INTELLIGENCE MONOGRAPH SECRECY VS DISCLOSURE: A STUDY IN SECURITY CLASSIFICATION

• Document Type: CREST
• Collection: General CIA Records
• Document Number (FOIA) /ESDN (CREST): CIA-RDP80-00630A000300010001-6
• Release Decision: RIPPUB
• Original Classification: C
• Document Page Count: 118
• Document Creation Date: December 9, 2016
• Document Release Date: January 5, 2001
• Sequence Number: 1
• Content Type: STUDY

CONFIDENTIAL
Approved For Release 2001/09/04: CIA-RDP80-0063OA000300010001-6
OGC HAS
REVIEWED.
INTELLIGENCE
MONOGRAPH
A STUDY IN SECURITY CLASSIFICATION
a ~ CENTER FOR THE STUDY OF INTELLIGENCE
CENTRAL INTELLIGENCE AGENCY
TR/IM 76-06
Approved For Release 2001/09/04: CIA-RDP80-00630A000300010001-6
CONFIDENTIAL
 Approved For Release 2001/09/04: CIA-RDP80-0063OA000300010001-6
THE CENTER FOR THE STUDY OF INTELLIGENCE IN OTR OPERATES
A RESEARCH AND DISCUSSION PROGRAM KEYED TO THE PROCESSES
AND FUNCTIONS OF INTELLIGENCE. THE OBJECTIVE OF THE CENTER
IS TO CONTRIBUTE TO THE PROFESSIONAL UNDERSTANDING AND
TO THE RECORD OF THE ART OF INTELLIGENCE. RESEARCH PROJECTS
ARE UNDERTAKEN BY INTELLIGENCE "FELLOWS"--VOLUNTEER
OFFICERS FROM ACROSS THE AGENCY ON FULL-TIME DETAIL TO
THE CENTER. INQUIRIES ABOUT THE CENTER PROGRAM, OR COMMENTS
ON THIS REPORT ARE INVITED BY THE DIRECTOR/CSI, EXTENSION 2193.
Classified by 031484
Exempt from general declassification schedule
of E.O. 11652, exemption category:
? 5B(2)
Automatically declassified on:
date impossible to determine
Approved For Release 2001/09/04: CIA-RDP80-0063OA000300010001-6
 Approved For Release 2001/09/04: CIA-RDP80-0063OA000300010001-6
Page
INTRODUCTION
i
PRINCIPAL CONCLUSIONS AND RECOMMENDATIONS
v
DEVELOPMENT OF THE CLASSIFICATION SYSTEM
1
Criticism of the Classification System
4
THE FUNCTIONING OF CLASSIFICATION WITHIN CIA
12
Misclassification
13
Related Issues
17
Reasons for Misclassification
21
Anomalies in the Handling of Top Secret
23
Material
Exemption, Downgrading, Declassification
26
Impact of Freedom of Information and
30
Privacy Acts
Scope of Noncompliance with E.O. 11652
34
Effect of Classification on CIA
37
The Protection of Sources and Methods
42
Statutory and Constitutional Barriers
51
to Disclosure
CONCLUSIONS AND RECOMMENDATIONS
60
FOOTNOTES
80
ANNEX A - CLASSIFICATION CRITERIA
94
ANNEX B - PRINCIPAL CHANGES EFFECTED BY
102
E.O. 11652
ANNEX C - 13REAKDOWN OF THE CLASSIFICATION
105
OF THIS STUDY
Approved For Release 2001/09/04: CIA-RDP8O-0063OA000300010001-6
 Approved For Release 2001/09/04: CIA-RDP80-00630A000300010001-6
CONFIDENTIAL
SECRECY VS. DISCLOSURE - A STUDY IN
SECURITY CLASSIFICATION
"We have an obligation to
provide as much information
as possible on an unclassi-
fied basis, but without
derogation of the necessity
to protect sensitive sources
and methods and to protect
information which truly
requires sensitive treatment."
Director of Central Intelligence
Guiding Ptc,ine.ipte4 6ote the
IfteUigenee Community.
INTRODUCTION
This Study is the first in a series undertaken by the
1
Center on the subject of compartmentation, the basic purpose
of which in the Agency and elsewhere is to preserve secrecy.
The principal technique for doing this, the one on which the
others repose, is the federal classification system. This
Study looks first at the historical development of classifica-
tion, seeking to isolate its endemic problems and to gain a
fresh perspective on a procedure that has become hackneyed
for most of us. The Study then focuses on classification
practices within CIA--a major factor in Agency compartmentation.
CONFIDENTIAL
Approved For Release 2001/09/04: CIA-RDP80-00630A000300010001-6
 Approved For Release 2001/09/04: CIA-RDP80-00630A000300010001-6
CONFIDENTIAL
It also looks at the interrelationship of classification
and the Freedom of Information and Privacy Acts, at the
statutory protection of sources and methods, and at the
other legal barriers to the disclosure of foreign intelli-
gence information.* The preservation of secrecy through
classification and the other techniques of compartmentation
ultimately depends ors the legal power to withhold informa-
tion and to deter unauthorized disclosure.
The protection of classified information in our society
has always involved a tug of war between the conflicting
requirements of secrecy and publicity. From the earliest
days there has been conflict between the Executive and the
Legislature over Executive confidentiality. Events, moreover,
since the adoption of the present classification system in
1972--particularly Vietnam, Watergate, and various disclosures
and investigations--have intensified these historical tensions,
generating, both inside and outside the government, sharp
25X1A *The author of this Study was (DDO/AF) assisted
by the staff and fellows of t e Center for the Study of Intel-
ligence. The methodology of the Study involved interviewing key
personnel in the Agency involved with classification and related
matters and extensive research in the available literature.
Congressional hearings and studies are the most important source
of the latter. Although there have been numerous studies of
codeword compartmentation within CIA and the Intelligence
Community, we are not aware of any single, published, compre-
hensive study of the classification system done within the
Intelligence Community. We have, therefore striven for compre-
hensiveness in the Study both as a means of solving the immediate
problems at hand and of facilitating the further research of
others concerned with classification.
Approved For Release 200/:IAIfkbP80-00630A000300010001-6
 Approved For Release 2OD1F/IDS:I@I[A-RDP80-00630A000300010001-6
criticism of classification practices. Whenever classification
decisions have lacked an aura of credibility for Congress and
the public, the information they protected has been vulnerable
to disclosure. No statute or Executive order has long sufficed
to protect information that Congress or the public has come to
regard as pseudo-secrets or cover-up of illegality. The amended
Freedom of Information Act (FOIA) and the Privacy Act have
brought the executive branch to the realization that in a
democracy the "right to know" of Congress and the public is
just as imperious as the "need to know" of the bureaucrat.
This counterpoint between secrecy and disclosure and, in a
closely related sense, between secrecy and publicity, is the
main theme of this Study.
The developments referred to above have led the executive
branch to reexamine the full implications of secrecy and clas-
sification. National Security Study Memorandum (NSSM) 229,
dated 16 August 1975, created an ad hoc interagency group,
under the chairmanship of the Deputy Assistant to the
President for National Security Affairs, charged with
making a comprehensive study of the classification system
and submitting, if appropriate, recommendations for the
revision of E.O. 11652 and possible legislation on the
management of classified information. Symptomatic of CIA
concern with classification problems was the establishment
on 13 January 1976 of an interagency group chaired by
Approved For Release 2@/p . RDP80-00630A000300010001-6
 Approved For Release 2001/09/04: CIA-RDP80-00630A000300010001-6
CONFIDENTIAL
Dr. Edward Proctor, charged with finding "a basis for designing
a simplified classification system which will provide the neces-
sary protection of intelligence sources and methods and will
facilitate access to the results of the intelligence process
by all consumers who need the intelligence product."* Finally,
E.O. 11905, issued 18 February 1976, directed the DCI "to
establish a vigorous program to downgrade and declassify
foreign intelligence information as appropriate and consistent
with E.O. 11652."
The Center hopes that this Study will contribute to the
efforts of those officers engaged in carrying out the above
orders and will be of use to others interested in the impact
of a vital adjunct of the intelligence process within the CIA.
A summary of its principal conclusions and recommendations is
set forth in the next section. (The full text of the recom-
mendations is contained in the final section.)
In an effort to provide an independent perspective on clas-
sification problems, this Study has not been linked with the
results of the Proctor group.
Approved For Release 20019((4)E-DP80-00630A000300010001-6
 Appro ADMINISTIVE01/091/RILR~0-M~OA000300010001-6
PRINCIPAL CONCLUSIONS AND RECOMMENDATIONS
Throughout the history of the classification system,
going back even to its nineteenth century precursors, certain
negative traits--overclassification, unnecessary classifica-
tion, vagueness of the classification criteria, and accumula-
tion of vast quantities of classified paper--have always been
present. Unfortunately, they are all part of the classifica-
tion experience of the CIA today. In recent years, the
negative traits have attracted public and congressional
attention and have resulted in a series of studies, regula-
tions, and reforms intended to improve the system. There
has been some improvement, but not much.
- Although recognizing the primary responsibility
of the National Security Council and the ICRC
in the area of classification, this Study recom-
mends that CIA take a leading role in the develop-
ment of classification theory and in the reform
of the present classification system. In this
sense, six recommendations are offered in the
Study for the reform of E.O. 11652.
Their principal thrust is to:
- increase the specificity of the classification
criteria,
- prevent misclassification for reasons of
administrative privacy,
- replace the ICRC with a greatly strengthened
interagency classification board,
- restrict the establishment of compartments
that have repercussions beyond the originating
department, and
ApprovecfPW; hL i/09/0 .~8W-AbP 10%'M000300010001-6
 Approved For Release 2001/09/04: CIA-RDP80-00630A000300010001-6
ADMINISTRATIVE - INTERNAL USE ONLY
- raise the level of classification-consciousness
throughout the government.
The Study finds significant Agency noncompliance with
the provisions of E.O. 11652. The principal areas are over-
and unnecessary classification, excessive use of the exemp-
tion provision, and anomalous handling of Top Secret cables.
Due to bureaucratic pressures and the force of precedent,
classification errors made in the past tend to be perpetuated.
Original classifiers, who are usually busy supervisors, often
abdicate their classification responsibilities to subordinates.
The "play-if-safe" mentality is a powerful contributing factor
to overclassification. As remedies, the Study recommends:
- the establishment of a CIA Classification Board
to replace the present CIA Information Review
Committee,
- new guidance and stepped-up training in
classification,
- the beginning of classification of documents
by section or paragraph,
- the preparation of internal CIA guidelines for
de novo classification of information taken
from raw reports for use in finished intelligence,
- the handling of Top Secret cables in the same
manner as other Top Secret collateral material,
- the abolition of pre-printed Secret forms,
- a thorough review of the number and level of
authorized classifiers within the Agency,
- a new regulation on administrative privacy,
- research into means of controlling the reproduc-
tion of classified paper.
Approved FAq0?jTgqj(p
VQ9/D4rlW8(Q60,00300010001-6
 Approved For Release 2001/09/04: CIA-RDP80-00630A000300010001-6
ADMINISTRATIVE - INTERNAL USE ONLY
In assessing the effect of the classification system on
day-to-day operations within CIA, the Study concludes that,
except for Top Secret, the flow of information within the
Agency, even when accompanied by overclassification, is only
slightly impeded in reaching proper users.
- But the cumulative effects of overclassification
are more insidious: Bad secrets devalue good
ones, setting in motion a process of escalating
protection for the "real" secrets which drains
the classification criteria of their prescribed
meaning.
- The regular classification categories have thus
become the underpinning for a vast superstructure
of supplementary protection. The latter includes
not only the codeword compartments, but also the
array of document control markings and the mini-
compartments employed by the Operations Directorate
to protect sources and methods.
The computer system of the Central Reference Service (CRS)
recognizes some 1400 combinations of document control markings.
The latter not only complicate retrieval of information, but
sometimes preclude it altogether when, for reasons of sensitivity,
documents are not indexed in the Agency's central reference
system.
- The Study concludes that a DDO dissemination,
stamped Secret and invested with some of the
more restrictive handling controls, probably
enjoys greater protection than the average
Top Secret collateral or Top Secret codeword
document. These access-limiting and distribu-
tion-control devices, and the special handling
devices within the DDI, have a very tangible
effect on the flow of information within the
Agency.
ADMINISTRATIVE - INTERNAL USE ONLY
Approved For Release 2001/09/04: CIA-RDP80-00630A000300010001-6
 Approved For Release 2001/09/04: CIA-RDP80-0063OA000300010001-6
ADMINISTRATIVE - INTERNAL USE ONLY
- Although not taking a stand against these tech-
niques, the Study points out that the lack of
formal criteria for their use and their irregular
application sometimes deprive analysts and operators
of the information they need to do their job.
FOIA and the Privacy Act are having a significant impact
on the Agency and on the functioning of the classification
system within the Agency. The statutory and judicial aspects
of these Acts place a premium on strict compliance with the
classification and exemption provisions of E.O. 11652 and on
quick discernment of overclassification in disputed cases.
In the long run, they should have a corrective influence on
classification abuses.
After examining the origins and judicial history of the
sources-and-methods provision of the 1947 National Security
Act, the Study concludes that, as it stands, the provision is
an imperfect vehicle for protecting intelligence information
from disclosure. Other statutory and constitutional barriers
to disclosure are also examined and found wanting.
- The Study judges that the Atomic Energy Act of
1954 establishing the category of "Restricted
Data" is perhaps the most successful security
program in the government and may have value
as a paradigm for statutory protection of
foreign intelligence information.
The Study recommends, after reviewing legislative pro-
posals for establishing a statutory classification system, as
well as proposals for codifying the Espionage and related
statutes and for strengthening the sources-and-methods
Approved Re1e2QQ1,/Q9/04I lqtAE8HR6MQP0300010001-6
 Approved For Release 2001/09/04: CIA-RDP80-00630A000300010001-6
ADMINISTRATIVE - INTERNAL USE ONLY
provision, that the Agency sponsor legislation amending the
sources-and-methods provision, defining it, and establishing
a new category of legally protected information, overlapping
with, but independent of, the classification system.
- It recommends finally that the DCI, acting
within the NFIB structure, create a committee
of legal experts to study the revision of the
Espionage Statutes.
Approv P I s1eI A1/0~/ RC A-RDP80-OO0630A000300010001-6 USE
 Approved For Release 2001/09/04: CIA-RDP80-00630A000300010001-6
DEVELOPMENT OF THE CLASSIFICATION SYSTEM
Information has been protected in one way or another from
the very beginning of this country. The Constitutional Con-
vention of 1787 conducted its proceedings in secret, and not
until 1920 were the records of the Convention made public.
The 1775 Articles of War, and legislation since 1776, have
prohibited soldiers from corresponding with the enemy and
civilians from spying in wartime.
During the Civil War, the government suspended habeas
corpus, made numerous political arrests, and censored the
press. After the Civil War, there was a gradual recognition
of the importance of protecting military installations and
the information concerning them from the prying eyes of
foreign intelligence. Up to World War I, there was only
one "classification"--Confidential--used to designate certain
types of defense information requiring protection. But,
unlike its modern namesake, the word "Confidential" had no
prescribed meaning in terms of the contents of a document,
but was rather a means of limiting the distribution to
specified addressees or categories of addressees. Nor was
it generally used to stamp documents.
Influenced by British and French example, the American
Expeditionary Force in 1917 adopted multiple levels of clas-
sification--Secret, Confidential, For Official Circulation
1
Approved For Release 2001/09/04: CIA-RDP80-00630A000300010001-6
 Approved For Release 2001/09/04: CIA-RDP80-0063OA000300010001-6
Only--which were soon incorporated in War Department Regulations.
But the use criteria for these document markings were based on
a vague determination that a given piece of information was
"more or less secret" and on a prescribed narrowing of the
addressees according to the determined level of sensitivity.
Though some see in this World War I experience the beginnings
of the present classification system, its lineaments are much
clearer in the 1936 revision of Army Regulations, No. 330-5,
where, for the first time, the levels of classification were
defined in terms of damage to national security, and the
protection of the markings was extended to non-defense
information affecting national security. The category,
Restricted, however, which had now replaced Official Use
Only, was really not a national security marking but rather
a means of ensuring administrative privacy.
Although there are no close parallels to the present
classification system in the pre-World War I period, there
are nevertheless archetypes of several closely related
security practices still in vogue today. Thus, General
Orders, No. 3 of the War Department, 16 February 1912,
prescribed these measures for safekeeping military records
concerning seacoast defenses: They were to be "classed" as
Confidential, kept under lock, given serial numbers, logged,
subject to copy control, shown only to "trusted employees"
according to the "exigencies of service." War Department
Approved For Release 2001/09/04: CIA-RDP80-0063OA000300010001-6
 Approved For Release 2001/09/04: CIA-RDP80-00630A000300010001-6
Regulations of 1913 ordered double wrapping of telegrams and
other communications, with Confidential stamped on the inner
envelope only. The Chief of Artillery complained bitterly
in 1907 about the indiscriminate use of Confidential on Army
issuances, citing in particular one that contained the formula
for making whitewash. (This is perhaps the first documented
instance of "unnecessary" classification.)
While the military were protecting their secrets by means
of classification, the State Department, up to 1958, was still
relying on the "housekeeping" statute of 1789 to protect dip-
lomatic secrets. The first Executive order dealing with clas-
sification, President Roosevelt's E.O. 8381 of 1940, was
essentially a wartime measure and had no lasting influence
on classification. President Truman's E.O. 10290 authorized
any government agency to classify "official information the
safeguarding of which is necessary in the interest of national
security and which is classified for such purposes by appropriate
classifying authority." Because of the vagueness of the
"national security" standard, Truman's order was immediately
attacked as an infringement of the First Amendment.
It was superseded two years later by President Eisenhower's
E.O. 10501, issued 5 November 1953, which lasted almost twenty
years until replaced by the present system. It redefined
Confidential, Secret, and Top Secret, reduced the number of
Approved For Release 2001/09/04: CIA-RDP80-00630A000300010001-6
 Approved For Release 2001/09/04: CIA-RDP80-00630A000300010001-6
agencies authorized to classify, and set a twelve-year limit,
subject to some exceptions, on maintenance of information in
the Top Secret category. President Kennedy in 1961 (E.O. 10964)
added provisions for automatic downgrading and declassification.
President Nixon ushered in the present system on 8 March 1972
with E.O. 11652 entitled, "Classification and Declassification
of National Security Information and Material."
A National Security Council Directive implementing this
Order was issued on 17 May 1972. The Order itself, which took
effect on 1 June 1972, refined the criteria for the three
levels of classification, replaced the "national defense"
formulation of E.O. 10501 with that of "national security,"
a collective term for "national defense" and "foreign relations,"
established a general declassification schedule with four
exemptions, and created an Interagency Classification Review
Committee (ICRC) to assist the National Security Council
(NSC) in monitoring the system.
Criticism of the Classification System
It is hardly an exaggeration to say that the classifica-
tion systems created by Executive order have been under almost
continuous attack by Congressional and other bodies since
their inception. The chief critic has been the House Sub-
committee on Information, established in 1955, and joined
Approved For Release 2001/09/04: CIA-RDP80-00630A000300010001-6
 Approved For Release 2001/09/04: CIA-RDP80-0063OA000300010001-6
later in that role by the Senate Subcommittee on Intergovern-
mental Relations. The Freedom of Information and Privacy Acts
were by-products of this close monitoring of government secrecy.
A principal target of the House subcommittee in its early years
was the functioning of the classification system within the
Department of Defense. Reacting to Congressional pressure,
the Pentagon established the Coolidge Committee to conduct
an in-house investigation. This committee, and others since,
directed attention to shortcomings that are still evident in
classification practices today.
In its report, the Coolidge Committee concluded that the
two major shortcomings of the system at that time were over-
classification and deliberate unauthorized disclosure. The
latter was a reference to the spate of disclosures prompted
by interservice rivalry. The report found a tendency on the
part of Pentagon officials to "play-it-safe" and overclassify;
abuse of security in the classification of administrative
matters; attempts to classify the unclassifiable; confusion
stemming from classification procedures that were based on
shifting foreign policy; and a failure to declassify material
that was no longer secret.
