should also carry a GPWS--after all, they run the same risk of crashing--the FAA didn't act. By 1986 it was clear that this was a mistake. A study conducted by the National Transportation Safety Board (NTSB), an independent government agency that investigates accidents on highways, on railroads, and in the air, found that after the GPWS systems were put into larger planes, "controlled flight into terrain" accidents (as they are dryly called) dropped by 75 percent. Even more to the point, the report concluded that three recent commuter crashes, responsible for 25 deaths, easily could have been prevented if the devices had been installed.

Still, the FAA didn't move--and the crashes continued. In January 1988, a commuter plane slammed into a Colorado ridge, killing nine. The next month, another commuter crashed, killing 12. That fall, according to LA. Times reporter Brazil, an internal FAA memo concluded that, of 280 commuter crashes between 1972 and 1985, 60 crashes might have been avoided with the GPWS. Not until April 1990, though, did the FAA say that it intended to require them in smaller planes. Then it gave the airlines two years to comply. "For every six months we delay the GPWS, we can expect an accident," wrote an FAA official in another memo Brazil uncovered. Indeed, four months before the deadline, another 18 died.

For an outsider, such sluggishness by an agency ostensibly created to protect public safety is downright baffling. Part of the explanation lies in the FAA's mission or, rather, its dual mission. The legislation that set up the agency mandated that the FAA not only regulate the airlines to ensure the safety of the flying public, but that it also promote the aviation industry. If these two goals sound contradictory, that's because they are, and too often, the instinct to help the airlines--or, in the soft-pedaled jargon of the beltway, to "be sensitive to their concerns"--has softened the FAA's resolve to protect safety.

Recently, for example, the FAA was trying to decide whether to take action on the Boeing 757, which, like many large planes, leaves a vortex in its wake that can endanger smaller planes trailing behind. Two accidents and 13 deaths had already been linked to the problem. Why would the FAA even think twice about taking action? Well, the LA. Times's Brazil uncovered a memo that listed an unusual item among the reasons for not aggressively pursuing the safety problem: worry about damaging the public's perception of Boeing. That's kind of like not removing E. coli-tainted beef from Jack-In-the-box because cheeseburger sales might drop.

Usually the reasons for the FAA's laggardness are more subtle. Before the FAA can issue a new regulation, the law requires the FAA to show that the benefits (in lives saved, equipment not damaged, etc. It will outweigh the costs of complying. There is nothing wrong with this in theory, but in practice it is abused. "Those who are affected by any recommendations often hide behind the cost considerations," explains NTSB board member John Goglia, a gregarious former aircraft mechanic who doesn't pull punches. "They're very good at making sure that any and all direct and indirect costs ... get included." If the industry can show a change will cost too much, there's no change.

The FAA's conflicted identity contrasts sharply with that of the NTSB. Step into the NTSB lobby near the Mall in Washington, DC, and their single-minded purpose--to stop accidents from happening--is obvious. On the walls is a dramatic collection of accident photographs: U.S. Air Flight 5050, which took a slide off the end of the La Guardia runway; the tail of a small plane protruding from the second floor of a suburban home; even the distinctive smoke trails of the Space Shuttle Challenger accident.

The NTSB's simple mandate is to find out why accidents like these happened and to determine how they can be prevented in the future. The agency is small, with only 350 employees and a $38 million budget--"enough to run the [Department of Transportation] for about eight hours," jokes Chairman James Hall.

Yet it is extremely effective; over the years, its public bullying has prodded the FAA to make a host of important safety changes. "It's not that these guys are corrupt," says former NTSB Chairman James Burnett of FAA officials. "But a lot of them come from industry ... and a lot of them contemplate going back to industry ... which can make them a little more conservative."

Another problem, says Hall, is that too often it takes an accident to spur the FAA to fix problems: As early as the 1970s, safety critics began urging the FAA to do something about the problem of "runway incursions," accidents caused when a plane, or even a service vehicle, wanders onto an active runway. One report counted 1,210 near-collisions on runways between May 1978 and September 1983. In February 1991, U.S. Air Flight 1493 landed on top of a Skywest commuter plane at Los Angeles International Airport, killing 34 people. It was the third fatal runway collision in 13 months. Yet the FAA has been slow to respond. The ASDE-3 radar, a system to warn controllers of possible problems, has experienced significant delays. Twenty-one radar systems were promised by 1990. Now 1996 is the target.

* The NTSB has been lobbying the FAA for years to require more sophisticated "black boxes" in planes. The boxes (which are actually orange) record a plane's vital signs: airspeed, velocity, the positions of its various controls, etc. The older boxes record only the most rudimentary information, while the newer ones record more than 100 separate pieces of information, making it much easier to determine the cause of accidents and prevent future ones. For example, two Boeing 737s have mysteriously crashed recently. There may be a design flaw in the plane's tail which causes the pilot to lose control of the rudder--if true, a serious design flaw. Yet the two flight recorders were of an older design, so investigators cannot say for sure.

