Learning from Disasters

Sept. 14, 2005 -- -- In an ideal world, no mistake would ever be repeated, since lessons learned from mistakes (especially those ending in tragic aviation accidents) would be immediately and permanently applied to change the safety system.

In fact, when you climb on any commercial airliner, this is a concept you trust your airline believes in to the depth of its corporate existence.

Uncovering such safety "lessons" is what the National Transportation Safety Board does. The NTSB is a mostly independent government agency that investigates all air, highway, ship, pipeline and rail accidents. It is based in Washington, D.C., and consists of more than 300 highly professional staff members. Their profession is to squeeze every ounce of useful, corrective information from every accident and not rest until all of it has been applied to make the safety system better.

In other words, the NTSB is there to make sure we don't make the same mistakes twice.

But, air safety in particular costs money. Untold barrels of money. A simple change, for instance, in the structure of an airliner that can improve the margins of safety can also cost a half million dollars per airplane. When a financially troubled airline is flying 500 or 600 such airplanes, paying for such changes is often a practical impossibility.

This is the intersection at which all the noble intentions of airline leaders and governmental regulators and NTSB investigators sometimes collide. The NTSB can recommend, but cannot order, the airlines to make changes for safety. The Federal Aviation Administration, which can issue such orders, is also required by Congress to be careful not to damage the airline industry's ability to stay solvent. And, of course, while the airlines want perfect safety, there will always be disagreements over what it takes to achieve it.

There is no clear, visible dividing line between "safe" and "unsafe." So when an airline is hemorrhaging cash, the temptation to redefine how much safety is enough becomes all but overwhelming. It's very hard for a beleaguered management team to keep from pressuring the very expensive maintenance and training functions of its airline for savings while mugging the employees for concessions and dancing around outside bankruptcy court like a child delaying a trip to the bathroom.

The good news is that for the most part -- and despite gigantic losses in the tens of billions that have in effect wiped out every dollar of profit ever earned by the airline industry -- America's airlines have done a superlative job behind the scenes of maintaining their expenditures and focus on safety. After all, the airlines invented near-perfect safety, with the FAA cheering them on, not creating the condition. (The industry has traditionally kept itself far above the FAA minimum standards, which is why we're approaching four years in the U.S. without a major airline accident).

When it comes to really knowing what's going on in the heart of an airline, large or small, we have a societal problem: Only the airlines themselves are really watching. The FAA is not a police force and simply can't be everywhere. Just as was true in the bloody '80s when airline deregulation sparked a long series of accidents, the FAA's main emphasis is on paperwork, and there have been examples of unscrupulous smaller companies who have maintained good paperwork for the FAA while leaving their actual airplanes in somewhat less superlative condition -- a problem the FAA didn't or couldn't discover. With red ink flowing through the hallways of most major legacy airlines, perfect safety compliance must be inspected, not just expected.

But if there's no accident, there's no problem, right?

Wrong. And that is the entire point of this column: "The absence of an accident in no way proves the presence of safety." That cogent statement was made in the late '80s by Dr. John Lauber, formerly a board member of the NTSB, to describe why we desperately need ways of measuring and monitoring the true safety levels of an airline, instead of waiting for accidents to tell us something is wrong.

One of the most heart-wrenching lessons occurred in the predawn darkness of Dec. 12, 1985, in Gander, Canada, when a chartered DC-8 carrying 248 American servicemen and women of the 101st Airborne home for the holidays crashed on takeoff, killing all aboard.

While there is still some controversy over whether ice on the wings made sustained flight impossible or whether something else occurred, the hearings that followed on Capitol Hill revealed a chilling truth: The Pentagon had routinely depended on the FAA to monitor the health of airlines chartered to carry military members around the world, but the FAA was doing little more than confirming that an airline's certificate to operate was current. No one was inspecting or watching.

In the case of Gander, intense post-crash investigation of the airline uncovered a pervasive pattern of questionable maintenance, training, operational and philosophical compliance levels which, if known to the Defense Department, would have put it on a blacklist. From that tragedy grew a method of inspecting and monitoring the real safety levels of airlines used by the Pentagon, and that method should be (and isn't yet) the gold standard for FAA's civilian oversight.

Called the Air Carrier Survey and Analysis Office, the fledgling operation began at Scott Air Force Base in Illinois and rapidly began teaching the FAA how to monitor the safety health of an airline by looking carefully and constantly at the metal airplanes, the crews, the philosophy of compliance, the finances and attitudes of the leaders. Because they didn't stop with the paperwork, the ACSAO teams rapidly became the most feared in the industry, far beyond the FAA. If you had an airline and wanted to charter to the Pentagon, you had to pass ACSAO's stringent scrutiny, and that wasn't easy. In addition, it set up methods of monitoring on a daily basis any reports of financial or other operational problems from a wide variety of sources with the ability to swoop in and suspend any carrier if necessary.

Yes, there have been attempts to incorporate this much-expanded method of safety monitoring into the mainstream FAA's techniques, but those attempts need to expand to a major overhaul of FAA oversight, especially now, when the river of red ink from airline losses is once more threatening to put pressure on the very things that make airline flying safer than any other form of transportation. Even the reality that a depressed work force is less safe must be considered.

We have an amazingly reliable level of safety in commercial aviation, but we absolutely must consider safety problems from financial pressure a possibility the same way we should have dealt with the potential of flooding in New Orleans as a clear and present danger.

Costly lessons should be learned by a wise people only once.