Reflecting the security concerns of the McCarthy era,
the Congress in 1955 established a Commission on Government
Security (known as the Wright Commission) that included
Approved For Release 2001/09/04: CIA-RDP80-0063OA000300010001-6
 Approved For Release 2001/09/04: CIA-RDP80-0063OA000300010001-6
classification among the subjects it investigated. The section
of the Commission's 807-page report, issued in 1957, dealing
with classification, recommended "that the Confidential clas-
sification be abolished. The Commission is convinced that
retention of this classification serves no useful purpose
which could not be covered by the Top Secret or Secret clas-
5
sification." It based this recommendation on a finding that
overuse of Confidential was impeding the free exchange of
scientific and technical information and thus hampering
national security. (As will be seen in a later section,
the abuse of Confidential then was similar to the abuse of
the Secret label now.)
The Commission urged also the enactment of legislation
making it a crime to release to an unauthorized person infor-
mation classified Secret or Top Secret, knowing it to have
been so classified, for any reason whatsoever. The bill was
aimed at persons outside the government, such as newsmen,
6
and aroused a storm of protest. Noting that there were
1.5 million employees in the government authorized to
classify documents, the Commission also recommended a major
reduction in the number of classifiers.
Meanwhile, the House Subcommittee on Information (known
as the Moss Committee after its first Chairman, Representative
John E. Moss), having finished its investigation, delivered
Approved For Release 2001/09/04: CIA-RDP80-0063OA000300010001-6
 Approved For Release 2001/09/04: CIA-RDP80-00630A000300010001-6
its report in June 1958. The following excerpt is indicative
of the tenor of the report:
In a conflict between the right to know and
the need to protect true military secrets from a
potential enemy, there can be no valid argument
against secrecy. The right to know has suffered,
however, in the confusion over the demarcation
between secrecy for true security reasons and
secrecy for "policy" reasons. The proper
imposition of secrecy in some situations is
a matter of judgment. Although an official
faces disciplinary action for the failure to
classify information which should be secret,
no instance has been found of an official being
disciplined for classifying material which should
have been made public. The tendency to "play it.
safe" and use the secrecy tamp, has, therefore,
been virtually inevitable.
The committee's recommendations, many subsequently embodied in
the present E.O. 11652, called for a reduction in the number
of classifying agencies and individual classifiers, an accelera-
tion of the process of downgrading and declassifying, and
8
establishment of an appeals procedure against misclassification.
The final days of E.O. 10501 were characterized by spec-
tacular leaks and acute administrative difficulties. In
June 1971, a Top Secret study on Vietnam, the "Pentagon
Papers," was published with impunity. This was undoubtedly
the most flagrant and brazen violation of security in the
history of the American classification system. That much
of it was overclassified and that one would be hard put to
prove that the disclosure caused "exceptionally grave damage
to the nation" (to measure it against the then criterion for
Approved For Release 2001/09/04: CIA-RDP80-00630A000300010001-6
 Approved For Release 2001/09/04: CIA-RDP80-00630A000300010001-6
Top Secret) is beside the point. In December 1971, there was
a leak in connection with the India-Pakistan war; in early
1972, a National Security Council memorandum revealing policy
conflicts over Vietnam war strategy appeared. All three
instances exemplify unauthorized disclosure motivated by
opposition to government policy. Their counterpart, and
perhaps their inspiration, was the executive branch practice
of selective disclosure and premeditated leaking to promote
government policy.
On the administrative side, Congressman William S.
Moorhead, the new Chairman of the Subcommittee on Government
Operations and Government Information, referred to "mountains
of classified documents" that had accumulated, contending that
"the inevitable was finally recognized in early 1971 as massive
overclassification and other abuses of the system created a
classification crisis and the virtual breakdown of our system."
Although one can doubt that the situation was as cata-
clysmic as this, it was sufficiently serious to prompt
President Nixon to appoint, on 15 June 1971, an interagency
committee charged with making a searching review of the
classification system. The committee was headed by then
Assistant Attorney General, William H. Rehnquist. After
more than a year of study, the committee, early in 1972,
presented a draft issued shortly thereafter as Executive
Order 11652.
8
Approved For Release 2001/09/04: CIA-RDP80-00630A000300010001-6
 Approved For Release 2001/09/04: CIA-RDP80-00630A000300010001-6
Since the issuance of E.O. 11652, Congressional criticism
10
has made such points as the following: (These points have
figured prominently in the Congressional hearings on clas-
sification and, as we shall see further on, some of them are
reflected in proposed legislation. The Freedom of Information
and Privacy Acts, also discussed below, embody some of the
thinking implicit in these criticisms.)
- "It confuses the legal meaning of the term
'national defense' and 'national security'
and the terms 'foreign policy' and 'foreign
relations' while failing to provide an
adequate definition of these terms." (The
reference is to the use of the term "national
security" in E.O. 11652 in place of "national
defense" used in E.O. 10501 and similar wording
in Exemption (b)(1) of the Freedom of Informa-
tion Act: "matters under Executive order
required to be kept secret in the interest of
national defense or foreign policy.")
- There are no specific penalties for over-
classification and unnecessary classification.
- It permits defense and foreign policy errors
to be concealed for at least three 4-year
Presidential terms, and up to 30 years or
longer under the exemptions.
- The ICRC is not accountable to Congress.
Section 9 legitimizes and broadens authority
for the use of special categories of "clas-
sification governing access and distribution
Of classified information beyond the three
specified categories." (This refers to the
authorization of supplementary protection
such as codeword compartments and document
control markings.)
- An independent regulatory body, with admin-
istrative, enforcement, and adjudicatory
Approved For Release 2001/09/04: CIA-RDP80-00630A000300010001-6
 Approved For Release 2001/09/04: CIA-RDP80-00630A000300010001-6
powers, should be established in the executive
branch to police all aspects of the classifica-
tion system.
- A statutory classification system should replace
the practice of Executive orders governing
classification.
Congressional critics of the classification system have
paid particular attention to the number of authorized clas-
sifiers. Seeing a causal connection between the expanding
volume of classified information and the number of classi-
fiers, they have demanded sharp reductions in the number of
the latter. Thus, the executive branch was induced to reduce
the number of classifiers from 1.5 million in 1957 to 55,000
in 1971, the year prior to the issuance of E.O. 11652. The
55,000 figure consisted of 2,849 Top Secret classifiers,
11
18,029 Secret, and 34,122 Confidential. (The CIA figures
for classifiers, given below, are the reverse of these, with
the Confidential category being the smallest and the Secret
the largest.)
But there is little evidence of a corresponding decline
in the birth rate of classified paper.* Actually, no one
knows the exact number of classified documents in the execu-
tive branch, not even the number of Top Secret ones. But
the order of magnitude is ascertainable. Hearings of the
*We note, however, that the ICRC Progress Report for 1975
claims a 2 percent reduction in the number of classification
actions in 1974 and 6 percent in 1975.
Approved For Release 2001/09/04: CIA-RDP80-00630A000300010001-6
 Approved For Release 2001/09/04: CIA-RDP80-0063OA000300010001-6
House Subcommittee on Information in 1971 and 1972 provide some
bench marks: William G. Florence, a retired Air Force clas-
sification official, estimated that the Defense Department had
at least 20 million classified documents and was of the opinion
that 99 percent of them did not merit classification in terms
of a strict interpretation of the Executive order. A State
Department witness gave an estimate of 35 million classified
documents for his department. And Dr. James B. Rhoads, the
United States Archivist, said he was responsible for 470 million
pages of classified material covering the period 1939-1954.
It would probably be safe to say that total classified holdings
of the executive branch at this time have long since passed
the billion mark.
There are grounds, therefore, for questioning the assump-
tion that limiting the number of classifiers necessarily
restrains the accumulation of classified paper. As indicated
later in the CIA context, strong bureaucratic forces tend to
undercut this assumption.
Approved For Release 2001/09/04: CIA-RDP80-0063OA000300010001-6
 Approved For Release 2001/09/04: CIA-RDP80-0063OA000300010001-6
CONFIDENTIAL
THE FUNCTIONING OF THE CLASSIFICATION SYSTEM WITHIN CIA
Our retrospective glance at the classification systems
has shown the persistence of the following negative, seemingly
ineradicable, characteristics: overclassification, unneces-
sary classification, vagueness of classification criteria,
unauthorized disclosure, and the relentless accumulation of
vast quantities of classified paper. Unfortunately, these
traits are all relevant to the classification experience of
CIA. In discussing classification within CIA,* the following
questions are addressed:
- What is the incidence of abuses of classification,
and the reasons therefor?
- How do the mandatory review provisions of E.O.
11652 affect the Agency?
- What is the impact of the Freedom of Information
and Privacy Acts on the Agency and on the func-
tioning of classification within the Agency?
To what extent, if any, has the Agency failed to
comply with the Executive order on classification?
- What is the effect of classification on the Agency?
The problem of unauthorized disclosure is considered from two
perspectives:
Faced with the problem of protecting foreign
intelligence information and preventing
unauthorized disclosure, what judicially
enforceable sanctions are available to the
Agency?
*The CIA regulation implementing E.O. 11652 is
revised 11 February 1975.
Approved For Release 2000/M EQI RPP80-00630A000300010001-6
 Approved For Release ~OC'lg$/hQN pIA-RDP80-00630A000300010001-6
- What can be done within the Agency to improve
the classification system, its management, and
the protection of foreign intelligence informa-
tion from unauthorized disclosure?
Misclassification
As of 31 December 1975, the CIA had 1,975 original clas-
sifiers, broken down as follows: 553 with Top Secret authority,
1,358 with Secret, and 64 with Confidential. The fact that
the wielders of the Secret stamp are clearly in greatest demand
suggests a positive correlation between the number of Secret
classifiers (which includes also the Top Secret ones) and the
overwhelming number of collateral documents bearing that clas-
sification. Up until about two years ago when the Director
reacted against the wholesale abuse of the Secret classifica-
tion, the use of the Confidential classification was relatively
rare in the Agency. There were offices at that time where one
could not find a Confidential stamp. It had become a reflex
action to place Secret upon a piece of paper. If the officer
forgot to do it, the secretary did so automatically. Given
this automatism, which is far from being extinct, the total
number of authorized classifiers tended to have a minimal
effect on the total amount of classified paper produced.
The hierarchical bias of the Agency suggests that the 1,975
classifiers mentioned above are by and large busy supervisory
personnel who have little time to devote to the subtleties of
Approved For Release 109704 TM -RDP80-00630A000300010001-6
 Approved For Release 2001/09/04: CIA-RDP80-0063OA000300010001-6
CONFIDENTIAL
the classification system. They tend to accept the decisions
of the subordinate non-classifiers. The number of de facto
classifiers, therefore, far exceeds the number of authorized
classifiers.
The problem confronting the conscientious classifier,
once he has decided that a document should be classified, is
to determine, according to the Order, whether the unauthorized
disclosure of the information could reasonably be expected to
cause: "exceptionally grave damage to the national security"
(Top Secret), or "serious damage to the national security"
(Secret), or "damage to the national security" (Confidential).
National security is viewed as a hybrid of national defense
and foreign relations. Examples are given in the Order for
Top Secret and Secret, but none for Confidential. "Examples
of 'exceptionally grave damage' include armed hostilities
against the United States or its allies; disruption of foreign
relations vitally affecting the national security; the com-
promise of vital national defense plans or complex cryptologic
and communcations intelligence systems; the revelation of sen-
sitive intelligence operations; and the disclosure of scientific
or technological developments vital to national security." A
strict application of these norms would substantially reduce
the number of Top Secret documents, both collateral and com-
partmented. Persons interviewed have cited such egregious
examples of overclassification as the following: NSA's Top
Approved For Release 2001/09/04: CIA-RDP80-0063OA000300010001-6
CONFIDENTIAL
 Approved For Release 2001/09/04: CIA-RDP80-00630A000300010001-6
CONFIDENTIAL
Secret city map of Moscow; a request for a computer terminal
from one CIA component stamped Top Secret; the Top Secret
NSCID 6 dealing with COMINT which, according to an Agency
COMINT specialist, should at most be Secret.
For Secret level classification, the Order gives these
examples of "serious damage to the national security": "dis-
ruption of foreign relations significantly affecting the
national security; significant impairment of a program or
policy directly related to the national security; revelation
of significant military plans or intelligence operations; and
compromise of significant scientific or technological develop-
ments relating to national security." One of the most pro-
lific causes of abuse of the Secret classification is the
widespread use in the Agency of pre-printed forms reading:
"Secret When Filled In." This leads to such patent mis-
classification as Secret time and attendance cards, pay
roll slips, etc. for overt employees. Should the Agency
telephone book containing only the names of overt employees
be classified Secret? Would its disclosure do "serious damage"
to national security, or simply "damage," or no damage at all?
Or should it be simply labeled "Internal Use Only?"
Fitness reports, investigatory reports, the whole field
of administrative correspondence, is a fertile field for over-
classification and often unnecessary classification. In this
field, there is undoubtedly a great deal of confusion between
Approved For Release 2001/09/04: CIA-RDP80-00630A000300010001-6
CONFIDENTIAL
 Approved For Release 2001/09/04: CIA-RDP80-00630A000300010001-6
CONFIDENTIAL
the legitimate needs of privacy and the classification criteria
that relate exclusively to the protection of information whose
disclosure would cause some degree of damage to national security.
The Clandestine Service has a propensity for Secret that tends
to blot out the other classifications. Some progress has been
made, it is true, in downgrading most administrative matters to
Confidential, but in the operational domain the Secret classifica-
tion is so ingrained that the use of Confidential is still rela-
tively rare. The mere presence of a cryptonym or pseudonym is
often sufficient to dictate the use of the Secret stamp. Some
analysts in both the DDI and the DDS$T have reported that it
is generally the practice in their shops to publish at the
highest level on the grounds that their high-level customers
have all the clearances and that this gives greater status to
the report. In one instance it was standard practice to
include some codeword material so that the report could be
printed the same day in a quick seventh-floor press setup,
rather than sent to the Printing Shop, which would have
entailed a delay of a day or so. This catalogue can be
concluded by noting that the Confidential NSSM 229, estab-
lishing an ad hoc group to study classification procedures,
is itself an example of unnecessary classification.
The most imprecise of the classifications is Confidential.
The Order gives no explanation of what constitutes "damage" to
national security, the criterion for Confidential. Still, it
Approved For Release 2001/09/04: CIA-RDP80-00630A000300010001-6
CONFIDENTIAL
 Approved For Release X00TM n4 J RDP80-00630A000300010001-6
requires little acumen to question the appropriateness of the
Confidential classification on Credit Union statements of overt
employees. Or the use of the Confidential stamp whenever an
analyst expresses an opinion in writing. It is true that some
opinions emanating from an official source could cause damage
to our foreign relations, but it is highly unlikely that this
is routinely the case. One senior DDI officer has expressed
the view that most OCI non-codeword publications could almost
be written at the unclassified level.
But, the reader may ask, what harm is there in overclas-
sifying? The classifications all give protection, and that
is what we are seeking, is it not? Unfortunately, classifica-
tion abuses have side effects, both short-term and long-term,
which cause significant difficulties for the CIA. Before
describing these side effects, some other aspects of mis-
classification must be examined first.
Related Issues
Implicit in the description of the three levels of clas-
sification is the inherent vagueness of the classification
criteria in the Executive order and the consequent play this
offers to the subjectivity of the classifier. "The problem,"
according to Professor Stanley Futterman, "is that the clas-
sification system rests entirely on subjective decisions
13
about contingent political effect." In a section entitled
Approved For Release 11lq%piTFA -RDP80-00630A000300010001-6
 Approved For Release 2001/09/04: CIA-RDP80-00630A000300010001-6
CONFIDENTIAL
"Resolution of Doubts," the NSC implementation directive for
E.O. 11652 recognizes the casuistical problems that may con-
front the conscientious classifier: "If the classifier has
any substantial doubt as to which security classification
category is appropriate, or as to whether the material should
be classified at all, he should designate the less restrictive
treatment." As early as 1961, former Defense Secretary Robert
McNamara offered a similar admonition: "...I suggest that we
14
follow this principle: when in doubt, underclassify." There
is unfortunately very little evidence that either of these
precepts has ever had a significant following.
In addition to the problems posed by the criteria as such,
the casuistry of classification must also contend with certain
esoteric aspects of secrecy itself. These revolve largely
around information that is unclassified in itself but for various
reasons, rightly or wrongly, is treated as secret. A safe
combination, not secret in itself, merits that status because
15
of what it gives access to. A Department of Defense affi-
davit given the court in connection with the "Pentagon Papers"
case against the Washington Post stated: "It is sometimes
necessary to classify a document in which no single piece or
16
part is itself classified." This is a valid point if the
organization of the material, the mosaic effect superimposed
on the unclassified pieces, meets the classification criteria.
Often, however, the argument is specious. This is particularly
Approved For Release 2001/09/04: CIA-RDP80-00630A000300010001-6
CONFIDENTIAL
 Approved For Release ? Mift0 f - ~At-RDP80-00630A000300010001-6
true of lists and compilations of unclassified items. E.O. 11652
contains this injunction: "Material containing references to
classified materials, which references do not reveal classified
information, shall not be classified." One may also be skeptical
of the contention of one component of the Clandestine Service
that the subjects of certain of its sensitive reports are not
only Top Secret, but so sensitive that they cannot be included
in the index of the Central Reference Service.
There is also secrecy by proximity or relationship, what
17
has been referred to as the "layered secret." A legitimate
secret is surrounded by successive layers of secrecy. One
must in these cases verify continually that the core secret
remains a secret and that the protective layers are necessary
and kept within reasonable limits. Some aspects of overhead
reconnaissance illustrate well both the good and bad features
of the layered secret.
There is the attempt to protect what really can't be
protected: the self-revealing "secret." The launching of
the Department of Defense Program 647 Early Warning and
Surveillance Satellite is a typical example. "No declas-
sification of information regarding the operation of that
important and highly expensive vehicle has been permitted,
even though news media and a Congressional committee have
reported publicly how it functions. The system is so exten-
sive in scope that it exposes itself."
Approved For Release 2001/09/04: CIA-RDP80-00630A000300010001-6
CONFIDENTIAL
 Approved For Release 2001/09/04: CIA-RDP80-00630A000300010001-6
CONFIDENTIAL
Finally, there is the "secret" which, through tacit agree-
ment between two countries, is designed to avoid confrontation.
The secret is based on political rather than security considera-
tions. The Soviet Union is aware of our overhead reconnaissance
program, just as it and the whole world is of our connection
with the Glomar Explorer. In neither case, however, does the
government officially acknowledge that connection. In the
latter cases, unlike the situation that prevailed before the
Soviets shot down the U-2 plane, the "secrets" are no longer
bilateral and related to national security, but universally
known "nonsecrets." The wisdom of the decision to avoid
official acknowledgement is not in question, but the for-
malistic use of the classification system to continue to
protect these "nonsecrets" is subversive of the system.
Although original classification for each level is in
principle limited to classifiers designated in writing for
that level by higher authority, there is no restriction on
the proliferation of classified material through carry-over
of classification from one document to another. To take an
extreme case, but not an unusual one, a classified item
representing perhaps one percent of an otherwise unclassified
document will suffice to confer its classification on the
whole document. Both the Order and the implementing directive
are rigid on this point. One can easily see the potential
for overclassification in this procedure. Typically, the
Approved For Release 2001/09/04: CIA-RDP80-00630A000300010001-6
CONFIDENTIAL
 Approved For Release F%1 49Lf FAA-RDP80-00630A000300010001-6
CIA analyst producing a piece of finished intelligence will
use information from several sources and, in the process,
often paraphrase or otherwise change the form of the original
information, thus effecting in many cases a de facto sanitiza-
tion of the original material. Yet he must carry over the
original classification and give to the document the highest
classification of the incorporated elements or go through the
time-consuming process of negotiating a change of classifica-
tion with the original classifiers. This problem is compounded
by the Agency's failure to comply with those provisions of the
Order and Directive requiring that, where practicable, docu-
ments be classified by paragraph. Thus a paragraph that
deserves no classification or at most a low one would have
to be taken over bearing the high overall classification of
the document from which it was extracted. Many concerned with
this problem have come to believe that a more rational approach
would be de novo classification of. finished intelligence products.