* The NTSB has had to constantly prod the FAA to redouble its efforts to keep bad pilots out of the cockpit. Alcohol and drug abuse is a real problem. One pilot who was found dead at the controls of a plane that crashed had a blood alcohol level of .16 percent, the rough equivalent of downing seven drinks in an hour. Despite incidents like these, it took years of prodding before the FAA took even the simple step of checking to see if pilots had lost their driver's licenses for driving while intoxicated.

The FAA has been--and continues to be--even more remiss in keeping unqualified pilots out of the cockpit. A number of planes have crashed due to pilots who were documented incompetents, most recently in North Carolina where 13 people were killed. For years the safety critics have lobbied the FAA to require that detailed personnel files be available to potential employers, but the agency has resisted, insisting that such a mandate would cause legal problems or violate pilots' right to privacy. What about the public's right to safety?

Beyond these specific problems, critics have urged the FAA, in the words of one recent GAO report, to take steps to "expand the margin of safety by being more proactive." The NTSB and others, for example, have suggested that the FAA systematically check flight data recorders--not just after an accident--to uncover problems with the planes' or pilots' performance. In Europe, checking the tapes is standard procedure. In the U.S., however, the pilots' unions have objected to such preventive investigation because they fear the information will be used against them. The FAA says it is still working on a compromise.

Being There

The FAA's job does not end with passing safety regulations, though. It must also enforce them. It is this area--making sure that the rules passed in Washington actually work in the field--that government agencies and the press are most likely to overlook. We've paid billions of dollars to bail out failed savings and loans, for example, not only because the government deregulated the industry, but because the agencies responsible for overseeing the S&Ls stopped monitoring what was happening outside Washington.

Keeping an eye on what's happening in the field is especially important for the FAA, particularly since the airline industry deregulation of 1978. In the old days, competition was tightly limited, which made the industry stable if not necessarily efficient. Since deregulation, however, competition has become far more cut-throat; old carriers go bankrupt, and new carriers start up every year. Such keen competition can tempt some carriers--especially upstarts--to cut comers on safety. Although the industry's overall safety record has continued to improve over the past two decades, deregulation has put more pressure on the FAA and its inspectors. "It's like a basketball game when you've suddenly got these two aggressive teams," says former NTSB chairman Burnett. "You've got to call it a lot closer."

The problem is that the FAA's referees haven't always been at the game. Consider the program for "designated mechanics examiners." To work on aircraft, a mechanic must be certified. Because there are so many mechanics who need to take this test, the FAA "designates" some mechanics, giving them the authority to test the others. But when the DOT's Office of Inspector General looked into the program, it found that the FAA was not keeping an eye on the examiners. And, like a kid given the key to the candy store, they were letting all their friends in: 99 percent of everyone who took the test passed.

Perhaps, you might object, 99 percent deserved to pass. But when someone from the inspector general's office sat in on the testing, the pass rate dropped to 58 percent. When an FAA official sat in as well, the pass rate dropped to 40 percent. Goglia sees this incident as representative of a broader problem--FAA inspectors just aren't in the field enough. "When I first came into the business in the early sixties, you always saw the inspectors. Twenty years later, you can go for months and months without seeing them."

Part of the problem, inspectors say, is an increase in paperwork, which keeps them in the office. There is also another classic bureaucratic problem: Middle management has swelled at the expense of actual employees in the field. According to the Office of Personnel Management, there is one manager for every six employees at the FAA, compared to a government-wide average of close to eight. (And the government-wide average leaves plenty to be desired. The inspectors also complain that they are tied to the office because they have inadequate clerical support. "The FAA took a big hit in the downsizing, and a lot of the people were clerical workers," says Rep. James L. Oberstar (D-Minn.), the former chairman of the aviation subcommittee and a longtime air safety expert.

Being on-site is valuable not just because of problems inspectors might cover. Their mere presence inspires--or frightens--airline personnel to stay on their toes. "If you are around--and everyone knows you're around--it has an impact," says the NTSB's Goglia. With inspectors stuck in the office, not only do the airline workers have a tendency to get sloppy, but the inspectors themselves lose their eye. "I have personally watched an FAA inspector do a ramp check of an airplane and walk by two major violations," Goglia says. One violation--using flexible instead of steel tubing inside the wheel well--could have caused a dangerous fire. Inspectors don't just have to be there; they have to know what to look for."

