INFORMATION SECURITY OVERSIGHT OFFICE (ISOO) ANNUAL REPORT TO THE PRESIDENT

• Document Type: CREST
• Collection: General CIA Records
• Document Number (FOIA) /ESDN (CREST): CIA-RDP89G00643R001100020015-7
• Release Decision: RIPPUB
• Original Classification: C
• Document Page Count: 42
• Document Creation Date: December 23, 2016
• Document Release Date: October 14, 2011
• Sequence Number: 15
• Publication Date: June 18, 1987
• Content Type: REPORT

-tr~~v?teet~,a~-
25X1
25X1
25X1
25X1
25X1
25X1
ROUTING AND
RECORD SHEET
SUBJECT: (Optional)
INFORMATION SEC
URITY
OVERSIGHT OFFICE (IS00)
ANNUAL
REPOR
T TO THE PRESIDENT
FROM:
EXTENSION
NO. DD A Re ist
A
enc
Sec
urit
lassif
i
cation Officer
g
y
y
-
Office of
Informio
at
DATE Jli"
1$ ~ ly,"
TO: (ORlcer designation,
building)
room er, and
numb
OFFICER'S
COMMENTS (Number each comment to show from whom
1 9
REOEIVED
FORWARDED
INITIALS
to whom. Draw a line across column after each comment.(
1
Executive
- If
Assistant t h e
/q
Deputy Director for A minis
ratio
Attached is the Information
7018 Hq
Security Oversight Office's
2.
(IS00) FY 86 Annual Report to
the President. IS00's Annual
Report is based on
3.
classification statistics
14
!j 7
provided by the Agency and
other Executive branch
4.
CMS
_
I
~
n2 agencies.
y/
~
,
0
Also attached is a brief
5.
summary of ISOO's oversight
role and the Agency's FY 86
statistical report to IS00.
6'
0/DDA/IMO
Twice annually each Agency
office is obliged to count
7.
the number of classified
documents it produces. This
is a laborious and time
8.
consuming task for all
concerned and I would like
to take this opportunity
v
to express my appreciation
for the cooperation and
support of your office
10
articularly that of~
F ODDA Information
Management Officer (IMO).
11.
12.
13.
14.
15.
FORM 610 USE aIIONS P 4 U.s. Government Mating Office: 1uer14eae/eelee
i_,e l\ I
Sanitized Copy Approved for Release 2011/10/14 CIA-RDP89GO0643RO01100020015-7
 INFORMATION SECURITY OVERSIGHT OFFICE (IS00)
ANNUAL REPORT TO THE PRESIDENT
1. The I500 Annual Report reports on Executive branch
compliance with Executive Order (EO) 12356 - National Security
Information. The EO defines the types of information that must
be protected by national security classification, the procedures
to ensure its protection and, finally, authorizes ISOO to
oversee Executive branch compliance. The IS00 Report is based
on statistics collected from the Executive branch agencies and
IS00's on-site inspections.
2. The Agency provides IS00 with information on
classification related matters, on Mandatory Review requests and
on Agency inspections or surveys done in relation to the
information security program. The Classification Management
Branch (CMB) is responsible for collecting the information and
for providing it to I500 using Standard Form (SF) 311. A brief
summary of Agency reporting follows:
a. Sections 5, 6 and 7 of SF 311 concerns the
number of persons authorized to make original
classification decisions and the number of classified
documents created during FY 86. Agency data concerning the
number of classified documents created is compiled by
taking an actual count of all classified documents produced
by the Agency during two one-week periods. Agency
Information Management Officers (IMO) are responsible for
insuring that all classified documents originated during
the two one-week periods are counted. The first count is
taken in March and the second count taken in September.
The totals from the two counts are combined and projected
for the entire year. Using this process, we reported a
total of 2,497,456 classified documents originated in the
Agency during FY 86. The classification count is the most
labor intensive part of the ISOO reporting requirement.
b. Section 8 of SF 311 concerns Mandatory Review (MR)
requests. ISOO requires reporting on the number and status
of the MR requests received. Most Agency MR requests are
from Presidential libraries or from the National Archives
and Records Administration (NARA). The Information and
Privacy Division is responsible for processing Agency MR
requests and for providing CMB with information for
inclusion in the Agency's annual report to ISOO.
c. Sections 10, 11 and 12 of SF 311 refer to the
number of self-inspections and surveys carried out on
various aspects of our information security program such as
improper storage of classified information or unauthorized
access, or improper classifications or classification
markings. Agency information provided to IS00 on
self-inspections and surveys relates to Office of Security
investigations of security violations and overseas security
surveys. Information relating to document classification
activities is provided by CMB.
 STAT
AGENCY INFORMATION
A. FROM
B. TO
SECURITY PROGRAM DATA
1 Oct. 85
30 Sen. Rf
Central Intelligence Agency
Office o
n ormation Services
irect
f Administration
A. TOP SECRET
B. SECRET
C. CONFIDENTIAL
D. TOTAL
0
0
0
0
ORIGINAL
7. CLASSIFICATION
DECISIONS
DATE OR EVENT
DETCRIGINATING AGENCY'S ERMINATION REQUIRED
DE'IVATIVE
(O.A. D. R.1
a
l
IbI
Ib1
A. TOP SECRET
104
11,596
64,428
B. SECRET
1,S86
101 400
' 029 638
C. CONFIDENTIAL
832
20,150
403.390
8. MANDA-
CASES FOR WHICH AGENCY IS RESPONSIBLE FOR FINAL DECISION
TORY
CASES
DECLASSIFICATION DECISIONS
CASES
REVIEW
CARRIED
NEW
(Report in cases, documents, and pages)
CARRIED
OVER FROM
CA
ES
REQUESTS
PREVIOUS
S
RECEIVED
GRANTED IN FULL
GRANTED IN PART
DENIED
OVER TO
NEXT
AND
PERIOD
(cI
(d)
(el
PERIOD
APPEALS
(a)
(b)
CASES
ROCS.
PAGES
CASES
, ROCS.
, PAGES
CASES
I DOGS.
PAGES
(fl
AREQUESTS
123
363
5J
121
530
191
3165
I
11936
8
] 201
I
2159
151
B.APPEALS
11
19
31
J
4J
203
12
*This figure reflects two completed cases not accounted for under "declassification
decisions"; both were cancelled.
11, NUMEER OF INFRACTIONS INVOLVING.
B. UNOERCLASSIFICATION IC. CLASSIFICATIONNITHOUT
AUTI-*OR [TV
311102 STANDARD FORM 311 (REV. 4-93)
Sanitized Copy Approved for Release 2011/10/14 : CIA-RDP89GO0643RO01100020015-7 rescrlbea by GSA/ISOO
-.O. 12356
 Sanitized Copy Approved for Release 2011/10/14: CIA-RDP89G00643R001100020015-7
Information
Security
Oversight
Office
Annual
Report
to the
President
FY 1986
Sanitized Copy Approved for Release 2011/10/14: CIA-RDP89G00643R001100020015-7
 Thank you for your latest report on the Government-
wide information security program established in
Executive Order 12356. On many occasions I have
stressed the need for vigilance in protecting infor-
mation essential to our national security, while
ensuring that only the information that requires
such protection is classified. I am pleased to
note the progress we are making in both areas.
I take particular interest in the findings of your
special document reviews, especially in the evidence
that the problem of overclassification is not as wide-
spread as has been suggested in some quarters. Despite
this positive development, I again urge those officials
who are authorized classifiers to make every effort not
to classify information unless the national security
demands it. The integrity of the entire information
security program depends upon these efforts.
Once again, I want to express my appreciation to
you, your staff, and all the others throughout
Government and industry involved in the protection of
national security information. With perseverance, the
system we have established will continue to meet
successfully America's need in this vital area.
Mr. Steven Garfinkel
Director
Information Security Oversight
office
18th and F Streets, N.W.
Washington, D.C. 20405
 Sanitized Copy Approved for Release 2011/10/14 : CIA-RDP89GO0643RO01100020015-7
inrormauon oecunry uversigni urnce
Washington, DC 20405
I am pleased to submit the Information Security Oversight
Office's (IS00) 1986 Report to the President.
This is IS00's fourth Report on the system that you
established in Executive Order 12356, "National Security
Information." By necessity a system that is primarily concerned
with protecting our vital secrets, it operates under your firm
instruction to limit classification to its necessary minimum.
This Report notes many achievements in support of this mandate.
Among the reported data are a substantial decrease in the total
number of classification actions and a substantial increase in
the declassification of historically valuable records.
In addition to the data that ISOO collects from the
agencies, for the first time this Report also includes the
results of special reviews of classified documents conducted by
IS00's analysts in FY 1986. These reviews provide further
evidence of a system that is generally working very well. Just
as important, however, they reveal that a little extra effort can
eliminate almost all the technical errors that comprise most of
the problems that we uncovered.
Throughout the executive branch people are working to meet
your goal of enhanced protection for national security
information without excessive classification. Your commitment to
an effective oversight program offers us the opportunity to have
a positive impact upon their efforts.
