How Did Hijackers Get Past Airport Security?

As 19 hijackers made their way along the concourses at three East Coast airports on Sept. 11, bent on executing the deadliest terrorist attack in history, they were subjecting the U.S. aviation security system to its most critical test.

At almost every step along the way, the system posed no challenge to the terrorists--not to their ability to purchase tickets, to pass security checkpoints while carrying knives and cutting implements nor to board aircraft.

The system worked the way it was intended, according to all the available evidence. For three decades, it has been preoccupied with looking for guns and explosives rather than for dangerous people. That, security experts and aviation professionals say, was its vulnerability. The terrorists did not breach the nation’s airport security system; they slipped through its loopholes.

Nothing in Federal Aviation Administration rules and regulations--even assuming they were followed to the letter--would have prevented the hijackers from carrying out their baleful missions, considering what is known so far about the plot to hijack jetliners and suicide-dive them into the World Trade Center and the Pentagon.

“The absolute truth is I have not been informed of any breaches in security that took place on Sept. 11,” said Susan Baer, general manager at Newark International Airport in New Jersey, where one of the hijacked planes took off. She echoed similar assessments by law enforcement and airline industry officials of the hijackers’ success at also boarding planes at Boston and Virginia airports.

The governing assumption has long been that a hijacker’s principal goal is to use the aircraft as a bargaining chip, whether for political purposes or to reach an unscheduled destination.

Until the very morning of Sept. 11, leaders of the security community had not focused on the threat that actually materialized: squads of lightly armed hijackers seizing airliners as instruments of suicide and destruction.

But an attack similar to this month’s, if not on the same horrific scale, was not entirely unforeseen. On May 7, Brian Sullivan, a former FAA special agent at Logan International Airport in Boston, expressed his concerns in a letter to Sen. John F. Kerry (D-Mass.).

“What protection is there against a rogue terrorist? And with the concept of jihad, do you think it would be difficult for a determined terrorist to get on a plane and destroy himself and all other passengers? . . . With our current screening system, this is more than possible, almost likely.”

Kerry passed the letter on to the General Accounting Office, the congressional investigative agency, which was examining security issues.

Why the nation’s aviation security system failed to anticipate a threat that, in retrospect, might easily have been theorized by experts will weigh heavily on U.S. efforts to upgrade the security arsenal. Senate and House panels started that process last week.

Only Friday, Inspector General Kenneth M. Mead of the Transportation Department recommended to a House subcommittee that a federal agency or government entity should be entrusted with all screening of passengers, aviation employees, baggage and cargo; that it take over the Federal Air Marshal program (a small cadre of undercover officers who fly as passengers); and that it develop advanced security equipment.

Fighting the new threat requires radically different security systems, assumptions and even equipment, including new screening technologies that will subject passengers to unprecedented intrusiveness; changes in the interior construction of airliners and the operating procedures of cockpit crews confronted with disturbances; and much more aggressive profiling of passengers to identify potential terrorists.

But aviation experts say none of these measures will do much good if more fundamental flaws in the system aren’t quickly repaired.

Indeed, one of the U.S. security system’s principal weaknesses is that individual pieces of the overall program are uncoordinated, so that information gathered by one sector is not communicated to others. Although a little-known profiling program used by most U.S. carriers identifies potentially high-risk passengers, for example, gate agents are not required to subject those travelers to special scrutiny.

“To say no one can use a knife today even in an airport restaurant is to shut the door after the horse is gone,” says Issy Boim, a former Israeli security official who worked for 17 years with El Al, that country’s national air carrier. More crucial, he says, is designing counter-terrorism measures in a system of interlocking and complementary rings. “In aviation, there’s no one solution.”

A team of Times reporters has traced the paths of the 19 suspected hijackers through the nation’s aviation security system as it existed before Sept. 11 and found:

* Airport security checkpoints, even if they did detect such weapons as small knives and cutting tools, typically did not consider these as threatening.

* The sophisticated profiling system designed to screen for high-risk passengers apparently overlooked the hijackers because, based on the history of domestic hijackings, it did not contemplate a suicide mission.

* Commercial aircraft design offered almost no protection from cockpit takeovers. Such attempts, in fact, have been growing more frequent in recent years.

The Arrival

Sometime before 8 a.m. Sept. 11, five men strolled through the airy, spacious lobby of Logan International’s Terminal C. As they approached the security checkpoint, they encountered the gentle jingling of the concourse’s artistic centerpiece, a Rube Goldberg-like contraption housed in a 6-foot glass cube.