Reasons for Misclassification
The foregoing discussion has assumed a conscientious,
rational classifier contending with the arcana of classifica-
tion and the rigidities of the present system. Such a clas-
sifier is probably the exception. The persistent phenomena
of overclassification and unnecessary classification can
only be understood in terms of the classifier beset by
Approved For Release VdcJ) T -RDP80-00630A000300010001-6
 Approved For Release 2001/09/04: CIA-RDP80-00630A000300010001-6
ADMINISTRATIVE - INTERNAL USE ONLY
bureaucratic pressures and the powerful force of precedent.
Decisions made in the past tend to be perpetuated indefi-
nitely. How much easier it is to take over perfunctorily
the classification of a reference or of similar material
already published or in project files than to deliberate
each time on the degree of damage to national security.
There thus emerges what has been termed "inherited clas-
19
sification." This mindless process probably accounts
for a major portion of misclassified material.
There is also, on the part of some, an exaggerated
conception of sources and methods which tends to see almost
everything connected with the Agency as classified. The
elitism referred to above in connection with publishing for
a very select audience tends to counteract the pressure to
publish at the lowest possible level.
Why is it, one may ask, that these abuses of classifica-
tion are not corrected by the more perceptive in the Agency?
The Order enjoins that if a holder of classified information
believes "that there is unnecessary classification, that the
assigned classification is improper, or that the document is
subject to declassification under this order, he shall so
inform the originator who shall thereupon re-examine the
classification." But, in practice, this is never done.
Instead, there is a kind of vicious circle described in the
Approved fp,~Cgl l g ~~/~p9(04I 1 ~~8g-M6a$AOY00300010001-6
 Approved For Release 2001/09/04: CIA-RDP80-00630A000300010001-6
CONFIDENTIAL
illustrations above, with the supervisors too often relying
on their subordinates, and the subordinates on tradition and
precedent.
The basic reason, however, the reason that subsumes all
these considerations, is that the present system is tilted
in favor of overclassification. The calculus of risk for the
classifier in over- or unnecessary classification is less than
in underclassifying. "And since everyone believes that over-
protection has less serious results than under-protection,
overclassification is tolerated and shortfalls in judgment
20
are excused." There are, it is true, pro forma instances
of persons in the Agency who have been notified that their
actions were in violation of the terms of the Order and a
few who have received an "administrative reprimand" for
21
repeated abuse, but the overall effect has been slight.
Nor has the National Security Council, charged with moni-
toring the implementation of the Order, nor the Interagency
Classification Review Committee (ICRC), created to assist
the NSC in this task, been any more successful in rooting
out misclassification. Playing-it-safe has thus become
the rule of the game.
Anomalies in the Handling of Top Secret Material
There are approximately 129,500 collateral (non-codeword)
Top Secret documents in CIA, and the Operations Directorate
Approved For Release T:IA-RDP80-00630A000300010001-6
 Approved For Release 2001/09/04: CIA-RDP80-0063OA000300010001-6
CONFIDENTIAL
accounts for more than 90,000 of them. These figures relate
to documents other than cables. (For the latter there are no
reliable estimates since they are not subject to mandatory
logging and copy control.) In the neighborhood of 1100 Top
Secret collateral documents are generated by the Agency each
year. In addition, during 1975 CRS handled 444,870 codeword
disseminations, of which perhaps 20 percent, in any case
many times the number of Top Secret collateral cables, bore
the Top Secret classification.
It is anomalous that the CIA Top Secret Control Officer
has no responsibility whatsoever for Top Secret cables or
codeword material. Once a Top Secret cable leaves the Cable
Secretariat, it ceases to be the responsibility of the latter
and is not subject to centralized record keeping, as are Top
Secret collateral documents. Nor is there mandatory logging
or copy control of Top Secret material within the codeword
compartments. There is a computer program for keeping
track of Top Secret collateral material, but neither Top
Secret cables nor codeword material form part of the pro-
gram. Unlike Top Secret collateral, where the recipient
is supposed to acknowledge by his signature having seen a
document, there is generally no record of those who see code-
word material. If the often-made contention be accepted as
true that Top Secret collateral is better protected than Top
Secret compartmented information, then we have a further
anomaly: a system designed to provide supplementary protection
24
Approved For Release 2001/09/04: CIA-RDP8O-0063OA000300010001-6
CONFIDENTIAL
 Approved For ReleaseCfflyfl TACiIA-RDP80-00630A000300010001-6
per Section 9 of the Order providing less rather than more.
How much protection do Top Secret collateral procedures
really provide? When followed to the letter, a considerable
degree. But there are indications of deficiency on several
fronts: First, it is not uncommon for the secretary of a
unit to sign the control sheet of the document, which is then
read anonymously by others in the unit. Copy control is often
subverted by the Xerox machine. This was particularly true
during the Congressional investigations when a considerable
number of unnumbered copies of Top Secret documents were
turned over to the committee investigators, before the prac-
tice was ended. Annual inventorying of Top Secret collateral
documents has, in the past, been somewhat erratic. Even at
present there are more than 16,000 Top Secret documents in
the Agency which can't be located.
This discussion would be incomplete without passing
reference to the pseudo-classification, "Top Secret Sensitive."
Totally devoid of any authoritative basis, its only rationale
is convenience. Circumventing the codeword compartments, it
brings together in one place all-source information in a
collateral format. The uses that have come to the attention
of this Study are the following: The daily cable summary for
the DCI (seen by nine persons in the Agency); the Intelligence
Checklest (until recently sent to five committees of Congress);
and use as a classification on codeword material released to
Approved For Release 2001/09/04: CIA-RDP80-0063OA000300010001-6
 Approved For Release 2001/09/04: CIA-RDP80-00630A000300010001-6
CONFIDENTIAL
foreign chiefs of state (when authorized by the DCI on a
highly selective basis). Its proponents argue that it is
the only feasible means, in these and similar situations,
of transcending, in timely fashion, the complexity of cumula-
tive codewords and dissemination controls.
Exemption, Downgrading, and Declassification
In a totalitarian society, downgrading and declassifica-
tion of information have little or no meaning. It is only
when the ideal of an informed citizenry is accepted and acted
upon that these procedures have any chance of catching the
attention of management. The Freedom of Information Act was
passed in 1966, but was relatively ineffective until teeth
were put into it in 1974. Since February 1975, the effective
date of the amended Act, the whole subject of classification,
particularly declassification, has become an important con-
cern of U.S. Government management. The new DCI emphasized
this point on 13 May 1976 in his Guiding Principles for the
Intelligence Community: "We have an obligation to provide
as much information as possible on an unclassified basis,
but without derogation of the necessity to protect sensitive
sources and methods and to protect information which truly
requires sensitive treatment." The Privacy Act (PA), which
became effective on 27 September 1975, has also had the
effect of refining the Agency's perception of classification.
Approved For Release 200~8?/pf - TP80-00630A000300010001-6
D ENT
 Approved f$h}T"A~PIVE 9fflINTERNAL USEI66RM00300010001-6
It is against this background that the conscientious clas-
sifier described above must approach another duty prescribed by
E.O. 11652. In addition to determining the degree of damage to
national security that disclosure would entail, he must gauge
how long the information will require this level of protection.
His options are these: to subscribe to the timetable of the
General Declassification Schedule (GDS), to specify a date or
an event earlier than the GDS for declassification, or, judging
that the information requires greater protection than that pro-
vided by the GDS, to exempt it from the operation of the GDS.
This he may do if the information falls into one of four
exemption categories specified in Section 5(B) of the Order
and providing he specifies the category and, unless impossible,
a date for automatic declassification. Exemption 5B(1) protects
confidential information provided by foreign governments or
international organizations; 5B(2) deals with information or
material "specifically covered by statute, or pertaining to
cryptography, or disclosing intelligence sources and methods";
exemption 5B(3), the broadest of the categories, protects
information "disclosing a plan, installation, project or
specific foreign relations matter the continuing protection
of which is essential to the national security"; exemption
5B(4) exempts information "the disclosure of which would
place a person in immediate jeopardy." (Hereafter for con-
venience we shall refer to the exemptions as El, E2, E3, and
Approved 6 1109104NNNP80 D6 PO00300010001-6
 Approved For Release 2001/09/04: CIA-RDP80-00630A000300010001-6
ADMINISTRATIVE - INTERNAL USE ONLY
E4.) The Order requires that "the use of the exemption authority
shall be kept to the absolute minimum consistent with national
security requirements..."
Despite the foregoing, downgrading and declassification,
except as triggered by external requests under FOIA, PA, or
under E.O. 11652 for classification review, are relatively rare
within the Agency. This is because almost the entire classified
product of CIA is routinely exempted, usually under El and/or
23
E2. The propriety of this wholesale use of the exemption
clause is difficult to assess directly, but it is at the very
least highly suspect. When few in the Agency can recall ever
having seen a classified piece of CIA paper without E.(1), or
E2 IMPDET on it, there are grounds for doubting that the use
of the exemption authority is being kept to an absolute minimum.
The truth is that the automatism that we found in the case of
classification is even stronger and more prevalent in this
domain. Although the Order stipulates that only a Top Secret
classifier may exempt, it is commonplace throughout the Agency
for secretaries to affix E2 IMPDET to just about every clas-
sified piece of paper they type. Even granting that most of
the paper generated by the Agency deserves to be exempted,
this still leaves a large amount that doesn't, that simply
falls victim to another play-it-safe routine. There is no
gainsaying the short-run advantages of this approach, though
Approved J Ye'9/04 !TA-#MfP6f 6i0O0300010001-6
 Approved For Release 2001/09/04: CIA-RDP80-00630A000300010001-6
ADMINISTRATIVE - INTERNAL USE ONLY
contrary to the Order, in terms of convenience, economy of time,
and fool-proof protection of information.
By invoking the exemption authority the classifier by-
passes the automatic downgrading timetable of the GDS (Top
Secret to Secret two full calendar years after origination,
Secret to Confidential two, Confidential to declassified six)
but, in the long run, the exempted material still comes
up against mandatory review for declassification. This may
occur at any time ten years after the date of the document's
origination, upon request of a member of the public or another
department. If the information no longer qualifies for exemp-
tion, it must be declassified. All classified information or
material 30 or more years old must be declassified also unless
the head of the originating department personally determines
in writing that the information requires continued protection
for a specified period. In both instances, the Directive
provides that, if the final departmental decision on declas-
sification is negative, a member of the public may appeal the
decision to the ICRC, and, in that case, "the burden of proof,"
the Directive adds, "is on the originating Department to show
that continued classification is warranted within the terms
of the Order."
The wholesale exemption of information, the failure to
conduct systematic reviews before retiring files to the Records
Center, and the failure to label the unclassified portions of
Approved' P ftI Re1e O'O /09/OBI:TCUNRDPP8JDOBBOA000300010001-6
 Approved For Release 2001/09/04: CIA-RDP80-00630A000300010001-6
ADMINISTRATIVE - INTERNAL USE ONLY
documents--these practices have merely served to postpone
decisions that must eventually be made on destruction, historical
preservation, and declassification. With thirty-year review these
decisions come home to roost, but magnified in number and com-
plexity. For several years now a task force has been at work
culling OSS documents that have passed their thirtieth year,
and next year the first CIA documents, dating from 1947, must
be reviewed. In the time frame--1947 through 1966--there are
13,579 cubic feet of records in the Agency Archives. Since
these records are by and large segregated chronologically,
the problem is not so much finding what has to be reviewed,
as perusing thirty years later a vast accumulation of often:
unfamiliar documents and making decisions that could more
easily have been made in the past. Without belaboring the
point, it is clear that the task of the reviewer of old
Agency files would have been considerably less complicated
and less time-consuming if the files had been thinned out
periodically by downgrading and declassification, if worth-
less documents had been destroyed, and if the remaining
documents of historical interest had been sectionally
classified.
Impact of Freedom of Information and Privacy Acts
The following figures for 1975 give some idea of the
scope of the Agency's FOIA/PA program (the figures also
Approved Fri R1 tta ItLOO1/M/04 :IQI RPfi0t63g4p9O300010001-6
 Approved For Release 2001/09/04: CIA-RDP80-0063OA000300010001-6
ADMINISTRATIVE - INTERNAL USE ONLY
include requests for information under E.O. 11652): 5,859
requests for information were processed; 100 persons worked
full time processing them; the more tangible costs totalled
24
$1,392,000. Since these are not full-year figures for
either FOIA or PA, the figures for 1976 will probably be
significantly higher. (The amended FOIA took effect in
February 1975 and the PA in September 1975. From February
1975 until September 1975 privacy-type requests were processed
under FOIA.)
Under the Privacy Act any American citizen or resident
alien may levy a request on CIA for information concerning
himself and, unless its disclosure would compromise sources
and methods or cause damage to national security, it must be
granted. On the other hand, anyone anywhere may request
information under the provisions of FOIA and, unless the
information falls in one of nine exempt categories, it must
be released. Six of these categories are particularly rele-
vant to CIA. Exemption (b)(1)--information properly classified
pursuant to an Executive order--in effect gives statutory
recognition to E.O. 11652,; (b)(3) exempts information
specifically exempted from disclosure by statute. The
other four, however, go considerably beyond the Order.
Thus, the following categories of information, whether clas-
sified or not, may be withheld: (b)(2)--personnel rules and
guidance;(b)(5)--records of deliberations; (b)(6)--personnel
Approve AdL ? R E01/09/101: CIA-RDP8E- 063OA000300010001-6
 Approved For Release 2001/09/04: CIA-RDP80-00630A000300010001-6
ADMINISTRATIVE - INTERNAL USE ONLY
and medical files; (b)(7)--investigatory files of five speci-
fied types. These exemptions are of interest not only in their
own right, but also because they provide protection for infor-
mation that, as we have seen before, has often been erroneously
classified under E.O. 11652. Seven of the exemptions (all
except (b)(1) and (b)(3)) are discretionary in nature, but
the Attorney General has ruled that information should be with-
held only when it is "necessary or desirable in the national
interest."
Under the terms of both laws, if information is withheld
and the requestor seeks legal redress, the district court may
examine the propriety of both the classification of the infor-
mation withheld and of the exemption invoked. Section 552 (a)
(4)(B) of the FOIA declares that the district court has juris-
diction "to enjoin the agency from withholding agency records
and to order the production of any agency records improperly
withheld from the complainant. In such a case the court shall
determine the matter de novo, and may examine the contents of
such agency records in camera to determine whether such records
or any part thereof shall be withheld under any of the exemp-
tions set forth in subsection(b) of this section, and the
burden is on the agency to sustain its action."
The following table is a breakdown of the action taken
by CIA on the 5,859 requests for information processed during
Approved FA Jog qQ1, ]p9/D4r 1A E8 6%gD0300010001-6
 Approvecf99 ",ML1/091%;I I RPPJR~08RQ~000300010001-6
25
1975. (The vast majority of these requests concerned, in
order, the operations Directorate, the Office of Security,
and the Office of Personnel.)
FOIA
PA
EO*
Total Final Responses during
CY1975:
5,479
196
184
a.
Granted in full
300
4
63
b.
Granted in part
428
3
88
c.
Denied in full
174
0
28
d.
No record available, etc.
4,577
189
68
*This includes requests from a member of the public or another
government department under the mandatory review provisions of
Section 5(C) of the Order.
The 886 requests granted in whole or in part involved
declassification of documents that had been classified up
through Top Secret and almost without exception exempt from
downgrading.
As of 30 September 1976, litigation stemming from requests
for information had produced these results at the district
court level: ten cases won (appeals pending on three of them),
none lost. Two dealt with the Agency's refusal to release
information concerning its budget; one with the Glomar Explorer;
one with the "Bay of Pigs"; one with Congressional documents
held by the CIA; and five with requests for information con-
tained in CIA personnel files. Commenting on these results
ADMINISTRATIVE - INTERNAL USE ONLY
Approved For Release 2001/09/04: CIA-RDP80-00630A000300010001-6
 Approved For Release 2001/09/04: CIA-RDP80-00630A000300010001-6
CONFIDENTIAL
and our experience to date, an officer who follows these pro-
grams closely remarked: "We've been forthcoming, but haven't
given any secrets away."
The statutory and judicial aspects of FOIA and the Privacy
Act place a premium on strict compliance with the classifica-
tion and exemption provisions of E.O. 11652 and on discernment
of overclassification in disputed cases. The leavening effect
of FOIA/PA should eventually have a remedial influence on the
abuses of classification discussed earlier. But for the
Intelligence Community as a whole, they have created a new
environment dominated by the interaction of three powerful
26
forces: Privacy, Publicity, Secrecy. Any reform of clas-
sification must reckon with these conflicting forces and
attempt to establish a dynamic equilibrium among them.
Scope of Noncompliance with E.O. 11652
This is a convenient point at which to sum up the extent
of Agency noncompliance with the Order and its implementing
Directive and, at the same time, to consider the extent to
26a
which compliance with the Order is feasible. Unless otherwise
indicated, the references below are to sections of the Order.
- Section 1: The failure to interpret strictly
tie definitions of Top Secret, Secret, and
Confidential is the root cause of over- and
unnecessary classification in the Agency.
It does not appear that Top Secret, espe-
cially in the codeword compartments, has been
used with "the utmost restraint."
Approved For Release 2001! : CIARDP80-00630A000300010001-6
 Approved For Rye"Fsg&q1TRL04 : CIA-RDP80-00630A000300010001-6
- Section 3: As we have seen, there is practically
no downgrading or declassifying until a member of
the public makes a request for information, or
until the obligatory review of thirty-year-old
documents occurs.
Section 4: This section lays down rules designed
to prevent abuses of classification. Contrary to
this section, a person is rarely held "accountable
for the propriety of the classification attributed
to him." Although the failure to enforce this
section has probably contributed to the spread of
over- and unnecessary classification, it has not
drawn in its wake, as far as we can judge, those
other abuses proscribed by this section: clas-
sifying information "in order to conceal ineffi-
ciency or administrative error, to prevent
embarrassment to a person or Department, to
restrain competition or independent initiative..."
We have seen that the Agency does not show "to
the extent practicable," which "portions" of
documents are classified and which are not
classified, as required by Section 4. The
Directive specifies: "Whenever a classified
document contains either more than one security
category or unclassified information, each
section, part or paragraph should be marked
to the extent practicable to show its clas-
sification category or that it is unclassified."
Although onerous, this requirement is hardly
infeasible. Other departments, especially
the Department of Defense, have substantially
complied with it.
Section 5: There is noncompliance on two counts:
wholesale exemption of information and failure
to declassify promptly exempted material no
longer requiring protection.
Section 6: Subsection (E): There is account-
ability for Top Secret collateral material, but
not uniformly for Top Secret codeword material;
35
CONFIDENTIAL
Approved For Release 2001/09/04: CIA-RDP80-00630A000300010001-6
 Approved For Release 2001/09/04: CIA-RDP80-00630A000300010001-6
CONFIDENTIAL
there is accountability generally in trans-
mitting material outside the Agency, but not
generally for Secret and Confidential within
components. Subsection (F): There is no
systematic review of classified material "no
longer needed in current working files or for 77
reference or record purposes" for destruction.
The counterpart of such a review is the iden-
tification of material that should be preserved
and, where appropriate, marked for eventual
declassification.
- Section 7(B)(3): Based on what we have seen
above, the training and orientation programs
for employees concerned with classified infor-
mation appear inadequate; periodic reorienta-
tion programs during employment are generally
nonexistent. (Formal Agency instruction on
classification consists of: 45-60 minutes
given to new employees by the Office of Security;
about 20 minutes of a one-hour lecture also
covering the FOIA and the Privacy Act, in the
Office of Training's Operation Records I course.)
- Section VII of Directive: This section requires
each department to undertake the establishment
of a "Data Index System" for Top Secret, Secret,
and Confidential material of sufficient historical
or other value for preservation (the categories to
be approved by the ICRC). The ICRC has accepted
the AEGIS system of the Central Reference Service
as compliance with this provision. AEGIS, however,
contains only disseminated intelligence; the whole
operational product of the Agency is excluded and,
although much of the latter may not be of historical
interest, this is certainly not true of material
dealing with covert action, and many other types
of operations.