Respectfully,
Steven Garfinkel
Director
The President
The White House
Washington, DC 20500
 Table of Contents Letter to the President ............................... 1
Agency Acronyms or Abbreviations ..................... 5
Summary of FY 1986 Program Activity ................... 7
The Information Security Program FY 1986 ................ 9
Program Reviews and Inspections ..................... 9
Statistical Reporting ............................... 10
Use of Sampling Systems ........................... 10
Original Classification Authorities ..................... 12
Original Classification .............................. 13
Derivative Classification ............................ 16
Combined Classification Activity ...................... 19
Mandatory Review for Declassification .................. 22
Systematic Review for Declassification ................. 25
Agency Self-Inspections and Infractions ................ 27
Appendix A: ISOO Special Document Reviews .............. 28
B: Classified Information Nondisclosure Agreement .. 35
C: ISOO Standard Forms Program ............... 36
D: ISOO Information Security Briefing Series ....... 39
1. Original Classifiers ............................... 12
2. Original Classifiers by Level
FY1986 ....................................... 12
3. Original Classification Activity
FY 1985-1986 .................................. 13
4. Original Classification Level Assignments .............. 14
5. Original Classification Activity by Agency
FY 1985-1986 ..................................14
6. Original Classification Activity
Major Agencies
FY1986 ....................................... 15
7. Original Classification/Declassification Assignments
FY 1986 .......................................15
8. Derivative Classification Activity
FY1985-1986 ..................................16
9. Derivative Classification Level Assignments ............. 17
10. Derivative Classification Activity by Agency
FY1985-1986 ..................................18
11. Derivative Classification Activity
Major Agencies
FY1986 .......................................18
12. Combined Classification Activity
FY1985-1986 .................................. 19
13. Combined Classification Level Assignments ............ 20
14. Combined Classification Activity by Agency
FY 1985-1986 ..................................20
15. Combined Classification Activity
Major Agencies
FY 1986 .......................................21
 Exhibits (Cont.) 16. Mandatory Review
Requests Received
FY 1979-1986 ................................... 22
17. Mandatory Review
Pages Processed
FY 1983-1986 ..................................22
18. Mandatory Review
Action Taken
FY 1986 .......................................23
19. Mandatory Review
Actions by Agency
FY 1986 .......................................23
20. Mandatory Review Appeals
Pages Processed
FY1983-1986 ..................................24
21. Mandatory Review Appeals
Workload in Pages
FY 1986 .......................................24
22. Systematic Review for Declassification
Pages Reviewed
FY 1973-1986 ...................................25
23. Systematic Review for Declassification
Percentage of Pages Declassified
FY 1973-1986 ...................................25
24. Systematic Review for Declassification
Actions by Agency
FY1986 .......................................26
25. Agency Self-Inspections ........................... 27
26. Infractions
FY 1983-1986 ..................................27
27. ISOO Special Document Reviews
The Sample
By Document Type/Classification Level ................ 28
28. ISOO Special Document Reviews
Classification Levels .............................. 30
29. ISOO Special Document Reviews
Basis for Classification ............................ 31
30. ISOO Special Document Reviews
Original/Derivative Classification ..................... 31
31. ISOO Special Document Reviews
Duration of Classification .......................... 32
32. ISOO Special Document Reviews
Discrepancies .................................. 33
 Agency Acronyms or
Abbreviations Used in This
Report
ACDA
Arms Control and Disarmament
Labor
Department of Labor
Agency
MMC
Marine Mammal Commission
AID
Agency for International
NARA
National Archives and Records
Development
Administration
Air Force
Department of the Air Force
NASA
National Aeronautics and Space
Army
Department of the Army
Administration
BIB
Board for International
Navy
Department of the Navy
Broadcasting
NLRB
National Labor Relations Board
CEA
Council of Economic Advisers
NRC
Nuclear Regulatory Commission
CIA
Central Intelligence Agency
NSA
National Security Agency
Commerce
Department of Commerce
NSC
National Security Council
DARPA
Defense Advanced Research
NSF
National Science Foundation
Projects Agency
OA, EOP
Office of Administration, Executive
DCA
Defense Communications Agency
Office of the President
DCAA
Defense Contract Audit Agency
OJCS
Organization of the Joint Chiefs of
DIA
Defense Intelligence Agency
Staff
DIS
Defense Investigative Service
OMB
Office of Management and Budget
DLA
Defense Logistics Agency
OMSN
Office for Micronesian Status
DMA
Defense Mapping Agency
Negotiations
DNA
Defense Nuclear Agency
Overseas Private Investment
DOD
Department of Defense
Corporation
DOE
Department of Energy
OPM
Office of Personnel Management
DOT
Department of Transportation
OSD
Office of the Secretary of Defense
ED
Department of Education
OSTP
Office of Science and Technology
EPA
Environmental Protection Agency
Policy
EXIMBANK
Export-Import Bank
OVP
Office of the Vice President
FBI
Federal Bureau of Investigation
PC
Peace Corps
FCA
Farm Credit Administration
PFIAB
President's Foreign Intelligence
FCC
Federal Communications
Advisory Board
Commission
President's Intelligence Oversight
FEMA
Federal Emergency Management
Board
Agency
SBA
Small Business Administration
FHLBB
Federal Home Loan Bank Board
SDIO
Strategic Defense Initiative
FMC
Federal Maritime Commission
Organization
FRS
Federal Reserve System
Securities and Exchange
GSA
General Services Administration
Commission
HHS
Department of Health and Human
SSS
Selective Service System
Services
State
Department of State
HUD
Department of Housing and Urban
Treasury
Department of the Treasury
Development
TVA
Tennessee Valley Authority
ICC
Interstate Commerce Commission
USDA
Department of Agriculture
ISOO
Information Security Oversight
USIA
United States Information Agency
Office
USPS
United States Postal Service
Interior
Department of the Interior
USTR
Office of the United States Trade
ITC
International Trade Commission
Representative
Justice
Department of Justice
Veterans Administration
 Summary of FY 1986 Program
Activity
The FY 1986 Report to the President is the fourth to examine the infor-
mation security program under E.O. 12356. The statistics concerning
classification decisions include DOD figures that reflect its use of a
revised sampling system introduced in FY 1985. The following data
highlight ISOO's findings:
Classification Activities ? The number of original classification authorities decreased
significantly to 6,756.
? Original classification decisions increased 6%, to 1,221,110.
? By classification level, 2% of original classification decisions were
Top Secret, 53% were Secret, and 45% were Confidential.
? Derivative classification decisions decreased 32%, to 9,548,538.
? The total of all classification actions, 10,769,648, marked a surprising
29% decrease from the prior year.
? Among executive branch agencies, DOD accounted for 68% of all
classification decisions; CIA 24%; Justice 5%; State 2%; and all
others less than 1 %.
Declassification Activities ? Agencies received 4,081 new mandatory review requests.
? Agencies processed 3,991 cases, 10% more than in FY 1985;
declassified in full 119,504 pages; and declassified in part 39,911
additional pages.
? Agencies received 497 new mandatory review appeals, 76% more
than in FY 1985.
? Agencies acted on 575 appeals, 10% more than in FY 1985, and
declassified in whole or in part 46,529 pages in addition to those
released in the initial mandatory review process.
? Under the systematic review program, agencies reviewed 16,373,035
pages of historically valuable records, an increase of 57% over
FY 1985; and declassified 14,272,268 pages, 76% more than in
FY 1985.
? Agencies conducted 27,361 self-inspections, a slight decrease from
FY 1985.
? Agencies reported 13,812 infractions, 9% fewer than in FY 1985.
 Information Security
Oversight Office
The Information Security
Program
FY 1986
Under Executive Order 12356, the
Information Security Oversight
Office (ISOO) is responsible for
monitoring the information
security programs of those exec-
utive branch activities that create
or handle national security infor-
mation. Originally established by
Executive Order 12065, ISOO
continues to be the primary over-
sight organization in the system
prescribed by President Reagan's
Order of April 2, 1982. In this role,
ISOO oversees the information
security programs of approxi-
mately 65 departments, inde-
pendent agencies and offices of
the executive branch. E.O. 12356
also requires the Director of ISOO
to report annually to the President
about the ongoing implementa-
tion of the Order's provisions.
This Report summarizes Govern-
ment-wide performance during
FY 1986, the system's fourth year.
ISOO is an administrative com-
ponent of the General Services
Administration but receives its
policy direction from the National
Security Council. The Administra-
tor of General Services appoints
the ISOO Director, whose
appointment must be approved by
the President. The ISOO Director
appoints the staff, which numbers
between 13-15 persons. For
FY 1986, ISOO's budget was
$686,000.
ISOO accomplishes its mission
through a number of different
oversight activities. First, it
develops and issues implement-
ing directives and instructions
regarding the Order. Second,
ISOO conducts on-site inspec-
tions or program reviews of agen-
cies that create or handle national
security information. During
FY 1986, ISOO analysts also
continued to monitor the agencies'
implementation of the Classified
Information Nondisclosure Agree-
ment, Standard Form 189, pre-
scribed by National Security
Decision Directive 84 (NSDD 84),
which all cleared personnel must
sign as a condition of access to
classified information. Appendix
B, p. 35, reports on the status of
implementation of this require-
ment by each agency. Third, ISOO
gathers, analyzes, and reports
statistical data on agencies' pro-
grams. Fourth, it evaluates, devel-
ops, or disseminates security
education materials and pro-
grams. Having begun in FY 1986,
ISOO is now completing revision
of its Information Security Briefing
series. Appendix D, p. 39, briefly
describes this program. Fifth,
ISOO receives and takes action
on suggestions, complaints, dis-
putes, and appeals from persons
inside or outside the Government
on any aspect of the administra-
tion of the Order. In this area,
ISOO serves as the final appellate
authority for the mandatory
declassification review of presi-
dential materials. Sixth, it con-
ducts special studies on identified
or potential problem areas and on
programs to improve the system.