About 50 yards past the sculpture is the security checkpoint leading to Gates C-11 through C-21. Airport officials say the hijackers of United Airlines Flight 175, bound for Los Angeles, passed through here on their way to Gate 19. In the two hours after their 7:58 a.m. departure, they would use weapons believed to have been carried past the security screeners to commandeer the plane and steer it into the south tower of the World Trade Center.

Fourteen other men followed similar paths that morning through security posts at Logan’s Terminal B, serving American Airlines, and on airport concourses in Newark, N.J., and at Dulles International near Arlington, Va. Reports differ on whether they checked bags.

Each was likely subjected to the same screening familiar to the 1.6 million passengers who fly U.S. skies every day: a perfunctory X-ray scan of carry-on bags; a walk through a magnetometer portal tuned to detect metallic objects; and very occasionally, a hand search of personal belongings.

Whether the magnetometers could or should have detected the knives or box cutters that investigators believe the hijackers carried on board is unknown. But security professionals do regard the devices, which detect aluminum and other metals containing iron, as far from foolproof. Operators can turn down their sensitivity to speed up passengers’ movement through the line.

“The FAA establishes criteria [for the settings], but when the airlines apply pressure, they often turn them down [below FAA specifications],” according to a senior FAA security source.

As for the X-ray equipment used to screen carry-ons, these are easily snookered because they transmit scanning beams vertically. Thus, a knife laid on edge can look as slender as a wire. Most major airports are equipped with more capable scanning systems using computer tomography similar to hospital CAT scanners. These produce three-dimensional images, but are slower and used only on a small percentage of checked bags and an even smaller percentage of carry-ons.

The security companies that operate the screening posts under contract with American and United airlines--Argenbright Holdings Ltd. at the Dulles and Newark terminals, Huntleigh USA at Logan’s United terminal and Global Aviation at Logan’s American terminal--all say the checkpoints were fully staffed that morning.

“All checkpoints were open, all were operational and fully staffed,” said Bill Barber, president of Argenbright, referring to the Newark and Dulles stations.

In any event, many experts agree that the training, expertise and work patterns of staff at airport security checkpoints have been inadequate. Some workers have complained to union officials that rules mandating breaks at least every half-hour to keep them alert are routinely violated.

Mead, the Transportation Department’s inspector general, testified Friday that inspectors successfully entered off-limits areas at eight major U.S. airports in 68% of their attempts in 1998 and 1999. They were able to board aircraft 117 times without going through security, often because airport and airline employees didn’t close security doors behind them.

FAA enforcement and oversight of the checkpoints also has come under fire. Although agency inspectors regularly run undercover tests of screening stations at airports nationwide, they give screeners the benefit of the doubt.

“For a time, the inspector general was forced to use official FAA test weapons, which were easily recognized by screeners,” says Mary Schiavo, a former inspector general of the Department of Transportation and a vocal critic of the FAA. “It was the same stupid Samsonite briefcase every time. They all recognized it--it might as well have had the FAA seal on it.”

Weapons and explosives cached inside the bag were designed to be hard to miss: A fake bomb looked like sticks of dynamite attached to an alarm clock by long, curly wires.

“It was like the bomb from Acme Supply on the Road Runner cartoons,” Schiavo says. When she suggested taking the fake bomb apart and secreting the pieces, “the screeners said that wasn’t fair. There were actually negotiations with the screeners over what kind of items they should be expected to pick up.”

Argenbright has admitted to sloppy practices. The company pleaded guilty last year to hiring more than 1,300 untrained screeners--including dozens of criminals whose backgrounds had not been checked--to work at Philadelphia International Airport from 1995 to 1998.

Argenbright was fined $1 million and ordered to pay $350,000 in restitution to the airlines it defrauded. Argenbright’s Barber called this an “isolated incident” and said the company responded by adopting a code of conduct, ethics training and an internal audit system.

Even if the terrorists’ baggage received unusually sharp-eyed scrutiny as it passed through the X-rays, even if the men were checked with manual wands because the magnetometers detected metal, this much has been known for years to security experts: The checkpoint screeners would have been looking for the wrong things.

The Threat Vector

The nature of the hijackers’ arsenals is still not fully known, and may never be, but cell phone calls from passengers on three of the hijacked jets indicate most were armed with cutting implements. Passengers on United Flight 93, which crashed in a Pennsylvania field, also reported that their hijackers claimed to have a bomb and that one displayed a box strapped to his chest. It isn’t known whether the box was a genuine bomb or a prop.

