This week the Federal Aviation Administration will rule on the most contentious issue to emerge from the tragic explosion of TWA Flight 800: what more to do to prevent other center tanks - empty of fuel but filled with fumes - from igniting into a deadly blast.
The question has caused a rare public rift between the FAA, which regulates the airlines, and the National Transportation Safety Board, which is responsible for investigating and preventing accidents. Last week, it also prompted the FAA to issue new guidelines for dealing with certain 747 fuel pumps and wiring. And it has once again raised one of the most difficult questions for guardians of the nation's air-traffic system: How safe is "safe"?
"By all objective measures, the accident rate is lower than it's ever been. The system is quite safe," says Richard Golaszewski, executive vice president of GRA Consulting, a transportation consulting firm based in suburban Philadelphia. "But it's not clear the public thinks it's safe enough."
In 1960, there were 28 airline accidents per 1 million departures. In the United States today, there is fewer than 1 accident per million departures.
But such statistics have done little to quell public concerns about the safety of flying. The ValuJet crash of May 1996, in which 110 people died, and the TWA explosion two months later, in which 230 people died, were extraordinary disasters. But both served to undermine public confidence in the air system.
"The FAA is always talking about how safe the system is, but it's like the public got a slap on one cheek, then the other," says David Stempler, president of the Air Travelers Association, a Washington-based passengers-advocacy group.
Early speculation about the cause of each crash, and the media attention that followed, helped feed the public's dismay. Soon after the ValuJet crash, then-Secretary of Transportation Federico Pea insisted the airline was so safe he'd fly on it anytime. Soon after, however, a Department of Transportation memo surfaced that indicated the low-cost carrier had one of the worst safety records in the industry, by certain measures. The cause of that crash was later determined to be oxygen tanks that had been improperly loaded into the plane's cargo bay.
The FAA responded by banning the transport of certain oxygen generators in passenger aircraft, and requiring fire-detection and suppression systems to be installed in all planes that didn't have them.
With TWA 800 in July 1996, initial fears that a terrorist bomb or a missile caused the explosion prompted a review of the nation's security systems.
It turns out, in Mr. Stempler's words, the systems "were a sieve." Despite the December 1988 terrorist attack against Pan Am Flight 103 over Lockerbie, Scotland, in which powerful plastic explosives were placed inside a radio cassette player, most American airports had only metal detectors that identify things like handguns or knives. Anyone with a pocketful of plastic explosives could still walk onto almost any airplane and take a seat.
In the aftermath of Pan Am 103, a presidential commission recommended that machines capable of identifying plastic explosives be installed at all airports, that passengers be profiled to identify security risks, and that bags be matched to passengers on the plane before they're loaded. None was implemented. Then TWA Flight 800 exploded.
Another presidential commission - known as the Gore Commission - made almost identical recommendations last winter. Many are now being implemented. Passengers meeting certain criteria are now being profiled, a pilot "baggage match" program is in the process of being expanded nationally, and more than 50 new detectors that can trace plastic explosives have been or will soon be installed in some of the nation's largest airports.
Over the weekend, the FAA also began a process that could result in passengers being able to bring with them less carry-on baggage.
To help pay for the safety improvements, Congress just granted the FAA an additional $100 million to "improve safety" and $44 million for research.
But with a missile or a bomb ruled out as the cause of the TWA 800 explosion, the focus has shifted back to the difficult, and to some, more unsettling problem of mechanical failure.
On Dec. 8, the National Transportation Safety Board (NTSB) will hold week-long hearings on its TWA 800 investigation. How it proceeds could be determined, at least in part, by what the FAA does this week.
Almost a year ago, the NTSB recommended that airlines keep the tanks of their planes partially full of fuel or completely full of inert gases. The airlines balked at the cost. They reminded the safety board that the cause of the deadly spark has not yet been determined.
Soon after the crash, the FAA required the immediate and regular inspection of fuel-pump wiring, but it, too, hesitated at the NTSB's more sweeping recommendations. It said there was no scientific data to show that such changes would improve overall safety - and the agency insisted it needed to know more about the way jet fuel reacts in tanks at different temperatures before it could make a decision.
"The FAA advised the board that its recommendations could have sweeping ... impacts, potentially affecting approximately 5,400 aircraft in the US alone, including Boeing 737s, 747s, 757s, 767s, as well as aircraft manufactured by other companies," wrote the FAA last July in response to criticism over its handling of the NTSB proposals.
Last April, the FAA asked for public comment on those proposals. Last week, it announced immediate changes to some fuel pumps and asked for comment on proposed wiring changes.
NTSB chairman James Hall said in a statement he applauds "any safety initiative." But neither FAA action addressed the NTSB's recommendations. In fact, the safety board had not proposed either move.
The tension between the two agencies, with their different responsibilities, is not new. But the public nature of this dispute is extremely rare - and that, say analysts, works to the disadvantage of the FAA.
"There's a long history of the NTSB making safety recommendations irrespective of what they cost - in some sense that's their job," says Clint Oster, a professor at Indiana University at Bloomington who researches aviation safety. "Unfortunately for the FAA, its job is to take a much harder look at those recommendations and determine which ones make sense from a broader perspective."
That broader perspective includes calculating how much proposed changes would cost, how significantly they would increase overall air safety, and the overall costs and benefits of each.
To some people, that approach is short-sighted. They'd like to see every safety recommendation implemented. Their goal: a zero-accident rate.
"They may be setting a very, very high standard - to eliminate all of the risk," says Mr. Golaszewski, "the implications of which would be a much more costlier air-transportation system."