The first airplane built to U.S. government specifications flew at Fort Myer, Virginia on September 17, 1908, with Orville Wright at the controls. The plane crashed.
Since then, more than 60,000 people have died in U.S. air disasters. And the government has reached the point of spending more than $2.5 billion a year to promote flight safety, through funding of the Federal Aviation Administration.
Air travel has a relatively good safety record. But that record, ironically, has become an obstacle to obtaining a better one. With more than 100,000 people airborne over the United States at any given moment, it’s assumed that some accidents are inevitable. This has led to a kind of benign neglect at FAA. Inertia, inaction, delay, and complacency have cost many lives — because while it may be true that absolute safety is an impossible dream, it’s equally true that there’s still room for much improvement in the way airplanes are built and operated.
The history of air safety confirms that needed changes are usually brought about only after a disaster generates publicity and political pressure. This is called the “blood-on-the-runway” approach to safety. This year we had the spectacle of a DC-10 crashing in Chicago after an engine fell off; later, the plane’s engine mounts came under scrutiny, as did FAA’s certification and inspection proced- ures. But this was only the most recent episode in a regulatory record punctuated by disaster.
Something of a classic in blood-on-the-runway regulating was demonstrated in the case of a 1974 crash that ranks as the worst single-plane accident in aviation history—a tragedy that gave two years’ advance warning before it happened.
The prologue took place on June 12, 1972, in the skies over Windsor, Ontario. What happened there to Bryce McCormick, piloting an American Airlines DC-10 at 11,750 feet, was like driving on an expressway and having the steering wheel come off in your hands. An explosion tore through the plane and most of the controls suddenly became useless. By manipulating the throttles, Captain McCormick wrestled the big jet safely to the ground.
The National Transportation Safety Board—the independent federal agency that investigates air accidents—immediately began an inquiry. There was no question about the cause of this near-disaster: the plane’s control cables had snapped in a chain reaction that began when an improperly designed latch allowed a cargo door to blow off.
The NTSB immediately urged FAA to require that the flaws be corrected on all DC-10s. This could have been accomplished by issuing an Airworthiness Directive, which would have made it a legal requirement to correct the problem. Instead, FAA negotiated a “gentlemen’s agreement” with the plane’s manufacturer, McDonnell Douglas, in which the company promised to modify the planes.
The promised modifications had not been performed on the cargo door latches of a Turkish Airlines DC-10 that took off from Paris on March 3,1974. In flight, a latch gave way and a door blew off. This time there was no piloting miracle. The plane plunged into the French countryside, killing the 346 passengers and crew.
This was not the only time when FAA complacency and foot-dragging have contributed to an avoidable disaster. A few years ago, it was estimated that half the crashes in the world were cases of “controlled flight into terrain”—in which planes were flown into the ground because of poor visibility or guidance error. Devices were available to prevent this by signaling the pilot that he was flying too low. NTSB, among others, repeatedly urged FAA to require these devices on all airliners. No action was taken, even after a Lockheed TriStar skimmed into the Everglades in clear weather.
In 1973, NTSB made the recommendation again, calling the warning systems “the single most effective means of preventing approach and landing accidents.” FAA responded that the system was not yet perfected, and would cost $11,000 per plane. Further study was needed, FAA said. Then, on December 1, 1974, a plane approaching Dulles Airport slammed into a mountaintop, killing all 92 passengers and crew. Two weeks later, FAA adopted a rule requiring the warning systems.
Then there are the cases where even repeated crashes provoke no effective FAA response. The search for a means to prevent midair collisions is an example. Part of the problem is that aviation is an arena of competing interests. Privately owned planes vie with the larger air transport craft for access to the airways, pilots are often at odds with air traffic controllers, airports with the neighboring communities. In the absence of FAA leadership, the contestants just snipe at each other. So it has been with collision avoidance. The pilots want a cockpit-based collision avoidance system—which would sense an impending collision and direct evasive action. The controllers, who happen to be FAA employees, want the system based on the ground with them. The result: no system at all, even after last year’s San Diego collision between a Boeing 727 and a private plane. The latest estimate from the FAA is that something will be ready for operation by the mid-1980s.