But an FBI official told The Times that investigators have not found evidence that any of the hijackers used weapons that were prohibited aboard airlines under longtime FAA guidelines.

In any event, as the attacks made clear, “you don’t need a weapon to hijack an aircraft,” says Stephen Luckey, a former Northwest Airlines pilot who is chairman of the national security committee of the Air Line Pilots Assn. “There are any number of improvised weapons. You could shove a pencil in someone’s ear. If you’re big and mean and grab someone by the throat, you can gain access to an airplane.”

Security experts have been warned about the shortcomings of the U.S. security system’s preoccupation with locating dangerous objects. In 1996, Dan Issacharoff, a former security chief at El Al, told a National Research Council panel on airline security that, although El Al inspects all baggage, it also conducts face-to-face questioning of all passengers.

“There’s a difference of philosophy,” says Robert Berkebile, a panel member. “U.S. security efforts aim to catch the bomb; El Al’s is to catch the bomber.”

Issacharoff acknowledged in his 1996 briefing that the practices of El Al--widely regarded as one of the best carriers in terms of security--could be impractical in the U.S., which serves twice as many airports and carries millions more passengers a year.

To a certain extent, U.S. security focused on baggage because it is easier to track and examine than people. The instrument that carries potential peril is technically known as the “threat vector,” whether that is an explosive, a vehicle or a team of individuals.

“Our thoughts were that if you have a madman willing to kill himself, frankly there’s very little you could do about it,” Berkebile says. “There are so many threat vectors at the airport that a person willing to take his own life is almost impossible to stop.”

Still, there is no doubt that even the U.S. efforts to focus on dangerous objects have been flawed and inconsistent. FAA policy about knives in carry-on baggage is one example.

A policy barring knives with blades longer than 4 inches dates to the early 1970s, when airport screening first began, according to Gerald Dillingham, director of civil aviation issues for the GAO. The policy, he said, evolved as a blend of local laws and cultural norms, such as the routine practice of carrying pocketknives.

Another person familiar with the history of the regulations said many items that fell within the permissible standard, such as double-edged razor blades and craft knives, were considered harmless. Also, he said that detection of blades shorter than 4 inches would have been nearly impossible because they are so easily masked by other items.

The policy, says Cathal Flynn, the FAA’s associate administrator for civil aviation security from 1993 until 2000, was “common-sensical borne out by experience,” adding that he could not recall a single serious hijacking incident involving a knife prior to Sept. 11.

The FAA’s weapon guidelines did not specifically mention razor-sharp box cutters--among the instruments said to be wielded by the hijackers. Some security screeners said they prohibited them; others said they permitted them as being within the 4-inch knife standard.

Profiling the Passengers

Once through security, the hijackers made their way to their gates--boarding a mobile shuttle at Dulles, ambling along the corridors at the other airports--and approached gate agents. As agents typed their names into airline computers and called up their reservations, each also was checked against a little-known but highly sophisticated database.

This is CAPS, for “computer-assisted passenger screening.” Developed in the mid-1990s by Northwest Airlines and deployed by most major domestic carriers by 1998, CAPS is nothing less than a profiling program designed to distinguish passengers who pose no threat to safety from those who might.

CAPS has its limitations. Among other things, because its profiling criteria are based on previous hijackings, it missed the hijackers--a flaw that is already being addressed, according to people familiar with the program. Nevertheless, CAPS is regarded by security professionals as perhaps one of the most powerful tools in the aviation safety arsenal. Yet, as with so many other elements of the system, it has been employed ineffectually.

CAPS is designed to be triggered every time a customer purchases a ticket on a major U.S. carrier. In the case of the hijackers, it probably noted that (as detailed in an FBI document obtained by The Times) all five suspects on United Flight 175 had purchased one-way tickets. Of those boarding American Flight 11 at Logan, all had made their reservations within two days of each other, four had paid with Visa cards and one with cash.

Of those boarding American Flight 77 at Dulles, two had reserved their round-trip flights in August using the airline’s Web site, purchasing their tickets with cash 10 days later at Baltimore-Washington International Airport, and recording their frequent-flier program numbers. Two made their reservations via the online service Travelocity and paid for them, apparently online, with Visa. (Ticketing information on the hijackers of United Flight 93 from Newark is not known.)

CAPS would examine these facts and many others. The specific data collected are among the most closely held secrets in aviation security, but it is understood that they would probably include passengers’ addresses and flying histories, whether they were traveling with a family or alone and possibly their national origin. Each piece of information is given a numerical score to be tallied.