During FY 1986, ISOO conducted
special document reviews of the
major classifying agencies in
addition to its regular inspections.
Appendix A, p. 28, reports the
cumulative results of these
reviews. Also in FY 1986, ISOO
continued its program to develop
and issue standard forms to pro-
mote uniform implementation of
the Order, and to reduce Govern-
ment costs by eliminating unnec-
essary duplication. Appendix C,
p. 36, describes the standardized
forms that ISOO has issued to
date. Seventh, ISOO maintains
continuous liaison with monitored
agencies on all matters relating to
the information security system.
This Report is based upon pro-
gram reviews and inspections
conducted by the ISOO staff and
the compilation and analysis of
statistical data regarding each
agency's program activity.
Program Reviews and
Inspections
ISOO's program analysts serve as
liaison to specific agencies to
facilitate coordination and to pro-
vide for continuity of oversight
operations. The analysts must
stay abreast of relevant activities
within each agency's information
security program; coordinate with
assigned agency counterparts on
a continuing basis; and conduct
formal inspections of the
agency's program in accordance
with a planned annual inspection
schedule. These inspections may
include visits to selected field
activities as well as offices in the
Washington metropolitan area.
These on-site surveys encom-
pass all aspects of the informa-
tion security program, including
classification, declassification,
 safeguarding, security education,
and administration. The inspec-
tions include detailed interviews
with agency security personnel,
classifiers, and handlers of
national security information. To
the extent possible, ISOO ana-
lysts review a sampling of classi-
fied information in the agency's
inventory to examine the propriety
of classification, the existence of
necessary security markings and
declassification instructions, and
compliance with safeguarding
procedures. ISOO analysts also
monitor security training pro-
grams to determine if the agen-
cies adequately educate person-
nel about classifying, declassify-
ing, marking, and safeguarding
national security information.
When weaknesses in an agency's
program are identified, ISOO ana-
lysts recommend corrections,
either on-the-spot or as part of a
formal inspection report. Critical
reports require immediate reme-
dial attention by the agency prior
to a follow-up inspection by ISOO.
These inspections provide spe-
cific indicators of agency com-
pliance or noncompliance with
E.O. 12356 that are not apparent
simply from the analysis of statis-
tical data.
Statistical Reporting
To gather relevant statistical data
regarding each agency's informa-
tion security program, ISOO
developed the Standard Form 311,
which requires each agency
to report annually the following
information:
1. The number of original classi-
fication authorities;
2. the number of declassification
authorities;
3. the number of original classi-
fication decisions, including the
classification level of those
decisions and the duration of
classification;
4. the number of derivative classi-
fication decisions by classifica-
tion level;
5. the number of requests
received for mandatory review
for declassification and agency
actions in response to these
requests in terms of cases,
documents, and pages;
6. the number of pages of
national security information
reviewed during the year under
systematic declassification pro-
cedures and the number
declassified;
7. the number of formal self-
inspections conducted by the
agency; and
8. the number of security infrac-
tions detected by the agency
within its own program.
The statistics reflected in this
Report cover the period October 1,
1985, through September 30,
1986. For most of the agencies
that ISOO monitors, the statistics
reported each year are based on
an actual count in each category.
Because of the enormous volume
of classification activity in some
of the larger agencies, they must
calculate their classification
actions on the basis of approved
sampling systems. All other data,
including classification author-
ities, declassification actions,
self-inspections, and infractions,
are based on actual counts for all
agencies.
The sampling system originally
developed by DOD, and in use
since ISOO began collecting pro-
gram activity statistics, was
based entirely on electronically
transmitted message traffic. DOD
then extrapolated the data
regarding electronically transmit-
ted messages in an effort to esti-
mate classification data for other
types of documentation. At the
time, it was believed to be the
only feasible means for DOD to
sample its classification activity.
Although ISOO approved the
message traffic system, ISOO and
DOD were never satisfied that it
was producing the most accurate
data, except to the extent that a
year by year analysis permitted
the observation of trends.
To improve the accuracy of its
classification statistics, DOD
agreed to develop a revised sam-
pling system that would produce
more reliable data. The revised
method varies considerably from
the prior one. First, the sample is
no longer based exclusively on
message traffic data. It includes
all other documentary types, such
as letters, memoranda, and
reports. Second, the totals are
based on data supplied from a
greater number of DOD compo-
nents, including all of the major
activities of the military depart-
ments, the DIA, and NSA. The
revised method requires the DOD
activities to count classification
actions over a one week period.
The numbers obtained are then
multiplied by 52. Appendix A to
last year's ISOO Report to the
President described the revised
sampling system in greater detail.
FY 1985 was the year of transi-
tion for DOD. It utilized both sam-
pling systems in reporting its
statistics to ISOO. In the body of
its FY 1985 Report, ISOO utilized
the figures provided by the old
message traffic sampling system.
This was done to allow for more
meaningful comparisons with
DOD's classification data from
prior years. However, at that time,
ISOO indicated its intention to use
the data provided by the revised
system in future reports, because
they provide more accurate num-
 bers. For this reason, the classi-
fication data in this Report include
those provided by DOD's new
sampling method. As a result, the
accompanying exhibits on Gov-
ernment-wide classification activ-
ity contain comparisons for
FY 1985 and FY 1986 only, since
they are the only years for which
comparable data are available.
The remainder of the exhibits,
relating to classification author-
ities, declassification activities,
self-inspections, and infractions,
usually include data from earlier
years as well.
To refine further the reporting
system for FY 1986, ISOO recom-
mended that DOD and CIA con-
duct their sampling of classifi-
cation activity on more than one
occasion during the year. Both
agencies agreed to do so in
FY 1986, and sampled their clas-
sification activity for two separate
weeks at different times of the
year. In the past, ISOO had
expressed concern that relying
exclusively on a one week period
may have resulted in skewed
numbers. For example, during
one year, the week selected might
be unusually slow in terms of the
volume of classified information
generated. Thus, the numbers
reported would be too low. On the
other hand, the week selected
another year might occur during a
crisis, and result in unrealistically
high figures.
 Original Classification
Authorities Decrease
Significantly
(Exhibits 1 and 2)
Original classifiers are those indi-
viduals designated in writing,
either by the President or by
selected agency heads, to clas-
sify information in the first
instance. During FY 1986, partly
in response to ISOO's continuing
entreaties to keep them to a mini-
mum, the agencies decreased the
number of individuals with original
classification authority. Limiting
the number of original classifiers
may have a significant impact on
controlling the volume of overall
classification activity. ISOO will
continue to impress upon agen-
cies the importance of regular
surveys of their original
classifiers.
The number of executive
branch employees authorized to
classify originally has decreased
dramatically since FY 1971, when
the figure was 59,316. In FY 1986,
there were 6,756 individuals with
original classification authority.
This is also considerably less
than the total of 7,014 reported in
FY 1985; in fact, it represents the
FY 1971 - 1986
E.O. 10501 E.O. 11652
21,277
it
13,976
0 7,056
I
fewest number of original classi-
fiers since ISOO began collecting
such data. ISOO commends the
four agencies which primarily
account for the reduction. They
are CIA, DOD, DOE, and NRC.
Their decreases more than offset
increases at several other
agencies.
FY 1986 Ig~1 Igo
"Top Secret" Authorities: I ^
"Secret" Authorities: MM
I I! 1
I~1
"Confidential" Authorities: I
A
ISOO's FY 1985 Report
expressed concern about the
increased number of original clas-
sification authorities over
FY 1984, especially at the Top
Secret level. In FY 1986, improve-
ments were achieved at all levels,
with Top Secret, Secret and
Confidential authorities decreas-
ing by 4%, 3%, and 5%, respec-
tively. Although some of the major
classifying agencies have signifi-
cantly reduced their authorities,
ISOO will continue to encourage
each to make a concerted effort
to reduce further the number of
(1,502) original classifiers. This is espe-
cially true for those agencies
reporting increases in FY 1986. In
(4,147) its program reviews, ISOO will
t = 500 Authorities
ask those agencies to justify
these increases or to reverse
them. ISOO continues to believe
that some designations are based
solely on the purported prestige
of being an original classification
authority. This will always be an
unacceptable basis.
 Original Classification
Decisions Increase Modestly
(Exhibits 3 through 7)
Original classification is an initial
determination by an authorized
classifier that information requires
protection against unauthorized
disclosure in the interest of
national security. This process
includes both the determination
of the need to protect the infor-
mation and the placement of
markings to identify the informa-
tion as classified. The act of clas-
sifying information originally is the
prelude to all other aspects of the
information security system.
