Home Page > Polygraph News

17 October 2003 Teller on Polygraph
Steven Aftergood of the Federation of American Scientists' Project on Government Secrecy comments in his Secrecy News e-mail publication:


In the perennial dispute over the legitimacy of polygraph testing as a tool for security screening of government employees, the late physicist Edward Teller sided with his scientific colleagues at the national laboratories in opposition to the polygraph.

"Together with many others, I believe that its negative effects ... by far outweigh the conceivable advantages [and] the rather dubious evidence that the tests may give," Teller wrote to the Secretary of Energy.

Teller's October 27, 1999 letter on polygraph was obtained by researcher/reporter Michael Ravnitzky under the Freedom of Information Act. See:

Dr. Edward Teller (1908-2003), a past director of Lawrence Livermore National Laboratory, led the development of the hydrogen bomb.

5 October 2003 Lie Detector "Testing" for Guantanamo Bay Naval Station Employees?
In an article titled, "Investigators open Guantanamo probe," the Associated Press reports that the commander of the Camp Delta detention center at the Guantanamo Bay Naval Station is considering polygraph interrogations for personnel working at the facility. Excerpt:

GUANTANAMO BAY NAVAL STATION, Cuba - Two dozen investigators began searching for possible security breaches yesterday at the U.S. prison camp, where espionage charges have heightened tensions among soldiers.

Investigators from the Miami-based U.S. Southern Command reported here Wednesday, a day after the arrival of five non-American-born Arabic interpreters contracted by the firm that employed an American translator accused of spying.

Sources familiar with the investigation said two more arrests could be imminent.

Investigators will try to establish how a translator already under investigation got secret clearance and was allowed onto the base, and how a second translator managed to leave with classified information. In addition, a Muslim chaplain is under investigation after allegedly leaving with diagrams of the prison layout.

The translators, from San Diego-based Titan Corp., arrived as officials boosted security by closely monitoring e-mail messages, asking troops to report suspicious behavior, and postponing the assignment of another Muslim chaplain.

Army Maj. Gen. Geoffrey D. Miller, who commands the detention mission, said he is increasing baggage checks and considering lie detector tests.

5 October 2003 Canada: "Cops praise new lie-detector"
Ajay Bhardwaj reports for the Edmonton Sun.

Cops and the City of Camrose are lauding a new lie-detector-type tool even though critics say it gives police the power to intimidate people accused of committing crimes.

The computer voice-stress analyser (CVSA), which purports to pick up tremors in the voice when a person is lying, is already in use by the Camrose Police Service. Cops say it's another investigational tool.

"If you didn't do anything wrong, there's nothing to worry about," said Camrose Mayor Norman Mayer. "I suppose it would be intimidating if you did something wrong because you know that something's going to show up. But maybe you should be intimidated if you're causing problems."

Staff Sgt. Peter Ratcliff of the Edmonton Police Association agreed.

"The results can lead you to other avenues for investigation," said Ratcliff. "There's probably been a lot of people who've confessed on a polygraph because they feel pressure they put upon themselves. I don't see this thing going any farther than the polygraph right now. I think it's a good idea. "

It's another one of those things that we have to use to get evidence to solve crimes. It's intimidating if you've done something."

Smaller police departments are using the $10,000 US tool, which many say is more accurate than a polygraph.

But Sanjeev Anand, a professor of criminal law at the University of Alberta, said any evidence gained from a CVSA test wouldn't be admissible in court.

"It's one possible tool but if it's not probitive enough to be admissible in court you wonder how much they should be relying on it, even in their investigations," said Anand.

Ratcliff said anyone who is convicted using a CVSA is likely to challenge it in court.

Camrose Mayor Norman Mayer is either woefully misinformed, a fool, or both if he truly believes that "if you didn't do anything wrong, there's nothing to worry about." Voice stress analysis, like polygraphy, has no scientific basis whatsoever. See the CVSA and Other Voice Stress Analysis Applications forum of the message board for further discussion of this pseudoscience.

2 October 2003 "Former Trooper Can File Suit Against Former Polygraph Supervisor" reports:

BRIDGEPORT, Conn. -- A retired state trooper's lawsuit against a former supervisor in the state police polygraph unit is being allowed to proceed. U.S. District Judge Stefan Underhill refused this week to dismiss retired Trooper Adrienne Lamorte's lawsuit that claims she was subject to retaliation after raising some questions about the accuracy of some lie detector tests.

Lamorte alleged that a supervisor in the state police polygraph unit was issuing faulty results.