Over the years, this procedure has sounded civil liberties alarms. “Profiling is a naughty, naughty word,” Berkebile says.

A 1997 report of the White House Commission on Aviation Safety and Security, headed by then Vice President Al Gore, submitted the profiling standards to a panel of experts in privacy and foreign relations and representatives of such groups as the American Civil Liberties Union and the Council on American-Islamic Relations.

The panel asked that the profiling standards not be based on anything of “a constitutionally suspect nature,” such as race, religion or national origin of U.S. citizens. The commission further recommended that the airlines and government be forbidden from maintaining permanent databases on so-called selectees--people selected by the CAPS system for closer scrutiny--and that the entire system should only remain in place until “explosive detection systems are reliable and fully deployed”--another sign of the naive, prevailing assumption before Sept. 11 that explosives, not people, represented the real threat.

The extent to which the recommendations were adopted by the FAA is unknown.

The profiling process assigns a score to each passenger, similar to the scoring system lenders use to judge mortgage applicants’ credit-worthiness. In this case, passengers who score below a certain threshold are judged low-risk and cleared to board without further scrutiny.

Everyone else is termed a “selectee” and is supposed to be examined more closely. Selectees may be patted down or questioned but not necessarily kept off the aircraft if they are not found to be in violation of safety rules. Their checked bags should be subjected to CAT scans and other systems designed to find weapons or explosive material. They should be subject to “positive bag matching,” in which their bag is not loaded on a plane until they are confirmed to have boarded.

The idea is not so much to identify risky passengers as to weed out the low-risk ones who can be passed quickly through security checkpoints, allowing screeners to concentrate on everybody else.

“CAPS doesn’t look for bad people,” says Luckey, who helped develop the system at Northwest. It will flag some travelers, for example, simply because too little is known about them.

“Let’s say you’re a new kid just out of high school flying to your first military posting,” he says. “Uncle Charlie buys you a ticket at the last minute. CAPS would tag you as an unknown, and you’d get looked at.”

In this sense, CAPS is different from the profiling system used by El Al, which has developed psychological profiles of passengers who could pose a threat and interrogation methods to identify them at the airport. El Al’s system divides selectees into five categories, ranging from “naive”--those who may be unaware that a friend or acquaintance has planted an explosive in his or her luggage--to suicide terrorists.

Experts in the U.S. emphasize that a high ratio of CAPS-selected passengers on a plane does not indicate that the flight is unsafe. Nor does a low rate mean the opposite: Whether by luck or design, Luckey says, the hijackers apparently evaded selection.

“There are technical reasons why the old CAPS [the system in place prior to Sept. 11] wouldn’t have selected these guys.” Luckey says. He noted that many of the suspected hijackers had been in the country for some time “and had a lot of commonalities not germane to tripping CAPS.”

Another flaw is that CAPS was never well integrated with the rest of the security system. Passengers without checked baggage, for example, would already have passed through a carry-on checkpoint before gate agents would be alerted by their computer screens that the passengers were selectees.

In practice, the gate agents often disregarded the alert, assuming that the passengers had already been screened, according to airline sources. “I’d guess that before Tuesday, the gate agent would not have done anything further,” says Reynold Hoover, a Jacksonville, Fla., aviation security professional.

Many professionals expect CAPS to play a much greater role in airport security, especially if it can be well integrated with other elements of the system.

Asked what he would expect if CAPS noted that several “selectees” boarding a flight had purchased one-way tickets or paid cash, one security official at a major airport responded: “I’d expect a team of sky marshals to be deployed on that flight.”

Apart from the CAPS program, the only airline passengers whose names are routinely screened against law enforcement lists are foreigners entering the country. Their names are run through an Immigration and Naturalization Service database of individuals who law enforcement agencies have determined should be questioned or detained because they are criminals or a security risk.

There is no such systematic screening for individuals traveling within the United States. When law enforcement and intelligence agencies deem someone a potential threat to airlines, they may provide a name and other information to the FAA or the airlines.

Former FAA security official Flynn said that, in his experience, the FAA “would appropriately and promptly inform the air carriers and sometimes direct them to take appropriate actions, depending on the information.”

In the case of two of the suspected hijackers aboard American Flight 77, the FBI had given their names in the weeks prior to the attack to the INS to include on its watch list because they were linked to terrorist groups, only to learn that the men were already in the country. The FBI did not provide their names to the airlines, according to people with direct knowledge of the matter.

