Page 1 of 2

Security Meltdown

Who is watching the people who are watching our nukes?

"Gentlemen, we have a problem," said Air Force Brig. Gen. B.C. Harrison upon convening a highly classified meeting at the Pentagon. The year was 1959, and Harrison was surely guilty of understatement. The Pentagon had just learned that an American sergeant stationed at a Royal Air Force base in Sculthorpe, England, had held the base hostage by placing a .45-caliber pistol to the warhead of a nuclear bomb. After six hours, the officer had surrendered. It was soon learned that he was being treated for serious depression and that his psychiatrist had not known that the man worked directly with nuclear weapons.

Advertise on MotherJones.com

If you weren't willing to push the button, you wouldn't stay in the program -- Tommy Metcalf, former nuclear submarine technician and convicted killer

Eli Flyer, then a personnel researcher for the U.S. Air Force and a participant in the meeting, recalls that in the aftermath of the incident there was some dispute within the Pentagon about what would have happened if the sergeant had in fact fired his weapon. At a minimum, it was concluded, the bullet's impact would likely have detonated the high explosives contained in the bomb and thus scattered nuclear debris into the atmosphere. But even disclosure of the incident -- which was hushed up by the military and not reported until 1962 -- would have been a public relations disaster. At that very time, Air Force officials had been assuring the public that U.S. military personnel were rigorously screened to ensure that none would intentionally provoke a nuclear disaster. "Disclosure of the event would certainly have knocked us out of England," says Flyer, who provided Mother Jones with the first inside account of the meeting.

Responding to the near disaster at Sculthorpe, in 1962 Flyer helped the Air Force develop a system to screen candidates for nuclear-sensitive jobs. By 1965 a version of this system had been adopted service-wide and called the Personnel Reliability Program (PRP). It is supposed to guarantee that only "competent, stable, and dependable individuals" have access to America's nuclear arsenal.

Since PRP was implemented, no nuclear accidents can be directly traced to mentally unreliable personnel. But Mother Jones -- along with the National Security News Service, which assisted in researching this story -- has learned of numerous cases of decidedly unreliable service members receiving PRP clearance. In the last 11 years, three different men who were approved to work with nuclear weapons were convicted of committing murders that occurred while they were on active duty.

Of course, any program responsible for screening tens of thousands of applicants is likely to make mistakes in at least a handful of cases. But critics charge that PRP's approach is cursory at best. "The screening process looks for self-announced kooks but isn't good at ferreting out less sensational cases," says Herbert L. Abrams, a PRP expert at Stanford University's Center for International Security and Cooperation. "As a result, thousands of unstable people have been certified for PRP-approved posts."

Indeed, in one case, the Navy granted PRP clearance to a man whom it knew to have been a suspect in an unsolved murder, and who was caught in a multitude of lies during his screening. Three years later, in 1989, that man, Tommy Harold Metcalf, a fire control technician on the USS Alaska nuclear submarine, brutally murdered an elderly couple. Metcalf, now serving a life sentence at Walla Walla prison in Washington state, claims that at the time of his arrest he had the knowledge to override the controls on the Alaska and launch a nuclear attack.

Weapons designers have gone to great lengths since the Sculthorpe affair to ensure that no individual is able to single-handedly set off a nuclear weapon. However, several experts consulted by Mother Jones say that the possibility of such a disaster remains.

One concern is that some nuclear warheads, such as the high-yield W88 on the Navy's Trident D-5 missile, continue to use conventional high explosives to trigger a nuclear reaction, as opposed to less detonable "insensitive high explosives." According to a 1994 study co-authored by John Harvey, now a deputy assistant secretary of defense, an accident (such as dropping a missile during handling) or act of sabotage with such a warhead could "create impact pressures and temperatures sufficient to cause motor detonation or fire." The resulting "catastrophic" dispersal of toxic plutonium would cause an environmental disaster and increased cancer deaths for the people exposed to the fallout. "PRP remains a crucial part of our effort to ensure that accidental or unauthorized use of our nuclear weapons does not occur," says Scott Sagan, a colleague of Abrams' at the Center for International Security and Cooperation.

PRP is a two-step process that includes an initial screening and postapproval monitoring. Military investigators look for a variety of traits, including "good social adjustment," "emotional stability," and a "positive attitude toward nuclear weapons duty." If problems emerge on the job, individuals can be temporarily or permanently barred from duties that require PRP clearance. (About 7,000 people were decertified between 1990 and 1996.) The number of armed forces personnel with PRP certification has dropped from an average of about 100,000 during the 1980s, when the Cold War was still raging, to just 19,042 by 1996.

During the initial screening, PRP candidates undergo a medical evaluation and are interviewed by certifying officials. The candidate's personnel file is reviewed, and military investigators conduct a background check to examine professional, educational, and personal histories. As part of this process, investigators may interview family and friends, and former employers and colleagues.

But critics of the screening say it includes no routine psychological testing and that a candidate's entire medical evaluation may be limited to an examination of old medical records. James Bush, a retired submarine commander, says that the screening's primary purpose was to make sure that the candidates were not criminals. "My interview was oriented toward making sure they'd fire the weapon if I told them to," he says.

Although a criminal past is grounds for rejection, candidates can lie about their records with little chance of being caught. Steve Sellman, director of recruiting policy for the Pentagon, estimates that FBI background checks pick up only about "5 to 8 percent of people who've had trouble with the law, beyond kids who self-divulge that information." Sellman admits that such figures leave him "concerned" that bad apples are receiving high-level security clearances, including PRP certification.

It is hard to know exactly how bad the problem is because the Pentagon, citing privacy concerns, refuses to provide details on specific cases in which PRP-certified personnel have committed serious crimes. When asked if anyone in PRP has been found guilty of murder during the past few years, Col. Dale Landis, a staff officer in the office of the assistant secretary of defense, who helped monitor PRP, would say only that he is unaware of any such incidents. (Since speaking with Mother Jones, Landis has retired.)

However, information compiled by Mother Jones gives credence to Sellman's worst fears. The Pentagon's annual status report on PRP for 1996 shows that 758 people were kicked out of the program that year. Of those, 169 were expelled due to "conviction by a military or a civilian court of a serious offense" or "a pattern of behavior indicative of a contemptuous attitude toward the law or other duly constituted authority."

Page 1 of 2