Inspectors complain--and a number of auditors have confirmed--that they are not adequately prepared for their job and do not get enough training to keep them "current." According to a story by Adam Bryant in The New York Times, only four percent of inspectors of the ATR commuter plane, the one that crashed in Roselawn, were fully qualified to fly the planes themselves, which makes it pretty hard to determine what might go wrong during a flight. According to Bryant's story, the poor training of some FAA inspectors has contributed to several accidents. For example, after a Northwest Airlink turbo-prop crashed in December 1993, NTSB investigators found that the FAA's principal operations inspector did not have any experience flying the planes Airlink used. Similarly, after a September 1985 crash of a Midwest Express DC-9 in which 31 people died, the NTSB found that the principal operations inspector had no experience with jet aircraft. The problem is not a simple lack of money: According to the GAO, the FAA did not even fill all of its training slots last year.

Another problem in the field, where the dictates of the home office ought to be turned into action, is a lack of accountability. Most of the FAA's employees are hard-working, intelligent, and dedicated. But lax standards in some areas hurt agency performance and worker morale. Consider, for example, a harrowing story in Bryant's piece in The New York Times. Earlier this year, the FAA found a laundry list of safety violations at Arrow Air, a Miami-based charter company with a checkered past (including a 1985 crash in Newfoundland that killed hundreds of Americans heading home for the holidays). The violations included an out-of-date training manual, tools that were missing or improperly calibrated, and planes being okayed to fly without having made safety modifications the FAA had ordered.

Arrow was shut down for more than two months. Since the inspectors had either knowingly ignored or just missed these egregious and dangerous violations, they should have been fired or at the least retrained. Instead, they were merely reassigned.

Then there was the long report a FAA researcher wrote warning about metal fatigue. "It is possible for a number of cracks, each not easily inspected, to suddenly join together and form a long critical crack," the report read. "A number of fleets are currently operating at double their initially anticipated design life goals." Well after the report, as the LA. Times recounted, the top half of a Boeing 737 ripped off over Hawaii--because of metal fatigue. There is, a DOT Inspector General report concluded, "no formal system ... to ensure aircraft problems do not fall into a `black hole'."

Some black holes develop because employees in the field don't want to upset the status quo. Recently, mechanics for a major carrier looked at a 727 flap track, a device that holds one of the plane's control surfaces, which had arrived labeled ready-to-install. But the flap track was covered with corrosion pits and other visible problems. Outraged that a repair station would okay the part, the mechanics called in an FAA investigator. He nodded in agreement at the obvious problems, but when the mechanics asked what he was going to do, he said "nothing." The frustrated mechanics then called the FAA's regional flight standards office, but it, too, refused to act. That kind of response is "typical," says Goglia, "I've seen it over and over."

Another place accountability can break down is in the air traffic control tower. Air traffic controllers guide thousands of planes through crowded airspace, often under perilous conditions, so it's understandable that they are supposed to stay on top of the latest rules and developments by reading a "read and initial" folder. But controllers are never tested on it; that's why some jokingly refer to it as the "read and ignore" folder.

Directing Traffic

There's no question that ensuring orderly air traffic is one of the FAA's most critical services to the public. It's hard not to be impressed by the scene i a tower like the one at Washington's National Airport. Behind several sets of electronically locked steel doors, controllers with radio headsets peer out the thick, angled windows of the tower, directing a carefully orchestrated dance of landings and take-offs across three runways. And while their success often goes unnoticed, it is quite remarkable: This tower alone handles 1,000 take-offs and landings a day; the system nationwide handles 58 million operations a year.

But the record's not perfect. The major problem, as any controller can tell you, is that the FAA has done a terrible job updating the system. At a press conference, President Clinton held up a vacuum tube that looked as if it came from a sci-fi set: "We actually have to buy these vacuum tubes for [the FAA's] old computers and radar systems from other countries because they're not even produced here any more." Not surprisingly, these ancient computers have been breaking down. The computers have crashed at Dallas-Fort Worth, Oakland, and other big city airports. At Chicago, the main computer once went out for a full 29 hours.

In the Washington tower, one of the controllers showed me what happens when the computer fails. One minute, you see a radar scope with each plane identified by flight number, altitude, airspeed, and so on. Suddenly, all the identifying information drops off, the screen is filled with little green spots. The controllers are trained to work without the computers, and there have not been any accidents. But after at least one breakdown, shaken controllers told NTSB investigators they did not feel they were trained well enough.