Lamorte filed the federal lawsuit against John Leonard in 2001, claiming Leonard disciplined her after she lodged a complaint about Sgt. Randolph Howell's conclusions on polygraph results.

Lamorte claimed Howell, a sergeant and commanding officer of the polygraph unit, was "conducting polygraph examinations in an incompetent manner." the lawsuit states.

The lawsuit alleges Howell would report an examinee was truthful when the test scores indicated the opposite.

Lamorte was concerned about the consequences of false results so she relayed her concerns about Howell to Leonard, the lawsuit states.

Lamorte claims she was threatened in September 2000 with a disciplinary transfer unless she agreed to an administrative transfer out of the polygraph unit, according to the lawsuit.

She chose the administrative transfer and later retired.

1 October 2003 Canada: "Lie detector 'useful interrogation prop'" reports:

VANCOUVER (CP) - Lie detector technology less costly than controversial polygraphs is catching on among smaller Canadian police forces, scaring some doctors who say it's even less reliable.

The computerized voice stress analyser (CVSA), which purports to pick up tremors in the voice when a person is lying, "isn't more than 50 per cent accurate," said University of Toronto psychiatrist John Furedy.

"It is however, a very useful interrogation prop."

The device was in the news this week for helping to cause another scandal within the Vancouver Police Department.

A Vancouver police officer who took a CVSA test in a job interview with another force admitted to lying under oath in a criminal trial, keeping goods that should have been turned in and withholding information about possible brutality by other officers.

To use this high-tech method of lie detection, the person being interviewed speaks into a microphone and software, run on a laptop, measures voice modulations.

The problem, Furedy said, is the operators of these machines don't really know what they're measuring.

"Voice can vary for a number of reasons and we don't know whether it's relative to certain stimuli or not," he said.

The CVSA can be used on taped interviews and claims to give officers fresh leads on cold cases.

The $10,000 US device is used by more than 1,000 law enforcement and government agencies south of the border, and recently Canadian forces have been buying it, said a spokesman for its manufacturer, the National Institute for Truth Verification.

In the past two years departments in Lethbridge and Camrose, Alta., Edmunston and Rothsay, N.B., Vancouver and Saanich, B.C., in the Victoria area, have purchased the system from the West Palm Beach, Fla., company.

"Canada just recently started to really get involved," said executive director David Hughes.

Hughes said the company only sells the CVSA to law enforcement and government agencies and has buyers sign a contract prohibiting them from passing the equipment on to anyone else.

There are fears that it could fall into the hands of organized criminals and be used on suspected undercover officers.

People taking the test don't have to be hooked up to the machine, as is the case with a polygraph, and therefore wouldn't know it was being performed on them unless told.

Marshall Chalmers, chief of the Camrose police force, said his investigators always inform people being tested because they have the right to refuse it.

"We say, OK, if you're really telling the truth, prove it to us," he said.

Chalmers said that in the first six months after purchasing the CVSA last August, it was used to solve 42 criminal cases.

"That's a dramatic increase. It's an integral part of our operation now."

He said people have been brought in on charges and in the course of their CVSA test, other crimes they had committed came to light.

Before buying the device, Camrose police had to share a $20,000 polygraph machine with a number of municipal organizations. Only suspects of major crimes were eligible and there was a three-week wait list.

"That gave the suspects plenty of time to change their minds about taking the test," Chalmers said.

The chief read about the CVSA in a police magazine and ordered one of his officers to research its effectiveness.

"This is leading edge technology and it is advancing quickly across the U.S.," Chalmers said.

"We talked to numerous departments that are using it and were convinced of its accuracy. The CVSA is 100 per cent accurate. With the polygraph, there's a grey area."

Furedy said "it's pretty scary" that police have that kind of faith in the machine.

"It's a sign of a fundamentally superstitious society, certainly here in North America. The reality is that you can never determine with 100 per cent accuracy whether someone is telling the truth."

Lie-detection devices are a scare tactic he said, that can sometimes push people into confessing to crimes they didn't commit.

He has testified at many criminal trials and convinced juries to throw out such admissions.

One was the trial of a 74-year old crossing guard who was accused of molesting a girl. He was under accusation for a month and his community was suspicious of him, Furedy said.

He failed the test and in a post-exam interview that lasted several hours Furedy said the man got so angry he confessed in order to put an end to the questions.

"He just wanted to get out of there and figured he would talk to his lawyer and sort it out."

The confession was eventually thrown out and the man was found not guilty.

Furedy said all lie-detection tests are subjective and results vary depending on the way the questions are worded and the relationship the interviewer has with the interviewee.