The Insecure Cockpit

Clear weather, early morning flying times: The elements were in place for smooth departures Sept. 11 for the four targeted airliners. All took off with the cockpit doors closed and locked--standard operating procedure on FAA-regulated flights. Within an hour of takeoff, investigators believe, each flight had been commandeered.

How the hijackers entered the cockpits may never be precisely known. As The Times previously reported, at some point early in United Flight 93, which left Newark 42 minutes after its scheduled 8 a.m. departure, a ground controller heard someone in the cockpit exclaim: “Hey, get out of here.”

That suggests an intruder took the pilots by surprise, which may explain why pilots in at least two planes had no time to broadcast an emergency “7500” code indicating a hijacking in progress. On American Flight 11, four of the five suspected hijackers booked seats along the Boeing 767’s left aisle, which leads directly to the cockpit door. One suspect’s seat was 2B, which provides one of the best vantages for observing and quickly reaching the cockpit.

Security experts say that among the most shocking elements of the hijackings was the apparent ease with which the hijackers gained control of the cockpits. History says the experts should not be surprised.

“Cockpit incursions,” to use the industry term, were hardly unknown in civil aviation even before this month’s hijackings. At least five times in the last two years, passengers bent on mayhem were able to burst into an aircraft’s cockpit on a commercial flight. In several cases, the intruder’s goal was to attack a crew member, and in some cases, control of the plane was compromised.

Perhaps the best-known attempt occurred in August 2000, when a passenger trying to break into the cockpit of a Southwest Airlines flight from Las Vegas to Salt Lake City was fatally beaten by others on board. None of the passengers was charged in the death.

In March 2000, a man burst into the cockpit on a flight from Tenerife, Spain, to Berlin and tried to crash the plane before the crew wrestled the controls away from him.

That same month, a passenger on an Alaska Airlines flight from Puerto Vallarta to San Francisco broke through the cockpit door and attacked the pilots before passengers pounced on the attacker.

And in 1999, a Japanese passenger pulled an 8-inch knife on a flight attendant on the upper deck of an All Nippon Airways Boeing 747 jumbo jet before storming the cockpit. The hijacker locked the co-pilot out of the cockpit and ordered the pilot to steer toward the U.S. military’s Yokota Air Base in western Tokyo. When the pilot refused, the hijacker stabbed him in the neck and seized the controls. The incident ended when the co-pilot and another All Nippon pilot traveling in the cabin kicked down the cockpit door and subdued the hijacker.

In December, a British Airways passenger invaded the cockpit of a jumbo jet traveling from London to Nairobi, causing the plane to plunge more than 1,000 feet before the crew subdued the intruder and regained control. At the time, British Airways allowed pilots to leave the cockpit door open during the flight, so that curious passengers could peek inside. After the incident, the airline ordered cockpit doors locked at all times.

In the U.S., the cockpit doors are required to be locked only during takeoff and landing; at all other times, they may be opened at the crew’s discretion.

Even locked doors, however, represent scant protection against determined intruders. FAA officials and aircraft manufacturers say agency specifications require cockpit doors to be sturdy but “frangible”--that is, breakable--so the pilots can escape into the cabin in an emergency or passengers can use a cockpit window to flee the plane.

The cockpit doors are designed to be kicked open from either side. Although they typically have locks, these are largely for show and can easily be broken, according to an official at Boeing Co.

That might not even have been necessary on the four fated flights. On most U.S. airliners, flight attendants, typically the purser, carry a cockpit door key, often on a chain around their neck. (Some airlines allow it to be stowed in different places in the airplanes.)

The point is to allow the cabin crew to check on the cockpit during long flights to ensure that the pilots have not become incapacitated and to bring them meals. But the easy availability of the key might explain how the cockpit of United Flight 93 was taken by surprise.

One of the few exceptions to these practices among the world’s airlines is El Al, whose fleet is equipped with reinforced cockpit doors and whose crews are under orders to keep the cockpit secure at all times. At least some El Al planes are equipped with double doors resembling an airlock, so that even if a crew member needs to leave or enter the cockpit, it is never open to the cabin.

Until Sept. 11, it was not unusual for pilots on commercial U.S. flights to emerge from the cockpit to respond to a cabin emergency or chat with the passengers.

When a hijacking did occur, said the GAO’s Dillingham, “instruction to pilots and to flight attendants has tended to be: ‘Be passive and cooperative.’ The thought was that the hijackers just wanted to go somewhere.”

