SEARCH: Search Options
The Issues
Federal Page
Post Series
Columns - Cartoons
Live Online
Photo Galleries
Other News:
Editorial Page
Real Estate
Home & Garden
News Digest
Print Edition


Click here!
Can Polygraphs Detect Spies?
Panel Says No, and Worries About Blemishing the Innocent

_____TechNews.com Archive_____
* Lie-Detecting Devices: Truth or Consequences? (The Washington Post, Aug 18, 2002)
* Today's Political News
* Elections 2002 Coverage
* Daily E-mail Updates
_____Full Coverage_____
* Espionage
_____From TechNews.com_____
* New E-mail:Get weekly tech policy and security updates
* Tech Policy Section
E-Mail This Article
Printer-Friendly Version
Subscribe to The Post
By Shankar Vedantam
Washington Post Staff Writer
Wednesday, October 9, 2002; Page A29

Polygraph tests are ineffective in catching spies and have probably tarred thousands of innocent government employees and applicants with unwarranted suspicion, a top scientific panel has concluded.

While lie detectors may have some utility in criminal investigations, where subjects can be tested on specific questions about a crime, they tend to be unreliable in countering espionage, where large numbers of people are asked general questions about whether they have done anything wrong, the scientists said in a report.

"Too many loyal employees may be falsely judged as deceptive, or too many major security threats could go undetected," the scientists said, warning against reliance on the tests.

The study was commissioned by the Department of Energy in the wake of controversy over the case of scientist Wen Ho Lee, who was wrongly accused of passing U.S. secrets to China. Lee was subjected to polygraph testing, and controversy has swirled around how the tests were administered.

Polygraph tests have long been controversial; they are largely rejected by the courts and most countries, the scientists said. Still, investigators use them routinely, in part because of the belief that simply administering the tests make subjects more compliant and amenable to making confessions.

"We stress, though, that no spy has ever been caught using the polygraph," said Kathryn Laskey, a researcher at George Mason University in Fairfax and a member of the study team assembled by the National Research Council, which is part of the National Academy of Sciences.

For the study, the 17-member panel conducted the most exhaustive review to date of published studies on polygraphs and current government polygraph procedures.

The CIA and FBI now administer polygraph tests to all prospective employees, and both agencies and the national energy labs administer periodic tests to employees with access to secret material.

Spokesmen at the FBI, the CIA and the Department of Energy said they would evaluate the report. Linton F. Brooks, acting director of the National Nuclear Security Administration (NNSA), which oversees the nation's nuclear stockpile and government labs, said the tests were not used on a stand-alone basis, but as part of a larger investigative fabric.

Sens. Jeff Bingaman (D-N.M.), chairman of the Energy and Natural Resources Committee, and Pete V. Domenici (R-N.M.) yesterday called on the department to abolish the tests. The senators sponsored legislation requiring Energy Secretary Spencer Abraham and the administrator of the NNSA to implement a new polygraph rule based on the new report.

Edward J. Curran, a former director of counterintelligence for the Energy Department, disputed the findings. "To have them say it's ineffective is irresponsible," Curran said. "You can ask the prisoners in the jail what they think of the jail and you will get the same answer. You are asking scientists who don't want to take the tests" to evaluate them.

The scientific review came on the heels of two celebrated failures involving polygraphs: Aldrich H. Ames, who was a spy for Moscow at the CIA, "passed" his polygraph test after he talked polygraph investigators into believing he was innocent. Wen Ho Lee, who worked at Los Alamos National Laboratory in New Mexico, was wrongly told that he had "failed" a polygraph test, when in fact the machine had indicated he was telling the truth about being innocent.

Polygraph tests, which have become popularized as "lie-detector tests," involve asking subjects a series of questions while hooked up to a machine that measures physiological changes such as blood pressure and heart rate. No machine can detect deception itself, but the polygraph is said to pick up signs of anxiety and tension that may be associated with lying, signs that test-takers themselves cannot control.

Law enforcement officials like Curran insist that polygraphs are a deterrent, not just for keeping scientists from becoming spies, but in getting people to closely follow rules. Polygraphs are also widely used as an interrogation tool; in Lee's case, for instance, FBI agents told Lee the machine said he was lying, even though it had not, in the hope it would prompt him to confess to passing secrets to China. Lee spent months in solitary confinement before the case against him unraveled.

"The polygraph is not worthless, it is worse than worthless," said Alan Zelicoff, a physicist, physician and senior scientist at Sandia National Laboratories in Albuquerque. Zelicoff said his colleagues at the lab have described the tests as "mental rape," and an excuse for investigators to conduct racial profiling and "inquisitions."

Yesterday, the chairman of the National Research Council panel, Stephen Fienberg, a professor of statistics and computer science at Carnegie Mellon University, said that the scientists would not tell the government to halt the tests altogether because no superior alternative was available.

But he warned that subjects and administrators of the tests had been lulled into a false belief that they were effective, raising the real possibility that spies who clear the tests could seriously damage national security.

Citing movies where lie detectors are shown as effective tools to read other people's minds, Fienberg said, "the power of an image of a needle going up and down really fast evokes a response of overconfidence in this as a tool."

2002 The Washington Post Company

Related Links

More National News

Full Coverage