Saanich Police Insp. John Charlton, who heads the department's detective division, said he only uses the CVSA to point him in the right direction.

"Every result we've gotten through the CVSA we then go out and corroborate with evidence and we've found it to be very effective. If there were any credibility issues with the device, we wouldn't use it."

1 October 2003 "White House won't rule out polygraphs"
CNN reports. Excerpt:

WASHINGTON (CNN) -- The White House pledged full cooperation Wednesday with a Justice Department probe into the leak of a CIA operative's name and has followed proper procedures so far, a spokesman told reporters Wednesday.

Press secretary Scott McClellan strongly intimated that President Bush would expect White House aides to take polygraph examinations if the Justice Department asked for them.

"Full cooperation is full cooperation," McClellan said.

All White House employees have been instructed to preserve documents dating to February 2002 that could be relevant to the investigation.

1 October 2003 "Is . . . John Ashcroft willing to ask Karl Rove to submit to a polygraph?"
As reported by Wayne Washington of the Boston Globe in an article titled, "US Begins criminal probe of CIA leak," Senator Dick Durbin raised the question on the floor of the U.S. Senate in the context of the Justice Department investigation into who leaked Valerie Plame's identity as a CIA employee:

Speaking on the Senate floor, Senator Richard J. Durbin, Democrat of Illinois, described how FBI agents, at the behest of the Bush administration, asked that he submit to a lie detector test to help them uncover who leaked classified information that had been provided to the Senate Intelligence Committee, of which Durbin is a member.

He argued that the Justice Department won't be as aggressive in pursuing White House officials. "Is . . . John Ashcroft willing to ask Karl Rove to submit to a polygraph?" Durbin asked.

1 October 2003 "Summit elections worker fails polygraphs"
Akron, Ohio Beacon Journal staff writer Julie Wallace reports. Excerpt:

A Summit County Board of Elections employee quizzed as part of a probe into a candidate's missing petitions failed two polygraph tests after admitting using drugs and alcohol before the tests, law enforcement officials said.

Several sources familiar with the investigation identified the employee -- who was not named in a news release issued Tuesday by the Summit County Sheriff's Office -- as Deputy Director John Schmidt, a Democrat who holds the No. 2 position in the board's office.

Schmidt declined to comment. His lawyer, Carmen Roberto, said Schmidt did not fail the polygraphs; he said the results were inconclusive on both tests.

Sheriff Drew Alexander, a Republican, said 11 full-time employees were asked to submit to lie-detector tests in the investigation into the June disappearance of election petitions belonging to Akron City Councilman Joe Finley, D-2. Finley is a maverick often at odds with his party's local leaders.

Eight of those 11 full-time employees easily passed the polygraph.

Three others were tested twice. Two of them had results that initially were labeled inconclusive but later were determined by polygraph examiner Bill Evans to be truthful.

The third employee also underwent two voluntary tests -- showing up and acknowledging to Evans that he had used marijuana and alcohol prior to the appointments, Alexander said.

Alexander, who declined to confirm that the employee is Schmidt, said no charge would be filed against the employee over his admitted drug use because the tests were voluntary.

``Both times, he failed miserably at over $500 a pop,'' Alexander said. ``That's enough polygraph examinations. I have no confidence that he'd come in and take a third or fourth and not try to beat it.''

Although it is not mentioned in this article, the Summit County Sheriff's Office's news release accuses the employee who "failed" the test of having employed countermeasures:


30 September 2003 DoD Agency Trashes Polygraph
Steven Aftergood of the Federation of American Scientists' Project on Government Secrecy comments in his Secrecy News e-mail publication:


The Defense Advanced Research Projects Agency (DARPA) has been battered and bruised by controversies surrounding several of its more questionable programs, including the defunct Terrorism Information Awareness.

But the Agency has not received the credit to which it is arguably entitled for conducting those programs in an unclassified form, in which they can be freely debated, criticized and attacked.

Now DARPA has published a complete descriptive summary of all of its (unclassified) programs, where they can be reviewed in some context. It is an intriguing collection, with numerous items of interest.

Describing its "Deception Detection" initiative to develop new "lie detector" methods, for example, DARPA renders an unusually harsh official judgment concerning the polygraph:

"Current screening techniques are flawed, enabling many deceivers to avoid detection and falsely accusing large numbers of innocent people. An effective method to assess intent will decrease both the missed detections and the false alarms." (Page 49).

