<<

. 79
( 118 .)



>>

when a government investigator knocked on his door in the mid-1980s
and began quizzing him about his neighbor, a high school student
looking for a summer job at NSA. The questions started off routine, he
said, but soon turned very personal. "Do you know if he's a homosexual?
Does he use drugs or alcohol? Does he go to church frequently? What
can you tell me about his home life? Does he get along with his parents?"
"Appropriate character" is what NSA was looking for, according to Bill
Shores, in charge of NSA's college recruitment program at the time. That
someone is homosexual or a drug user, per se, "does not mean [he or
she] can't come to work for NSA," he said, but "a person that has
something to hide would not be a good security risk."
NSA officials are fighting a new proposal by the Defense Security
Service to abandon neighborhood in-person visits in favor of simple
telephone calls. The DSS argues that it can no longer afford such costly


448
and time-consuming procedures. Pointing out that NSA is only one of its
customers, DSS officials say that they must conduct more than 250,000
background investigations of government and contractor personnel each
year, leading to tremendous backlogs. At the same time, the agency is
behind on tens of thousands of five-year updates required for the 3
million federal employees and contractors who hold active security
clearances. Thus, by 2000, DSS's total backlog was a whopping 900,000
investigations. On top of those problems, DSS personnel have been cut
back about 40 percent in recent years, from 4,300 employees in the mid-
1980s to 2,500 in 1998.
A survey done in 1999 discovered that 94 percent of the background
investigations DSS conducted for NSA were incomplete and not up to
federal standards. That same year, a routine reinvestigation polygraph
examination resulted in the arrest of Daniel King, a Navy petty officer
working for NSA. King, an eighteen-year veteran, was arrested on
October 28 and charged with espionage for allegedly confessing to
mailing a computer disk to the Russian embassy five years earlier. The
disk allegedly contained supersensitive details on NSA's undersea cable-
tapping operations against the Russians.
After the SSBI is completed, the results are sent back to NSA for
evaluation.
The next phase takes them down a narrow passageway to an area of
small offices that sends shivers down the backs of most candidates:
Polygraph Services. Within the tiny beige offices, new, computerized
polygraph machines sit on wood-grain desktops and are attached to
monitors that display the recruit's physiological responses in a variety of
formats. Among the data recorded, according to an NSA document, are
the individual's "respiration, electro-dermal responses, pulse rate, pulse
amplitude, vascular volume, capillary volume, vascular pressure,
capillary pressure, and bodily movement as recorded by pneumograph,
galvanograph, cardiosphygmograph, plethysmograph and cardio activity
monitors, which are sections of polygraph instruments." Watching the
graphics form sharp peaks and deep valleys is one of the agency's several
dozen certified examiners. Many of the questions they ask come from the
results of the SSBI.
On the other side of the desk, the applicant sits in a large, heavily
padded, executive-type swivel chair. Electrodes are attached to the
fingers; rubber tubes are strapped around the chest; and a bulky blood
pressure cuff fits around the upper arm. What the examiners are looking
for are significant changes from the subject's baseline chart. These may
be as dramatic as a total cessation of breathing or a major increase in
blood pressure”or as subtle as a slight decrease in skin resistance.
The Armed Forces Security Agency began the polygraph program in