Therefore, the number of original
classification decisions is proba-
bly the most important statistic
reported by ISOO. Because ISOO
now includes the data derived
from DOD's revised sampling sys-
tem, it can compare only the clas-
sification data for FY 1985 and
FY1986. Comparable numbers are
not available for prior years.
In FY 1986, the number of orig-
inal classification decisions rose
by 73,757 (+6%), to 1,221,110. A
comparison of classification
activity at each level between
FY 1985 and FY 1986 shows that
the number of Top Secret actions
decreased by 17,118 (-41 %), to
24,207; Secret actions rose by
72,918 (+13%), to 652,290; and
Confidential actions increased by
17,957 (+3%), to 544,613.
Although the agencies demon-
strated a concerted effort to
reduce the number of original Top
Secret actions, the continuing
increase in Secret actions is one
indicator of a trend that has devel-
oped to protect more classified
information at this level. Last year
ISOO stated that it would watch
carefully to ensure that any such
movement was justified. Accord-
ingly, during FY 1986, ISOO initi-
Sec
85
Sec
86
Conf Conf
85 86
ated a special document review
program to examine, among other
areas, how agencies are comply-
ing with proper classification prin-
ciples. Appendix A, p. 28,
contains more detailed informa-
tion on these document reviews.
Although these reviews provide
evidence that the classification
process is generally working well,
they confirm that classifiers are
assigning the Secret level more
frequently each year. In part, this
appears to result from a lack of
confidence by classifiers in the
Confidential level. The Secret
level also appears to offer the
comfortable "middle ground" in
making a classification assign-
ment. ISOO will continue to
monitor this area closely.
Total
85
Total
86
 ORIGINAL CLASSIFICATION LEVEL
ASSIGNMENTS
A comparison of original clas-
sification level assignments
between FY 1985 and FY 1986
shows that the relative percent-
ages remained nearly constant. In
FY 1985, 4% of the agencies'
original classification actions
were Top Secret; 50% were
Secret; and 46% were Con-
fidential. During FY 1986, the
percentage of Top Secret actions
decreased to 2%; Secret
increased to 53%; and Con-
fidential decreased to 45%. The
significant decrease in original
Top Secret decisions is made
more important by the sur-
prisingly high number of Top
Secret derivative actions
reported below. If both original
and derivative Top Secret actions
had increased significantly, it
would strongly indicate a trend in
that direction. However, the
reported decrease in original Top
Secret decisions is one indicator
that suggests that this year's
derivative Top Secret numbers
are in all likelihood a statistical
aberration. This would be con-
sistent with all other evidence
available to ISOO regarding the
distribution of classification level
assignments.
A comparison of the data of
each of the major original classify-
ing agencies indicates that DOD
and State registered increases in
FY 1986, while CIA and Justice
reported decreased activity. DOD
reported 812,058 original deci-
sions (+16%) in FY 1986. State's
actions increased by 14,124
(+8%), to 199,844. The CIA
reported a large decrease in the
number of original classification
decisions during FY 1986, from
181,688, to 135,668 (-25%). Jus-
tice also reported decreased orig-
inal classification activity for
FY 1986. The 65,022 actions are
9% fewer than for FY 1985.
Among the agencies with more
modest levels of classification
activity, ISOO commends the fol-
lowing for achieving substantial
reductions in original classifica-
tion: AID (-67%); NRC (-80%);
and Treasury (-6%).
ORIGINAL CLASSIFICATION ACTIVITY
BY AGENCY
FY 1985 - 1986
In Thousands
702 FY 198FY 1986
10 0 136
loom 65 9
CIA Justice All
Others
CIA Justice All
Others
 Consistent with prior years, the
same four agencies account for
approximately 99% of the original
classification decisions within the
executive branch. In FY 1986,
DOD accounted for 67%; State
16%; CIA 11 %; and Justice 5%.
As part of the original classi-
fication process, the classifier
must determine a time frame for
the protection of the information.
This is commonly referred to as
the "duration" of classification.
E.O. 12356 provides classifiers
with two means of designating
declassification instructions for
national security information.
First, the information may be
marked for declassification upon
a specific date or event. For
example, a classifier may deter-
mine that the information's sen-
sitivity will cease upon the
completion of a particular project.
That event would be noted on the
face of the document. Only if a
specific date or event cannot be
determined at the time of classi-
fication does the classifier mark
the document with the notation
"Originating Agency's Determina-
tion Required" ("OADR").
"OADR" indicates that the infor-
mation must be reviewed by the
originating agency before any
declassification action is taken.
For FY 1986, using the new
sampling method, DOD reported
that 18% of the documents it orig-
inally classified contained a date
or event for declassification,
down from 22% in FY 1985. As a
result, the overall rate within the
executive branch decreased from
15% in FY 1985, to 13% in
FY 1986. Despite the lower figure
reported by DOD, it continues to
lead the major agencies in the
proportion of items assigned a
date or event for declassification.
Agencies improving their rate dur-
ing the year include the CIA (1 % in
FY 1985, to 2% in FY 1986); Jus-
tice (0A% in FY 1985, to 1 % in
ORIGINAL CLASSIFICATION ACTIVITY
MAJOR AGENCIES
FY 1986); and Treasury (9% in
FY 1985, to 15% in FY 1986).
State's rate decreased from 8% in
FY 1985, to 6% in FY 1986.
During the course of their on-
site inspections and the special
document reviews, ISOO analysts
have observed a number of docu-
ments marked "OADR" that could
have specified a date or event for
declassification. For example,
they have seen a number of visit
itineraries, sensitive only for the
duration of the visit, marked
"OADR:" ISOO remains con-
ORIGINAL CLASSIFICATION/
DECLASSIFICATION ASSIGNMENTS
vinced that this is an area in
which improvements are achieva-
ble. It will continue to press agen-
cies to use a date or event
whenever possible, since auto-
matic declassification instructions
can result in substantial savings
to the Government in terms of the
resources needed to maintain
sufficient declassification pro-
grams, and in terms of the costs
associated with the safeguarding
of classified information for
unnecessarily long periods of
time.
FY 1986
Agency
% Assigned Date
or Event for
Declassification
% OADR (Must be
Reviewed Before
Declassification)
0/b
"TS"
0/0
"S"
%
"C
DOD
18
82
1
55
44
CIA
2
98
9
76
15
State
6
94
0
22
78
Justice
1
99
6
84
10
Treasury
15
85
1
6
93
All Others
5
95
2
23
75
 Derivative Classification
Decreases Substantially
(Exhibits 8 through 11)
Derivative classification is the act
of incorporating, paraphrasing,
restating or generating in new
form classified source informa-
tion. Information may be deriva-
tively classified in two ways: (a)
through the use of a source docu-
ment, usually correspondence or
publications generated by an orig-
inal classification authority; or (b)
through the use of a classification
guide. Only executive branch or
Government contractor employ-
ees with the appropriate security
clearance who are required by
their work to restate classified
source information may classify
derivatively.
For FY 1986, executive branch
agencies reported making
9,548,538 derivative classification
decisions. This figure represents
a tremendous decrease of
4,424,407 fewer (-32%) deriva-
tive classification decisions than
in FY 1985. It also represents the
first time since ISOO has been
collecting, analyzing, and report-
ing classification data that the
number of derivative decisions
has decreased from one year to
the next. While ISOO is not sur-
prised that the number of classi-
fication decisions did not increase
from FY 1985 to FY 1986, the
amount of the decrease is very
Sec
85
Sec
86
Conf Conf
85 86
surprising. Equally surprising is
the reported number of Top
Secret derivative decisions. While
reported Secret derivative deci-
sions decreased by 895,444
actions (-13%), and reported
Confidential derivative decisions
decreased by an enormous
4,467,974 actions (-65%), the
number of reported Top Secret
derivative actions rose by 939,011
actions (+259%). As a result, Top
Secret actions comprised an
unprecedented 14% of all deriva-
tive actions reported for FY 1986.
Total
85
Total
86
 ISOO strongly suspects that
both the huge decrease in deriva-
tive actions and the increased
percentage of Top Secret deriva-
tive actions are statistical aberra-
tions. Both are primarily the result
of the large variation in the num-
bers reported by DOD from
FY 1985 to FY 1986. DOD is
unable to cite any specific reason
related to its classified programs
that accounts for these changes.
However, these are the first two
years of the revised DOD sam-
pling system. Given the magni-
tude of DOD's world-wide
operations, the successful imple-
mentation of this system is a
mammoth task. ISOO believes
that variations of this dimension
are very likely to disappear as the
improved sampling system is bet-
ter established over time.
While ISOO cannot recalculate
through alternative means the
amount of classification activity
that takes place in a given year, it
does have alternative means to
estimate the breakdown in classi-
fication activity among classifica-
tion levels. No other indicator in
this or any other year has ever
suggested a percentage of Top
Secret actions that comes even
close to that reported for deriva-
tive actions in FY 1986. For exam-
ple, ISOO's special document
reviews in FY 1986 (see Appendix
A, p.28) revealed a Top Secret
rate of less than 7%, even though
the reviews concentrated on
activities that would be expected
to produce a higher than average
proportion of Top Secret actions.