See "Fact File: A Compendium of DARPA Programs," August 2003 (thanks to GP):

15 September 2003 "Truth tests are flawed, so cut 'em out, DOE"
The Albuquerque Tribune comments on polygraph screening in this editorial, cited here in full:

Truth tests are flawed, so cut 'em out, DOE

First, give the Department of Energy some credit. Then yank its chain - until it altogether stops using polygraph tests to try to find spies in its nuclear weapons program.

The oft-troubled department last week announced it will reduce the number of employees at its nuclear weapons laboratories - including Sandia and Los Alamos in New Mexico - who will be required to take polygraph tests.


Its reasoning was as sound as the strong reasoning of protesters that greeted DOE's efforts to increase polygraph testing of employees in recent years. Those efforts followed congressional pressure and reports that said classified nuclear weapons information had been compromised.

For years, scientists - most notably, Dr. Alan Zelicoff, a physician and scientist at Sandia National Laboratories in Albuquerque - were vocal in warning DOE that polygraph technology and protocols are scientifically unreliable, unsound and essentially useless. They contended that relying on the tests could provide a false sense of security and therefore prove a threat to national security.

Many thousands of employees at Sandia and Los Alamos laboratories and at Lawrence Livermore National Laboratory east of San Francisco were subjected to the policy, and several thousands still will be - though polygraph testing has no credibility.

We applaud Deputy Energy Secretary Kyle McSlarrow's testimony to Congress last week that DOE will "substantially lower" the number of workers - until now as many as 20,000 - required to take polygraphs as part of the department's anti-espionage program. The chance of inaccurately implicating innocent workers with falsely positive polygraph results is just too great to ignore, he reported.

Good - but not good enough.

If, as the highly respected National Academy of Sciences has reported, polygraph tests don't work, why use them at all? Why use them, when McSlarrow acknowledged that the academy is right in concluding that the technology is incapable of reliably distinguishing between people who are telling the truth and those who are lying?

Why use them, when the academy report found that polygraph machines determine that truthful people are being deceptive and inaccurately labels them as "security risks"?

Why use them at all, if, as McSlarrow states, such uncertainty actually risks "undermining the very national security goals we hope to attain"?

Why use them, if, as New Mexico senior Sen. Pete Domenici observed last week, "We hold our scientists' work to the highest standard of accuracy and reliability, and then we impose on them something as sloppy and subjective as polygraph tests. That practice is indefensible"?


It is good to know that DOE will no longer cancel security clearances of employees in the nuclear weapons program based only on failed polygraph tests, as McSlarrow said. But why tarnish employees by branding them for having failed such a test when we all know the test isn't to be trusted?

Certainly DOE has as grave a responsibility to protect the nation's nuclear weapons information, technology and materials as it does to ensure our nuclear arsenal is safe, secure and functional.

But, in doing so, DOE should never rely upon so weak, uncertain, unreliable and dysfunctional a technology as polygraphs have proved to be.

12 September 2003 "Bull Moose Music settles lawsluit"
Doug Harlow reports for the Kennebec, Maine Sentinel on a lawsuit brought under the Employee Polygraph Protection Act. Excerpt:

WATERVILLE -- Bull Moose Music and a former employee have come to terms on a combined lawsuit filed after the store allegedly was robbed by a masked man in August 2001.

Vickie Shuman, 20, now of Oakland, then of Fairfield, was working alone at the store on Elm Plaza two years ago when she told police she had been robbed of approximately $5,000 as the store was closing for the night.

Store officials suggested that Shuman had manufactured the story about the robbery and in fact had taken the money herself.

When Shuman refused to take a lie-detector test, she was fired. She then filed a lawsuit against the music store seeking reinstatement and payment of lost wages and benefits.

She also sought damages for emotional distress and unspecified punitive damages, according to the lawsuit filed in Kennebec County Superior Court in December 2001.

Through its lawyer, Kevin J. Beal of Lewiston, Bull Moose defended Shuman's termination and countersued, charging her with fraud and theft of $5,000 cash. The company also sought punitive damages from Shuman.

"There was a settlement and the charge of theft against her was dismissed, pursuant to the agreement. She was fully satisfied with the settlement," Shuman's lawyer Jason M. Jabar of Waterville said Wednesday. "I would say the conclusion is that her name is cleared.

"She feels her name has been cleared. She got a label she didn't think was fair."

In his response to the claim, Beal took Bull Moose's defense a step further, charging in the written document that it was Shuman herself who robbed the store and took the money.

Shuman "planned and conspired to steal some or all of the cash" from cash registers, deposit bags, the store safe or other locations, according to the counterclaim.