There is evidence of considerable violation of Section
VI G.(4) of the Directive which prohibits the reproduction
of Top Secret information without the consent of the originating
office.
- Section 13 (Administrative and Judicial Action):
The pro forma enforcement of the penalties
Approved For Release 2000Q4JEptAgWP80-00630A000300010001-6
 Approved For ReleaseR1PB:I~l~-RDP80-00630A000300010001-6
prescribed by this section--notification and
administrative reprimand--has been noted above.
Effect of Classification on the Agency
In terms of the classification Directive and Order, there
are clearly shortcomings in the Agency's compliance. But, in
terms of the collection and production of intelligence, what
impact do our classification practices have? Do they sig-
nificantly undercut the effective performance of our mission?
One observer of the classification system has written that
there "is no threat to the Republic in the overclassification
28
of trivia." Does this dispose of the question? We have
seen that indirectly, through statutory requests for informa-
tion from members of the public, the effect on the Agency's
classification practices has been considerable; and we have
seen that henceforth review of Agency documents on their
thirtieth anniversary will be immensely complicated. But
what has been the effect on day-to-day operations? In an
organization where everyone is Top Secret-cleared, simple
classification, with the exception of Top Secret, restricts
only slightly the flow of paper. In terms of access and
storage, there is not much difference in the handling of
Secret and Confidential documents. And the relative paucity
of Top Secret collateral documents limits their quotidian
impact. In this restricted context overclassification is
probably not a "threat to the Republic."
Approved For Release o (*. 4 -RDP80-00630A000300010001-6
 Approved For Release 2001/09/04: CIA-RDP80-0063OA000300010001-6
CONFIDENTIAL
There is, however, another context in which the effects,
although subtle and elusive, are cumulatively more serious.
In the long run overclassification saps the defenses of clas-
sification. Bad secrets depreciate good secrets. Thus, many
within the Agency have come to believe that Secret alone has
lost much of its meaning and some, out of zeal for security,
buttress an item requiring only the protection of Secret with
supplementary controls and markings or even find a way of
insinuating it into one of the codeword compartments. This
escalating protection accorded the "real" secrets, tends to
drain the classification criteria of their prescribed meaning.
Simple classification has become the scaffolding for
labyrinthine systems of supplementary protection. Paradoxi-
cally, the cause is not just overclassification, but also the
permissive language of Section 9 of E.O. 11652 which, as we
saw above, sanctions additional protection. In this context
we think immediately of the codeword compartments, but there
are other techniques, such as document control markings and
the mini-compartments erected by the Operations Directorate
to protect intelligence sources and methods, which are highly
29
effective and, sometimes, more so than the former.
The automatic data processing system of the Central
Reference Service (CRS) recognizes some 1400 permutations
of document control markings used in the Intelligence
30
Community or by our allies. This, despite the fact that
Approved For Release 2001/09/04: CIA-RDP80-0063OA000300010001-6
CONFIDENTIAL
 Approved For Releast3R~ B Q IA-RDP80-00630A000300010001-6
DCID 1/7, dated 5 October 1975, prescribes only five restric-
tive markings: ORCON, NFIB Departments Only, NOCONTRACT,
PROPIN, and NOFORN. Of these, ORCON-"DISSEMINATION AND
EXTRACTION OF INFORMATION CONTROLLED BY ORIGINATOR"- is the
most restrictive and the most controversial. It impinges not
only on Agency customers who complain vociferously about it,
but also, to a lesser degree, on CIA producers of intelli-
gence. The markings in their totality have complicated
enormously the job of CRS. A CIA Inspector General team
concluded in a 1973 study that "the various elements of the
Agency, in observing or instituting the handling restrictions
affixed on intelligence documents, have compromised the
retrievability of a significant amount of paper by excluding
it from the Agency's central reference system. Such with-
holding has tended to assume a permanent character, without
reference to either the changing sensitivity of the particular
operation or the paper produced by it, or CRS's ability to
31
ensure the application of the most stringent controls."
Considerably more constricting are the DDO procedures
for protecting sensitive information. These include: "RYBAT"
for "highly sensitive operational correspondence," the
"Prescribed and Limited Distribution" (P and L) for "excep-
tionally sensitive material," and "Restricted Handling" (RH),
not covered by regulation but set forth in a book dispatch
Approved For Release 1Jc3 1Z A-RDP80-00630A000300010001-6
 Approved For Release 2001/09/04: CIA-RDP80-00630A000300010001-6
CONFIDENTIAL
for field stations, for the most sensitive material. The
distinctive feature of the latter is that incoming traffic
is not processed by the Cable Secretariat, but held by the
32
Signal Center for pickup by designated persons. Only a
small number of persons, sometimes not more than 20 in the
Agency whose names are entered on a bigot list, have access
to RH traffic on a sensitive operation. The objective is to
protect not only the identity of the source, but the very
existence of the operation. (There has been considerable
erosion in the effectiveness of the RYBAT procedure and some
would add in that of P and L also.) Up to this point RYBAT,
P and L, and RH concern primarily the DDO. But the intelli-
gence information generated by the operations these procedures
protect must be disseminated. This, in turn, gives rise to
an array of access-limiting techniques that include multiple
source descriptions, the "Particularly Sensitive Report" that
is handcarried to addressees and receipted for, quasi-com-
partments with codewords for authorized recipients.
"Exclusive For" reports can be as restrictive or more
restrictive than P and L or RH-generated disseminations.
The IG study mentioned above estimated that between 1,500
and 3,000 reports a year emanating from P and L and RH cases
fail to reach the CRS because of sensitivity. This poses a
serious retrieval problem, since this information is usually
not indexed in CRS either.
Approved For Release 20(C'4Q/QAEAF DP80-00630A000300010001-6
 Approved For Release 2001/09/04: CIA-RDP80-00630A000300010001-6
CONFIDENTIAL
Summing up, we come to the ironical conclusion that a
DDO dissemination stamped Secret and surrounded by some of
the more restrictive devices described above probably enjoys
greater protection than the average Top Secret collateral or
Top Secret compartmented document. In. contrast with simple
classification, these access-limiting and distribution-control
devices have a very palpable effect on the flow of information
within the Agency and on the production of intelligence. They
are outgrowths, at least in part, of the lack of flexibility
of the classification system and of the failure to use it
correctly. These DDO devices and the special handling ones
within the DDI and DDSFTT restrict information to small numbers
of officers, to certain classes of employees, even to certain
elements within Divisions or Offices. This is not to say that
they should not be used, but merely to emphasize that their
inevitable spotty application and the absence of formal
criteria for their use can and has led to situations in which
analysts and operators lack the information needed to do their
jobs.
Approved For ReleascC? 1I/ /MM 1J54A-RDP80-00630A000300010001-6
 Approved For Release 2001/09/04: CIA-RDP80-00630A000300010001-6
ADMINISTRATIVE - INTERNAL USE ONLY
THE PROBLEM OF DISCLOSURE
The Protection of Sources and Methods
The classification system applies to the executive branch
as a whole, but the protection of "sources and methods" applies
33
to CIA in a very special, almost unique way. It overlaps
with classification, but has an independent life; it is another
means of protecting foreign intelligence information. Its
statutory basis is Section 102(d)(3) of the National Security
Act of 1947: "And provided further, that the Director of
Central Intelligence shall be responsible for protecting
intelligence sources and methods from unauthorized disclosure;
Referring back to this latter provision, Section 6 of
the CIA Act of 1949 exempts the Agency from the provisions of
any other law which requires "the publication or disclosure
of the organization, functions, names, official titles,
salaries, or numbers of personnel employed by the Agency...
Note that it is the Director personally who is charged with
the responsibility of protecting sources and methods and that
there is no explicit grant of powers to be exercised in
carrying out this responsibility. Nor is there a definition
of the scope of "sources and methods." E.O. 11652 refers
twice to "sources and methods": E2 excludes information
"disclosing intelligence sources and methods" and Section 9
Approved Fw i$q aQOM9/04 f l 8Qjg 6 Qp0300010001-6
 Approve AdDWN eVT fa, f (91/091 NTERNALDUSE 08~jQff0300010001-6
authorizes supplementary protection for intelligence sources and
methods. The criteria for Top Secret and Secret mention respec-
tively "sensitive intelligence operations" and "intelligence
operations."
The origin of the "sources and methods" concept is some-
what nebulous. The earliest known references occur in military
planning papers dealing with the establishment of a "central
intelligence service." General William J. Donovan had sub-
mitted to the President in November 1944 recommendations for
a post-war intelligence service and the President had instructed
the Joint Chiefs of Staff to study them and to prepare a draft
Executive order for his signature. In a memorandum, dated
18 January 1945, the Joint Strategic Survey Committee, com-
menting on a proposed draft recommended: "With a view to
emphasizing the importance of protecting certain methods and
sources of obtaining information the following should be
added to paragraph 6 of the draft directive: 'In the inter-
pretation of this paragraph, the National Intelligence
Authority and the Central Intelligence Agency will be
responsible for fully protecting intelligence sources and
methods which, due to their nature, have a direct and highly
important bearing on military operations.'"
This wording was incorporated into the draft Executive
order that the Joint Chiefs of Staff sent to the President
Approve el 'bTr AW '2Ug1/69IDNT p 0O QA000300010001-6
 Approved For Release 2001/09/04: CIA-RDP80-00630A000300010001-6
ADMINISTRATIVE - INTERNAL USE ONLY
circa 18 September 1945. The pertinent portion of paragraph 7
of the draft reads: "As approved by the National Intelligence
Authority, the operations of the departmental intelligence
agencies shall be open to inspection by the Central Intelligence
Agency in connection with its planning function. In the inter-
pretation of this paragraph, the National Intelligence Authority
and the Central Intelligence Agency will be responsible for
fully protecting intelligence sources and methods which, due
to their nature, have a direct and highly important bearing
35
on military operations." There is circumstantial evidence
that this sources-and-methods formulation may have originated,
at least indirectly, with the Navy, in particular with the
Director of Naval Communications who expressed concern that
the availability of military communications intelligence to
the Central Intelligence Agency would be detrimental to
military operations and therefore recommended inclusion in
the draft directive of language permitting each department
or agency to withhold such information if it felt that dis-
closure "will be inimical to the functions of such department
36
or agency." The sources-and-methods obligation was apparently
the result of a compromise with those in the military demanding
discretionary authority to withhold sensitive information from
CIA. In any event, it is fairly clear that the wording was
designed to ensure that CIA adequately protected military
secrets.
Approved FkliARR1d 2.QfUQ9/04I=AR%MQJ6U,63 YO~O0300010001-6
 ApproveAcl ~ bQ1/09441 I-jRQ6ff-OQOR?OYA000300010001-6
"Sources and methods" were not mentioned in the CIA
sections of the draft bill sent to Congress by President
Truman in 1947. The White House felt that the CIA section
should be kept as short as possible to avoid controversy and
not jeopardize the main thrust of the bill which involved
the unification of the armed services. Congress felt other-
wise. The Central Intelligence Group accordingly submitted
to Congress its recommendations (originally sent to the White
House) containing the sources-and-methods language. As
incorporated in the 1947 Act, the latter is in the form of
a proviso, one of three provisos restricting the powers of
CIA. The other two provisos respectively deny to the Agency
police powers and, by authorizing departmental intelligence,
a monopoly in the intelligence field. Although the explicit
reference to military secrets found in the old NSSC version
is dropped, the contextual implication of the obligation to
protect sources and methods is almost that of a condition of
access to the intelligence of other departments.
However this may be, CIA legal thinking on the sources-
and-methods obligation has seen in it a significant grant
of implicit authority to the Director that goes beyond the
mere protection of classified information. "The Congress
use of the term 'sources and methods,"' writes the Assistant
General Counsel, "indicates its recognition of the existence
Approv@d*b T I2O1I0gV014'E RDJ 9-OO8 A000300010001-6
 Approved For Release 2001/09/04: CIA-RDP80-00630A000300010001-6
ADMINISTRATIVE - INTERNAL USE ONLY
of a special kind of data encompassing a great deal more than
38
what is usually termed 'classified intelligence information."'
And the CIA General Counsel in a letter to Senator Muskie, dated
13 August 1974, in connection with the Senate hearings on clas-
sification, declared: "...it is conceivable that certain intel-
ligence sources and methods information would require protection
under 403(d)(3) of Title 50 [United States Code designation of
Section 102(d)(3)] even though it would not also warrant clas-
sification under the Executive order. Information protected
under that subsection, whether or not classified, is not
subject to the mandatory disclosure provisions of the Freedom
of Information Act since that Act does not apply to matters
that are specifically exempt from disclosure by statute."
From these citations two thoughts emerge: that sources-
and-methods information is not synonymous with classified
intelligence information, and that it may even embrace infor-
mation not classifiable in terms of the Executive order. It
follows that sources-and-methods information has a specific
character distinguishable from substantive intelligence
information. One might define it as embracing:
a) information or material revealing or tending to
reveal the identity and association with CIA of
any person, group, organization, or governmental
entity, whether witting or unwitting, that pro-
vides foreign intelligence information or intel-
ligence-related services, as well as the iden-
tificiation and connection with CIA of any
intelligence-producing device; and
Approved F j a 0pj1A9/04 i , L80 63 0300010001-6
 Approved For Release 201/0940 $t D~g~-08H~A000300010001-6
ADMINISTRATIVE E
b) information or material revealing or tending
to reveal the means, techniques, and procedures
by which foreign intelligence is collected,
processed, and analyzed, including those used
to support and protect foreign intelligence
activities, to the extent that these means,
techniques, and procedures are subject to
countermeasures, or revelatory of intelli-
gence intentions and capabilities. It must
be broad enough to include all forms of clan-
destine activity, as well as scientific and
technical intelligence. And, of course, it
must include sources-and-methods information
furnished by foreign governments. Unlike
much other sensitive information, it is dif-
ficult to prescribe in advance the life span
of sources-and-methods information. 0
The question naturally occurs, How have the sources-and-
methods provisions fared in the courts? In the United States
v. Jarvinen, a 1952 case, the United States District Court
for the Western District, State of Washington, rejected the
argument that two CIA employees, acting on instructions from
the DCI under Section 102(d)(3), could refuse to testify in
court concerning an informant of the CIA
They were sentenced to two weeks in jail, but later received
a Presidential pardon. Because of the defective fact situa-
tion and the danger of creating an unfavorable precedent, the
Agency decided not to appeal the decision of the district
41
court.
On the other hand, in Heine v. Raus, an action filed in
1964, there was a clear vindication of the Director's role in
the protection of sources and methods. Confirming the decision
Approve~d4%i4iU4 '20'(l1/69/b4T$M DP> 0OISWA000300010001-6
 Approved For Release 2001/09/04: CIA-RDP80-00630A000300010001-6
ADMINISTRATIVE - INTERNAL USE ONLY
of the Maryland District Court, the Fourth Circuit Court of
Appeals said, "action here to protect the integrity of sources
42
of foreign intelligence was explicitly directed by Congress."
Sources and methods figured also in the Marchetti case.
Although the Supreme Court had refused the government an
injunction in the "Pentagon Papers" case, the U.S. District
Court for the Eastern District of Virginia in Alexandria issued
an injunction on 18 April 1972 enjoining Victor Marchetti from
public disclosure of any intelligence information, particularly
that relating to intelligence sources and methods, and requiring
him to submit his manuscript to the CIA for review before
releasing it "to any person or corporation." The Fourth
Circuit Court of Appeals restricted the injunction to "clas-
sified information" acquired by Marchetti during his employment
by CIA.
In its final position on the Marchetti manuscript the
Agency insisted on 168 deletions. The district court upheld
only 26 of them, but, on appeal, the Fourth Circuit Court
sustained the remaining 142 deletions and remanded the case
to the district court "for such further proceedings as might
be necessary." In his opinion of 7 February 1975 the chief
judge of the Fourth Circuit Court, Judge Haynsworth, took
note of the DCI's statutory responsibility to protect sources
and methods, but based his decision on the classified nature
of the information.
Approved F MWg 29/041 PL8~~R6~QAQP0300010001-6
 Approva~MF f "~ff 0l /09 f Q,E~J&FDUPS8h0-Q3$A000300010001-6
What was particularly noteworthy about the Marchetti case,
however, was the willingness of the courts, under certain narrowly
defined circumstances, to accept "prior restraint"--in this
instance, because of the contractual nature of the secrecy
43
agreement signed by Marchetti as a CIA employee.
Since December 1975 there have been at least seven cases
in which U.S. district courts have recognized the sources-and-
methods provisions of the 1947 and 1949 Acts as a statutory
basis, under exemption (b)(3) of the FOIA, for withholding
information. Moreover, in most cases the courts have accepted
testimony and affidavits rather than insisting on in camera
examination of documents. Reaffirming the legal force of the
sources-and-methods provisions, Judge William P. Gray of the
U.S. District Court for the Central District of California
declared in Stanley D. Backrack v. CIA, William Colby, a case
decided on 13 May 1976: "While there is a strong public
interest in the public disclosure of the functions of govern-
ment agencies, there is also a strong public interest in the
effective functioning of an intelligence service, which could
be greatly impaired by irresponsible disclosure." Through
the decisions of these district courts a series of precedents
is emerging which have already greatly enhanced the legal
stature of sources and methods as an independent means of
protecting intelligence information--at least in the context
of FOIA requests for information.
49
Approvaaffi RTi16I o1104614TE8W-k -633A000300010001-6
 Approved For Release 2001/09/04: CIA-RDP80-00630A000300010001-6
ADMINISTRATIVE - INTERNAL USE ONLY
The Agency's experience with the Marchetti case revealed
certain weaknesses from a judicial review aspect in generalized
appeals to Section 102(d)(3) as a means of preventing the dis-
closure of sources-and-methods information. To give greater
legal solidity to future use of this Section, the Office of
the General Counsel has drawn up a catalogue of sources and
methods, hopefully broad enough and specific enough to prove
convincingly in case of litigation that a disputed piece of
information falls clearly in a category previously designated
45
by the Director pursuant to his statutory authority. Com-
plementary to this is the draft DCID (1/19) entitled "Non-
disclosure Agreements for Intelligence Sources or Methods
Information." Paragraph 1 sets forth the policy: "All
members of the Executive Branch and its contractors given
access to information containing sources or methods of intel-
ligence shall, as a condition of obtaining access, sign an
agreement that they will not disclose that information to
persons not authorized to receive it." The agreement is to
make specific reference to Section 102(d)(3), and each pro-
tected document is to bear the marking: "Warning Notice:
Sensitive Intelligence Sources and Methods Involved." When
finally implemented, these two steps should go a long way
toward filling loopholes that judicial challenges might
otherwise have found. The shortcoming of both these steps,
ADMINISTRATIVE - INTERNAL USE ONLY
Approved For Release 2001/09/04: CIA-RDP80-00630A000300010001-6
 AffMd1,rlgtlease Fffi/RR/4 :11-~RE~0-00630A000300010001-6
however, is the absence of a definition of sources and methods.
A catalogue excludes what it fails to include. Undefined,
sources-and-methods information runs the risk of becoming as
jargonized and abused a concept as that of "national security."
Statutory and Constitutional Barriers to Disclosure
"American culture is a populistic culture. As
such, it seeks publicity as a good in itself.
Extremely suspicious of anything which smacks of
holding back, it appreciates publicity, not merely
as a curb on the arrogance of rulers, but as a con-
dition in which the members of society are brought
into a maximum of contact with each other. Favoring
the exposure of practically every aspect of life, it
is uneasy in the presence of those who appear to be
withholding something."46
It is against this ethos that the torrent of unauthorized
disclosures in the seventies must be viewed. These included
the "Pentagon Papers" in 1971; the Marchetti and Marks expose
in 1974; Agee's damaging book in 1975; and the Village Voice
publication of the Secret report of the House Select Committee
on Intelligence in 1976.