449
May 1951 with the hiring of six examiners at annual salaries of $6,400.
The program was introduced because the agency was growing so quickly
that background investigations could not be completed fast enough for
the hiring program. More than 1,000 people had been hired but could
not be cleared until their background investigation was finished, which
because of the Korean War was taking from nine to eighteen months. By
1953 NSA was giving polygraphs to all job applicants. The questioning
was originally conducted in a well-guarded, ominous-looking building at
1436 U Street, NW, in Washington, before the office moved to the
Operations Building and then to FANX.
The polygraph remains the most dreaded part of NSA's admission
ritual. "Polygraph! The word alone is enough to set your nerves on edge,"
began one article on the machine in NSA's in-house newsletter. It is also,
by far, the most important part of that ritual. According to a study at
NSA, 78 percent of all information used in evaluating an applicant as a
security risk comes from the polygraph reports. Only 22 percent of the
information is based on the background investigations.
From July 1983 to June 1984 the agency administered a total of
11,442 examinations. Of those, 4,476 were given to job applicants. From
that group, 1,875 dropped out voluntarily for a variety of reasons. Of the
remaining 2,601, 793 were rejected by the agency's Applicant Review
Panel, composed of personnel, security, and medical managers. As an
example of the power of the box, a whopping 90 percent of those (714 of
793) were booted because of bad polygraph results. During the first half
of 1984 a total of 1,202 contractors were strapped to the machine, and
167 were shown the door after leaving the polygraph room.
The polygraph sessions earned a black eye during the 1950s and early
1960s because of the agency's heavy dependence on the EPQ, or
embarrassing personal question. EPQs are almost inevitably directed
toward intimate aspects of a person's sex life and bear little relationship
to his or her honesty or patriotism. Following a congressional
investigation and an internal crackdown, the personal questions became
somewhat tamer but abuses have occasionally continued.
"The worst experience of my life," said one former NSA Russian
linguist, "was the lie-detector test." After starting out with questions
about shoplifting, the polygraph operator quickly turned to sex, asking if
she was into bestiality. "If you have sex, they want to know how much. If
you have too much sex, they get scared. If you don't have sex, they think
you're gay. At the time I wasn't dating anybody and they kept wanting to
know, 'Why don't you have a boyfriend?' " That test was given in 1993.
More recently, NSA claims, the questions have been less intrusive.
Contractor employees were first required to take polygraphs in 1957.
And in 1982, following a damaging spy scandal at Britain's GCHQ,



450
military personnel assigned to NSA were first required to be strapped to
the box. The military entrance polygraph is conducted by the military
services on military assignees before their acceptance for a position at
NSA and is directed toward counterintelligence questions.
At the same time, a five-year reinvestigation polygraph examination,
which also focuses on counterintelligence-related questions, was
introduced for all employees. Still another polygraph program, the special
access examination, was instituted to test employees about to be
assigned to especially sensitive programs within NSA. Those tested under
this program are asked both counterintelligence and, under certain
circumstances, "suitability"”personal”questions.
Finally, again in 1982, NSA instituted a dreaded policy of
unscheduled "aperiodic" counterintelligence polygraph examinations.
One purpose of these tests is to look for spies; another is to look for
leakers. According to a memo from the director, civilian employees who
refused to consent faced "termination of employment." The agency, said
one senior NSA official, asked the Justice Department to investigate
about four leaks a year during the first half of the 1980s.
Among the topics covered during NSA's counterintelligence polygraph
examination are the following.


• knowledge of, participation in, or commission of acts of
espionage or sabotage against the United States
• knowledge of, approaches to, or giving or selling any classified
information or material to unauthorized persons
• unauthorized or unreported foreign contacts


The idea of suddenly being called from an office, strapped to a
machine, and asked whether you have been selling secrets to the
Russians or leaking information to the press might leave "the work force
at NSA ... shocked," said Philip T. Pease, the chief of the Office of
Security at NSA. As a result, employees were called to the Friedman
Auditorium for a series of town meetings during which the new
procedures were discussed.
Under the aperiodic exam program, the agency, without notice, pulled
1,770 people into the polygraph rooms in 1983. Of those, 1,699 were
thanked and sent on their way. Seventy-one, however, were asked to
come back for a further interview, which cleared all but four. They
returned for a third round of drilling but were eventually also allowed to
return to work, presumably a few pounds lighter. According to the chief
of the Polygraph Division, Norman Ansley, the problems ranged across