The company alleges that Shuman stole the money and gave it to a coconspirator who had agreed to assist her, then falsely reported the robbery to Waterville police.

Jabar said at the time that leveling such a charge without proof or evidence was a dangerous move on the part of Bull Moose and its lawyer.

Shuman had told police she heard a knock on the back door the night of the alleged robbery. When she opened it, she reportedly was confronted by a man standing about 6 feet tall, wearing a ski mask.

The alleged robber said to Shuman: 'Give me the money or I'll shoot you,' although he never displayed a weapon, police said.

Deputy Police Chief Joseph Massey said at the time that police officers were near Elm Plaza when the robbery call came in and responded in less than a minute and found no one matching the description.

The police investigation turned up neither the money nor the culprit in the robbery. Shuman had worked at the store for 2 1/2 years.

On Sept. 27, 2001, according to the lawsuit, Bull Moose officials ordered Shuman to take a lie-detector test as a condition of her continued employment at the store.

The claim argued that Shuman's firing and the order to take a lie-detector test in the first place were violations of United States Code.

11 September 2003 "FBI Agent Helps Clear Men Held After 9-11"
Associated Press writer Kimberly Hefling reports how an innocent suspect failed his FBI polygraph "test," while his lying accuser passed. Excerpt:

EVANSVILLE, Ind. (AP) - An FBI agent who helped arrest eight Egyptian men in the weeks after the Sept. 11 attacks became the driving force behind clearing their names, an arduous process that eventually led to a rare public apology from the FBI.

FBI Agent Thomas Van Wormer said he felt a responsibility to help the men, whose names cropped up on a far-reaching database - paired with the word ``terrorism'' - even after they were cleared of suspicion.

The database had nightmarish consequences for the men: One missed two flights home from an overseas trip when his name showed up on a no-fly list. Others had problems getting public housing and immigration cards.

``I want to commend them for the way they acted,'' Van Wormer said from his office in downtown Evansville. ``It made it so you wanted to help them. Their demeanor was to their benefit.''

Van Wormer began investigating the men after one of their wives called a law-enforcement hotline set up after Sept. 11, 2001. She said one of the men ``would be traveling to Chicago's O'Hare airport, was going to engage in a suicide crash, and that he would be dead on Oct. 12, 2001,'' according to federal documents.

The woman passed a lie-detector test, but her husband - using an interpreter to take the test - showed some signs of deception, although he did not blatantly lie, Van Wormer said.

The woman then alleged other men were also involved. Authorities became more suspicious when they learned one of the men, Tarek Albasti, had taken pilot lessons.

The men were taken into custody on material-witness warrants and spent about a week in a Chicago detention center. Investigators released the group after deciding the woman's statements were not true. Van Wormer said cultural and language barriers were likely to blame for the man's lie-detector test results.

Actually, the inherent unreliability of polygraphy is more likely to blame for the man's lie detector "test" results. In addition, bias introduced because of the man's wife having "passed" was also a likely contributing factor. The case of the Evansville Eight is a good example of how our government's misplaced reliance on the pseudoscience of polygraphy leads to investigative misdirection. For further examples, see Chapter 2 of The Lie Behind the Lie Detector.

10 September 2003 "New Mexico May Lose Polygraph"
This UPI article, published by the Washington Times, is cited here in full:

ALBUQUERQUE, N.M., Sept. 10 (UPI) -- A New Mexico judge has recommended the state discontinue using lie detector results in court, saying the tests are too unreliable.

New Mexico is the only state to routinely permit the results of polygraph, or lie detector, tests to be entered into evidence.

The New Mexico Supreme Court, which makes rules that other New Mexico courts must follow, had asked Albuquerque District Judge Richard Knowles to hear evidence on polygraphs in five consolidated cases from around the state.

The Albuquerque Journal said Knowles subsequently decided polygraph testing lacks standards, is not based on an overarching theory and employs techniques not based upon well-recognized scientific principles.

Knowles wrote: "Because of the inherently subjective nature of the test procedure, the polygraph examination cannot be repeated. Successful repetition of a test is the cornerstone of the scientific method."

Opponents of the use of lie detectors noted in New Mexico, polygraphers receive eight weeks of training while barbers need 1,000 hours of training to be licensed.

10 September 2003 "Elections board idea ignites sparks: Summit Democrat suggests members take polygraphs; GOP chief objects"
Akron, Ohio Beacon Journal staff writer Lisa A. Abraham reports. Excerpt:

The chairman of the Summit County Board of Elections wants all board members to take polygraph tests voluntarily to protect the board's integrity.

Russell Pry made the suggestion at a board meeting Tuesday night.