Besides the American penchant for publicity, disclosure
at any particular time may be triggered by such factors as a
disgruntled or disaffected member of the executive branch,
sharp cleavages in the body politic, a confrontation between
the Congress and the President, and secrets that have lost
47
their credibility. The "Pentagon Papers," which has been
referred to perversely as "citizen disclosure," belongs to
ADMINISTRATIVE - INTERNAL USE ONLY
Approved For Release 2001/09/04: CIA-RDP80-00630A000300010001-6
 Approved For Release 2001/09/04: CIA-RDP80-00630A000300010001-6
ADMINISTRATIVE - INTERNAL USE ONLY
the species designed to change government policy. As such,
it has a number of precedents in our early history, although
certainly not in scale or impact. In 1795 Senator Mason of
Virginia, feeling that the people had a right to know the
terms of a treaty that Washington had laid before the Senate
in secret session, sent a copy to the Philadelphia Aurora;
Senator Tappan of Ohio did the same thing in 1844 with a
treaty calling for the annexation of Texas which President
Tyler had presented to the Senate in secret session, sending
the text to the New York Evening Post; and in 1848 the New
York Herald Tribune published the Treaty of Guadaloupe Hidalgo
ending the Mexican war, while the Senate was debating it in
48
executive session. The other three cases of recent dis-
closure illustrate the tension between publicity and secrecy
at its tautest point, that is, as it relates to covert intel-
ligence activities. In three of the four cases an essential
ingredient was a disgruntled or disaffected employee or former
employee of the executive branch.
It is striking that the compromises of classified infor-
mation in the seventies have been overwhelmingly due to public
disclosure rather than espionage. The legal defenses of secrecy,1
like the Maginot Line, have been so singlemindedly directed
against espionage that they have been repeatedly outflanked
by public disclosure. The First Amendment to the Constitution,
in severely restricting the use of "prior restraint" against F
ADMINISTRATIVE - INTERNAL USE ONLY
Approved For Release 2001/09/04: CIA-RDP80-00630A000300010001-6
 Approv a -IF nfR01/0?Tf LRI O 8R 3PA000300010001-6
the press, has also been, it is true, conducive to disclosure.
In view of this situation, how much protection then do statutory
and other judicially enforceable principles provide for foreign
intelligence information?
Before there was a classification system or espionage
laws, the executive branch protected secrets by virtue of the
principle of "executive privilege." The authority for issuing
Executive orders on classification derived from the exercise
49
of this privilege. E.O. 11652 refers only indirectly to
the Espionage Statutes. The doctrine of executive privilege
is an unwritten, implicit power that is usually derived from
Article 1, Section 2 of the Constitution--the separation-of-
powers provision--and, as it relates to national security,
from the powers of the President as Commander in Chief and
as the principal representative of the State in the conduct
of foreign affairs. In the New York Times Company v. the
United States, in 1971, even while refusing to support the
government's position on the "Pentagon Papers," Justice Potter
Stewart gave a ringing affirmation of executive privilege:
"It is clear to me that it is the constitutional duty of the
executive--as a matter of sovereign prerogative and not as a
matter of law as the courts know law--through the promulgation
and enforcement of executive regulations, to protect the con-
fidentiality necessary to carry out its responsibilities in
50
the fields of international relations and national defense."
ApproAkM~VI akaiJYbol/0it$TtEPdALRt[P?O-Mb,'OA000300010001-6
 Approved For Release 2001/09/04: CIA-RDP80-00630A000300010001-6
ADMINISTRATIVE - INTERNAL USE ONLY
But the extravagant claims of executive privilege in
connection with the Watergate experience and their rejection
by the Congress and the courts have made incontrovertibly
clear that the executive has no absolute power to withhold
information for national security, or any other reasons, being
subject in the exercise of executive privilege to legislative
and judicial checks and balances. Executive privilege still
remains a valid doctrine, but the courts are more likely to
support the Executive in withholding valid state secrets than
in preventing their publication once they have escaped from
51
executive control. This is certainly borne out by recent
disclosure history.
One statutory barrier to disclosure--the sources-and-
methods provisions of the 1947 and 1949 Acts--has been dis-
cussed in the preceding section. Exemption (b)(1) of the
Freedom of Information Act gives statutory sanction to the
protection of information properly classified in accordance
with an Executive order.
A particularly operative statute is the Atomic Energy
Act of 1954, which offers this definition of Restricted
Data: "all data concerning (1) design, manufacture, or
utilization of atomic weapons; (2) the production of special
nuclear material; or (3) the use of special nuclear material
in the production of energy, but shall not include data
Approved Ubi.WNAVGI B9/04L?n PQ8g&Q6?PAQ00300010001-6
 Approved For SD-1 - T TIVE01/09/NT RNAL USED-B? A000300010001-6
ADMIN
declassified or removed from the Restricted Data category
pursuant to Section 142." As perhaps the most effective
security program in the government, Restricted Data merits
our attention. A unique feature of Restricted Data that
marks it off from national security information classified
under E.O. 11652 is that by law any information falling
within the definition above, no matter where it originates,
is automatically classified, or, to use the expression of
the Energy Research and Development Administration (ERDA)
52
Handbook, is "born classified." The open-ended nature
of the definition is brought within reasonable bounds by
Section 142 of the Atomic Energy Act which obliges the
Commissioners collectively to declassify all Restricted
Data that can be published without undue risk to the common
defense and security.
Sections 142(c) and 142(d) require the Department of
Defense and the AEC to determine jointly what Restricted
Data relates primarily to the military utilization of atomic
weapons; this information can then be "transclassified" and
protected as national security information. This decompart-
mented atomic information then is known as "Formerly Restricted
Data," and its passage to foreign countries, except by special
agreement, is precluded. Section 142(e) permits the trans-
classification of atomic energy information pertaining to
ADMINISTRATIVE - INTERNAL USE ONLY
Approved For Release 2001/09/04: CIA-RDP80-00630A000300010001-6
 Approved For Release 2001/09/04: CIA-RDP80-00630A000300010001-6
ADMINISTRATIVE - INTERNAL USE ONLY
other countries if the Commission and the Director of Central
Intelligence jointly determine that the information is neces-
sary for the intelligence process and can be adequately pro-
tected as national security information. This flexibility,
coupled with ERDA's strict compliance with the declassifica-
tion provisions of the Act, has contributed greatly to its
success as a security program.
Next to Restricted Data, cryptographic information has
probably been the category of classified information most
successfully protected by statute. It is protected under
the Espionage Statutes, which are codified in Sections 792-
799 of Title 18 of the United States Code. Section 798 deals
with cryptographic information. It criminalizes the publica-
tion or transmission to an unauthorized person of classified
information "(1) concerning the nature, preparation, or use
of any code, cipher, or cryptographic system of the United
States or any foreign government; or ... (4) obtained by the
processes of communications intelligence from the communica-
tions of any foreign government, knowing the same to have
been obtained by such processes..." It goes on to define
"communications intelligence" as "all procedures and methods
used in the interception of communications and the obtaining
of information from such communications by other than the
intended recipients." Section 798 specifically bans
ADMINISTRATIVE - INTERNAL USE ONLY
Approved For Release 2001/09/04: CIA-RDP80-00630A000300010001-6
 Approved FoIRelease 2 01/094E:RIRp 0-q 10A000300010001-6
publication of cryptologic information in unequivocal terms
and omits the "intent" criteria of culpability which weaken
the other sections. As it relates to the "procedures and
methods" of communications intelligence, this statute over-
laps to some extent with the sources-and-methods provision
of the 1947 Act.
Section 793 of the Espionage Statutes penalizes a series
of specified actions undertaken "for the purpose of obtaining
information respecting the national defense with intent or
reason to believe that the information is to be used to the
injury of the United States, or to the advantage of any
foreign nation." The underlining, which we have added, is
the guilt criterion common to this section and Section 794.
Subsection (a) of Section 794 punishes with death or imprison-
ment anyone who delivers, or attempts to deliver, to a foreign
person or government, information relating to the national
defense with the intent formulation underlined above. Sub-
section (b) imposes the same penalties on anyone who, in time
of war, "with intent that the same shall be communicated to
the enemy, collects, records, publishes, or communicates..."
information on troop movements, defense dispositions, etc.
This is the only place in the Espionage Statutes where the
"publishing" of defense information is specifically mentioned.
It is limited in its application to time of war and to com-
munications intended for the enemy. "If this intent requirement
ADMINISTRATIVE - INTERNAL USE ONLY
Approved For Release 2001/09/04: CIA-RDP80-00630A000300010001-6
 Approved For Release 2001/09/04: CIA-RDP80-00630A000300010001-6
ADMINISTRATIVE - INTERNAL USE ONLY
is read to mean conscious purpose--a construction suggested by
the absence of the "reason to believe" standard used in the
culpability formulation of 794(a)--then prosecution of normal
publication under Section 794(b) is a virtual impossibility." 54
Returning to Section 793 of the Espionage Statutes, there
is no definition of "intent," "reason to believe," "damage"
or "advantage" in the guilt criterion. Although 793 is adequate
to convict a person guilty of espionage, its applicability to
55
a person who publishes defense information is less clear.
Professor Benno C. Schmidt, Jr., an authority on the Espionage
Statutes, sums up his study of them as follows:
In my reading of the Espionage Statutes, publication
of defense information not animated by a purpose to
communicate to a foreign country is not prohibited,
except for the narrow range of cryptographic infor-
mation covered by Sections 952 and 798. This reading
admittedly makes heavy use of legislative history in
construing the culpability provisions of subsections
794(b), 793(a) and 793(b). My conclusion rests also
on the belief--perhaps speculation would be a better
word--that courts will refuse to apply Sections 793(d)
and (e) to acts preparatory to publication, either by
finding some very narrow reading that conforms the
provision to the pattern of the other Espionage
Statutes, or--preferably as it seems to me--by
striking the provisions from TTitle 18 on grounds
of vagueness and overbreadth.55a
It is noteworthy that the United States did not invoke the
Espionage Statutes against the New York Times in connection
with the publication of the "Pentagon Papers."
In addition to the defects described above, the Espionage
Statutes have two other major weaknesses when viewed in terms
Approved R eg gQlVQ9/D4Y 1 ?8 p6~ ( 00300010001-6
 ApprovRffrISTn$n;/D 1/09Jf,EKNA 9FA0-~W31QA000300010001-6
of the protection of intelligence information, and in particular,
the sources-and-methods information of CIA. First, the meaning
of "defense information" contained therein is probably not broad
enough to embrace the whole of sources-and-methods information
or much of foreign relations information. In Gorin v. the
United States, the Court declared: "In short, the phrase
'information connected with National Defense' as used in the
context of the Espionage Act, means broadly, secret or con-
fidential information which has its primary significance in
relation to the possible armed conflicts in which the nation
56
might be engaged." Second, proving in a court of law
"intent or reason to believe" that the information in question
was to be used to the "injury of the United States, or advantage
of a foreign nation" will often be more costly in terms of
security than the violation to be punished. Referring to
court decisions that the government must present proof of
these points to a jury, the CIA Assistant General Counsel
wrote: "These rulings have left the government in the posi-
tion of having to reveal in court the very information it is
trying to keep secret, or else not prosecute those who steal
information and use it to the injury of the nation. To
invoke the law's protection of the secret, the secret must
be told."
ADMINISTRATIVE - INTERNAL USE ONLY
Approved For Release 2001/09/04: CIA-RDP80-00630A000300010001-6
 Approved For Release 2001/09/04: CIA-RDP80-0063OA000300010001-6
ADMINISTRATIVE - INTERNAL USE ONLY
CONCLUSIONS AND RECOMMENDATIONS
General
Some positive steps should be taken to re-establish the
credibility of the classification system. This thought is
implicit in the Director's recently enunciated guiding prin-
ciples for the Intelligence Community: "Improvement of the
public perception of U.S. Intelligence will be given con-
tinuing attention.... Within the constraints of legitimate
security requirements, the Intelligence Community should
strive to better public understanding of our mission and of
our product....The Intelligence Community should bexas respon-
sive as possible to Congressional inquiries. Congressional
support is essential to sustain the effectiveness of the U.S.
intelligence effort..."
A prerequisite in re-establishing the credibility of the
classification system is the reduction of secrecy to an
absolute minimum. This would have the effect of upgrading
legitimate secrets and thereby better protecting them.
"Secrecy practices which were taken for granted before
Vietnam and Watergate need to be adjusted to the processes
59
of re-establishing faith in our institutions." Although
some may dismiss the reduction of secrecy as merely hortatory
and not very practical, it is fundamental to all reform of
Approved For ReleaseR2001/09/04INCIA-RDP80-D0630kv00300010001-6
 Approved MIN ReTleas j2001/091 Ff 2~P Q-Q0 A000300010001-6
the classification system. The means of achieving it are
strict construction of the classification criteria in
E.O. 11652, placing the burden of proof on the classifier,
and, in accordance with the NSC Directive, choosing the less
restrictive classification when in doubt.
In its long-standing defensive posture against the demands
of Congress and the public for its closely held secrets, the
executive branch has neglected to devote much attention to
the theory and practice of its own classification system.
Its sporadic examinations have been reactive to Congressional
pressure. As a result, the only serious ongoing study of the
classification system, although predominantly adversary in
nature, has been done by Congress in investigative hearings.
Recommendation 1
CIA should take a leading role in the government's
development of classification theory and reform of the clas-
sification system. This would promote greater Agency sensi-
tivity to the mood of Congress and the public. This recom-
mendation is made in full awareness that the National Security
Council and the Interagency Classification Review Committee
have primary responsibility in this area. The intimate con-
nection between the classification system and the protection
of foreign intelligence information justifies such an initia-
tive. Executive Order 11905 lists as a function of the DCI
Approve'&KJPl~gl LT tb6I /fl9T 4T:EC"DP8B-0D QA000300010001-6
 Approved For Release 2001/09/04: CIA-RDP80-00630A000300010001-6
ADMINISTRATIVE - INTERNAL USE ONLY
ensuring "the establishment, by the Intelligence Community,
of common security standards for managing and handling foreign
intelligence systems, information and products, and for
granting access thereto."
Suggestions for Reform of E.O. 11652
Classification Criteria. Any radical reform of the present
classification system would involve redefinition of the criteria
for each level of classification. The vagueness of the criteria
has been a primary source of system abuses. A study of the
definitions employed throughout its history shows a gradual
evolution toward greater specificity. (See Annex A.) The
NSSM 229 group studying the classification system has thus
far been unable to improve the current criteria for the three
levels. It is true, of course, that no conceivable recasting
of the criteria will ever totally remove the judgmental ele-
ment from the act of classification.
On the other hand, there is nothing sacrosanct about the
three present levels of classification. At one point there
was only one classification--Confidential--and, at other
times, there have been as many as four. Some have recommended
scrapping the three existing levels and substituting a single
60
new category, Secret Defense Data. Tinkering with the three
present levels, which have existed for some twenty-five years,
is more likely, however, to generate confusion than reform.
Approved For Release 20 1109/041: IA 6 86- 696i (600300010001-6
 Approve RRRM?RhQ1-"I`NTERN ALL USE ONLYA000300010001-6
In the first place, it would be costly, it would require giving
maximum protection to all information and conducting full
background investigations (now only required for Top Secret)
on all those granted access to the single level. And it
would have a disruptive effect qn agreements with our
61
Allies in which the three levels have been incorporated.
Recommendation 2
Revise the first sentence of section 1 of the Order as
follows:
a. "Official information or material which requires
protection against unauthorized disclosure in
the interest of the national defense, foreign
relations of the United States, or the protection
of foreign intelligence sources and methods
(hereinafter collectively termed "national
security") shall be classified in one of three
categories..." (The addition is indicated by
underlining.)
b. An attempt should be made to cite under the
criteria for each level of classification one
or two telling examples of what should not be
classified at that particular level, chosen
on the basis of frequent and flagrant
misclassification.
c. Under the criteria for Confidential, of what
Confidential is, examples should be provided.
Administrative Privacy. A frequent cause of over- and
unnecessary classification is the bona fide desire to protect
information falling generally in the category of "administra-
tive privacy." There is no question about the desirability
of protecting this type of information if the government is
to go about its business, but this information has nothing
Appro dYt 1PA-W'9901101/9#ElR~LRIN~0 BR~JDA000300010001-6
 Approved For Release 2001/09/04: CIA-RDP80-0063OA000300010001-6
ADMINISTRATIVE - INTERNAL USE ONLY
to do with national security as defined in E.O. 11652 and
should not be classified thereunder. Such information is
amply protected from disclosure by the exemption provisions
of FOIA and, in certain specific respects, by Sections 5
and 6 of the CIA Act of 1949.
Recommendation 3
a. Insert in Section 4 of E.O. 11652 dealing with
abuses of classification a warning against the classifica-
tion of information falling in the category of "administra-
tive privacy" and reference the pertinent provisions of the
FOIA and the Privacy Act.
b. In the interest of unformity and avoidance of
misclassification, consideration should be given to the
issuance of an Executive order or NSC directive dealing
with administrative privacy and prescribing a uniform
marking for this type of material (Internal Use Only,
Administrative Use Only, etc.) dissimilar to present
national security markings. It should reference the
relevant exemptions of the FOIA.
c. Insert a caveat in the first pargraph of Section 4
of the Order warning that no information, regardless of
national security considerations, may be classified if the
activity it protects contravenes the U.S. Criminal Code, a
statute or an Executive order.
Approved Fq 1M4WRAQVQ2/04 : f4* R~0-99~3W(P300010001-6
 Approve r ~Lr ?~~11091P4TLyNALDUSE 08~1(A000300010001-6
Monitoring the Effectiveness of the Classification System.
Section 7 of the Order provided for the establishment of
the ICRC to assist the National Security Council in monitoring
the classification system. The Committee consists of repre-
sentatives of the seven principal agencies, including CIA,
concerned with classification. It has a permanent staff of
eight persons. By and large the Committee has been ineffec-
tive in preventing abuses of classification and has not pro-
vided dynamic leadership to the executive branch in the field
of classification theory and practice. This is no fault of
the members, but rather of the part-time nature of the assign-
ment, of the heavy responsibilities each member continues to
exercise in his parent organization, and of the inability of
the members to act independently of the organizations they
62
represent. The NSC directive implementing E.O. 11652 imposed
on the classifying departments and agencies the onerous task of
forwarding to the ICRC five types of quarterly reports. What-
ever utility these reports may have, one annual report would
probably provide it. There is need for a strong independent
body in the executive branch to monitor the classification
system and the codeword compartments.
Recommendation 4
Amend Section 7 of the Order and Section IX of the
Directive as follows:
ApprovaRIA~ T IY( 01709I/6X~&fi R a- abA000300010001-6
 Approved For Release 2001/09/04: CIA-RDP80-00630A000300010001-6
ADMINISTRATIVE - INTERNAL USE ONLY
Abolish the ICRC and establish in its place a
quasi-judicial board with investigative powers.
Its membership would consist of senior officers
from the Departments of State, Defense, and
Justice, the Energy Research and Development
Administration, the Central Intelligence Agency,
the National Security Council Staff, and a
chairman designated by the President. The
members would serve for at least two years,
and would not represent their parent organiza-
tions. The Board, responsible only to the NSC,
would have cognizance in all matters relating
to security classification, and the FOIA and
the Privacy Act as they impinge on the clas-
sification system.
Special Departmental Arrangements. Section 9 of the
Order authorizes originating departments to impose special
requirements with regard to access, distribution, and pro-
tection of classified information. This is the charter for
the codeword compartments and the plethora of dissemination
and other controls that have proliferated in the Intelligence
Community. Where these arrangements have repercussions beyond
the originating department (as is generally the case), there
should be some form of central supervision and control over
departmental initiatives. Their inhibiting effect on the
flow of information is imcomparably greater than that of
simple classification.
Recommendation 5
Amend Section 9 of the Order and VI F of the Directive
to require the concurrence of the Director of Central
Intelligence whenever an NFIB member establishes a compartment
Approved For Release zv~~lI 9/04 MITAE- 80 3 X60300010001-6
 Approv pd o R r leaass 2001/09/04: CIA--RDU80 00630A000300010001-6
that inhibits the flow of information to other members of the
Intelligence Community.
