Agencies, employees spar over lie detector tests
In Richardson's case, the decision to go ahead with lie detector tests was prompted by missing computer hard drives at the Los Alamos nuclear weapons lab in New Mexico; Reno's headache was a leaked report from the Justice Department task force investigating Vice President Al Gore's fund-raising activities. Sen. Robert G. Torricelli, D-N.J., recommended the polygraph tests to the Attorney General at a Senate Judiciary Committee hearing on June 27, but it remains an open question whether such testing will actually be carried out at Justice. Spokeswoman Obern Rainey says that because the campaign finance investigation is ongoing, the department won't comment on the issue.
Richardson's swift resort to polygraph testing was meant to plug the holes in federal security and quell controversy. Instead, it launched a debate about the accuracy of the tests and about whether federal agencies should use them on their employees. The issue is now being argued in the nation's nuclear labs, courts, and scientific communities. Critics of the polygraph say that at best it is unreliable, and at worst it's nothing more than government-sanctioned interrogation and intimidation. Supporters acknowledge that the test has its limitations, but say it's a necessary tool in safeguarding national security.
Random polygraph testing of employees is routine at many federal agencies, including the FBI, the CIA, and the Secret Service, where authorities are trying to uncover illegal drug use, theft, and disclosures of classified information. In 1994, FBI Director Louis Freeh authorized the use of polygraph examinations for all job applicants on the grounds that the conventional investigative methods weren't always "capable of detecting national security risks," said Donald M. Kerr, the director of the FBI's lab division.
In October 1999, following reports of espionage by the People's Republic of China that led to the arrest of Wen Ho Lee, a scientist at the Los Alamos lab, Congress created the National Nuclear Security Administration and authorized Richardson to conduct, in consultation with the FBI, counterintelligence polygraph examinations of Energy Department employees who have access to high-risk programs. The testing effort was to focus on the staffs of Los Alamos and Sandia, Energy's nuclear weapons labs in New Mexico, and Lawrence Livermore lab in California.
Congress's original mandate envisioned testing 13,000 lab employees, but, according to Edward J. Curran, the director of counterintelligence at the Energy Department, the number was reduced to just 868 people in sensitive positions. Rep. Heather Wilson, R-N.M., who generally supports the use of polygraph examinations as a counterintelligence tool, criticizes Energy's response: "The Department of Energy did a terrible job of implementing the polygraph program. By doing it for public relations reasons rather than security, they ensured failure." She adds that Richardson "acted in a way that was capricious and unnecessary, rather than one that would be designed to handle the intelligence threat."
At the two labs, some employees greeted the policy with "Just Say No to Polygraphs" buttons. The Federation of American Scientists decried the program. FAS senior research analyst Steven Aftergood said, "Congress has a closed mind to the subject and imposed the requirements without considering the validity, reliability, or consequences [of polygraph testing]."
Criticism of polygraphs is nothing new, according to Brandeis University professor Leonard Saxe, who served as a staff member of the Polygraph Validity Advisory Panel, created by Congress in 1983. The panel concluded that polygraphs' ability to detect lies was only slightly better than random chance, and that errors were possible. In 1998, the panel's findings were cited in a U.S. Supreme Court case, United States vs. Scheffer, as support for the conclusion that "to this day, the scientific community remains extremely polarized about the reliability of polygraph techniques." The American Medical Association says that more research is needed, and until additional studies are completed, testing "shouldn't be undertaken in the private or public sector."
Alan Zelicoff, a senior scientist at Sandia, says that the use of a nonscientific tool to screen for deception has caused "a palpable sense of decline in morale, that we are widgets rather than highly trained scientists and engineers." In October 1999, he and eight other senior scientists conducted their own review of polygraphs. They concluded, based on an analysis of the published research, that it is impossible to predict the test's error rate. They also found that a subtle but significant part of polygraphy is the reliability of the examiner. "All humans, even polygraph examiners, have biases of one sort or another that can create errors in polygraph test interpretations," they said.