The county Democratic Party chairman said he would put a formal motion on the agenda of the board's next meeting.

Pry said he thinks the tests would serve to clear the board's reputation in light of recent ``bad press'' it has been getting over mistakes made by elections workers. He noted the ongoing criminal investigation into the lost elections petitions of Akron City Councilman Joe Finley, D-2.

Finley had to sue to get his name on Tuesday's primary ballot after the board lost his petitions. The Summit County Sheriff's Office is investigating.

Republican board member Alex Arshinkoff questioned whether the sheriff wants board members to take the tests.

Pry said the sheriff had not asked for that. The sheriff, however, has been testing elections employees over the missing petitions.

``I'm not going to take a polygraph test with Diane Evans' husband,'' said Arshinkoff, who is the chairman of the county Republican Party.

Evans, a staff writer for the Akron Beacon Journal, is married to Bill Evans, a local private investigator and one of the area's leading polygraph administrators. Arshinkoff has been an outspoken critic of past Beacon Journal coverage.

5 September 2003 "Polygraph Roulette: DOE Has Mastered 'The Expectation Game'"
Bob Park of the American Physical Society comments on Department of Energy polygraph policy in his weekly What's New column:

A two-year study by the National Academy of Sciences, "The Polygraph and Lie Detection," showed polygraph testing to be less than worthless (WN 18 Apr 03). You might have expected at least a token decrease in testing by the Department of Energy. Instead DOE boldly reissued the old policy, which would subject about 20,000 employees to random character assassination. There was an immediate outcry from employees, and Sen. Pete Domenici (R-NM) convened an Energy Committee oversight hearing on Thursday, where DOE announced that a mere 4,500 employees with top-secret clearance or positions in intelligence will now be subject to having their careers trashed by polygraph roulette. It was a victory for Sen. Domenici, who praised DOE for its enlightened policy. But nothing in the NAS study says the polygraph works better if you have top-secret clearance.

5 September 2003 "DOE Restructures Polygraph Program"
Steven Aftergood of the Federation of American Scientists' Project on Government Secrecy comments in his Secrecy News e-mail publication:


Under pressure from members of Congress, scientific critics and others, the Department of Energy said that it will reduce its reliance on polygraph testing to screen its employees.

"The approach I am recommending would have the effect of reducing the number of individuals affected from well in excess of potentially 20,000 under the current rule to approximately 4,500 under this new program," said Deputy Secretary of Energy Kyle E. McSlarrow at a September 4 hearing.

Remarkably, however, Mr. McSlarrow also announced a separate random polygraph testing initiative, which is "an entirely new proposed element" of the DOE polygraph program. Random testing would apply in an unpredictable manner to an additional 6,000 persons not covered by mandatory polygraph screening.

This is truly novel. It is sort of a homeopathic approach to security policy, in which the mere specter of a polygraph test, not even an actual test, is believed to have a deterrent effect and to enhance security.

Under the random screening program, the affected personnel "would be subject to random selection for polygraph examinations at any time, at any frequency.... even though it is possible that an individual in such a position may never actually be selected through the random process," said Mr. McSlarrow. In fact, it seems doubtful that the majority of the 6,000 employees in the proposed random testing program would ever be tested.

Although firm numbers were not immediately available, the Department has probably not come close to testing the 20,000 persons who were nominally subject to the program since it began a few years ago. Consequently, the new "reduction" in the scope of its polygraph program is more apparent than real.

See the testimony from the September 4 hearing of the Senate Energy and Natural Resources Committee on the DOE polygraph program here:

5 September 2003 "Labs will scale back polygraphs"
Andrea Widener reports for the Contra Costa Times. Excerpt:

Under mounting pressure, the Department of Energy will scale back its massive polygraph testing program to half its previous size, a move that may halt mandatory screening tests for some nuclear weapons workers.

The decision is a sudden turnaround for the DOE, which this summer had refused to acknowledge a polygraph testing study by a prestigious scientific panel that was critical of the DOE's policy.

That study, by the National Academy of Sciences, said polygraph tests do not help screen out spies and may even be dangerous, because they give counterintelligence officials a false sense of security.

At a Senate hearing Thursday, Undersecretary of Energy Kyle McSlarrow acknowledged the test's problems and its impact on morale of employees at nuclear weapons laboratories, including the Bay Area's Lawrence Livermore and Sandia/California.

"If you can't eliminate it, you have to manage it," he said in unveiling the plan.

Polygraph critics were pleased with the acknowledgment of the test's problems but remained skeptical about its use as a screening tool.