Sanctions for Noncompliance with the Order
Historically, the principal shortcoming of the classifica-
tion systems has been the lack of credible, deterring sanctions
against abuses and unauthorized disclosure. This explains the
repetitive nature of the criticisms aimed at the various sys-
tems. Still, we are forced to the conclusion that, by and
large, the proximate cause of the failure of the sanctions
prescribed by the Order to deter abuses is not so much the
weakness of the penalties, as their sporadic imposition.
As far as overclassification is concerned, we have seen
that the system is tilted in favor of overclassification. If
management wishes to, however, it should not be more difficult
to make people classification-conscious than cost-conscious.
One Agency critic of the Order has charged that undue
importance is given to the number of persons authorized to
classify. This has had the effect of forcing busy, higher-
level people to make decisions that they do not have time to
make. They, therefore, accept uncritically the decisions of
subordinates. He sees in this a "forced pattern of non-com-
63
pliance." The rationale behind limiting the number of
classifiers was to reduce the amount of classified paper.
As noted, this has had little effect on the volume of
ADMINISTRATIVE - INTERNAL USE ONLY
Approved For Release 2001/09/04: CIA-RDP80-0063OA000300010001-6
 Approved For Release 2001/09/04: CIA-RDP80-00630A000300010001-6
ADMINISTRATIVE - INTERNAL USE ONLY
classification material within the Agency. Consideration
should perhaps be given to increasing selectively the number
of classifiers in certain components, but only as part of an
overall revamping of classification practices.
Recommendation 6
Add to Section 13(A) of the Order a requirement that
supervisors rate all classifying officers on their knowledge
and exercise of the classifying function. It is recommended
that CIA adopt this practice independently.
CIA Management of Classification
In the section dealing with the functioning of the clas-
sification system within CIA, certain abuses, anomalies, and
nonconformity with certain provisions of E.O. 11652 were noted.
The interaction of FOIA and the Privacy Act with the classifica-
tion system was also examined. Although no evidence was found
of the use of classification to conceal inefficiency or admin-
istrative error, various investigations of the Agency, internal
and external, have shown that classification served in some
instances to conceal illegality, often perceived as such only
retrospectively. There has emerged from the investigations
what might be termed the doctrine of the "valid secret." To
be valid, a secret, even if properly classified in accordance
with the national security criteria, must not be tainted by
ADMINISTRATIVE - INTERR~~AAII~
Approved For Release 2001/09/04: CIA-RDP80U0 63BW0300010001-6
 Approved For Release 2001/09/04: CIA-RDP80-00630A000300010001-6
ADMINISTRATIVE - INTERNAL USE ONLY
illegality, which has the effect of nullifying the classifica-
tion. As remedial steps in countering CIA noncompliance with
the Order, eight steps under the recommendation below are
offered.
Recommendation 7
To correct the abuses, to provide centralized direction
of classification and related matters, to monitor compliance
with E.O. 11652 in the future, and to develop a body of doc-
trine, it is recommended that a CIA Classification Board,
with the following features be established:
a. A five-member board of senior officers
directly responsible to the Deputy
Director of CIA. The members would
serve full-time for tours of at least
two years. Each Directorate would name
a member; the chairman would be appointed
by the DCI. The members would not repre-
sent their parent organizations, nor be
answerable to them for their decisions.
b. The Board would have investigative,
adjudicatory, and research functions
within the limits of its responsibilities.
c. The Board would replace the p6jsent CIA
Information Review Committee.
d. Besides monitoring compliance with E.O. 11652,
the Board would provide policy guidance to
the office dealing with FOIA and the Privacy
Act, to the Special Security Center, and to
the Systems Classification Branch in the
Office of Joint Computer Services. It
would monitor the operation of the codeword
compartments within CIA. Its concurrence
would be required on the establishment of
ADMINISTRATIVE - INTERNAL USE ONLY
Approved For Release 2001/09/04: CIA-RDP80-00630A000300010001-6
 Approved For Release 2001/09/04: CIA-RDP80-00630A000300010001-6
ADMINISTRATIVE - INTERNAL USE ONLY
any compartment, control, or marking affecting
the flow of information beyond the originating
Directorate.
e. The Board would prepare handbooks, and guides
on classification, declassification, sanitiza-
tion, and decompartmentation as these matters
fall within the province of CIA.
f. The Board would serve as a central clearing
office for information on formal and informal
compartmentation, providing guidance to
researchers and others on special clearances
required to gain access to information needed
to do their job. For this purpose, the Board
would be briefed on every compartment existing
within the Intelligence Community which is
relevant to the mission of CIA.
The Board would be provided with a small
permanent secretariat designed to provide
research and secretarial support.
Recommendation 8
a. A handbook dealing with not only the basics, but
also the more subtle aspects of classification, its inter-
relationships with FOIA and the Privacy Act, should be
prepared. The ERDA handbook on classification could serve
as a model.
b. There is need to establish training programs on
classification, FOIA, and the Privacy Act, especially for
original classifiers. A program of periodic reorientation
of employees on classification, as required by the Order,
should be instituted.
ADMINISTRATIVE - INTERNAL USE ONLY
Approved For Release 2001/09/04: CIA-RDP80-00630A000300010001-6
 Approved For Release 2001/09/04: CIA-RDP80-00630A000300010001-6
ADMINISTRATIVE - INTERNAL USE ONLY
Recommendation 9
a. A beginning should be made on compliance with that
section of the Order enjoining classification by portion or
paragraph, where feasible. This would facilitate declassifica-
tion and minimize overclassification in those cases where infor-
mation is extracted and incorporated in other reports. (It
should be noted that at the present time cabled intelligence
summaries prepared by OCI are classified by paragraph.)
b. In addition to paragraph marking, an attempt should
be made to devise guidelines that would permit de novo clas-
sification, under specified conditions, of information extracted
from raw reports for incorporation in finished intelligence.
Recommendation 10
Top Secret cables should be handled in the same manner
as other Top Secret collateral material once they leave the
Cable Secretariat and should be brought under the supervision
of the Top Secret Control Officer. The problem of better
control of Top Secret codeword material requires further study.
Recommendation 11
Pre-printed forms marked; Secret When Filled In, are
a prime cause of overclassification. They should be abolished
and replaced with forms that offer other classification options,
including Unclassified.
71
ADMINISTRATIVE - INTERNAL USE ONLY
Approved For Release 2001/09/04: CIA-RDP80-00630A000300010001-6
 Approved For Release 2001/09/04: CIA-RDP80-00630A000300010001-6
ADMINISTRATIVE - INTERNAL USE ONLY
Recommendation 12
There should be a thorough review, at least annually,
of the number and level of classifiers required by each com-
ponent. Both the tables of organization upon which the alloca-
tion of original classifiers is usually based and the lists of
authorized classifiers appear in need of updating. Selective
increases in the number of classifiers seem warranted in some
components. These increases could probably be offset by
removing from the lists those who exercise their classifica-
tion authority infrequently or not at all. To dispel con-
fusion in the minds of many officers, those authorized to
classify should be reminded of the level at which they are
authorized to classify at least annually and those whose
authorization has been terminated should be notified immediately.
Recommendation 13
A regulation should be issued on the handling of infor-
mation falling in the category of "administrative privacy."
It should reference the pertinent provisions of FOIA and the
Privacy Act. It should prescribe uniform marking and pro-
tection for this information, emphasizing that it should not
be classified under E.O. 11652 unless it has unmistakable
claims to this protection in terms of national security
and/or the protection of sources and methods. (Section 8f
of CIA Regulation dealing with dissemination controls
ADMINISTRATIVE - INTERNAL USE ONLY
Approved For Release 2001/09/04: CIA-RDP80-00630A000300010001-6
 Approv-.ri A?~ T RRI109 C1AARhD6pp-OM~~A000300010001-6
for foreign intelligence, lists as additional controls: CIA
Internal Use Only, Administrative-Internal Use Only, and For
Official Use Only. Administrative-Internal Use Only, which
comes closest to a marking for administrative privacy, is
defined as a control that may be used "for unclassified, non-
sensitive administrative information that should not be dis-
seminated outside of CIA." If this definition were broadened
to include the types of unclassified information exempted by
the FOIA and if protective measures were prescribed, this
control would satisfy the sense of this recommendation.)
Recommendation 14
To bring the current exponential proliferation of clas-
sified paper through Xeroxing under control, research should
be conducted into the feasibility of a system similar to the
following: Programming the Xerox copying machine so that it
will not function unless certain information is input into
it--the name or badge number of the employee, whether the
document to be copied is unclassified or draft and, if clas-
sified, the classification, subject, dispatch symbol, etc.
If classified, (the machine could be programmed to recognize
this from the marking), the machine could assign a control
number and provide a record of copies made. Something of
this sort should serve two purposes: deter Xeroxing and,
where necessary, provide an inventory of classified documents
that have been Xeroxed.
Approv6NVb~ '&1f09lOi4 ECHA-BD 1-@ A000300010001-6
 Approved For Release 2001/09/04: CIA-RDP80-00630A000300010001-6
ADMINISTRATIVE - INTERNAL USE ONLY
The Problem of Disclosure
This Study has examined the statutory and constitutional
barriers to disclosure--the Espionage Statutes, executive
privilege, the classification system, certain provisions of
FOIA and the Privacy Act, the sources-and-methods clause--and
found them wanting in important respects. Before suggesting
any remedies, it is first necessary to consider various pro-
posals that have been made for a statutory solution.
During the last Senate hearings on classification in
1974 there were six bills pending before Congress designed
65
to create a statutory classification system. The motiva-
tion behind these bills was not to give greater protection
to government secrets by providing legal underpinning but
rather, through various devices, to accelerate the declas-
sification of information and insure Congressional access to
executive branch secrets. One bill (S. 1520) would set up
a National Committee on Executive Secrecy; another (S. 3399)
would create a Classification Review Commission. In the
case of the latter, Congress would propose the names of six
of the nine members to the President for appointment. This
thinly veiled intent to share executive power is probably
unconstitutional. The proposed Classification Review
Commission would be empowered to overrule the President
on classification matters. As mentioned above, two bills
ADMINISTRATIVE - INTERNAL USE ONLY
Approved For Release 2001/09/04: CIA-RDP80-00630A000300010001-6
 Appro ,%$ ~VI I Tg 01/Oi1V'1'ERNALRO~EO(92~OA000300010001-6
(S. 1726 and S. 2451) would set up a single classification--
Secret Defense Data--that S. 2451 defines as information the
unauthorized disclosure of which "would adversely affect the
ability of the United States to protect itself against overt
or covert hostile action." The bills generally avoid the
"national security" formulation of E.O. 11652 in favor of
an undefined "national defense." None was found satisfactory
by the executive branch, and none contains features that would
be helpful in drafting a constructive statutory vehicle for
the classification system.
More deserving of our attention is Subchapter C, Chapter 11
of S.1, a Senate bill entitled "The Criminal Justice Reform
Act of 1975." This subchapter, a part of a monumental codifica-
tion of Federal criminal law, over twenty years in the making,
presents a revision of the Espionage and related statutes
discussed earler. Sections 1121-1123 deal with espionage
and the disclosing and mishandling of "national defense
information." Section 1124 penalizes the disclosure to an
unauthorized person of any type of classified national
defense information. The vague "intent or reason to believe"
wording of Sections 792-794 of the Espionage Statutes has
been replaced throughout Subchapter C by the following guilt
formulation: "knowing that defense information could be used
to the prejudice of the safety or interest of the United
States, or to the advantage of a foreign power."
Approv Moiio /lW bAIR60-0T3XbA000300010001-6
 Approved For Release 2001/09/04: CIA-RDP80-00630A000300010001-6
ADMINISTRATIVE - INTERNAL USE ONLY
A key element in the codification of these statutes is the
definition of "national defense information" in Section 1128(f).
The latter is essentially an enumeration of ten types of infor-
mation, the first five of which relate to military secrets.
Item (6) reads: "intelligence operations, activities, plans,
estimates, analyses, sources, or methods, of the United States."
The remaining four specify intelligence concerning a foreign
power, communications intelligence and cryptographic informa-
tion, atomic Restricted Data, and any matter involving the
security of the U.S. in time of war.
Section 1123 would indirectly permit the prosecution of
anyone deliberately publishing national defense information,
but the burden of proof would be upon the government to con-
vince the jury that the information was in fact national
defense information. Thus, we are faced again with the
dilemma encountered earlier, the requirement of proof that
would compound the damage already done to national security.
Although Section 1224 embraces "all" classified informa-
tion and applies to former as well as present government
employees, it specifically exempts from prosecution the
unauthorized person to whom national defense information
has been communicated. Since the unauthorized recipient
may well be a member of the press, Section 1124 is not an
effective deterrent to the publication of classified information.
ADMINISTRATIVE - INTERNAL USE ONLY
Approved For Release 2001/09/04: CIA-RDP80-00630A000300010001-6
 ApprovR~R f mI2 1 /09#9 l&-L I USE f j A000300010001-6
Moreover, by bringing Section 798 of the Espionage Statutes
(the section protecting communications intelligence, etc.,
and the most effective of the Espionage Statutes) under its
wing, Section 1124 vitiates 798.
In short, Subchapter C represents an advance in terms of
simplicity of language, procedure, and careful definition of
terms, but a step backwards in exempting from prosecution the
unauthorized recipient of classified information and in using
the concept, "national defense information," rather than that
of "national security information." Foreign relations infor-
mation, an essential element of the latter, is apparently
66
excluded from its purview.
Those who decry the lack of effective statutory deter-
rents to disclosure often look enviously at the British
Official Secrets Act. The latter, however, with its D-notices
and catchall language that protects any "official information"
whether related to national security or not, is not only
inconceivable politically in the context of American freedoms
but would undoubtedly be struck down by the courts as
67
Recommendation 15
The Agency should propose legislation along the following
lines to protect sources-and-methods information. The proposed
legislation should be patterned on the Atomic Energy Act of
1954. It should:
77
Appro,~JW kSkAig~b0?/OJM RE)PS01 9.6:90A000300010001-6
 Approved For Release 2001/09/04: CIA-RDP80-00630A000300010001-6
ADMINISTRATIVE - INTERNAL USE ONLY
a. amend section 102(d)(31 of the National Security
Act of 1947 for the purpose of defining sources
and methods and creating a new category of
legally protected information, overlapping wi~hh
but independent of the classification system. 8
Like Restricted Data, Foreign Intelligence Data
or Sources and Methods Information would be
"born classified."
b. reference exemption (b)(3) of the Freedom of
Information Act.
c. provide for an energetic program of removal
of nonsensitive or sanitized sources-and-
methods information from the special category
to the protection of simple classification,
or, if appropriate, declassification. These
procedures would usually apply to the product
of sources and methods and, only rarely, to
sources and methods per se. There is a partial
analogy to Formerly Restricted Data.
d. prescribe a full background investigation as a
condition of access to sources-and-methods
information.
e. make clear that the proposal is not directed
against Congress and that all appropriate
information will be made available to con-
cerned committees of Congress in secret
session. There has been no known leak of
Restricted Data by Congress and hopefully
the same would be true of sources-and-
methods information.
f. ban all controls, markings, or compartments,
other than the three levels of classification,
within the Foreign Intelligence Data category,
unless approved by the DCI. In general, sup-
plementary protection, other than through a
strict application of need-to-know, should be
unnecessary within the special category.
Where applicable, the three levels of classification of
E.O. 11652 should continue to be utilized within the special
category. By definition, however, sources-and-methods infor-
mation would be protected, whether classified or not.
78
Approved kbMMFE&tb OO1JO9f04INCMRW M 6Mb'000300010001-6
 Approved For Release 2001/9IW~j -&- 89 30A000300010001-6
ADMINISTRATIVE
Recommendation 16
The DCI, acting within the NFIB structure, should
designate an ad hoc committee of legal experts to study
the revision of the Espionage Statutes. In particular,
sections 792 through 796 should be consolidated, simpli-
fied, and clarified. The intent criteria should be
abolished and other loopholes plugged. As an alternative
to Recommendation 15, section 798 dealing with cryptographic
information could be expanded to include sources-and-methods
information, or a separate section could be added for this
purpose. (Because of its primary responsibility in this
area, the establishment of such a committee would, of course,
have to be closely coordinated with the Justice Department.)
ADMINISTRATIVE - INTERNAL USE ONLY
Approved For Release 2001/09/04: CIA-RDP80-00630A000300010001-6
 Approved For Release 2001/09/04: CIA-RDP80-00630A000300010001-6
CONFIDENTIAL
FOOTNOTES
1. The dictionary defines compartmentation as division
into separate sections or units, giving as an example, the
elaborate compartmentation of submarines wnich reduces the
danger of sinking. Within a government agency, and par-
ticularly within an intelligence organization, compartmenta-
tion takes on a special coloration. In this context, it may
be defined as the deliberate restriction of the free flow of
information for the purpose of minimizing the risk of unau-
thorized disclosure of designated types of information whose
protection is deemed essential to the organization.
One may picture compartmentation as a series of
rings starting with the badge that excludes those who are
not members of the organization, and working inward and
upward through the levels of document classification to
the ever smaller rings of the codeword compartments. If
the application of compartmentation were absolute in the
sense of the submarine analogy, the organization would
suffocate. What enables the organization to survive is
the operation of the countervailing principle of "need-to-
know" permitting a directed flow of information.
Interacting with deliberate compartmentation are
various bureaucratic practices that tend to clog the need-
to-know channels and magnify difficulties in the compart-
mented systems. In the series of Center studies, compart-
mentation has been viewed as "formal" as it relates to the
codeword systems, "informal" as it pertains to the gener-
alized need-to-know found in the classification system or
the more rigid need-to-know practiced in the Clandestine
Service, and "bureaucratic" as it evolves accidentally
from organizational phenomena. Forthcoming studies in
this series will deal with formal and bureaucratic com-
partmentation as well as other aspects of informal
compartmentation.
2. A unique source of information on the pre-World
War II origins of the classification system is a typescript
by Dallas Irvine, Origins of Defense-Information Markings
in the Army and Former War Department. Washington: National
Archives and Records Service, 1972. Particularly valuable
are the appendices to this work reproducing the original
documents referred to in the text. An excellent overall
view is given by Harold C. Relyea, "The Evolution of
Government Information Security Classification Policy: A
Brief Overview (1775-1973)." It appears as a supplement
in: Government Secrecy: Hearings before the Subcommittee
Approved For Release 20 Qay PM#B L MDP80-00630A000300010001-6
 Approved For Releas1FQI/$R 4IA~IA-RDP80-00630A000300010001-6
on Intergovernmental Relations of the Committee on Government
Operations.. .on S. T5 0, S. 1726, S. 451,-S. 2738-,--S. 3393,
and S. 3399. May 22, 23, 29, 30, 31, and dune 10, 1974.
Washington: U.S. Govt. Print. Office, 1974. The following
two works cover much the same ground, but are useful for
recent developments and bibliographic references: William G.
Phillips. "The Government's Classification System." Ch. 2
in None of Your Business-Government Secrecy in America.
Ed. Normal Dorsen and Stephen Gillers. New Yor : T e
Viking Press, 1974; Representative William S. Moorhead.
"Operation and Reform of the Classification System in the
United States." Ch. 6 in Secrecy and Foreign Policy. Eds.
Thomas M. Franck and Edward Weisband. New Yor : Oxford
University Press, 1974.
3. Other changes effected by E.O. 11652 are summarized
in Annex B.
4. Quoted in H. Rept. 93-221, Executive Classification
of Information-Security Classification Problems Involving
Exemption (b)(1) of the Freedom of Information Act (5 U.S.C.
552). 93d Congress. 2d Sess. Washington: U.S. Govt. Print.
Mice,, 1973, p. 16.
5.
Quoted by Relyea, op. cit., p. 861.
6.
Ibid., pp. 875-876. For the text of the proposed
bill
tains
see
the
Appendix III of Relyea's study. Appendix II con-
text of other recommendations.
7.
Quoted by Relyea, ibid., p. 863.
8.
Ibid., pp. 863-865.
9.
Moorhead, op. cit., p. 98.
10. Ibid., pp. 77-78. See also H. Rept. 93-221, op. cit.
Ch. VI, and Senate Hearings on Government Secrecy, op. cit.,
passim.