Delays stretching to decades are particularly incongruous in an agency devoted to aviation — an industry that loves to rhapsodize about its rapid development. The agency responsible for the safety of this industry often moves, as one critic puts it, “in a range between slow and very slow.” Despite the deadly consequences of failing to act swiftly, new rules or changes in existing ones can take years to be passed, even when an unmistakable hazard is involved. In 1975, the General Accounting Office reported that it takes an average of 24 months after a rulemaking is approved for the final change to be issued. GAO found cases where the delay was more than five years.
While FAA is vested with enormous powers to “promote safety of flight,” these powers have been exercised with a benevolent restraint. So most of the thousands of inspections that each airplane must pass are performed not by FAA employees but by employees of the aircraft manufacturers, authorized by FAA to pass judgment on the quality of their own products. So, when the first DC-10 cargo door blew off, no order was issued requiring modification, for fear of inflicting financial hardship on McDonnell Douglas. And so, when a rule requiring the “ground proximity warning system” was finally issued, airline pressure held the compliance deadline off for two years.
As one former chief of FAA’s evaluation division put it, “FAA is really a democracy, in that if those it regulates don’t like a proposed rule, FAA will abandon it.”
The National Transportation Safety Board is one of the smallest agencies in the government; its $16-million annual budget would keep FAA going for two days. Created in 1966, NTSB was made 52 completely independent in 1974 after word reached Congress of a Nixonian effort to mute the Board and bring it under White House control. Even though it remains in the Department of Transportation building, with FAA offices above and below it, the Safety Board fastidiously guards its position as an independent enclave.
Most observers agree that NTSB has been successful at maintaining its credibility over the years. It is also one of the few government agencies in which at least some of the politically appointed members must have actual qualifications for the job. Of the board’s five members, the act requires that “no less than two. . . shall be individuals who have been appointed in the field of accident reconstruction, safety engineering, or transportation safety.”
Aircraft accidents are usually the result of a series of errors and failings, and NTSB investigators are masters at piecing together the story from the most elusive of clues. Audio specialist Paul Turner, for example, interpreted a single click on the tape from a cockpit voice recorder to pinpoint a switch that had been mistakenly turned off, leading to a crash.
Not all NTSB investigations are greeted with acclaim, however. Its findings are sometimes disputed, often by the Air Line Pilots Association, the union that represents the pilots of most U.S. carriers. ALPA generally takes the position that when an investigation ends with an accusation of “pilot error,” the investigators haven’t gone far enough. ALPA’s objections are given serious consideration, because most airline pilots are tough safety advocates. (“We’re always the first ones at a crash,” they like to point out.) If a pilot makes an error, ALPA wants to know what was in the system that permitted him to make that error and permitted the error to lead to disaster.
When an Eastern Airlines DC-9 crashed into a cornfield on approach to Charlotte, North Carolina, there was no doubt that the pilot had been flying too low. Pilot error—but why? At least part of the answer lay with the plane’s altimeter, which could be easily misread. Called a drum-pointer, this confusing instrument is still in wide-spread use, even though it has been implicated in several other crashes and even though an altimeter with an unmistakable digital display is available.
If planes are found to be flying with obsolete and dangerous instruments, why doesn’t somebody flat-out order them replaced? If not FAA, why not NTSB? But NTSB has legal authority over neither industry nor FAA.
“We regulate by raised eyebrows,” says NTSB vice chairman Elwood Driver. There is actually a sensible idea at work in this arrangement; the theory is that enforcement and investigation should be kept separate, so that the enforcer cannot be placed in the awkward position of investigating his own adequacy. But NTSB is frequently criticized for not doing more to cross the subtle line between passive and active recommendations—it could be much more persistent (and public) about badgering FAA when FAA is sluggish about responding to NTSB’s recommendations. NTSB says that on 72 per cent of its recommendations FAA has taken action it finds acceptable. But that’s cumulative, going all the way back to the day NTSB opened for business in 1967. To get a better idea of the situation, look at the record for any single year. Of the 71 NTSB aviation recommendations issued in 1977, for example, 56 were still awaiting FAA action halfway through 1978.