Therefore, ISOO assumes that
the number of Top Secret deriva-
tive actions reported for FY 1986
largely overstates their actual
number. It also assumes that in
future years this percentage will
return to a figure between 2%
and 5%.
Despite the enormous drop in
the number of derivative
Confidential decisions, which
also is a result of the figure
reported by DOD, ISOO does not
assume that this is as obvious a
statistical aberration. All other
indicators now available to ISOO
suggest that Confidential
decisions account for between
30% and 40% of the total. While
the reported figure of 25% Confi-
dential derivative actions seems
quite low, it is not as abnormal as
the reported number of Top
Secret derivative actions.
Because ISOO has no means
available to recalculate the
number of classification actions,
it cannot state with certainty the
comparative accuracy of the
disparate figures in derivative
classification reported by DOD in
FY 1985 and FY 1986, respec-
tively. It suspects, however, that
the actual number may be some-
where in between the two. Alter-
natively, the figure reported for
FY 1985 was skewed by an
unusually high level of derivative
classification activity during the
week sampled, or the figure
reported for FY 1986 was skewed
by an usually low level of deriva-
tive classification activity during
the two weeks sampled.
DERIVATIVE CLASSIFICATION
LEVEL ASSIGNMENTS
 For FY 1986, DOD reported
6,516,146 derivative decisions,
over 4 million fewer (-38%), than
the number reported in FY 1985.
CIA also reported a significant
drop in derivative classification
activity, from 3.25 million to 2.5
million (-23%). While applauding
both agencies for their sincere
efforts to control classification
activity, ISOO also believes that
these decreases resulted in part
from improved sampling systems.
DOD's situation is discussed
above. In CIA's case, the agency
conducted its sampling over two
separate time periods rather than
one, as it had done in the past.
This lessened the possibility that
the numbers would be skewed by
unusual classification activity in
one sampling period.
DOD's and CIA's reported
decreases more than offset a
large increase in derivative classi-
fication activity reported by
Justice, from 108,930 actions in
FY 1985, to 487,879 actions in
FY 1986 (+348%). Again, how-
ever, this large variation appears
to result from a change in the
sampling system used to collect
DERIVATIVE CLASSIFICATION
ACTIVITY BY AGENCY
FY 1985 - 1986
In Millions
1.1.5
the data to be reported to ISOO.
Here the change took place at the
FBI, which accounts for most of
the classification activity within
Justice. In prior years, the FBI
took an actual count of its Head-
quarters classification activity
over a one week period and
multiplied that figure by 52. Head-
quarters then estimated the
DERIVATIVE CLASSIFICATION ACTIVITY
MAJOR AGENCIES
amount of additional classification
that took place in FBI Field
Offices, and added that to the
Headquarters total. In FY 1986,
both Headquarters and the Field
Offices conducted an actual count
over a one week period and
multiplied this amount by 52. It
turned out that the FBI Field
Offices account for a far greater
number of derivative actions than
was previously estimated.
Among the agencies with more
modest levels of classification
activity, ISOO commends the fol-
lowing for achieving significant
reductions in derivative classifica-
tion: ACDA (-70%); NSC (-12%);
Treasury (-35%); USDA (-26%);
and USPS (-44%).
Once again, DOD and CIA
account for the overwhelming
number of derivative actions
within the executive branch, 68%
and 26%, respectively. With its
reported increase in derivative
activity, Justice now accounts for
5% of the total. All other agencies
reported 47,057 derivative
actions, less than 1 % of the total.
 Combined Classification
Activity Decreases Significantly
(Exhibits 12 through 15)
For FY 1986, the number of
original and derivative classifi-
cation decisions combined was
10,769,648. This represents a
decrease of 4,350,650 actions
(-29%) from FY 1985. The
decrease is based, of course, on
the tremendous decrease in
derivative actions discussed
above. By classification level, the
number of combined Top Secret
actions increased from 404,330,
to 1,326,223 (+228%); the num-
ber of Secret actions decreased
from 7,289,093, to 6466,567
(-13%); and the number of
Confidential actions decreased
from 7,426,875, to 2,976,858
(-60%). Again, the significant
changes in the Top Secret and
Confidential levels are the result
of those for derivative classifica-
tion, also discussed above.
7,289,093 6,466,567
Sec
85
Sec
86
Conf
85
Conf
86
Total
85
Total
86
 For FY 1986, the breakdown of
combined classification by classi-
fication level is Top Secret, 12%;
Secret, 60%; and Confidential,
28%. As discussed in the section
on derivative classification,
above, ISOO believes that these
proportions probably differ from
what would be revealed if an item
by item breakdown were possible.
All other indicators suggest that
the actual percentage of Top
Secret actions is between 2%
and 5%, and the number of Con-
fidential actions is between 30%
and 40%.
COMBINED CLASSIFICATION
LEVEL ASSIGNMENTS
COMBINED CLASSIFICATION ACTIVITY
BY AGENCY
FY 1985 - 1986
In Millions
The overall decrease in
combined classification activity
reported for FY 1986 reflects the
significant decreases reported by
the classification arena's two
largest players, DOD and CIA.
The CIA reported decreases in
both original and derivative
actions. DOD's large decrease in
derivative actions dwarfed its
reported increase in original
actions. The significant
decreases in classification activity
at DOD and CIA more than offset
the large increase in reported
actions by Justice, and the
modest increase reported by
State.
 As has been true in all other
ISOO reporting periods, four
agencies account for over 99% of COMBINED CLASSIFICATION ACTIVITY
the combined classification MAJOR AGENCIES
activity in FY 1986: DOD, 68%;
CIA, 24%; Justice, 5%; and State, FY 1986
2%. Of the 10,769,648 reported
all other
classification decisions
,
agencies reported that they made DOD 68% All Others 1 %
S 1 i
I MADE A MISTAKE cLA5;IFy = IT WAS A
ON 1HIS OEPbRTit , FpETTy go
SIR , 'SEzEf CZt 2F
THEN
CLASSIFY
IT 1ToP
SFC125T
CROCK by Bill Rechin and Don Wilder
Copyright by and permission of North America Syndicate, 1987
 Mandatory Review
Remains Strong
(Exhibits 16 through 21)
Under E.O. 12356, the mandatory
review process allows agencies or
citizens to require an agency to
review specified national security
information for purposes of
seeking its declassification.
These requests must be in writing
and must describe the informa-
tion with sufficient detail to permit
the agency to retrieve it with a
reasonable amount of effort.
Mandatory review remains
popular with some researchers as
a less contentious alternative to
Freedom of Information Act
requests. It is also used to seek
the declassification of presidential
papers or records, which are not
subject to the Freedom of Infor-
mation Act.
The number of mandatory
review requests received in
FY 1986 increased by 44, to 4,081.
Agencies experiencing large
increases include ACDA, CIA, and
NSC. When the 1,946 cases
MANDATORY REVIEW
REQUESTS RECEIVED
4,192 3,945 4,037 4,081
2,246
carried forward from the prior of 5,560, during FY 1985. In
year are added to the new cases FY 1986, agencies acted on 3,991
received, agencies had a total cases, 10% more than in FY 1985.
caseload of 6,027, during FY 1986.
This compares to a caseload
Since FY 1983, ISOO has
collected data on agency actions
in response to mandatory review
Exhibit 17 requests in terms of cases,
documents, and pages. The
FY 1983 - 1986
In Thousands
Granted
in Full
PAGES PROCESSED
Granted
in Part
Denied
in Full
3,991 cases processed durng
FY 1986 comprised 57,674
documents totaling 176,563
pages. The number of pages
processed under mandatory
review was 153,382 fewer than in
FY 1985.
 Of the 3,991 cases completed
in FY 1986, 1,952 were granted in
full, 1,582 were granted in part,
and 457 were denied in full.
These totals are consistent with
the favorable results in prior MANDATORY REVIEW ACTION TAKEN Exhibit 18
years.
Of the 57,674 documents acted FY 1986
on in FY 1986, 49,056 were
granted in full, 6,188 were granted
in part, and 2,430 were denied in
full. As a percentage of the total,
agencies granted in full 85% of 40%
the documents, granted in part 11% --( 4%
11%, and denied in full 4%.
Although the percentage of docu-
ments declassified in whole or in
part (96%) is down slightly from
FY 1985, its magnitude empha-
sizes the continued vitality of the
Of the 176,563 pages
processed, 119,504 (67%) were
granted in full, 39,911 (23%) were
granted in part, and 17,148 (10%)
were denied in full. Again, FY 1985, mandatory review
although the percentage of pages remains a highly successful
declassified in whole or in part mechanism for the declassifica-
(90%) is down slightly from tion of information.
MANDATORY REVIEW
ACTIONS BY AGENCY
Total Cases We Granted % Granted
Agency Acted On in Full in Part
The sharp decline in the
number of pages reviewed in
FY 1986 is attributable to a reduced
Exhibit 19 workload at DOD. During the year,
it received 114 fewer cases
(-15%), and acted on 214,645
fewer pages than in FY 1985.
We Denied Additionally, several requests in
in Full FY 1985 involved a large quantity
State
1152
45
45
10
NSC
747
37
58
5
DOD
628
59
22
18
NARA
527
50
35
15
Justice
409
90
7
3
CIA
333
17
58
25
All Others
195
52
43
5
of cables, which are generally
easier to review and declassify
than the more substantive classi-
fied reports and memoranda.