11. Moorhead, op. cit., pp. 100-101.
12. Phillips, op. cit., pp. 72, 96, 101. Dr. Ray S.
Cline, testifying at the 1974 Senate hearings estimated that
"there probably is 15 to 25% of the material that is clas-
sified by the Government which, at least at the time, is
legitimately classified." Senate Hearings on Government
Secrecy, op. cit., p. 56.
CONFIDENTIAL
Approved For Release 2001/09/04: CIA-RDP80-00630A000300010001-6
 Approved For Release 2001/09/04: CIA-RDP80-00630A000300010001-6
CONFIDENTIAL
13. Stanley Futterman. "What is the Real Problem with
the Classification System?" Ch. 3 in None of Your Business,
op. cit., p. 103.
14. Quoted by Phillips, op. cit., p. 69.
15. "Secrecy and Intelligence in a
Free Society," an unpublished study done for the Center
for the Study of Intelligence. Some of the highlights of
this study appear in: "Secrecy and Intelligence in a
Free Society-The Dilemma of Security vs. Openness," Studies
in Intelligence. Vol. 19, No. 2, summer 1975.
16. Quoted by William G. Florence. "Issues in
Classifying and Protecting National Defense Information."
Ch. 10 in Surveillance and Espionage in a Free Society.
Ed. Richar um, New York: Praeger u is ers, 972,
p. 132.
17. . cit., pp. 91-92.
18. Florence, op. cit., p. 132.
19. Apposite descripti-on used by Dr. Edward Proctor
in a talk to the Senior Seminar on 3 May 1976.
20. op. cit., pp. 9-10.
21. During the period 1 January 1973 to 31 March 1974,
there were 33 recorded abuses of classification in CIA, 26
of which were attributable to overclassification and two to
unauthorized disclosure, source unknown. The reprimands
given were characterized as more educational than punitive
in tone and were not entered in personnel files. Colby
Letter to Senator Edmund S. Muskie, 24 May 1974. Senate
Hearings on Government Secrecy, op. cit., pp. 114, 469.
22. A memorandum from the chairman of the ICRC, dated
22 March 1976, to the departments authorized to classify
information under E.O. 11652 stresses that the use of such
designations as "Top Secret/Sensitive," "White House
Confidential," and "Conference Confidential" is a violation
of the Order and should he discontinued.
23. Senator Muskie used figure of 96% for the amount
of CIA information exempted. 'senate Hearings on Government
Secrecy, op. cit., p. 102. Commenting on this at a later
date, he remarked: "That is tc~ say the Executive order is a
82
CONFIDENTIAL
Approved For Release 2001/09/04: CIA-RDP80-00630A000300010001-6
 Approved For Release Q 1 ffR ,I.RIA-RDP80-0063OA000300010001-6
dead letter in its application to that institution." Ibid.,
p. 252. At the same hearings the following exemption figures
for other agencies were given: FBI, over 99%; Defense Depart-
ment, 50%; State Department, less than 50%. Ibid., p. 162.
24. CIA Annual Report to the Congress for the Year
1975 on the Freedom of Information Act.
26. See Edward A. Shils, The Torment of Secrecy,
New York: Free Press, 1956, for a sociological analysis
of this triad.
26a. In 1974 the Office of the Inspector General con-
ducted a review to determine "whether Agency officers with
classification authority understand classification criteria
and are complying with guidance...." In pursuing this
objective, the review did not consider "substantive intelli-
gence" nor cable traffic, limiting itself essentially to
"administrative material." The principal conclusions of
its four-page report were these:
- The Agency has made progress in responding to
E.O. 11652. There is "more discretion in clas-
sification decisions."
- There is some overclassification in the Agency,
but it is "not a problem of any serious magnitude."
Much remains to be done, however, in correcting
the overclassification of pre-printed forms. And,
the report adds, "given the nature of the Agency's
business, the tendency to overclassify will remain
with us for some time."
- In some components no distinction is made between
the authority to classify and the authority to
exempt.
- Training for classification officers is virtually
non-existent.
- There is apparently no periodic review of clas-
sified material for downgrading and declassification.
- The report recommends "a periodic training program"
for classification officers and the designation of
an officer in each component to review at regular
CONFIDENTIAL
Approved For Release 2001/09/04: CIA-RDP80-0063OA000300010001-6
 Approved For Release 2001/09/04: CIA-RDP80-0063OA000300010001-6
CONFIDENTIAL
intervals chrono files for classification abuses.
See: Memorandum For: Chief, Information Systems
Analysis Staff. Subject: Classification/Declas-
sification of Information. DD/M and S 74-1319,
dated 12 April 1974.
27. The National Archives and Records Service (NARS)
has general supervisory responsibility for the destruction
of useless official records. Each government Agency is
required to provide the NARS with a records schedule setting
forth the disposition of its records. Actual destruction
may then be carried out by each agency in accordance with
the schedule.
28. Futterman, op. cit., pp. 93-94.
29. The rationale for all forms of supplementary
protection is sensitivity. The security classification
system itself is a three-step sensitivity scale. Top
Secret information is information requiring "the highest
degree of protection." Leaving aside practical considera-
tions such as managerial efficiency, there is obviously
a logical redundancy in providing supplementary protection
to Top Secret information.
30. Memorandum For: Executive Secretary, CIA
Management Committee. Subject: Central Storage and
Retrieval of Sensitive Intelligence Documents, 6 August
1973, p. 5 of Attachment to report entitled "Security
of the CRS System."
31. Ibid., p. 3 of report.
Approved For Release 2001/09/04:CIA-RDP80-00630A000300010001-6
 Approved For Release e8'P l --RDP80-00630A000300010001-6
33. The Rockefeller Commission Report notes: "In
connection with the statutory responsibility of the Director
of Central Intelligence for the protection of intelligence
sources and methods from unauthorized disclosure, the
National Security Council has directed that each agency or
department be responsible for the protection of its own
sources and methods, and that the Director call upon these
other bodies as appropriate to investigate any unauthorized
disclosures and report to him. The Director has, in turn,
delegated these responsibilities to the Security Committee
of the United States Intelligence Board [now the National
Foreign Intelligence Board]..." Report to the Commission
on CIA Activities within the Unite States. Washington:
Govt. Print. Office, June 1975, p. 56. The precise extent
to which the DCI's statutory responsibility extends to
other members of the Intelligence Community is unclear.
34. JSSC memorandum, dated 18 September 1945,
entitled "Proposed Establishment of a Central Intelligence
Service. Report of the Joint Strategic Survey Committee."
It references Joint Chiefs of Staff (JCS) memorandum 1181
(Donovan's recommendations). The National Intelligence
Authority was the predecessor of the National Security
Council.
35. JCS 1181/5 (amended). "Establishment of a
Central Intelligence Service Upon Liquidation of O.S.S.
Directive Regarding the Coordination of Intelligence
Activities." For text see Appendix R, Donovan and the
CIA: A Histor of the Establishment of the Central
Intelligence Agency, by CIA, 1975 SECRET).
36. Memorandum from Director of Naval Communications
to Chief of Naval Operations, dated 8 January 1975. Subject:
Establishment of a National Intelligence Service-Necessity
for Safeguarding the Security of Military Intelligence in
Connection Therewith.
Approved For Release ~kWWT 8!A-RDP80-00630A000300010001-6
 Approved For Release 2001/09/04: CIA-RDP80-0063OA000300010001-6
CONFIDENTIAL
37. This seems borne out by these words in the Central
Intelligence Group (CIG) draft for the CIA section of the
1947 Act: "Be responsible for fully protecting sources and
methods used in the collection of foreign intelligence infor-
mation received by the Agency..." And also in the draft for
a separate CIA Act of 10 March 1947: "Be responsible for
taking measures to protect sources and methods used in the
collection and dissemination of foreign intelligence infor-
mation received by the Agency..." The Rockefeller Commission
Report, op. cit., p. 53, expresses a similar view: "This
language [sources and methods] was originally inserted in
the early drafts of the Act in response to the expressed
concern of some military officials that a civilian agency
might not properly respect the need for secrecy. Congress
was also aware of the concern that the United States espionage
laws were ineffective in preventing unauthorized disclosure of
classified information."
38. "The Protection of Intelligence
Data." S u ies in n e igence. Vol. 11, No. 2, p. 72.
39. Letter fro to Senator Muskie,
Senate Hearings on Government Secrecy, op. cit., p. 115.
40. This definition draws on some of the concepts
contained in the OGC catalog of sources and methods and
the Agency-sponsored bill dealing with sources and methods.
See footnotes 42 and 66. "Foreign intelligence information"
itself would, of course, also require definition. The
Rockefeller Commission Report observes that "'foreign intel-
ligence' is a term with no settled meaning. It is used but
not defined in National Security Council Intelligence
Directives. Its scope is unclear where information has
both foreign and domestic aspects." Op. cit., p. 59. It
adds its belief that "...congressional concern is properly
accommodated by construing 'foreign intelligence' as infor-
mation concerning the capabilities, intentions, and activities
of foreign nations, individuals or entities, wherever the
information can be found. It does not include information
on domestic activities of United States citizens unless there
is reason to suspect they are engaged in espionage or similar
illegal activities on behalf of foreign powers." Ibid., p. 59.
41. For a short account of this case, see: Guide to
CIA Statutes and Law, p. 16. Also Lawrence R. Houston.
"U.S. v. Jarvinen." Studies in Intelligence. Vol. 15,
No. 1.
CONFIDENTIAL
Approved For Release 2001/09/04: CIA-RDP80-0063OA000300010001-6
 Approved For Release MI1/ / VILA-RDP80-00630A000300010001-6
42. Guide to CIA Statutes and Law, pp. 16-18.
43. On this point Judge Haynsworth reiterated his
previous holding "that the First Amendment is no bar against
an injunction forbidding the disclosure of classifiable infor-
mation within the guidelines of the Executive Orders when
(1) the classified information was acquired, during the
course of his employment, by an employee of a United States
agency or department in which such information is handled
and (2) its disclosure would violate a solemn agreement
made by the employee at the commencement of his employment.
With respect to such information, by his execution of the
secrecy agreement and his entry into the confidential
employment relationship, he effectively relinquished his
First Amendment rights." Opinion, p. 15. For a brief
summary of the Marchetti case prior to the final appeal
see'Victor Marchetti and John D. Marks. The CIA and the
Cult of Intelligence. New York: Alfred A. Knopf-, - f97 .
Introduction by Melvin L. Wulf, legal director of ACLU
and Marchetti's defense lawyer. The Supreme Court refused
to review the Marchetti case.
44. Backrack was suing for all information on the
relations of Nicholas de Rochefort (deceased) with CIA and
its predecessor organizations. Paragraph 9 of Judge Gray's
opinion is particularly noteworthy: "Since it is concluded
that the exemptive provisions of 5 U.S.C. 552(b)(3) [that
is, the sources-and-methods provisions under the FOIA
statutory exemption] are applicable herein, the Court has
no occasion to consider whether the sought information, if
it exists, would also be exempt from disclosure by the pro-
visions of U.S.C. 552(b)(1) [information properly classified
pursuant to an Executive order]." The other cases referred
to in the text are: Harriet A. Phillippi V. CIA, et. al.,
1 December 1975, a case in w ic-h in camera examination of
documents with the plaintiff's lawyer present was denied;
William B. Richardson v. J.T. S ahr et. al., 30 January 1976,
a consolidation of three suits demanding CIA financial
records; Gary A. Weissman v. CIA et. al., 14 April 1976,
in which the plaintiff requestedthe CIA security file on
himself; Jonathan A. Bennett v. DOD, CIA, et.al., 13 September
1976, requesting information on all missions sent into Cuba
by DOD; and Morton H. Halperin v. William E. Colby, et. al.,
4 June 1976, a request fogetary information (Although
the sources-and-methods provisions were cited by the. judge,
the case was decided on the basis of exemption (b)(1)); and
Anthony V. Vecchiarello v. Edward Levi, et. al. (CIA),
1 June 1976. The District Court decided that the disputed
information was properly withheld under the FOIA exemptions.
CONFIDENTIAL
Approved For Release 2001/09/04: CIA-RDP80-00630A000300010001-6
 Approved For Release 2001/09/04: CIA-RDP80-00630A000300010001-6
CONFIDENTIAL
45. The sources-an.-methods catalogue, OGC 76-03333,
dated 12 December 1975, is entitled "Aspects of Intelligence
Sources and Methods of the Central Intelligence Agency That
Require Protection from Unauthorized Disclosure." It is
divided into twelve sections with a total of 126 "aspects."
Approved by the DCI on 12 January 1976, it is to be issued
as an Agency regulation. An 0GC staff memorandum explaining
the rationale of the sources-and-methods catagloue notes that
in those instances where the district court decided against
the Agency's deletions from Marchetti's book, it was generally
because the Agency was unable to document prior determinations
concerning the classification of the contested item. The
nondisclosure agreement for sources-and-methods information
will be in addition to the secrecy agreement that employees
now sign for the protection of classified information.
46. Shils, op. cit., p. 41. He adds: "With its
[America's] devotion to publicity on such a scale, it could
scarcely be expected that in its normal state Americans
would have much sympathy with secrecy, particularly govern-
ment secrecy." Ibid., p. 42. And again: "No society has
ever been so extensively exposed to public scrutiny as the
United States in the twentieth century." Ibid., p. 39.
47. Writing of the disclosures of CIA cover and funding
operations in 1967, - concludes: "Habits of thinking
within the Agency and the Executive had become outmoded, and
preserved from change by secrecy." Op. cit., p. 64. In
other words, covert operations that had been appropriate
and credible in the fifties had ceased to be so in 1967,
but were not recognized as such until it was too late.
Secrecy often tends to breed insensitivity to change and
public opinion.
48. Cited by Professor Arthur Schlesinger, Jr., Senate
Hearings, op. cit., pp. 40-41.
49. The preamble to E.O. 11652 takes note of the
section of the Freedom of Information Act (552(b)(1) of
Title 5, U.S.C.) exempting properly classified information
from disclosure, but the Executive order does not expressly
derive its authority from that Act.
50. Quoted by Stanley Futterman. "What is the Real
Problem with the Classification System?" Ch. 3 in None of
Your Business, op. cit., p. 102.
CONFIDENTIAL
Approved For Release 2001/09/04: CIA-RDP80-00630A000300010001-6
 Approved For Release F1 ~
L ( 41 T A-RDP8O-OO63OAOOO3OOO1 OOO1-6
CUN
51. Speaking of the decision in the "Pentagon Papers"
case, New York Times v. United States, Ralph E. Erickson,
then Assistant Attorney General, expressed this view: "While
the Justices applied a number of different standards, it seems
clear that injunctive relief against publication of classified
material already in the hands of the press will be granted
only in the most extreme circumstances, at least in the
absence of specific legislation." Hearings on the Proper
Classification and Handling of Government Information
Involving the National Security and H.R. 9853, a Related
Bill. Special Subcommittee on Intelligence. House Armed
Services Committee. 92d Congress. 2d Sess. March and
May 1972. H.A.S.C. No. 92-79, p. 17472.
52. The then Atomic Energy Commission (AEC) Classifica-
tion Handbook and excerpts from the Atomic Energy Act are
reproduced in the Senate Hearings on Government Secrecy,
op. cit., pp. 364-467.
53. The House Committee Report on Executive Classifica-
tion, op. cit., p. 99, makes this interesting comment on the
atomic energy program:
Like other executive agencies the AEC also
functions within the Executive order classifica-
tion system, as well as its own statutory system.
The committee notes, however, the sharp'contrast
between the apparent efficient operation of the
AEC classification system and the administrative
failures that have marked the operation of the
Executive order system within the past 20 years.
It is true that the highly technical type
of information that is subject to classification
within AEC's own statutory system and its limited
scope of applicability makes it more manageable.
Moreover, scientific development in the atomic
energy field usually provides more precise
benchmarks for measuring the necessity to con-
tinue classification of AEC information at a
particular level than is generally true in the
fields of foreign policy or defense information.
54. Benno C. Schmidt, Jr. "The American Espionage
Statutes and Publication of Defense Information." Ch. 11
in Secrecy and Foreign Policy, op. cit., p. 188. By dropping
the wor "intent" and retaining "reason to believe" and by
substituting for foreign person or power "any person not
CONFIDENTIAL
Approved For Release 2001/09/04: CIA-RDP8O-0063OA000300010001-6
 Approved For Release 2001/09/04: CIA-RDP80-00630A000300010001-6
CONFIDENTIAL
entitled to receive it," subsections (d) and (e) of 793 come
closest to embracing press disclosure of defense information.
Subsection (d) prescribes penalties for one lawfully in pos-
session of defense information who refuses to deliver it on
demand to an officer or employee of the United States "entitled
to receive it"; subsection (e) covers a person who, being in
unauthorized possession of defense information retains it,
communicates it to another unauthorized person, or refuses to
surrender it to an officer or employee of the United States
entitled to receive it. Prior to 1950 there was only a
section (d) applying to government employees; the addition of
(e)- was done as a result of the Whittaker Chambers "pumpkin
papers" case to criminalize retention by non-goverment personnel.
Ibid., p. 188.
55. Ibid., p. 198. The present Espionage Statutes
comprise in the main legislation enacted in 1911, 1917, and
1950. The most recent provision, Section 799, daals with
the protection of NASA secrets.
55a. Ibid., p. 201. Section 952 (18 U.S.C. 952) imposes
penalties on a government employee who publishes or makes
unauthorized disclosure of information concerning or trans-
mitted by a foreign diplomatic code.
56. Quoted in CIA publication entitled "Title 18,
U.S. Code. Sections 792, 793, 794, 795, 796, 797, and
798 with an Interpretation of the Internal Security Act
of 1950." p. 6.
57. op. cit., p. 75. Mention should also
be made of another statutory barrier to disclosure, sub-
section (b) of the Internal Security Act of 1950 (50 U.S.C.
783). This subsection makes it a crime "for any officer or
employee of the United States" to communicate to a foreign
agent "any information of a kind which shall have been
classified by the President as affecting the security of
the United States..." Quoted by Ralph E. Erickson in
statement to the House Subcommittee on Intelligence,
H.A.S.C. No. 92-79, op. cit., p. 17471. It apparently
does not apply to former government employees.
58. Guiding Principles for the Intelligence Community,
13 May 1976. NFIB-D-1/49.
59. ~ & op. cit., p. 10.
60. See Senate bills S. 1726 and S. 2451. Texts and
analysis of these and other bills dealing with classification
Approved For Release 2O ~ $/ JfAl DP80-00630A000300010001-6
 Approved For Release 0001-6
will be found in Legislation on Government Secrecy. Sub-
committee on Intergovernmental Relations. Committee on
Government Operations. U.S. Senate. Washington: Govt.
Print. Office, 1974. Florence, a strong protagonist of a
single-tier system, wrote, op. cit., p. 131: "Consistent
with the urgent need for narrowing the scope of any execu-
tive order or :law for the protection of national defense
information, there should be only one category of such
information. Internal distribution designators could be
used to limit the dissemination of a given item, but there
should be only one classification marking to indicate the
application of law or, in the absence of law, such an
executive order as might exist."
61. See testimony of Fred Buzhardt, General Counsel,
Department of Defense, before House Subcommittee on
Intelligence, op. cit., pp. 17376-17377 and that of
William Blair, Deputy Assistant Secretary of State, ibid.,
pp. 17562-17464. The Rehnquist Committee considered and
rejected suggestions for both a one-tier and a two-tier
classification system.
62. "The ICRC is charged with duties it cannot perform,
given the size of the staff and the volume of non-ICRC duties
of its members. Realistically, it is not possible to allocate
sufficient resources (money and people) to provide the neces-
sary expertise and support services. Further, conceptually,
a Committee of departmental employees cannot monitor and
regulate the policies and practices of the heads of their
own departments." Memorandum to ICRC Chairman, dated
9 March 1976, from CIA member,
OGC-76-1148. The memorandum is entitled: ssues and
Problems with Executive order 11652."
64. The CIA Information Review Committee consists of
the DDA, who is the chairman, the DDI, the DDO, the DDSET,
the Deputy for National Intelligence, and the General
Counsel. It is obvious that a high-level body such as
this, which also handles FOIA appeals, can only address
itself to the most urgent problems posed by the classifica
tion system and not to radical reform of the latter.