In threatening situations now, “the policy is to assume a fortification and capsulation policy to protect the cockpit,” says Luckey, who said the policy has been communicated by the pilots association to its 60,000 members at 47 domestic airlines. If a disturbance occurs in the cabin, he said, “we do not come out for any reason but try to land the plane as quickly as possible.”

But one commercial pilot, with years of experience flying domestic and international routes, says the major U.S. airlines have offered few specific instructions regarding how best to counter a rush on the cockpit. This has persisted, he said, despite the events of Sept. 11.

“We call it yo-yo: You’re on your own,” said the pilot, who spoke on a condition of anonymity.

New World

Experts have known for years that U.S. aviation security needed serious upgrading, owing to a dysfunctional system and complacency at almost every level.

In the wreckage of the World Trade Center and the Pentagon lies several old assumptions that have hobbled previous attempts to improve security. Perhaps foremost among them are the notions that the risk of an airborne attack was too low to justify the added expense of more security and that passengers would not abide the additional inconvenience.

Schiavo relates how, during her tenure at the Department of Transportation, the FAA resisted her recommendations for multibillion-dollar fixes to airport security systems.

“They said that Pan Am 103 [which exploded over Lockerbie, Scotland, in 1988] cost $3 billion, and that if we had another Pan Am in 10 years, that would be cheaper than making all the changes I proposed. They said it just wasn’t worth it because we had never had a major domestic terrorist incident, and no one would put up with security measures or pay for them.”

A National Research Council panel on aviation security concluded in 1996 that passengers were likely to rebel against equipment that projected graphic images of their bodies to inspectors, as opposed to equipment that simply sounded an alarm at a suspicious object. Any screening system for carry-on bags that took more than 6 seconds per bag--the standard for X-ray machines found at most airports--would be regarded as an imposition by travelers and a logistical bottleneck by airlines, the panel reported.

“The degree to which people will accept the inconvenience, discomfort, delays, embarrassment . . . health risks and invasions of their privacy” would depend on many factors, first and foremost “the nature, extent and likelihood of the actual threat,” the panel reported.

Measuring the “actual threat” in hindsight meant waiting for a catastrophe.

Meanwhile, the entire system is fragmented into the jurisdictions of multiple agencies and private contractors, many working at cross purposes. Turf battles rage incessantly among these entities.

“You can’t think of security as just a screening device,” says Thomas Hartwick, a physicist and chairman of the National Research Council aviation panel. “It’s a system--in fact, a system of systems. You have to optimize the way the whole thing works. Suppose you had CAPS selecting someone as a suspicious character; wouldn’t you want to flow that information forward to other points in the system, like the screeners?”

Indeed, all security for Israeli airports and its national carrier, El Al, is the purview of Shin Bet, the government’s internal security agency.

It should be noted that the U.S. is not unique in delegating essential levels of aviation security to nongovernmental entities. About 50 countries, including Britain, allow airlines or private companies to manage airport security, according to Denis Chagnon of the International Civil Aviation Organization.

Still, screeners and inspectors in many other industrialized nations are held to higher standards and paid better than their American counterparts. Some require inspectors to be citizens or meet a minimum residency threshold. In Belgium, screeners earn up to $15 an hour, more than twice the average U.S. wage for screeners.

In Japan, most screeners hold college degrees at Narita New Tokyo International Airport, according to an airport authority spokesman. They go through 150 hours of classroom training and six months of supervision. By contrast, the average seniority of screeners at the worst U.S. airports is less than three months, according to a GAO assessment issued in January.

“We wanted a security system on the cheap,” says Jim Hall, former chairman of the National Transportation Safety Board. “No one wanted to pay for it, not the government, not the airlines.”

American aviation security experts have been heartened by proposals to place checkpoint inspectors under federal jurisdiction and to improve such long-neglected efforts as the Federal Air Marshal program, which in the 1970s deployed undercover agents on all “high-risk” international flights and many domestic routes. That would strengthen security not merely because the government might ensure that training standards and work rules are better enforced--which could be a faulty assumption--but because lines of authority and accountability will be much clearer.

“Whatever the threat is,” Hartwick says, “you want to keep it off the airplane. The airline’s job is to fly the airplane, so let some other authority take care of everything else.”

Creating a coordinated system, he argues, might be the single most important enhancement to security.

“Individual pieces are working well in many cases,” Hartwick says, “but the totality is not in place yet.”

*

Times staff writer Mark Magnier and Times researchers Rie Sasaki in the Tokyo bureau and Janet Lundblad in Los Angeles contributed to this report.