"As a nation, we should not allow ourselves to continue to be blinded by the polygraph," said Stephen Fienberg, chairman of the National Academy study, who also testified Thursday.

Under the proposed plan, 4,500 nuclear weapons workers and DOE employees in the most sensitive jobs will receive mandatory polygraph tests, down from about 20,000 who had been eligible for the previous tests.

6,000 others will be subject to random polygraph tests, a new part of the program that McSlarrow said was meant to retain the deterrent effects of the tests -- something critics say is unproven.

At Lawrence Livermore, that will likely mean no more than 500 people will get mandatory polygraph tests and about 1,000 may be eligible for the random tests, said lab spokeswoman Susan Houghton. That is significantly lower than the number of employees who were getting tests before.

Some observers, including lab employees, wondered if the move was any change at all, as nowhere near 20,000 employees had been polygraphed even in the height of testing.

"It is certainly good that they're scaling back the numbers, but if this is wrong for 20,000 people then it is wrong for 4,500," said Lawrence Livermore scientist Jeff Colvin, who has been following the issue for the Society of Professional Scientists and Engineers.

While praising the scale-back, the New Mexico senators who conducted the hearings remained skeptical.

"It still seems to me that a large number of scientists in our employ or in the employ of (lab) contractors will be placed under suspicion," Sen. Jeff Bingaman, a Democrat, said at the hearing.

"It makes little difference to the scientist at our labs if the polygraphs are administered to 20,000 or 6,000 when all it takes is one false-positive to ruin a career," Rep. Ellen Tauscher, D-Alamo, said in a statement.

5 September 2003 "Polygraph testing at labs may be reduced"
Staff writer Ian Hoffman reports for the Oakland Tribune. Excerpt:

In an about-face, top U.S. nuclear weapons officials said they plan on scaling back polygraph testing for thousands of weapons scientists and workers, in part to avoid driving them away.

Yet the U.S. Energy Department's proposed reduction of polygraph testing still would leave enough top scientists at labs such as Lawrence Livermore and Sandia subject to a flawed test that falsely could brand hundreds as security risks.

Senators commended DOE deputy secretary Kyle McSlarrow for recommending his agency retreat from testing 20,000 workers, to routine testing of 4,600 scientists and random testing for 6,000 others.

His proposal would mandate testing for scientists who have high-level knowledge of nuclear weapons designs or access to top-secret intelligence. It would eliminate testing mostly for workers who handle, guard, transport or authorize the shipment of nuclear weapons materials.

At Livermore alone, the new policy would relieve almost 600 employees of routine polygraph testing once it takes effect, most likely in several months.

But lawmakers and scientists questioned why the agency still is relying on a test that a National Academy of Sciences panel found was likely to miss real spies and terrorists while mistakenly flagging honest defense scientists as disloyal.

"It is still a voodoo test that jeopardizes these people's careers without being an accurate or reproducible test," said Livermore nuclear safety engineer Bill O'Connell, former president of a lab union, the Society of Professional Scientists and Engineers. "This reform still does not get at the basic program of why they're using an inaccurate test for screening even 5,000 employees."

5 September 2003 "Energy Dept. to Limit Polygraphs' Use"
Associated Press correspondent Robert Gehrke reports in this article published in Newsday. Excerpt:

WASHINGTON -- The Energy Department plans to use fewer polygraph tests to detect espionage at energy labs after a study said employees could be unjustly accused -- in effect reversing a policy that grew out of the Wen Ho Lee investigation.

The department will continue to use the so-called lie detector tests to screen a smaller number of workers with access to the most critically sensitive material -- roughly 4,500 instead of more than 20,000 -- Deputy Energy Secretary Kyle McSlarrow told members of the Senate Energy and Natural Resources Committee.

A National Academies of Science study found that the lie detector tests were not an effective means to screen for spies and would almost certainly result in "false positives" -- innocent lab workers mistakenly coming under suspicion for espionage.

That may be the case in nearly one in six cases, based on the NAS study, and could damage morale at the labs and discourage top-tier scientists from working there, said Sen. Jeff Bingaman, D-N.M., a longtime opponent of the polygraph program.

If 20,000 people were tested, 3,000 would fail the test, Bingaman said.

"We believe national security is too important to be left with such a blunt instrument," said Stephen Fienberg, chairman of the National Research Council committee that reviewed the use of polygraphs.

Senate Energy and Natural Resources Committee Chairman Pete Domenici, R-N.M., said McSlarrow's revisions mark a major improvement, but "I'm still skeptical about the effect of what they're going to have."

