Pittsburg Post-Gazette science editor Byron Spice reports. Excerpt:
Who knows what evil lurks in the hearts of men?
Not polygraph examiners, at least not those charged with finding spies and other security risks within the ranks of federal employees, a new National Research Council report concludes.
Lie detectors simply aren’t accurate enough to ferret out of what is presumably a handful of spies amid the tens of thousands of federal employees who undergo polygraph examinations each year, said Stephen Fienberg, a Carnegie Mellon University statistician who headed the study.
“The polygraph itself has not caught a single spy,” Fienberg said yesterday following the release of his panel’s two-year study on the use of polygraphs for screening employees. Lie detector results, he explained, “are better than chance, but well below perfection.”
That can make lie detectors an effective interrogation tool for police trying to choose from among a handful of suspects, but even then, a large part of their effectiveness owes to their mystique, Fienberg said. The widespread belief that they work — or might work — can lead to confessions.
But when used as a screening tool, polygraphs can be expected to miss many, if not most, spies, while misidentifying untold numbers of loyal citizens as suspicious.
Even if the test were designed to catch eight of every 10 spies, it would produce false results for large numbers of people. For every 10,000 employees screened, Fienberg said, eight real spies would be singled out, but 1,598 innocent people would be singled out with them, with no hint of who’s a spy and who isn’t.
Yet use of lie detectors as screening tools has expanded, particularly at the Department of Energy’s nuclear weapons laboratories following the Wen Ho Lee “spy” case at Los Alamos National Laboratory.
“But it’s not just the DOE,” Fienberg said. “It’s the CIA, the FBI, the NSA. It’s every three-letter agency. Even the Postal Service is using polygraphs.” The Department of Defense alone administers 10,000 polygraph examinations every year.
“The polygraph is one of many tools we use to protect some of the nation’s most sensitive secrets,” said Linton Brooks, acting administrator of DOE’s National Nuclear Security Administration. “It is used not on a stand-alone basis, but as part of a fabric of investigative and analytical reviews to help security personnel determine who should have access to classified information.”
Still, law enforcement and security personnel seem to trust the polygraph more than other tools, Fienberg said. The study panel concluded that this is a misplaced trust that leads to overconfidence. “And this overconfidence can lead to a false sense of security,” he added.
Al Zelicoff, a bioweapons expert at Sandia National Laboratories in Albuquerque, N.M., and an outspoken critic of DOE’s polygraph use, agreed that the cloud of suspicion almost always lifts once an individual has passed a lie detector test. “That’s a very bad assumption,” he added, given that no one claims that polygraphs are perfect.
But the new report makes clear that the polygraph “is worse than worthless when used in a screening mode,” said Zelicoff, a Pittsburgh native and graduate of the University of Pittsburgh School of Medicine. Lots of honest people are unfairly placed under suspicion and precious resources are diverted from other security measures, he maintained.