Similar opinions are shared by two prominent researchers in the lie-detection field-Sheila Reed and Charles R. Honts-both of whom worked at the Department of Defense Polygraph Institute. The institute provides basic and continuing education for all federal polygraph examiners. Reed, the clinical psychologist responsible for developing and standardizing the test format and operator's manual currently used by the Department of Energy, told National Journal that government-trained examiners don't understand psychology, physiology, and electronics, and that their procedures are "unethical." In addition, she said, her preliminary research at the institute showed that polygraph examiners do have biases that can affect results.
Honts, now at Boise State University's department of psychology, was assigned to analyze polygraph research data at the Department of Defense Polygraph Institute in the late 1980s. He says the results showed that the test doesn't work. He also noted that the procedure had always been expected to be prone to false positives, or false accusations of deception. But he was surprised to discover the opposite result: high false-negative rates, or deception going undetected. He says that the government is suffering from a "major case of institutional denial" and that it has allowed "policy to drive science, not science driving policy."
Drew Richardson, a supervisory special agent and scientist for the FBI, testified before the Senate Judiciary Committee in September 1997 that polygraph tests are of no value in national security matters. "I believe there is virtually no probability of catching a spy with the use of polygraph screening techniques," he said. "To the extent that we place any confidence in the results of polygraph screening, and as a consequence shortchange traditional security vetting techniques, I think our national security is severely jeopardized." (In response to a request by National Journal for comment from the FBI on its use of polygraph exams, spokesman Paul E. Bresson said, "We can't answer any questions at this time.")
But not all experts hold the polygraph in such low regard. Michigan State University professor Frank Horvath, a former president of the American Polygraph Association, the largest professional polygraph organization, strongly disagrees with the naysayers and adds that critics of polygraphy have no better evidence on the issue of reliability than do proponents. Horvath says that the evidence shows polygraphs to be accurate up to 90 percent of the time. He also says that polygraph results are not the sole determining factor in the government's efforts to uncover deception. The Energy Department's Curran agrees: "We know it's not scientific, but it's not the only thing we use to determine whether you get access to classified information."
David Renzelman, the Energy Department's polygraph program manager, also defends the test. He said the polygraph is "like a camera, only a recording device" to take down responses to standardized questions. According to Renzelman, examiners are well trained in analyzing the responses and determining the significance of any discomfort a question might cause. John Furedy, a professor of psychology at the University of Toronto who has published nearly 300 articles related to the subject, disagrees: "The polygraph is not a test, because, unlike other psychological tests, like IQ tests, it is not standardized.... The examiner makes up some of the questions on the spot.... It is really an interrogative interview."
The polygraph is now under scrutiny in the U.S. District Court for the District of Columbia, says Mark Zaid, an attorney representing a group that has filed a class action suit against the government. The plaintiffs say that the FBI, the Secret Service, and the Drug Enforcement Agency unfairly denied them employment as a result of polygraph exams. In its response to the suit, the U.S. Attorney's office says the fact that a screening tool is imperfect does not mean that it should be prohibited: "For example, if comparable examinations were given to brain surgeons, one would be hard-pressed to argue that the surgeon who failed the 'flawed' examination should still be hired and allowed to operate."
Curran is determined that the polygraph program at the Energy Department succeed: "I have a level of confidence in the polygraph, not a high level. My experience with the polygraph is that you've got to have the best examiners.... They are key." Renzelman says that Energy's polygraph program is the best in the government. It has been certified by the Department of Defense Polygraph Institute in an inspection and, he says, employs only the most experienced polygraph examiners with proven records.
What lies ahead? Aftergood, of the Federation of American Scientists, says: "The mania over security has turned the labs into fortresses-that is not a fruitful environment for innovation." Rep. Wilson says that she is not sure that the Department of Energy can recover from this "ill-conceived plan."