65. See Legislation on Government Secrecy, op. cit.
66. See S. 1, Chapter 11, Subchapter C, pp. 68-74.
The bill is dated 15 January 1975. See also Report of the
Approved For Release V/b T:IuAk-RDP80-00630A000300010001-6
 Approved For Release 2001/09/04: CIA-RDP80-00630A000300010001-6
CONFIDENTIAL
Committee on the Judiciary United States Senate to Accompany
S.1, 94th Congress, 1st Sess., Rpt. No. 94, Washington: U.S.
Rout. Print. Office, 1975, particularly pp. 229-263. Sub-
chapter D takes over the provisions of the Atomic Energy Act,
as amended, but does not appear to weaken them. The National
Commission that worked on the codification of Federal criminal
law used the term "national security information." which the
Judiciary Committee replaced with "national defense information."
67. D-Notices are administrative warnings by the
Defense, Press and Broadcasting Committee that information
someone proposes to publish is secret and concerns national
security. For a discussion of the Official Secrets Act see
Franck and Wiesband, op. cit., pp. 325-331.
op. cit., pp. 70-71, explains that the British acts "are
based on the theory of privilege, according to which all
official information, whether or not related to the national
defense and security is the property of the crown. It is
therefore privileged, and those who receive it officially
may not divulge it without the crown's authority." And
he concludes: "In short, the Official Secrets Acts would
seem to be in important respects unconstitutional in this
country and therefore cannot be relied on as examples of
means by which we could protect intelligence data." Ibid.,
p. 72.
68. An Agency-sponsored bill, H.R. 12006, introduced
by Representative Robert McClory on 19 February 1976, aims
at strengthening the sources-and--methods clause by adding a
new subsection (g) to Section 102 of the National Security
Act of 1947. Its main features are these:
--It criminalizes the communication of classified
sources-and-methods information to an unauthorized
person or the general public.
--It extends to anyone who is or has been in authorized
control or possession of classified sources-and-
methods information, but not to the unauthorized
recipient of such information.
--It requires in camera hearings when a court wishes
to determine whether the information was properly
classified and designated as sources-and-methods
information (the court's determination is then a
point of law).
--It provides for a temporary or permanent injunction
when in the judgment of the DCI a violation is
CONFIDENTIAL
Approved For Release 2001/09/04: CIA-RDP80-00630A000300010001-6
 Approved For Release W/%Tn - PIA-RDP80-00630A000300010001-6
RTIAL
imminent. Unless there is presumptive evidence
of improper classification, the court is directed
not to hold an in camera hearing prior to granting
an injunction.
Information relating to sources and methods is defined
as "any information, regardless of its origin, that is clas-
sified pursuant to the provisions of a statute or Executive
order, or a regulation or a rule issued pursuant thereto as
information requiring a specific degree of protection against
unauthorized disclosure for reasons of national security and
which, in the interest of the foreign intelligence activities
of the United States, has been specifically designated by a
department or agency of the United States Government which is
authorized by law or by the President to engage in foreign
intelligence activities for the United States as information
concerning--(A) methods of collecting foreign intelligence;
(B) sources of foreign intelligence, whether human, technical,
or other; or (C) methods and techniques of analysis and evalua-
tion of foreign intelligence."
According to the Office of the General Counsel, the
Justice Department insisted on the provision in the bill
restricting protected sources-and-methods information to
that which has been classified.
The bill's exemption from prosecution of the unauthorized
recipient of sources-and-methods information was apparently
inspired by similar language in Section 1124 of S.1 (dealing
with classified. national defense information). Because of
this exemption, H.R. 12006 affords considerably less protection
to sources-and-methods information than Section 798 of the
Espionage Statutes (also limited to classified information)
affords to communications intelligence.
By limiting sources-and-methods information to that
which has been classified it tends also to undercut the theory
that sources-and-methods information has statutory claims to
protection under Section 102(d)(3), independent of the clas-
sification system. Unlike the Atomic Energy Act, this bill
does not create a unique category of information that is
"born classified."
On the other hand, it breaks new ground in the statutory
protection of information with its provisions for unjunctive
relief and in camera hearings. The injunctive provision could
apparently be invoked against a newsman who planned to publish
sources-and-methods information, but the newsman would not be
actionable under the provisions of the bill if he managed to
publish the information before being restrained.
Approved For ReleascC2 JW 4IANA-RDP80-00630A000300010001-6
 Approved For Release 2001/09/04: CIA-RDP80-00630A000300010001-6
A SELECTION OF CLASSIFICATION CRITERIA*
(Arranged Chronologically)
21 November 1907. Circular No. 78, War Department.
Paragraph 1 of the Circular specified that Confidential is
permitted only "where the subject matter is intended for the
sole information of the person to whom addressed." Paragraph 2
required that issuances be accompanied by a statement indicating
the class or classes of individuals to whom the contents might
be disclosed. Paragraph 4 stated that issuances marked
Confidential in the past were for the use of Army officers,
enlisted men, and government employees "when necessary in
connection with their work." (Irvine, op. cit., p.14.)
British Classification Criteria
The criteria are based on the degree of damage which
disclosure would entail:
Top Secret--"exceptionally grave damage to the nation."
Secret--"serious injury to the interests of the nation."
Confidential--"prejudicial to the interests of the nation."
Restricted--"undesirable in the interests of the nation."
(Franck and Wiesband, op. cit., p. 336.)
*Full identification of the references cited after each entry
will be found in the footnotes.
94
Approved For Release 2001/09/04: CIA-RDP80-00630A000300010001-6
 Approved For Release 2001/09/04: CIA-RDP80-00630A000300010001-6
21 November 1917. General Orders, No. 64, General Hqrs., AEF.
General Orders, No. 64, General Headquarters, American
Expeditionary Force, established three classifications for
"official information."
Confidential: "Confidential matter is restricted for
use and knowledge to a necessary minimum of persons, either
members of this expedition or its employees."
Secret: "The word 'secret' on a communication is intended
to limit t- e use or sight of it to the officer into whose hands
it is delivered by proper authority, and, when necessary, a
confidential clerk. With such a document no discretion lies
with the officer or clerk to whom it is delivered, except to
guard it as SECRET in the most complete understanding of that
term. There are no degrees of secrecy in the handling of
documents so marked. Such documents are completely secret."
For Official Circulation Onl : "Orders, pamphlets of
instruction, maps, diagrams, intelligence publications, etc.,
from these headquarters.. .which are for ordinary official
circulation and not intended for the public, but the acci-
dental possession of which by the enemy would result in no
harm to the Allied cause; these will have printed in the
upper left hand corner, "For Official Circulation Only."
This order also prescribed the following circulation
controls:
Not To Be Taken into Front Line Trenches; Not To Be Reproduced;
Not To Go Below Division Headquarters; Not To Go Below Regi-
mental Headquarters. (Irvine, op. cit., pp. 24-25. The
markings, "Secret" and "Confidential," were borrowed from the
French., The British had a marking "For Official Use Only."
14 December 1917. Compilation of Orders, No. 6, War Department
Section 176 of the Compilation of Orders reads:
A document or map marked 'Secret' is for the personal
information of the individual to whom it is offi-
cially entrusted and of those officers under him
whose duties it affects. The officer to whom it
is entrusted is personally responsible for its safe
Approved For Release 2001/09/04: CIA-RDP80-00630A000300010001-6
 Approved For Release 2001/09/04: CIA-RDP80-0063OA000300010001-6
custody and that its contents are disclosed to those
officers mentioned above, and to them only. The
existence of such a document or map will not be dis-
closed by the officer to whom it is entrusted, ..."
"A document or map marked 'Confidential' is of less
secret a nature than one marked 'Secret,' but its
contents will be disclosed only to persons known
to be authorized to receive them or when it is
obviously in the interest of the public service
that they receive them."
"The information contained in a document or map
marked 'For Offical Use Only' will not be com-
municated to the public or to the press, but may
be communicated to any person known to be in the
service of the United States simply by virtue of
his official position."
Section 176 also required that catalogues of classified
documents be classified. (Irvine, op. cit., Appendix K.)
22 January 1921, Army Regulations, 330-5.
"A document will be marked 'Secret' only when the
information it contains is of great importance and
when the safeguarding of that information from
actual or potential enemies is of prime necessity.
A document will be marked 'Confidential' when it
is of less importance and of less secret nature
than one requiring the mark of 'Secret,' but which
must, nevertheless, be guarded from hostile or
indiscreet persons.
A document will be marked 'For Official Use Only'
when it contains information which is not to be
communicated to the public or to the press, but
which may be communicated to any person known to
be in the service of the United States whose duty
it concerns, or to persons of undoubted loyalty
and discretion who are cooperating with Government
work." (Irvine, op. cit., pp. 33-34.)
Approved For Release 2001/09/04: CIA-RDP80-0063OA000300010001-6
 Approved For Release 2001/09/04: CIA-RDP80-00630A000300010001-6
12 February 1935, Changes No. 3 in Army Regulations, 850-25.
A fourth classification "Restricted" was added.
"Whenever the chief of an arm or service which is
charged with a research project or design, develop-
ment, test, and production of a unit of military
equipment or component thereof, shall determine
that the maintenance of secrecy regarding any such
project is sufficiently important to the national
defense of the United States to warrant it, he may
declare it a "-Restricted' project. Information
regarding a 'Restricted' Project will be considered
to be information affecting the national defense
within the meaning of the provisions of the
Espionage Act (sections 1 and 2, Title I,...
"During the period that a project has a restricted
status, all documents-containing technical infor-
mation regarding it will be identified by being
marked substantially as follows: Notice This
document contains?information affect ni g the
national defense of the United States within the
meaning of the Espionage Act (U.S.C. 50:31, 32).
The transmission of this document or the revela-
tion of its contents in any manner to any
unauthorized person is prohibited." (Irvine,
op. cit., pp. 38-39.)
1936. Army Regulations, 330-5.
"For Official Use Only" was dropped from Army Regulations.
The following new definitions were added:
"A document will be classified and marked 'Secret'
only when the information it contains is of such
nature that its disclosure might endanger the
national security, or cause serious injury to the
interests or prestige of the Nation, an individual,'
or any government activity, or be of great advantage
to a foreign nation."
"A document will be classified and marked 'Confidential'
when the information it contains is of such a nature
that its disclosure, although not endangering the
national security, might be prejudicial to the interests
or presitge of the Nation, an individual, or any govern-
ment activity, or be of advantage to a foreign nation."
Approved For Release 2001/09/04: CIA-RDP80-00630A000300010001-6
 Approved For Release 2001/09/04: CIA-RDP80-0063OA000300010001-6
"A document will be classified and marked 'Restricted'
when the information it contains is for official use
only or of such a nature that its disclosure should
be limited for reasons of administrative privacy, or
should be denied to the general public." (Irvine,
op. cit., pp. 39-40.)
28 September 1942, OWI Regulation No. 4.
"Secret information is information the disclosure of
which might endanger national security, or cause
serious injury to the Nation or any government
activity thereof."
"Confidential information is information the dis-
closure of which although not endangering the
national security would impair the effectiveness
of government activity in the prosecution of war."
"Restricted information is information the disclosure
of which should be limited for reasons of administra-
tive privacy, or is information not classified as
confidential because the benefits to be gained by a
lower classification, such as permitting wider dis-
semination where necessary to effect the expeditious
accomplishment of a particular project, outweigh the
value of the additional security obtainable from the
higher classification." (Relyea, op. cit., p. 855.)
5 November 1953, E.O. 10501.
E.O. 10501 described classified information as "official
information affecting the national defense." Top Secret was
defined as "'defense information' requiring the highest degree
of protection whose disclosure would cause "exceptionally grave
damage to the nation."
Secret information was information whose disclosure would
cause "serious damage to the nation." The words "vital" and
"important" were used to qualify the examples given of Secret
Approved For Release 2001/09/04: CIA-RDP80-0063OA000300010001-6
 Approved For Release 2001/09/04: CIA-RDP80-0063OA000300010001-6
information. Confidential was defined as information whose
unauthorized disclosure "could be prejudicial to the defense
interests of the nation." No examples of Confidential were
given.
21 June 1957. Legislation Proposed by the Wright Commission.
"(1) the term 'top secret' or 'atomic top secret'
means any information affecting the national
defense of the United States in such degree that
.its unauthorized disclosure could result in
exceptionally grave damage to the Nation; and
"(2) the term 'secret' or 'atomic secret' means
any information affecting the national defense
of the United States in such degree that its
unauthorized disclosure could result in serious
damage to the Nation." (Quoted by Relyea,
op. cit., p. 876.)
March 9, 1972. Defense Department Guidelines for Classification.
" ..a determination to classify shall be made only
when one or more of the following considerations
are present and the unauthorized disclosure of
the information could result in a degree of harm
to the national security:
1. The information provides the United States,
in comparison with other nations, with a
scientific, engineering, technical, opera-
tional intelligence, strategic or tactical
advantage related to the national defense.
2. Disclosure of the information would weaken
the international position of the United
States; create or increase international
tensions contrary to U.S. interests; result
in a break in diplomatic relations; or lead
to hostile economic, political, or military
action against the United States or its
allies, thereby adversely affecting the
national defense.
Approved For Release 2001/09/04: CIA-RDP80-0063OA000300010001-6
 Approved For Release 2001/09/04: CIA-RDP80-00630A000300010001-6
3. Disclosure of the information would weaken
the ability of the United States to wage war
or defend itself successfully, limit the
effectiveness of the Armed Forces, or make
the United States vulnerable to attack.
4. There is sound reason to believe that other
nations do not know that the United States
has, or is capable of obtaining, certain
information or material which is important
to the international posture or national
defense of the United States vis-a-vis
those nations.
S. There is sound reason to believe that the
information involved is unique, and is of
singular importance or vital to the national
defense.
6. The information represents a significant
breakthrough in basic research which has
an inherent military application potential
in a new field or radical change in an
existing field.
7. There is sound reason to believe that knowledge
of the information would provide a foreign
nation with an insight into the war potential
or the war or defense plans or posture of the
United States; allow a foreign nation to
develop, improve or refine a similar item of
war potential; provide a foreign nation with
a base upon which to develop effective counter-
measures; weaken or nullify the effectiveness
of a defense or military plan, operation,
project, or activity which is vital to the
national defense.
The criteria will be reexamined in light of the
new Executive order [that is, E.O. 11652].
(Statement of Joseph J. Liebling, Deputy Assistant
Secretary of Defense. Senate Hearings on Clas-
sification, op. cit., pp. 17409-17410.)
Approved For Release 2001/09/04: CIA-RDP80-00630A000300010001-6
 Approved For Release 2001/09/04: CIA-RDP80-0063OA000300010001-6
1974. S-2451, a bill introduced by Senator William D. Hathaway.
Secret Defense Data, a single classification defined as
information the disclosure of which would adversely affect the
ability of the United States to protect itself afainst overt
or covert hostile action. (Senate Hearings on Government
Secrecy, op. cit., p. 77).
19 February 1976. Definition of Classified Sources and Methods
of Information. See footnote 68.
Approved For Release 2001/09/04: CIA-RDP80-0063OA000300010001-6
 Approved For Release 2001/09/04: CIA-RDP80-00630A000300010001-6
ANNEX B
PRINCIPAL CHANGES EFFECTED BY E.O. 11652
The Order:
tightens the classification criteria.
reduces the number of agencies authorized to classify
from 47 under E.O. 10501 to 26, only thirteen of which
,have Top Secret classifying authority.
establishes schedules for automatic downgrading and
declassification of all information except that
falling in four exemption categories. (Under E.O.
10501 only one of four groups of information was
eligible for downgrading and declassification.)
It reduces the downgrading interval from three to
two years, and the declassification cycle from
12 years for all levels under the former Order to
10 for Top Secret, eight for Secret, and six for
Confidential.
subjects exempted information to mandatory review
for declassification ten calendar years after its
origination upon request of a member of the public
or of another government agency and to automatic
declassification thirty years after its origination
unless the head of the originating agency determines
102
Approved For Release 2001/09/04: CIA-RDP80-00630A000300010001-6
 Approved For Release 2001/09/04: CIA-RDP80-00630A000300010001-6
in writing that the exempted information requires
additional protection.
charges the National Security Council with monitoring
the implementation of the Order and, to assist the
NSC in this function, envisages the creation of an
Interagency Classification Review Committee (ICRC).
(The ICRC was established by E.O. 11714, dated
24 April 1973, with the following composition: a
chairman appointed by the President; representatives
of the Departments of State, Defense, and Justice,
of the Atomic Energy Commission, the Central
Intelligence Agency, and the National Security
Council Staff.)
refers specifically in the preamble to the Freedom
of Information Act, stressing that the interests of
the United States are best served by making informa-
tion readily available to the public. In the accom-
panying statement President Nixon sums up well the
dichotomy implicit in FOIA and the classification
system: "Clearly the two principles of an informed
public and of confidentiality within the Government
are irreconcilable in their purest form, and a
balance must be struck between them."
- inveighs, in Section 4, against the besetting sins
of the classification system throughout its
Approved For Release 2001/09/04: CIA-RDP80-00630A000300010001-6
 Approved For Release 2001/09/04: CIA-RDP80-00630A000300010001-6
history--overclassification and unnecessary
classification. Offenders "shall be notified"
and, in case of repeated abuses, shall be subject
to "administrative reprimand." Heads of agencies
are to take "prompt and stringent administrative
action" against those guilty of unauthorized
disclosure.
provides, in Section 9, a charter for supplementary
protection of classified information and material,
in particular for codeword compartmentation and
the informal compartmentation techniques employed
within the Clandestine Service. It reads: "Special
Departmental Arrangements. The originating Department
or other appropriate authority may impose, in con-
formity with the provisions of this order, special
requirements with respect to access, distribution
and protection of classified information and material,
including those which presently relate to communica-
tions intelligence, intelligence sources and methods
and cryptography."
Approved For Release 2001/09/04: CIA-RDP80-00630A000300010001-6
 Approved For Release 2001/09/04: CIA-RDP80-00630A000300010001-6
BREAKDOWN OF THE CLASSIFICATION OF THIS STUDY
OVERALL CLASSIFICATION ..... CONFIDENTIAL E2 IMPDET
BY SECTIONS:
INTRODUCTION ..... CONFIDENTIAL (Unclassified if asterisked footnote
of page 1 is removed.)
PRINCIPAL CONCLUSIONS AND RECOMMENDATIONS .....ADMINISTRATIVE -
INTERNAL USE ONLY
DEVELOPMENT OF THE CLASSIFICATION SYSTEM ..... UNCLASSIFIED
THE FUNCTIONING OF CLASSIFICATION WITHIN CIA ..... CONFIDENTIAL
Misclassification ..... CONFIDENTIAL
Related Issues ..... Confidential
Reasons for Misclassification .....ADMINISTRATIVE - INTERNAL
USE ONLY
Anomalies in the Handling of Top Secret Material ..... CONFIDENTIAL
Exemption, Downgrading, and Declassification ..... ADMINISTRATIVE -
INTERNAL USE ONLY
Impact of Freedom of Information and Privacy
Acts ..... ADMINISTRATIVE - INTERNAL USE ONLY
Scope of Noncompliance with E.O. 11652 ..... CONFIDENTIAL
Effect of Classification on the Agency ..... CONFIDENTIAL
THE PROBLEM OF DISCLOSURE ..... ADMINISTRATIVE - INTERNAL USE ONLY
CONCLUSIONS AND RECOMMENDATIONS ..... ADMINISTRATIVE - INTERNAL USE
ONLY
FOOTNOTES ..... CONFIDENTIAL (Except for footnote 32, which is
Confidential, footnotes as a whole are unclassified,
Administrative - Internal Use Only
ANNEX A ..... UNCLASSIFIED
ANNEX B ..... UNCLASSIFIED
ANNEX C. . . . , UNCLASSIFIED
Approved For Release 2001/09/04: CIA-RDP80-00630A000300010001-6
 CONFIDENTIAL
Approved For Release 2001/09/04: CIA-RDP80-00630A000300010001-6
Approved For Release 2001/09/04: CIA-RDP80-00630A000300010001-6
CONFIDENTIAL