Still more revealing is a look at which types of recommendations receive the slowest FAA response. According to Driver, most of these involve factors that determine your chances of surviving a crash.
An airplane falls out of the sky and the impact destroys it, instantly killing everyone on board. To most people, that’s what a plane crash is all about.
But in fact this description fits only perhaps 30 per cent of crashes. There is another type of accident—the “sur- vivable” crash, in which the plane is only damaged initially and most if not all of the people on board survive the impact. What kills in a survivable crash is what comes after the impact: fire, smoke and toxic gas.
Typically, a survivable crash goes like this: The impact in a landing mishap ruptures a fuel tank, touches off a spark, knocks out the power, and tears some of the passengers’ seats loose from their moorings. The lights in the cabin go out, the escape routes are blocked by the scattered seats, and fire and smoke fill the plane in less than a minute. The odds of getting out of such a crash alive can be so bad as to suggest that “survivable” is a mislead- ing adjective. The fire potential alone is awesome. A fully loaded 747 may carry 40,000 gallons of highly flammable aviation fuel at takeoff.
But burning fuel is not the only thing to fear in an aircraft fire. Next time you fly, take a look around you as you settle back in your seat. Survey the tasteful, restful cabin interior—and reflect that virtually everything you see is potentially lethal. If they began to burn, the seat cushions and carpeting would give off hydrogen cyanide, the wall panels would give off hydrogen fluoride, and carbon monoxide would be generated by just about everything in sight. The cabin of a typical wide- body jetliner contains nearly five tons of mostly synthetic materials capable of giving off huge volumes of smoke and toxic gas when they burn.
Crash investigators have been worrying about this aspect of “survivable” accidents for years. When a Lockheed Constellation crashed in Virginia in 1961, for example, investigators found that all 79 aboard had survived the impact — but 77 died within minutes from carbon monoxide poisoning. This was a pattern repeated in 1965 at Salt Lake City, in 1970 at Anchorage, in 1979 near Athens—in more than 300 accidents over the past 15 years.
The only way to save lives in “non-survivable” crashes is to prevent them from happening. But “survivable” crashes can be made less deadly by reducing the threat of fuel fires and explosions, and by using cabin materials that won’t burn readily and that emit little smoke and poisonous gas when they do burn. Although FAA has been struggling with this problem for more than 30 years, it has little to show except a regulation that requires cabin materials to be self-extinguishing — which means a dropped cigarette won’t trigger an inferno. But this regulation (adopted in 1946) doesn’t address the problem of coping with flames, smoke, and toxic gas when the plane’s fuel catches fire. Equipment manufacturers have lobbied hard and successfully thus far to keep such regulations from being imposed, arguing (in the words of DuPont) that “compartment materials are not a major… problem [compared to] crash impact and jet fuel.” In other words, the fuel tanks will get you before the seat cushions do. NTSB’s files, however, are full of evidence that thousands of people have died from toxic gas — and then there are grisly cases like the Varig 707 that crash-landed near Paris with 123 dead people on board, killed by a cabin fire that never spread to the fuel system. Thus far, industry has managed to persuade FAA that what’s needed is something called a “combined hazard index”—a method for figuring out the synergistic effects of everything that can kill you. It sounds appealing but, conveniently, it’s beyond the current state of the art. So FAA has given $397,000 to McDonnell Douglas to think about it.
State of the art was also the reason given for withdrawing the rule on fuel system protection—another “survivability” problem that has gone uncorrected for decades. In 1963, a Pan Am 707 was destroyed in flight by a fuel tank explosion caused by a lightning strike, and the same thing happened to an Iranian Air Force 747 in 1976. Tank explosions can also destroy planes on the ground—as the takeoff crash of a TW A 707 in Rome demonstrated when fire set the tanks off.