451
the board. One individual had kept a classified manual at his residence
for several years. Another person knew of the improper destruction of
crypto keying material. Still another described a suspicious approach by
foreign personnel but had failed to report the incident at the time it took
place.
After the test, the examiner reviews the individual's charts and makes
a final decision on the results. "NSR" (no significant response) means
that there were no unresolved issues. "SPR" (specific physiological
response) signifies that the individual reacted consistently to a specific
question. "INC" (inconclusive) means that the test results could not be
interpreted. And "incomplete" signifies that the test was not finished.
When issues are unresolved, the individual is requested to return to the
box for retesting.
Once completed, the examiner's report is forwarded to quality control
for an independent review of accuracy and analysis and to ensure that all
issues have been covered. From there it travels to the Clearance Division
for adjudication. The polygraph examiner does not make any clearance
decisions. His or her sole purpose is to verify the validity of the
information being provided during the interview and to resolve any
matters that are causing the person difficulty in passing the test.
A unique insight into the NSA polygraph program comes from an
analysis of 20,511 applicants between 1974 and 1979. Of those, 695 (3.4
percent) admitted to the commission of a felony. In nearly all cases the
perpetrator had gone undetected. The admissions included murder,
armed robbery, forcible rape, burglary, arson, embezzlement, hit-and-run
driving with personal injury, thefts of expensive items or large amounts
of money, smuggling, and wholesale selling of illegal drugs.
One person who applied to NSA proved to be a fugitive who, during
questioning under the polygraph, admitted firing a rifle into his
estranged wife's home in an attempt to murder her. Another confessed to
firing his shotgun at six people, and hitting all of them. He had been
charged with attempted murder but not tried because of lack of evidence.
Still another told of setting fire to the trailer in which his ex-wife and
their child lived. A veteran admitted to a polygraph operator that while in
Vietnam he had murdered a young girl. On a later occasion he stabbed a
stranger in the face with a knife in an argument over some beer. And an
applicant for an engineering position”who was employed as an engineer
by another government agency”blurted out that he had shot and
wounded his second wife and that his present wife was missing under
unusual circumstances, which he would not explain. He also suddenly
declared that his engineering degree was phony.
Even espionage has turned up during polygraph examinations. One
applicant with access to Top Secret/Codeword intelligence who was



452
about to retire from the military described making several visits to the
Soviet embassy to make arrangements to defect to the Soviet Union. The
Russians took copies of his classified documents and when they found
out he had applied to NSA for employment, they encouraged him to
continue.
Another applicant, who had access to classified information while in
the military, confessed that he would sell classified information to a
foreign intelligence service if he could get enough money. And one person
looking for a job at NSA eventually admitted that much of his
background was falsified and that he had worked as a scientific adviser
to the chief of a foreign military intelligence agency.
Most significantly, the NSA study indicated serious questions about
highly cleared military personnel assigned to NSA's Central Security
Service. At the time, service members were not subject to the polygraph.
During the five-year period of the survey, 2,426 of these SCI-cleared
military personnel applied for employment with NSA as civilians. Of that
number, thirteen admitted that either they themselves or someone they
knew had been involved in espionage. Another twenty-five told of passing
classified information to Communists or terrorists.
In the early 1990s NSA became the first intelligence or defense agency
to completely computerize its polygraph program”the first major change
in the art of polygraphy since 1940. According to NSA officials, the
computerized polygraph equipment was found to be more accurate than
conventional methods because it could record signals at maximum
sensitivity. The computer also allows the examiner to change how the
data are displayed on the screen without changing the base data, thus
protecting the validity of the test.
The agency is currently working on ways to almost completely
eliminate all human involvement in the polygraph process. In the near
future an Orwellian computer, programmed with an individual's history,
will ask the questions, analyze the answers, and decide whether a person
is lying or telling the truth.
In 1991 the Office of Security Services, working with the Mathematics
Branch of the Research and Sigint Technology Division, issued a contract
aimed at elevating the computer from simply a passive display to an
active analyst. The lead contractor on the project, Johns Hopkins
University's Applied Physics Laboratory, was able to develop a system
called Polygraph Assisted Scoring System (PASS) using a briefcase-sized
AXCITON computerized polygraph. Unlike evaluation by humans, which
sometimes took days or even weeks to produce a final decision, the
computerized procedure finishes within two or three minutes of the
exam. Using both a history of the individual's past tests and his or her
own physiological makeup, the computer comes up with a statistical



453
probability concerning the meaning of test results.
Although with the PASS system examiners would still make the final
determination, their future does not look bright. Early testing indicates
that computerized analysis is more accurate and produces fewer
inconclusive results than human-administered tests. According to one
NSA document, "In the near future, it may even be possible for the
computer to ask the test questions”eliminating any possibility of the
examiner's affecting the test results."
But despite the growing dependence on the polygraph, the box is far
from infallible, as Norman Ansley, chief of NSA's Polygraph Division
during the 1980s, once admitted. Asked whether someone addicted to
drugs and alcohol could beat the box, his answer was "Possibly," if that
person "had practiced dissociation by thinking of something else." Which
is precisely why many both inside and outside government distrust the
machines. "Polygraphing has been described as a 'useful, if unreliable'
investigative tool," said the Senate Select Committee on Intelligence in
1999. Given such questionable data, the panel asked CIA director George
J. Tenet and FBI director Louis J. Freeh to assess "alternative
technologies to the polygraph." The newest agency to use the polygraph

<<

. 79
( 118 .)



>>