It's not that the FAA hasn't tried to upgrade--it's been trying since 1981. A new system was supposed to be in place in 1992; the most recent estimate is 2002--or later. And not only are we stuck with the old system until the 21st century, the project is also billions of dollars over budget. Fortunately, FAA administrator Hinson has shown signs of getting a handle on the problems. Instead of following the usual pattern--making an already unrealistic project even more ambitious--Hinson has downscaled the upgrade to its bare essentials. He has also chosen strong, able managers and delegated the authority to keep the project in line. "It does seem to be back on track," says Oberstar, "and Hinson deserves full credit for that." The first step to improving the agency's operations--putting capable, committed people in charge--seems to have been taken.

In the long term, though, the FAA will not be able to keep its air traffic system up-to-date--and safe--without more basic changes. A confusing array of reform proposals is making the rounds in Washington, from privatization to simply removing the FAA from under the purview of the Department of Transportation, but the smart ones share three basic principles: First, keep the agency a government entity. Letting the airlines handle their own air traffic control would be like letting Johnson & Johnson control drug regulation.

Second, change the antiquated and cumbersome procurement rules that have obstructed efforts to purchase up-to-date equipment. The FAA should be able to run its finances more like a business: It should be able to make smart capital investments up front, rather than fighting with Congress over its budget every year. It also should be able to buy good products off the shelf without a tangle of paperwork.

Third, make the agency's leadership more stable. As it is, the administrator has changed every two years or so, and every new crop of political appointees has tried to start from scratch. Administrators need to be there long enough to hold their own in highly technical debates, not just long enough to cash in on a terrific resume builder. Hinson plans on serving for five years; some even think that the term should be seven years long.

But while these changes are important, they do not address the most fundamental question: What is the mission of the FAA? To be truly effective, an organization must have a clear sense of what it is trying to accomplish. Consider the case of NASA in the 1960s. President Kennedy told them to put a man on the moon, and bring him back safely, by the end of the decade. So that's what they did. In the case of the FAA, it seems pretty clear that it should drop "promoting aviation" from its mandate. Why does the aircraft industry need any special help? Instead, the FAA's mission should be simple and straightforward: to do everything in its power to make flying in this country as safe as reasonably possible.

Hinson, to his credit, is trying to move the FAA in this direction. In December, he announced that the FAA was issuing a regulation that will hold commuter airlines to the same safety standards as larger aircraft. Before Hinson, this critical change--a long--time demand of safety critics--had been the subject of enormous resistance from the industry, and, accordingly, from the FAA. Hinson also recently led a follow-up meeting on a "safety summit" held last year, even though the public's sense of emergency has passed. Hinson's relationship with the NTSB is better than in past administrations; he has a better record of accepting NTSB recommendations. And the rough draft of the 1996 FAA "Strategic Plan" goes out of its way to emphasize that safety is the priority.

More importantly, Hinson has put Chris Hardt, a well-respected safety expert and former NTSB board member, in charge of a safety office that reports directly to Hinson. The move is essential because it will give him the power to call the administrator's attention to problems that, in the past, have been buried in the FAA's bureaucracy.

Sometimes, though, it's not the FAA's bureaucracy that is at fault; it's the rest of the government's. Even after the FAA has decided it would to issue a new safety regulation, it must undergo a tortuous approval process. The proposed moves, as slowly as molasses, from the to the Department of Transportation, to the Office of Management and Budget (OMB), then to the DOT and, finally, back to the FAA. At any point it can be derailed, or sent back to the beginning--the federal government's version of Monopoly's "Return to Go." "There are twelve levels of sign-offs," Rep. Oberstar complains. "That's far too much."

How do we speed up the approval process? Well, that's an ideal story for a reporter who wants to investigate how to improve the FAA. There are plenty of other stories:

* To have a good organization, inside or outside the government, you need good people. One GAO report voiced concerns over the FAA's ability to attract talented people in high-tech fields. That's worth investigating and so is the broader question: How good is the FAA at recruiting good people? And does it inspire them to stay?

* After many years of steady decline, the accident rate for private pilots started to climb in 1990. It has been climbing ever since. Why? There is little to prevent a pilot from flying drunk, and, often enough, they do not even have to file a flight plan. General aviation does not get much critical attention--until, that is, some pilot takes to the air reeking of bourbon and then dumps his Cessna into somebody's living room. Has anyone considered spot checks, just like drunk driver road-blocks? (One can't help but wonder if the power of the private pilots' lobby--pilots tend to be a wealthy, influential lot--has something to do with the lax standards.

The challenge of journalism is to find important stories like these and make them interesting. Send a good reporter into the Department of Education, for example, and find out what they are actually doing. Is the Department making schools better? Are the schools making good citizens? (Local papers, which dutifully record every school board meeting but almost never walk into a classroom, would do well to get into the act too. If the government is doing a terrible job, you've got a dramatic story. And if it is doing a good job, then you have an uplifting one. Either way, you have a story that matters.