According to both FAA and the industry, a satisfactory method toprevent fire and explosion has not yet been devised. There have been attempts at explosion-proofing fueltanks by replacing the air in them with nitrogen or filling them with plastic honeycomb, and some of these systems are in use today in military aircraft— including Air Force One. But there were added costs, weight penalties and maintenance problems that the commercial aviation industry—and FAA—called unacceptable.
Furthermore, these systems offer only limited protection when fuel tanks rupture in a crash. FAA is currently testing an additive that prevents fuel from turning to mist when it sprays out from a ruptured tank—important research, because fuel ignites much more readily as a mist than as a liquid. Initial tests of the additive have been successful, but FAA is quick to point out that more testing is needed. How much more? FAA’s best estimate: several years.
Race, Sex, and Safety
After the industries that would have been bound by crash survivability standards convinced FAA that such standards were beyond the state of the art, FAA met privately with industry representatives to discuss what to do instead. Ralph Nader’s Aviation Consumer Action Project objected, because closed-door sessions between a regulator and the industry it regulates are illegal. The industry, specifically the airframe manufacturers, then came up with the idea of a public advisory committee in which both government and industry would be represented, along with a smattering of representatives from consumer interests, academe, and professional organizations. It was called SAFER (Special Aviation Fire and Explosion Reduction Advisory Committee), and it was established to “consider all aspects of post-crash aircraft fires and… recommend a program of corrective action to the FAA Administrator.” To SAFER’s supporters, the committee is the first coordinated approach to aircraft fire safety, the first to consider all the complexities in drafting a set of standards, the first to draw on all available expertise. To skeptics, it’s another delaying tactic dreamed up by industry to forestall regulation.
So far the skeptics have largely been proved right. SAFER was given a two- year mandate, beginning June 1978. Selection of the members was to be completed by October 31, and within 30 days after that, FAA administrator Langhorne Bond promised in a letter, the committee was to hold its first meeting. SAFER didn’t meet until May of the following year. The delay, FAA claimed, resulted from carrying out the time-consuming task of collecting information on the race, sex, and political affiliation of the advisory committee members.
But the problem is not so much SAFER itself as the fact that it has taken the place of FAA rulemaking. Research should go on, if need be, under SAFER’s aegis. But new regulations—of the sort FAA was supposed to start developing in 1969—should not have to wait for the completion of that research. The optimum cabin material is still to be found, perhaps, but FAA’s own research has identified some of the worst materials for use in an aircraft interior, and those at least should be banned. Better test methods may be discovered later, but the ones in hand are good enough for at least a first set of standards. And as for the relationship among smoke, gas, and flammability, say critics, new materials are already available with better characteristics in all three areas.
SAFER has in fact made a few recommendations for an interim standard to cover smoke, gas, and flame in cabin materials. But FAA is under a tight deadline if it wants such rules to save many lives. The reason is that a new generation of passenger aircraft will appear in the early 1980s. When the last generation, the wide-bodies, began to appear a decade ago, it was a time for upgrading regulations.
Stiffer flammability standards were issued for the new planes. If there is no regulatory spring cleaning at this time, planes such as Boeing’s 757 and the McDonnell Douglas DC-9-80 will be manufactured according to no standard for smoke or gas and in line with an aging flammability standard. These planes may be flying well into the next century, and if they start coming off the production line before the new standards are in effect, it will be many years before their cabins are upgraded.
Into the Harbor
Survival in a plane crash does not depend on fire protection alone; it’s also a matter of rapid and safe evacuation. Here, too, FAA regulations are at issue.
