The Pentagon's push for polygraphs. Congress may allow more tests; critics say careers may be wrongly damaged

The mood in Congress may be shifting in the Pentagon's favor on the controversial polygraph issue, in part because of a growing concern about Soviet spies infiltrating the US military. The Pentagon has long pushed for wider authority to administer lie-detector tests to employees as a means of supplementing background security checks. They would be aimed at helping to expose Soviet and other ``moles'' in the government.

In the past, Congress has been reluctant to grant the Pentagon such broad authority. This is partly out of concern that the careers and constitutional rights of loyal government employees may be jeopardized by a machine that has not been scientifically proved to be 100 percent accurate. (Polygraph results are inadmissible as evidence in US federal courts.)

Defense officials say polygraphs have a 75 to 90 percent accuracy rate. Some critics stress, however, that even with a 90 percent accuracy rate, 1 out of every 10 people who take the test runs the risk of being wrongly identified as a security threat.

If the polygraphs were administered to about 450,000 Navy personnel, as Secretary Lehman proposes, roughly 45,000 people could come under suspicion -- falsely.

Nevertheless, reluctance may be shifting to acquiescence in Congress in the wake of potential damage done to United States national security by the alleged spy ring headed by John A. Walker Jr., say observers in Congress.

On Friday, Sen. Phil Gramm (R) of Texas introduced a bill with a provision that would grant authority to the Pentagon to conduct random counterintelligence polygraph screening for virtually everyone with a Defense Department security clearance. Currently some 4.3 million people hold such clearances.

But congressional critics of the Pentagon's polygraph plans are concerned that defense officials may use the current wave of public indignation over the Walker spy case to gain a freer hand to administer polygraph tests.

``They [Pentagon officials] have the ability to get a major polygraph screening program installed for the first time,'' says a congressional aide. ``This is a major blow to our efforts.''

In the past some members of Congress have viewed such tests as ``fishing expeditions'' aimed more at plugging leaks to the news media than at catching Soviet spies.

Now with the vivid disclosure in the Walker case that a former Navy communications specialist was apparently able for 20 years to assemble a network of family and friends to help him spy and that naval counterintelligence was unable to detect it, a new case is being made for the importance of polygraphs as an adjunct to more traditional counterintelligence methods.

Some observers say they believe that Walker would have been either deterred or detected much sooner had the Navy maintained a more rigorous system of background security checks and random counterintelligence polygraph exams.

``We believe that the polygraph is a valid investigative tool in connection with our overall personnel security program,'' says retired Gen. Richard G. Stilwell, former deputy undersecretary for defense policy. General Stilwell, who has been the Pentagon's most vocal advocate for wider use of polygraphs, was recently appointed to head a Pentagon commission to evaluate Defense Department security procedures with special attention to the Walker case.

Under a congressionally approved pilot program, the Pentagon currently has authority to carry out 3,500 counterintelligence screenings. The screenings are confined to only those Defense employees and contractors with access to the country's most sensitive secrets.

The results of the screenings were to be presented in December in a report to Congress prior to a final decision on whether to permit wider use of the polygraph. But the sense of urgency created by the Walker case may cause Congress to take up the issue sooner.

In response to the case, Navy Secretary John F. Lehman last week announced a proposed plan that would eventually require everyone in the Navy holding a security clearance -- even relatively low-level ``classified'' clearances -- to submit to random polygraph tests.

At present there are 900,000 Navy personnel with security clearances of some kind. Under the Lehman proposal, that number would be reduced by half, meaning that some 450,000 Navy personnel would ultimately face random polygraph tests.

The Central Intelligence Agency and the National Security Agency require extensive polygraph exams as a condition of employment. Intelligence officials have observed that while the producers of intelligence information at the CIA and NSA must take polygraphs, the consumers or analyzers of that information in the Defense Department traditionally have not had to. The Pentagon would like to eliminate the discrepancy, establishing a more uniform system of security checks for all government workers with access to sensitive information.

Rep. Jack Brooks (D) of Texas, chairman of the House Government Operations Committee, sees the Pentagon's efforts as a dangerous precedent.

``The Department of Defense's push for increased polygraph use is a quick fix, and it will in all likelihood backfire,'' he says. ``The fact is that polygraph screening has no scientific basis.''

Pentagon officials have stressed that action against an individual wouldn't be taken solely on the basis of the results of a polygraph exam. They add that exam results would not be included in personnel files.

But others suggest that once suspicion is aroused about an individual working in a highly sensitive area, it can have a profoundly adverse impact on a career.

Several members of Congress are concerned that the Defense Department may become too reliant and trusting of polygraphs, rather than using them to supplement personal interviews and extensive background checks.

Critics of the Pentagon's position say that all major studies of polygraph accuracy are based on exams given in the course of specific criminal investigations. They note that there is no reliable measure of the accuracy of polygraphs in conducting counterintelligence screenings.

1. Have you ever engaged in espionage or sabotage against the US? 2. Have you ever given or sold any classified materials to unauthorized persons?

3. Do you have knowledge of anyone who has given or sold classified materials to unauthorized persons?

4. Have you had any unauthor- ized contact with representatives of a foreign government?

You have just taken an abbreviated version of a counterintelligence national-security polygraph screening test. The important exception is that you weren't hooked up to a machine that carefully records your slightest physiological responses to each question.

Even if you answered ``no'' to each of the above questions, that's not what counts to a polygraph examiner. What counts is what registers on the machine.

There is considerable debate about whether such questions could ultimately trip up a well-trained Soviet spy who had burrowed deep within the US military bureacracy and who had no particular feelings of guilt about his mission in the US.

Christopher Boyce, who sold US satellite secrets to the Soviets in the mid-1970s, says he was told by his Soviet control agent in Mexico City that US polygraph tests posed no obstacle for KGB agents. ``Oh, that's no problem, we have ways of having people pass polygraphs,'' the agent is said to have told Boyce.

Jack Donnelly, director of counterintelligence and investigative programs at the Defense Department, doubts such KGB claims. ``They have given the same assurance to a number of people,'' he says, ``But to my knowledge they have never given anyone such training.''

He notes that even if they had, polygraph operators are trained to spot countermeasures. ``I don't know of any `mole' [spy] that has been able to beat the machine,'' Mr. Donnelly says. But ``there have been a number of cases where the polygraph prevented an infiltration.''

Nonetheless, there have been reports of a special school in Czechoslovakia to teach KGB and East-bloc agents how to beat polygraphs.

``If I were running the KGB, I would certainly be doing that,'' says Paul Ekman, a professor of psychology at the University of California Medical School in San Francisco.

``In the Soviet Union they have recruited their best scientific talent to work in the area of lie detection,'' says Professor Ekman, author of the book ``Telling Lies.'' ``In this country, for a variety of reasons, our best scientific talent isn't being used.''

Intelligence officials have pointed out that polygraphs are most useful not in catching people lying, but rather in provoking a confession prior to the actual test. The confessions are premised on the belief -- the fear -- that the polygraph machine will detect deception.

Critics of the polygraph maintain that sophisticated moles or spies will not confess.