The Energy Department began requiring employees take lie detector tests several years ago in the aftermath of the Wen Ho Lee controversy at the department's nuclear weapons laboratory in Los Alamos, N.M. Lee was accused in 1999 of mishandling nuclear weapons codes; the case ended with a plea bargain that freed the Taiwanese-born scientist.

Concerns that the tests were inaccurate prompted congressional demands for the NAS review and that the Energy Department incorporate the results into their polygraph program.

"Polygraph testing yields an unacceptable choice," the NAS report stated. "Its accuracy in distinguishing actual or potential security violations from innocent test takers is insufficient to justify reliance on its use in employee security screening in federal agencies."

5 September 2003 "Energy Dept. to cut use of lie detectors"
Richard Willing reports for USA Today. This short article is cited here in full:

WASHINGTON Citing ongoing doubts about the accuracy of lie detectors, a top Energy Department official said Thursday the department plans to eliminate routine screening of most employees.

About 20,000 Energy Department workers, including many with access to secret weapons programs, are subject to random lie detector tests. Those tests were imposed after the Wen Ho Lee controversy at the department's Los Alamos (N.M.) National Laboratory. Lee was accused in 1999 of mishandling nuclear weapons codes; the case ended with a plea bargain that freed the scientist.

Under the new plan, the random tests would be limited to 4,500 employees with greatest access to secure programs. Workers whose polygraph tests show they are being deceptive will be investigated, but they will not be fired or lose access to secret programs unless the investigation confirms that they are a security risk.

"The bottom line is that we intend that a polygraph screen serve (as) a 'trigger' that may often be useful for subsequent investigations," Deputy Secretary of Energy Kyle McSlarrow told a hearing of the Senate's energy committee.

Critics who say lie detectors often produce faulty results applauded the policy but suggested the tests should be eliminated. "National security is too important to be left to such a blunt instrument," Stephen Fienberg, professor of statistics at Carnegie Mellon University in Pittsburgh, said.

Polygraphs measure changes in breathing patterns and other biological indicators of stress. They are used in criminal investigations and to screen for security risks.

Fienberg was the principal author of a study last year by the National Academy of Sciences that found polygraphs used to screen potential spies are often inaccurate. The study estimated that in a group of 10,000 employees, about 1,600, or 16%, would test as "false positives." McSlarrow said testing by the Department of Energy has produced a "far smaller" rate, but the figure is classified.

5 September 2003 "Government to Give Fewer Lie Detector Tests"
New York Times correspondent William J. Broad reports. Excerpt:

The Energy Department said yesterday in a surprise announcement that it was sharply cutting the number of lie detector tests it would give to people with access to nuclear secrets, particularly at the nation's weapons laboratories.

The deputy energy secretary, Kyle E. McSlarrow, said at a Senate hearing that the new policy was likely to reduce the number of people given polygraph tests to 4,500, mainly in sensitive arms and intelligence posts, from some 20,000 now.

"No one," Mr. McSlarrow said, "has suggested that we abandon their use, or that we hire people and entrust them with national defense information with no prior checks or reviews whatsoever."

But he acknowledged that the department, which runs the nation's nuclear weapons laboratories, had faced many criticisms of the polygraph technique in the last year and had come to agree with some of them. Thus, he said, officials proposed "substantial changes" in the tests' routine use for trying to ferret out spies.

In 2001, Congress instructed the Department of Energy to adopt widespread polygraph screening in reaction to the case of Wen Ho Lee, the scientist at the Los Alamos weapons laboratory in New Mexico who was suspected of being a spy but was freed from jail in September 2000 after admitting to a security violation. That order raised an outcry from experts who ridiculed lie detector tests as pseudoscientific and a potential threat to national security.

Last October, in a report requested and paid for by the Energy Department, a panel convened by an arm of the National Academy of Sciences said polygraph testing was too flawed to use for security screening. The panel said lie detector tests did a poor job of identifying national security risks and were likely to produce accusations against innocent people.

The department's retreat came as a surprise to members of Congress and scientific experts who were preparing to criticize the widespread testing yesterday before the Senate Committee on Energy and Natural Resources. At the hearing, the panel's chairman, Senator Pete V. Domenici, whose New Mexico constituents include thousands of employees of the Los Alamos and Sandia nuclear arms laboratories, praised the department's announcement.

"This is a smart decision," Mr. Domenici said. "I have been appalled by the D.O.E.'s continued massive use of polygraph tests in the wake of a national study condemning the reliability of these tests. Our national scientists deserve better." Home Page > Polygraph News