The Aviation Consumer Action Project and the Association of Flight Attendants have petitioned FAA to require that airplane seats be able to withstand at least as much as the people who sit in them. There’s no such requirement today. When a 20-seat Twin Otter crashed in Maine, there was no smoke or fire, and the cabin was not seriously deformed; yet all but three of the seats ripped loose from their moorings, and all but one of the 18 passengers were killed.
The current seat standards were adopted in 1952, and have been upgraded only slightly. FAA admits that improvement is needed, but says that research must precede rule-making. And so a “man/ seat dynamic computer model” was developed to study the problem. In a 1976 FAA report, that model was reported to be completed; yet the agency says that at least four more years of research are needed. There are also dangerous problems with the inflatable slides that passengers and crew are supposed to use in an emergency to get from the doorway to the ground. In 1978 a DC-10 at Los Angeles aborted a takeoff after three tires failed, and one leg of the landing gear gave way before the plane came to a stop. With the plane closer to the ground than normal, one of the emergency slides extended at the wrong angle. Overloaded with passengers who couldn’t get out fast enough at the bottom, the slide tore loose. The rest of the slides failed because of exposure to heat and flame. Even how many airports are located next to large bodies of water, and wonder why flotation cushions aren’t standard on all flights.
FAA claims to have found a new concern with crash survivability. “I think our focus became too devoted to accident prevention,” says FAA administrator Bond. He has promised that “more attention and effort will be directed toward helping people survive accidents.” The agency’s experiments with the anti-mist gas additive are offered as one example of this concern, along with a program for fire-testing cabin materials in a full-scale airplane interior, using the fuselage from an old Air Force cargo plane. New evacuation systems are to be tested there as well. Yet, when you add up all these projects, it turns out that less than $10 million of FAA’s $80-million research and development budget is now spent on fuel system and cabin safety research, with the balance still going for accident prevention projects.
To be sure, some safety improvements are appearing even without government pressure. The cabin of the DC-9-80, for example, is to have aluminum sidewalls, rather than the more flammable plastic-based ones. “If you look at the proposals that Boeing now has for their new aircraft,” says Joseph Del Balzo, acting director of FAA’s National Aviation Facilities Experimental Center, “standards they’ve set for themselves are higher than the standards being used today.” But letting the rules lag behind industry performance hardly seems something to boast about. FAA’s own ad hoc committee to investigate certification of the DC-10 after the 1974 Paris crash warned against letting “regulatory policy … continue to fall behind the state-of-the-art in today’s rapidly advancing civil aircraft design and production technology.”
And despite all the concern about state of the art, it may be that cabin safety is more an economic problem than a technological one. There are companies already able to develop cabin materials superior to those found in today’s airplanes. At least initially, however, these new materials would cost more, and so before going into development and production, the companies need assurance that the aircraft manufacturers will be customers. The aircraft manufacturers, in turn, need assurance that the airlines will buy airplanes with the new materials.
Without rules requiring it, there’s little incentive for the airlines to pay the additional initial cost. Accidents prevented don’t show up on a balance sheet, and neither do lives saved in a survivable crash. (The airlines can’t even promote their safety efforts in their advertising. Under unwritten agreement among the companies, an airline may tout your chances of enjoying one of its flights, but not your chances of surviving it. When Pacific Airlines broke the taboo by hiring humorist Stan Freberg to design some ads aimed at reassuring fearful passengers, the executives responsible lost their jobs.)
This is precisely the sort of situation that cries out for strong, effective, creative regulation. Creative? Yes, and in an economic sense. If FAA issued clear-cut safety requirements, an incentive would be created to produce such materials at reasonable cost, anda market would automatically be created for the new products.
But this is not the way FAA thinks. And that is why few people, outside the air industry, are happy with the Federal Aviation Administration. W hen you read safety proposals written ten years ago, and realize they are still pending today, you get a sense of time suspended. At FAA headquarters in Washington, you can hear the jets approaching and departing from National Airport, and it is a matter of statistical record that the great majority will complete their flights uneventfully. The sounds of those successful flights are the sounds of complacency at FAA, the sounds of a regulatory lullaby.