ISOO commends the several
agencies that processed consid-
erably more pages in FY 1986,
than in FY 1985. These include
NARA (+43,406), NSC (+5,966),
and State (+9,173).
 Requesters may also appeal
mandatory review denials to
officials of the denying agencies,
or, with respect to classified
presidential materials, to the
ISOO Director. During FY 1986,
agencies received 497 new
appeals, 215 (+76%) more than in
FY 1985. When these are added
to the 539 carried over from the
previous year, agencies had an
appeals' caseload of 1,036. Of
these, agencies completed 575 in
FY 1986. This represents a 10%
improvement over FY 1985, and a
30% improvement over FY 1984.
As in FY 1985, Justice was the
agency primarily responsible for
the improved figure in FY 1986.
MANDATORY REVIEW APPEALS
PAGES PROCESSED
FY 1983 - 1986
In Thousands
MANDATORY REVIEW APPEALS
WORKLOAD IN PAGES
FY 1986
? Granted in Full
3%
Of the 575 appeals completed,
177 (31 %) were granted in full,
334 (58%) were granted in part,
Exhibit 21 and 64 (11 %) were denied in full.
These appeals totaled 11,166
documents and 47,995 pages. Of
the documents reviewed on
appeal, 3,294 (29%) were
released in full, 7,663 (69%) were
released in part, and only 209
(2%) were denied in full. Of the
47,995 pages reviewed, 11,429
(24%) were declassified in full,
35,100 (73%) were declassified in
part, and only 1,466 (3%)
remained fully classified. These
impressive numbers suggest that
researchers can anticipate even
greater returns in declassified
information if they pursue the
mandatory review appeal process.
 Systematic Review Results
Encouraging
(Exhibits 22 through 24)
"Systematic review for
declassification" is the program,
first introduced in 1972, in which
classified, permanently valuable
(archival) records are reviewed for
purposes of declassification after
the records reach a specific age.
Under E.O. 12356, NARA is
required to conduct a systematic
review of its classified holdings as
they become 30 years old, except
for certain intelligence or
cryptologic file series, which are
to be reviewed as they become 50
years old. While other agencies
are not required to conduct a
systematic review program, ISOO
encourages them to do so if
resources are available.
SYSTEMATIC REVIEW FOR
DECLASSIFICATION
PAGES REVIEWED
FY 1973 - 1986
In Millions
SYSTEMATIC REVIEW FOR
DECLASSIFICATION
PERCENTAGE OF PAGES DECLASSIFIED
ISOO is pleased to report that
during FY 1986, the product of
the systematic review program
showed its first significant
Exhibit 23 increase in recent years. During
FY 1986, agencies reviewed 164
million pages, up almost 6 million
pages (+57%), from FY 1985. Of
the pages reviewed, 87% were
declassified, a significant increase
from the 78% rate reported for
FY 1985. As a result of the
greater number of pages
reviewed and the improved
declassification percentage rate,
14.3 million pages were declassi-
fied under the systematic review
program in FY 1986, over 6 million
more than in FY 1985, a 76%
increase.
 SYSTEMATIC REVIEW FOR
DECLASSIFICATION
ACTIONS BY AGENCY
Agency
Pages
Reviewed
NARA
7,461,550
7,123,364
95
DOD
6,875,907
5,708,603
83
AID
1,859,167
1,347,327
72
State
39,955
34,586
87
DOT
36,100
6,130
17
Justice
6,576
1,266
19
All Others
93,780
50,992
54
Almost all of the increase in
systematic review is due to the
figures reported by NARA. In
FY 1986, it reviewed 7.5 million
pages, an increase of 4.3 million
(+137%), from FY 1985. The
dramatic change is the result of
several projects undertaken by
NARA during FY 1986. These
included efforts at the regional
archives in Los Angeles, San
Francisco, Fort Worth, and
Atlanta; coordination of the
review of the Office of Strategic
Services (OSS) operational files;
and continuation of the review by
NARA, on an interagency agree-
ment basis, of particular State
and AID records. Both the
regional records and the OSS
operational files are of World
War II vintage and generally are
much easier to review than post-
war records. This is due primarily
to the age of the documents, and
the availability of specific declas-
sification guidelines provided by
the major agencies. In addition, in
the case of the OSS operational
files, the records were first
reviewed by the CIA over the
course of the preceding years,
with NARA's responsibilities
limited to examining the docu-
ments for the interests of foreign
governments and other agencies.
While ISOO is encouraged by
the figures reported by NARA for
FY 1986, they are, unfortunately,
unlikely to signal a trend. Over
2 million pages of the records
reviewed in FY 1986 were suscep-
tible to bulk declassification
methods that are not viable for
most postwar records in NARA's
custody. In order to maintain the
increases achieved during
FY 1986, NARA will have to
commit additional resources to its
systematic review program.
Other agencies showing
increased systematic review
activity during FY 1986 include
DOD (+1 %), AID (+427%), NASA
(+294%), NSC (+100%), and
DOT (+81 %). ISOO applauds the
efforts of these agencies since
they are not required to conduct
systematic review programs. It
urges their continuation.
 Agency Self-Inspections
Decline Marginally
(Exhibits 25 and 26)
Executive Order 12356 mandates
that agency heads establish and
maintain "an active oversight and
security education program." In
this regard, the agencies report
to ISOO the number of self-
inspections and the number and
type of infractions found during
the year. Infractions are minor
violations of the order, the imple-
menting ISOO Directive, or
agency regulations. These statis-
tics do not include the more
serious security violations that
agencies must report to ISOO as
they occur.
For FY 1986, agencies reported
that they had conducted 27,361
self-inspections. This is a 3%
decrease from FY 1985, and
approximates the total registered
for FY 1984. Despite the overall
decrease, some agencies
reported significantly higher
numbers of self-inspections and
merit special commendation.
These include CIA (+62%),
Infraction
Total
FY 83
Total
FY 85
Total
FY 86
Unauthorized Access
620
440
560
Mismarking
10,849
6,642
5,177
Unauthorized
Transmission
1,294
1,688
1,969
Improper Storage
3,844
5,089
4,850
Unauthorized
Reproduction
249
143
107
Overclassification
220
164
157
Underclassification
317
265
250
Classification w/o
Authority
238
109
185
Improper Destruction
581
322
346
Other
132
292
211
DOE (+32%), and DOT (+86%).
ISOO continues to be
concerned not only with the
quantity of self-inspections that
the agencies undertake, but also
with their quality. This concern
results from the fact that during
the self-inspections conducted in
FY 1986, agencies found 1,342
fewer infractions than in FY 1985.
The total of 13,812 infractions is
FY 1985 FY 1986
9% less than the figure reported
for the previous year. The average
number of infractions discovered
per inspection also fell, from .54
in FY 1985, to.50 in FY 1986.
The figures reported for
FY 1986 again call into question
the thoroughness of agency self-
inspections. From ISOO's experi-
ence, thorough inspections, even
in organizations with outstanding
information security programs,
disclose a far greater number of
infractions than those routinely
reported by the agencies. See, for
example, the results of ISOO's
special document reviews con-
ducted during FY 1986, Appendix
A, p. 28. ISOO strongly suspects
that a major reason for the
absence of reported infractions is
the failure by most agencies to
examine periodically a sampling
of its classified product. ISOO
urges a heightened effort by the
agencies to increase both the
number and quality of the self-
inspections they are conducting.
Only in this manner will the
agencies themselves be able to
evaluate employee compliance
with the Order's provisions, and
provide the basis for revising their
security training programs.
 Appendix A
Findings of Special ISOO
Reviews of Classified
Documents
Background
One of ISOO's major functions is
to consider and respond to
complaints and suggestions from
within and outside the executive
branch concerning the adminis-
tration of the information security
system. Understandably, many
complaints, whether directed to
ISOO or not, concern the quality
of the classified product. Not
infrequently, executive branch
officials, members of Congress,
journalists, researchers or others
express their dissent concerning
the classification or marking of
national security information.
Most often, this dissent involves
the allegation that information
that is classified should not be
classified, the circumstance that
is popularly referred to as "over-
classification."
Examining the classified
product, therefore, is an essential
element of an effective oversight
program. In addition to its ad hoc
consideration of specific com-
plaints, ISOO routinely examines
a small sample of an agency's
classified product during its
program reviews or inspections.
Further, the agencies themselves,
which are responsible for the
predominant oversight of their
information security programs,
should include reviews of their
classified product within their
internal monitorship. However,
because ISOO's small size limits
the scope of its document
reviews during its routine
inspections, and because ISOO
has found that agency self-
inspections too often do not
include an examination of the
classified product, ISOO has
sought alternative means to
answer the question, "How good
or how bad is the classified
product that is being created?"
During FY 1986, ISOO
conducted the first group of what
it hopes will be a series of special
document reviews. The underly-
ing purpose of these reviews is to
provide an additional means to
determine the extent of compli-
ance with the classification princi-
ples and marking procedures out-
lined in Executive Order 12356,
implementing ISOO directives and
agency regulations.
In this first group of special
document reviews, IS00 analysts
examined a sample of 3,025
documents generated by selected
units of twelve of the most signifi-
cant players in the classification
arena: Air Force, Army, CIA, DIA,
DOE, OJCS, Justice, NSA, Navy,
OSD, State and Treasury. ISOO
does not purport that these
documents comprise a fully
representative sample of the
universal classified product;
indeed, ISOO is uncertain
whether it is even possible to
construct a sample that is fully
representative. To be sure, ISOO
is aware of several features of its
sample that may skew its findings
somewhat:
(a) most of the documents were
created within headquarters
units, rather than within field
units;
(b) intelligence and policy-
making units were more than
proportionally represented, while
operational units were less than
proportionally represented; and
ISOO SPECIAL DOCUMENT REVIEWS
THE SAMPLE: BY DOCUMENT TYPE/
CLASSIFICATION LEVEL
 (c) almost all of the documents
examined were created recently,
and were maintained in active
files.
Despite or even because of
these features, ISOO is convinced
that the results provide valid
indicators of the state of the
classified product that has been
created in recent years and will be
created in the next few years.
Because the sample contained
large numbers of recent, relatively
high level headquarters docu-
ments, it can be reasonably
assumed that they and docu-
ments like them will influence
significantly the contemporary
classified product, especially the
derivative product. Also, the
results, when broken down by
agency, parallel the findings of
ISOO's program reviews.
Two other factors merit consid-
eration. First, the ISOO reviewers,
who have all served as ISOO
analysts for at least three years,
were placed in three teams. Each
team reviewed the documents of
four of the twelve agencies. All
the reviewers met before, during,
and after the reviews to seek as
much consistency as possible in
their methodology and analysis.
Nevertheless, as would be true in
any situation that involves individ-
ual judgment, it was clear that
each team took a slightly different
approach toward its reviews and
analyses.
Second, ISOO has cumulated
its findings for this report in order
to present a clearer picture of the
classified product as a whole.
When broken down by agency,
however, the variation in results
was often quite extensive. For
example, a few agencies
accounted for a significantly dis-
proportionate share of the dis-
crepancies described below,
while the classified product of a
few others revealed remarkably
few discrepancies. Again, these
findings very much parallel those
of ISOO's program reviews.
The data collected from the
sample fall into two broad cate-
gories: (1) information about the
classification of each document,
and (2) discrepancies in classifi-
cation or marking noted by the
analysts. The first category
included information about the
generic type of document; the
classification level; the basis for
or source of the classification;
and the assigned duration of clas-
sification. In identifying the
discrepancies, the analysts
worked with the following
definitions:
"Overclassification" -(a) Clear-
cut: The information in the docu-
ment does not meet the
standards necessary for classifi-
cation; (b) Questionable: While
the question of meeting classifi-
cation standards is arguable,
classification does not appear to
be necessary to protect our
national security; (c) Partial: A
portion(s) of the document
appears to be unnecessarily clas-
sified, although the overall clas-
sification of the document is
correct.
"Overgraded" - All or some of
the information in the document
appears to be classified at a
higher level than justified.
"Undergraded" - All or some of
the information in the document
appears to be classified at a lower
level than necessary.
"Declassification" -The docu-
ment appears to have improper
declassification instructions or no
declassification instructions at all.
"Duration" - The duration of
classification is marked "OADR;'
when a specific date or event for
declassification appears feasible.
"Original/Derivative" -The
document is marked and treated
as an original classification action
although the classified informa-
tion appears to be derived from a
guide or other source(s).
"Marking" -The document
appears to have improper classifi-
cation markings or lacks required
markings, including instances in
which the document fails to cite
or cites improperly the classifica-
tion source. This does not include
marking errors more specifically
included under another discrep-
ancy category.
"Portion Marking" -The docu-
ment appears to lack required
portion markings.
"Multiple Sources" -The official
file copy of the document cites
"multiple sources" as the basis
for classification, but does not list
these sources.
Findings
Types of Documents Reviewed:
Of the 3,025 documents reviewed
by ISOO, 1,328 (43.9%) were
memoranda or letters; 748
(24.7%) were cables or messages;
and 949 (31.4%) were some other
type of document, predominantly
reports. There is no statistical
significance to the breakdown
other than ISOO's efforts to
examine sufficient numbers of
each type of document that might
be classified. In terms of the
levels of classification within each
type, however, there were several
significant variations. A document
in the "other" category, i.e., a
report, was 1.6 times as likely to
be classified Top Secret than a
document generally, while a cable
or message was 1.6 times as
likely to be classified Confi-
dential. This disproportionate
representation of cables or mes-
 sages is quite significant to
ISOO's reports in prior years con-
cerning the proportions of Top
Secret, Secret and Confidential
actions. In those years the DOD's
sampling system was based on
message or cable traffic exclu-
sively, the totals for which were
then extrapolated to include other
types of documents. Because
cable or message traffic appears
to be classified at the Confi-
dential level disproportionately to
the classification level of other
types of documents, and because
DOD accounts for such a signifi-
cant portion of classification
actions, it now appears that
ISOO, based on the data avail-
able, reported overstated per-
centages of Confidential actions,
understated percentages of
Secret actions, and slightly
understated percentages of Top
Secret actions for the years
1979-1984. Only when DOD
altered its sampling system in
FY 1985, did ISOO receive data
that are not skewed by the dispro-
portion in classification levels
represented by cable or message
traffic.
Classification Levels: Of the
3,025 documents reviewed, 208
(6.9%) were classified Top
Secret; 1,912 (63.2%) were
classified Secret; and 905
(29.9%) were classified Confi-
dential. These proportions
resemble the distribution (TS:
12.3%; S: 60.0%; C: 27.6%) of all
the classification actions reported
to ISOO by the agencies for
FY 1986 (see p. 20, supra). Based
on all the data available to ISOO, it
would appear that three out of
five contemporary classification
actions are at the Secret level,
and that this ratio is gradually
increasing. The increasing rate of
Secret classification assignments
suggests that some classifiers
feel more secure in choosing the
"middle ground" for assigning a
classification level, because they
lack confidence in the Confi-
dential level.
ISOO SPECIAL DOCUMENT REVIEWS
CLASSIFICATION LEVELS
 Classification Basis or Source: Of
the 3,025 documents, the review-
ers determined that 575 (19%)
were original classification
actions; 996 (32.9%) were deriva-
tively classified based on a classi-
fication guide; 673 (22.3%) were
derivatively classified based on
multiple sources; and 133 (4.4%)
were derivatively classified based
on a single source. The reviewers
could not determine the basis for
or source of the classification for
648 (214%) of the documents.
Most of these were cables or
messages, and electronically
transmitted information is not
required to include the classifi-
cation source. Therefore, ISOO
cites the absence of a classifica-
tion source as a marking discrep-
ancy, below, only when it pertains
to a document other than an elec-
tronically transmitted cable or
message.
ISOO SPECIAL DOCUMENT REVIEWS
BASIS FOR CLASSIFICATION DECISIONS Exhibit 29
Of the 2,377 documents for
which the reviewers could deter-
/SOO SPECIAL DOCUMENT REVIEWS mine the classification basis or
ORIGINAL/DERIVATIVE CLASSIFICATION Exhibit 30 source, ce, 575 (24.2%) were 75.8%i-
were classified derivatively. There
is a significantly higher percent-
Original Derivative age of original decisions in this
sample compared to the distribu-
tion (original: 11.3%; derivative:
88.7%) reported to ISOO by the
agencies for all classification
actions in FY 1986. ISOO attributes
this difference to the fact that its
reviewers concentrated on head-
quarters, policy-making units,
which are more likely to originate
classified information than field or
Sanitized Copy Approved for Release 2011/10/14 : CIA-RDP89GO0643RO01100020015-7
 ISOO SPECIAL DOCUMENT REVIEWS
DURATION OF CLASSIFICATION Exhibit 31
Duration of Classification: Of the
3,025 documents, 2,769 (91.5%)
were marked with the indefinite
instruction for the duration of
classification, "Originating
Agency's Determination
Required" ("OADR"). Only 82
(2.7%) were marked with an
assigned date or event for
declassification. Additionally, 174
(5.8%) contained no declassifica-
tion instruction at all, and are
included in the "declassification"
discrepancy, below. As is the case
with documents marked "OADR:'
when no declassification instruc-
tion appears on a document, the
originating agency must review it
before it may be declassified.
Of the 2,769 documents
marked "OADR:' the ISOO
reviewers identified 54 (1.8%) that
they believed should have been
marked for declassification at a
specific date or event, and these
are shown in the "duration" dis-
crepancy below. The percentage
of documents in ISOO's sample
with an indefinite duration of clas-
sification exceeds the percentage
reported to ISOO by the agencies
in FY 1986 (see p. 15, supra).
ISOO accounts for this difference
by the overrepresentation of intel-
ligence documents in its sample.
Classified information that per-
tains to intelligence activities,
sources or methods is seldom
marked for declassification on a
specific date or event.
 Discrepancies: In reviewing the
3,025 documents, ISOO's ana-
lysts noted a total of 1,090 discre-
pancies. As described below,
most of these discrepancies were
relatively minor technical deficien-
cies. A number of documents had
more than one discrepancy, some
as many as three or four, so that
the actual number of documents
that contained no discrepancies
whatsoever was 2,145 (71 %).
Given the detail of ISOO's scru-
tiny, the number of documents
with no discrepancies is note-
worthy, especially for the several
agencies in which the vast major-
ity of documents achieved this
status.
Of the 3,025 documents,
ISOO's analysts identified 52
(1.7%) that they believed clearly
should not have been classified;
75 (2.5%) that they believed were
unnecessarily classified although
there was an arguable basis for
classification; and 15 (.5%) that
contained classified portions that
the analysts believed should not
have been classified, although the
overall classification of the docu- improper declassification instruc-
ments was correct. The total of tions or no declassification
only 142 documents (4.7%) that instructions at all ("declassifica-
ISOO's analysts identified as over- tion"); 54 documents (1.8%) that
classified is quite commendable, were marked "OADR;' but which
and reflects well on the validity of should have expressed a specific
the classification system gener- date or event for declassification
ally. No doubt the fact that most ("duration"); 158 documents
of the documents were of recent (5.2%) that indicated that they
vintage impacted favorably on were original classification
these results. The national secu- actions although the classified
rity sensitivity of information ordi- information had clearly been
narily decreases over time. From derived from another source
ISOO's experience, most ("original/derivative"); 195 docu-
instances of overclassification ments (6.5%) that contained
occur in that large body of classi- some type of marking error not
fied information that is maintained included in one of the other dis-
over the years without being sub- crepancy categories, including
jected to prior public access the failure to cite a required
demands or some other means of source of classification ("mark-
initiating a declassification review. ing"); 175 documents (5.8%) that
With respect to the other dis- lacked required portion markings
crepancies, ISOO's analysts iden- ("portion marking"); and 182 doc-
tified eight documents (.3%) that uments (6%) that were deriva-
they believed were classified at tively classified on the basis of
too high a level ("overgraded"); multiple sources, but which failed
two documents (.1 %) that they to list these sources with the file
believed were classified at too low or record copy of the document
a level(" undergraded"); 174 docu- ("multiple sources").
ments (5.8%) that contained
ISOO SPECIAL DOCUMENT REVIEWS
DISCREPANCIES
Clear-cut Questionable Partial Over-
Overclas Clas Overclas graded
 nA
Sanitized Copy Approved for Release 2011/10/14 : CIA-RDP89G00643R001100020015-7
(a) The results of this exercise
suggest that the overclassifica-
tion of information is not as
serious a problem as some popu-
lar media report. Nevertheless,
because each publicized instance
of overclassification damages the
credibility of the information
security system far beyond its
quantitative significance, over-
classification will always be a
problem that merits concern and
diligent oversight.
(b) The variation in perform-
ance among the agencies very
closely parallels ISOO's experi-
ence in its ongoing monitorship
program. Further, it confirms one
of ISOO's constant themes: An
agency's commitment to a strong
information security program, pri-
marily in terms of its security edu-
cation and internal monitorship
programs, results in a classified
product that clearly reflects this
commitment. The absence of a
strong commitment results in a
considerably poorer classified
product.
(c) Too many classified docu-
ments contain relatively minor dis-
crepancies in marking procedures
or classification principles. With a
modest investment in improved
security education and internal
monitorship, agencies could elimi-
nate almost all these technical
deficiencies.
(d) Agencies would clearly ben-
efit from an internal monitorship
program that includes regular
examination by appropriate offi-
cials of a sample of the classified
product.
(e) ISOO's first group of special
document reviews provides evi-
dence of a classification system
that is generally performing well.
It answers the question, "How
good or how bad is the classified
product that is being created?",
with a "Pretty good, but with a lit-
tle bit of extra effort, it could be
very good!"
 Appendix B -
Classified Information
Nondisclosure Agreement
FY 1986
Full
Implementation
Full In Some
Agency Implementation Components
Paragraph 1(a) of National Security Decision Directive 84, "Safeguard-
ing National Security Information," of March 11, 1983, directed ISOO to
issue a standardized nondisclosure agreement to be executed as a
condition of access to classified information. In September 1983, ISOO
issued the Standard Form 189, "Classified Information Nondisclosure
Agreement," and directed agencies to work toward complete implemen-
tation as quickly as possible. The chart below provides an agency by
agency breakdown of progress to date.
implementation
for New
Agency-wide Employees anchor No
Implementation Reinvestigations Planning Apparent
In Progress Only implementation implementation
VA
Received waiver from NSC to use a substitute form that fully complies with NSDD 84
Sanitized Copy Approved for Release 2011/10/14 : CIA-RDP89G00643R001100020015-7
 Appendix C
ISOO Standard Forms
Program
Executive Order 12356 broad-
ened ISOO's authority to include
the issuance of standardized
forms relating to the protection of
national security information.
ISOO has since issued fourteen
standard forms for use by execu-
tive branch agencies that gener-
ate or handle national security
information. Each has been
designed, in conjunction with the
affected agencies, to enhance
safeguarding procedures while
seeking to reduce the costs asso-
ciated with competitive or dupli-
cative forms.
Use of these standard forms is
mandatory, unless an agency has
received a waiver to use an alter-
native form based upon special
security or cost requirements.
Supplies of these standard forms
are available through regular pro-
curement channels. Below are a
facsimile and brief description of
each of the forms ISOO has
issued to date.
SECURITY CONTAINER CHECK SHEET
ER CHECK SHEET
TO FIU ?n
FRON -- na .....o
.,row..,,
r---
-1-
7
-~ j
SF 702, Security Container
Check Sheet
The SF 702 records openings,
closings, and end-of-the-day
checks of containers that store
national security information.
SF 701, Activity Security
Check List
The SF 701 is a check sheet used
in the conduct of end-of-the-day
security inspections of work
areas in which classified informa-
tion is handled or stored.
 SF 189, Classified Information
Nondisclosure Agreement
SF 189-A, Classified Information
Nondisclosure Agreement,
(Industrial/Commercial /Non-
Government)
The SF 189 and SF 189-A are
nondisclosure agreements
between the United States and an
individual. An individual must ex-
ecute either the SF 189 or the SF
189-A, as appropriate, before the
Government may authorize that
individual access to classified
information.
SF 700, Security Container
Information
The SF 700 contains (1) informa-
tion on agency employees who
are to be contacted if the security
container to which the form per-
tains is found open and unat-
tended, and (2) a current record
of the security container's com-
bination, classified at the level of
the most sensitive information
stored in the container.
CLPSIFIEO INFORMATION NONDISCLOSURE AGREEMENT
sTRIRLICOMMERCIPLMOXdOVFRNMEHTI
m1 m;ro11-tt aptae~q?~ iw,ror, small o, I.~e~~. m.'f,-G-
moa
W ITNESS AND ACCEPTANCE
 SF 703, Top Secret Cover Sheet
SF 704, Secret Cover Sheet
SF 705, Confidential Cover Sheet
These cover sheets serve as
shields to protect Top Secret,
Secret, or Confidential informa-
tion, respectively, from inadver-
tent disclosure and to alert
observers that such information is
attached.
SF 711, Data Descriptor Label
This label is used to record addi-
tional safeguarding controls that
pertain to classified information
that is stored on ADP storage
media or other non-paper media.
SF 706, Top Secret Label
SF 707, Secret Label
SF 708, Confidential Label
These labels serve the same
purposes as the cover sheets,
except that they are placed on
automatic data processing media,
other non-paper media, and
equipment for which cover sheets
are inappropriate.
SF 709, Classified Label
This label is used to identify and
protect ADP storage media and
other media that contain classi-
fied information pending a deter-
mination by the original classifier
of the specific classification level
of the information.
QASSIRED
SF 710, Unclassified Label
In a mixed environment in which
both classified and unclassified
information are being processed
or stored, this label is used to
identify ADP storage media and
other media that contain un-
classified information.
 Sanitized Copy Approved for Release 2011/10/14: CIA-RDP89GO0643RO01100020015-7
Appendix D
Information Security Briefing
In its continuing efforts to assist
agencies in the presentation of
effective security training, ISOO
has just completed production of
an updated and improved version AWd
of its very successful information 4
Security Briefing series on Execu-
tive Order 12356.
This colorful and informative
audiovisual briefing is divided into
three modules: (1) An overview of
the information security system
established under the Order; (2) a
detailed treatment of proper
marking practices and pro-
cedures; and (3) highlights of the
basic safeguarding requirements.
The presentation is designed so
that each module can be shown
separately or as one unit. The
briefing is unclassified, and is
available at very reasonable cost
in both slide/tape and video cas-
sette formats. The briefing is non-
agency specific, and has been
designed for effective use by both
military and civilian agencies and
their contractors. For further
information, contact ISOO at
[phone].
Sanitized Copy Approved for Release 2011/10/14: CIA-RDP89GO0643RO01100020015-7
 Sanitized Copy Approved for Release 2011/10/14: CIA-RDP89GO0643RO01100020015-7
Information Security
Oversight Office
Washington, D.C. 20405
Sanitized Copy Approved for Release 2011/10/14: CIA-RDP89GO0643RO01100020015-7